key: cord-0771132-59cdlhge
authors: Rogiers, Aljosja; Pires da Silva, Ines; Tentori, Chiara; Tondini, Carlo Alberto; Grimes, Joseph M; Trager, Megan H; Nahm, Sharon; Zubiri, Leyre; Manos, Michael; Bowling, Peter; Elkrief, Arielle; Papneja, Neha; Vitale, Maria Grazia; Rose, April A N; Borgers, Jessica S W; Roy, Severine; Mangana, Joanna; Pimentel Muniz, Thiago; Cooksley, Tim; Lupu, Jeremy; Vaisman, Alon; Saibil, Samuel D; Butler, Marcus O; Menzies, Alexander M; Carlino, Matteo S; Erdmann, Michael; Berking, Carola; Zimmer, Lisa; Schadendorf, Dirk; Pala, Laura; Queirolo, Paola; Posch, Christian; Hauschild, Axel; Dummer, Reinhard; Haanen, John; Blank, Christian U; Robert, Caroline; Sullivan, Ryan J; Ascierto, Paolo Antonio; Miller Jr, Wilson H; Stephen Hodi, F; Suijkerbuijk, Karijn P M; Reynolds, Kerry L; Rahma, Osama E; Lorigan, Paul C; Carvajal, Richard D; Lo, Serigne; Mandala, Mario; Long, Georgina V
title: Clinical impact of COVID-19 on patients with cancer treated with immune checkpoint inhibition
date: 2021-01-19
journal: J Immunother Cancer
DOI: 10.1136/jitc-2020-001931
sha: cf4d076cfff394f0e17acabd4fbcf2438d76b027
doc_id: 771132
cord_uid: 59cdlhge

BACKGROUND: Patients with cancer who are infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are more likely to develop severe illness and die compared with those without cancer. The impact of immune checkpoint inhibition (ICI) on the severity of COVID-19 illness is unknown. The aim of this study was to investigate whether ICI confers an additional risk for severe COVID-19 in patients with cancer. METHODS: We analyzed data from 110 patients with laboratory-confirmed SARS-CoV-2 while on treatment with ICI without chemotherapy in 19 hospitals in North America, Europe and Australia. The primary objective was to describe the clinical course and to identify factors associated with hospital and intensive care (ICU) admission and mortality. FINDINGS: Thirty-five (32%) patients were admitted to hospital and 18 (16%) died. All patients who died had advanced cancer, and only four were admitted to ICU. COVID-19 was the primary cause of death in 8 (7%) patients. Factors independently associated with an increased risk for hospital admission were ECOG ≥2 (OR 39.25, 95% CI 4.17 to 369.2, p=0.0013), treatment with combination ICI (OR 5.68, 95% CI 1.58 to 20.36, p=0.0273) and presence of COVID-19 symptoms (OR 5.30, 95% CI 1.57 to 17.89, p=0.0073). Seventy-six (73%) patients interrupted ICI due to SARS-CoV-2 infection, 43 (57%) of whom had resumed at data cut-off. INTERPRETATION: COVID-19–related mortality in the ICI-treated population does not appear to be higher than previously published mortality rates for patients with cancer. Inpatient mortality of patients with cancer treated with ICI was high in comparison with previously reported rates for hospitalized patients with cancer and was due to COVID-19 in almost half of the cases. We identified factors associated with adverse outcomes in ICI-treated patients with COVID-19.

Coronavirus disease 2019 (COVID-19) is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). 1 As of October 2020, more than 35,000,000 people have been infected with SARS-CoV-2 worldwide with more than 1,000,000 deaths. 2 The clinical spectrum of COVID-19 varies enormously from asymptomatic individuals to critical illness and death. 3 Risk factors for severe disease include older age, male sex and comorbidities such as cardiovascular and pulmonary diseases, diabetes and cancer. 4 5 Mortality rates of COVID-19 infection in the cancer population range from 7.6% to 33% 5-9 compared with 1.4%-2.3% 3 10 in an unselected population.

