The Coronavirus Disease 2019 Rebound Study: A Prospective Cohort Study to Evaluate Viral and Symptom Rebound Differences in Participants Treated With Nirmatrelvir Plus Ritonavir Versus Untreated Controls

Abstract

Background

The uptake of nirmatrelvir plus ritonavir (NPR) in patients with coronavirus disease 2019 (COVID-19) has been limited by concerns around the rebound phenomenon despite the scarcity of evidence around its epidemiology. The purpose of this study was to prospectively compare the epidemiology of rebound in NPR-treated and untreated participants with acute COVID-19 infection.

Methods

We designed a prospective, observational study in which participants who tested positive for COVID-19 and were clinically eligible for NPR were recruited to be evaluated for either viral or symptom clearance and rebound. Participants were assigned to the treatment or control group based on their decision to take NPR. Following initial diagnosis, both groups were provided 12 rapid antigen tests and asked to test on a regular schedule for 16 days and answer symptom surveys. Viral rebound based on test results and COVID-19 symptom rebound based on patient-reported symptoms were evaluated.

Results

Viral rebound incidence was 14.2% in the NPR treatment group (n = 127) and 9.3% in the control group (n = 43). Symptom rebound incidence was higher in the treatment group (18.9%) compared to controls (7.0%). There were no notable differences in viral rebound by age, gender, preexisting conditions, or major symptom groups during the acute phase or at the 1-month interval.

Conclusions

This preliminary report suggests that rebound after clearance of test positivity or symptom resolution is higher than previously reported. However, notably we observed a similar rate of rebound in both the NPR treatment and control groups. Large studies with diverse participants and extended follow-up are needed to better understand the rebound phenomena.

In patients with acute COVID-19 infection, there was a similar rate of viral rebound between those treated with nirmatrelvir plus ritonavir versus untreated controls. Rebound after clearance of test positivity or symptom resolution was higher than previously reported.

Coronavirus disease 2019 (COVID-19) continues to be a global health crisis that is responsible for over 6 million reported deaths so far [1]. Nirmatrelvir plus ritonavir (NPR) is an oral antiviral medication for preventing progression of symptom severity in non-hospitalized patients [2, 3]. The EPIC-HR (Evaluation of Protease Inhibition for COVID-19 High-Risk Patients) study showed that high risk, unvaccinated participants treated with NPR had an 89% lower risk to develop severe symptoms, compared to placebo [4]. The government committed to purchasing 10 million NPR doses, followed by another tranche to support test-to-treatment programs [5]. However, NPR remains largely under-prescribed due to concerns around medication interactions and the peculiar COVID-19 rebound phenomenon.

Several publications have reported a return of COVID-19 symptoms or detectable viral load, after completion of the NPR treatment course [6–9]. This “rebound” phenomenon occurs in a subset of COVID-19 participants who have completed the 5-day course of NPR, subsequently testing negative for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus, and 2–8 days later experience a temporary return of symptoms with or without a positive rapid antigen or polymerase chain reaction (PCR) test [10]. Viral rebound and symptom rebound have also been seen in participants who did not receive any NPR treatment [11], imploring further investigation into this rebound phenomenon. There have been suggestions that rebound could be related to drug pharmacodynamics [12] or antiviral mediated interactions with viral immunologic response [13]; however, the peer-reviewed literature is limited to small, retrospective studies [7, 9, 14–16] and reports of rebound in untreated cohorts consists of small, prospective studies [11, 17].

Gaining a better understanding of rebound is important because it may result in unintended transmission after someone has tested negative, and a return of symptoms may discourage people from getting NPR to prevent severe illness. The purpose of this study is to prospectively compare the epidemiology of COVID-19 rebound in participants with acute COVID-19 infection who receive NPR compared to medication eligible controls who independently choose to not take NPR.

