key: cord-0026817-8ep0o4ec
authors: Blayney, Douglas W.; Mohanlal, Ramon; Adamchuk, Hryhoriy; Kirtbaya, Dmitry Valikovich; Chen, Michael; Du, Lihua; Ogenstad, Stephan; Ginn, Greg; Huang, Lan; Zhang, Qingyuan
title: Efficacy of Plinabulin vs Pegfilgrastim for Prevention of Docetaxel-Induced Neutropenia in Patients With Solid Tumors: A Randomized Clinical Trial
date: 2022-01-27
journal: JAMA Netw Open
DOI: 10.1001/jamanetworkopen.2021.45446
sha: 0940fd0e29b12e895f64fd47ed4046e398f4ee91
doc_id: 26817
cord_uid: 8ep0o4ec

IMPORTANCE: Prevention of chemotherapy-induced neutropenia (CIN) and its clinical consequences is an unmet need for which plinabulin, a selective immunomodulating microtubule-binding agent, is being tested. OBJECTIVE: To demonstrate noninferiority between plinabulin and pegfilgrastim for days of severe neutropenia in cycle 1 in patients with solid tumors treated with docetaxel. DESIGN, SETTING, AND PARTICIPANTS: The Plinabulin vs Pegfilgrastim for the Prevention of Docetaxel-Induced Neutropenia in Patients With Solid Tumors (PROTECTIVE-1) double-blind phase 3 randomized clinical trial was performed in multiple centers in China, Russia, Ukraine, and the US. Participants included patients with breast, prostate, or non–small cell lung cancer treated with single-agent docetaxel chemotherapy. Data were collected from June 1, 2018, to January 31, 2019. The database was locked on February 18, 2021. Data analysis was based on intention to treat and safety and performed from October 5, 2018, to February 23, 2021. INTERVENTIONS: Plinabulin, 40 mg, plus placebo or pegfilgrastim, 6 mg, plus placebo. MAIN OUTCOMES AND MEASURES: The primary end point was day of severe neutropenia in cycle 1. Additional end points included clinical consequences of CIN (febrile neutropenia, hospitalizations, infections, antibiotic use, and modifications of chemotherapy dose), patient-reported outcomes for bone pain score, markers for immune suppression (neutrophil-to-lymphocyte ratio [NLR] of >5), immature neutrophils (band, promyelocyte, and myelocyte counts >0), and safety. RESULTS: Among the 105 patients included in the analysis (65 [6.19%] women; median age, 59 [range, 31-81] years), the primary end point was met within a noninferiority margin of 0.65 days, with a mean difference of 0.52 days (98.52% CI, 0.40-0.65 days). Grade 4 neutropenia frequency in cycle 1 was not significantly different. Plinabulin had earlier onset of action with less grade 4 neutropenia in week 1 of cycle 1. Plinabulin had fewer adverse clinical consequences with rates of febrile neutropenia (0 of 52 vs 1 of 53 [1.9%]), infections (4 of 52 [7.7%] vs 8 of 53 [15.1%]), chemotherapy dose delay of more than 7 days (2 of 52 [3.8%] vs 3 of 53 [5.7%]), and permanent chemotherapy discontinuation (7 of 52 [13.5%] vs 14 of 53 [26.4%]). Patients receiving plinabulin had significantly less bone pain (difference, −0.67 [95% CI, −1.17 to −0.16]; P = .01) and a better immunosuppressive profile (NLR >5 at day 8, 2 of 52 [3.8%] vs 24 of 51 [46.0%]; P < .001). Plinabulin was well tolerated, with comparable safety to pegfilgrastim. CONCLUSIONS AND RELEVANCE: Plinabulin has comparable efficacy to pegfilgrastim for the prevention of CIN, with better safety and a better immunosuppressive profile. Plinabulin’s same-day dosing compared with pegfilgrastim’s next-day dosing offers distinct advantages, including reducing use of health care services. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT03102606

IND serial number updated.

Footers were changed to reflect the amendment number and date.

Latin America and Australia removed and South America and Central America added to the list of investigational sites. Part of inclusion criterion 3 revised for Phase 3 from "[patients with] advanced or metastatic NSCLC after failing platinum-based therapy" to "[patients with] advanced or metastatic NSCLC who have previously received platinum based therapy and have either disease progression, are intolerant of platinum-based therapy, or in the opinion of the investigator, would benefit from docetaxel chemotherapy". For Phase 3 only, pain inventory (short version) removed and patient bone pain scale and pain medication assessment added at Days 1 to 8 for Cycle 1. For Phase 3, secondary objectives/endpoints separated into those to be assessed in Cycle 1 and those to be assessed in Cycles 1 to 4. Mean bone pain score and change in bone pain score added as secondary objectives/endpoints for Phase 3; statistical methodology updated accordingly. Incidence of Grade 4 neutropenia in Cycle 1 added as a Cycle 1 secondary objective/endpoint for Phase 3; statistical methodology updated accordingly. Incidences of use of pegfilgrastim or filgrastim as treatment for neutropenia removed from the Phase 3 secondary objectives/endpoints . Area under the curve using the trapezoidal quadrature method for bone pain, from Day 1 through Day 8 in Cycle 1, proportion of patients who needed bone pain medication, and time to first use of bone pain medication added as exploratory objectives/endpoints for Phase 3; statistical methodology updated accordingly. Incidence of Grade 4 neutropenia on any of the Days 8 and 15 in Cycle 1 and on Day 8 in Cycles 2 to 4 moved from a secondary objective/endpoint to an exploratory objective/endpoint for Phase 3. For Phase 2, safety objective/endpoint to assess the "incidence, occurrence and severity of bone pain" updated to "incidences of bone pain". For Phase 3, safety objective/endpoint to assess the incidence, occurrence and severity of bone pain removed.

The Phase 2 exploratory objectives were updated to include:

For selected countries only: to investigate the following cytokine panel: IL-1beta, IL-6, IL-12p70, IL-12p40, IL-17A, IL-23, G-CSF, GM-CSF, IFN-alpha, IFNgamma, TNF-alpha, IL-2, FLT-3 ligand, and IL-8. The Phase 3 primary objective was amended from "[patients with]… advanced or metastatic non-small cell lung cancer (NSCLC) who have previously received platinum based therapy and have either disease progression, are intolerant of platinum-based therapy, or in the opinion of the investigator, would benefit from docetaxel chemotherapy…" to "[patients with] …locally advanced or metastatic non small cell lung cancer (NSCLC) after platinum therapy failure…" The Phase 3 secondary and exploratory objectives were amended following preliminary results from the Phase 2 portion of the study, to include new objectives regarding healthcare utilization and to re-order for hierarchical analysis.

Secondary efficacy objectives were amended to evaluate: o Change in platelet count from baseline in Cycle 1 o The proportion of patients with neutrophil-to-lymphocyte ratio > 5 during Cycle 1 o Area under the concentration-time curve using the trapezoidal quadrature method for bone pain, from Day 1 through Day 8 (pegfilgrastim will be administered on Day 2) in Cycle 1, based on the pain score from the patient bone pain scale o Change in estimated mean bone pain in Cycle 1 o The proportion of patients with thrombocytopenia (all grade) in Cycles 1 to 4 Exploratory efficacy objectives (Cycle 1) were amended to evaluate:

o The proportion of patients in Cycle 1 with: Thrombocytopenia (all grade) Grade 4 neutropenia (absolute neutrophil count [ANC] < 0.5 × 10 9 /L) Grade 3 neutropenia (ANC < 1 × 10 9 /L) Grade 3 (ANC < 1 × 10 9 /L) and Grade 4 neutropenia (ANC < 0.5 × 10 9 /L) Bands > 0 after Day 7 through Day 15 Promyelocytes plus myelocytes > 0 after Day 7 through Day 15 Lymphocyte-to-monocyte ratio < 3. It was clarified that cytokine panel testing will be done using plasma samples collected at the pharmacokinetic (PK) time points and that reconsenting of the patients will be required before analysis is performed. It was specified that unused samples (urine/serum plasma) will also be saved for future potential biomarker research, with patient consent. Timings were included for the new endpoints associated with new objectives (eg healthcare utilization).

The following statement was added: "In order to facilitate balanced treatment arms with respect to cancer type, once either arm reaches at least 1/3 (of total) of patients with that cancer type, it will be closed to that cancer type and enrollment will continue for patients with the other cancer types, up to the planned maximum number of patients. " It was clarified throughout the protocol that patients who withdraw for progressive disease and who will continue to another chemotherapy regimen will complete the end of treatment (EOT) visit at Cycle X Day 21. They then continue to another chemotherapy regimen and do not need to have the safety follow-up visit (where X is the last cycle prior to progression, and is 4 or less). Additionally, if, in the opinion of the investigator, the patient will benefit from more than 4 cycles of docetaxel and open label pegfilgrastim, then the fifth cycle will not start until completion of the EOT visit (in this instance, the EOT visit will be Cycle 4 Day 21) . It was clarified throughout the protocol that patients without progressive disease at the EOT visit will undergo a follow-up visit every 2 months until the occurrence of either disease progression or death. The visits may be conducted per telephone or other means. Inclusion Criterion 7 was amended to include a footnote as follows "Results are from the central laboratory. Local laboratory results may be accepted on a case by case basis after discussion with the Medical Monitor, however in this case central laboratories must also be taken within the screening time window" Exclusion Criterion 17 was revised from "Active Hepatitis B virus (HBV) infection which requires antiviral treatment or the patient has detectable Hepatitis B surface Antigen (HBsAg) "…to "Active Hepatitis B virus (HBV) infection which requires antiviral treatment. Patients with detectable Hepatitis B surface Antigen (HBsAg) may be eligible provided the patient has a negative viral load. Patients with a positive HBsAg must have a negative viral load before each chemotherapy administration... " It was clarified throughout the protocol that new NSAIDs are prohibited but patients already receiving NSAIDs are allowed to continue taking treatment as long as the dose remains stable. It was clarified that if a patient develops a positive viral load during the study, chemotherapy should be held for a maximum of 21 days. If the viral load returns to zero treatment can be restarted, otherwise the patient should be withdrawn from the study. Throughout the protocol, the following text was added to clarify the PK analysis for plinabulin "Plinabulin PK will be characterized using the population PK approach. Population PK parameter estimates, and individual PK parameters, including AUC and Cmax will be summarized." It was clarified that the interim analysis will be performed by an independent statistician. The requirement to document vital signs (position selected, arm, and blood pressure cuff used) and temperature location on the vital signs CRF was removed. 

Investigational sites (to be determined) in North America, South America, Central America, Europe, or Asia Pacific.

Patients will be on study for approximately 5 months, including screening, on treatment, and follow-up. Phase 2/3

Plinabulin pharmacokinetic (PK) and pharmacodynamic (PD) assessments will be made to enable a PK/PD analysis. Primary objective:

To establish the Recommended Phase 3 Dose (RP3D) based on PK/PD analysis.

To assess duration of severe neutropenia (DSN) in treatment Cycle 1 in patients treated with docetaxel (75 mg/m 2 ) + plinabulin (5, 10 or 20 mg/m 2 ) or with docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg). Severe (Grade 4) neutropenia is an absolute neutrophil count (ANC) <0.5 × 10 9 /L. ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose) and during Cycle 1 on Days 1, 2, 6, 7, 8, 9, 10 , and 15 (pre-dose on dosing days; times equivalent to pre-dose on other days) Primary Safety Pharmacodynamic Objective:

To assess blood pressure semi-continuously with 15-minute intervals, starting 15 minutes pre-plinabulin dose and lasting 4.5 hours after start of infusion with plinabulin (Arms 2 to 4) or for 4.75 hours starting 15 minutes after the end of docetaxel infusion (Arm 1).

To characterize the pharmacokinetic profile of plinabulin and docetaxel

To characterize the exposure-response relationships between measures of plinabulin exposure and the pharmacodynamic endpoint DSN To characterize the exposure-safety relationships between measures of plinabulin exposure and safety events of interest Exploratory Objectives:

To assess CD34+ at screening, and on Days 2, 6, and 8 in Cycle 1 and Day 1 in Cycle 2

Health-related Quality of Life (QoL) questionnaire evaluated with EORTC QLQ-C30 and EQ-5D-5L

To collect data on disease progression Protocol BPI-2358-105 IND cytokine panel: IL-1beta, IL-6, IL-12p70,  IL-12p40, IL-17A, IL-23, G-CSF, GM-CSF, IFN-alpha, IFN-gamma, TNF-alpha, IL-2, FLT-3  ligand, and IL-8  Safety Objectives Incidence, occurrence, and severity of adverse events (AEs)/serious adverse events (SAEs)

Safety and tolerability (physical examination and safety laboratory assessments)

Primary Efficacy Objective (Cycle 1):

To assess DSN in treatment Cycle 1 in patients with advanced or metastatic breast cancer, who have failed < 5 prior lines of chemotherapy; locally advanced or metastatic non small cell lung cancer (NSCLC) after platinum therapy failure; or hormone refractory (androgen independent) metastatic prostate cancer (HRPC) treated with docetaxel (75 mg/m 2 ) + plinabulin (40 mg) (Arm 2) versus docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) (Arm 1). ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose) and during Cycle 1 on Days 1, 2, 6, 7, 8, 9, 10, and 15 . Blood draws for ANC will be taken approximately the same time as the time of the pre-dose sample on Day 1, and will be taken by preference in the morning. DSN should be calculated as the number of consecutive days from the first day when a patient's ANC is below 0.5 x l09/L until the patient reaches an ANC> 0.5 x 109/L, in Cycle 1. For patients who do not experience any severe neutropenia in Cycle 1, the DSN is set to 0. For patient's experiencing several episodes, the number of days of DSN will be summed up.

Platelet count in Cycle 1: maximum decrease from baseline (prior to Cycle 1 docetaxel dose)

Proportion of patients with neutrophil-to-lymphocyte ratio (NLR) > 5 after Day 7 through Day 15 in Cycle 1

Area under the concentration-time curve (AUC) using the trapezoidal quadrature method for bone pain, from Day 1 through Day 8 (pegfilgrastim will be administered on Day 2) in Cycle 1, based on the pain score from the patient bone pain scale To compare the time (in days) to first use of bone pain medication between the treatment groups To assess DSN in Cycle 1 in patients with locally advanced or metastatic NSCLC after platinum therapy failure To assess DSN in Cycle 1 in patients with HRPC To assess DSN in Cycle 1 in patients with advanced or metastatic breast cancer who have failed < 5 prior lines of chemotherapy To assess DSN in Cycle 1 in patients with locally advanced or metastatic NSCLC after platinum therapy failure or HRPC Platelet count at least 30% change from baseline at any time during Cycle 1 ANC nadir during Cycle 1

For selected countries only: to investigate the following cytokine panel: IL-1beta, IL-6, IL-12p70, IL-12p40, IL-17A, IL-23, G-CSF, GM-CSF, IFN-alpha, IFN-gamma, TNF-alpha, IL-2, FLT-3 ligand, and IL-8.

To evaluate the proportion of patients in Cycles 1 to 4 with:

Febrile neutropenia (FN) (ANC <1.0 × 10 9 /L AND a single temperature of >38.3°C or a sustained temperature of ≥38°C for more than 1 hour).

Grade 4 neutropenia (ANC < 0.5 × 10 9 /L). o Health-related QoL questionnaire evaluated with European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30 and EQ-5D-5L Safety Objectives:

Incidence, occurrence, and severity of adverse events (AEs)/serious adverse events (SAEs) Safety and tolerability (physical examination and safety laboratory assessments)

This is a multicenter, randomized study with an open label phase 2 portion and a double blind phase 3 portion. Approximately 190 patients will be enrolled in this study.

In the Phase 3 portion, on Day 1 of each cycle, 1.5 hours (± 10 minutes) after the start time of docetaxel infusion (i.e., approximately 30 minutes after the end of docetaxel infusion), patients will get a single dose of plinabulin or placebo intravenously over 30 minutes (± 5 minutes). On Day 2 of each cycle, ≥24 hours after completing chemotherapy, patients will receive a single dose of pegfilgrastim (6 mg) or placebo (subcutaneous injection). If a chemotherapy cycle is delayed by more than 3 weeks, the patient will be withdrawn from the study. If a critical AE (refer to Section 10.5) occurs during the cycle, the dosage of docetaxel may be reduced 20% in the next cycle. Only one docetaxel dose reduction is allowed (refer to Taxotere® (Prescribing Information) . No dose reductions are allowed with plinabulin or pegfilgrastim. All patients, including patients who withdraw from the study early, will complete a safety follow-up visit 30 days (± 2 days) after the last dose of study drug. Patients who withdraw for progressive disease and who will continue to another chemotherapy regimen will complete the End of Treatment (EOT) visit at Cycle X Day 21. They then continue to another chemotherapy regimen and do not need to have the safety follow-up visit (where X is the last cycle prior to progression, and is 4 or less). If, in the opinion of the investigator, the patient will benefit from more than 4 cycles of docetaxel and open label pegfilgrastim, then the fifth cycle will not start until completion of the EOT visit (in this instance, the EOT visit will be Cycle 4 Day 21) . Follow-up visits will be required to monitor for ongoing treatment-related AEs. All patients experiencing drug-related toxicities of Grade 2 at the EOT visit should be followed-up at least monthly until the AE(s) resolves to Grade 1, the event is considered to be chronic, or the patient receives other anti-cancer therapy. The method of follow-up assessment will be at the Investigator's discretion (for example, patient site visit or telephone call). All deaths which occur within 30 days of study drug administration regardless of relationship to the study drug must be reported to the Sponsor immediately and within 24 hours of becoming aware of the event. Patients without progressive disease at the EOT visit will undergo a follow-up visit every 2 months until the occurrence of either disease progression or death. These visits may be conducted per telephone or other means. Laboratory test results (hematology and serum chemistry) will be collected via a central laboratory. Safety laboratory tests are required prior to treatment on Day 1 of each cycle and can be collected by a local laboratory; however, all other scheduled blood samples as per the schedule assessments and procedure table must also be obtained for central laboratory assessment. Urinalysis will be performed at screening only. CD34+ counts will be established through a fluorescence-activated cell sorting (FACS) method as described in Table 6 .

During the phase 2 study if at any given cohort, 3 patients or more have Grade 4 or 5 toxicity not related to underlying disease (with the exception of neutropenia), accrual to that cohort will be halted and the study will be continued at the lower dose cohorts in phase 2 (for example if 3 patients at the 20 mg/m 2 cohort develop Gr. 4 toxicity the accrual to that cohort will be stopped and the study will continue as planned with the accrual of the two remaining open cohorts). Interactive web response system (IWRS) will be utilized to assign patients to a lower dose cohort in phase 2. Study sites will be instructed to call IWRS when a Grade 4 or 5 toxicity event occurs.

Approximately 40 patients will be enrolled in Phase 2 and approximately 150 patients are planned to be enrolled in Phase 3.

1. At least ≥ 18 years of age (male or female) at the time of signing the informed consent form. 2. ECOG performance status of 0 or 1.

Phase 2 only:

Advanced or metastatic NSCLC failing platinum-based therapy Phase 3 only:

Advanced or metastatic breast cancer, who have failed < 5 prior lines of chemotherapy (Note that study treatment may be the first chemotherapy treatment for advanced or metastatic cancer) Locally advanced or metastatic NSCLC after platinum therapy failure HRPC (Note that study treatment may be the first chemotherapy treatment). Protocol BPI-2358-105 IND 4. Pathology confirmation of cancer is required. 5. Patients with ≥1 of the following risk factors, at the initiation of docetaxel chemotherapy, that would require neutropenia prophylaxis per National Comprehensive Cancer Network (NCCN) guidelines (version 2, 2016) Myeloid Growth Factors (refer to Appendix A): a. Prior chemotherapy or radiation treatment b. Bone marrow involvement by tumor c. Surgery and/or open wounds within 4 weeks of first administration of study drug d. Age > 65 years of age and receiving full chemotherapy dose intensity 6. Life expectancy of 3 months or more. 7. The following laboratory results assessed within 14 days prior to study drug administration: 1 Hemoglobin 9 g/dL independent of transfusion or growth factor support ANC 1.5 x 10 9 /L independent of growth factor support Serum total bilirubin 1.5 times the upper limit normal (ULN), unless the patient has a diagnosis of Gilbert's disease in which case direct bilirubin < 1.5 times ULN of the direct bilirubin.

Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) 2.5 x ULN ( 1.5 x ULN if alkaline phosphatase is > 2.5 x ULN) Serum creatinine 1.5 x ULN 8. Prothrombin time (PT) and International Normalized Ratio (INR) ≤ 1.5 × upper limit of normal (ULN), activated partial thromboplastin time (PTT) ≤ 1.5 × ULN, based on central laboratory results. 1 9. Female subjects of childbearing potential have a negative pregnancy test at screening. Females of childbearing potential are defined as sexually mature women without prior hysterectomy or who have had any evidence of menses in the past 12 months. However, women who have been amenorrhoeic for 12 or more months are still considered to be of childbearing potential if the amenorrhea is possibly due to prior chemotherapy, anti-estrogens, or ovarian suppression.

Women of childbearing potential (i.e., menstruating women) must have a negative urine pregnancy test (positive urine tests are to be confirmed by serum test) documented within the 24hour period prior to the first dose of study drug.

Sexually active women of childbearing potential enrolled in the study must agree to use two forms of accepted methods of contraception during the course of the study and for 3 months after their last dose of study drug. Effective birth control includes (a) intrauterine device (IUD) plus one barrier method; (b) on stable doses of hormonal contraception for at least 3 months (e.g., oral, injectable, implant, transdermal) plus one barrier method; (c) 2 barrier methods. Effective barrier methods are male or female condoms, diaphragms, and spermicides (creams or gels that contain a chemical to kill sperm); or (d) a vasectomized partner.

For male patients who are sexually active and who are partners of premenopausal women: agreement to use two forms of contraception during the treatment period and for at least 3 months after the last dose of study drug. Current use of strong cytochrome P450 (CYP) 3A4 inhibitors, within 3 days of the first administration of study drug, and 7 days after treatment with taxanes OR requires use of strong CYP3A4 inhibitors (refer to Section 10.6.2). 6. Received an investigational agent or tumor vaccine within 2 weeks before the first dose of study drug; patients must have recovered from toxicity of prior treatment and have no > Grade 1 treatment emergent AEs. 7. Receiving any concurrent anticancer therapies. 8. Received a prior bone marrow or stem cell transplant. 9. Has a co-existing active infection or received systemic anti-infective treatment within 72 hours before the first dose of study drug. 10. Prior radiation therapy within the 4 weeks before the first dose of study drug. 11. Prior use of pegfilgrastim or filgrastim within 4 weeks before the first dose of study drug. 12. Presence of any serious or uncontrolled illness including, but not limited to: uncontrolled diabetes, ongoing or active infection, symptomatic congestive heart failure, unstable angina pectoris, uncontrolled cardiac arrhythmia, uncontrolled arterial thrombosis, symptomatic pulmonary embolism, or psychiatric illness that would limit compliance with study requirements, or any other conditions that would preclude the patient from study treatment as per the discretion of the Investigator. 13. Significant cardiovascular history: History of myocardial infarction or ischemic heart disease within 1 year (within a window of up to 18 days less than 1 year) before first study drug administration; Uncontrolled arrhythmia;

History of congenital QT prolongation; Electrocardiogram (ECG) findings consistent with active ischemic heart disease;

New York Heart Association Class III or IV cardiac disease;

Uncontrolled hypertension: blood pressure consistently >150 mm Hg systolic and > 100 mm Hg diastolic in spite of antihypertensive medication. 14. History of hemorrhagic diarrhea, inflammatory bowel disease, or active uncontrolled peptic ulcer disease.

(Concomitant therapy with ranitidine or its equivalent and/or omeprazole or its equivalent is acceptable). History of ileus or other significant gastrointestinal disorder known to predispose to ileus or chronic bowel hypomotility. 15. Any other malignancy requiring active therapy. 16 . Known human immunodeficiency virus (HIV) seropositivity.

Hepatitis B surface Antigen (HBsAg) may be eligible provided the patient has a negative viral load. Patients with a positive HBsAg must have a negative viral load before each chemotherapy administration. Hepatitis B surface antibody (anti HBs) without detectable HBsAg does NOT exclude patients from the study. Hepatitis C infection (Hepatitis C antibody reactive) which requires treatment also excludes patients from the study. 18. Female subject who is pregnant or lactating. 19. Unwilling or unable to comply with procedures required in this protocol Study Treatments Phase 2 (approximately 10 patients per arm with advanced or metastatic NSCLC after failing platinum-based therapy): Arm 1: Docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) Arm 2: Docetaxel (75 mg/m 2 ) + plinabulin (20 mg/m 2 ) Arm 3: Docetaxel (75 mg/m 2 ) + plinabulin (10 mg/m 2 ) Arm 4: Docetaxel (75 mg/m 2 ) + plinabulin (5 mg/m 2 ) Protocol BPI-2358-105 IND Phase 3 (a planned 75 patients per arm with advanced or metastatic breast cancer, who have failed < 5 prior lines of chemotherapy; locally advanced or metastatic NSCLC after platinum therapy failure; or HRPC): Arm 1: Docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) + placebo matching plinabulin Arm 2: Docetaxel (75 mg/m 2 ) + plinabulin (40 mg) + placebo matching pegfilgrastim All patients must be pre-medicated with oral corticosteroids such as dexamethasone 16 mg per day (e.g., 8 mg bid) for 3 days starting 1 day prior to docetaxel administration in order to reduce the incidence and severity of fluid retention as well as the severity of hypersensitivity reactions (refer to Taxotere® (Prescribing Information) ). For hormone-refractory metastatic prostate cancer, given the concurrent use of prednisone, the recommended premedication regimen is oral dexamethasone 8 mg, at 12 hours, 3 hours and 1 hour before the docetaxel infusion (refer to Taxotere® (Prescribing Information)).

Patients who experience an FN event in Cycle 1 should be discussed with the medical monitor. The blinding will be broken (if Phase 3), and patients assigned to plinabulin will receive pegfilgrastim in subsequent cycles. If a patient was originally assigned to the pegfilgrastim arm, patients must be treated at a lower dose of docetaxel, or taken off study at the discretion of the investigator. If the patient develops an FN event on subsequent cycles, the patient should be discussed with the medical monitor and either treated with a lower dose of docetaxel, or taken off study at the discretion of the investigator. Febrile neutropenia should be treated with antibiotics per institutional standard of care (refer to Section 10.6.1). If a patient is hospitalized, the procedure for reporting Serious Adverse Events (Section 13.2) should also be followed.

Patients will receive treatment with study drug for up to 4 cycles in this study, a treatment cycle is 21 days; thereafter, patients may continue receiving docetaxel and pegfilgrastim at the Investigator's discretion. After completion of 4 cycles, patients will complete a safety follow-up visit 30 days (± 2 days) after the last dose of study drug (see Table 6 and Table 7 Study Assessments and Procedures Schedule) . If, in the opinion of the investigator, the patient will benefit from more than 4 cycles of docetaxel and open label pegfilgrastim, then the fifth cycle will not start until completion of the EOT visit (in this instance, the EOT visit will be Cycle 4 Day 21) . Patients with progressive disease requiring another treatment will also complete the EOT visit at Cycle X Day 21 and can then continue on another chemotherapy regimen (where X is the last treatment cycle prior to progression, and is 4 or less). Treatment up to 4 cycles of study drug in this study will continue until any 1 of the following occurs:

Dose limiting toxicity or critical adverse event as described in the docetaxel package insert (Taxotere® (Prescribing Information)) Need for a protocol-prohibited dose reduction or study drug delay greater than 21 days

Initiation of a protocol-prohibited concomitant medication or non-protocol chemo/biological therapy for treatment of his or her disease Development of a AE/SAE, illness, or condition that may interfere with the patient's participation or require treatment discontinuation Investigator opinion

Voluntary withdrawal of consent

Institutional guidelines should be followed in the event of infusion/hypersensitivity reaction. Diphenhydramine and dexamethasone infusion may be administered in the event of infusion reaction. All patients must be premedicated with oral corticosteroids (see below for HRPC) such as dexamethasone 16 mg per day (e.g., 8 mg twice daily) for 3 days starting 1 day before the docetaxel administration in order to reduce the incidence and severity of fluid retention as well as the severity of hypersensitivity reactions (refer to Taxotere® (Prescribing Information)). For HRPC, given the concurrent use of prednisone, the recommended premedication regimen is oral dexamethasone 8 mg at 12 hours, 3 hours, and 1 hour before docetaxel infusion (refer to Taxotere® (Prescribing Information)). The prophylactic use of antibiotics is allowed at the discretion of the Investigator. The use of antibiotics will be recorded by patient and summarized by treatment arm (refer to Table 5 ). Corticosteroids (except as described for premedication) and new non-steroidal anti-inflammatory drugs (NSAIDs) are prohibited except for the treatment of AEs and as premedication. Patients already receiving NSAIDs are allowed to continue taking treatment as long as the dose remains stable. The use of strong CYP3A4 inhibitors as concomitant medications will be prohibited because docetaxel exposure increases by approximately 2-fold (Section 10.6.2) with the use of strong CYP3A4 inhibitors. Docetaxel is low emetic risk (10% to 30% frequency of emesis) and appropriate anti-emetic prophylaxis should be given prior to study medication per institutional policy. In this study, dexamethasone is used as anti-emetic prophylaxis as well as to minimize docetaxel associated fluid retention.

If "breakthrough" emesis (e.g. emesis and nausea after Day 1) occurs, the general strategy is to add one agent from a different drug class to the "rescue" anti-emesis regimen. Useful anti-emetic agents for rescue include the benzodiazepines, cannabinoids (e.g. dronabinol or nabilone), haloperidol, metoclopramide, scopolamine, the phenothiazines (e.g. prochlorperaxine or promethazine) and 5HT3 receptor agonists. If a 5HT3 receptor agonist is needed, palonosetron (which is not known to prolong QT/QTc intervals) is safest and must be chosen over other 5HT3 receptor agonists. Tropisetron is an acceptable substitute 5HT3 receptor agonist (Yavas et al, 2008) . Because of the potential interference with QT/QTc interval, in Cycle 1 between Day 1 and Day 2, the 5HT3 receptor agonists ondansetron, granisetron and dolenasetron, and the atypical antipsychotic olanzapine, are prohibited until the triplicate ECGs are completed. After Cycle 1, triplicate ECGs are not obtained, and therefore no restrictions in the use of anti-emetics apply in those cycles, thus any 5HT3 receptor agonist can be used. If nausea and /or vomiting of Grade 2 and higher occurs, it must be treated with "rescue" anti-emetics during mid-cycle, and on subsequent cycles, the prophylactic anti-emetic regimen should be modified. Aprepitant is an acceptable prophylactic anti-emetic in this situation [Marbury et al, 2009] ). If diarrhea of Grade 1 and higher occurs, it must be treated. Grade 1 diarrhea is less than 4 bowel movements a days without any signs of hypotention, dehydration. Antidiarrheals such as loperamide (or diphenoxylate/atropine) must be prescribed for diarrhea. Suggested loperamide use: 4 mg orally after first loose stool, then 2 mg after each stool not to exceed 16 mg in 24 hours. The use of anti-emetics and anti-diarrheals will need to be recorded on the CRF.