One key question specific for the cancer population pertains to the potential impact of immune checkpoint inhibition (ICI) on the clinical course of COVID-19. Programmed cell death 1 (PD-1)-based immunotherapy releases the brakes of immune tolerance mechanisms leading to effective anti-tumor responses. 11 Adaptive immune cells involved in this process, in particular CD8 + and CD4 + T cells, are also essential to control and establish immunity against viruses, and ICI may enhance immunologic control of viral Open access infections. 12 13 It is therefore theoretically possible that ICI offers protection against the development of severe COVID-19 illness. Nevertheless, these immune cellseither through direct cytotoxicity or cytokine releasecan also contribute to inflammation and may aggravate the clinical course of COVID-19. An example of an immune-mediated consequence of SARS-CoV-2 is acute respiratory distress syndrome which is the leading cause of mortality in COVID-19. 14 15 ICI has been associated with severe disease in some 9 16 17 but not all 18 studies, and the numbers of ICItreated patients in these studies were small. Although vigilance is certainly warranted in the ICI-treated patient population, unnecessary treatment delays may compromise cancer-related outcomes and some patients may not be offered ICI therapy in areas of high COVID-19 prevalence due to concerns of infection and severe illness. With the pandemic continuing, recommendations are needed to inform ICI-related treatment decisions. In this multicentric study, we describe the clinical course, treatment and outcomes of COVID-19 infection in patients treated with ICI across different tumor types and across different geographic regions.

We conducted a multicenter, retrospective, cohort study in 19 centers across 9 countries (Australia, Canada, France, Germany, Italy, Switzerland, The Netherlands, UK and USA). Between March 5 and May 15, 2020, we included 110 adult (aged ≥18 years) patients with any type of solid malignancy who had undergone treatment with ICI and had laboratory-confirmed positive SARS-CoV-2. The test could either be nucleic acid detection based (nasopharyngeal swabs) or serological. Asymptomatic patients found positive for SARS-CoV-2 were included in this study. These patients were tested according to local policies after exposure to a person with confirmed SARS-CoV-2 (transmission tracking and contact tracing). All patients must have received at least one cycle of ICI within 12 months prior to testing positive for SARS-CoV-2. Patients who received chemotherapy within 12 weeks prior to COVID-19 diagnosis were excluded as chemotherapyinduced immunosuppression may have confounded analyses. Combinations of ICI with anti-VEGF agents were allowed. Clinical and laboratory data were obtained from the medical records from each of the centers. Local Institutional Review Board approval was required for each center. All study procedures were in accordance with the precepts of Good Clinical Practice and the declaration of Helsinki.

Clinical data were extracted from medical records and de-identified for analysis. Data were divided into the following categories: demographics and patient characteristics (age, sex, geographic region, Eastern Cooperative Oncology Group (ECOG) performance status), cancer characteristics (cancer type, American Joint Committee on Cancer stage, treatment setting), comorbidities (cardiovascular, pulmonary, renal disease or diabetes mellitus), ICI treatment (anti-PD-1/anti-PD-L1, combination anti-PD-1 plus anti-CTLA-4 or other), COVID-19 symptoms (fever, cough, dyspnea), laboratory tests (leukocytes, lymphocytes, C reactive protein (CRP) and creatinine), management (oxygen therapy, mechanical ventilation, use of vasopressors, renal replacement therapy, antivirals, antibiotics, glucocorticoids, anti-IL-6 agents and intravenous immunoglobulins) and clinical outcomes (hospital admission, intensive care unit admission and mortality).

Patient demographics and clinical characteristics are summarized by their frequency and proportion. Primary outcomes were hospital admission (coded yes/no) and overall survival (time to event). Measures for association between demographic and clinical characteristics on one hand and hospital admission on the other hand were assessed using univariate and multivariate logistic regression. ORs and associated 95% CIs are reported. Overall survival was assessed using univariate Cox proportional hazard regression; given the low number of deaths observed, multivariate Cox was not considered. HRs and associated 95% CIs are reported. For associations between demographic and clinical characteristics and intensive care unit admission, only frequencies are reported given the low absolute number of events. The selection of symptoms and laboratory findings was based on previous reports. 19 

Between March 5 and May 15, 2020, 110 patients with cancer on ICI only who tested positive for SARS-CoV-2 were included in this study. The median follow-up since COVID-19 diagnosis was 82 days (range 1-119 days). The median age was 63 years (range 27-86) and most patients were male (65%). Seventy (64%) patients were treated in Europe (48 in Italy) and 36 (33%) in North America. Eightythree (75%) patients had advanced cancer, including 64 (58%) with melanoma, 17 (16%) with non-small cell lung cancer (NSCLC) and 10 (9%) with renal cell carcinoma (RCC) (table 1). Most patients did not have an underlying comorbidity (55%); the most common comorbidities were cardiovascular disease (27%), diabetes mellitus (15%), pulmonary disease (12%) and/or renal disease (5%). Overall, for cancer, 90 (82%) patients were treated with anti-PD-(L)1 monotherapy (nivolumab, pembrolizumab, spartalizumab, atezolizumab or durvalumab) and 16 (15%) with combination anti-PD-(L)1 and anti-CTLA-4 (nivolumab-ipilimumab, durvalumab-tremelimumab or pembrolizumab-MK1308). Twenty-five (23%) patients had chemotherapy prior to ICI. The median time between last chemotherapy and COVID-19 diagnosis was 40 weeks 