METHODS

Study Design and Recruitment

We designed a decentralized, prospective observational study developed in collaboration between Scripps Research Translational Institute (SRTI) and eMed, a virtual care platform that assists users through proctor-guided at-home testing that, when positive, triages them to a telemedicine visit where they may be prescribed NPR treatment, if eligible (this process is called “test-to-treat”). Individuals qualified for the study if they lived in the United States, spoke English, were ≥18 years of age, had a positive rapid antigen test for SARS-CoV-2 (verified by eMed), and were prescribed NPR through the telehealth visit regardless of whether they intended to take the medicine. Users who were offered NPR then received an email within hours with a link to the study materials, including informed consent.

Study Procedures and Assessments

Participants were split into 2 arms based on their independent decision to take NPR. All study tasks were identical in both arms (NPR and control; Figure 1). Enrolled participants completed a baseline demographic and preexisting conditions survey. Participants were shipped a kit overnight with 12 eMed telehealth proctored rapid antigen home tests. The antigen tests used for this study were BinaxNOW COVID-19 antigen home tests (Abbott Laboratories, Chicago, Illinois, USA). Ideally, participants in the NPR arm completed their first antigen test and first symptom surveys on days 2 and 5 of the 5-day NPR course (day 2 was the quickest the study kits could be delivered to participants) and then every other day through day 16. After the 16-day period, participants completed a persistent/long COVID symptoms survey at 1-, 3-, and 6-month intervals.

Figure 1.

Study design. After consent, participants were asked to take a COVID-19 rapid antigen test and a symptom survey on day 2, day 5, day 7, day 9, day 11, day 13, and day 15. Participants were then given an end of 16-day period survey, followed by a long COVID symptom survey at the 1-, 3-, and 6-month periods. Abbreviations: COVID-19, coronavirus disease 2019; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

End Points

The primary endpoint was the incidence of viral (testing) and symptom rebound within the NPR and control groups after acute COVID-19 infection. Secondary endpoints were time-to-test negativity, time to symptom resolution, and time from symptom onset to test negativity. Additionally, time from negative test to rebound positive test, and no symptoms to symptom rebound were calculated. Frequency of symptoms in both groups during the acute phase and the 1-month mark were evaluated.

Oversight

The study was conducted in accordance with Good Clinical Practice Guidelines and was approved by the Scripps Institutional Review Board. Electronic informed consent was obtained from all participants. Safety oversight was performed by SRTI. Study data were collected and managed using Research Electronic Data Capture (REDCap) tools [18, 19].

Statistical Analysis

For the incidence analysis, participants with <2 surveys and/or test results was excluded. Any participant with a positive rapid antigen test observed after a negative antigen test was categorized as having viral (testing) rebound, and everyone else in the cohort was categorized as having no viral rebound. Any participant who reported resolution of symptoms and then a recurrence on subsequent symptom surveys within the acute 15-day study period was categorized as having symptom rebound and everyone else was categorized as having no symptom rebound. Any participant reporting taking NPR at any time during their acute illness (regardless of completing their course or missing doses) was grouped into the treatment arm.

Kaplan-Meier survival curves, stratified by NPR arm, compared time to events in each arm. Participants were censored at the time of 2 missing tests or missing symptom surveys in a row (for test-related results or symptom-related results, respectively) or by day 15 (last test day) if they did not have an event. A log-rank test was used to compare the survival curves of the NPR and control groups.

Participant characteristics and COVID-19 recovery categories were compared (using a χ² test for categorical and t test for continuous variables) between the 2 groups as well as differences between participants who did and did not rebound, regardless of treatment group and reported symptoms at month 1 (Tables 1–3). All analyses were done in SAS version 9.4. Of note, because this is the first report to our knowledge of our ongoing study, the current report is not fully powered to robustly evaluate statistical significance. Thus, we do not attempt to draw statistical conclusions of group differences but focus, rather, on reporting observations and incidence.

Table 1.

