key: cord-0963527-13akkmcq
authors: Lee, James T; Hesse, Elisabeth M; Paulin, Heather N; Datta, Deblina; Katz, Lee S; Talwar, Amish; Chang, Gregory; Galang, Romeo R; Harcourt, Jennifer L; Tamin, Azaibi; Thornburg, Natalie J; Wong, Karen K; Stevens, Valerie; Kim, Kaylee; Tong, Suxiang; Zhou, Bin; Queen, Krista; Drobeniuc, Jan; Folster, Jennifer M; Sexton, D Joseph; Ramachandran, Sumathi; Browne, Hannah; Iskander, John; Mitruka, Kiren
title: Clinical and Laboratory Findings in Patients with Potential SARS-CoV-2 Reinfection, May–July 2020
date: 2021-02-18
journal: Clin Infect Dis
DOI: 10.1093/cid/ciab148
sha: 81e7b92176ec33526bc395cc1437b2f41ddcf9dc
doc_id: 963527
cord_uid: 13akkmcq

BACKGROUND: We investigated patients with potential SARS-CoV-2 reinfection in the United States during May–July 2020. METHODS: We conducted case finding for patients with potential SARS-CoV-2 reinfection through the Emerging Infections Network. Cases reported were screened for laboratory and clinical findings of potential reinfection followed by requests for medical records and laboratory specimens. Available medical records were abstracted to characterize patient demographics, comorbidities, clinical course, and laboratory test results. Submitted specimens underwent further testing, including RT-PCR, viral culture, whole genome sequencing, subgenomic RNA PCR, and testing for anti-SARS-CoV-2 total antibody. RESULTS: Among 73 potential reinfection patients with available records, 30 patients had recurrent COVID-19 symptoms explained by alternative diagnoses with concurrent SARS-CoV-2 positive RT-PCR, 24 patients remained asymptomatic after recovery but had recurrent or persistent RT-PCR, and 19 patients had recurrent COVID-19 symptoms with concurrent SARS-CoV-2 positive RT-PCR but no alternative diagnoses. These 19 patients had symptom recurrence a median of 57 days after initial symptom onset (interquartile range: 47 – 76). Six of these patients had paired specimens available for further testing, but none had laboratory findings confirming reinfections. Testing of an additional three patients with recurrent symptoms and alternative diagnoses also did not confirm reinfection. CONCLUSIONS: We did not confirm SARS-CoV-2 reinfection within 90 days of the initial infection based on the clinical and laboratory characteristics of cases in this investigation. Our findings support current CDC guidance around quarantine and testing for patients who have recovered from COVID-19.

As of January 3rd, 2021, more than 83 million cases of coronavirus disease 2019 (COVID- 19) have been confirmed worldwide, including 20 million cases in the United States [1] . For most diagnosed cases, the U.S. Centers for Disease Control and Prevention (CDC) recommends completing isolation at 10 days after symptom onset with resolution of fever for at least 24 hours [2] . This recommendation is based on the absence of replication-competent virus for 10 days following symptom onset in mild to moderately severe cases. Retesting is not recommended for 90 days among persons who remain asymptomatic after recovery because SARS-CoV-2 RNA can be detected in their upper respiratory specimens for up to 12 weeks [2] [3] [4] [5] . Patients who do develop new symptoms within 90 days of recovery can be considered for retesting after investigation of alternative diagnoses. However, a better understanding of the duration and robustness of immunity to SARS-CoV-2 and the potential for reinfection would guide public health actions [6] .

Reinfection has been documented among other species of human coronaviruses (HCoV), both experimentally [Callow] and in surveillance cohorts of community members [7, 8] . However, reinfection with SARS-associated coronavirus 1 (SARS-CoV-1) and Middle East respiratory syndrome coronavirus (MERS-CoV), two coronaviruses that can cause severe disease, has not been demonstrated, possibly related to the limited scope of these outbreaks [10, 11] . Recent case reports [12; 13] demonstrating phylogenic differences in SARS-CoV-2 genomes isolated from initial and recurrent episodes of COVID-19 in the same patients raise concern for reinfection; however, questions remain about the frequency and timing of such cases.