Patients who have FN should receive antibiotics per standard of care (refer to Rescue Treatment section and Table 5 for prohibited medications). The use of granulocyte colony-stimulating factor (G-CSF) as a treatment option during hospitalization for FN is strongly discouraged, since G-CSF is not approved for the treatment of FN, and is not likely to be effective. If, however, G-CSF treatment for FN is considered, the Investigator should contact the Medical Monitor prior to its use. FN is defined ANC < 1.0 × 10 9 /L AND a single temperature of >38.3°C or a sustained temperature of ≥38°C for more than 1 hour (as per CTCAE v4.03).

Plinabulin PK will be characterized using the population PK approach. Population PK parameter estimates, and individual PK parameters, including AUC and Cmax will be summarized. Patients in Phase 3 will follow the plinabulin and docetaxel PK sampling schedules from Phase 2.

Patients in phase 2 will participate in the PD assessments. The PD assessments include blood pressure, DSN, and area over the neutropenia curve (AOC) in cycle 1 of the phase 2 portion of the study.

Phase 2 only: A blood sample for exploratory marker evaluation will be collected for CD34+ which will be established by FACS. This test will be performed in selected countries participating in the study, via central laboratory. cytokine panel: IL-1beta, IL-6, IL-12p70, IL-12p40, IL-17A, IL-23, GCSF,  GMCSF, IFN alpha, IFN-gamma, TNF-alpha, IL-2, FLT-3 ligand, and IL-8 analysis. This test will be performed in selected countries participating in the study, using unused plasma samples collected at PK time points.

Note: Patients in Phase 3 will be stratified based on his or her tumor type (breast cancer, NSCLC, HRPC) and region (Asia, non-Asia).

Phase 2 (Open Label) Plinabulin pharmacokinetic (PK) and pharmacodynamic (PD) assessments will be made to enable a PK/PD analysis.

The primary endpoint is to establish the Recommended Phase 3 Dose (RP3D) based on PK/PD analysis

DSN in treatment Cycle 1 in patients treated with docetaxel (75 mg/m 2 ) + plinabulin (5, 10 or 20 mg/m 2 ) or with docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg). ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose) and during Cycle 1 on Days 1, 2, 6, 7, 8, 9, 10 , and 15 (pre-dose on dosing days; times equivalent to pre-dose on other days). The Negative Binomial Regression (NBR) model, on the integer number of days of severe neutropenia, will be used to analyze the DSN endpoint during the fixed time window outlined above, with the treatment arm as the only covariate.

The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used. Contrasts for pairwise comparisons between Arm 1 and the plinabulin arms will be used for the evaluation. A closed testing procedure by Hommel (Hommel 1988, Hommel and Bernhard 1999) will be used for the multiple comparisons.

To assess blood pressure semi-continuously with 15-minute intervals, starting 15 minutes pre-plinabulin dose and lasting 4.5 hours after start of infusion with plinabulin (Arms 2 to 4) or for 4.75 hours starting 15 minutes after the end of docetaxel infusion (Arm 1). See the Pharmacometrics Analysis Plan for the analysis methods.

To characterize the pharmacokinetic profile of plinabulin and docetaxel. See the Pharmacometrics Analysis Plan for the analysis methods. To characterize the exposure-response relationships between measures of plinabulin exposure and the pharmacodynamic endpoint DSN. Robust linear regression (Bablok et al, 1988) with DSN as the dependent variable and plinabulin exposure as the independent variable. The method will also be used to construct point estimates and confidence intervals. The NCSS version 12 statistical software will be used. To characterize the exposure-safety relationships between measures of plinabulin exposure and safety events of interest. Robust linear regression (Bablok et al, 1988) with the number of safety events as the dependent variable and plinabulin exposure as the independent variable. The method will also be used to construct point estimates and confidence intervals. The NCSS version 12 statistical software will be used.

To assess CD34+ at screening, and on Days 2, 6 and 8 in Cycle 1 and Day 1 in Cycle 2. A repeated measures mixed linear model with the Day 1 value and treatment arm as covariates will be used to analyze this endpoint. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS version 9.4 or later will be used for the analysis. Contrasts Protocol BPI-2358-105 IND for pairwise comparisons between Arm 1 and the plinabulin arms will be used for the evaluation. A closed testing procedure by Hommel (Hommel 1988, Hommel and Bernhard 1999) will be used for the multiple comparisons.

Health-related QoL questionnaire evaluated with EORTC QLQ-C30 and EQ-5D-5L. The Wilcoxon Rank Sum test will be used to analyze the responses to the individual questions. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis. Contrasts for pairwise comparisons between Arm 1 and the plinabulin arms will be used for the evaluation. A closed testing procedure by Hommel (Hommel 1988, Hommel and Bernhard 1999) will be used for the multiple comparisons.

Data collection on disease progression. The log-rank test (LIFETEST procedure in SAS) will be used to compare time to disease progression between the treatment groups. The method will also be used to construct point estimates and confidence intervals. Contrasts for pairwise comparisons between Arm 1 and the plinabulin arms will be used for the evaluation. A closed testing procedure by Hommel (Hommel 1988, Hommel and Bernhard 1999) will be used for the multiple comparisons.

For selected countries only: to investigate the following cytokine panel: IL-1beta, IL-6, IL-12p70, IL-12p40, IL-17A, IL-23, G-CSF, GM-CSF, IFN-alpha, IFN-gamma, TNF-alpha, IL-2, FLT-3 ligand, and IL-8.

Incidence, occurrence, and severity of AEs/SAEs. These endpoints will be presented descriptively in tables and listings.

Incidences of bone pain. The NBR model will be used to analyze the endpoint during the fixed time window from pre-dose Day 1 through Day 8 in Cycle 1, with the treatment arm as the only covariate. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used. Contrasts for pairwise comparisons between Arm 1 and the plinabulin arms will be used for the evaluation. A closed testing procedure by Hommel (Hommel 1988, Hommel and Bernhard 1999) will be used for the multiple comparisons. Safety and tolerability (physical examination and safety laboratory assessments). These endpoints will be presented descriptively in tables and listings.

DSN in Cycle 1 in patients treated with docetaxel (75 mg/m 2 ) + plinabulin (40 mg) (Arm 2) compared with patients treated with docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) (Arm 1). ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose) and during Cycle 1 on Days 1, 2, 6, 7, 8, 9, 10, and 15 . Blood draws for ANC will be taken approximately the same time as the time of the pre-dose sample on Day 1, and will be taken by preference in the morning. The NBR model, on the integer number of days of severe neutropenia, will be used to analyze the DSN endpoint during the fixed time window outlined above, with the treatment arm as the only covariate. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used.

Maximum decrease from baseline (prior to docetaxel dose) in platelet count for each patient in Cycle 1. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

Proportion of patients with NLR > 5 after Day 7 through Day 15 in Cycle 1. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis. AUC using the trapezoidal quadrature method for bone pain, from Day 1 through Day 8 in Cycle 1,  based on the bone pain score from the patient bone pain scale. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

Estimated mean pain score from the patient bone pain scale from pre-dose Day 1 through Day 8 in Cycle 1. A repeated measures mixed linear model with the pre-dose Day 1 value and treatment arm as covariates will be used to analyze this endpoint. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS version 9.4 or later will be used for the analysis.

Proportion of patients with thrombocytopenia (all grade) in Cycles 1 to 4. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

Proportion of patients in Cycle 1 with:

Thrombocytopenia (all grade). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

Grade 3 (ANC < 1 × 109/L) and ANC >= 0.5 × 109/L) and Grade 4 neutropenia (ANC < 0.5 × 109/L). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

Grade 4 neutropenia (ANC < 0.5 × 109/L). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

Grade 3 neutropenia (ANC < 1 × 109/L) and ANC >= 0.5 × 109/L). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

Bands > 0 after Day 7 through Day 15. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

Promyelocytes plus myelocytes >0 after Day 7 through Day 15. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

LMR < 3.2 after Day 7 through Day 15; time course of percentage of patients with LMR < 3.2 over time in Cycle 1. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

PLR > 200 after Day 7 through Day 15; time course of percentage of patients with PLR > 200 over time in Cycle 1. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

For the below bone pain measurements The Barnard's test will be used to evaluate the difference Proportion of patients in Cycle 1 who needed bone pain medication (defined as any medication reported on the pain medication assessment from Day 1 through Day 8) . The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

Time (in days) to first use of bone pain medication. The log-rank test (LIFETEST procedure in SAS) will be used to compare time to first use of bone pain medication between the treatment groups. Separate analyses will be made with respect to narcotic and non-narcotic analgesics.

DSN in Cycle 1 in patients with locally advanced or metastatic NSCLC after platinum failure. The analysis will be done as per the primary endpoint.

DSN in Cycle 1 in patients with HRPC. The analysis will be done as per the primary endpoint.

DSN in Cycle 1 in patients with advanced or metastatic breast cancer who have failed < 5 prior lines of chemotherapy. The analysis will be done as per the primary endpoint.

DSN in Cycle 1 in patients with locally advanced or metastatic NSCLC after platinum failure or HRPC. The analysis will be done as per the primary endpoint.

Platelet count at least 30% change from baseline at any time during Cycle 1. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis. ANC nadir during Cycle 1. The Wilcoxon Rank Sum test will be used to analyze the endpoint. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis. For selected countries only: maximum change from baseline in the following cytokine panel:  IL-1beta, IL-6, IL-12p70, IL-12p40, IL-17A, IL-23, G-CSF, GM-CSF, IFN-alpha, IFN-gamma, TNF-alpha, IL-2, FLT-3 ligand, and IL-8. The Wilcoxon Rank Sum test will be used to analyze the endpoint. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

Proportion of patients in Cycles 1 to 4 with: FN (ANC <1.0 × 10 9 /L AND a single temperature of >38.3°C or a sustained temperature of ≥38°C for more than 1 hour). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

Grade 4 neutropenia (ANC < 0.5 × 10 9 /L). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis. Healthcare utilization endpoints: o Incidence of 30-day rehospitalizationsall cause. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used. o Incidence of all cause hospitalizations. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used. o Duration of all cause hospitalizations. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis. o Incidence of all cause ER visits. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used. o Incidence of all cause ICU stays. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used. o Duration of all cause ICU stays. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis. o Incidence of all cause docetaxel dose delay, dose reduction, or dose discontinuation. The NBR model will be used to analyze the endpoints, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used: The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis o o Incidence of transfusions due to thrombocytopenia. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used. o Incidence of antibiotics use. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used. Data collection on disease progression. The log-rank test (LIFETEST procedure in SAS version 9.4 or later) will be used to compare time to disease progression between the treatment groups. The method will also be used to construct point estimates and confidence intervals.

QoL endpoints will be analyzed in safety analysis set: o Health-related QoL questionnaire evaluated with EORTC QLQ-C30 Item 30 (response to question "How do you rate your overall quality of life during the past week") in Cycle 1 to 4 will be analyzed using linear mixed model for repeated measures (MMRM) with the Cycle 1 Day 1 value and treatment arm as covariates. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS v9.4 or later will be used for the analysis. The summary statistics in terms of counts, means, standard deviations, medians, minimums, and maximums for individual questions will be summarized in a tabular format. o Health-related QoL questionnaire evaluated with EORTC QLQ-C30 will also be analyzed with the summary scores. The EORTC QLQ-C30 scoring method combines the 30 questions from the questionnaire into 15 scores: global health state/QoL, functional scale (5 items), and symptoms scale (9 items) (Fayers et al, 2001) . Three summary measures will be constructed: the QLQ-C30 Summary Score, which combined the symptom and functional scales (excluding the financial difficulties item), the symptom summary score (excluding the financial difficulties item), and the functional summary score (Hinz et al, 2012) . The symptom summary score will be analyzed using linear mixed model for repeated measures (MMRM) with the Cycle 1 Day 1 value and treatment arm as covariates. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS v9.4 or later will be used for the analysis. The summary statistics in terms of counts, means, standard deviations, medians, minimums, and maximums for the 15 scores, the QLQ-C30 summary score, symptom score and the functional summary score will be summarized in a tabular format. o The response to the question "Your Health Today" from EQ-5D-5L in Cycle 1 to 4 will be analyzed using linear mixed for model repeated measures (MMRM) with the Cycle 1 Day 1 value and treatment arm as covariates. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS v9.4 or later will be used for the analysis. The summary statistics in terms of counts, means, standard deviations, medians, minimums, and maximums for the question "Your Health Today" will be summarized in a tabular format. o The EQ-5D-5L data will be converted into a health utility score using the interim EQ-5D-5L crosswalk value set for the United States (van Hout et al, 2012) unless a validated EQ-5D-5L value set for the United States becomes available. The EQ-5D-5L health utility score in Cycle 1 to 4 will be analyzed using linear mixed model for repeated measures (MMRM) with the Cycle 1 Day 1 value and treatment arm as covariates. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS v9.4 or later will be used for the analysis. The summary statistics in terms of counts, means, standard deviations, medians, minimums, and maximums for the health utility score will be summarized in a tabular format.

Incidence, occurrence, and severity of AEs/SAEs. These endpoints will be presented descriptively in tables and listings. Safety and tolerability (physical examination and safety laboratory assessments). These endpoints will be presented descriptively in tables and listings.

Primary efficacy endpoint will be imputed by multiple imputation method as sensitivity analysis.Analysis Sets Phase 2

The intent-to-treat analysis set for Phase 2 is comprised of all Phase 2 patients that have been randomized in the study and have received at least one dose of study medication. The analysis of all endpoints, unless noted otherwise, will be conducted on the intent-to-treat analysis set. 

The safety analysis set will be the same as the intent-to-treat analysis set for Phase 2.

All patients who received at least 1 dose of plinabulin or docetaxel and had pre-dose sample collection and at least 1 PK sample collected postdose will be included in the PK analysis set.

All patients who had blood pressure and DSN collected at any time during the study will be included in the PD analysis set.

The intent-to-treat (ITT) analysis set for Phase 3 is comprised of all Phase 3 patients that have been randomized in the study. The analysis of all endpoints, unless noted otherwise, will be conducted on the intentto-treat analysis set.

The safety analysis set is comprised of all Phase 3 patients that have been randomized in the study and have received at least one dose of study medication.

Primary efficacy analysis considerations DSN was not measured in the previous Phase 2 study ([Study NPI-2358-101] . The study will assume that the shape of the time/neutrophil recovery curve in plinabulin-treated patients is indistinguishable from the time/neutrophil recovery curve for filgrastim and its biosimilars. In a study with filgrastim and its biosimilar, time course of ANC in Cycle 1 for the Per Protocol dataset was published by . Mean values and standard deviations of ANC during the 21-day follow-up period were readily available. This information was used to write a computer simulation program that would generate random ANC data that asymptotically has the same means and standard deviations for the 21-day follow-up period as the publication. The simulation would then also generate the projected number of days with severe neutropenia, (i.e., the DSN). Deming regression was used to calculate the linear relationship between simulated nadir and DSN. The rank correlation between simulated nadir and DSN was used to calculate the DSN with plinabulin (+ docetaxel) and docetaxel alone. In the Phase 2 study, ANCs were obtained on Day 8, which approximately coincides with the time that the neutrophil nadir occurs after docetaxel administration. The observed Day 8 neutrophil (nadir) values were computed into the linear relationship (Deming regression), mentioned above to calculate DSN for each patient. Using these methods, calculated mean DSN was 0.065 days for the plinabulin+ docetaxel arm, and 1.076 days for the docetaxel alone. Based on published data with filgrastim in patients receiving docetaxel (Alexopoulos K et al, 1999) , the assumption is that Grade 4 neutropenia in Cycle 1 would occur in a 2 times higher frequency with G-CSF+docetaxel versus plinabulin+docetaxel, resulting in a presumed mean DSN of 0.13 days for the G-CSF+ docetaxel combination. This non-inferiority trial design will utilize a difference (arm 2 minus arm 1) of 0.65 days (non-inferiority margin) in DSN in Cycle 1 as the largest acceptable difference between plinabulin and pegfilgrastim. The non-inferiority test will evaluate the null hypothesis H0: true difference (arm 2 minus arm 1) ≥ 0.65 against the alternative hypothesis H1: true difference (arm 2 minus arm 1) < 0.65. Plinabulin will be considered non-inferior to pegfilgrastim if in Cycle 1, the upper limit of the 2-sided 95% confidence interval for the true difference in mean duration of Grade 4 neutropenia was < 0.65 days. A sample size of patients was based on sample size considerations as outlined. (Aarts M et al, 2013) , which would translate into a DSN of 1 day according to . Based on these data, it is assumed that the median DSN for docetaxel monotherapy + G-CSF will be approximately 1 day. The frequency of FN with docetaxel monotherapy (without G-CSF) has been reported to be 11% in Cycle 1 (17% over all cycles) docetaxel dose of 100 mg/m 2 ) and 19.8% over all cycles at a lower docetaxel dose of 60 mg/m 2 . Hanna N et al, 2004 reported an FN percentage of 12.7% with 75 mg/m 2 docetaxel. Based on this range of FN, the relationship established by between FN and DSN, we make the assumption that, with docetaxel monotherapy at a dose of 75 mg/m2 without G-CSF, the median DSN is estimated to be 4-5 days.

In the Zarxio® briefing document, 2015, the margin was selected based on the fact that Taxotere/Adriamycin/cyclophosphamide (TAC) chemotherapy is known to induce a median DSN of 7 days in breast cancer patients receiving no G-CSF treatment , while G-CSF treatment reduces the mean DSN for this chemotherapy to 1.4 days (95% CI: 1.07 -1.69) as shown in pegfilgrastim (Neulasta®) Study 20020778 (Kaufman et al, 2004) . Based on this a non-inferiority limit of 1 day was derived.

As an extension of this reasoning, it is argued for our study, a non-inferiority margin of 0.65 would be reasonable and correspond to approximately a median of 4.5 days of DSN, as a ratio of 1 day to 7 days of DSN in the Zarxio® briefing document, 2015.

A non-inferiority margin of 0.65 days can also be justified, because a difference of 0.65 days is not considered to be clinically meaningful.

Plasma plinabulin and docetaxel concentrations will be measured using validated methods and PK parameters will be summarized using descriptive statistics. Individual and mean serum plinabulin and docetaxel concentration versus time profiles will be plotted on both linear and logarithmic scales. Exposure-Neutropenia Relationship A sequential population PK/PD modeling approach will be used to characterize the exposure neutropenia relationship. The semi-physiological model described by Friberg (2003) will be used to characterize the time course of neutropenia in the exposure-neutropenia model. The developed exposure-response models will be used to simulate ANC profiles for the computation of key PD endpoints such as DSN, DMSN, and area over the ANC curve (AOC). The latter will be calculated as the area below the threshold of <0.5 x 10 9 cells/L and above ANC-time response curve in a chemotherapy cycle. Exposure-Blood Pressure Analysis A sequential PK-PD model analysis will be performed. Since the ambulatory blood pressure measurements (ABPM) are subject to circadian variations, ABPM will be modeled to describe the circadian rhythm in blood pressure over the 4.5-hour observation period. The diurnal rhythm in the AMBP data will be described with the cosine function (Hempel 1998) in the exposure ABPM models. Exposure-QTc Analysis Either a direct or sequential PK-PD model analysis will be performed. The choice of the modeling approach will be informed by the matching or non-matching of PK sampling times with ECG observation times. QT measurements corrected with the Fridericia formula will be used in the exposure-QTc analysis.

The study design is group sequential with 1 interim analysis (after 50 patients in each treatment arm have completed at least 1 cycle in each of the treatment arms docetaxel + plinabulin (40 mg) versus docetaxel + pegfilgrastim, with matching placebos) and 1 final analysis at the completion of the study. These results assume that 2 sequential tests are made using the O'Brien-Fleming spending function to determine the test boundaries.

If non-inferiority is determined from the statistical testing, then also the hypothesis of superiority will be tested, and if it is concluded that the plinabulin treatment is superior to the pegfilgrastim treatment, with respect to DSN, then the study will be stopped. Since the design allows for stopping for inferiority, it might be decided at that occasion that the study will be stopped. Since this is a hierarchical testing procedure no penalty with respect to overall significance will be paid. The statistical testing will be performed by an independent statistician and will be reviewed by the Data Safety Monitoring Board at the interim analysis.

In the Phase 2 (40 patients; approximately 10 patients per arm), patients with advanced or metastatic NSCLC will be enrolled.

Approximately 150 patients are planned to be enrolled with 1 of the following diagnoses: advanced or metastatic breast cancer, NSCLC, or HRPC. A sample size of 75 patients in each of the treatment arms docetaxel + plinabulin (40 mg) versus docetaxel + pegfilgrastim, with matching placebos achieve at least a 90% power to reject the null hypothesis of 0.65 day of inferiority in DSN between the treatment means with standard deviations of 0.75, at a significance level (alpha) of 0.05 two-sided two-sample zero-inflated Poisson model.

The software PASS version 15.0.1 has been used for the calculations referencing .

Arm 1: Docetaxel + Pegfilgrastim 6 mg 10

Arm 2: Docetaxel + Plinabulin 20 mg/m 2 10

Arm 3: Docetaxel + Plinabulin 10 mg/m 2 10 Arm 4: Docetaxel + Plinabulin 5 mg/m 2 1 0 PK/PD Analysis (to determine RP3D)

Arm 1: Docetaxel + Pegfilgrastim 6 mg 50 75

Arm 2: Docetaxel + Plinabulin (40 mg) 50 75

Protocol 

The sponsor will obtain approval to conduct the study from the appropriate regulatory agency in accordance with any applicable country-specific regulatory requirements before the study is initiated at a study center in that country.

The Investigator will ensure that this study is conducted in full conformance with the principles of the "Declaration of Helsinki" (as amended in Tokyo, Venice, Hong Kong, South Africa, and Edinburgh) or with the laws and regulations of the country in which the research is conducted, whichever affords the greater protection to the individual. The study must fully adhere to the In other countries where "Guideline for Good Clinical Practice" exists the Sponsor and the Investigators will strictly ensure adherence to the stated provisions.

This Protocol and any accompanying material provided to the patient (such as patient information sheets or descriptions of the study used to obtain informed consent) as well as any advertising or compensation given to the patient, will be submitted by the Investigator to an Institutional Review Board (IRB)/Institutional Ethics Committee (IEC). Approval from the Committee must be obtained before starting the study, and should be documented in a letter to the Investigator specifying the date on which the committee met and granted the approval.

Any modifications made to the Protocol after receipt of the IEC approval must also be submitted by the Investigator to the Committee in accordance with local procedures and regulatory requirements.

When no local review board exists, the Investigator is expected to submit the Protocol to a regional committee. If no regional committee exists, the Sponsor will assist the Investigator in submitting the Protocol to an appropriate Ethics Review Committee.

It is the understanding of the Sponsor that this Protocol (and any modifications) as well as appropriate consent procedures will be reviewed and approved by an IRB. This board must operate in accordance with the current Federal Regulations. A letter or certificate of approval will be sent by the Investigator to the Sponsor before initiation of the study, and also whenever subsequent modifications to the Protocol are made.

It is the responsibility of the Investigator, or a person designated by the Investigator (if acceptable by local regulations), to obtain written informed consent from each patient participating in this study, after adequate explanation of the aims, methods, anticipated benefits, and potential hazards of the study. For patients not qualified or incapable of giving legal consent, written consent must be obtained from the legally acceptable representative. In the case where both the patient and his/her legally acceptable representative are unable to read, or unable to understand due to language barriers, an impartial witness should be present during the entire informed consent discussion. After the patient and representative have orally consented to participation in the study, the witness' signature on the form will attest that the information in the consent form was accurately explained and understood. The Investigator or designee must also explain that the patients are completely free to refuse to enter the study or to withdraw from it at any time, for any reason. The Case Report Forms (CRFs) for this study contain a section for documenting informed patient consent, and this must be completed appropriately. If new safety information results in significant changes in the risk/benefit assessment, the consent form should be reviewed and updated if necessary. All patients (including those already being treated) should be informed of the new information, given a copy of the revised form and give consent to continue in the study. No study related activities should be conducted on a patient until after obtaining informed consent. Protocol BPI-2358-105 IND 

This study will be conducted at sites (to be determined) in North America, South America, Central America, Europe, or Asia Pacific.

The name, telephone and fax numbers of the medical monitor and other contact personnel at the Sponsor are listed in the regulatory binder provided to each site. Protocol BPI-2358-105 IND 

Myelosuppression is the primary toxicity of many chemotherapy regimens which often limits applicability. Both the duration of Grade 4 neutropenia and the depth of the neutrophil nadir have been correlated to severe and life-threatening infections (Pizzo et al, 1993; Crawford et al, 2004) . As a result, the prevention of neutropenia is a major goal for oncology practitioners for both safety and cost-efficiency (Green et al, 2004; Aarts et al, 2013; Dinan et al, 2015) .

Neutropenia is a frequent and potentially life-threatening complication of cytotoxic myelosuppressive chemotherapy. Research has shown that patients who develop neutropenia are more susceptible to infections which often required treatment with antibiotics and in severe cases require hospitalization (Kawatkar et al, 2017) . Moreover, severe neutropenia often necessitates modification of the chemotherapy regimen, thereby compromising the ultimate success of the anticancer treatment plan (Waller et al, 2010) .

Febrile neutropenia (FN) is a potentially life-threatening condition characterized by the development of fever (a single temperature of >38.3°C or a sustained temperature of ≥38°C for more than 1 hour) in addition to chemotherapy-induced Grade 3 or 4 neutropenia (absolute neutrophil count [ANC] < 1.0 × 10 9 /L).

The risk of severe neutropenia including FN is mitigated by reducing chemotherapy dosages or extending the dosing interval of the agents. However, research has shown these measures are directly correlated to lower long-term survival rates because of the relative reduction in the dose intensity of the chemotherapy (Zhang et al, 2018) . Therefore, granulocyte colony-stimulating factor (G-CSF) also known as filgrastim (Neupogen ® ) or pegfilgrastim (Neulasta ® ), is given as standard of care to manage chemotherapy-induced severe neutropenia and to allow chemotherapy to be administered more effectively. Guidelines for the use of G-CSF based on the risk of FN have been established by several groups worldwide such as the National Cancer Institute (NCI) Network, American Society of Clinical Oncology, National Comprehensive Cancer Network (NCCN), European Organization for Research and Treatment of Cancer, and Japanese Society of Clinical Oncology (JSCO). According to these guidelines, prophylactic G-CSF use is recommended for patients at significant risk of FN based on the chemotherapy regimen and patient specific risk factors (Kosaka et al, 2015) . However, the prophylactic use of G-CSF has some significant limitations in terms of safety, cost and convenience of use. Treatment should not be administered within 14 days of chemotherapy initiation. Moreover, G-CSF therapy cannot be initiated until 24 hours after the last dose of chemotherapy for each treatment cycle and is generally administered once per chemotherapy cycle (requires baseline complete blood count [CBC] and platelet count during therapy). The concern with administering G-CSF on the day of chemotherapy is that increasing growth of myeloid cells may increase sensitivity to cytotoxic chemotherapy agents. Since cytotoxic chemotherapy causes the most damage to rapidly growing cells, giving an agent that causes myeloid cells to grow faster while chemotherapy is present may cause more toxicity. Duration of G-CSF therapy is to attenuate chemotherapy-induced neutropenia and is dependent on the myelosuppressive potential of chemotherapy regimen employed. Patients are required to either self-administer the drug or return to the center for treatment and evaluation which is often difficult and costly for the patient. Cost constraints in health care have been reported to be factor in access to filgrastim and pegfilgrastim for some patients. In countries in which patients are required to contribute to treatment costs, high drug prices have resulted in reduced compliance. In Europe, the availability of filgrastim biosimilars (which are more cost effective) have been accompanied by increased use suggesting physicians are more likely to use a cost-effective product .

Warnings and precautions for pegfilgrastim include splenic rupture, acute respiratory distress syndrome, allergic reactions including anaphylaxis, fatal sickle cell crisis, glomerulonephritis, capillary leak syndrome, and leukocytosis. The most common adverse reactions are bone pain and pain in an extremity which occurred in 31% and 9% of patients, respectively. Additional notable adverse events include acute febrile neutrophilic dermatosis, cutaneous vasculitis and injection site reactions (Neulasta® Package Insert).

Plinabulin has the potential to offer a new treatment option that would ameliorate chemotherapy-related severe neutropenia (including FN) as well as have a better safety profile (much less bone pain) and be more convenient for the patient by reducing the number of required patient visits and potentially also reducing the burden to the healthcare system. Most importantly, plinabulin can be given 1 hour after a chemotherapy cycle as opposed to 24 hours after the completion of the cycle (as prescribed by pegfilgrastim).

Two clinical studies with plinabulin have been conducted to date, a Phase 1 monotherapy study (Study NPI-2358-100) and a Phase 1/2 combination study (Study NPI-2358-101) with docetaxel. A total of 141 patients with solid tumors received plinabulin. Laboratory data did not uncover any clinically significant deleterious changes in hematology or chemistry laboratory parameters; however, there was a significantly lower incidence of neutropenia in patients receiving plinabulin plus docetaxel compared with the docetaxel monotherapy arm.

Plinabulin (BPI-2358) is a synthetic, low molecular weight, new chemical entity originally developed (previous name NPI-2358) by Nereus Pharmaceuticals, Inc., and now by BeyondSpring Pharmaceuticals, Inc. It belongs to the diketopiperazine class of compounds.

The chemical name of plinabulin is 2, 5-piperazinedione, 3-

Plinabulin is a yellow to orange solid and the clinical drug product is supplied as a solution in 40% Kolliphor HS 15 (formerly known as polyoxyl 15 hydroxystearate, or Solutol HS-15) ® /60% propylene glycol in an amber vial containing 80 mg in 20 mL (4 mg/mL). Each vial is designated for single use.

Plinabulin is intended for intravenous (IV) infusion and is diluted with dextrose 5% in distilled water (D5W) and given over 30 minutes (± 5 minutes) at an initial dose of 30 mg/m 2 .

Plinabulin which inhibits the polymerization of tubulin monomers, also has multiple mechanisms of action that inhibit tumor growth. Plinabulin targets angiogenesis and the existing tumor vasculature and also directly induces cancer cell apoptosis via the JNK pathway (Nicholson et al, 2006; Singh et al, 2011; Kennedy et al, 2003) . Plinabulin stimulates the tumor-related immune system by means of dendritic cell maturation and enhances the antitumor activity of checkpoint inhibitors in an immune-competent mouse model . Based on the nonclinical and clinical studies conducted to date, plinabulin has several advantages compared to agents that Protocol BPI- target existing tumor vasculature, including inhibition of angiogenesis, induction of tumor cell apoptosis, sustained suppression of tumor growth after treatment, inhibition of tubulin dimerization and thus new microtubule formation without alteration of microtubule dynamic instability (reducing the risk of peripheral neuropathies associated with taxanes) (Gornstein, 2014) . The safety profile of plinabulin appears to be superior to that of other agents with immune-oncology effects, such as checkpoint inhibitors (e.g., nivolumab), providing a major advantage in cancer therapy. Thus, plinabulin may prove to be efficacious in the management of cancers such as advanced non-small cell lung cancer (NSCLC). A Phase 3 global trial with plinabulin in combination with docetaxel is underway in NSCLC patients. (Study NPI-2358-103).

