SARS-CoV-2 Rebound With and Without Use of COVID-19 Oral Antivirals

Early treatment with a first-line therapy (nirmatrelvir/ritonavir [Paxlovid] or remdesivir) or second-line therapy (molnupiravir) prevents hospitalization and death among patients with mild-to-moderate COVID-19 who are at risk for severe disease and is recommended by the National Institutes of Health COVID-19 Treatment Guidelines. On May 25, 2023, the Food and Drug Administration approved nirmatrelvir/ritonavir for treatment of adults at high risk for severe disease. Although antiviral therapies are widely available, they are underutilized, possibly because of reports of SARS-CoV-2 rebound after treatment. To enhance current understanding of rebound, CDC reviewed SARS-CoV-2 rebound studies published during February 1, 2020– November 29, 2023. Overall, seven of 23 studies that met inclusion criteria, one randomized trial and six observational studies, compared rebound for persons who received antiviral treatment with that for persons who did not receive antiviral treatment. In four studies, including the randomized trial, no statistically significant difference in rebound rates was identified among persons receiving treatment and those not receiving treatment. Depending on the


Introduction
COVID-19 has caused approximately 6.5 million hospitalizations and 1.1 million deaths in the United States.*Institutes of Health (NIH) COVID-19 Treatment Guidelines (1).The two oral antivirals, nirmatrelvir/ritonavir and molnupiravir, are widely available but underutilized (5).The limited use of these antivirals might be partially attributable to reports of rebound after treatment, especially with nirmatrelvir/ritonavir. § However, rebound was reported before the advent of COVID-19 antivirals and was related to immunity and individual level factors (6,7).

SARS-CoV-2 Rebound
SARS-CoV-2 rebound is typically described as recurrence of signs or symptoms or a new positive viral test result after initial recovery from COVID-19.In May 2022, CDC issued a health advisory alert that described case reports of SARS CoV-2 rebound among patients who completed the recommended 5-day course of nirmatrelvir/ritonavir and noted that rebound was also described among persons who were not treated.¶ On May 25, 2023, the Food and Drug Administration (FDA) approved nirmatrelvir/ ritonavir, which was authorized for emergency use in December 2021, for treatment of mild to moderate COVID-19 among adults aged ≥18 years who are at high risk for severe disease.**In their review of data from Evaluation of Protease Inhibition for COVID-19 in High Risk Patients (EPIC-HR), a phase 2/3 § https://www.medscape.com/viewarticle/987121?form=fpf ¶ https://stacks.cdc.gov/view/cdc/117609** https://www.fda.gov/news-events/press-announcements/fda-approves-first-oralantiviral-treatment-covid-19-adultsrandomized controlled trial that examined the efficacy of nirmatrelvir/ritonavir, FDA concluded that there was no consistent association between treatment and rebound (8).To enhance current understanding of rebound, CDC reviewed recent literature comparing rebound among COVID-19 patients who did and did not receive antiviral treatment.
(Table 1) (Supplementary Table, https://stacks.cdc.gov/view/cdc/137156).Seven studies compared rates of rebound among patients who did and did not receive COVID-19 antiviral treatment (Table 2) (10)(11)(12)(13)(14)(15)(16).Findings from two studies examining infectivity, resistance, and immune response were summarized (11,17).Individual case data from three studies that used the same definition of viral rebound were examined to estimate days to onset of viral rebound and rebound duration (18)(19)(20).Median days to rebound and resolution of acute and rebound illness were calculated.Pearson's chi-square or Fisher's exact test were used to compare proportions for studies that did not report the test statistic.This activity was reviewed by CDC, deemed not research, and was conducted consistent with applicable federal law and CDC policy.§ §
Initial analysis of EPIC-HR trial data showed that viral rebound rates were low and similar between the treated and untreated groups (Table 2) (10).In addition, rebound was not associated with low nirmatrelvir/ritonavir levels, hospitalization or death, severe symptom relapse, vaccination or serologic status, or emergent mutations (8,10).

Infectivity, Resistance, and Immune Response
One observational study demonstrated that duration of shedding of infectious virus was longer among persons with rebound (14 days) compared with those without rebound (3 days), but found no evidence of resistance-associated mutations using genomic sequencing (11).Another study of biomarkers among six patients with rebound after treatment with nirmatrelvir/ritonavir demonstrated that a robust immune response was present during rebound, likely reducing risk for disease progression (17).This study also found no evidence of resistance.

Onset and Duration of Rebound
Among 22 patients (from three studies) with available virologic data and who received treatment, median time to negative test results was 6 days (IQR = 5-7 days) after initial positive test result (18-20) (Figure 2).Median time to viral rebound was 9 days (IQR = 9-13 days) after diagnosis, and to resolution was 16 days (IQR = 16-19 days) into the viral illness.Rebound occurred during the course of illness when there was variability in viral load because of host factors (21).