Participant Characteristics and COVID-19 Recovery by NPR Treatment, Frequency (%)

	NPR Treatment (n = 127)	Control (n = 43)	P value
Participant characteristics	…	…
Race/ethnicity	…	…
White	111 (88.1)	26 (61.9)	.0002
African American	9 (7.1)	8 (19.1)	.03
Asian	5 (4.0)	3 (7.1)	.40
Latino	10 (7.9)	6 (14.3)	.22
Hawaiian/Pacific Islander	1 (0.8)	0 (0.0)	.56
American Indian	5 (4.0)	3 (7.1)	.40
Age Category (years old)	…	…	.42
18–44	51 (40.2)	22 (51.2)
45–64	58 (45.7)	17 (39.5)
≥65	18 (14.2)	4 (9.3)
Gender	…	…	.48
Female	72 (56.7)	27 (62.8)
Male	55 (43.3)	16 (37.2)
Preexisting conditions	…	…
Asthma	21 (16.5)	8 (18.6)	.76
Cancer	6 (4.7)	1 (2.3)	.49
Heart disease	3 (2.4)	2 (4.7)	.44
Heart failure	0 (0.0)	1 (2.3)	.08
COPD	0 (0.0)	1 (2.3	.08
High blood pressure	34 (26.8)	12 (27.9)	.88
Chronic bronchitis	1 (0.8)	1 (2.3)	.42
Other lung condition	2 (1.6)	0 (0.0)	.41
Diabetes	10 (7.9)	2 (4.7)	.48
Autoimmune condition	12 (9.5)	1 (2.3)	.13
Emphysema	0 (0.0)	1 (2.3)	.08
None of the above	56 (44.1)	25 (58.1)	.11
Received at least 1 COVID-19 vaccine dose	121 (95.3)	41 (95.3)	.98
COVID-19 recovery	…	…
Viral (testing) rebound	18 (14.2)	4 (9.3)	.41
Symptom rebound	24 (18.9)	3 (7.0)	.06
Consistently positive	5 (3.9)	0 (0.0)	.20
First positive test to first negative test, days (SD)	7.1 (3.4)	7.0 (3.9)	.85
Symptom start to first day of no symptoms, days (SD)	10.5 (4.9)	10.7 (4.7)	.80
Symptom start to test negative, days (SD)	6.8 (3.4)	6.1 (2.9)	.25

Abbreviations: COPD, chronic obstructive pulmonary disease; COVID-19, coronavirus disease 2019; NPR, nirmatrelvir plus ritonavir; SD, standard deviation.

Table 3.

Symptoms Reported at Month 1 Between NPR Treatment and Control Groups, Frequency (%)

	NPR Treatment (n = 109)	Control (n = 34)	P value
Palpitations	3 (2.8)	0 (0.0)	.33
Neurological	5 (4.6)	0 (0.0)	.20
Sleep Disturbance	2 (5.9)	2 (1.8)	.21
Mental	3 (2.8)	0 (0.0)	.33
Menstrual	1 (0.9)	0 (0.0)	.58
Headache	10 (9.2)	2 (5.8)	.55
Body aches	9 (8.3)	4 (11.8)	.53
Cough	19 (17.4)	7 (20.6)	.68
Congestion	11 (10.1)	6 (17.7)	.23
Sore Throat	4 (3.7)	2 (5.9)	.57
Headache	10 (9.2)	2 (5.9)	.55
Fatigue	19 (17.4)	8 (23.5)	.43
SOB	9 (8.3)	1 (2.9)	.29
Tinnitus	1 (0.9)	1 (2.9)	.38

Abbreviations: NPR, nirmatrelvir plus ritonavir; SOB, shortness of breath.

RESULTS

Participants

Between 4 August 2022 and 1 November 2022, 247 participants consented to participate in the study, and 188 participants had completed the 16-day study procedures. Participants with <2 completed surveys or test results (or who completed the surveys outside the outlined survey schedule) were excluded from the analysis, leaving 170 participants: 127 participants in the NRP arm and 43 in the control arm for the analysis.

The majority of study participants were under age 65 (86% of the NPR group; 91% of the control group), and roughly 70% of participants in each group had at least 1 preexisting condition. In the NPR group, 104 (82%) participants were fully vaccinated (ie, received at least 1 booster after the initial dose), 17 (13%) were partially vaccinated (ie, received initial dose with no booster), and 6 (5%) were unvaccinated. In the control group, 30 (70%) participants were fully vaccinated, 11 (25%) were partially vaccinated, and 2 (5%) were unvaccinated. The NPR and control arms had no notable differences by age, gender, or preexisting conditions (Table 1). There was a higher frequency of White participants in the NPR group compared to controls (P = .0002). Five participants (4%) in the NPR treatment group remained positive throughout the 15-day period and thus were not eligible for evaluation for rebound and were excluded from the tables but were included in the time to viral clearance and symptom resolution analyses.