The mechanism by which plinabulin exerts its beneficial effect in neutropenia is still under investigation. Preclinical evidence shows that plinabulin induces maturation of dendritic cells, resulting in the release of the cytokines interleukin (IL)-1β, IL-6 and IL-12 from monocytes/dendritic cells . In particular IL-6 is mediated in the prevention of neutrophil apoptosis (Asensi et al, 2004) and IL-1β with increased neutrophil count (Dinarello, 2011) .

Phase 1 (Study NPI-2358-100) and Phase 1/2 (Study NPI-2358-101) studies with plinabulin have been completed. Study NPI-2358-100 was a Phase 1, open-label, dose-escalation study to determine the maximum-tolerated dose and/or recommended Phase 2 dose (RP2D) of plinabulin monotherapy in patients with advanced solid tumor malignancies or lymphoma, whose disease had progressed after treatment with standard approved treatments. Plinabulin was administered once per week as an IV infusion for 3 successive weeks in repeating 4-week cycles. Doses ranged from 2 to 30 mg/m 2 . A total of 38 patients received plinabulin monotherapy.

Study NPI-2358-101 was a Phase 1/2, open-label study to evaluate plinabulin in combination with docetaxel in patients with advanced NSCLC that had progressed after treatment with at least 1 chemotherapy regimen. In the Phase 1 part of the study, patients received escalating doses of plinabulin (13.5 mg/m 2 to 30 mg/m 2 ) in combination with a standard dose of docetaxel (75 mg/m 2 ). No drug-drug interaction was detected between plinabulin and docetaxel. The RP2D of plinabulin administered with docetaxel was determined to be 30 mg/m 2 .

In the Phase 2 part, patients were randomized to receive either docetaxel in combination with plinabulin or docetaxel alone (active control group). A docetaxel dose of 75 mg/m 2 was administered to all patients. The study was stratified into 2 segments. A stratum that compared plinabulin 30 mg/m 2 plus docetaxel (study arm DN) to docetaxel alone (study arm D) followed by a second stratum of plinabulin 20 mg/m 2 plus docetaxel (DN arm) to docetaxel alone (D arm). Study drug was administered by IV infusion on Day 1 (docetaxel plus plinabulin) and Day 8 (plinabulin) of each 3-week cycle. A total of 103 patients receive plinabulin plus docetaxel and 73 patients received docetaxel alone.

In the Phase 2 clinical trial, patients were randomized to receive docetaxel 75 mg/m 2 alone (n=73) or docetaxel 75 mg/m 2 followed by plinabulin (Study NPI-2358-101) at 30 mg/m 2 (n=50) or at 20 mg/m 2 (n=40), repeated every 3 weeks (clinicaltrials.gov NCT00630110). Plinabulin was given by a 30-minute intravenous (IV) infusion, starting 1 hour after administration of docetaxel. The primary efficacy endpoint was median overall survival. Secondary endpoints Protocol BPI- included safety assessments, such as complete blood count measurements, on Days 1, 8, and 15 of each cycle.

In Figure 1 , plinabulin and docetaxel combination has a much lower incidence of Grade 4 neutropenia versus docetaxel alone (4% versus 33% in the first cycle, based on hematological laboratory values) for both the 20 mg/m 2 and 30 mg/m 2 plinabulin cohorts combined, and a benefit in neutropenia prevention was maintained over Cycles 2, 3, and 4.

In Figure 2 , the percentage of patients with Grade 4 neutropenia (ANC < 0.5 x 10 9 /L) on Day 8 of Cycle 1, in the patients receiving docetaxel monotherapy without plinabulin ("Docetaxel"), docetaxel +20 mg/m 2 plinabulin, and docetaxel+ 30 mg/m 2 plinabulin are shown. The reduction in Grade 4 neutropenia with either plinabulin dose was statistically highly significant, and this reduction was comparable for the 20 and 30 mg/m 2 plinabulin dose. The benefit in Grade 4 neutropenia reduction on Day 8 in Cycle 1 was observed prior to patients receiving their plinabulin dose on Day 8, which provides evidence that a single plinabulin dose per cycle, given on Day 1 after docetaxel administration, is sufficient to achieve the neutropenia benefit. Further evidence that a single 20 mg/m 2 dose of plinabulin after each dose of docetaxel is sufficient to prevent myelosuppressive regimens-induced neutropenia is demonstrated in Figure 3 in which the nadir following a single dose (measured on Day 8 prior to the second dose of plinabulin) is similar to that measured in all blood samples in a cycle. There were 10% of patients (5 of 50 patients) in the plinabulin 30 mg/m 2 plus docetaxel treatment arm who required docetaxel dose reductions and 18.2% of patients (10 of 55 patients) in the docetaxel arm who required dose reductions. In the 20 mg/m 2 cohort, similar findings were observed, in which a lower proportion of patients required docetaxel dose reductions when treated with the combination (2.5%) than the companion docetaxel arms (22.2%) (refer to Table 1 ). Abbreviations: D = docetaxel; DN = docetaxel plus plinabulin; G-CSF = granulocyte colony stimulating factor. Note: G-CSF included the following concomitant medications: pegfilgrastim, filgrastim, Neupogen, and Neulasta. P = 0.0013 (G-CSF use % in combined DN arm versus combined D arm).

The Phase 2 data suggest that a single dose of 20 mg/m 2 plinabulin per cycle is highly effective in preventing docetaxel-induced neutropenia, and that the 20 mg/m 2 plinabulin dose appears equally effective as the 30 mg/m 2 group for the prevention of docetaxel-induced neutropenia. Consistent with the benefit of neutropenia prevention, in the plinabulin groups fewer patients required G-CSF treatment, and fewer patients required docetaxel dose reductions.

The safety profile of plinabulin was consistent with what has been reported in earlier Phase 1 studies. The combination appeared to have a similar adverse event (AE) profile as single-agent docetaxel. Increases in both systolic and diastolic blood pressure were seen in 20% to 30 % of patients. The incidence of hypertension in the 20 mg/m 2 (9 patients, 22.5%; 5% with Grade 3) and 30 mg/m 2 (16 patients, 32.0%; 20% with grade 3) cohorts was higher compared with the docetaxel group (1 patient 5.6% and 2 patients, 3.6%), respectively. These events typically resolved within 2 hours after plinabulin infusion. No events of Grade 4 hypertension had occurred. Grade 3 hypertension had occurred in 2 cases with the 20 mg/m 2 dose and 7 cases with the 30 mg/m 2 dose. In both cases, the patients had a previous medical history of hypertension, which makes a causal relationship with plinabulin unlikely. In the 30 mg/m 2 , there were 7 cases with Grade 3 hypertension, of which 4 cases had a prior medical history of hypertension; a causal relationship with plinabulin could not be determined in these 4 cases. Hypertension typically resolved within 2 hours after the plinabulin infusion. Ileus and intestinal obstruction were also reported as AEs. It was not clear if they were a direct effect of plinabulin or caused by some concomitant medications such as opioids.

Overall, plinabulin + docetaxel appeared to have a more favorable safety profile at the 20 mg/m 2 plinabulin dose versus the 30 mg/m 2 plinabulin dose, as demonstrated by a smaller proportion of patients who discontinued treatment due to a treatment emergent adverse event ([TEAE] 3 patients, 7.5% versus 11 patients, 22.0%) and a smaller proportion experienced 1 or more serious adverse events ([SAEs] 14 patients, 35.0% versus 31 patients, 62.0%).

The proportion of patients in the plinabulin plus docetaxel 20 mg/m 2 arm who discontinued treatment due to a TEAE and experienced 1 or more SAEs was similar to that reported for the combined docetaxel alone arms (4 patients, 5.5% and 23 patients, 31.5%, respectively), thus the 20 mg/m 2 plinabulin dose appears to represent a safe plinabulin dose.

Laboratory data did not uncover any clinically significant deleterious changes in hematology or chemistry laboratory parameters; however, there was a lower incidence of neutropenia in patients Protocol BPI- in the DN 30 mg/m 2 and 20 mg/m 2 arm compared with its companion docetaxel arm as discussed above.

Adverse events of concern were atrial fibrillation, ileus/intestinal obstruction, and reversible posterior leukoencephalopathy syndrome (see current Investigator Brochure).

In specific therapeutic settings as described by the NCCN guidelines, docetaxel is an accepted treatment regimen for breast cancer, NSCLC, and hormone refractory prostate cancer (HRPC) (https://www.nccn.org/professionals/physician_gls/f_guidelines.asp#site).

The pharmacokinetics (PK) of plinabulin was evaluated in a Phase1 monotherapy clinical study in advanced solid tumor malignancies or lymphoma. Compartment-independent PK analysis revealed a linear increase in systemic exposure (are under the curve [AUC] tot ) with increasing dose of plinabulin. Inter-patient variability in AUC tot was large, suggesting individual differences in metabolism of the parent drug. Maximal plasma concentration was observed at the end of infusion for most patients. Distributive volumes (Vz and Vss) were similar for all patients, and were indicative of the drug reaching deep compartments or being stored in peripheral compartments.

There was no evidence of any interaction between plinabulin and docetaxel with regard to plasma levels in the Phase 1b portion of Study NPI-2358-101.

Pharmacokinetic analysis of plinabulin indicated Cmax and AUC were dose proportional over the range of 2 to 30 mg/m 2 without evidence of drug accumulation. The mean t½ is 6.35 hours, the mean CL is 31 L/hour, and the mean Vz is 208 L for plinabulin as a single agent.

Plinabulin is a small molecule with tumor-inhibiting and immune-enhancing effects. Plinabulin induces dendritic cell maturation and cytokines interleukin-1β (IL-1β), IL-6, and IL-12 production, all of which are important in neutrophil survival. In preclinical studies, plinabulin prevented docetaxel-or cyclophosphamide-induced neutropenia via a mechanism of action different from that of G-CSF analogues. In Phase 1 and 2 solid tumor trials of plinabulin, which included >140 patients, routine safety laboratory assessments revealed an unexpected protective effect against neutropenia.

Compared to docetaxel treatment alone, the addition of plinabulin to docetaxel significantly (p < 0.0003) reduced the proportion of patients with Grade 4 neutropenia from 33.3% to 4.6% in Cycle 1. Figure 2 shows the proportions of patients with Grade 4 neutropenia (ANC < 0.5 x 10 9 /L) on Day 8, the approximate day after docetaxel administration corresponding to the largest reduction in ANC . Plinabulin also reduced the clinical sequelae associated with docetaxel-induced neutropenia (sepsis, infections, hospitalizations, need for docetaxel dose reductions, and G-CSF use). Bone pain was reported in 4% of patients receiving plinabulin. Plinabulin had a favorable safety profile; the most prominent finding was Grade 3 (Blayney D, et al, 2016) .

Plinabulin is a novel small molecule that is being developed for the mitigation of chemotherapy-induced neutropenia. Administered by IV infusion on the same day of (approximately 1 hour after) chemotherapy, plinabulin will be given in a single dose to be determined per cycle. Plinabulin has the potential to be an effective, safe (with much less bone pain), cost-effective, and convenient alternative to G-CSF for the prevention of chemotherapy-induced neutropenia (Kuderer et al, 2006) .

Generalized bone pain is a risk associated with the use of G-CSF and is correlated with the mechanism of action (Lambertini et al, 2014). A meta-analysis of 5 controlled trials indicates that the overall incidence of bone pain with G-CSF can range from 20% to 50% (Lambertini et al, 2014; Kubista et al, 2003) . The overall incidence of mild-to-moderate bone pain attributed to G-CSF is approximately 24%, and overall incidence of severe bone pain can range from 1% to 5% of patients (Lambertini et al, 2014; Kubista et al, 2003) . The G-CSF-induced bone pain can lead to the discontinuation of this growth factor, which in turn can reduce the chemotherapy dose intensity (Kirshner et al, 2012) .

Generally, bone pain with pegfilgrastim reached its peak at Day 3 after dosing with pegfilgrastim on a scale of 1 to 10 (Kirshner et al , 2012). An AUV analysis of bone pain over at least 5 days after dosing with pegfilgrastim was previously employed to evaluate bone pain with pegfilgrastim (Kirshner et al, 2012).

G-CSFs mobilize hematopoietic progenitor cells from the bone marrow into the systemic circulation (Lambertini et al, 2014) . This progenitor-cell-mobilizing effect is an important reason why G-CSFs should wait 24 hours after the chemotherapy, and if administered sooner (ie, on the same day of chemotherapy), G-CSFs are known to lead to a paradoxical worsening of the neutropenia (Burris et al, 2010) . The possible biologic explanation for the increase in FN observed with same-day pegfilgrastim dosing, relates to G-CSFs release of bone marrow-derived hematopoietic progenitor cells, and that the stimulation of these myeloid progenitor cells by G-CSFs makes these progenitor cells more sensitive to the effects of cytotoxic chemotherapy (Burris et al, 2010) . Therefore, pegfilgrastim is not recommended to be administered on the same day but on the next day of chemotherapy (Burris et al, 2010) .

A therapeutic intervention that is effective though a mechanism that is independent from mobilization of progenitor cells from the bone marrow has the potential to have superior efficacy over G-CSFs, as this therapeutic intervention can be administered on the same day, shortly after the chemotherapy. Since no bone marrow effects are involved, this agent should therefore also not cause bone pain.

The absence of an effect on neutrophil counts with plinabulin monotherapy (in the absence of chemotherapy), and the relative absence of bone pain with plinabulin, is indicative of differential mechanisms of action between plinabulin and G-CSF in the prevention of chemotherapy induced neutropenia. In addition, in contrast to G-CSFs, plinabulin has anti-cancer effect, which potentially represents an important advantage over G-CSFs, which do not have anti-cancer effects, or may even promote cancer growth, being a growth factor. Based on the bone pain data from the completed Phase 2 study NPI-2358-101 in 90 patients on plinabulin (at the dose of 30 or 20 mg/m 2 ), a significant bone pain benefit is expected with plinabulin compared to G-CSF, while having an at least comparable benefit on chemotherapy-induced neutropenia.

The patient reported outcome bone pain is included as a secondary endpoint in the Phase 3 portion of the present study.

A model-based approach was used to characterize the effect of plinabulin on docetaxel-induced neutropenia. The population pharmacokinetic/pharmacodynamic model developed was used to perform simulations. Daily sampling for neutrophils was implemented to enable accurate delineation of the efficacy metrics, duration of severe neutropenia (DSN; number of days with an ANC of < 0.5 x 10 9 /L), duration of moderate and severe neutropenia (DMSN; number of days with an ANC of < 1.0 x 10 9 /L ), and area over the curve (AOC), which measure the duration and severity of neutropenia. The results were summarized by dose groups 5, 10, and 20 mg/m 2 studied in the Phase 2 portion of the present study.

Based on AOC, DMSN, and DSN, 20 mg/m 2 of plinabulin performed better than the 5 and 10 mg/m 2 doses, and was therefore the recommended dose for the Phase 3 study. In recommending the dose, the minimal dose-independent effect of plinabulin on ambulatory blood pressure monitoring, and negligible effect on QT interval are taken into account.

Simulations were also performed to determine the fixed dose of plinabulin that is similar to the body surface area (BSA)-based dose of 20 mg/m 2 . Simulations were performed with plinabulin fixed doses of 20, 40, 60, and 80 mg and compared with the selected BSA-based dose of 20 mg/m 2 as the reference, assuming a similar population as that studied in the Phase 2 portion of the study and used in the BSA-based dosing simulations. The 40 mg fixed dose yielded similar peak concentration and AUC as the body surface area-based dose of 20 mg/m 2 , indicating that subjects are not predicted to be at any risk of over or under exposure to plinabulin at this dose.

The efficacy performance of the 40 mg fixed dose is predicted to be similar to that of the BSA-based dose of 20 mg/m 2 . The 60 and 80 mg fixed doses yielded results similar to the 40 mg fixed dose. Thus, a fixed plinabulin dose of 40 mg is predicted to yield a similar efficacy as a BSA-based dose of 20 mg/m 2 . 

Plinabulin pharmacokinetic (PK) and pharmacodynamic (PD) assessments will be made to enable a PK/PD analysis.

To establish the Recommended Phase 3 Dose (RP3D) based on PK/PD analysis.

To assess DSN in treatment Cycle 1 in patients treated with docetaxel (75 mg/m 2 ) + plinabulin (5, 10 or 20 mg/m 2 ) or with docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg). Severe (Grade 4) neutropenia is an absolute neutrophil count (ANC) <0.5 × 10 9 /L ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose) and during Cycle 1 on Days 1, 2, 6, 7, 8, 9, 10 , and 15 (pre-dose on dosing days; times equivalent to pre dose on other days).

To assess blood pressure semi-continuously with 15-minute intervals, starting 15 minutes pre-plinabulin dose and lasting 4.5 hours after start of infusion with plinabulin (Arms 2 to 4) or for 4.75 hours starting 15 minutes after the end of docetaxel infusion (Arm 1).

To characterize the pharmacokinetic profile of plinabulin and docetaxel To characterize the exposure-response relationships between measures of plinabulin exposure and the pharmacodynamic endpoint DSN To characterize the exposure-safety relationships between measures of plinabulin exposure and safety events of interest

To assess CD34+ at screening, and on Days 2, 6, and 8 in Cycle 1 and Day 1 in Cycle 2 Health-related Quality of Life (QoL) questionnaire evaluated with EORTC QLQ-C30 and EQ-5D-5L To collect data on disease progression For selected countries only: to investigate the following cytokine panel: IL-1beta, IL-6, IL-12p70, IL-12p40, IL-17A, IL-23, G-CSF, GM-CSF, IFN-alpha, IFN-gamma, TNFalpha, IL-2, FLT-3 ligand, and IL-8.

Incidence, occurrence, and severity of AEs/SAEs Incidences of bone pain Safety and tolerability (physical examination and safety laboratory assessments) Protocol BPI- 

To assess DSN in treatment Cycle 1 in patients with advanced or metastatic breast cancer, who have failed < 5 prior lines of chemotherapy; locally advanced or metastatic non small cell lung cancer (NSCLC) after platinum therapy failure; or hormone refractory (androgen independent) metastatic prostate cancer (HRPC) treated with docetaxel (75 mg/m 2 ) + plinabulin (40 mg) (Arm 2) versus docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) (Arm 1). ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose) and during Cycle 1 on Days 1, 2, 6, 7, 8, 9, 10, and 15 . Blood draws for ANC will be taken approximately the same time as the time of the pre-dose sample on Day 1, and will be taken by preference in the morning. DSN should be calculated as the number of consecutive days from the first day when a patient's ANC is below 0.5 x l09/L until the patient reaches an ANC> 0.5 x 109/L, in Cycle 1. For patients who do not experience any severe neutropenia in Cycle 1, the DSN is set to 0. For patient's experiencing several episodes, the number of days of DSN will be summed up.

Platelet count in Cycle 1: maximum decrease from baseline (prior to Cycle 1 docetaxel dose)

Proportion of patients with neutrophil-to-lymphocyte ratio (NLR) > 5 after Day 7 through Day 15 in Cycle 1 AUC using the trapezoidal quadrature method for bone pain, from Day 1 through Day 8 (pegfilgrastim will be administered on Day 2) in Cycle 1, based on the pain score from the patient bone pain scale To compare the time (in days) to first use of bone pain medication between the treatment groups

To assess DSN in treatment Cycle 1 in patients with locally advanced or metastatic NSCLC after platinum therapy failure

To assess DSN in treatment Cycle 1 in patients with HRPC

To assess DSN in treatment Cycle 1 in patients with advanced or metastatic breast cancer who have failed < 5 prior lines of chemotherapy

To assess DSN in treatment Cycle 1 in patients with locally advanced or metastatic NSCLC after platinum therapy failure or HRPC Platelet count at least 30% change from baseline at any time during Cycle 1.

For selected countries only: to investigate the following cytokine panel: IL-1beta, IL-6, IL-12p70, IL-12p40, IL-17A, IL-23, G-CSF, GM-CSF, IFN-alpha, IFNgamma, TNF-alpha, IL-2, FLT-3 ligand, and IL-8.

To evaluate the proportion of patients in Cycles 1 to 4 with:

o Febrile neutropenia (FN) (ANC <1.0 × 10 9 /L AND a single temperature of >38.3°C or a sustained temperature of ≥38°C for more than 1 hour).

o Grade 4 neutropenia (ANC < 0.5 × 10 9 /L).

To evaluate the following healthcare utilization objectives: 

This is a multicenter, randomized study with an open label phase 2 portion and a double blind phase 3 portion. Approximately 190 patients will be enrolled in this study.

The decision to complete the Phase 2 portion of the study as open label was made to reduce the unnecessary complexities of study conduct (such as placebo infusions and injections).

All patients will receive docetaxel at a dose of 75 mg/m 2 .

In Phase 2, patients only with advanced or metastatic NSCLC after failing platinum-based therapy will be enrolled.

In Phase 3, patients with one of the following will be enrolled: advanced or metastatic breast cancer, who have failed < 5 prior lines of chemotherapy; locally advanced or metastatic NSCLC after platinum therapy failure; or hormone refractory (androgen independent) metastatic prostate cancer.

The eligibility of all patients will be determined during a 28-day screening period.

Approximately 40 patients with advanced and metastatic NSCLC will be enrolled. Patients are randomly assigned, with approximately 10 patients enrolled in each arm, with the arm designation and planned intervention as follows: The study will be temporarily closed to enrollment when 40 patients have been enrolled and completed at least 1 treatment cycle in each arm in phase 2. The Sponsor will notify the study sites when this occurs.

Once the study is temporarily closed to enrollment in phase 2, a PK/PD analysis will be performed to determine the RP3D. The PK/PD analysis will be done by an independent party at the time 40 patients in Phase 2 have completed at least Cycle 1.

Phase 3 will not begin until RP3D has been determined based on the phase 2 PK/PD analysis as mentioned above; the RP3D will be the only plinabulin dose administered in Phase 3.

A fixed dose of 40 mg has been selected as the RP3D following the Phase 2 PK/PD analysis (details of dose selection are provided in Section 6.2.3).

Approximately 150 patients are planned to be enrolled in the Phase 3 with one of the following diagnosis: advanced or metastatic breast cancer, who have failed < 5 prior lines of chemotherapy; locally advanced or metastatic NSCLC after platinum therapy failure; or hormone refractory (androgen independent) metastatic prostate cancer. Each eligible patient will be Protocol BPI- stratified according to his or her tumor type (breast cancer, NSCLC, or HRPC) and region (Asia, non-Asia). Patients will be randomly assigned with equal probability (1:1 ratio) or 75:75, with the arm designation and planned intervention as follows:

Arm 1: Docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) + placebo matching plinabulin Arm 2: Docetaxel (75 mg/m 2 ) + plinabulin (40 mg) + placebo matching pegfilgrastim

In order to facilitate balanced treatment arms with respect to cancer type, once either arm reaches at least 1/3 (of total) of patients with that cancer type, it will be closed to that cancer type and enrollment will continue for patients with the other cancer types, up to the planned maximum number of patients.

Data from all patients receiving the RP3D plinabulin dose in Phase 2 and Phase 3 will not be pooled for assessing the primary and secondary study endpoints, but analyzed separately.

In Phase 2 and Phase 3, Cycles 1 to 4 will consist of docetaxel 75 mg/m 2 administered by IV infusion on Day 1 over 60 minutes (±5 minutes) every 21 days.

In the phase 2 portion, on Day 1 of each cycle, 1.5 hours (± 10 minutes) after the start time of docetaxel infusion (i.e., approximately 30 minutes after the end of docetaxel infusion), patients assigned to a plinabulin arm (arms 2-4) will get a single intravenous infusion of plinabulin at their assigned dose over 30 minutes (± 5 minutes). Thus the wait time between end of docetaxel infusion and start of the plinabulin infusion is approximately 30 minutes. On Day 2 of each cycle, ≥24 hours after completing chemotherapy, patients assigned to pegfilgrastim (arm 1) will receive a single dose of pegfilgrastim (6 mg) (subcutaneous injection).

In the phase 3 portion, on Day 1 of each cycle, 1.5 hours (± 10 minutes) after the start time of docetaxel infusion (i.e., approximately 30 minutes after the end of docetaxel infusion), patients will get a single dose of plinabulin or placebo intravenously over 30 minutes (± 5 minutes). On Day 2 of each cycle, ≥24 hours after completing chemotherapy, patients will receive a single dose of pegfilgrastim (6 mg) or placebo (subcutaneous injection).

If a chemotherapy cycle is delayed by more than 3 weeks, the patient will be withdrawn from the study. If a critical AE occurs during the cycle, the dosage of docetaxel may be reduced 20% in the next cycle. Only one docetaxel dose reduction is allowed (refer to Taxotere® (Prescribing Information)). No dose reductions are allowed with plinabulin or pegfilgrastim.

All patients, including patients who withdraw from the study early, will complete a safety follow-up visit 30 days (± 2 days) after the last dose of study drug. Patients who withdraw for progressive disease and who will continue to another chemotherapy regimen will complete the EOT visit at Cycle X Day 21. They then continue to another chemotherapy regimen and do not need to have the safety follow-up visit (where X is the last cycle prior to progression, and is 4 or less).

If, in the opinion of the investigator, the patient will benefit from more than 4 cycles of docetaxel and open label pegfilgrastim, then the fifth cycle will not start until completion of the EOT visit (in this instance, the EOT visit will be Cycle 4 Day 21). Follow-up visits will be required to monitor for ongoing treatment-related AEs. All patients experiencing drug-related toxicities of Grade 2 at the EOT visit should be followed-up at least monthly until the AE(s) resolves to Grade 1, the event is considered to be chronic, or the patient receives other anti-cancer therapy. The method of follow-up assessment will be at the Investigator's discretion (for example, patient site visit or telephone call). All deaths which occur within 30 days of study drug administration regardless of relationship to the study drug must be reported to the Sponsor immediately and within 24 hours of becoming aware of the event.

Laboratory test results (hematology and serum chemistry) will be collected via a central laboratory. Safety laboratory tests are required prior to treatment on Day 1 of each cycle and can be collected by a local laboratory; however, all other scheduled blood samples as per the schedule assessments and procedure table must also be obtained for central laboratory assessment. Urinalysis will be performed at screening only.

For both Phase 2 and Phase 3, once a patient has completed all cycles of study treatment they will enter a post-treatment (non-interventional) assessment phase. On a monthly basis, the patients will be contacted to assess whether they have started to receive another anti-cancer therapy, which could be cytotoxic chemotherapy, a programmed death-ligand 1 (PD-L1) inhibitor, radiation therapy, or other therapy. This assessment will continue until the patient starts another anti-cancer therapy, the patient chooses to have no further therapy, the patient dies, or the patient becomes lost to follow-up. This information will be used to analyze the time to next treatment.

This will be performed in addition to any follow up for monitoring of ongoing treatment-related AEs, as described above.

Patients without progressive disease at the EOT visit will undergo a follow-up visit every 2 months until the occurrence of either disease progression or death. These visits may be conducted per telephone or other means.

The schedule of assessments is presented in Table 6 and Table 7 . Protocol BPI-2358-105 IND Hemoglobin 9 g/dL independent of transfusion or growth factor support ANC 1.5 x 10 9 /L independent of growth factor support Serum total bilirubin 1.5 times the upper limit normal (ULN), unless the patient has a diagnosis of Gilbert's disease, in which case direct bilirubin less than or equal to 1. 9. Female subjects of childbearing potential have a negative pregnancy test at screening. Females of childbearing potential are defined as sexually mature women without prior hysterectomy or who have had any evidence of menses in the past 12 months. However, women who have been amenorrhoeic for 12 or more months are still considered to be of childbearing potential if the amenorrhea is possibly due to prior chemotherapy, anti-estrogens, or ovarian suppression.

Women of childbearing potential (i.e., menstruating women) must have a negative urine pregnancy test (positive urine tests are to be confirmed by serum test) documented within the 24-hour period prior to the first dose of study drug. Sexually active women of childbearing potential enrolled in the study must agree to use two forms of accepted methods of contraception during the course of the study and for 3 months after their last dose of study drug. Effective birth control includes (a) intrauterine device (IUD) plus one barrier method; (b) on stable doses of hormonal contraception for at least 3 months (e.g., oral, injectable, implant, transdermal) plus one barrier method; (c) 2 barrier methods. Effective barrier methods are male or female condoms, diaphragms, and spermicides (creams or gels that contain a chemical to kill sperm); or (d) a vasectomized partner.

For male patients who are sexually active and who are partners of premenopausal women: agreement to use two forms of contraception during the treatment period and for at least 3 months after the last dose of study drug.

Patients will not be entered in the study if they meet any of the following criteria: 10. Prior radiation therapy within the 4 weeks before the first dose of study drug.

11. Prior use of pegfilgrastim or filgrastim within 4 weeks before the first dose of study drug.

12. Presence of any serious or uncontrolled illness including, but not limited to: uncontrolled diabetes, ongoing or active infection, symptomatic congestive heart failure, unstable angina pectoris, uncontrolled cardiac arrhythmia, uncontrolled arterial thrombosis, symptomatic pulmonary embolism, or psychiatric illness that would limit compliance with study requirements, or any other conditions that would preclude the patient from study treatment as per the discretion of the Investigator. 

Patients will receive treatment with study drug up to 4 cycles in this study; thereafter patients may continue receiving docetaxel and pegfilgrastim at the Investigator's discretion. Patients will complete a safety follow-up visit 30 days (± 2 days) after the last dose of study drug. If, in the opinion of the investigator, the patient will benefit from more than 4 cycles of docetaxel and open label pegfilgrastim, then the fifth cycle will not start until completion of the EOT visit (in this instance, the EOT visit will be Cycle 4 Day 21). The safety follow-up visit does not need to be performed in this case. The same situation applies for patients with progressive disease who need another chemotherapy treatment, they also will perform the EOT visit at Cycle X Day 21 and continue to the next chemotherapy regimen (where X is the last cycle prior to progression, and is 4 or less). The safety follow-up visit is not applicable in these cases.

If a patient develops a positive viral load during the study, chemotherapy should be held for a maximum of 21 days. If the viral load returns to zero treatment can be restarted, otherwise the patient should be withdrawn from the study.