Discussion
Current evidence, including randomized controlled trial and observational data, suggests that SARS-CoV-2 rebound occurs initially as a mild illness 3-7 days after resolution of the initial acute illness, occurs in both treated and untreated patients, and is not associated specifically with receiving nirmatrelvir/ ritonavir.Moreover, rebound occurs when there is variable,  host-mounted immune response to infection during the course of illness.Finally, no hospitalizations or deaths were reported among outpatients who experienced rebound.Some observational studies demonstrated a higher frequency of rebound among treated persons (10%-14%) (11,14,22) than reported by the randomized controlled trial, EPIC-HR (8,10) (Supplementary Table, https://stacks.cdc.gov/view/cdc/137156).Viral rebound might occur in persons on antiviral treatment because they are at high risk for severe disease and might have host factors, such as immunosuppression, that contribute to the natural variability in viral dynamics (21).Risk factors for rebound appear to be similar to risk for severe disease, but further studies are needed to understand whether persons with certain characteristics or underlying medical conditions are predisposed to experiencing rebound.Another important consideration is that persons receiving antiviral treatment might be at higher risk for experiencing rebound given the viral suppression related to use of treatment early in the disease course and resumption of viral replication after completion of treatment because of delayed viral clearance.This elevated risk could be due to early discontinuation of antiviral treatment or the need for longer courses of treatment among certain persons, such as those who are immunocompromised (14).Two ongoing clinical trials of nirmatrelvir/ ritonavir will further characterize the frequency of rebound after different durations of nirmatrelvir/ritonavir treatment among immunocompromised subjects ¶ ¶ and the potential ¶ ¶ https://clinicaltrials.gov/ct2/show/NCT05438602 benefit of nirmatrelvir/ritonavir retreatment among subjects with posttreatment rebound.***Rebound does not likely represent reinfection or resistance to treatment (12); however, further studies are needed to confirm this finding.The FDA analysis identified potential treatmentassociated mutations that were not clinically relevant among two treated patients because rebound symptoms resolved without hospitalization (8).It is important to ensure that use of antivirals does not accelerate viral evolution and result in resistant mutations, such as through counseling patients to complete antiviral treatment and monitoring for resistance using molecular analyses.Two studies demonstrated shedding of infectious virus during rebound (8,11).Comparisons of genomic strains present in both acute and rebound episodes and viral culture to determine infectiousness are important to understanding the clinical implications of rebound.In addition, a large assessment of innate and adaptive immunity and monitoring biomarkers of inflammation and cytokine storm would contribute to understanding of the underlying pathophysiology of recurrence.

Limitations
The findings in this report are subject to at least five limitations.First, standardized definitions for symptom, viral, and clinical rebound were not used across studies.Using standard definitions to accurately reflect outcomes could improve interpretability and comparisons of data across studies and settings.Most studies examined symptom or viral rebound.A definition that requires reemergence of virus after complete resolution of illness, which takes 7-10 days for a healthy adult, and a negative viral test result after resolution of initial symptoms would allow for examination of clinical implications of rebound or recrudescence, such as a dysregulated immune response (23).
Second, publications about recurrences and viral kinetics might have been missed given the narrow search.Third, a major limitation of observational studies is the difficulty in verifying whether antiviral treatment courses were completed and whether vaccination status and previous infection were documented accurately.Fourth, few studies correlated symptoms with viral load, which makes the significance of recurrence of mild symptoms difficult to understand because symptoms are subjective and might not represent viral reactivation.Finally, ascertainment bias is also possible given that persons receiving antiviral treatment are closely followed, and more likely to report recurrent symptoms, which would explain the early case reports being associated with nirmatrelvir/ritonavir, the most commonly used oral antiviral in the United States.
factors, such as immunosuppression, delayed viral clearance, and overall immune response.The current literature review, along with a recently published randomized trial (8), suggests the substantial benefit of antiviral treatment among persons at risk for severe disease outweighs the risk for rebound, because rebound resolves quickly and is not associated with an increase in severity of recurring signs and symptoms.Increased education and awareness among practitioners and patients about rebound not increasing risk for hospitalization or death might increase use of COVID-19 treatment.According to NIH COVID-19 Treatment Guidelines, rebound should not deter providers from prescribing life-saving antiviral treatments when indicated to prevent morbidity and mortality from COVID-19 (1).

TABLE 2 . Summary of seven SARS-CoV-2 rebound studies among persons who did and did not receive antiviral treatment with nirmatrelvir/ ritonavir or molnupiravir -February 1, 2020-November 29, 2023* Study author, year Study type Definition of rebound Sample size Treatment Rebound prevalence, % (no./No.) p-value Study conclusions and key limitations With treatment Without treatment Anderson et al., 2022 †
1362 US Department of Health and Human Services | Centers for Disease Control and Prevention | MMWR | December 22, 2023 | Vol.72 | No. 51

Timing of viral rebound and resolution during SARS-CoV-2 infection among 22 patients* , † -February 1, 2020-November 29, 2023
Median time to negative test result was defined as day of first negative viral test result (polymerase chain reaction or antigen) after initial positive test result.Viral rebound was defined as the first positive viral test result (polymerase chain reaction or antigen) after a negative test result.Resolution was defined as first negative viral test result after day 1 of viral rebound.† Timing and duration of viral rebound generated using data from 22 patients in three studies that used a virologic definition of rebound and had complete data: * https://doi.org/10.1093/cid/ciac512,https://doi.org/10.1056/NEJMc2206449,and https://doi.org/10.1016/j.jinf.2022.06.011.