During March-April 2020, several reports described cases of recurrent or prolonged SARS-CoV-2 RT-PCR positivity among individuals who had recovered from COVID-19 [14, 15] . An investigation in South Korea reported the absence of both viable virus and secondary transmission from investigated cases of recurrent SARS-CoV-2 RT-PCR positivity [4] . To better understand the clinical and public health implications of similar cases in the United States, in May 2020, CDC initiated an investigation of cases of potential reinfection. Here, we summarize the clinical characteristics and available laboratory findings of cases of potential reinfection reported to CDC by clinicians and public health officials.

We conducted case finding through the Emerging Infections Network (EIN), a provider-based sentinel network of over 1100 actively practicing infectious diseases professionals mainly from North America. EIN is administered by the Infectious Disease Society of America under a CDC cooperative agreement to help identify and understand emerging infectious diseases or clinical manifestations [16] . On May 13, 2020, we published a post on the EIN listserv soliciting reports about potential cases of reinfection from members. We also publicized this EIN posting to state, tribal, territorial, and local health departments. Cases of interest included patients with laboratory-confirmed COVID-19 who had any of following after clinical recovery: 1) Recurrent COVID-19 symptoms and reverse transcription PCR (RT-PCR) positive result for SARS-CoV-2 2) Two documented negative RT-PCR results followed by a positive RT-PCR result for SARS-CoV-2 3) Persistently positive RT-PCR for SARS-CoV-2 results for >30 days M a n u s c r i p t Recovery was further defined as occurring at least 10 days after symptom onset and accompanied by resolution of fever for at least 24 hours, without the use of fever-reducing medications, and with improvement of other symptoms [17] . Clinicians and public health personnel could describe patients who met these criteria by completing a brief web-based form on the EIN website. We reviewed entries to assess if cases met potential reinfection criteria and contacted submitters via phone or email for any needed clarifications and to respond to questions.

During May 13, 2020 -June 19, 2020, we included patients for further clinical and laboratory characterization if they met the criteria listed in the EIN post. During June 20, 2020 -July 17, 2020, we narrowed the inclusion criteria to patients with recurrent COVID-19 symptoms [18] and a concurrent positive SARS-CoV-2 RT-PCR result without an alternative diagnosis. We excluded patients who were aged ≤18 years, not meeting the definition of recovery, or without sufficient clinical information to characterize the COVID-19 clinical course.

This investigation was determined to be non-research and was exempt from further institutional review board (IRB) review at CDC, and the University of Iowa IRB determined the investigation to be nonresearch. This activity was conducted consistent with applicable federal law and CDC policy. 1

For included cases, we requested deidentified medical records pertaining to COVID-19 care and SARS CoV-2 test results (RT-PCR testing laboratory, platform, and cycle threshold [Ct] values; serology). Data on demographics, comorbidities, clinical course, and laboratory testing were abstracted from the medical records and entered into a secured, electronic database (REDCap [Research Electronic Data Capture]) [19] . In addition to data validation checks, three clinician authors reviewed each case to ensure agreement on abstracted data fields, with a focus on SARS-CoV-2 test results and dates of onset and recovery of each clinical episode.

After chart abstraction, three clinician authors classified all patients into four categories based on their clinical course after recovery from an initial episode: 1) Recurrent COVID-19 symptoms without an alternative diagnosis, with concurrent positive SARS-CoV-2 RT-PCR result; 2) Recurrent COVID-19 symptoms with an alternative diagnosis (including potential complications following COVID-19), with concurrent positive SARS-CoV-2 RT-PCR result; 3) Asymptomatic with recurrent positive SARS-CoV-2 RT-PCR after two negative results in specimens collected 24 hours apart; 4) Asymptomatic with recurrent positive SARS-CoV-2 RT-PCR at ≥30 days after recovery, without two negative results in specimens collected 24 hours apart.

We report descriptive statistics to characterize cases according to the four case categories. We expressed continuous variables as medians and interquartile ranges and summarized categorical variables as counts and percentages. All data were analyzed using R software, version 3.61 [19] .