Treatment up to 4 cycles of study drug in this study will continue until any 1 of the following occurs:

Dose limiting toxicity or critical adverse events as described in the docetaxel package insert (refer to Taxotere® (Prescribing Information) and Table 4 ). Need for a protocol-prohibited dose reduction or study drug delay greater than 21 days Initiation of a protocol-prohibited concomitant medication or non-protocol chemo/biological therapy for treatment of his or her disease Development of an AE/SAE, illness or condition that may interfere with the patient's participation or require treatment discontinuation Investigator opinion Sponsor decision Voluntary withdrawal of consent Protocol BPI- 

The study drugs under evaluation are plinabulin and pegfilgrastim. Placebo-matching plinabulin and pegfilgrastim will also be used as part of the study design.

Patients will be included in either Phase 2 or Phase 3 parts of the study (Table 2 and Table 3 In the phase 2 portion, on Day 1 of each cycle, 1.5 hours (± 10 minutes) after the start time of docetaxel infusion (i.e., approximately 30 minutes after the end of docetaxel infusion), patients assigned to a plinabulin arm (arms 2-4) will get a single intravenous infusion of plinabulin at their assigned dose over 30 minutes (± 5 minutes). Thus the wait time between end of docetaxel infusion and start of the plinabulin infusion is approximately 30 minutes. On Day 2 of each cycle, ≥24 hours after completing chemotherapy, patients assigned to pegfilgrastim (arm 1) will receive a single dose of pegfilgrastim (6 mg) (subcutaneous injection).

Phase 3 (Double Blind) (a planned 75 patients per arm with advanced or metastatic breast cancer, who have failed < 5 prior lines of chemotherapy; locally advanced or metastatic NSCLC after platinum therapy failure; or hormone refractory [androgen independent] metastatic prostate cancer):

Arm 1: Docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) + placebo matching plinabulin Arm 2: Docetaxel (75 mg/m 2 ) + plinabulin (40 mg) + placebo matching pegfilgrastim

All patients with HRPC also are given prednisone 5 mg orally twice daily continuously in addition to docetaxel (see Taxotere® (Prescribing Information) as well as other assigned study drugs.

In the phase 3 portion, on Day 1 of each cycle, 1.5 hours (± 10 minutes) after the start time of docetaxel infusion (i.e., approximately 30 minutes after the end of docetaxel infusion), patients will get a single dose of plinabulin or placebo intravenously over 30 minutes (± 5 minutes). On Day 2 of each cycle, ≥24 hours after completing chemotherapy, patients will receive a single dose of pegfilgrastim (6 mg) or placebo (subcutaneous injection).

Patients who experience an FN event in Cycle 1 should be discussed with the medical monitor. The blinding will be broken (if Phase 3), and patients assigned to plinabulin will receive pegfilgrastim in subsequent cycles. If a patient was originally assigned to the pegfilgrastim arm, Protocol BPI- patients must be treated at a lower dose of docetaxel, or taken off study at the discretion of the investigator.

If the patient develops an FN event on subsequent cycles, the patient should be discussed with the medical monitor and either treated with a lower dose of docetaxel, or taken off study at the discretion of the investigator. Febrile neutropenia should be treated with antibiotics per institutional standard of care (refer to Section 10.6.1). If a patient is hospitalized, as a result of a Serious Adverse Event, the procedure for reporting Serious Adverse Events (Section 13.2) should also be followed.

Treatments Administered: Dose modifications for docetaxel for specific Grade 3/4 AEs are provided in Section 10.5, Table 4 .

During the phase 2 study if at any given cohort, 3 patients or more have Grade 4 or 5 toxicity not related to underlying disease (with the exception of neutropenia), accrual to that cohort will be halted and the study will be continued at the lower dose cohorts in phase 2 (for example if 3 patients at the 20 mg/m 2 cohort develop Gr. 4 toxicity the accrual to that cohort will be stopped and the study will continue as planned with the accrual of the two remaining open cohorts). Interactive web response system (IWRS) will be utilized to assign patients to a lower dose cohort in phase 2. Study sites will be instructed to call IWRS when a Grade 4 or 5 toxicity event occurs.

The Sponsor will supply plinabulin during the study treatment. If docetaxel/pegfilgrastim is not available at study sites, this will be supplied by the Sponsor.

Refer to Taxotere® (Prescribing Information) and Neulasta® Package Insert for details on docetaxel and pegfilgrastim.

Details are provided for plinabulin below.

Instructions for pharmacy drug preparation can be found in the study Pharmacy Manual.

Plinabulin ( 

Details are provided for plinabulin only. Refer to the Taxotere® (Prescribing Information) and Neulasta® Package Insert for details on docetaxel and pegfilgrastim.

Instructions for pharmacy drug preparation can be found in the study Pharmacy Manual.

Plinabulin is supplied as a solution in 40% Kolliphor HS 15 (formerly known as polyoxyl 15 hydroxystearate or Solutol HS-15 ® )/60% propylene glycol in amber vials containing 80 mg in 20 mL (4 mg/mL). Each vial is designated for single use. The labeled storage condition for the drug product is stored between 15° and 25°C (59° and 77°F). Vials contain 80 mg drug in 20 mL (4 mg/mL) and are labeled with other information as per local regulatory requirements.

The contents of the label will be in accordance with all applicable regulatory requirements.

The study drug will be dispatched to a study center only after receipt of the required documents in accordance with applicable regulatory requirements and the sponsor's procedures.

Vials of plinabulin should be stored at room temperature (between 15°C and 25°C [59° F and 77°F]) and must be protected from light. Protection from light must be maintained throughout the drug administration process. The stability of the drug meets all specifications under controlled room temperature storage conditions (25°C, 60% relative humidity). Using a previous manufacturing process, drug product stability was demonstrated for up to, and including, 48 months. Maintenance of plinabulin purity, potency, and impurity profile has been demonstrated at 6 months at elevated temperature and humidity (40°C/75% relative humidity) and after freeze-thaw cycles.

Instructions for the preparation of prescribed doses are provided in the Pharmacy Manual.

Plinabulin will be provided by an agent of the Sponsor and shipped directly to the Investigator or the designated pharmacist. If docetaxel/pegfilgrastim is not available at study sites, this will be supplied by the Sponsor.

The Investigator or designated pharmacist will acknowledge receipt of the shipment and note content and condition of the shipment on the clinical material shipping form. The Sponsor or its representatives will supply the appropriate forms.

The pharmacist or person responsible for dispensing the study drug at the site will maintain an accurate and current record of all drug supplies received from the repository and dispensed to study patients. The dispensing record should contain the protocol number and information regarding the amount/vial(s) dispensed; date dispensed, lot #, patient identifier number, patient initials, and the initials of the person dispensing the medication.

All partially used or empty vial drug counts should be verified by the Sponsor's monitor. The site will contact and discuss with the Sponsor the method of study drug destruction to determine whether the study drug will be shipped to a designated facility contracted by the Sponsor or destroyed at the study center according to the site's Standard Operating Procedures (SOP). If it is determined that the study drug will be destroyed on site, written confirmation of vial destruction and a copy of the institutional SOP must be provided to the Sponsor or its representative.

At the end of the study, all expired or unused medication will be returned to the contract repository with an inventory of returned clinical materials or destroyed on site according to site procedures. The Sponsor will be notified before shipment or destruction.

Patients will be identified by a patient number.

Patients in Phase 3 will be stratified based on his or her tumor type (breast cancer, NSCLC, HRPC) and region (Asia, non-Asia). Patients will be randomized using IWRS to 1 of the following treatment groups:

Phase 2 (approximately 10 patients in each arm): Protocol BPI- Phase 3 (75 patients enrolled in each arm):

Arm 1: Docetaxel (75 mg/m 2 ) + placebo matching plinabulin + pegfilgrastim (6 mg)

Arm 2: Docetaxel (75 mg/m 2 ) + plinabulin (40 mg) + placebo matching pegfilgrastim

Phase 2 Open Label:

During randomization, patients and all personnel involved in the conduct of the study and interpretation of results, including investigators, site personnel, and sponsor staff will be notified of treatment allocation. A master list of all treatments and the patient numbers associated with the treatments will be maintained by the clinical supply vendor, the IWRS vendor, and the sponsor.

During randomization, patients and all personnel involved with the conduct and interpretation of the study, including Investigators, site personnel, and sponsor staff will be blinded to the treatment codes. Randomization data will be kept strictly confidential, filed securely by an appropriate group with the sponsor or contract research organization (CRO) and accessible only to authorized persons.

A master list of all treatments and the patient numbers associated with the treatments will be maintained in a sealed envelope by the clinical supply vendor, the IWRS vendor, and the sponsor. In the event that emergency conditions require knowledge of the study treatment given, the blind may be broken via the code breaker facility within the IWRS. Emergency procedures for revealing drug codes are provided in the study manual. If possible, before breaking the blind, the Investigator should consult with the sponsor to ascertain the necessity of breaking the code.

In the Phase 2 portion, plinabulin will be administered at a dose of 5 mg/m 2 , 10 mg/m 2 , and 20 mg/m 2 . In the Phase 3 portion of the study, plinabulin will be administered at a fixed dose of 40 mg and the study matching placebo of plinabulin will be administered in an equal volume.

In the Phase 2 portion, pegfilgrastim will be administered at a dose of 6 mg as a single dose syringe. In the Phase 3 portion, pegfilgrastim will be administered at a dose of 6 mg as a single dose syringe and the study matching placebo of pegfilgrastim will be administered in an equal volume. The Investigator and study site staff should be experienced in the use of pegfilgrastim and familiar with the pegfilgrastim prescribing information provided by the manufacturer. Pegfilgrastim should be obtained from the institutional pharmacy and prepared per institutional protocol. United States Food and Drug Administration (US FDA) approved source of pegfilgrastim is required for use in this study. The recommended dosage of pegfilgrastim is a single subcutaneous injection of 6 mg administered once per chemotherapy cycle in adults. Do not administer pegfilgrastim between 14 days before and 24 hours after administration of cytotoxic chemotherapy. Pegfilgrastim or matching placebo will be administered 24 hours after completing chemotherapy. Visually inspect parenteral drug products for particulate matter and discoloration prior to administration, whenever solution and container permit. Do not administer pegfilgrastim if discoloration or particulates are observed. NOTE: The needle cover on the single-use prefilled syringe contains dry natural rubber (latex); persons with latex allergies should not administer this product. Follow institutional guideline/practice for prevention/assessment/management of hypersensitivity/infusion reaction with pegfilgrastim administration (refer to Neulasta® Package Insert for further details).

Docetaxel will be administered at a dose of 75 mg/m 2 .

As a standard approved chemotherapy agent, the Investigator and study site staff should be experienced in the use of docetaxel and familiar with the docetaxel prescribing information provided by the manufacturer. Docetaxel should be obtained from the institutional pharmacy and prepared per institutional protocol. US FDA approved source of docetaxel is required for use in this study. Docetaxel may be provided by a distributer in some regions by agreement with the Sponsor. Administration should be carried out with a 1-hour IV infusion per institutional protocol at the dose prescribed by this clinical study protocol (75 mg/m 2 ). Dexamethasone (16 mg per day administered as 8 mg twice daily, or as per institution standard) will be given on the day before, the day of (Day 1), and the day following docetaxel infusion (Day 2). Follow institutional guideline/practice for prevention/assessment/management of hypersensitivity/infusion reaction with docetaxel administration (refer to Taxotere® (Prescribing Information) for further details).

Docetaxel when given on a standard regimen (e.g., 75 mg/m 2 IV every 3 weeks in non-small cell lung cancer) has an intermediate (10%-20%) risk of causing febrile neutropenia. The NCCN guidelines recommend that physicians consider (but do not mandate) myeloid growth factor support in patients with ≥1 risk factor with docetaxel treatment (see below). Patients without risk factors may be observed for their initial treatment cycle. If the patient experiences an episode of febrile neutropenia or a dose limiting neutropenic event (a nadir or a day of treatment count impacting the planned dose of chemotherapy) during the previous treatment cycle, the patient should continue on study, and have dose adjustments as specified in Sections 10.5 and 10.6. 

All adverse events should be assessed according to the Common Terminology Criteria for Adverse Events (CTCAE, v4.03) . In the event of multiple toxicities, dose delays and modifications should occur in accordance with the highest adverse events observed.

All patients, including patients who withdraw from the study early, will complete a safety follow-up visit 30 days (± 2 days) after the last dose of study drug. Patients who withdraw for progressive disease and who will continue to another chemotherapy regimen will complete the EOT visit at Cycle X Day 21. They then continue to another chemotherapy regimen and do not need to have the safety follow-up visit (where X is the last cycle prior to progression, and is 4 or less).

If, in the opinion of the investigator, the patient will benefit from more than 4 cycles of docetaxel and open label pegfilgrastim, then the fifth cycle will not start until completion of the EOT visit (in this instance, the EOT visit will be Cycle 4 Day 21). Follow-up visits will be required to monitor for ongoing treatment-related AEs. All patients experiencing drug-related toxicities of Grade 2 at the EOT visit should be followed-up at least monthly until the AE(s) resolves to Grade 1, the event is considered to be chronic, or the patient receives other anti-cancer therapy. The method of follow-up assessment will be at the Investigator's discretion (for example, patient site visit or telephone call). All deaths which occur within 30 days of study drug administration regardless of relationship to the study drug must be reported to the Sponsor immediately and within 24 hours of becoming aware of the event.

Patients without progressive disease at the EOT visit will undergo a follow-up visit every 2 months until the occurrence of either disease progression or death. These visits may be conducted per telephone or other means. Table 4 , the occurrence of specific Grade 3 or 4 AEs during chemotherapy will require a dose reduction, delay, or discontinuation. Only one docetaxel dose reduction is allowed (refer to Taxotere® (Prescribing Information) for further details). If a chemotherapy cycle is delayed by more than 3 weeks, the patient will be withdrawn from the study.

No dose reductions are allowed with plinabulin or pegfilgrastim.

Dose reduction and treatment delay for docetaxel will follow guidance described in docetaxel prescribing information as summarized in Table 4 .

Hematologic support, such as erythropoietin, darbopoetin, granulocyte colony-stimulating factor (Neupogen ® or Neulasta ® ), or red blood cell or platelet transfusions prior to 4 weeks of the first dose of study drug in order to meet entry criteria is not permitted.

Patients may receive hemopoietic support or transfusions during the study as clinically indicated. Please note the special instructions regarding Neupogen ® or Neulasta ® use during study treatment in Section 10.6.1.

Any delay of docetaxel dosing will result in the delay of plinabulin administration. For each cycle, the dose of plinabulin is always administered 30 minutes after the docetaxel infusion. If docetaxel is discontinued, plinabulin will also be discontinued. Protocol BPI-2358-105 IND 

If the patient has trouble sleeping, anxiety, stomach upset, nausea, vomiting, diarrhea, headache or swelling, they have the option to skip the 3rd day of dexamethasone (or steroid equivalent) dosing, on the day after the day of docetaxel infusion. On the day before and on the day of docetaxel infusion, the patient must take their dexamethasone (or steroid equivalent) as per institution standard.

During the phase 2 study if at any given cohort, 3 patients or more have Grade 4 or 5 toxicity not related to underlying disease (with the exception of neutropenia), accrual to that cohort will be halted and the study will be continued at the lower dose cohorts in phase 2 (for example if 3 patients at the 20 mg/m 2 cohort develop Gr. 4 toxicity the accrual to that cohort will be stopped and the study will continue as planned with the accrual of the two remaining open cohorts).IWRS will be utilized to assign patients to a lower dose cohort in phase 2. Study sites will be instructed to call IWRS when a Grade 4 or 5 toxicity event occurs.

Patients have the right to withdraw from the study at any time for any reason. The Investigator also has the right to withdraw patients from the study if it is in the best interest of the patient. The sponsor may also decide to withdraw a patient. All efforts should be made to complete and report the observations as thoroughly as possible as described in Section 11.7.

Every reasonable effort should be made to determine as completely as possible the reason for the withdrawal, including contacting the patient either by telephone or through a personal visit, or contacting a responsible relative. A complete final evaluation at the time of the patient's withdrawal should be made with an explanation of why the patient is withdrawing from the study. If the reason for removal of a patient from the study is an adverse event or an abnormal laboratory test result, the principal specific event or test will also be recorded on the eCRF.

Patients with clearly documented progressive disease will be taken off treatment.

Concomitant Medications and Non-drug Therapies 10.6.1.

Institutional guidelines should be followed in the event of infusion/hypersensitivity reaction. Diphenhydramine and dexamethasone infusion may be administered in the event of infusion reaction.

All patients must be premedicated with oral corticosteroids (see below for HRPC) such as dexamethasone 16 mg per day (e.g., 8 mg twice daily) for 3 days starting 1 day before docetaxel administration in order to reduce the incidence and severity of fluid retention as well as the severity of hypersensitivity reactions (refer to Taxotere® (Prescribing Information)). If the patient has trouble sleeping, anxiety, stomach upset, nausea, vomiting, diarrhea, headache or swelling, they have the option to skip the 3rd day of Dexamethasone (or steroid equivalent) dosing, on the day after the day of docetaxel infusion. On the day before and on the day of docetaxel infusion, the patient must take their dexamethasone (or steroid equivalent) as per institution standard.

For HRPC, given the concurrent use of prednisone, the recommended premedication regimen is oral dexamethasone 8 mg at 12 hours, 3 hours and 1 hour before docetaxel infusion (refer to Taxotere® (Prescribing Information).

The prophylactic use of antibiotics is allowed per discretion of the treating physician. The use of antibiotics will be recorded per treatment arm (refer Table 5 for prohibited medications).

Corticosteroids (except as described for premedication) and new non-steroidal anti-inflammatory drugs (NSAIDs) are prohibited except for the treatment of AEs and as premedication as described above. Patients already receiving NSAIDs are allowed to continue taking treatment as long as the dose remains stable.

The use of strong CYP3A4 inhibitors as concomitant medications will be prohibited because docetaxel exposure increases by approximately 2-fold (Section 10.6.2) with the use of strong CYP3A4 inhibitors.

Patients who have FN should receive antibiotics per standard of care (refer to Section 10.1.1 and Table 5 for prohibited medications). The use of G-CSF as a treatment option during hospitalization for FN is strongly discouraged, since G-CSF is not approved for the treatment of FN, and is not likely to have efficacy. If, however, G-CSF treatment for FN is considered, the Investigator should contact the Medical Monitor prior to its use. FN is defined as ANC <1.0 × 10 9 /L AND a single temperature of >38.3°C or a sustained temperature of ≥38°C for more than 1 hour (as per CTCAE v4.03). 

Corticosteroids (except as described in Section 10.6.1) and new NSAIDs are prohibited except for the treatment of AEs. Patients already receiving NSAIDs are allowed to continue taking treatment as long as the dose remains stable. Topical formulations of prohibited medication(s) are permitted. The effect of grapefruit juice varies widely among brands and is concentration-, dose-, and preparation-dependent. studies have shown that it can be classified as a "strong CYP3A4 inhibitor" when a certain preparation was used (e.g., high dose, double strength) or as a "moderate CYP3A4 inhibitor" when another preparation was used (e.g., low dose, single strength). Withdrawn from the US market. CYP = cytochrome P450 Strong CYP3A4 inhibitors are not permitted as they may alter docetaxel exposure; consider therapeutic substitutions for these medications. Approval from the Medical Monitor is required in these situations. Please contact the Sponsor for review and approval if questions on any concomitant medication before patient enrollment or during the study treatment.

Docetaxel is low emetic risk (10% to 30% frequency of emesis) and appropriate anti-emetic prophylaxis should be given prior to study medication per institutional policy. In this study, dexamethasone is used as anti-emetic prophylaxis as well as to minimize docetaxel associated fluid retention.

If "breakthrough" emesis (e.g. emesis and nausea after Day 1) occurs, the general strategy is to add one agent from a different drug class to the "rescue" anti-emesis regimen. Useful anti-emetic agents for rescue include the benzodiazepines, cannabinoids (e.g. dronabinol or nabilone), haloperidol, metoclopramide, scopolamine, the phenothiazines (e.g. prochlorperaxine or promethazine) and 5HT3 receptor agonists. If a 5HT3 receptor agonist is needed, palonosetron (which is not known to prolong QT/QTc intervals) is safest and must be chosen over other 5HT3 receptor agonists. Tropisetron is an acceptable substitute 5HT3 receptor agonist (Yavas et al, 2008 ). Because of the potential interference with QT/QTc interval, in Cycle 1 between Day 1 and Day 2, the 5HT3 receptor agonists ondansetron, granisetron and dolenasetron, and the atypical antipsychotic olanzapine, are prohibited until the triplicate ECGs are completed. After Cycle 1, triplicate ECGs are not obtained, and therefore no restrictions in the use of anti-emetics apply in those cycles, thus any 5HT3 receptor agonist can be used.

If nausea and /or vomiting of Grade 2 and higher occurs, it must be treated with "rescue" anti-emetics during mid-cycle, and on subsequent cycles, the prophylactic anti-emetic regimen should be modified. Aprepitant is an acceptable prophylactic anti-emetic in this situation [Marbury et al, 2009] ).

Diarrhea:

If diarrhea of Grade 1 and higher occurs, it must be treated. Grade 1 diarrhea is less than 4 bowel movements a days without any signs of hypotention, dehydration.

Antidiarrheals such as loperamide (or diphenoxylate/atropine) must be prescribed for diarrhea. Suggested loperamide use: 4 mg orally after first loose stool, then 2 mg after each stool not to exceed 16 mg in 24 hours.

Patients should also be cautioned to avoid dehydration, and of the importance of drinking water and electrolyte containing fluids throughout the day when diarrhea occurs. If IV fluids are needed, their administration must be recorded on the CRF.

In patients who develop diarrhea, use of a motility enhancing agent such as metoclopramide as part of the anti-emetic regimen in subsequent treatment cycles should be avoided.

Prophylaxis with bowel motility agents should follow institutional practice as applied to drugs such as vincristine, including the use of agents such as stool softeners, bulking agents, stimulating agents and/or dopamine antagonists. The use of opiates should be limited to when clearly indicated and prophylaxis for opiate induced constipation with agents such as methylnaltrexone should be administered. If significant constipation develops, it should be managed immediately and plinabulin administration should be delayed until resolution. Careful observance for signs of ileus and early diagnostic evaluation with radiographic and/or ultrasound studies is recommended.

The use of anti-emetics and anti-diarrheals will need to be recorded on the CRF.

If an increase in systolic blood pressure to > 160 mmHg is observed after administration of plinabulin or placebo, oral amlodipine 10 mg or an equivalent calcium channel blocker should be administered before each subsequent dose. Increases in systolic blood pressure above 200 mmHg should be managed with nitroprusside or similar regimen per institutional practice. If hypertension can be successfully managed, patient can continue in the study at the discretion of the investigator. The Investigator should be experienced in the use of docetaxel and familiar with the docetaxel prescribing information provided by the manufacturer. According to the prescribing information for docetaxel, there have been no formal clinical studies to evaluate the drug interactions of docetaxel with other medications. In vitro studies have shown that the metabolism of docetaxel may be modified by the concomitant administration of compounds that induce, inhibit, or are metabolized by CYP3A4, (refer to medications should be avoided when patients receive docetaxel as there is a potential for a significant interaction. For patients receiving these medications before the study entry, these medications must be discontinued before docetaxel administration and sufficient time for drug clearance must be provided.

Any other medication which is considered necessary for the patient's welfare including bisphosphonates, and which is not expected to interfere with the evaluation of the study drug, may be given at the discretion of the Investigator.

All medications will be recorded in an appropriate section of the CRF.

No other cancer therapies or investigational agents are permitted during the entire duration of the study treatment (from 21 days before the first administration until the End of Treatment evaluation). 

The study-specific assessments and procedures for Phase 2 and Phase 3 are shown in Table 6 and  Table 7 . Protocol 

Adverse events y X X X X X X X X X X X X X X X X X Abbreviations: ECG = electrocardiogram, EORTC = European Organization for Research and Treatment of Cancer, HIV = human immunodeficiency virus, QoL = Quality of Life. a. EOT is defined as the last assessment for the protocol-specified treatment post cycle 4 (Day 22 [+ 7 days]) of the study for an individual patient. b. If a patient discontinues the study, procedures should be performed within 21 days of the last dose of study drug. c. All patients, (including patients who withdraw from the study early), will complete a safety follow-up visit 30 (+7) days after the last dose. Follow-up visits will be required to monitor for ongoing treatment-related adverse events. All patients experiencing drug-related toxicities of Grade 2 at the End of Treatment visit should be followed-up at least monthly until the adverse event(s) resolves to Grade 1, the event is considered to be chronic or the patient receives other anti-cancer therapy. The method of follow-up assessment will be at the Investigator's discretion (for example, patient site visit or telephone call). All deaths which occur within 30 days of study drug administration regardless of relationship to the study drug must be reported the Sponsor immediately and within 24 hours of becoming aware of the event. d. Demographic data will include gender, date of birth (or age), and race/ethnicity. e. Background characteristics will include a history of disease and current disease status, bone marrow involvement, sites of disease, prior anticancer therapies, and prior medications/significant non-drug therapies. f. Patients must be in a supine position in a rested and calm state for at least 5 minutes before blood pressure is assessed. If the patient is unable to be in the supine position, the patient should be in the most recumbent position possible. The position selected for a patient, the same arm, and same blood pressure cuff should be kept the same throughout the study. Two methods will be used to collect blood pressure. Method 1 in Phase 2 ONLY, Cycle 1 Day 1 ONLY, using blood pressure devices provided by the sponsor (SpaceLabs 90217 ambulatory blood pressure monitor): Day 1 (Arm 1): Blood pressure and heart rate will be measured semi-continuously starting at 15 minutes after completion of docetaxel infusion, and at every 15 minutes thereafter for 4.5 hours. Day 1 (Arms 2 to 4): Blood pressure and heart rate will be measured semi-continuously starting at 15 minutes before plinabulin infusion (15 minutes after completion of docetaxel infusion), and at every 15 minutes thereafter for 4.5 hours after start of infusion with plinabulin. Method 2 in Phase 2 in Cycles 1, 2, 3, and 4 as follows:

A standard cuff will be used to measure blood pressure (heart rate will also be measured):

Cycle 1 (Arms 1 to 4): At screening, Day 1 (pre-docetaxel), Days 2, 6, 7, 8, 9, 10 , and 15 (once, prior to blood draw). Cycles 2 to 4 (Arm 1): Day 1 pre-docetaxel dose, and on Day 8 prior to blood draw Cycles 2 to 4 (Arms 2 to 4): On Day 1 at pre-dose, 30 min, 1 hour, 2 hours post-infusion with plinabulin, and on Day 8 g. Temperature is to be taken every time a blood draw is taken for neutrophil count and can be taken orally or in the ear; however, the same method (ear or oral) should be used throughout the study for each patient; thus if the ear method was used the first time for a given patient, the ear method should be used throughout the study for that patient. h. Height (cm) will be measured at screening. i. A single 12-lead ECG will be performed at screening, EOT, Early Discontinuation and 30 Day Safety Follow-Up. All other ECGs will be performed in triplicate. In Cycle 1 Day 1, ECG will be collected before docetaxel infusion, immediately before plinabulin infusion, 5-minutes before end of plinabulin infusion, 30 minutes and 4.5 hours after start of infusion with plinabulin. In Cycle 1 Day 2, ECG will be performed in triplicate prior to the blood draws on Day 2 in Arms 2 to 4. For Arm 1, the triplicate ECGs will not be performed. j. Laboratory test samples (hematology and serum chemistry) will be collected and sent to the protocol central laboratory. Safety laboratory tests are required prior to treatment on Day 1 of each cycle and can be collected by a local laboratory and will be used to determine docetaxel dosing; however, all other safety (e.g. protocol specified) blood samples as per the schedule assessments and procedure table must also be obtained for central laboratory assessment. In addition a central laboratory blood draw needs to be taken on the day of dosing on Day 1 of each cycle, prior to the docetaxel dosing. Neutrophils are to be collected on time points as indicated in this schedule; neutrophils must be collected at pre-dose on Day 1 of each cycle. During Cycle 1, neutrophils count will be assessed at baseline (prior to Cycle 1 docetaxel dose), pre-dose on Day 1 and on Days 2, 6, 7, 8, 9, 10, and 15. k. Analyzed at a central laboratory. l. CD34+ will be analyzed using FACS via a central laboratory. m. Samples for CD34+ analysis to be collected at Day 1 Cycle 2 only. Do not collect at cycle 3 and 4 visits. n. Hepatitis B surface antigen reactive, hepatitis B core antibody, hepatitis B surface antibody, and hepatitis C antibodies o. Pregnancy tests will be done using urine samples in women of childbearing potential. Subject must have a negative urine pregnancy test documented within the 24-hour period prior to the first infusion. Confirm with serum testing (central laboratory) if urine sample is positive. p. Plasma samples (5 mL each) for plinabulin and docetaxel PK. All patients will be sampled for PK via a central laboratory. For PK collection schedule refer to Table 13 and Table 14 (see Section 11.9). During the phase 2 open label portion of the study patients randomized to pegfilgrastim will not have samples collected for plinabulin PK analysis. Cytokine panel testing will be done using unused plasma samples collected at the PK time points; reconsenting of the patients will be required before analysis is performed. q. Investigator opinion of progression (yes/no) at End of Treatment (EOT) recorded in CRF. For example if the patient completes two cycles of docetaxel and study drug, and in the opinion of the investigator per institutional practice the cancer is growing and a new treatment is required, then the EOT evaluation will be performed as specified, and the "disease progression" will be scored as "yes." As an another example, if after four cycles of docetaxel and study drug, the cancer is stable or responding, and the patient receives further docetaxel, then the EOT evaluation will be completed as specified, and the "disease progression" will be scored as "no". Patients without progressive disease at the EOT visit will undergo a follow-up visit every 2 months until the occurrence of either disease progression or death. These visits may be conducted per telephone or other means. r. Docetaxel infusion: 75 mg/m 2 docetaxel will be administered via IV infusion over 1 hour on Day 1 of each cycle. All patients must be premedicated with oral corticosteroids such as dexamethasone 16 mg per day (e.g., 8 mg bid) for 3 days starting 1 day prior to docetaxel administration in order to reduce the incidence and severity of fluid retention as well as the severity of hypersensitivity reactions (refer to Taxotere ® Package Insert). s. Cycles 1 to 4 will consist of docetaxel 75 mg/m 2 administered by IV on Day 1 over 60 minutes (±5 minutes) each 21 day cycle. Patients will get a single dose of plinabulin intravenously over 30 minutes (± 5 minutes) 30 minutes after the end of the docetaxel infusion if in arms 2-4. On Day 2 of each cycle ≥ 24 hours after completing chemotherapy, patients will receive a single dose of pegfilgrastim (6 mg) if in arm 1. t. The bone pain questionnaire should be completed prior to docetaxel infusion and at the site if possible, if not the questionnaire needs to be returned to the site at the next scheduled visit. u. Bone pain questionnaire to be completed prior to pegfilgrastim and plinabulin pre-dose, Cycle 1 v. Bone pain questionnaire to be collected at Cycle 2: Pre-dose Day 1, only. Do not collect at Cycle 3 and 4 visits. w. Health-related QoL questionnaire evaluated with EORTC QLQ-C30 and EQ 5D-5L will be collected prior to docetaxel infusion on Day 1 of each cycle.

x. All concomitant medicines (dose, schedule, and duration of treatment) and in particular analgesics as well as antibiotics should be entered in the eCRF. y. All hospitalizations should be entered in the eCRF. 

