This Covid-19 Rebound Study showed that in a real-world, observational cohort of participants who tested positive for COVID-19 and were eligible to receive NPR, the NPR treatment group had triple the 15-day incidence rate of COVID-19 viral rebound and double the incidence rate of symptom rebound compared to the control group. These findings persisted in our secondary analysis to adjust for baseline characteristics that may confound the association between NPR and COVID-19 rebound. In our preliminary report14, which was only a quarter of the overall size (n = 170), and released due to the need for expedience of results in the pandemic, we showed probability of viral rebound after initial viral test clearance (18% NPR, 8% control) and symptom rebound after initial symptom resolution (42% NPR, 11% control) but these did not meet statistical significance – likely owing to being underpowered (Figure 3 in preliminary report14). Now, at study completion, we see a clear statistically significant positive association of both viral rebound (18% NPR, 6% control) and symptom rebound (44% NPR, 18% control) with intention to take NPR (Figure 2).
The estimates in our study are comparable with other studies, which have reported cumulative incidence of viral rebound in NPR-treatment populations from 4% to 27%, and symptom rebound ranges from 0.8% to 32%.16-18 While variability could be due to several factors including rebound definitions, eligibility criteria, and overall study design, our study is one of the few prospective, real-world observational cohorts adding to the literature. Studies with lower estimates tended to be secondary analyses of randomized controlled trials or retrospective analyses of electronic medical record data, which may only capture more severe cases.12,17-21 Studies with the higher estimates tended to use more sensitive measures of viral load and in some cases did not require full resolution of symptoms to be eligible for symptom rebound.16,22 A strength of our study is that it was designed with rebound as the primary outcome and employed frequent every two days telehealth proctor verified home testing, and symptom surveys to get incidence rate calculations. This allowed us to calculate both overall risk of rebound from the time of initial NPR treatment decision, and clearance-dependent rebound risk, to address the question of risk of rebound after starting to feel better or testing negative. Further, in contrast to most other studies, we did not exclude participants who took partial doses of NPR, allowing for more accurate real-world estimates of rebound. Interestingly, the highest rates of viral rebound were among those who reported taking the full NPR prescription of 5 days, while the highest rates of symptom rebound were among those who only completed a partial prescription, suggesting a possible dissociation between viral clearance and symptoms. Symptoms such as fevers and congestion often represent immune processes. Symptoms following rebound may reflect the ramping up of the immune system to ensure the virus remains clear after a crutch afforded by NPR to help clear the virus is removed.
We found that there were minimal overall differences in time to initial viral or symptom clearance in the NPR and control groups but increased rebound, suggesting further investigation into the discordance between how NPR reduces the potency of the virus to ultimately reduce the progression to severe COVID-19 as found in previous studies, versus the increased rates of rebound or persistent virus as detected here. Blood samples collected during the second phase of our study will be used to ask this question and will be part of a separate report.
Time to initial symptom clearance was longer than viral clearance, which is consistent with another prospective study that found the cumulative probability of clearance within 10 days was 34% for symptoms versus 63% for viral load, without significant differences between their NPR group and propensity matched controls.16 The tendency for symptoms to remain after rapid antigen tests become negative is supported by other studies as well as the congruence of positive tests and symptoms across the acute phase in our study (Supplementary Table S9), adding support to the idea that a longer course of NPR may be needed to fully clear virus and prevent resurgence of both virus and symptoms. One important caveat to note here is that we captured self-reported data on the presence or absence of symptoms, but given the study design it was not feasible for us to collect symptom severity data. Therefore, while we did not find differences between study groups on time to initial symptom clearance, symptom severity might yield different results.
An additional finding from our study, is that at 5 days after enrollment (corresponding to up to 7 after first turning positive), over one third of all participants remained positive on rapid antigen tests, regardless of NPR (Figure 2C). Although CDC guidance has largely recommended ending isolation after 5 days our findings suggest that individuals should remain cautious even after 5 days, as many could still remain infectious.