We requested available respiratory specimens that tested RT-PCR positive for SARS-CoV-2 from two time points -the initial diagnosis and at the recurrence of COVID-19 symptoms or recurrent test positivity. Respiratory specimens first underwent RT-PCR testing using the SARS-CoV-2 CDC assay M a n u s c r i p t protocol; Ct values were reported for the N1 and N2 viral nucleocapsid protein gene regions (CDC

). If the Ct value on the respiratory specimen was ≤34, we attempted: 1) viral culture using Vero-CCL-81 cells; 2) whole genome sequencing (WGS) of extracted nucleic acid [22] ; and 3) detection of subgenomic viral RNA transcripts by RT-PCR (i.e., Ct < 40 for both subgenomic spike and nucleocapsid RNA). Specimens also underwent further testing at higher Ct values if laboratory capacity was available. Available serum specimens from after the initial diagnosis were analyzed by chemi-immunoluminiscent assay (CIA) to detect total anti-SARS-CoV-2 spike receptor-binding domain. Detailed laboratory methods are described in Supplementary Text 1.

From May 13, 2020, through July 17, 2020, 296 potential cases of reinfection were submitted through the EIN. After initial review, 75 cases did not meet our initial investigation criteria, and another 51 did not meet our narrowed investigation criteria in place after June 19. We requested records for 170 cases and received complete records for 93 cases; upon review of these records, another 20 did not meet investigation criteria (Supplementary Figure 1 ). In our report, 73 cases were ultimately included. Among these cases, the most common symptoms of the initial COVID-19 episode were respiratory followed by constitutional, and the most common underlying medical condition category was cardiovascular (Table 1) .

Of the 73 patients, 49 (67.1%) developed recurrent COVID-19 symptoms after recovery, including 30 patients whose symptoms were explained by an alternative diagnosis, identified either clinically, through laboratory evaluation, or based on treatment response. Of these 30 patients, 8 (26.6%) had cardiac/circulatory (e.g., congestive heart failure leading to shortness of breath) diagnoses at their subsequent episode, 7 (23.3%) had a bacterial infection (e.g., pneumonia improved with antibiotics), 5 (16.7%) had non-infectious pulmonary diagnoses (e.g., asthma exacerbation), 3 (10.0%) had gastrointestinal diagnoses, and 3 (10.0%) had neurological dysfunction; each of the remaining 4 (13.3%) patients had diagnoses related to either autoimmune, endocrine, urological disorders, or fever of unknown origin (Case #24, Table 2 ). During the recurrent versus initial episode, a lower proportion of patients had respiratory (60.0% vs 90.0%) and constitutional symptoms (40.0% vs 73.3%), and a higher proportion had other symptoms (e.g., headache and chest pain) (70.0% vs 46.7%) ( (Figure 1c ). Among the remaining ten asymptomatic patients who remained PCR positive without the intervening two negative results, three were healthcare workers, while others were tested prior to elective medical procedures or as part of a test-based strategy to discontinue isolation. These patients tested positive a median of 56.5 days (IQR: 47.3-66.5) and up to 71 days after their initial symptom onset date (Figure 1d ).

We received paired specimens from the initial and recurrent COVID-19 symptomatic episodes for nine cases, all of which had tested RT-PCR positive at outside labs under different assays and protocols. Of these, six patients had recurrent COVID-19 symptoms without an alternative diagnosis, and three patients had an alternative diagnosis. Table 2 shows the demographic, clinical presentation, and laboratory test results of these nine cases. All nine initial episode specimens tested positive for SARS CoV-2 on repeat RT-PCR using the CDC assay. Five initial episode specimens underwent additional testing and all were positive for sub-genomic SARS-CoV-2 RNA PCR and viral culture. The remaining four specimens did not undergo additional testing either due to high Ct values (3 cases) or because the specimen had been received prior to the start of our investigation (1 case).