Adverse events x X X X X X X X X X X X X X X X X Abbreviations: ANC = absolute neutrophil count; ECG = electrocardiogram, EORTC = European Organization for Research and Treatment of Cancer, HIV = human immunodeficiency virus, QoL = Quality of Life. a. EOT is defined as the last assessment for the protocol-specified treatment post cycle X (Day 22 [+ 7 days]) of the study for an individual patient, regardless at what time the patient discontinues the study. b. All patients, (including those who discontinued from the study) will complete a safety follow-up visit 30 (±2) days after the last dose. Patients who withdraw for progressive disease and who will continue to another chemotherapy regimen will complete the EOT visit at Cycle X Day 21 (where X is the last cycle prior to progression, and is 4 or less). They then continue to another chemotherapy regimen and do not need to have the safety follow-up visit. investigator, the patient will benefit from more than 4 cycles of docetaxel and open label pegfilgrastim, then the fifth cycle will not start until completion of the EOT visit (in this instance, the EOT visit will be Cycle 4 Day 21). Follow-up visits will be required to monitor for ongoing treatment-related adverse events. All patients experiencing drug-related toxicities of Grade 2 at the End of Treatment visit should be followed-up at least monthly until the adverse event(s) resolves to Grade 1, the event is considered to be chronic or the patient receives other anti-cancer therapy. The method of follow-up assessment will be at the Investigator's discretion (for example, patient site visit or telephone call). All deaths which occur within 30 days of study drug administration regardless of relationship to the study drug must be reported the Sponsor immediately and within 24 hours of becoming aware of the event. c. Demographic data will include gender, date of birth (or age), and race/ethnicity d. Background characteristics will include a history of disease and current disease status, bone marrow involvement, sites of disease, prior anticancer therapies, and prior medications/significant non-drug therapies e. Patients must be in a supine position in a rested and calm state for at least 5 minutes before blood pressure is assessed. If the patient is unable to be in the supine position, the patient should be in the most recumbent position possible. The position selected for a patient, the same arm, and same blood pressure cuff should be kept the same throughout the study.

A standard cuff will be used to measure blood pressure (heart rate will also be measured), with a ±15 minute window: Cycle 1: On Day 1 at pre-dose docetaxel, pre-dose plinabulin, 30 min, 1 hour, 2 hours post-infusion with plinabulin or placebo, and on Days 2, 6, 7, 8, 9, 10, and 15 (once, prior to blood draw). Cycles 2 to 4: On Day 1 at pre-dose docetaxel, pre-dose plinabulin, 30 min,1 hour, 2 hours post-infusion with plinabulin or placebo and on Day 8 f. Temperature is to be taken every time a blood draw is taken for ANC. The same temperature location selected for a patient (ear, oral, or axillary) should be used throughout the study for each patient. g. Physical examination will include height (cm) at screening. h. A single 12-lead ECG will be performed at screening, EOT, and 30 Day Safety Follow-Up. All other ECGs will be performed in triplicate. In Cycle 1 Day 1, ECG will be collected before docetaxel infusion, immediately before plinabulin/placebo infusion, 5-minutes before end of plinabulin/placebo infusion, 60 minutes and 4.5 hours after start of infusion with plinabulin/placebo. In Cycle 1 Day 2, ECG will be performed in triplicate prior to pegfilgrastim/placebo injection. i. Laboratory test samples (hematology and serum chemistry) will be collected and sent to the protocol central laboratory. Safety laboratory tests are required prior to treatment on Day 1 of each cycle and can be collected by a local laboratory and will be used to determine docetaxel dosing; however, all other safety (e.g. protocol specified) blood samples as per the schedule assessments and procedure table must also be obtained for central laboratory assessment. In addition a central laboratory blood draw needs to be taken on the day of dosing on Day 1 of each cycle, prior to the docetaxel dosing. Neutrophils are to be collected on time points as indicated in this schedule; neutrophils must be collected at pre-dose on day 1 of each cycle. During cycle 1, ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose), on Days 1, 2, 6, 7, 8, 9, 10, and 15. Blood draws for ANC will be taken approximately the same time as the time of the pre-dose sample on Day 1, and will be taken by preference in the morning. If Grade 4 neutropenia is observed in a patient, absolute neutrophil count will be evaluated daily (central and local laboratories) during the first cycle, until Grade 4 neutropenia is resolved. j. Analyzed at a central laboratory. k. Hepatitis C antibody, hepatitis B surface antigen (HBsAg), viral load at local laboratory if HBsAg is positive, hepatitis B surface antibody (anti-HBs). Hepatitis serology can be performed using laboratory standard methodology, by local laboratory or by the central laboratory if local testing is not available. l. If HIV testing is done (at the discretion of the investigator), the test will be conducted at the local site. m. Pregnancy tests will be done using urine samples in women of childbearing potential. Subject must have a negative urine pregnancy test documented within the 24-hour period prior to the first infusion (either on Day -1 or prior to infusion on Day 1). Confirm with serum testing (local or central laboratory) if urine sample is positive. Additional testing may be performed at the discretion of the investigator. n. Patients can be randomized on Cycle 1 Day -1 or Cycle 1 Day 1, to allow site flexibility to dispense pre-medication treatment. o. Plasma samples (5 mL each) for plinabulin and docetaxel PK. All patients will be sampled for PK via a central laboratory. Patients in Phase 3 will follow the plinabulin and docetaxel PK sampling schedules from Phase 2. For PK collection schedule refer to Table 13 and Table 14 (see Section 11.9). Cytokine panel testing will be done using unused plasma samples collected at the PK time points; reconsenting of the patients will be required before analysis is performed. p. Investigator opinion of progression (yes/no) at End of Treatment (EOT) recorded in CRF. For example if the patient completes two cycles of docetaxel and study drug, and in the opinion of the investigator per institutional practice the cancer is growing and a new treatment is required, then the EOT evaluation will be performed as specified, and the "disease progression" will be scored as "yes." As an another example, if after four cycles of docetaxel and study drug, the cancer is stable or responding, and the patient receives further docetaxel, then the EOT evaluation will be completed as specified, and the "disease progression" will be scored as "no." Patients without progressive disease at the EOT visit will undergo a follow-up visit every 2 months until the occurrence of either disease progression or death. These visits may be conducted per telephone or other means. q. Docetaxel infusion: 75 mg/m 2 docetaxel will be administered via IV infusion over 1 hour on Day 1 of each cycle. All patients must be premedicated with oral corticosteroids (see below for prostate cancer) such as dexamethasone 16 mg per day (e.g., 8 mg bid) for 3 days starting 1 day prior to docetaxel administration in order to reduce the incidence and severity of fluid retention as well as the severity of hypersensitivity reactions (refer to Taxotere ® Package Insert). For hormone-refractory metastatic prostate cancer, given the concurrent use of prednisone, the recommended premedication regimen is oral dexamethasone 8 mg, at 12 hours, 3 hours and 1 hour before the docetaxel infusion (refer to Taxotere ® Package Insert). r. Cycles 1 to 4 will consist of docetaxel 75 mg/m 2 administered by IV on Day 1 over 60 minutes (±5 minutes) each 21 day cycle. Patients will get a single dose of plinabulin or placebo intravenously over 30 minutes (± 5 minutes) 30 minutes after the end of the docetaxel infusion. On Day 2 of each cycle ≥ 24 hours after completing chemotherapy, patients will receive a single dose of pegfilgrastim (6 mg) or placebo (subcutaneous injection). s. The bone pain scale and pain medication assessment should be completed on Day 1 prior to docetaxel infusion and on each remaining day through Day 8 at approximately the same time (Moore et al, 2017) . On Days 2, 6, 7, and 8 the patient will complete the bone pain scale and pain medication assessment at the investigational site t. Bone pain assessment to be completed pre-dose. u. On Day 2 the patient will be provided with a further copy of the assessment to be completed at home on Day 3, Day 4, and Day 5 and returned to the study center at the next visit (Day 6). v. Health-related QoL questionnaire evaluated with EORTC QLQ-C30 and EQ 5D-5L will be collected on Day 1 (prior to the infusion) and on Day 8 of each cycle (Giesinger et al, 2014). w. All concomitant medicines (dose, schedule, and duration of treatment) and in particular analgesics as well as antibiotics should be entered in the eCRF.

x. All hospitalizations should be entered in the eCRF. Adverse events will be collected at all study visits starting with the first dose of study drug. Serious adverse events are collected from the time of signing the informed consent form and up to 30 days following the last dose of the study drug. 

Demographic data will include gender, date of birth (or age), and race/ethnicity.

Medical history findings (i.e., previous diagnoses, diseases, or surgeries) not pertaining to the study indication, started before signing the informed consent, and considered relevant for the patient's study eligibility will be collected and captured in the eCRF.

Baseline characteristics will include a history of disease and current disease status, staging, bone marrow involvement, sites of disease, prior anticancer therapies, and prior medications/significant non-drug therapies will be collected.

Information will also be collected regarding child-bearing potential and any other assessments that are done for the purpose of eligibility for inclusion into the study (physical examination, vital signs, hematology and blood chemistry, urinalysis, pregnancy test, and ECG). For further details on eligibility assessments, see Table 6 and Table 7 .

The following measurements for vital signs must be performed: systolic/diastolic blood pressure, heart rate, and respiratory rate. The patient must be in a supine position in a rested and calm state for at least 5 minutes before blood pressure assessments are conducted. If the patient is unable to be in the supine position, the patient should be in the most recumbent position possible. The position selected for a patient, the same arm, and same blood pressure cuff should be the same throughout the study.

Two methods will be used to collect blood pressure.

Method 1: using blood pressure devices provided by the sponsor (SpaceLabs 90217 ambulatory blood pressure monitor) to measure blood pressure (heart rate will also be measured)

Method 2: using a standard cuff to measure blood pressure (heart rate will also be measured)

Vital assessments will be performed as detailed in Table 8 , with a ± 15 minute window for the standard cuff measurements. 

The temperature location (ear,oral, or axillary) selected for a patient should be the same throughout the study.

If abnormalities are found and they are considered an AE, record on the AE summary page. If an isolated elevated temperature is not clinically significant, it should not be reported as an AE. If an elevated temperature is part of an AE that is reported elsewhere, elevated temperature (fever) should not be reported as an AE.

Physical examinations (comprehensive [including neurological examination] or symptom directed) will be performed as described in Table 6 and Table 7 . A comprehensive physical examination will include evaluations of the head, eyes, ears, nose, throat, neck, chest (including heart and lungs), abdomen, limbs, skin, and a complete neurological examination. A urogenital examination will only be required in the presence of clinical symptoms related to this region. Documentation of the physical examination will be included in the source documentation at the site. Significant findings at the screening Visit will be recorded on the CRF. Changes from the screening physical examination findings that meet the definition of an AE will be recorded on CRF.

Height will be measured in centimeters once at screening. Weight will be measured in kilograms at screening, prior to Day 1 dosing in all cycles and at the end of treatment (EOT) visit. For Cycle 1, body weight will also be measured on Days 2 and 6. The body weight measurements are made to enable the calculation of the plinabulin dose and to monitor for potential weight increase due to the docetaxel pre-medication. Both measurements will be performed without the patient wearing shoes.

Patients will be graded according to the ECOG Performance Status scale and criteria as described in Table 9 . 

A single 12-lead ECG will be performed at screening and triplicate ECGs will be taken for Phase 2 (Table 10) and Phase 3 (Table 11) .

ECGs are to be performed using a standardized method before blood draws or any other procedures. The patient must be in a supine position in a rested and calm state for at least 5 minutes before the ECG assessment is conducted. If the patient is unable to be in the supine position, the patient should be in most recumbent position as possible.

In Phase 2 and Phase 3, single safety ECGs will be obtained during screening, at the EOT, Early Discontinuation (Phase 2 only), and at 30-day safety follow-up visit for all patients. ECG in triplicate with 2-3 minutes between measurements will be obtained in Phase 2 (Arms 2 to 4 only) and Phase 3 on Cycle 1 Day 1 (in both arms) before docetaxel infusion, immediately before plinabulin infusion, 5 minutes before end of plinabulin infusion, 60 minutes (30 minutes in Phase 2) and 4.5 hours after start of infusion with plinabulin or matching placebo. On Cycle 1 Day 2, ECG in triplicate will be obtained prior to blood draws in Arms 2 to 4 (Phase 2) or prior to pegfilgrastim/placebo injection in both arms (Phase 3).

For Arm 1 of Phase 2, the triplicate ECGs will not be obtained; so the only assessments will be the single ECGs at screening, end of treatment, early discontinuation, and follow-up.

The ECG must include the following measurements: heart rate, QRS, QT, and PR intervals. 

Health-related QoL questionnaire evaluated with EORTC QLQ-C30 and EQ-5D-5L is a validated questionnaire developed to assess the QoL of cancer patients.

It is a copyrighted instrument, which has been translated and validated in over 90 languages and used in more than 3,000 studies worldwide.

Refer to the study manual for specific module and how to administer.

The incidences of bone pain will be recorded through a validated questionnaire. The bone pain questionnaire should be completed prior to docetaxel infusion and at the site if possible, if not the questionnaire needs to be returned to the site at the next scheduled visit

The incidences, occurrences, and severity of bone pain will be recorded through a patient bone pain scale during Cycle 1, at the times presented in Table 7 .

The bone pain scale will be based upon the validated Wong Baker Faces ® Pain Rating Scale (http://wongbakerfaces.org).

The bone pain scale and pain medication assessment should be completed on Day 1 prior to docetaxel infusion and on each remaining day through Day 8 at approximately the same time. On Days 2, 6, 7, and 8 the patient will complete the bone pain scale and pain medication assessment at the investigational site. If the patient is discharged from hospital prior to Day 8, the bone pain scale and pain medication assessment pages for the remaining days will be handed to the patient with the date field completed by the site personnel for each of the remaining days. The patient will be instructed to fill in the documents at home, approximately at the same time as on Day 1. On each day following the patient's discharge from the hospital when the patient returns to the site for study procedures, they will return any available completed assessments.

Increases of bone pain severity are assessed as an objective during Phase 3 of the study, using the bone pain scale ("Please rate your bone pain by circling the one number below the faces that best describes your pain at its worst in the last 24 hours"). The severity of pain is marked on a scale from 0 ('no hurt') to 10 ('hurt worst').

Concomitant pain medication use will be recorded via a pain medication assessment. Information recorded for concomitant pain medication use will include date, name of medication, route of administration, dosage, total taken in one day, and reason medication was taken.

Efficacy assessments and the time when they will be performed are presented in Table 6 and  Table 7 .

Patients will be evaluated for disease progression in accordance with institutional practice. For example, the investigator may identify target lesions at screening, and at the EOT visit will assess the patient's response to the docetaxel and study drug. This will be designated the patient's treatment response. If no reliable target lesions are identifiable (e.g. patients with boneonly cancer metastases), institutional response criteria will be used for recording disease progression.

Investigators should exercise their clinical judgment in performing studies (including computed tomography and positron emission tomography scans) to assess disease progression, and can end study treatment if a patient is clearly having disease progression prior to the completion of the planned 4 cycles of chemotherapy. If a patient is taken off-study prior to completing 4 cycles of docetaxel and study drug, the screening and end of study images, when clinically indicated, will be performed. If the patient completes four cycles of docetaxel and study drug, and there is not disease progression and the docetaxel is continued, then the EOT assessment of disease progression will be recorded as "no."

The investigator will submit supporting documentation for the treatment response determination, which could include reports from cross-sectional imaging, laboratory values, and performance status estimates (before and at the end of study treatment), and this data will support the investigator's determination in the CRF as disease progression "yes" or "no". Source documents supporting the investigator's judgment of disease progression will be reviewed, but images (such as CT scan, MRI etc.) or other documents will not be collected for central review. The disease progression will be judged by the investigator.

The assessment of disease progression during the follow-up period for patients assessed as non-progressive at the EOT visit does not need to be reported, only the determination of disease progression will be collected.

(Note: a requirement for a target lesion or for "measurable disease" is not required for study entry.) Protocol BPI- 

Safety assessments should be performed at all visits to the study center and throughout the study. The list of events and the time when they will be performed are presented in Table 6 and Table 7 .

Adverse events observed by the Investigator or reported by the patient will be collected at all study visits starting with the first dose of study drug.

Serious adverse events are collected from the time of signing the informed consent form and up to 30 days following the last dose of the study drug.

All AEs, SAEs, treatment emergent AEs, treatment emergent SAEs, and treatment emergent deaths regardless of the relationship to the study drug, will be collected.

Disease progression or deterioration of the malignancy under study (including new sites of metastasis due to disease progression) will be recorded as part of the subject's disease status and should not be reported as an AE/SAE. If death is determined to be due to disease progression, the SAE form will attribute the death to disease progression and document that disease progression was determined to have caused the event(s). These events will not be included in the safety analysis.

The Sponsor will collect all deaths on study regardless of its potential relationship to disease progression and up to 30 days after the last dose of plinabulin on an SAE form.

All hospitalizations should be documented in the eCRF.

All concomitant medications and in particular all analgesics, pain medications, and antibiotics should be recorded with date of onset and discontinuation, dose, and frequency will be entered into the eCRF.

Information on all cause hospitalizations (including ICU) and ER visits will be collected at all visits to the study center and throughout the study, at end of study and during any follow-up post study (except for FU visits for patients without disease progression at the EOT visit).

Chemistry, coagulation tests, hematology, and urinalysis will be performed using a central laboratory and serum pregnancy testing and hepatitis serology will be performed using a local or central laboratory; results will be used to determine docetaxel dosing. Safety laboratory tests are required prior to treatment on Day 1 of each cycle and can be collected by a local laboratory; however, all other safety (e.g. protocol specified) blood samples as per the schedule of assessments (refer to Table 6 and Table 7 ) must also be obtained for central laboratory assessment. In addition a central laboratory blood draw needs to be taken on the day of dosing on Day 1 of each cycle, prior to the docetaxel dosing. The Sponsor or the central laboratories will supply containers for sample collection, preparation, packaging, and shipping. Detailed instructions for sample collection, processing, and shipping are provided in the central laboratory manual and/or the Sponsor will provide training materials. The date and time of sample collection will be recorded in the source documents at the site. Table 12 outlines the specific analytes that will be assessed during the study at time points outlined in the Schedule of Assessments (Table 6 and Table 7 ). Assessments for UDP glucuronosyltransferase (UGT) 1A1 genotyping and total bilirubin, conjugated bilirubin, and unconjugated bilirubin may be performed at the discretion of the investigator.

Pregnancy tests will be done using urine samples in women of childbearing potential. Subject must have a negative urine pregnancy test documented within the 24-hour period prior to the first Protocol BPI-2358-105 IND infusion. Confirm with serum testing (local or central laboratory) if urine sample is positive. Additional testing may be performed at the discretion of the investigator.

Hepatitis B/C serologic markers (hepatitis C antibody, HBsAg, viral load in case HBsAG is positive and anti-HBs) will be tested. If HBsAg is positive but viral load is negative, the patient should be tested prior to each treatment cycle to ensure that viral load remains negative.

Hepatitis (hepatitis C antibody, HBsAg, viral load and anti-HBs) serology can be performed using laboratory standard methodology, by local laboratory or by the central laboratory if local testing is not available (not for viral load).

If HIV testing is done (at the discretion of the investigator), the test will be conducted at the local site using standard methodology.

Phase 2 only: a blood sample for exploratory marker evaluation will be collected for CD34+ which will be established by FACS. This test will be performed in selected countries participating in the study, via central laboratory.

Phase 2 and Phase 3: Plasma cytokine panel: IL-1beta, IL-6, IL-12p70, IL-12p40, IL-17A, IL-23, GCSF, GMCSF, IFN alpha, IFN-gamma, TNF-alpha, IL-2, FLT-3 ligand, and IL-8 analysis. This test will be performed in selected countries participating in the study, using unused plasma samples collected at PK time points.

Unused samples (urine/serum plasma) will also be saved for future potential biomarker research, with patient consent.

EOT is defined as the last assessment for the protocol-specified treatment post Cycle X (Day 22 [+ 7 days]) of the study for an individual patient, regardless at what time the patient discontinues the study.

All patients, including patients who withdraw from the study early, should complete a safety follow-up visit 30 (± 2) days after the last dose (refer to Table 6 and Table 7 ). Patients who withdraw for progressive disease and who will continue to another chemotherapy regimen will complete the EOT visit at Cycle X Day 21. They then continue to another chemotherapy regimen and do not need to have the safety follow-up visit (where X is the last cycle prior to progression, and is 4 or less). If, in the opinion of the investigator, the patient will benefit from more than 4 cycles of docetaxel and open label pegfilgrastim, then the fifth cycle will not start until completion of the EOT visit (in this instance, the EOT visit will be Cycle 4 Day 21). Follow-up visits will be required to monitor for ongoing treatment-related AEs. All patients experiencing drug-related toxicities of Grade 2 at the EOT visit should be followed-up at least monthly until Protocol BPI-2358-105 IND the adverse event(s) resolves to Grade 1, the event is considered to be chronic, or the patient receives other anti-cancer therapy. The method of follow-up assessment will be at the Investigator's discretion (for example, patient site visit or telephone call). All deaths which occur within 30 days of study drug administration regardless of relationship to the study drug must be reported the Sponsor immediately and within 24 hours of becoming aware of the event.

Patients without progressive disease at the EOT visit will undergo a follow-up visit every 2 months until the occurrence of either disease progression or death. These visits may be conducted per telephone or other means.

The assessment of disease progression during the follow-up period for patients assessed as non-progressive at the EOT visit does not need to be reported, only the determination of disease progression will be collected.

All patients in Phase 2 and Phase 3 will participate in the PK assessments.

All patients in Phase 2 and Phase 3 will have samples taken for docetaxel PK on Cycle 1 Day 1 for up to 24 hours after the start of the docetaxel infusion. 

6.0 hours after start of infusion** (+/-15 min)

on Day 1 (+/-6 hr) * Sample to be taken 30 minutes after the end of the 60-minute docetaxel infusion. ** Sample to be taken 5 hours after the end of the 60-minute docetaxel infusion. Sample can be taken at the same time as the plinabulin 4.5 hour sample.

All patients randomized to a plinabulin treatment arm in Phase 2 and Phase 3 will have samples taken for plinabulin PK on Cycle 1 for up to 24 hours after the start of the plinabulin infusion. Patients in Phase 2 randomized to pegfilgrastim during open label treatment will not have samples collected for plinabulin PK. Cytokine panel testing will be done using unused plasma samples collected at the PK time points. Protocol BPI-2358-105 IND 

on Day 1 (+/-6 hr) * Sample to be taken 30 minutes after the end of the 30-minute infusion. ** Sample to be taken 4 hours after the end of the 30-minute infusion.

All safety assessments used in this study are standard, i.e., widely used and generally recognized as reliable, accurate, and relevant, either in clinical practice or specifically in cancer patients. Health-related QoL questionnaire evaluated with EORTC QLQ-C30 and EQ-5D-5L is validated for use in this population.

The patient bone pain scale will be based upon a validated pain rating scale (the Wong-Baker Faces ® Pain Rating Scale (http://wongbakerfaces.org). 

Data management will be the responsibility of the Sponsor. CRFs and edit checks will be designed and validated based on protocol requirements for data collection and with input from the statistical, data management and clinical operations staff.

Data will be captured by using an online Electronic Data Capture (EDC) system. Data collected in patient source documents will be entered onto the eCRFs by site study staff, and subject to an audit trail of changes made to the eCRF. For each patient enrolled, an eCRF must be completed and electronically signed by the principal investigator or authorized delegate from the study staff. The investigator should ensure the accuracy and completeness of the data reported to the sponsor/ CRO in the eCRFs. Data will be stored in files that can be accessed through listing exports.

A CRO will be responsible for data management, including quality checking of the data. During the data entry and verification process, edit check programs will identify data discrepancies and, automatically generate queries. Query reports will be sent to sites or to laboratories for resolution. The sites will correct the data as needed to resolve the queries. Any data captured electronically (such as laboratory results) will be transferred to the database electronically and edit checks will be programmed to search for missing and out of range data. For electronically transferred data, the laboratory is expected to re-send the data transferred with the correction applied. Per Sponsor's or designee's operating procedures, an audit trail will be maintained.

Throughout the study, the Study Management Team will review data according to the Edit Specifications Document as described in the Data Management Plan.

The database will be audited for quality assurance by an outside vendor based on a predefined study audit plan to ensure acceptable accuracy and completeness. Protocol BPI-2358-105 IND 

The investigator is responsible for the detection and documentation of events meeting the criteria and definition of an AE or SAE as provided in this protocol.

An AE is any untoward medical occurrence in a patient or clinical investigation patient administered a pharmaceutical product and which does not necessarily have to have a causal relationship with this treatment. An AE can, therefore, be any unfavorable and unintended sign (including an abnormal laboratory finding, for example), symptom, or disease temporally associated with the use of a medicinal product, whether or not considered related to the medicinal product. Pre-existing conditions, which worsen during a study, are to be reported as AEs.

All AEs encountered during the clinical study, irrespective of the likelihood that the event is related to disease progression, will be reported on the AE page of the CRF.

Whenever possible, the intensity of clinical AEs will be graded according to NCI Common Terminology Criteria for Adverse Events (CTCAE) (v 4.03) grading system. Adverse events not listed on the NCI-CTCAE grading system will be graded on a 5-point scale (mild, moderate, severe, life-threatening, and death) as described below, and reported in detail as indicated on the CRF (Study Manual).

Grade 1: Mild -Transient or mild discomfort; no medical intervention or therapy required.

Moderate -Moderate discomfort or limitation in activity -some assistance may be needed or minimal medical intervention or therapy required.

Severe -Marked limitation in activity, some assistance required; medical intervention or therapy required; hospitalization possible.

Life Threatening -Extreme limitation in activity; significant assistance required; significant medical intervention or therapy required; hospitalization probable.

The investigator will assess the possible relationship of a AE or SAE to the use of study drug. An investigator's causality assessment is the determination of whether there exists a reasonable possibility that the investigational product caused or contributed to an AE; generally, the facts (evidence) or arguments to suggest a causal relationship should be provided.

The following grading of causality will be used for this study: Protocol BPI- Related: There is a reasonable causal relationship between the study drug and the event, and the event occurred within a plausible time relationship to drug administration, and the event cannot be explained by the condition under study, concurrent disease, other drugs or chemicals, or other circumstances. The event responds to withdrawal of study drug (dechallenge) and recurs with rechallenge (if clinically feasible to rechallenge).

Probable: There is reasonable causal relationship between the event and the study drug, the event occurred within plausible time relationship to drug administration, the event is unlikely to be attributed to the condition under study, concurrent disease, other drugs or chemicals, or other circumstances. The event follows a clinically reasonable response on withdrawal of study drug.

Possible: There is a reasonable causal relationship between the event and study drug, the event occurred within a plausible time relationship to study drug administration, but the event could also possibly be explained by the condition under study, concurrent disease, other drugs or chemicals, or other circumstances. Dechallenge information is lacking or unclear.

Unlikely: There is a temporal relationship of the event to study drug but not a reasonable causal relationship, or there is no temporal relationship to study drug administration or the condition under study, concurrent disease, other drugs or chemicals, or other circumstances provide a plausible explanation for the event.

Unrelated: There is no temporal relationship between the event and study drug administration (too early or late or study drug not administered). There is no reasonable causal relationship between the event and the study drug. The condition under study, concurrent disease, other drugs or chemicals, or other circumstances provides a plausible explanation for the event.

For the purposes of regulatory reporting a causality assessment of related, probable, or possible, will be treated as related.

If the investigator's causality assessment is "unknown but not related to investigational product", this should be clearly documented on study records.

In addition, if the investigator determines an SAE is associated with study procedures, the investigator must record this causal relationship in the source documents and CRF and report such an assessment in accordance with the SAE reporting requirements.

For the purpose of this study, AEs irrespective of causality will be collected from first study drug administration (Study Day 1) and all SAEs irrespective of causality will be collected from the time the Informed Consent Form is signed. AEs and SAEs will be collected until 30 days after the last infusion of study treatment or initiation of another anti-cancer therapy. Thereafter, all SAEs which are considered to be drug-related should be reported, regardless of time elapsed since the last dose of study drug (even if the study has stopped). 

Laboratory test results will be recorded on the laboratory results pages of the eCRF. Laboratory test value abnormalities as such should not be reported on the AE page of the CRF as AEs unless they result in a clinically significant condition as judged by the investigator.

The definition and reporting requirements of International Council on Harmonisation (ICH) Guideline for Clinical Safety Data Management, Definitions and Standards for Expedited reporting, Topic E2 will be adhered to.

An SAE is any experience that suggests a significant hazard, contraindication, side effect or precaution. With respect to human clinical experience, this includes any experience which:

is fatal, is life-threatening, requires in-patient hospitalization or prolongation of existing hospitalization, results in persistent or significant disability/incapacity, is a congenital anomaly/birth defect, or is considered medically significant by the investigator or requires intervention to prevent 1 or other of the outcomes listed above, Medical and scientific judgment should be exercised in deciding whether expedited reporting to the Sponsor is appropriate in other situations, such as important medical events that may not be immediately life-threatening or result in death or hospitalization but may jeopardize the patient or may require intervention to prevent 1 of the outcomes listed in the definitions above. These situations should also usually be considered serious.

Examples of such events are intensive treatment in an emergency room or at home for allergic bronchospasm; blood dyscrasia or convulsions that do not result in hospitalization; or development of drug dependency or drug abuse. The term severe is a measure of intensity, thus an SAE is not necessarily serious. For example, nausea of several hours' duration may be rated as severe, but may not be clinically serious.

Hospitalization or prolongation of hospitalization is used as an indicator of the seriousness of the AE and should ONLY be submitted as an SAE where the event truly fits this definition (see regulatory definition above) and NOT for hospitalizations associated with non serious reasons for hospitalization or prolongation of hospitalization. If a patient is hospitalized or hospitalization is prolonged due to planned procedures such as administration of scheduled IV therapy or minor treatment such as hydration; administration of growth factors; elective surgery; social reasons and respite care; or economic reasons in the absence of any deterioration in the patient's general condition, these hospitalizations do not meet the criteria of seriousness from a medical perspective and do not need to be reported as SAEs. Hospitalizations or prolonged hospitalizations that are required for an SAE but also include a portion of hospitalization for nonmedical reasons should be reported as an SAE. However, AEs that do not meet criteria for Protocol BPI-2358-105 IND All relevant information, either initial or follow-up, has to be reported on the SAE report forms.

For serious and all other AEs, the following must be assessed and recorded on the AE page of the CRF: event, intensity, relationship to test substance, action taken, and outcome to date.