Numerous mechanisms have been hypothesized for why rebound occurs. The constant push and pull between virus and immunity drives “predator-prey dynamics” where the virus’ presence triggers the activation of immune defenses, and virus clearance signals these immune defenses to shift towards a recovery mode.23 If the shift is too early, the relaxed defenses can allow remaining virus to rapidly expand once more. Regarding NPR and a higher probability of rebound, it is possible that very early treatment initiation could work to pharmaceutically drive down virus growth while abrogating the need to raise a robust immunological defense. So, upon removal of NPR, remaining virus could replicate at a time scale that outpaces immune defenses, leading to a virus rebound until defenses catch up. Such a mechanism would likely present as an increase in COVID-19 rebound among those who either start NPR very early before immune defenses escalate or those who terminate the course prematurely, and could explain why remarkably high rates of rebound have been noted among individuals who might be undergoing particularly high frequency surveillance testing ergo early detection and therapy initiation. This has also been supported by mathematical model of viral load dynamics that support the hypothesis that a 5-day course of NPR started early is not always sufficient to completely clear virus and that delayed or longer courses may be needed to combat COVID-19 rebound while maintaining the important role of NPR on reducing risk of severe COVID-19 outcomes.24
Like many of the symptoms of COVID-19 in an era of widespread immunity that represent activating immune defenses (i.e. fever and congestion), the symptoms of rebound generally reflect mounting immune defenses. Thus, if upon removal of treatment with NPR, immunity must ramp up to clear remaining virus, it would be expected that a large fraction of individuals might experience symptoms, leading to a more common symptom rebound phenomenon and a less common full virus rebound, which is what we see in our results.25,26
The COVID-19 Rebound Study was not designed to investigate the efficacy of NPR. There is clear pre-existing evidence from randomized trials and clinical guidance supporting the impact of NPR on reducing COVID-19 related hospitalizations and progression to severe COVID-19 outcomes.3-7,15 However, there is a pressing need to better understand the pathophysiology of how NPR works to clear the viral load and the answer may lie in what happens to the virus in rebound patients. Based on our observations of NPR treatment outcomes, our recommendation to clinical providers prescribing NPR is to inform their patients that there is a higher likelihood of viral and symptom rebound, that we do not have enough evidence around the severity of these symptoms, but the evidence supporting reduction in hospitalizations and progression to severe COVID-19 is strong.
This was a non-randomized prospective observational study that may inherently be limited by selection bias, recall bias, and an unbalanced sample size; therefore, even though primary outcome findings remained positive after adjusting for potential confounders, external validity may be limited. Another limitation is that in attempting to make the cohort as real-world as possible, we required enrollment within 48-hours of a positive rapid antigen test; however, participants may have had symptoms and the disease for longer durations prior to self-testing and/or asymptomatic participants may have tested at an unknown time point after infection, thereby leading to underestimates on time to viral or symptom clearance. Timing estimates are also impacted by the length of the study as not everyone experienced initial clearance, or re-clearance after rebound by Day 15. The relatively large numbers of participants in both groups helps to mitigate but not necessarily eliminates the impact of these potential biases in timing estimates as they pertain to relative differences between groups. As mentioned previously, our insights on symptomatology are limited by only having binary presence or absence of distinct symptom types, with no detail around severity of symptoms between groups or phases of recovery. The social media recruitment arm was opened partway through the study primarily to increase the number of participants who were not intending to take NPR, as those recruited through the eMed platform may have been drawn to the test-to-treat virtual care paradigm.
The analysis sample size was reduced by the exclusion of 27.0% of consented participants due to the completion of less than three surveys before being censored for missing more than two sequential surveys. Those excluded from the analyses tended to be younger, less vaccinated, and had a higher proportion of controls and non-white participants than the included group (Supplementary Table S10). Additionally, the association between NPR and rebound may be confounded by baseline risk, since the self-selected treatment group tended to be older and have more preexisting conditions. Given this was a prospective cohort there was also non-response or dropout during the study, whereby 67.6% of the analysis sample completed all acute study phase surveys, and 83.7% completed the final acute phase survey on Day 15.
In summary, this prospective observational study of COVID-19 positive outpatients showed that NPR treatment decision was associated with a substantial increased risk of both viral and symptom COVID-19 rebound. There was no difference found in the overall time to initial clearance of viral test positivity or symptoms between the NPR and control groups, although the NPR group included in our analyses was older and had more preexisting conditions at baseline. With this important caveat for interpretation of our findings, there is clearcut need for more durable antiviral therapies or regimens, and knowledge for optimal patient selection.