Virus and Symptom Clearance

Time to viral clearance, defined as time from first positive antigen test to first negative antigen test was similar in the NPR treatment and control groups (mean 7.1 days vs 7.0 days; P = .85). Similarly, time from symptom onset to first symptom resolution (mean 10.5 days vs 10.7 days; P = .80), and time from symptom onset to first negative antigen test (mean 6.8 days vs 6.1 days; P = .25) was similar in the NPR treatment and control group, respectively (Figure 2 and Table 1). Additionally, these durations included participants who were still positive or symptomatic at day 15 (the end of follow-up), which makes the numbers an underestimate of the true durations.

Figure 2.

Survival curves comparing NPR treatment and control groups. Comparisons of (A) first positive test to first antigen negative test; (B) symptom start to first report of no symptoms; (C) symptom start to first antigen negative test. Horizontal axes are in days. Abbreviation: NPR, nirmatrelvir plus ritonavir.

Although not a primary outcome of this study, our study design allowed an additional evaluation of the proportion of participants who still tested positive at day 5, a time when current Centers for Disease Control and Prevention (CDC) guidance would have COVID-19 infected individuals exit isolation: either 5 days from first recorded antigen test positivity, or 5 days after symptom onset. Approximately 60% of participants in both groups remained positive after 5 days from first antigen test positive result, and approximately 50% of participants in both groups remained antigen test positive at 5 days after symptom onset (Figure 2A and 2C). Nearly 20% of participants in both groups remained positive 10 days after first turning antigen test positive.

Rebound Incidence

Virus testing rebound incidence trended higher in the NPR treatment group (18/127; 14.2%) than in the control group (4/43; 9.3%) (P = .41) (Figure 3A and Table 1). Symptom rebound incidence was notably higher in the NPR treatment group (18.9%) than the control group (7.0%) (P = .06). The differences between groups for viral rebound and symptom rebound were not statistically significant. Participants who rebounded reported less body ache compared to those who did not rebound. There were no notable differences in viral rebound by age, gender, preexisting conditions, or symptom groups during the 16-day follow-up period. When comparing participants with viral rebound versus those who did not have viral rebound, frequency of Asian and Native American participants was lower among those with rebound (Table 2). Furthermore, no notable differences were observed in the NPR treatment and control group symptoms at the 1-month period (Table 3). Out of 143 participants who reported whether they sought care at their 1-month survey, 1 participant in the control group (who did not experience viral rebound) reported they were hospitalized. There were no deaths reported in either group.

Figure 3.

Survival curves between NPR treatment and control groups (A). Time from negative test to viral rebound (testing positive) grouped by NPR treatment. Only includes participants with event (tested negative after initial positives) from Figure 2A. (B) Time from no symptoms to symptom rebound grouped by NPR treatment. Only includes participants with event (reported no symptoms after initially having symptoms) from Figure 2B. Horizontal axes are in days. Abbreviation: NPR, nirmatrelvir plus ritonavir.

Table 2.

Participant Characteristics by Viral Rebound (Includes NPR Treatment and Control Groups), Frequency (%)