The Investigator must notify the Ethics Review Committee/IRB of such an event in writing as soon as is practical and in accordance with international and local laws and regulations.

Please contact the sponsor or designee to report all SAEs within 24 hours of learning of the SAE.

Please complete an SAE Report Form and scan the form and any supporting documentation (which includes laboratory data, hospital records and the results of relevant tests). The form and supporting documentation must be e-mailed to: ICON-Safety-CentralReceipt@iconplc.com within 24 hours. The preferred method for receiving SAEs is via email. In cases where submission through email is not possible, the site may report the SAEs through the following alternative number: Fax No.: +44 (0)208 100 5005 If an ongoing SAE changes in intensity or causal relationship to the investigational product, or if new information becomes available, a follow-up SAE report should be sent to ICON within 24 hours using the same procedure used for transmitting the initial SAE report. All SAEs should be followed-up until resolution, improvement, or stabilization.

The definition and reporting requirements of ICH Guideline for Clinical Safety Data Management, Definitions and Standards for Expedited reporting, Topic E2 will be adhered to.

A female patient of child-bearing potential must be instructed to immediately inform the Investigator if she becomes pregnant during the study. Pregnancies occurring up to 90 days after final administration of study drug must also be reported to the Investigator. The Investigator should report all pregnancies within 24 hours to the Sponsor. The Investigator should counsel the patient; discuss the risks of continuing with the pregnancy and the possible effects on the fetus. Monitoring of the patient should continue until conclusion of the pregnancy. Patients who Protocol BPI-2358-105 IND become pregnant while on study will be discontinued from the study treatment and all End of Treatment visit procedures will be performed. If the patient is a man who is capable of fathering a child, he must agree to use adequate birth control beginning immediately after he enrolls on this study until 3 months after his last dose of study drug. Pregnancy occurring in the partner of a patient participating in the study should also be immediately reported to the Investigator and the Sponsor. The partner should be counseled and followed as described above.

Pregnancy is not an SAE; however, the outcome of a pregnancy must be reported to detect a potential SAE (congenital anomaly, premature birth, or birth defect). All pregnancies must be initially reported and follow-up information must be reported on the pregnancy follow-up form.

The reporting timeframe to report a pregnancy to the Sponsor is from start of study drug up to 90 days after the last dose of study drug. Procedures and policies at the site and at the Sponsor, regarding pregnancies, will be followed to ensure that the safety and well-being of the study patient and fetus are appropriately followed through the pregnancy to birth. In the event that a pregnancy occurs in the female partner of a male patient, the Investigator will then (and only then) also be required to obtain her consent so that the Sponsor can hold her data on file. If the female partner is unwilling to sign the consent her data may not be held in the safety database. However, this will not affect the ability of the male patient to continue in the study.

An independent data safety monitoring board (DSMB) will be assembled and be governed by a DSMB charter, which will specify membership, frequency of meetings, and potential sample size adjustment. 

All statistical analyses will be performed by the sponsor or designee after the study is completed included in a separate statistical analysis plan (SAP).

Patients in Phase 3 will be stratified according to his or her by tumor type (breast cancer, NSCLC, HRPC) and region (Asia, non-Asia).

Data from all patients receiving the RP3D plinabulin dose in Phase 2 and Phase 3, will not be pooled for assessing the primary and secondary study endpoints, but analyzed separately.

Plinabulin pharmacokinetic (PK) and pharmacodynamic (PD) assessments will be made to enable a PK/PD analysis.

To establish the Recommended Phase 3 Dose (RP3D) based on PK/PD analysis

DSN in treatment Cycle 1 in patients treated with docetaxel (75 mg/m 2 ) + plinabulin (5,10 or 20 mg/ m 2 ) or with docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg). ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose) and during Cycle 1 on Days 1, 2, 6, 7, 8, 9, 10 , and 15 (pre-dose on dosing days; times equivalent to pre-dose on other days). The Negative Binomial Regression (NBR) model, on the integer number of days of severe neutropenia, will be used to analyze the DSN endpoint during the fixed time window outlined above, with the treatment arm as the only covariate. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used. Contrasts for pairwise comparisons between Arm 1 and the plinabulin arms will be used for the evaluation. A closed testing procedure by Hommel (Hommel 1988, Hommel and Bernhard 1999) will be used for the multiple comparisons.

To assess blood pressure semi-continuously with 15-minute intervals, starting 15 minutes pre-plinabulin dose and lasting 4.5 hours after start of infusion with plinabulin (Arms 2 to 4) or for 4.75 hours starting 15 minutes after the end of docetaxel infusion (Arm 1). See the Pharmacometrics Analysis Plan for the analysis methods.

To characterize the pharmacokinetic profile of plinabulin and docetaxel. See the Pharmacometrics Analysis Plan for the analysis methods.

To characterize the exposure-response relationships between measures of plinabulin exposure and the pharmacodynamic endpoint DSN. Robust linear regression (Bablok et al, 1988) with DSN as the dependent variable and plinabulin exposure as the independent Protocol BPI-2358-105 IND variable. The method will also be used to construct point estimates and confidence intervals. The NCSS version 12 statistical software will be used.

To characterize the exposure-safety relationships between measures of plinabulin exposure and safety events of interest. Robust linear regression (Bablok et al, 1988) with the number of safety events as the dependent variable and plinabulin exposure as the independent variable. The method will also be used to construct point estimates and confidence intervals. The NCSS version 12 statistical software will be used.

To assess CD34+ at screening, and on Days 2, 6, and 8 in Cycle 1 and Day 1 in Cycle 2. A repeated measures mixed linear model with the Day 1 value and treatment arm as covariates will be used to analyze this endpoint. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS version 9.4 or later will be used for the analysis. Contrasts for pairwise comparisons between Arm 1 and the plinabulin arms will be used for the evaluation. A closed testing procedure by Hommel (Hommel 1988, Hommel and Bernhard 1999) will be used for the multiple comparisons. Health-related QoL questionnaire evaluated with EORTC QLQ-C30 and EQ-5D-5L. The Wilcoxon Rank Sum test will be used to analyze the responses to the individual questions. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis. Contrasts for pairwise comparisons between Arm 1 and the plinabulin arms will be used for the evaluation. A closed testing procedure by Hommel (Hommel 1988, Hommel and Bernhard 1999) will be used for the multiple comparisons. Data collection on disease progression. The log-rank test (LIFETEST procedure in SAS) will be used to compare time to disease progression between the treatment groups. The method will also be used to construct point estimates and confidence intervals. Contrasts for pairwise comparisons between Arm 1 and the plinabulin arms will be used for the evaluation. A closed testing procedure by Hommel (Hommel 1988, Hommel and Bernhard 1999) will be used for the multiple comparisons. For selected countries only: to investigate the following cytokine panel: IL-1beta, IL-6, IL-12p70, IL-12p40, IL-17A, IL-23, G-CSF, GM-CSF, IFN-alpha, IFN-gamma, TNFalpha, IL-2, FLT-3 ligand, and IL-8.

Incidence, occurrence, and severity of AEs/SAEs. These endpoints will be presented descriptively in tables and listings. Incidences of bone pain. The NBR model will be used to analyze the endpoint during the fixed time window from pre-dose Day 1 through Day 8 in Cycle 1, with the treatment arm as the only covariate. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used. Contrasts for pairwise comparisons between Arm 1 and the plinabulin arms will be used for the evaluation. A closed testing procedure by Hommel (Hommel 1988, Hommel and Bernhard 1999) will be used for the multiple comparisons. Safety and tolerability (physical examination and safety laboratory assessments). These endpoints will be presented descriptively in tables and listings.

Primary Efficacy Endpoint (Cycle 1):

DSN in Cycle 1 in patients treated with docetaxel (75 mg/m 2 ) + plinabulin (40 mg) (Arm 2) compared with patients treated with docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) (Arm 1). ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose) and during Cycle 1 on Days 1, 2, 6, 7, 8, 9, 10, and 15 . Blood draws for ANC will be taken approximately the same time as the time of the pre-dose sample on Day 1, and will be taken by preference in the morning. DSN should be calculated as the number of consecutive days from the first day when a patient's ANC is below 0.5 x l09/L until the patient reaches an ANC> 0.5 x 109/L, in Cycle 1. For patients who do not experience any severe neutropenia in Cycle 1, the DSN is set to 0. For patient's experiencing several episodes, the number of days of DSN will be summed up. The NBR model, on the integer number of days of severe neutropenia, will be used to analyze the DSN endpoint during the fixed time window outlined above, with the treatment arm as the only covariate. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used.

Maximum decrease from baseline (prior to docetaxel dose) in platelet count for each patient in Cycle 1. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

Proportion of patients with NLR > 5 after Day 7 through Day 15 in Cycle 1. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

AUC using the trapezoidal quadrature method for bone pain, from Day 1 through Day 8 in Cycle 1, based on the bone pain score from the patient bone pain scale. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

Estimated mean pain score from the patient bone pain scale from pre-dose Day 1 through Day 8 in Cycle 1. A repeated measures mixed linear model with the pre-dose Day 1 value and treatment arm as covariates will be used to analyze this endpoint. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS version 9.4 or later will be used for the analysis.

Proportion of patients with thrombocytopenia (all grade) in Cycles 1 to 4. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided Protocol BPI-2358-105 IND o Thrombocytopenia (all grade). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o Grade 3 (ANC >= 0.5 and < 1 × 109/L) and Grade 4 neutropenia (ANC < 0.5 × 109/L). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o Grade 4 neutropenia (ANC < 0.5 × 109/L). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o Grade 3 neutropenia (ANC < 1 × 109/L) and ANC >= 0.5 × 109/L). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o Bands > 0 after Day 7 through Day 15. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o Promyelocytes plus myelocytes >0 after Day 7 through Day 15. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o LMR < 3.2 after Day 7 through Day 15; time course of percentage of patients with LMR < 3.2 over time in Cycle 1. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o PLR > 200 after Day 7 through Day 15; time course of percentage of patients with PLR > 200 over time in Cycle 1. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o For the below bone pain measurements, The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis: The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

Time (in days) to first use of bone pain medication. The log-rank test (LIFETEST procedure in SAS) will be used to compare time to first use of bone pain medication between the treatment groups. Separate analyses will be made with respect to narcotic and non-narcotic analgesics.

DSN in Cycle 1 in patients with locally advanced or metastatic NSCLC after platinum failure. The analysis will be done as per the primary endpoint.

DSN in Cycle 1 in patients with HRPC. The analysis will be done as per the primary endpoint.

DSN in Cycle 1 in patients with advanced or metastatic breast cancer who have failed < 5 prior lines of chemotherapy. The analysis will be done as per the primary endpoint.

DSN in Cycle 1 in patients with locally advanced or metastatic NSCLC after platinum failure or HRPC. The analysis will be done as per the primary endpoint.

Platelet count at least 30% change from baseline at any time during Cycle 1. The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

ANC nadir during Cycle 1. The Wilcoxon Rank Sum test will be used to analyze the endpoint. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

For selected countries only: maximum change from baseline in the following cytokine panel: IL-1beta, IL-6, IL-12p70, IL-12p40, IL-17A, IL-23, G-CSF, GM-CSF, IFN-alpha, IFN-gamma, TNF-alpha, IL-2, FLT-3 ligand, and IL-8. The Wilcoxon Rank Sum test will be used to analyze the endpoint. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

Proportion of patients in Cycles 1 to 4 with:

o FN (ANC <1.0 × 10 9 /L AND a single temperature of >38.3°C or a sustained temperature of ≥38°C for more than 1 hour). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o Grade 4 neutropenia (ANC < 0.5 × 10 9 /L). The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and 2-sided confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis Healthcare utilization endpoints:

o Incidence of 30-day rehospitalizations -all cause. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used.

o Incidence of all cause hospitalizations. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used.

o Duration of all cause hospitalizations. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis. o Incidence of all cause ER visits. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used.

o Incidence of all cause ICU stays. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used.

o Duration of all cause ICU stays. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

o Incidence of all cause docetaxel dose delay, dose reduction, or dose discontinuation. The NBR model will be used to analyze the endpoints, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used: The Barnard's test will be used to evaluate the difference in proportions between the treatment arms. The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis o o Incidence of transfusions due to thrombocytopenia. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used.

o Incidence of antibiotics use. The NBR model will be used to analyze the endpoint, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to Protocol BPI-2358-105 IND construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used.

Data collection on disease progression. The log-rank test (LIFETEST procedure in SAS version 9.4 or later) will be used to compare time to disease progression between the treatment groups. The method will also be used to construct point estimates and confidence intervals.

QoL endpoints will be analyzed in safety analysis set: o Health-related QoL questionnaire evaluated with EORTC QLQ-C30 Item 30 (response to question "How do you rate your overall quality of life during the past week") in Cycle 1 to 4 will be analyzed using linear mixed model for repeated measures (MMRM) with the Cycle 1 Day 1 value and treatment arm as covariates. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS v9.4 or later will be used for the analysis. The summary statistics in terms of counts, means, standard deviations, medians, minimums, and maximums for individual questions will be summarized in a tabular format. o Health-related QoL questionnaire evaluated with EORTC QLQ-C30 will also be analyzed with the summary scores. The EORTC QLQ-C30 scoring method combines the 30 questions from the questionnaire into 15 scores: global health state/QoL, functional scale (5 items), and symptoms scale (9 items) (Fayers et al, 2001) . Three summary measures will be constructed: the QLQ-C30 Summary Score, which combined the symptom and functional scales (excluding the financial difficulties item), the symptom summary score (excluding the financial difficulties item), and the functional summary score (Hinz et al, 2012) . The symptom summary score will be analyzed using linear mixed model for repeated measures (MMRM) with the Cycle 1 Day 1 value and treatment arm as covariates. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS v9.4 or later will be used for the analysis. The summary statistics in terms of counts, means, standard deviations, medians, minimums, and maximums for the 15 scores, the QLQ-C30 summary score, symptom score and the functional summary score will be summarized in a tabular format. o The response to the question "Your Health Today" from EQ-5D-5L in Cycle 1 to 4 will be analyzed using linear mixed for model repeated measures (MMRM) with the Cycle 1 Day 1 value and treatment arm as covariates. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS v9.4 or later will be used for the analysis. The summary statistics in terms of counts, means, standard deviations, medians, minimums, and maximums for the question "Your Health Today" will be summarized in a tabular format. o The EQ-5D-5L data will be converted into a health utility score using the interim EQ-5D-5L crosswalk value set for the United States (van Hout et al, 2012) unless a validated EQ-5D-5L value set for the United States becomes Protocol BPI-2358-105 IND available. The EQ-5D-5L health utility score in Cycle 1 to 4 will be analyzed using linear mixed model for repeated measures (MMRM) with the Cycle 1 Day 1 value and treatment arm as covariates. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS v9.4 or later will be used for the analysis. The summary statistics in terms of counts, means, standard deviations, medians, minimums, and maximums for the health utility score will be summarized in a tabular format.

Incidence, occurrence, and severity of AEs/SAEs. These endpoints will be presented descriptively in tables and listings.

Safety and tolerability (physical examination and safety laboratory assessments). These endpoints will be presented descriptively in tables and listings.

The intent-to-treat analysis set for Phase 2 is comprised of all Phase 2 patients that have been randomized in the study and have received at least one dose of study medication.

The analysis of all endpoints, unless noted otherwise, will be conducted on the intent-to-treat analysis set.

The safety analysis set will be the same as the intent-to-treat analysis set for Phase 2.

All patients who received at least 1 dose of plinabulin or docetaxel and had pre-dose sample collection and at least 1 PK sample collected postdose will be included in the PK analysis set. These subjects will be evaluated for PK unless significant protocol deviations affect the data analysis or if key dosing, dosing interruption, or sampling information is missing. Patients in Phase 3 will follow the plinabulin and docetaxel PK sampling schedules from Phase 2.

All patients who had blood pressure and DSN collected at any time during the study will be included in the PD analysis set. For phase 3, PD data may be collected with a schedule of collection to be confirmed based on the emerging data to be determined. Exploratory PK/PD and exposure-response analyses will be conducted to evaluate the effects of plinabulin on safety and efficacy endpoints. Details of these analyses will be summarized in the statistical analysis plan, and may be reported outside of the main clinical study report. 

The intent-to-treat (ITT) analysis set for Phase 3 is comprised of all Phase 3 patients that have been randomized in the study.

The analysis of all endpoints, unless noted otherwise, will be conducted on the intent-to-treat analysis set.

The safety analysis set is comprised of all Phase 3 patients that have been randomized in the study and have received at least one dose of study medication.

Health economic and patient reported outcomes will be conducted on the safety analysis set, which is the same as the intent-to-treat analysis set for Phase 3.

Descriptive summaries will be generated to describe the disposition of all enrolled patients.

Demographic and other baseline characteristics will be described, but no hypothesis testing will be done.

Concomitant medications will be assigned an 11-digit code using the World Health Organization (WHO) Drug dictionary codes.

Data will be tabulated by treatment group, with data listings provided for all data captured in the eCRF as well as laboratory data. On treatment data, will be assessed descriptively as both observed values and as changes from pretreatment. When tabulated, data will be presented using descriptive statistics (e.g., mean, median, standard deviation, and range for continuously scaled parameters, and as number and percent for categorically scaled parameters). Statistical Analysis System Version 9.4 or higher will be used to perform the majority of the analyses; other software (e.g., NCSS) may be utilized to generate graphics or perform other analysis. A detailed statistical analysis plan (SAP) will be written and approved before unblinding the treatment allocation codes. Analyses will be performed based on observed data, and missing values will not be imputed unless otherwise stated in the SAP.

Primary efficacy endpoint will be imputed by multiple imputation method as sensitivity analysis .

Missing pain score data: In the case in which analgesic medication was taken during the treatment period, as sensitivity analyses, the pain scores will be imputed using the last . The study will assume that the shape of the time/neutrophil recovery curve in plinabulin-treated patients is indistinguishable from the time/neutrophil recovery curve for filgrastim and its biosimilars.

In a study with filgrastim and its biosimilar, time course of ANC in Cycle 1 for the Per Protocol dataset was published by . Mean values and standard deviations of ANC during the 21-day follow-up period were readily available. This information was used to write a computer simulation program that would generate random ANC data that asymptotically has the same means and standard deviations for the 21-day follow-up period as the publication. The simulation would then also generate the projected number of days with severe neutropenia (i.e., the DSN).

Deming regression was used to calculate the linear relationship between simulated nadir and DSN. The rank correlation between simulated nadir and DSN was used to calculate the DSN with plinabulin (+ docetaxel) and docetaxel alone. In the Phase 2 study, ANCs were obtained on Day 8, which approximately coincides with the time that the neutrophil nadir occurs after docetaxel administration. These observed Day 8 neutrophil (nadir) values were computed into the linear relationship (Deming regression), mentioned above to calculate DSN for each patient. Using these methods, calculated mean DSN was 0.065 days for the plinabulin+ docetaxel arm, and 1.076 days for the docetaxel alone. Based on published data with filgrastim in patients receiving docetaxel (Alexopoulos K et al, 1999) , the assumption is that Grade 4 neutropenia in Cycle 1 would occur in a 2 times higher frequency with G-CSF+docetaxel versus plinabulin+docetaxel, resulting in a presumed mean DSN of 0.13 days for the G-CSF+ docetaxel combination.

This non-inferiority trial design will utilize a difference (arm 2 minus arm 1) of 0.65 days (non-inferiority margin) in DSN in Cycle 1 as the largest acceptable difference between plinabulin and pegfilgrastim. The non-inferiority test will evaluate the null hypothesis H0: true difference (arm 2 minus arm 1) ≥ 0.65 against the alternative hypothesis H1: true difference (arm 2 minus arm 1) < 0.65. Plinabulin will be considered non-inferior to pegfilgrastim if in Cycle 1, the upper limit of the 2-sided 95% confidence interval for the true difference in mean duration of Grade 4 neutropenia was < 0.65 days. A sample size of patients was based on sample size considerations as outlined. Data suggest (http://www.neulastahcp.com/risk/duration-of-severe-neutropenia/) that FN is correlated with DSN. The frequency of FN with docetaxel monotherapy (100 mg/m 2 ) + G-CSF was reported to be 1% in cycle 1. FN frequency in Cycle 1 with docetaxel combined with doxorubicin and G-CSF was ~ 3 % (Aarts M et al, 2013) , which would translate into a DSN of 1 day according to . Based on this data, it is assumed that the median DSN for docetaxel monotherapy + G-CSF will be approximately 1 day.

The frequency of FN with docetaxel monotherapy (without G-CSF) has been reported to be 11% in cycle 1 (17% over all cycles) docetaxel dose of 100 mg/m 2 ) and 19.8% over all cycles at a lower docetaxel dose of 60 mg/m 2 between FN and DSN, we make the assumption that, with docetaxel monotherapy at a dose of 75 mg/m 2 without G-CSF, the median DSN is estimated to be 4-5 days.

In the Zarxio® briefing document, 2015, the margin was selected based on the fact that TAC chemotherapy is known to induce a median DSN of 7 days in breast cancer patients receiving no G-CSF treatment , while G-CSF treatment reduces the mean DSN for this chemotherapy to 1.4 days (95% CI: 1.07 -1.69) as shown in pegfilgrastim (Neulasta ® ) Study 20020778 (Kaufman et al, 2004) . Based on this a non-inferiority limit of 1 day was derived.

As an extension of this reasoning, it is argued for our study, a non-inferiority margin of 0.65 would be reasonable and correspond to approximately a median of 4.5 days of DSN, as a margin of 1 day to 7 days of DSN in the Zarxio® briefing document, 2015.

A non-inferiority margin of 0.65 days can also be justified, because a difference of 0.65 days is not considered to be clinically meaningful.

Plasma plinabulin and docetaxel concentrations will be measured using validated methods and PK parameters will be summarized using descriptive statistics. Individual and mean serum plinabulin and docetaxel concentration versus time profiles will be plotted on both linear and logarithmic scales.

Plinabulin PK will be characterized using the population PK approach. Population PK parameter estimates, and individual PK parameters, including AUC and Cmax will be summarized.

A sequential population PK/PD modeling approach will be used to characterize the exposure-neutropenia relationship. The semi-physiological model described by Friberg (2003) will be used to characterize the time course of neutropenia in the exposure-neutropenia model.

The developed exposure-response models will be used to simulate ANC profiles for the computation of key PD endpoints such as DSN, DMSN, and area over the ANC curve (AOC). The latter will be calculated as the area below the threshold of <0.5 x 10 9 cells/L and above ANC-time response curve in a chemotherapy cycle. Details will be included in the Pharmacometrics Analysis Plan (PAP). A sequential PK-PD model analysis will be performed. Since the ambulatory blood pressure measurements (ABPM) are subject to circadian variations, ABPM will be modeled to describe the circadian rhythm in blood pressure over the 4.5-hour observation period. The diurnal rhythm in the AMBP data will be described with the cosine function (Hempel 1998) in the exposure-ABPM models. Details will be included in the PAP.

Either a direct or sequential PK-PD model analysis will be performed. The choice of the modeling approach will be informed by the matching or non-matching of PK sampling times with ECG observation times. QT measurements corrected with the Fridericia formula will be used in the exposure-QTc analysis. Details will be included in the PAP.

Medical history and AE data will be coded by system organ class (SOC) and preferred term (PT), using the Medical Dictionary for Regulatory Activities (MedDRA) dictionary. Concomitant medication data will be coded by drug class and indication, using the WHO Drug dictionary.

All treatment emergent AEs will be graded according to NCI CTCAE version 4.03, and grouped by the MedDRA Preferred Term and System Organ Class, and summarized by worst grade severity per patient.

Treatment emergent AEs are those events that occur after first administration of any study therapy through 30 days post last dose of any study therapy, and/or any treatment-related AEs, regardless of the onset date. Dose delays, dose modifications and/or dose discontinuation of docetaxel due to safety concerns will be summarized for the 2 treatment groups.

Continuous variables and proportions will be analyzed using exact t-tests. Other categorical data will be analyzed using non-parametric statistical methods.

Treatment emergent deaths are those deaths within 30 days of last dose of any study therapy. Early deaths are those deaths within 60 days of the first dose of study therapy.

Treatment emergent and/or early deaths will be tabulated and summarized by treatment groups.

Refer to the SAP for details.

The AE verbatim descriptions (investigator terms from the CRF) will be classified into standardized medical terminology using MedDRA. AEs will be coded to MedDRA (Version 18.1 or higher) lower level term closest to the verbatim term. The linked MedDRA preferred term (PT) and primary system organ class (SOC) are also captured in the database. 

Descriptive summary statistics for laboratory values will be provide as well as the shift table..

Descriptive statistics tables will be used for the evaluations.

Electrocardiogram assessments will be provided in a listing.

In the Phase 2 (approximately 40 patients; approximately 10 patients per arm), patients with advanced or metastatic NSCLC will be enrolled.

Approximately 150 patients are planned to be enrolled with 1 of the following diagnoses: advanced or metastatic breast cancer, NSCLC, or HRPC. A sample size of 75 patients in each of the treatment arms docetaxel + plinabulin (40 mg) versus docetaxel + pegfilgrastim, with matching placebos achieve at least a 90% power to reject the null hypothesis of 0.65 day of inferiority in DSN between the treatment means with standard deviations of 0.75, at a significance level (alpha) of 0.05 two-sided two-sample zero-inflated Poisson model.

The software PASS version 15.0.1 has been used for the calculations referencing 14.5.

The study design is group sequential with 1 interim analysis (after 50 patients in each treatment arm have completed at least 1 cycle in each of the treatment arms docetaxel + plinabulin [40 mg] versus docetaxel + pegfilgrastim, with matching placebos) and 1 final analysis at the completion of the study (Table 15 ). These results assume that 2 sequential tests are made using the O'Brien-Fleming spending function to determine the test boundaries.

If non-inferiority is determined from the statistical testing, then also the hypothesis of superiority will be tested, and if it is concluded that the plinabulin treatment is superior to the pegfilgrastim treatment, with respect to DSN, then the study will be stopped. Since the design allows for stopping for inferiority, it might be decided at that occasion that the study will be stopped. Since this is a hierarchical testing procedure no penalty with respect to overall significance will be paid.

The statistical testing will be performed and will be reviewed by an independent DSMB at the interim analysis. Arm 1: Docetaxel + Pegfilgrastim 6 mg 10

Arm 2: Docetaxel + Plinabulin 20 mg/m 2 10

Arm 3: Docetaxel + Plinabulin 10 mg/m 2 10

Arm 4: Docetaxel + Plinabulin 5 mg/m 2 1 0 PK/PD Analysis (to determine RP3D)

Arm 1: Docetaxel + Pegfilgrastim 6 mg 50 75

Arm 2: Docetaxel + Plinabulin (40 mg) 50 75

During the phase 2 and phase 3 portions, randomized patients who withdraw before receiving the first dose will be replaced.

An independent DSMB will be utilized in this study and will be comprised of individuals who are not members of the clinical study team. At least 2 independent oncologists (external to the sponsor) will serve on the committee. The objective of the DSMB will be to ensure objective medical and/or statistical review of safety and/or efficacy issues in order to protect the ethical and safety interests of subjects and to protect the scientific validity of the study. See details in Section 14.5.

As indicated in periodic safety updates and notifications are the responsibility of the investigator and not of the sponsor.

The Investigator must assure that patients' anonymity will be maintained and that his or her identities are protected from unauthorized parties. On CRFs or other documents submitted to the Sponsor, patients should not be identified by his or her names, but by an identification code. The Investigator should keep a patient enrollment log showing codes, names and addresses. The Investigator should maintain documents not for submission to the Sponsor (e.g., patients' written consent forms, in strict confidence).

For each patient enrolled, an eCRF must be completed and electronically signed by the Principal Investigator or authorized delegate from the study staff. Once a patient has signed informed consent and any study related procedures are performed, an eCRF must be completed. If a patient is withdrawn from the study because of a treatment-limiting AE, thorough efforts should be made to clearly document the outcome. The Investigator should ensure the accuracy, completeness, legibility, and timeliness of the data reported to the Sponsor in the CRFs and in all required reports.

The investigator or designee (i.e., pharmacist) is responsible for ensuring adequate accountability of all used and unused study drug. This includes acknowledgment of receipt of each shipment of study product (quantity and condition), patient dispensing records and returned or destroyed study product.

The investigator is responsible for ensuring the study is conducted in accordance with the procedures and evaluations described in this protocol.

Protocol modifications to ongoing studies must be made only after consultation between an appropriate representative of the Sponsor and the Investigator. Protocol modifications must be prepared by a representative of the Sponsor and initially reviewed and approved by Chief Medical Officer.

All protocol modifications must be submitted to the appropriate IEC or IRB for information and approval in accordance with local requirements and to Regulatory Agencies if required. Approval must wait before any changes can be implemented, except for changes necessary to eliminate an immediate hazard to study patients, or when the change(s) involves only logistical or administrative aspects of the study (e.g. change in Monitor(s), change of telephone number(s). Protocol BPI-2358-105 IND 

The results of this study may be published or presented at scientific meetings. The Investigators agree to submit all manuscripts or abstracts to the Sponsor before submission. This allows the Sponsor to protect proprietary information and to provide comments based on information from other studies that may not yet be available to the Investigator.

Because this is a multicenter study, individual investigators may not publish the results of the study based on information from his or her sites until the complete study has been published in full (not abstract) form. If a joint publication has not been submitted within 8 months after the study has been completed or terminated at all sites, then individual sites may publish subject to the requirement to submit to the Sponsor before publication.

The Sponsor will prepare a clinical study report upon completion or termination of the study.

Both the Sponsor and the Investigator reserve the right to terminate the study at any time. Should this be necessary, both parties will arrange the procedures on an individual study basis after review and consultation. In terminating the study, the Sponsor and the Investigator will assure that adequate consideration is given to the protection of the patient's interests.

The Investigator must maintain adequate and accurate records to enable the conduct of the study to be fully documented and the study data to be subsequently verified. These documents should be classified into 2 different separate categories 1) Investigator's Study File, and 2) patient clinical source documents.

The Investigator's Study File will contain the Protocol/Amendments, sample CRFs, patient screening and enrollment logs, Independent Ethics Committee/IRB and governmental approval with correspondence, sample informed consent, study-drug records, staff curriculum vitae and authorization forms and other appropriate documents/correspondence, etc.

Patient clinical source documents (usually defined by the project in advance to record key efficacy/safety parameters independent of the CRFs) would include patient hospital/clinic records, physician's and nurse's notes, appointment book, original laboratory reports, ECG, X-ray, pathology and special assessment reports, signed informed consent forms and consultant letters. The Investigator must keep these 2 categories of documents on file for at least 15 years after completion or discontinuation of the study. After that period of time the documents may be destroyed, according to local regulations.

Should the Investigator wish to assign the study records to another party or move the study records to another location, the Sponsor must be notified in advance.

