Discussion:
In our study evaluating SARS-CoV-2 viral persistence in ICC, we found that the median time to the first of two negative tests was 6 weeks. Most children were tested in outpatient settings prior to admission or for procedural sedation, and over half of those tested were asymptomatic at the time of their first positive test. There was no significant difference in time to negativity among symptomatic versus asymptomatic children.  Among children who remained positive at day 30, half had CT values <30, suggesting persistence of moderate to high viral loads and assumed potential risk of transmission. Our study highlights the importance of PCR-based screening for ICC to guide isolation duration, irrespective of symptom presentation. These data support our current practice, with the need for at least two negative tests, given the potential for intermittent negative followed by positive results, and risk for ongoing, potential transmission even several weeks from initial infection.
The CDC recommends an isolation period of 10 days in healthy, non-immunocompromised individuals with SARS-CoV-2 infection based on studies that demonstrate the absence of viable/transmissible virus beyond this period.16 Immunocompromised patients can have a longer period of infectivity, but there is little available evidence in the literature to guide infection control practices. Our study revealed a prolonged period of viral persistence in ICC, with a longer period of positivity compared with a case series of ICC and a study of immunocompromised adults.9,17 Children with leukemia/lymphoma had longer viral persistence compared with children with other immunocompromising conditions, which we hypothesize is due to this population having the highest impairment of adaptive immunity. These children also represent the population of children with the greatest degree of lymphopenia, which has been shown in adults to be associated with more severe disease.18 Another study has also shown that patients critically ill from SARS-CoV-2 were more likely to have a decrease in CD3+CD4+ T-cells compared to non-critical patients.19 Additionally, a study of adult cancer patients found a general lack of T-cell response to CoV N, M and S proteins from SARS-CoV-2. While these findings suggested a lack of a protective T-cell response, the patients presented with mild illnesses despite relatively high viral loads.12 And although we had one death in our patient cohort post-bone marrow transplant, allogeneic, autologous and CAR-T therapy patients have overall favorable outcomes following SARS-CoV-2 infection.13
Similar to findings from adult studies, we did not see a relationship between clinical presentation and persistent positive SARS-CoV-2 status. An adult study of 3758 adults re-tested following initial SARS-CoV-2 infection found longer time to negativity to be associated with older age, multiple comorbidities and solid organ transplant but not by degree of immunocompromise or illness severity17. Our findings serve as a reminder that even asymptomatic children may have potentially transmissible virus several weeks after initial detection. Asymptomatic ICC are a potentially important group that pose high risk of transmissibility given their silent infection and prolonged infectious period.
While PCR is unable to differentiate actively replicating virus from viral shedding, CT values may serve as a useful surrogate for viral load and potential transmissibility. Other studies suggest that CT values of 25-30 represent moderate to high viral loads.  We evaluated CT values using the same PCR assay for a small sample of patients with available clinical samples, which demonstrated that half still had a moderately high viral load even 3-4 weeks from their initial viral infection, regardless of their presentation. Other studies evaluating samples with PCR, viral culture, and subgenomic RNA have shown that CT values similar to those found in our study were associated with viable, culturable virus, as well as the presence of subgenomic RNA, which indicates actively replicating virus.20 These data further support the need for extended isolation duration for ICC given the persistence of relatively low CT values, suggesting potentially transmissible virus. Low CT values indicating relatively high viral loads were observed even in asymptomatic ICC, reflecting a need for test-based isolation protocols, rather than symptom-based isolation.
There are several limitations of this single-center, retrospective study. A small sample size precluded adjusted analyses to account for factors that may influence viral persistence and increased the risk of a type II error. Given the retrospective nature of the study, there was non-standardized repeat testing, and the frequency and intervals were not consistent between study participants. Repeat testing at our institution was recommended at 3 weeks from the first positive test, which may have increased our median time to negative testing, but most children in our cohort were still positive at this timepoint. We did not have a cohort of non-ICC with repeat SARS-CoV-2 testing, so we were not able to compare the degree of viral persistence in ICC with a healthy population.  Finally, clinical SARS-\sout CoV-2 testing was conducted on different testing platforms, and there were a limited number of remaining sequential samples available for repeat testing to yield CT data.
In conclusion, our study demonstrates prolonged viral persistence of SARS-CoV-2 viral material in ICC, even in asymptomatic children, with moderate to high viral loads seen in the majority of children several weeks from initial positive testing. These findings highlight the importance of PCR-based screening for ICC to help guide isolation duration, irrespective of symptom presentation, given the risk for ongoing transmission even several weeks from initial infection.
Conflicts of Interest: Suchitra Rao and Samuel R. Dominguez received research support from BioFire Diagnostics; all other authors have no conflicts of interest to disclose.
Acknowledgements: We would like to acknowledge the Clinical Microbiology Laboratory at Children’s Hospital Colorado for their tireless work during the pandemic.