Median time between COVID-19 diagnosis and last ICI dose was 26 days (range 0-363). Sixty-seven (62%) patients were symptomatic at COVID-19 diagnosis. The most common COVID-19-related symptoms were fever (67%), cough (58%) and dyspnea (33%), and 10% of patients presented ECOG ≥2 at the time of COVID-19 diagnosis. Lymphocytopenia was the most common laboratory abnormality in 47% of the 79 patients who had laboratory tests available within 3 days of COVID-19 diagnosis, and 15 patients (14%) were on prednisone equivalent dose of ≥10 mg at diagnosis, mostly for ICI toxicity (table 2) . Thirty-five (32%) patients were admitted to hospital and 7 (6%) to intensive care (figure 1). For the patients managed in hospital, 20 received oxygen therapy and 3 mechanical ventilation. Antibiotics and antiviral agents were given to 24 and 5 admitted patients, respectively. Only 10 patients were treated with glucocorticoids for COVID-19; anti-IL6 and intravenous immunoglobulin were given to 2 patients and 1 patient, respectively ( 17 20 21 and was due to COVID-19 in almost half of the cases. These findings have implications for clinical decision-making for patients who are receiving immunotherapy. COVID-19 has had, and continues to have, an unprecedented impact on society and healthcare services. The increased risk for severe illness and mortality in patients with cancer has raised safety concerns among patients and clinicians alike. 7 8 17 The notion that a subgroup of patients with severe COVID-19 develop a cytokine storm syndrome, 14 possibly reflecting a high-risk inflammatory phenotype, is of particular concern in a context of ICI. 22 To better understand the impact of ICI on patients with laboratory-confirmed SARS-CoV-2, we integrated and analyzed data from 110 ICI-treated patients with different cancer types. 

Almost 40% of patients had asymptomatic COVID-19 infection, which is well in the range of the reported 30%-75% of individuals with asymptomatic COVID-19 infection in the community. 8 23-25 Although less permissive screening programs in some countries make it challenging to ascertain the true number of asymptomatic COVID-19 cases, patients with cancer are likely to be repeatedly tested. Our findings do not indicate a different symptomatic presentation at diagnosis of COVID-19 in ICI-treated patients. Considering the frequency at which patients on ICI visit healthcare facilities, this finding underscores the importance of transmission tracking and contact tracing in healthcare services to avoid spread of COVID-19 among vulnerable patients.

Just under a third of the patients with COVID-19 on immunotherapy were admitted to hospital in our study, most of whom were symptomatic. This is slightly lower than hospital admission rates reported for the cancer population in general. 7 The presence of dyspnea was, as expected, a predictor for both hospital admission and mortality. ICU admission rate in our study was low at only 6%. While this is in line with the notion that 5% of individuals without comorbidities develop severe illness from COVID-19 infection, 3 it is in stark contrast with the high mortality rates in the cancer population. In our study, 18 out of 110 patients died (16%), most not attributed to COVID-19. This fatality rate is clearly higher than the case fatality rate in the community (2.3%) 3 yet within the range of what has been reported for the cancer population (7.6%-33%). 5 6 None of the patients who received adjuvant ICI died. Of the 18 patients who died, 14 were not admitted to ICU. Although underlying malignancy and constrained healthcare systems were reported as factors in decision-making, further exploration of the perceptions of a cancer diagnosis will be essential in an era where increasing numbers of patients with advanced cancer achieve long-term responses to immunotherapy.