	Rebound (n = 22)	No Rebound (n = 148)	P value
Race/ethnicity	…	…
White	15 (71.4)	122 (83.0)	.20
African American	2 (9.5)	15 (10.2)	.92
Asian	3 (14.3)	5 (3.4)	.03
Latino	2 (9.5)	14 (9.5)	1.00
Hawaiian/Pacific Islander	0 (0.0)	1 (0.7)	.70
American Indian	3 (14.3)	5 (3.4)	.03
Age category (years old)	…	…	.09
18–44	9 (40.9)	64 (43.2)
45–64	7 (31.8)	68 (46.0)
≥65	6 (27.3)	16 (10.8)
Gender	…	…	.19
Female	10 (45.5)	89 (60.1)
Male	12 (54.6)	59 (39.9)
Preexisting conditions	…	…
Asthma	2 (9.1)	27 (18.2)	.29
Cancer	2 (9.1)	5 (3.4)	.21
Heart disease	0 (0.0)	5 (3.4)	.38
Heart failure	0 (0.0)	1 (0.7)	.70
COPD	0 (0.0)	1 (0.7)	.70
High blood pressure	7 (31.8)	39 (26.4)	.59
Chronic bronchitis	0 (0.0)	2 (1.4)	.58
Other lung condition	1 (4.6)	1 (0.7)	.12
Diabetes	1 (4.6)	11 (7.4)	.62
Autoimmune condition	1 (4.6)	12 (8.1)	.56
Emphysema	0 (0.0)	1 (0.7)	.70
None of the above	10 (45.5)	71 (48.0)	.83
Symptoms	…	…
Body ache	6 (27.3)	86 (58.1)	.007
Symptom groups	…	…
Systemic	15 (68.2)	124 (83.8)	.08
Neurologic	12 (54.6)	98 (66.2)	.29
Gastrointestinal	6 (27.3)	46 (31.1)	.72
Respiratory	22 (100.0)	146 (98.7)	.58
Received at least 1 COVID-19 vaccine dose	22 (100.0)	140 (94.6)	.26

Abbreviations: COPD, chronic obstructive pulmonary disease; COVID-19, coronavirus disease 2019; NPR, nirmatrelvir plus ritonavir.

Rebound Duration and Timing

In the control group, among the 3 individuals with symptom rebound, 2 had a symptom rebound that lasted <5 days and 1 had a rebound that was 5 days or longer. In the NPR treatment group, among the 24 with symptom rebound, 10 (42%) had a symptom rebound that was <5 days, 10 (42%) had symptom rebound that was 5 days or more, and 4 (16%) had multiple episodes of symptom rebound during the 16-day follow-up period.

Among the participants with symptom rebound, 13/27 (48%) reported a return of symptoms within 2 days of initial symptom resolution and among participants with viral rebound, 14/22 (64%) tested positive again within 2 days of initial viral clearance. Additionally, among the 18 and 4 participants with viral rebound in the NPR treatment and control groups, respectively, rebound duration was slightly increased in the NPR treatment group versus the control group (4.4 days vs 2.0 days, respectively). Three of the 24 participants had more than 1 episode of viral testing rebound, all of them in the NPR treatment group. Because some participants were still in a viral rebound at day 15, the end of test follow-up, these average durations are an underestimate.

DISCUSSION

The current indication for NPR is to treat mild-to-moderate COVID-19 in adults with positive SARS-CoV-2 viral testing and who are at a high risk for progression, including hospitalization or death. Our study is 1 of the first to our knowledge to prospectively evaluate COVID-19 rebound with a study population that was acutely positive with COVID-19 and comparable in age and gender to the EPIC-HR study; however, the vaccination status between the current study's cohort and the EPIC-HR cohort (the latter unvaccinated) is an important distinction that must be factored when comparing results [4]. This study demonstrates the feasibility of offering a virtual test-to-treatment approach, collecting patient-reported outcomes and clinical outcomes in a decentralized fashion, which complements public health prevention measures.

This study demonstrated an overall viral testing rebound incidence of 14% and a symptom rebound incidence of 19% among participants treated with NPR. Both incidences are higher than have been reported in prior retrospective studies, which ranged from 2% to 6% [20, 21]. However, we also show that both viral (9%) and symptom rebound (7%) occurs in the absence of treatment with NPR. Testing rebound in the control group in this study matches reported viral rebound incidence in other studies of untreated patients with COVID-19, which was 12% [11, 17]. A notable finding was that symptom rebound in the control cohort (7%) was lower than in the NPR treatment arm (19%). Although the small sample size limits our ability to draw statistical inference, this finding could be driven by the pharmacology of NPR. The current predominant hypothesis around rebound is that the immune system may not have the opportunity or need to fully ramp up upon infection with the virus, because NPR might suppress replication early in disease [10]. However, in our cohort, NPR treatment was not associated with a shorter time to symptom resolution or testing negative. Although viral rebound (ie, testing positive after converting to negative) has been much discussed as an NPR-related phenomenon, this study did not find a large difference in rebound based on a return to antigen test positivity. It is possible that there is not an increase in viral rebound after taking the medication, but rather more rebound is identified in this group due to increased testing owing to an increase in symptom rebound.