If the Investigator cannot guarantee this archiving requirement at the investigational site for any or all of the documents, special arrangements must be made between the Investigator and the Sponsor to store these in a sealed container(s) outside of the site so that they can be returned sealed to the Investigator in case of a regulatory audit. Where source documents are required for 

The Investigator shall supply the Sponsor on request with any required background data from the study documentation or clinic records. This is particularly important when CRFs are illegible or when errors in data transcription are suspected. In case of special problems and/or governmental queries or requests for audit inspections, it is also necessary to have access to the complete study records, provided that patient confidentiality is protected.

When the clinical study report is completed, the sponsor will provide the major findings of the study to the investigator.

It is understood that the responsible Sponsor's Monitor (or designee) will contact and visit the Investigator regularly and will be allowed, on request, to inspect the various records of the study (CRFs and other pertinent source data) provided that patient confidentiality is maintained in accord with local requirements.

It will be the Monitor's responsibility to inspect the CRFs at regular intervals throughout the study, to verify the adherence to the Protocol and the completeness, consistency and accuracy of the data being entered on the CRFs. The Monitor should have access to laboratory test reports and other patient records needed to verify the entries on the CRF. The Investigator (or his/her deputy) agrees to cooperate with the Monitor to ensure that any problems detected in the course of these monitoring visits are resolved. Refer to Monitoring Plan for further details.

The Investigator should understand that source documents for this study should be made available to appropriately qualified personnel from the Sponsor, or its designees, or to health authority inspectors after appropriate notification. The verification of the CRF data must be by direct inspection of source documents.

The purpose of this statistical analysis plan (SAP) is to describe the planned analyses and reporting for the Phase 3 portion of protocol BPI-2358-105 Protocol Amendment 5, dated 4 December 2018.

This SAP is being written with due consideration of the recommendations outlined in the most recent International Conference on Harmonization (ICH) E9 Guideline entitled Guidance for Industry: Statistical Principles for Clinical Trials and the most recent ICH E3 Guideline, entitled Guidance for Industry: Structure and Content of Clinical Study Reports (CSR).

This SAP describes the data that will be analyzed and the subject characteristics, efficacy, and safety assessments that will be evaluated. This SAP provides details of the specific statistical methods that will be used. The statistical analysis methods presented in this document will supersede the statistical analysis methods described in the clinical protocol. If additional analyses are required to supplement the planned analyses described in this SAP they may be completed and will be identified in the CSR.

To assess DSN in treatment Cycle 1 in patients with advanced or metastatic breast cancer, who have failed < 5 prior lines of chemotherapy; locally advanced or metastatic non-small cell lung cancer (NSCLC) after platinum therapy failure; or hormone refractory (androgen independent) metastatic prostate cancer (HRPC) treated with docetaxel (75 mg/m 2 ) + plinabulin (40 mg) (Arm 2)versus docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) (Arm 1). ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose) and during Cycle 1 on Days 1, 2, 6, 7, 8, 9, 10, and 15 . Blood draws for ANC will be taken at approximately the same time as the time of the pre-dose sample on Day 1 and will be taken by preference in the morning.

Platelet count in Cycle 1: maximum decrease from baseline (prior to Cycle 1 

To evaluate the proportion of patients in Cycles 1 to 4 with:

o Febrile neutropenia (FN) (ANC <1.0 × 10 9 /L AND a single temperature of >38.3°C or a sustained temperature of ≥38°C for more than 1 hour).

o Grade 4 neutropenia (ANC < 0.5 × 10 9 /L) on any of the Days 8 and 15 in Cycle 1 and on Day 8 in Cycles 2 to 4).

To evaluate the following healthcare utilization objectives: 

Incidence, occurrence, and severity of adverse events (AEs)/serious adverse events (SAEs)

Safety and tolerability (physical examination and safety laboratory assessments)

DSN in treatment Cycle 1 in patients treated with docetaxel (75 mg/m 2 ) + plinabulin (40 mg) compared with patients treated with docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg). ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose) and during Cycle 1 on Days 1, 2, 6, 7, 8, 9, 10, and 15 . Blood draws for ANC will be taken approximately the same time as the time of the pre-dose sample on Day 1, and will be taken by preference in the morning. The NBR model, on the integer number of days of severe neutropenia, will be used to analyze the DSN endpoint during the fixed time window outlined above, with the treatment arm as the only covariate. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used.

Maximum decrease from baseline (prior to docetaxel dose) in platelet count for each patient in Cycle 1. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

Proportion of patients with NLR > 5 after Day 7 through Day 15 in Cycle 1. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms . The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

AUC using the trapezoidal quadrature method for bone pain, from Day 1 through Day 8 in Cycle 1, based on the bone pain score from the patient bone pain scale. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis. Estimated mean pain score from the patient bone pain scale from pre-dose Day 1 through Day 8 in Cycle 1. A repeated measure mixed linear model with the pre-dose Day 1 value and treatment arm as covariates will be used to analyze this endpoint. The method will also be used to construct point estimates and confidence intervals. The MIXED procedure in SAS version 9.4 or later will be used for the analysis along with t-test at each day.

Proportion of patients with thrombocytopenia (all grade) in Cycles 1 to 4. . The Wilson score method will be used to evaluate the difference in proportions between the treatment arms . The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

Proportion of patients in Cycle 1 with:

o Thrombocytopenia (all grade). . The Wilson score method will be used to evaluate the difference in proportions between the treatment arms . The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

Protocol BPI-2358-105

Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

o Grade 3 (ANC < 1 × 10 9 /L) and Grade 4 neutropenia (ANC < 0.5 × 10 9 /L). The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .. The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o Grade 4 neutropenia (ANC < 0.5 × 10 9 /L). The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o Grade 3 neutropenia (ANC < 1 × 10 9 /L). The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o Bands > 0 after Day 7 through Day 15. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .. The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o Promyelocytes plus myelocytes >0 after Day 7 through Day 15. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o LMR < 3.2 after Day 7 through Day 15; time course of percentage of patients with LMR < o 3.2 over time in Cycle 1. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o PLR > 200 after Day 7 through Day 15; time course of percentage of patients with PLR > 200 over time in Cycle 1. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis.

o For the below bone pain measurements, the Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis: BeyondSpring Pharmaceuticals, Inc.

Proportion of patients in Cycle 1 who needed bone pain medication (defined as any medication reported on the pain medication assessment from Day 1 through Day 8) . The Wilson score method will be used to evaluate the difference in proportions between the treatment ) of patients who needed bone pain medication (defined as any medication reported on the pain medication assessment from Day 1 through Day 8 in Cycle 1).

Time (in days) to first use of bone pain medication. The log-rank test (LIFETEST procedure in SAS) will be used to compare time to first use of bone pain medication between the treatment groups. Separate analyses will be made with respect to narcotic and nonnarcotic analgesics.

DSN in Cycle 1 in patients with locally advanced or metastatic NSCLC after platinum failure. The analysis will be done as per the primary endpoint.

DSN in Cycle 1 in patients with HRPC. The analysis will be done as per the primary endpoint.

DSN in Cycle 1 in patients with advanced or metastatic breast cancer who have failed < 5 prior lines of chemotherapy. The analysis will be done as per the primary endpoint.

DSN in Cycle 1 in patients with locally advanced or metastatic NSCLC after platinum failure or HRPC. The analysis will be done as per the primary endpoint.

Platelet count at least 30% change from baseline at any time during Cycle 1. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms ).method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

ANC nadir during Cycle 1. The Wilcoxon Rank Sum test will be used to analyze the endpoint. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

For selected countries only: maximum change from baseline in the following cytokine panel: IL-1beta, IL-6, IL-12p70, IL-12p40, IL-17A, IL-23, G-CSF, GM-CSF, IFN-alpha, IFNgamma, TNF-alpha, IL-2, FLT-3 ligand, and IL-8. The Wilcoxon Rank Sum test will be used to analyze the endpoint. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

Proportion of patients in Cycles 1 to 4 with:

FN (ANC <1.0 × 10 9 /L AND a single temperature of >38.3°C or a sustained temperature of ≥38°C for more than 1 hour). The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis Grade 4 neutropenia (ANC < 0.5 × 10 9 /L). The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis Healthcare utilization endpoints:

o Incidence of 30-day rehospitalizations -all cause. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms Protocol BPI-2358-105 Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 13 of 39 .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis o Incidence of all cause hospitalizations. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis o Duration of all cause hospitalizations. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

o Incidence of all cause ER visits. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis o Incidence of all cause ICU stays. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis o Duration of all cause ICU stays. The Wilcoxon Rank Sum test will be used. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis.

o Incidence of all cause docetaxel dose delay, dose reduction, or dose discontinuation. The NBR model will be used to analyze the endpoints, with the treatment arm as the only covariate. An offset variable will be used to account for exposure time. The method will also be used to construct point estimates and confidence intervals. The GENMOD procedure in SAS version 9.4 or later will be used:

o Relative dose intensity (RDI) ≤ 85% o Docetaxel dose delays > 7 days o Regimen switching o Treatment discontinuation o Incidence of transfusions due to thrombocytopenia. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis o Incidence of antibiotics use. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms .The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis Data collection on disease progression. The log-rank test (LIFETEST procedure in SAS version 9.4 or later) will be used to compare time to disease progression between the treatment groups. The method will also be used to construct point estimates and confidence intervals.

Health-related QoL questionnaire evaluated with EORTC QLQ-C30 and EQ-5D-5L. The Wilcoxon Rank Sum test will be used to analyze the responses to the individual questions. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for the analysis. Protocol BPI-2358-105 Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 14 of 39

Incidence, occurrence, and severity of AEs/SAEs. These endpoints will be presented descriptively in tables and listings. Safety and tolerability (physical examination and safety laboratory assessments). These endpoints will be presented descriptively in tables and listings.

The null and alternative hypotheses for this non-inferiority trial design of duration of severe neutropenia (DSN) are as follows:

H0: True difference ([arm 2: Docetaxel (75 mg/m 2 ) + plinabulin (40 mg) + placebo matching pegfilgrastim] minus arm 1: [Docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) + placebo matching plinabulin]) in the mean DSN 0. 65 days H1: True difference (arm 2 minus arm 1) in the mean DSN < 0.65 days Plinabulin will be considered to be demonstrated to be non-inferior to pegfilgrastim if in Cycle 1, the upper limit of the 2-sided 95% confidence interval for the true difference in mean duration of Grade 4 neutropenia < 0.65 days. If non-inferiority is demonstrated, then the upper limit of the 2-sided 95% confidence interval will be compared to 0. If the upper limit < 0, the superiority of plinabulin to pegfilgrastim will be concluded. Specifics of the statistical tests are provided in Section 14.

This is a multicenter, randomized study with an open label Phase 2 portion and a double-blind Phase 3 portion. This SAP pertains to the Phase 3 portion of the study only.

Phase 3 will not begin until RP3D has been determined based on the phase 2 PK/PD analysis as mentioned above; the RP3D will be the only plinabulin dose administered in Phase 3.

A fixed dose of 40 mg has been selected as the RP3D following the Phase 2 PK/PD analysis (details of dose selection are provided in Section 6.2.3 of protocol amendment 5).

Approximately 150 patients are planned to be enrolled in the Phase 3 with one of the following diagnosis: advanced or metastatic breast cancer, who have failed < 5 prior lines of chemotherapy; advanced or metastatic NSCLC who have previously received platinum-based therapy and have either disease progression, are intolerant of platinum-based therapy or, in the opinion of the investigator, would benefit from docetaxel chemotherapy; or hormone refractory (androgen independent) metastatic prostate cancer. Each eligible patient will be stratified according to his or her tumor type (breast cancer, NSCLC or HRPC) and region (Asia, non-Asia). Patients will be randomly assigned within each stratum (diagnosis) with equal probability (1:1 ratio) or 75:75, with the arm designation and planned intervention as follows:

Arm 1: Docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) + placebo matching plinabulin Arm 2: Docetaxel (75 mg/m 2 ) + plinabulin (40 mg) + placebo matching pegfilgrastim Data from all patients receiving the RP3D plinabulin dose in Phase 2 and Phase 3 will not be pooled for assessing the primary and secondary study endpoints but analyzed separately.

Both Phase 2 and Phase 3, Cycles 1 to 4, will consist of docetaxel 75 mg/m 2 administered by intravenous (IV) infusion on Day 1 over 60 minutes (±5 minutes) every 21 days. In the phase 2 portion, on Day 1 of each cycle, 1.5 hours (± 10 minutes) after the start time of docetaxel infusion (i.e., approximately 30 minutes after the end of docetaxel infusion), patients assigned to a plinabulin arm (arms 2-4) will get a single intravenous infusion of plinabulin at their assigned dose over 30 minutes (± 5 minutes). Thus, the wait time between end of docetaxel infusion and start of the plinabulin infusion is approximately 30 minutes. On Day 2 of each cycle, ≥24 hours Protocol BPI-2358-105 Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 15 of 39 after completing chemotherapy, patients assigned to pegfilgrastim (arm 1) will receive a single dose of pegfilgrastim (6 mg) (subcutaneous injection).

In the phase 3 portion, on Day 1 of each cycle, 1.5 hours (± 10 minutes) after the start time of docetaxel infusion (i.e., approximately 30 minutes after the end of docetaxel infusion), patients will get a single dose of plinabulin or placebo intravenously over 30 minutes (± 5 minutes). On Day 2 of each cycle, ≥24 hours after completing chemotherapy, patients will receive a single dose of pegfilgrastim (6 mg) or placebo (subcutaneous injection).

If a chemotherapy cycle is delayed by more than 3 weeks, the patient will be withdrawn from the study. If a critical AE (refer to Section 10.5 of protocol amendment 5) occurs during the cycle, the dosage of docetaxel may be reduced 20% in the next cycle. Only one docetaxel dose reduction is allowed (refer to Error! Reference source not found. No dose reductions are allowed with plinabulin or pegfilgrastim.

All patients, including patients who withdraw from the study early, will complete a safety follow-up visit 30 days (± 2 days) after the last dose of study drug. If, in the opinion of the investigator, the patient will benefit from more than 4 cycles of docetaxel and open label pegfilgrastim, then the fifth cycle will not start until completion of the safety follow-up visit (in this instance, the safety follow-up visit will be Cycle 4 Day 21). Follow-up visits will be required to monitor for ongoing treatment-related AEs. All patients experiencing drug-related toxicities of Grade 2 at the End of Treatment visit should be followed-up at least monthly until the AE(s) resolves to Grade 1, the event is considered to be chronic, or the patient receives other anti-cancer therapy. The method of follow-up assessment will be at the Investigator's discretion (for example, patient site visit or telephone call). All deaths which occur within 30 days of study drug administration regardless of relationship to the study drug must be reported to the Sponsor immediately and within 24 hours of becoming aware of the event.

Laboratory test results (hematology and serum chemistry) will be collected via a central laboratory. Safety laboratory tests are required prior to treatment on Day 1 of each cycle and can be collected by a local laboratory; however, all other scheduled blood samples as per the schedule assessments and procedure table must also be obtained for central laboratory assessment. Urinalysis will be performed at screening only. CD34+ counts will be established through a fluorescence-activated cell sorting (FACS) method as described in Table 6 and Table 7 of protocol amendment 5.

The schedule of visits and assessments: BeyondSpring Pharmaceuticals, Inc.

Page 18 of 39 visit (in this instance, the EOT visit will be Cycle 4 Day 21). Follow-up visits will be required to monitor for ongoing treatment-related adverse events. All patients experiencing drug-related toxicities of Grade 2 at the End of Treatment visit should be followed-up at least monthly until the adverse event(s) resolves to Grade 1, the event is considered to be chronic or the patient receives other anti-cancer therapy. The method of follow-up assessment will be at the Investigator's discretion (for example, patient site visit or telephone call). All deaths which occur within 30 days of study drug administration regardless of relationship to the study drug must be reported the Sponsor immediately and within 24 hours of becoming aware of the event. c. Demographic data will include gender, date of birth (or age), and race/ethnicity d. Background characteristics will include a history of disease and current disease status, bone marrow involvement, sites of disease, prior anticancer therapies, and prior medications/significant non-drug therapies e. Patients must be in a supine position in a rested and calm state for at least 5 minutes before blood pressure is assessed. If the patient is unable to be in the supine position, the patient should be in the most recumbent position possible. The position selected for a patient, the same arm, and same blood pressure cuff should be kept the same throughout the study.

A standard cuff will be used to measure blood pressure (heart rate will also be measured), with a ±15 minute window:

Cycle 1: On Day 1 at pre-dose docetaxel, pre-dose plinabulin, 30 min, 1 hour, 2 hours post-infusion with plinabulin or placebo, and on Days 2, 6, 7, 8, 9, 10 , and 15 (once, prior to blood draw). Cycles 2 to 4: On Day 1 at pre-dose docetaxel, pre-dose plinabulin, 30 min,1 hour, 2 hours post-infusion with plinabulin or placebo and on Day 8 f. Temperature is to be taken every time a blood draw is taken for ANC. The same temperature location selected for a patient (ear, oral, or axillary) should be used throughout the study for each patient. g. Physical examination will include height (cm) at screening. h. A single 12-lead ECG will be performed at screening, EOT, and 30 Day Safety Follow-Up. All other ECGs will be performed in triplicate. In Cycle 1 Day 1, ECG will be collected before docetaxel infusion, immediately before plinabulin/placebo infusion, 5-minutes before end of plinabulin/placebo infusion, 60 minutes and 4.5 hours after start of infusion with plinabulin/placebo. In Cycle 1 Day 2, ECG will be performed in triplicate prior to pegfilgrastim/placebo injection. i.

Laboratory test samples (hematology and serum chemistry) will be collected and sent to the protocol central laboratory. Safety laboratory tests are required prior to treatment on Day 1 of each cycle and can be collected by a local laboratory and will be used to determine docetaxel dosing; however, all other safety (e.g. protocol specified) blood samples as per the schedule assessments and procedure table must also be obtained for central laboratory assessment. In addition a central laboratory blood draw needs to be taken on the day of dosing on Day 1 of each cycle, prior to the docetaxel dosing. Neutrophils are to be collected on time points as indicated in this schedule; neutrophils must be collected at pre-dose on day 1 of each cycle. During cycle 1, ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose), on Days 1, 2, 6, 7, 8, 9, 10, and 15 . Blood draws for ANC will be taken approximately the same time as the time of the pre-dose sample on Day 1, and will be taken by preference in the morning. If Grade 4 neutropenia is observed in a patient, absolute neutrophil count will be evaluated daily (central and local laboratories) during the first cycle, until Grade 4 neutropenia is resolved.

Analyzed at a central laboratory. k. Hepatitis C antibody, hepatitis B surface antigen (HBsAg), viral load at local laboratory if HBsAg is positive, hepatitis B surface antibody (anti-HBs). Hepatitis serology can be performed using laboratory standard methodology, by local laboratory or by the central laboratory if local testing is not available. l.

If HIV testing is done (at the discretion of the investigator), the test will be conducted at the local site. m. Pregnancy tests will be done using urine samples in women of childbearing potential. Subject must have a negative urine pregnancy test documented within the 24-hour period prior to the first infusion (either on Day -1 or prior to infusion on Day 1). Confirm with serum testing (local or central laboratory) if urine sample is positive. Additional testing may be performed at the discretion of the investigator. n. Patients can be randomized on Cycle 1 Day -1 or Cycle 1 Day 1, to allow site flexibility to dispense premedication treatment. o. Plasma samples (5 mL each) for plinabulin and docetaxel PK. All patients will be sampled for PK via a central laboratory. Patients in Phase 3 will follow the plinabulin and docetaxel PK sampling schedules from Phase 2. For PK collection schedule refer to Table 13 and Table 14 (see Section 11.9 of protocol Protocol BPI-2358-105 Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 19 of 39 amendment 5). Cytokine panel testing will be done using unused plasma samples collected at the PK time points; reconsenting of the patients will be required before analysis is performed. p. Investigator opinion of progression (yes/no) at End of Treatment (EOT) recorded in CRF. For example if the patient completes two cycles of docetaxel and study drug, and in the opinion of the investigator per institutional practice the cancer is growing and a new treatment is required, then the EOT evaluation will be performed as specified, and the "disease progression" will be scored as "yes." As an another example, if after four cycles of docetaxel and study drug, the cancer is stable or responding, and the patient receives further docetaxel, then the EOT evaluation will be completed as specified, and the "disease progression" will be scored as "no." Patients without progressive disease at the EOT visit will undergo a follow-up visit every 2 months until the occurrence of either disease progression or death. These visits may be conducted per telephone or other means.

Protocol BPI-2358-105

Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 20 of 39 q. Docetaxel infusion: 75 mg/m 2 docetaxel will be administered via IV infusion over 1 hour on Day 1 of each cycle. All patients must be premedicated with oral corticosteroids (see below for prostate cancer) such as dexamethasone 16 mg per day (e.g., 8 mg bid) for 3 days starting 1 day prior to docetaxel administration in order to reduce the incidence and severity of fluid retention as well as the severity of hypersensitivity reactions (refer to Taxotere ® Package Insert). For hormonerefractory metastatic prostate cancer, given the concurrent use of prednisone, the recommended premedication regimen is oral dexamethasone 8 mg, at 12 hours, 3 hours and 1 hour before the docetaxel infusion (refer to Taxotere ® Package Insert). r. Cycles 1 to 4 will consist of docetaxel 75 mg/m 2 administered by IV on Day 1 over 60 minutes (±5 minutes) each 21 day cycle. Patients will get a single dose of plinabulin or placebo intravenously over 30 minutes (± 5 minutes) 30 minutes after the end of the docetaxel infusion. On Day 2 of each cycle ≥ 24 hours after completing chemotherapy, patients will receive a single dose of pegfilgrastim (6 mg) or placebo (subcutaneous injection). s. The bone pain scale and pain medication assessment should be completed on Day 1 prior to docetaxel infusion and on each remaining day through Day 8 at approximately the same time (Moore et al, 2017) . On Days 2, 6, 7, and 8 the patient will complete the bone pain scale and pain medication assessment at the investigational site t. Bone pain assessment to be completed pre-dose. u. On Day 2 the patient will be provided with a further copy of the assessment to be completed at home on Day 3, Day 4, and Day 5 and returned to the study center at the next visit (Day 6). v. Health-related QoL questionnaire evaluated with EORTC QLQ-C30 and EQ 5D-5L will be collected on Day 1 (prior to the infusion) and on Day 8 of each cycle (Giesinger et al, 2014). w. All concomitant medicines (dose, schedule, and duration of treatment) and in particular analgesics as well as antibiotics should be entered in the eCRF. x. All hospitalizations should be entered in the eCRF. Adverse events will be collected at all study visits starting with the first dose of study drug. Serious adverse events are collected from the time of signing the informed consent form and up to 30 days following the last dose of the study drug.

A planned 75 patients per arm with advanced or metastatic breast cancer, who have failed < 5 prior lines of chemotherapy; advanced or metastatic NSCLC who have previously received platinum based therapy and have either disease progression, are intolerant of platinum-based therapy, or in the opinion of the investigator, would benefit from docetaxel chemotherapy; or hormone refractory [androgen independent] metastatic prostate cancer):

Arm 1: Docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg) + placebo matching plinabulin Arm 2: Docetaxel (75 mg/m 2 ) + plinabulin (40 mg) + placebo matching pegfilgrastim

All patients with HRPC also are given prednisone 5 mg orally twice daily continuously in addition to docetaxel (see Taxotere® (Prescribing Information) as well as other assigned study drugs.In the phase 3 portion, on Day 1 of each cycle, 1.5 hours (± 10 minutes) after the start time of docetaxel infusion (i.e., approximately 30 minutes after the end of docetaxel infusion), patients will get a single dose of plinabulin or placebo intravenously over 30 minutes ( 5 minutes). On Day 2 of each cycle, 24 hours after completing chemotherapy, patients will receive a single dose of pegfilgrastim (6 mg) or placebo (subcutaneous injection).

Screened subjects who sign an informed consent form (ICF) will be assigned a patient number that will be entered on the electronic case report form (eCRF). Stratified randomization will be used to avoid bias in the assignment of subjects to treatment, to increase the likelihood that known and unknown subject attributes (eg, demographics and baseline characteristics) are evenly balanced across treatment groups, to reduce the possible influence of covariates on the drug evaluation, and to Protocol BPI-2358-105 Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 21 of 39 enhance the validity of statistical comparisons. Once a subject meets all qualification criteria at Cycle 1 Day 1, he/she will be randomly assigned to study treatment, after stratification, according to his or her tumor type (breast cancer, NSCLC, HRPC) and region (Asia, non-Asia) through the use of an interactive web response system (IWRS).

This is a double-blind study. During randomization, patients and all personnel involved with the conduct and interpretation of the study, including Investigators, site personnel, and sponsor staff will be blinded to the treatment codes. Randomization data will be kept strictly confidential, filed securely by an appropriate group with the sponsor or contract research organization (CRO) and accessible only to authorized persons.

A master list of all treatments and the patient numbers associated with the treatments will be maintained in a sealed envelope by the clinical supply vendor, the IWRS vendor, and the sponsor. In the event that emergency conditions require knowledge of the study treatment given, the blind may be broken via the code breaker facility within the IWRS. Emergency procedures for revealing drug codes are provided in the study manual. If possible, before breaking the blind, the Investigator should consult with the sponsor to ascertain the necessity of breaking the code.

The overall randomization code will be broken only for final reporting purposes. This will occur once all final clinical data have been entered into the database, data queries have been resolved, and assignment of subjects to the analysis populations has been completed. The study team and database will remain blinded until the final database lock.

Approximately 150 patients are planned to be enrolled with 1 of the following diagnoses: advanced or metastatic breast cancer, NSCLC, or HRPC. A sample size of 75 patients in each of the treatment arms: docetaxel + plinabulin (40 mg) versus docetaxel + pegfilgrastim, with matching placebos achieve at least a 90% power to reject the null hypothesis of 0.65 day of inferiority in DSN between the treatment means with standard deviation of 0.75, at an overall two-sided significance level (alpha) of 0.05, using a two-sample zero-inflated Poisson model and an O'Brien-Fleming spending function to account for the interim analysis at 2/3rds (66.7%) of information.

The software PASS version 15.0.1 has been used for the calculations referencing .

Arm 1: Docetaxel + Pegfilgrastim 6 mg 50 75

Arm 2: Docetaxel + Plinabulin (40 mg) 50 75

Data management procedures, including database design, selection of the data dictionary, and coding of all adverse events and medications, will be performed by Statistics & Data Corporation (SDC). All reported study data will be recorded on the eCRFs supplied by SDC using Medidata Rave ® . Clinical personnel at the study center and CRO are responsible for ensuring that the protocol is followed and that the eCRFs are properly completed.

After data are entered into the clinical study database, electronic edit checks will be performed, including checks for missing data, out of range values, discrepancies within and across visits, and cross checks between different data tables. All data validation specifications and procedures are detailed in the Data Validation Manual (DVM), and manual data checks are documented in the Study Report Specifications (SRS). When the database has been declared to be complete and accurate, the Protocol BPI-2358-105 Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 22 of 39 database will be locked and treatment codes unmasked. Any changes to the database after that time can only be made with the approval of the Sponsor in consultation with SDC.

All final analyses outlined in this document will be performed after:

All data management requirements are met according to SDC's Standard Operating Procedures (SOP), including performance of edit and validation checks, documentation and resolution of data queries, and database lock with written authorization provided by appropriate SDC and Sponsor personnel; All protocol deviations have been classified as major or minor and the per protocol (PP) population has been determined; and The treatment codes have been unmasked.

The intent-to-treat (ITT) analysis set for Phase 3 is comprised of all Phase 3 patients that have been randomized in the study and have received at least one dose of study medication.

The analysis of all efficacy endpoints, unless noted otherwise, will be conducted on the intent-totreat analysis set.

The safety analysis set will be the same as the intent-to-treat analysis set for Phase 3.

Health economic and patient reported outcomes will be conducted on the safety analysis set, which is the same as the intent-to-treat analysis set for phase 3.

DSN was not measured in the previous Phase 2 study ([Study NPI-2358-101] . The study will assume that the shape of the time/neutrophil recovery curve in plinabulin-treated patients is indistinguishable from the time/neutrophil recovery curve for filgrastim and its biosimilars. In a study with filgrastim and its biosimilar, time course of ANC in Cycle 1 for the Per Protocol dataset was published by . Mean values and standard deviations of ANC during the 21-day follow-up period were readily available. This information was used to write a computer simulation program that would generate random ANC data that asymptotically has the same means and standard deviations for the 21-day follow-up period as the publication. The simulation would then also generate the projected number of days with severe neutropenia, (i.e., the DSN). Deming regression This non-inferiority trial design will utilize a difference (arm 2 minus arm 1) of 0.65 days (noninferiority margin) in DSN in Cycle 1 as the largest acceptable difference between plinabulin and pegfilgrastim. The non-inferiority test will evaluate the null hypothesis H0: true difference (arm 2 minus arm 1) ≥ 0.65 against the alternative hypothesis H1: true difference (arm 2 minus arm 1) < 0.65. Plinabulin will be considered non-inferior to pegfilgrastim if in Cycle 1, the upper limit of the 2-sided 95% confidence interval for the true difference in mean duration of Grade 4 neutropenia was < 0.65 days. A sample size of patients was based on sample size considerations as outlined. Data suggest (http://www.neulastahcp.com/risk/duration-of-severe-neutropenia/) that FN is correlated with DSN. The frequency of FN with docetaxel monotherapy (100 mg/m2) + G-CSF was reported to be 1% in Cycle 1. FN frequency in Cycle 1 with docetaxel combined with doxorubicin and G-CSF was ~ 3 % (Aarts M, et al. 2013) , which would translate into a DSN of 1 day according to . Based on these data, it is assumed that the median DSN for docetaxel monotherapy + G-CSF will be approximately 1 day. The frequency of FN with docetaxel monotherapy (without G-CSF) has been reported to be 11% in Cycle 1 (17% over all cycles) docetaxel dose of 100 mg/m 2 ) and 19.8% over all cycles at a lower docetaxel dose of 60 mg/m 2 . Hanna N et al, 2004 reported an FN percentage of 12.7% with 75 mg/m 2 docetaxel. Based on this range of FN, the relationship established by between FN and DSN, we make the assumption that, with docetaxel monotherapy at a dose of 75 mg/m 2 without G-CSF, the median DSN is estimated to be 4-5 days.

In the Zarxio® briefing document, 2015, the margin was selected based on the fact that Taxotere/Adriamycin/cyclophosphamide (TAC) chemotherapy is known to induce a median DSN of 7 days in breast cancer patients receiving no G-CSF treatment , while G-CSF treatment reduces the mean DSN for this chemotherapy to 1.4 days (95% CI: 1.07 -1.69) as shown in pegfilgrastim (Neulasta®) Study 20020778 (Kaufman et al, 2004) . Based on this a noninferiority limit of 1 day was derived.

As an extension of this reasoning, it is argued for our study, a non-inferiority margin of 0.65 would be reasonable and correspond to approximately a median of 4.5 days of DSN, as a ratio of 1 day to 7 days of DSN in the Zarxio® briefing document, 2015.