Interpreting overall mortality is inherently complicated by the divergent impact ICI may have on COVID-19related and malignancy-related outcomes. Furthermore, several factors could have influenced mortality rates in our study. First, the effect of tumor type on COVID-19 outcomes has been recognized: higher COVID-19 mortality rates have been reported for hematological 26 and thoracic malignancies. 6 Our cohort included more patients with melanoma than with NSCLC and the mortality rate of patients with melanoma was lower than that of patients with NSCLC. Second, treatment with anti-PD-(L)1 plus anti-CTLA-4 was associated with a significant higher admission but not mortality rate possibly reflecting the higher adverse events rate but better oncological outcomes associated with combination immunotherapy. Third, patients who had chemotherapy in the 12 weeks prior to COVID-19 were excluded from our study. If recent chemotherapy results in worse COVID-19-related outcomes, exclusion of these patients could contribute to a difference in observed outcomes; however, recent administration of chemotherapy has not been invariably associated with worse COVID-19 outcomes. 8 9 Finally, median follow-up in our study was 82 days, which is several times longer than earlier reports. 6 7 Of particular relevance is the notion, consistent with other reports, 9 that baseline corticosteroid use (at a prednisone equivalent dose of ≥10 mg) is associated with a higher risk for admission. The RECOVERY trial, which had not yet been published when patients in our study were treated, showed that the anti-inflammatory and immunosuppressive effects of dexamethasone were beneficial in patients requiring respiratory support. 22 While the use of corticosteroids may be confounding in an ICI context (immunotherapy-related adverse events), it remains important to note that dampening immune response with corticosteroids in the early stages of COVID-19 infection can be detrimental and may also compromise cancer-related outcomes. 27 In later stages, with an aberrant inflammatory response and less viral activity taking place, corticosteroids could be of benefit. This may be consistent with our finding that prednisone was associated with a higher risk for hospital admission but not with increased mortality. Lymphocytopenia, as reported by Yang et al, 19 and high CRP were predictors for higher mortality but may also impact malignancy-related outcomes. 28 It is, however, somewhat unexpected that a well-established risk factor such as cardiovascular disease was not associated with more severe COVID-19 illness. This is likely the result of a smaller population (ICI-only as opposed to unselected) and cardiovascular disease not being further specified. In our study, older age and male sex were also not associated with a higher mortality (4 out of 18 patients who died were younger than 50 years, 3 of whom were women).

Approximately one fourth of the patients did not interrupt ICI despite COVID-19 infection and most of these patients were asymptomatic. This subset of patients was too small to draw any conclusions regarding adverse outcomes for the patients themselves, yet it is, from a public health perspective, imperative to isolate patients who tested positive for SARS-CoV-2.

Our study has some limitations. First, there is no non-ICI control group; this would have been challenging given that treatment choices are inextricably linked to Figure 1 Venn diagram representing all patients (110), patients admitted to hospital (35), admitted to intensive care unit (ICU) (7) and patients who died (18) . 

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and coronavirus disease-2019 (COVID-19): the epidemic and the challenges

Johns Hopkins Coronavirus Resource Center. COVID-19 Map

Characteristics of and important lessons from the coronavirus disease 2019 (COVID-19) outbreak in China: summary of a report of 72 314 cases from the Chinese Center for Disease Control and Prevention

Clinical characteristics and risk factors associated with COVID-19 disease severity in patients with cancer in Wuhan, China: a multicentre, retrospective, cohort study

Report of the WHO-China joint mission on coronavirus disease

COVID-19 in patients with thoracic malignancies (TERAVOLT): first results of an international, registry-based, cohort study

Clinical impact of COVID-19 on patients with cancer (CCC19): a cohort study

COVID-19 mortality in patients with cancer on chemotherapy or other anticancer treatments: a prospective cohort study

Determinants of COVID-19 disease severity in patients with cancer

Clinical characteristics of coronavirus disease 2019 in China

Fundamental mechanisms of immune checkpoint blockade therapy

Functional exhaustion of antiviral lymphocytes in COVID-19 patients

Restoring function in exhausted CD8 T cells during chronic viral infection

COVID-19: consider cytokine storm syndromes and immunosuppression

The cytokine storm in COVID-19: an overview of the involvement of the chemokine/ chemokine-receptor system

A rapid fatal evolution of Coronavirus Disease-19 in a patient with advanced lung cancer with a long-time response to nivolumab

Patients with cancer appear more vulnerable to SARS-CoV-2: a multicenter study during the COVID-19 outbreak

Impact of PD-1 blockade on severity of COVID-19 in patients with lung cancers

Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a singlecentered, retrospective, observational study

Do patients with cancer have a poorer prognosis of COVID-19? An experience in New York City

Case fatality rate of cancer patients with COVID-19 in a New York hospital system

Dexamethasone in hospitalized patients with Covid-19 -preliminary report

Prevalence of asymptomatic SARS-CoV-2 infection: a narrative review

Lessons from mass-testing for COVID-19 in long term care facilities for the elderly in San Francisco

Prevalence of SARS-CoV-2 infection in residents of a large homeless shelter in Boston

COVID-19 prevalence and mortality in patients with cancer and the effect of primary tumour subtype and patient demographics: a prospective cohort study

Association of steroids use with survival in patients treated with immune checkpoint inhibitors: a systematic review and meta-analysis

Relationships between lymphocyte counts and treatment-related toxicities and clinical responses in patients with solid tumors treated with PD-1 checkpoint inhibitors

Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Patient consent for publication Not required.

Data availability statement All data relevant to the study are included in the article or uploaded as online supplemental information.Open access This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http:// creativecommons. org/ licenses/ by-nc/ 4. 0/.