Among nine recurrent episode specimens, repeat RT-PCR at CDC was negative in four specimens and inconclusive for one specimen ( Table 2) . Among the four specimens with positive RT-PCR results, two specimens had Ct values >34 for N1 and N2 gene targets with viral cultures and whole genome sequencing not attempted or unsuccessful when attempted, and negative sub-genomic RNA PCR. One specimen had a Ct value of 33.0 with negative viral culture, negative sub-genomic RNA PCR, and partial genomic sequencing (<200 base pairs). A fourth specimen had a Ct value of 32.4, and whole genome sequencing was unsuccessful.

Serological specimens were available for four out of nine cases -all were positive for antibodies against SARS-CoV-2 either immediately prior to or at the time of symptom onset for the recurrent episode.

Ct values were reported for an additional 16 cases using a variety of extraction techniques, platforms, and PCR targets applied at the diagnosing labs ( Figure 1 ). Among these cases, only one case (Case 48, Figure 1 ), a kidney transplant recipient with a complicated medical course, had a value <30 on the subsequent specimen.

In this investigation of 73 cases of potential SARS-CoV-2 reinfection reported from clinicians across the United States, we did not demonstrate reinfection within 90 days of the initial infection. Clinically, 70% of patients either had recurrent COVID-19 symptoms explained by alternative diagnoses or remained asymptomatic after recovery but were incidentally found to have recurrent or persistent RT-PCR positivity through surveillance and contact investigations. The remaining 19 patients, predominantly healthcare workers, were perhaps more concerning for reinfection because their recurrent COVID-19 symptoms, developing almost two months after recovery, had no alternative diagnoses. Further CDC laboratory investigations of nine available paired specimens from A c c e p t e d M a n u s c r i p t these cases could not confirm reinfection, with an absence of culturable SARS-CoV-2, sub genomic RNA, or complete genome sequence from the recurrent episode specimen. [9] . However, additional reports of potential cases of reinfection [24] [25] [26] did not present evidence of both distinct viral genomes and significant viral burden on reinfection.

To develop a common understanding of what constitutes reinfection, CDC has issued the Investigative Criteria for Suspected Cases of Reinfection [27] , which provides guidance on prioritizing cases with a higher index of suspicion for reinfection and genomic testing of paired specimens, including quality criteria for testing and levels of evidence for reinfection. The highest priority for investigation is suggested for person with detection of SARS-CoV-2 RNA (RT PCR Ct value <33 if known) ≥90 days after the first detection, with or without symptoms, and if paired respiratory specimens are available. For persons with COVID-19-like symptoms and detection of SARS-CoV-2 RNA 45-89 days since first SARS-CoV-2 infection, additional criteria are applied, including absence of an obvious alternative etiology for COVID-19-like symptoms or having had close contact with a person with laboratory-confirmed COVID-19. CDC's guidance is expected to be updated as evidence regarding the duration and robustness of immunity to SARS-CoV-2 emerge. A Common Investigation Protocol (CIP) [28] , has been available to support investigations into suspected SARS-CoV-2 reinfection cases.

This report has several limitations. This investigation was notable for difficulty in obtaining medical and laboratory reports, and specimens for further analysis at CDC, as most laboratories do not routinely retain positive specimens. These challenges contributed to the small sample of cases included in our investigation and tested at CDC. Further, our passive ascertainment of cases through EIN was subject to bias, with an over-representation of healthcare workers or long-term care facility residents, likely reflecting their increased access to SARS-CoV-2 testing. This investigation was focused on potential reinfection cases within 90 days of initial infection, thus we cannot generalize our findings to reinfection beyond 90 days, when waning immunity or divergent strains could increase susceptibility to reinfection.

Re-testing of specimens collected in routine clinical care can have significant variations in specimen collection techniques, timing of repeat testing, and sample degradation that could reduce the likelihood of subsequent WGS, viral culture, and subgenomic RNA, and potentially introduce variability in Ct values. While these issues with retesting of specimens at CDC could have played a role in the failure to culture and sequence recurrent episode specimens (almost all collected within 3 A c c e p t e d M a n u s c r i p t days of symptom onset), our laboratory successfully cultured and sequenced initial episode specimens. Because we did not collect patient identifiers, we were unable to collect exposure and secondary transmission information to support our laboratory findings. Finally, we were unable to rule out reinfection among patients who were asymptomatic with recurrent or persistently positive RT-PCR because we did not receive paired specimens from these cases.