A non-inferiority margin of 0.65 days can also be justified, because a difference of 0.65 days is not considered to be clinically meaningful.

Baseline measures are defined as the last non-missing measure prior to receipt of randomized study treatment.

All data analyses at the end of the trial will be performed by SDC. Statistical analyses will be reported using summary tables, figures, and data listings. Continuous variables will be summarized with counts, means, standard deviations, medians, minimums, and maximums. Categorical variables will be summarized by counts and by percentage of patients. Confidence intervals will be presented in the summary for continuous and/or categorical variables as deemed appropriate.

Formal inferential statistical analyses techniques will be discussed in subsequent sections of this SAP.

Individual patient data obtained from the case report forms (CRF), electrocardiogram (ECG), core laboratory, PK data and any derived data will be presented in by-patient listings sorted by study center and patient number.

Protocol BPI-2358-105

Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

All analyses and tabulations will be performed using SAS Version 9.4 or later. Tables, listings, and figures will be presented in Portable Document Format (PDF), and tables will be presented in Rich Text Format (RTF) as well. Upon completion, all SAS programs will be validated by an independent programmer. In addition, all program output will undergo a statistical review, a quality control specialist review and a senior level statistical review. The validation process will be used to confirm that statistically valid methods have been implemented and that all data manipulations and calculations are accurate. Checks will be made to ensure accuracy, consistency with this plan, consistency within tables, and consistency between tables and corresponding data listings. Upon completion of validation and quality review procedures, all documentation will be collected and filed by the project statistician or designee.

Missing data will not be imputed for the primary and key secondary endpoint analyses but will be used for sensitivity analyses. Sensitivity analyses will be performed to examine the potential impact of the missing data, by assigning worst case, best case and average case scenarios. Repeated measures mixed-effects models, weighted estimating equations and multiple-imputation models will also be used to determine the potential impact of missing data.

Missing pain score data: In the case in which analgesic medication was taken during the treatment period, as sensitivity analyses, the pain scores will be imputed using the last observation carried forward (LOCF), Worst Observation Carried Forward (WOCF), and Baseline Observation Carried Forward (BOCF) methods.

Adjustment of pain intensity data for bone pain medication usage: When bone pain medication is requested by the patient prior to receiving the bone pain medication a "pre-bone pain medication pain intensity score" will be obtained. This score will then be carried forward in the planned analyses, and the patient will continue to record their pain intensity over the rest of a pre-specified window of sampling time that is based on the PK of the bone pain medication. Pre-bone pain medication pain intensity scores are carried forward in the analysis record. The impact of multiple uses of bone pain medication and multiple pre-bone pain medication pain intensity scores will be included in the analysis records.

The study has two components to multiplicity, the first component is an interim analysis. The study design is group sequential with one interim analysis (after 50 patients in each treatment arm have completed at least one cycle in each of the treatment arms docetaxel + plinabulin [40 mg] versus docetaxel + pegfilgrastim, with matching placebos) and one final analysis at the completion of the study. These results assume that two sequential tests are made using the O'Brien-Fleming spending function to determine the test boundaries.

The second multiplicity component of the study is initially testing the primary endpoint to determine non-inferiority of plinabulin to pegfilgrastim. If non-inferiority is demonstrated, then superiority of plinabulin to pegfilgrastim will be tested. As this is a hierarchical testing procedure, no additional alpha adjustment is necessary.

At the interim analysis, if non-inferiority is demonstrated, then superiority will also be tested. If the plinabulin treatment is concluded to be superior to the pegfilgrastim treatment, with respect to DSN, then the study will be stopped. Additionally, since the design allows for stopping for inferiority, it might be decided at that occasion that the study will be stopped.

Subject disposition will be presented in terms of the numbers and percentages of subjects who were randomized, completed the study, and discontinued from the study. Subjects who are not discontinued from the study will be considered study completers. Disposition will be summarized by treatment group and for all subjects.

The number and percentage of subjects completing the study or who prematurely discontinued from the study and the reasons for study discontinuation will be summarized by treatment group for all Protocol BPI-2358-105 Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 25 of 39 randomized subjects. The reasons for study discontinuation prior to completion include: adverse event, drug related adverse event, physician decision, prohibited concomitant medication or therapy, protocol prohibited dose delay, protocol prohibited dose reduction, study terminated by sponsor, withdrawal by subject, and other. A subject listing will be provided that includes the date of completion or withdrawal, and reason for premature study discontinuation.

The number and percentage of subjects with protocol deviations will be summarized by treatment group for all randomized subjects. A subject listing will be provided that includes the date of the deviation, the deviation description and the classification of whether the deviation was judged to be major or minor.

In addition, subject listings will be provided that include informed consent date, randomization date, randomization stratum, inclusion and exclusion criteria violations.

A summary of age, sex, race, ethnicity and subject's cancer type will be presented using appropriate descriptive statistics by each treatment arm and overall total. The categorical (discrete) variables will be summarized using counts and percentages. The continuous variables will be summarized using counts, mean, median, standard deviation, and range (minimum, maximum). These summaries will include patients in the ITT analysis set.

All demographic and baseline characteristics will be listed by study center and subject number.

Cancer diagnosis, cancer staging, sample collection, bone marrow involvement, and disease progression will be summarized by using frequency counts and percentages for the ITT analysis set.

The child-bearing potential and any other assessments that are done for the purpose of eligibility for inclusion into the study (physical examination, vital signs, hematology and blood chemistry, urinalysis, pregnancy test, and ECG) will also be summarized using appropriate descriptive statistics by each treatment arm and overall total for the ITT analysis set.

Medical history will be summarized using discrete summary statistics and presented by treatment group at the subject and event level by system organ class (SOC) and preferred term (PT). If a subject reports the same PT multiple times within the same SOC, that PT will only be reported once within that SOC. As with the PT, if a subject reports multiple conditions within the same SOC, then that SOC will only be reported once. Medical history will be coded using Medical Dictionary for Regulatory Activities (MedDRA), version 20.1. The summaries will be based on the safety analysis set.

Listings of medical history will also be generated.

Prior cancer-specific surgery, radiotherapy, prior chemotherapy regimen, prior anti-cancer therapy for the investigational disease including drug name, dose, number of cycles completed and reason for therapy discontinuation and best overall response will be summarized using frequency counts and percentage.

All prior and concomitant medications, treatments and surgical procedures will be listed using the safety population including generic name, route of administration, start date, stop date, dosage, indication, related AE or medical history, and reason for treatment. Prior and concomitant medications will be coded using WHO Drug dictionary enhanced (WHO-DDE) + H B2, September 2017, to the appropriate Anatomical Therapeutic Chemical (ATC) classification and WHO generic term. Prior and concomitant treatments and surgical procedures will be coded using MedDRA, version 20.1.

Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Counts and percentages of prior and concomitant medications will be summarized using WHO-DDE ATC classification and preferred name. Summaries will be displayed by treatment group and for all subjects. Subjects with multiple medications in the same ATC class or preferred name will be counted only once for that respective ATC class or preferred name.

Study treatment exposure will be summarized in the safety population.

For each treatment arm for each product (docetaxel, doxorubicin, cyclophosphamide, pegfilgrastim, and plinabulin), the following will be summarized using descriptive statistics by study arm and overall:

Duration of exposure, calculated as (date of last dose -date of first dose) + 1 day.

Number of cycles received per patient.

Number of cycles with dose delay, dose adjustment, dose interruption.

Reasons for dose deviations from planned therapy.

All study drug administration data will be listed by study center and patient number.

The primary efficacy endpoint is DSN in treatment Cycle 1 in patients treated with docetaxel (75 mg/m 2 ) + plinabulin (40 mg) compared with patients treated with docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg). ANC will be assessed at baseline (prior to Cycle 1 docetaxel dose) and during Cycle 1 on Days 1, 2, 6, 7, 8, 9, 10, and 15 . The primary efficacy analysis will evaluate the difference of DSN between the treatment groups of docetaxel (75 mg/m 2 ) + plinabulin (40 mg) and docetaxel (75 mg/m 2 ) + pegfilgrastim (6 mg).

A Negative Binomial Regression (NBR) model on the integer number of days of severe neutropenia will be used to analyze the DSN endpoint during the fixed time window outlined above, with the treatment arm as the only covariate. This method will also be used to construct point estimates and confidence intervals. The NBREG, GLM and TABSTAT procedures in Stata v11.0 or later will be used as following: The primary efficacy analyses will be conducted on the ITT analysis set using observed data.

The secondary efficacy endpoints (Cycle 1) are the following:

Estimated mean pain score from the patient bone pain scale from pre-dose Day 1 through Day 8 in Cycle 1.

Change in bone pain score from pre-dose Day 1 through Day 8 in Cycle 1.

Incidence (as a binary variable) of Grade 4 neutropenia (ANC < 0.5 × 10 9 /L) in Cycle 1.

ANC nadir during Cycle 1.

Protocol BPI-2358-105

Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 27 of 39

The secondary efficacy endpoints for estimated mean pain score and change in bone pain score from pre-dose Day 1 through Day 8 in Cycle 1 will be analyzed by using a repeated measures mixed linear model with the pre-dose Day 1 value and treatment arm as covariates. The MIXED procedure in SAS will be used for these analyses as following: where indata is the name of the input dataset treatment is the treatment group; subjid is the subject identifier avisitn is the visit identifier baseline is the pre-dose Day 1 value; and resp is the response variable, e.g., pain score or change in bone pain score from pre-dose Day 1 through Day 8 in Cycle 1.

In addition, two-sample t-tests were used to test for differences between groups for each day. The Wilson score method will be used to evaluate the difference in proportions between the treatment arms in incidence (as a binary variable) of Grade 4 neutropenia (ANC < 0.5 × 10 9 /L) at Cycle 1. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY and FREQ procedure in SAS will be used for the analysis as following:

PROC FREQ DATA=INDATA; TABLES TREATMENT*OUTCOME / RISKDIFF (CL=(WILSON)); WEIGHT COUNT;

where indata is the name of the input dataset treatment is the treatment group; -outcome is the incidence (as a binary variable) of Grade 4 neutropenia (ANC < 0.5 × 10 9 /L) in Cycle 1; count is the frequency of outcome variable.

The Wilcoxon Rank Sum test will be used to analyze the endpoint for ANC nadir during Cycle 1. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS will be used for the analysis. where indata is the name of the input dataset treatment is the treatment group; outcome is the ANC nadir during Cycle 1;

The secondary efficacy endpoints (Cycle 1 to 4) are the following:

Incidences of FN (ANC <1.0 × 10 9 /L AND a single temperature of >38.3°C or a sustained temperature of ≥38°C for more than 1 hour) in Cycles 1 to 4.

Protocol BPI-2358-105 Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 28 of 39 Incidences of hospitalizations due to FN during Cycles 1 to 4; Duration of hospitalizations due to FN during Cycles 1 to 4;

Health-related QoL questionnaire evaluated with EORTC QLQ-C30 and EQ-5D-5L and EQ-5D-5L;

Incidences of antibiotic use;

Incidences of docetaxel dose delay, dose reduction, and/or dose discontinuation.

The Wilson score method will be used to evaluate the difference in proportions between the treatment arms . The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS version 9.4 or later will be used for the analysis. The variables are: incidences of FN, incidences of documented infections, incidences of hospitalizations due to FN during Cycles 1 to 4 and incidences of antibiotic use. The SAS codes are:

PROC FREQ DATA=INDATA; TABLES TREATMENT*OUTCOME / RISKDIFF (CL=(WILSON)); WEIGHT COUNT; RUN;

where indata is the name of the input dataset treatment is the treatment group; outcome is the ANC nadir during Cycle 1

The Wilcoxon Rank Sum test will be used to analyze the endpoints for duration of hospitalizations due to FN during Cycles 1 to 4 and Health-related QoL questionnaire evaluated with EORTC QLQ-C30 and EQ-5D-5L and EQ-5D-5L. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS will be used for the analysis. where indata is the name of the input dataset treatment is the treatment group; outcome is the duration of hospitalizations due to FN during Cycles 1 to 4 or Health-related QoL questionnaire evaluated with EORTC QLQ-C30 and EQ-5D-5L and EQ-5D-5L; count is the frequency of outcome variable.

The endpoints for incidences of docetaxel dose delay, dose reduction, and/or dose discontinuation will be presented descriptively in tables and listings.

The exploratory endpoints are the following:

AUC using the trapezoidal quadrature method for bone pain, from Day 1 through Day 8 in Cycle 1, based on the bone pain score from the patient bone pain scale;

Proportion of patients who needed bone pain medication (defined as any medication reported on the pain medication assessment from Day 1 through Day 8 in Cycle 1);

Time to first use of bone pain medication;

Incidence (as a binary variable) of Grade 4 neutropenia (ANC < 0.5 × 10 9 /L) on any of the Days 8 and 15 in Cycle 1 and on Day 8 in Cycles 2 to 4;

Protocol BPI-2358-105

Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 29 of 39

Time to disease progression.

The Wilcoxon Rank Sum test will be used to analyze AUC. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS will be used for the analysis.

PROC NPAR1WAY DATA=INDATA WILCOXON; CLASS TREATMENT; VAR OUTCOME; RUN;

where indata is the name of the input dataset treatment is the treatment group; outcome is AUC using the trapezoidal quadrature method for bone pain, from Day 1 through Day 8 in Cycle 1, based on the bone pain score from the patient bone pain scale; count is the frequency of outcome variable.

The Wilson score method will be used to evaluate the difference in proportions between the treatment arms in patients who needed bone pain medication (defined as any medication reported on the pain medication assessment from Day 1 through Day 8 in Cycle 1) and incidence (as a binary variable) of Grade 4 neutropenia (ANC < 0.5 × 10 9 /L) on any of the Days 8 and 15 in Cycle 1 and on Day 8 in Cycles 2 to 4. The method will also be used to construct point estimates and confidence intervals. The FREQ procedure in SAS will be used for the analysis as following:

PROC FREQ DATA=INDATA;

TABLES TREATMENT*OUTCOME/ RISKDIFF (CL=(WILSON)); WEIGHT COUNT; RUN;

where indata is the name of the input dataset treatment is the treatment group; outcome is the incidence of patients who needed bone pain medication (defined as any medication reported on the pain medication assessment from Day 1 through Day 8 in Cycle 1) or incidence (as a binary variable) of Grade 4 neutropenia (ANC < 0.5 × 10 9 /L) on any of the Days 8 and 15 in Cycle 1 and on Day 8 in Cycles 2 to 4; count is the frequency of outcome variable.

The log-rank test will be used to compare time to first use of bone pain medication and time to disease progression between the treatment groups. The LIFETEST procedure in SAS will be used for the analysis as following:

PROC LIFETEST DATA=INDATA; TIME TIME* TATUS(0); STRATA TREATMENT; RUN;

where indata is the name of the input dataset treatment is the treatment group; -Time is the time to first use of bone pain medication or time to disease progression between the treatment groups; -Status is the censoring indicator. A status of 1 indicates an event time, and a status of 0 indicates a censored time.

All patients will be evaluable for safety analysis if they receive at least 1 dose of study drug. The safety data will be presented by study arm in individual listings and summary tables, including frequency Protocol BPI-2358-105 Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 30 of 39 tables for AEs and frequency and shift tables for laboratory variables. All AEs and abnormal laboratory variables will be assessed according to the NCI CTCAE (v 4.03) grading system. Descriptive statistics will be used to summarize ECOG performance status. Vital signs will be reported in listings. AEs and SAEs will be reported in combined tables and SAEs will also be in tabulated in a separate table.

All safety information will be listed by study center and subject number.

An AE is defined as any untoward medical occurrence associated with the use of an investigational product (IP) in humans, whether or not considered IP-related. AEs that occur after the first use of IP are to be grouped as treatment-emergent adverse events (TEAE). AEs with unknown onset dates will be counted as TEAEs. All AEs will be assigned a toxicity grade from grade 1 to grade 5. Their relationship to IP will be classified as suspected (related, probable or possible relationship), or suspected (unlikely, unrelated or N/A -Drug Not Given). TEAEs with a missing (or unknown) relationship to study treatment will assume greatest relationship to study treatment. TEAEs with missing severity grades will be categorized as "missing" for tabulation of TEAEs by severity. Documentation of AEs will include onset date, toxicity grade, action(s) taken, IP relationship, outcome, resolution date, and seriousness. All AEs will be coded using MedDRA classifications with reference to SOCs and PTs (MedDRA version 20.1).

SAEs are collected from the time of signing the informed consent form and up to 30 days following the last dose of the study drug. An AE is considered serious if, in the view of either the investigator or sponsor, it results in any of the following outcomes:

Death; A life-threatening AE; Inpatient hospitalization or prolongation of existing hospitalization; A persistent or significant incapacity or substantial disruption of the ability to conduct normal life functions; A congenital anomaly/birth defect.

Important medical events that may not result in death, be life-threatening, or require hospitalization may be considered serious when, based upon appropriate medical judgment, they may jeopardize the subject and may require medical or surgical intervention to prevent one of the outcomes listed above.

An overall tabular summary of AEs will be presented that includes the number of events and the number and percentage of subjects who experienced at least one event, by treatment group and overall. This summary will also include breakdowns of AEs further categorized as SAEs, AEs by maximal severity, AEs leading to dose delays or discontinuation of the patients from the study and AEs resulting in death. AEs will be summarized using discrete summary statistics and presented by treatment group at the subject and event level by SOC and PT using the Safety population. If a subject reports the same PT multiple times within the same SOC, that PT will only be reported once within that SOC. As with the PT, if a subject reports multiple conditions within the same SOC, that SOC will only be reported once. The occurrence of AEs suspected to be related to IP will also be tabulated by SOC and PT. All AEs will be presented in a listing.

Physical examination will be conducted at screening, Day 1 in all cycles and at the end of treatment (EOT) visit. A comprehensive physical examination will include evaluations of the head, eyes, ears, nose, throat, neck, chest (including heart and lungs), abdomen, limbs, skin, and a complete neurological examination. A urogenital examination will only be required in the presence of clinical symptoms related to this region.

Height will be measured in centimeters once at screening. Weight will be measured in kilograms at screening, prior to Day 1 dosing in all cycles and at the end of treatment (EOT) visit. For Cycle 1, body weight will also be measured on Days 2 and 6. Physical examination will be listed by visit and treatment arms, but no tabular summaries will be produced.

Protocol BPI-2358-105

Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 31 of 39

Vital signs assessments will be conducted at screening, Day 1, 2, 6, 7, 8, 9,10 and 15 in Cycle 1, Day 1 and 8 in Cycle 2 to 4, end of treatment, early discontinuation, and 30-day safety follow-up, and include temperature, systolic blood pressure (SBP), diastolic blood pressure (DBP), heart rate, and respiratory rate.

Vital signs assessments and changes from baseline in vital signs assessments will be summarized by visit and treatment arm using quantitative summary statistics.

Safety laboratory data will include clinical chemistry, hematology and urinalysis. Descriptive summary statistics (mean, standard deviation, median, minimum, maximum, frequencies, and percentages, as appropriate) for laboratory values will be presented at baseline, the follow-up time points, and change from baseline for each study treatment arm.

Summaries in the form of shift tables for changes in CTCAE grades in key laboratory parameters showing the number and percentage of patients during the course of the study will be also be presented.

All laboratory data, values, units, normal reference range, and out-of-range flags collected in the clinical database will be included in by-patient listings. Graphs of key parametric clinical laboratory tests will be presented by treatment arm (mean ± standard deviation).

All laboratory values will be presented using conventional units.

ECOG will be obtained at screening visit for all patients. Patients will be graded according to the ECOG Performance Status scale and criteria as described in Table 2 . The ECOG will be summarized by grade and treatment arm using qualitative summary statistics and presented in listings. Restricted in physically strenuous activity but ambulatory and able to carry out work of a light or sedentary nature (e.g., office work or light house work) 2 Ambulatory and capable of all self-care but unable to carry out any work activities; up and about more than 50% of waking hours 3 Capable of only limited self-care, confined to bed or chair more than 50% of waking hours 4 Completely disabled; cannot perform any self-care; totally confined to bed or chair

ECGs will be obtained at screening, Day 1 and 2 in Cycle 1, EOT, and at 30-day safety follow-up visit for all patients. The ECG will include heart rate, QRS, QT, and PR intervals, and will be presented in listings.

Health-related QoL questionnaire evaluated with EORTC QLQ-C30 Item 30 (response to question "How do you rate your overall quality of life during the past week") and EQ-5D-5L Item (response to the question "Your Health Today") in Cycle 1 to 4 will be analyzed by using the Wilcoxon Rank Sum Protocol BPI-2358-105 Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Page 32 of 39 test. The method will also be used to construct point estimates and confidence intervals. The NPAR1WAY procedure in SAS version 9.4 or later will be used for these analyses.

The summary statistics in terms of counts, means, standard deviations, medians, minimums, and maximums for individual questions will be summarized in a tabular format.

The study design is group sequential with 1 interim analysis (after 50 approximately patients in each treatment arm have completed at least 1 cycle in each of the treatment arms docetaxel + plinabulin [40 mg] versus docetaxel + pegfilgrastim, with matching placebos) and 1 final analysis at the completion of the study. These results assume that 2 sequential tests are made using the O'Brien-Fleming spending function to determine the test boundaries.

If non-inferiority is determined from the statistical testing, then also the hypothesis of superiority will be tested, and if it is concluded that the plinabulin treatment is superior to the pegfilgrastim treatment, with respect to DSN, then the study will be stopped. Since the design allows for stopping for inferiority, it might be decided at that occasion that the study will be stopped. Since this is a hierarchical testing procedure no penalty with respect to overall significance will be paid.

The statistical testing will be performed and will be reviewed by an independent Data Safety Monitoring Board (DSMB) at the interim analysis.

The changes from the protocol-stated analyses are:

Protocol Amendment 5 SAP 1.0 NPAR1WAY procedure will be used to implement

Wilson score method FREQ procedure will be used to implement

Wilson score method and SAS codes are updated as well NBR model will be used for analysis of incidence rate or proportion type of endpoints

Wilson score method will be used for analysis of incidence rate or proportion type of endpoints Analysis of primary endpoint will be conducted using GENMOD procedure in SAS 9.4

Analysis of primary endpoint will be conducted using NBREG, GLM and TABSTAT procedures in Stata v11.0

Protocol BPI-2358-105

Statistical Analysis Plan, Phase 3

BeyondSpring Pharmaceuticals, Inc.

Zarixo® (filgrastim) FDA Oncologic Drugs Advisory Committee Meeting (FDA Advisory Committee Briefing Document). Jan 7, 2015.

Documentation of revision to the SAP will commence after approval of the Final version 1.0.

Tables that will be included in the topline delivery are shown in boldface font, the tables, listings, and figures for the interim analyses are italicized. 

Docetaxel and granulocyte colonystimulating factor in patients with advanced non-small-cell lung cancer previously treated with platinum-based chemotherapy: a multicenter Phase II trial

Primary granulocyte colony-stimulating factor prophylaxis

Statistical Models Based on Counting Processes

Cox's Regression Model Counting Process: A Large Sample Study

Vivo Interleukin-6 Protects Neutrophils from Apoptosis in Osteomyelitis. Infection and Immunity

A general regression procedure for method transformation. Application of linear regression procedures for method comparison studies in clinical chemistry, Part III

Comparison of EP2006, a filgrastim biosimilar, to the reference: a Phase III, randomized, double-blind clinical study in the prevention of severe neutropenia in patients with breast cancer receiving myelosuppressive chemotherapy

Plinabulin, a Novel Small Molecule That Ameliorates Chemotherapy-Induced Neutropenia, Is Administered on the Same Day of Chemotherapy and Has Anticancer Efficacy

Pegfilgrastim on the Same Day Versus Next Day of Chemotherapy in Patients With Breast Cancer, Non-Small-Cell Lung Cancer, Ovarian Cancer, and Non-Hodgkin's Lymphoma: Results of Four Multicenter, Double-Blind, Randomized Phase II Studies

Sample Size Calculations in Clinical Research. Marcel Dekker

Chemotherapy induced neutropenia. Risks, consequences and new directions for its management

Statistical adjustment of data

Management of chemotherapy-induced neutropenia: measuring quality, cost, and value

A clinical perspective of IL-1β as the gatekeeper of inflammation

Mechanistic models for myelosuppression

The paradox of paclitaxel neurotoxicity: Mechanisms and unanswered questions

Randomized phase III trial of pemetrexed versus docetaxel in patients with nonsmall-cell lung cancer previously treated with chemotherapy

Population pharmacokinetic-pharmacodynamic modeling of moxonidine using 24-hour ambulatory blood pressure measurements

More powerful procedures for multiple significance testing

Blinded, randomized, multicenter study to evaluate single administration pegfilgrastim once per cycle versus daily filgrastim as an adjunct to chemotherapy in high-risk stage II or stage III-IV breast cancer

Discrete sequential boundaries for clinical trials

Plinabulin: Evidence for an immune-mediated mechanism of action, AACR Tumor Microenvironment Event

Lack of effect of aprepitant or its prodrug fosaprepitant on QTc intervals in healthy subjects

Missing data in clinical trials: control based mean imputation and sensitivity analysis

Incidence of febrile neutropenia (FN) is directly related to duration of severe neutropenia (DSN) after myelosuppressive chemotherapy

Pegfilgrastim-Induced Bone Pain: A Review on Incidence, Risk Factors, and Evidence-Based Management

Anthracycline Combinations: Setting A New Standard in Breast Cancer?

Two-sided confidence intervals for the single proportion: comparison of seven methods

Interval estimation for the difference between independent proportions: comparison of eleven methods

Improved confidence intervals for the difference between binomial proportions based on paired data

NPI-2358 is a tubulin-depolymerizing agent: in-vitro evidence for activity as a tumor vascular-disrupting agent

A multiple testing procedure for clinical trials

Management of fever in patients with cancer and treatment induced neutropenia

Plinabulin Investigator Brochure

A novel vascular disrupting agent plinabulin triggers JNKmediated apoptosis and inhibits angiogenesis in multiple myeloma cells

Study NPI-2358-100: A Phase I Study of the Vascular Disrupting Agent NPI-2358 Administered via Intravenous Infusion in Patients with Advanced Solid Tumor Malignancies or Lymphoma

Study NPI-2358-101: A Phase 1/2 study of the Vascular Disrupting Agent NPI-2358 in combination with Docetaxel in patients with Advanced Non-Small cell Lung Cancer

Blinded Phase 3 Study of Second-or Third-Line Chemotherapy with Docetaxel + Plinabulin Compared to Docetaxel + Placebo in Patients with Advanced Non-Small Cell Lung Cancer with at Least One Measurable Lung Lesion

Prescribing Information)

Modeling Survival Data: Extending the Cox Model

First and subsequent cycle use of pegfilgrastim prevents febrile neutropenia in patients with breast cancer: a multicenter, double-blind, placebocontrolled phase III study

A Phase III Randomized Equivalence Study of Biosimilar Filgrastim versus Amgen Filgrastim in Patients Receiving Myelosuppressive Chemotherapy for

A randomized, double-blind, phase II study of ramucirumab plus docetaxel vs placebo plus docetaxel in Japanese patients with stage IV nonsmall cell lung cancer after disease progression on platinum-based therapy. Lung Cancer

Zarixo® (filgrastim) FDA Oncologic Drugs Advisory Committee Meeting (FDA Advisory Committee Briefing Document)

Impact of chemotherapy relative dose intensity on cause-specific and overall survival for stage I-III breast cancer: ER+/PR+, HER2-vs. triple-negative

8 2.2 Secondary Efficacy Objectives (Cycle 1 and Cycles 1 to 4)

Exploratory Efficacy Objectives (Cycle 1)

Exploratory Efficacy Objectives (Cycles 1 to 4)

10 3.2 Secondary Efficacy Endpoints (Cycle 1 and Cycles 1 to 4)

Cycle 1)

Exploratory Efficacy Endpoints (Cycles 1 to 4)

Method of Assigning Subjects to Treatment Groups

Health Economic and Patient Reported Outcomes

Missing or Inconclusive Data Handling

Confidential & Proprietary Page

25 13.2 Cancer History and Prior Cancer-Related Therapies

Prior and Concomitant Therapies (Medication, Treatments, and Surgical Procedures)

Health-Related Quality of Life Questionnaire (QoL) Evaluated with EORTC

Docetaxel and granulocyte colonystimulating factor in patients with advanced non-small-cell lung cancer previously treated with platinum-based chemotherapy: a multicenter Phase II trial

Primary granulocyte colony-stimulating factor prophylaxis during the first two cycles only or throughout all chemotherapy cycles in patients with breast cancer at risk for febrile neutropenia

Comparison of EP2006, a filgrastim biosimilar, to the reference: a Phase III, randomized, double-blind clinical study in the prevention of severe neutropenia in patients with breast cancer receiving myelosuppressive chemotherapy

Sample Size Calculations in Clinical Research. Marcel Dekker

Statistical adjustment of data

Blinded, randomized, multicenter study to evaluate single administration pegfilgrastim once per cycle versus daily filgrastim as an adjunct to chemotherapy in high-risk stage II or stage III-IV breast cancer

Randomized, double-blind, phase 2 study evaluating the safety of same vs next day administration of pegfilgrastim with docetaxel, doxorubicin and cyclophosphamide (TAC) in women with breast cancer

Discrete sequential boundaries for clinical trials

Missing data in clinical trials: control based mean imputation and sensitivity analysis

Incidence of febrile neutropenia (FN) is directly related to duration of severe neutropenia (DSN) after myelosuppressive chemotherapy

Anthracycline Combinations: Setting A New Standard in Breast Cancer?

Two-sided confidence intervals for the single proportion: comparison of seven methods

Interval estimation for the difference between independent proportions: comparison of eleven methods

Improved confidence intervals for the difference between binomial proportions based on paired data

A multiple testing procedure for clinical trials

First and subsequent cycle use of pegfilgrastim prevents febrile neutropenia in patients with breast cancer: a multicenter, double-blind, placebocontrolled phase III study

A randomized, double-blind, phase II study of ramucirumab plus docetaxel vs placebo plus docetaxel in Japanese patients with stage IV nonsmall cell lung cancer after disease progression on platinum-based therapy. Lung Cancer

List of Tables, Listings and Figures for Phase 3 Data

 Protocol BPI-2358-105 IND a. EOT is defined as the last assessment for the protocol-specified treatment post cycle X (Day 22 [+ 7 days]) of the study for an individual patient, regardless at what time the patient discontinues the study. b. All patients, (including those who discontinued from the study) will complete a safety follow-up visit 30 (±2) days after the last dose. Patients who withdraw for progressive disease and who will continue to another chemotherapy regimen will complete the EOT visit at Cycle X Day 21 (where X is the last cycle prior to progression, and is 4 or less). They then continue to another chemotherapy regimen and do not need to have the safety follow-up visit. If, in the opinion of the investigator, the patient will benefit from more than 4 cycles of docetaxel and open label pegfilgrastim, then the fifth cycle will not start until completion of the EOT X X X X X X X X X X X X X X X XAdverse events x X X X X X X X X X X X X X X X X