Discussion
In this case series, we identified 11 individuals who had viral RNA detected from their upper respiratory tract >70 days after their primary SARS-CoV-2 infection. We describe two groups: a probable SARS-CoV-2 re-infection group and a probable primary infection group with prolonged viral RNA shedding. Of these 11 individuals, 7 were consistent with probable re-infection, while 4 did not meet the criteria for probable re-infection and thus were considered primary infection with prolonged viral RNA shedding. We based this conclusion on several factors. First, two patients had ≥2 negative rtPCR tests prior to their probable SARS-CoV-2 re-infection and one patient (patient 3) had 1 negative test prior to their probable re-infection. Second, all 7 individuals had at least 90 days between their primary SARS-CoV-2 infection and their probable SARS-CoV-2 re-infection. We attempted to be more inclusive for re-infections occurring <90 days after initial symptom onset with primary infection, however, none were identified. Our data support the assertion by the CDC that re-infection is unlikely to occur within 90 days from onset of symptoms with primary SARS-CoV-2 infection.4,5
Definitive SARS-CoV-2 re-infection has been confirmed in a small number of cases.2,3 Confirmation via WGS demonstrated that the SARS-CoV-2 isolate detected with the re-infection event was sufficiently different from the isolate sequenced from the primary infection. In our population, we were unable to confirm SARS-CoV-2 re-infection cases using WGS or Sanger sequencing of the S gene because the viral RNA quantity isolated during the probable re-infection events was insufficient despite multiple sequencing attempts. In two cases, we had blood samples collected prior to and after their probable SARS-CoV-2 re-infection. In both cases, we did not observe a four-fold or greater rise in IgG anti-S antibody levels consistent with an infection. Lack of a significant antibody rise has been reported in another case report with confirmed SARS-CoV-2 infection.2,14 This lack of antibody response with re-infection was attributed to the patient’s mild illness, which is consistent with our observation that both of our patients were asymptomatic with their re-infection event.14
Early re-infection or prolonged viral RNA shedding with primary SARS-CoV-2 infection might be a manifestation of non-sterilizing immunity that is generated with the initial infection. Unlike other limited case reports, which have documented recurrence of infections after hospitalization due to SARS-CoV-2, this is an ambulatory population with relatively mild symptomatology.12 None of our 7 patients were hospitalized during their first episode and only 2 patients were symptomatic with their probable re-infection event. Patient 5 had the highest severity during her probable re-infection, and she was also the oldest patient in our cohort. Another study that examined viral kinetics of SARS-CoV-2 found that a higher severity of symptoms (ICU admission) significantly decreased a patient’s likelihood of prolonged shedding of SARS-CoV-2 RNA.6 This coupled with observations that individuals with severe cases of SARS-CoV-2 infection have increased IgG anti-S antibody responses compared with mild cases could explain why we see predominantly mild presentations in our cohort resulting in probable re-infections or primary infection with prolonged viral RNA shedding.13,15
Limitations
This was a single site cohort study that reported on approximately 7 months of rtPCR testing for SARS-CoV-2. The low viral RNA content in the samples collected from probable re-infection cases precluded our ability to confirm the re-infection event using sequencing.