In conclusion, this large public health investigation of potential reinfection contributed to our knowledge of the natural history of SARS-CoV-2 in the United States and informed a standard approach for assessing reinfection. We did not confirm SARS-CoV-2 reinfection within 90 days of the initial infection based on both clinical and laboratory characteristics of 73 cases. Our findings support CDC guidance around quarantine and testing for patients who have recovered from COVID-19 [2] . Additional systematic, prospective cohort investigations [29] are needed to better understand the clinical presentation, risk factors, and frequency of SARS-CoV-2 reinfection cases.

M a n u s c r i p t M a n u s c r i p t M a n u s c r i p t M a n u s c r i p t M a n u s c r i p t 

World Health Organization

Duration of Isolation and Precautions for Adults with COVID-19

Predicting infectious SARS-CoV-2 from diagnostic samples

Prolonged virus shedding even after seroconversion in a patient with COVID-19

Projecting the transmission dynamics of SARS-CoV-2 through the postpandemic period

The time course of the immune response to experimental coronavirus infection of man

Human Coronavirus NL63 Molecular Epidemiology and Evolutionary Patterns in Rural Coastal Kenya

Seasonal coronavirus protective immunity is shortlasting

Duration of Antibody Responses after Severe Acute Respiratory Syndrome

Middle East respiratory syndrome

Recurrence of positive SARS-CoV-2 in patients recovered from COVID-19

Symptomatic SARS-CoV-2 reinfection by a phylogenetically distinct strain

Polymerase Chain Reaction Assays Reverted to Positive in 25 Discharged Patients With COVID-19

The Infectious Diseases Society of America emerging infections network: bridging the gap between clinical infectious diseases and public health

Duration of Isolation and Precautions for Adults with COVID-19 | CDC

COVID-19) -Symptoms

The REDCap consortium: Building an international community of software platform partners

Centers for Disease Control and Prevention. CDC 2019-Novel Coronavirus (2019-nCoV) Real-Time RT-PCR Diagnostic Panel. 2020. Available at

Sensitive, Full-Genome Sequencing of Severe Acute Respiratory Syndrome Coronavirus 2

Duration of infectiousness and correlation with RT-PCR cycle threshold values in cases of COVID-19

Assessment of the risk of SARS-CoV-2 reinfection in an intense re-exposure setting

Genomic evidence for reinfection with SARS-CoV-2: a case study

Setting the criteria for SARS-CoV-2 reinfection -six possible cases

Centers for Disease Control and Prevention. Investigative Criteria for Suspected Cases of SARS-CoV-2 Reinfection (ICR). 2020. Available at

Centers for Disease Control and Prevention. Common Investigation Protocol for Investigating Suspected SARS-CoV-2 Reinfection. 2020. Available at

Antibody Status and Incidence of SARS-CoV-2 Infection in Health Care Workers

We thank the clinicians, infection control staff, and public health staff, who while immersed in caring for their patients and community generously reported information about cases to further our understanding of reinfection. We are especially grateful for the participation of the following individuals: Lilian Abho, Ashley Buys, Matthew Coon, Sherri Davidson, Todd Hulgan, Maria Napierskie, Jennifer Read, Maria Safdar, Alison Wheeler, Pamela Pontones who discussed cases with us, as well as coordinated records and specimens for analysis in this publication; and to CDC CLIA Testing Personnel Lili Punkova and Matthew Pauly. We appreciate Clint Paden and the SPHERES community for their support and helpful discussions. We thank the Emerging Infections Network, especially Susan Beekman for her invaluable support in processing case submissions.

All authors have completed and submitted the International Committee of Medical Journal Editors form for disclosure of potential conflicts of interest. No potential conflicts of interest were disclosed.The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

M a n u s c r i p t * Respiratory specimens with Ct ≥34 values were not to undergo further testing by protocol but could have been tested if laboratory capacity was available. **Further testing was not attempted on these specimens pre-existing in the CDC lab before the start of our investigation. A c c e p t e d M a n u s c r i p t A c c e p t e d M a n u s c r i p t Figure 1