In both the NPR treatment and control groups, the rebound symptoms are usually milder than symptoms of the COVID-19 infection [22]. Although NPR treatment reduces severe outcomes [4], the relatively high incidence of symptom rebound in the treatment group suggests that further investigation remains needed around the changes caused by NPR in the virologic and immunologic milieu of the pathogen and host.

We identified several interesting signals (acknowledging the study was not powered for statistical inference) in participant characteristics and acute phase symptoms among the NPR treatment and control groups. There was significantly higher NPR uptake among White participants and less among Black participants, which has also been noted by the CDC [20]. We also saw a significantly lower frequency of Asian Americans experiencing viral rebound and less rebound in participants who did not report body aches during the acute phase. In the future, identifying risk factors for developing rebound could help tailor treatment duration and testing guidance in specific subpopulations that have higher risk for rebound.

Positive rapid antigen tests are associated with levels of virus that remain infectious. In this study we observed nearly 60% of participants remained positive on a rapid antigen test at 5 days after first turning positive or becoming symptomatic, and 20% of participants remained positive even 10 days into infection. Thus, our study agrees with others that have found that the 5-day isolation period, as currently recommended by the CDC, is not adequate.

For our next phase, we will collect testing swabs for viral sequencing as well as serum from participants to understand any virus specific or host specific factors that provide insight into COVID-19 rebound. Retrospective analyses have reported the lack of an association between viral load rebound and low nirmatrelvir exposure or resistance to nirmatrelvir; however, they were done during the B.1.617.2 (Delta) variant rather than the B1.1.529 (Omicron) variant [23]. Notably, the EPIC-HR study focused on viral load based rebound only, which does not translate into the presence of infectious virus, and more importantly does not include symptom rebound [4]. The strengths of our study include prospectively collected data during the predominant omicron wave using a decentralized study model, which makes it pragmatic.

The primary limitations of the study were the unbalanced sample size in the control cohort and the largely White population. Specifically, based on the smaller sample size of the control group, we recognize the prevalence of viral rebound may be different in a larger study population. Additionally, 31% of consented study participants were excluded from the analysis due to missing test results and/or surveys, which resulted in a lower-than-expected sample size.

Because participants only needed 2 or more tests and surveys to be included in the incidence analysis, missing data likely biased our results to show a lower incidence of rebound (Table 1). For the survival analysis, 13 participants were censored when evaluating time to test negative (8 had 2 or more missing tests in a row, 5 still positive at day 15), and 69 were censored for time to symptom resolution (16 for having 2 or more missing symptom surveys in a row, and 53 were still symptomatic through day 15). The more stringent censoring for survival analysis may have resulted in higher rebound estimates by decreasing the denominator. A longer follow-up time may also be necessary to better understand rebound, because many participants remained symptomatic throughout the 15-day follow up period (Figure 2).

However, the primary goal of the current study was to understand the incidence of COVID-19 rebound and therefore we accepted the cohorts in a convenience sampling approach. Other sources of bias include recruitment through the eMed platform, which could introduce a selection bias to eMed users. Furthermore, participants were instructed to start testing right after they started taking NPR; however, some participants waited a few days to begin testing or started NPR after testing, which we adjudicated as best as possible. Finally, to balance participation burden with adherence, we asked participants to do every other day testing instead of daily testing and surveys, which could have led to missing data points. Despite the limitations, this study adds vital incidence information for COVID-19 rebound and spotlights the need to better understand the rebound phenomenon of our main pharmaceutical therapeutics options.

CONCLUSION

This preliminary report of our prospective study of both virus test-positive and symptomatic rebound suggests that rebound after clearance of test positivity or symptom resolution is higher than previously reported. Interestingly, however, we observed that the high rate of virus rebound is both in the NPR treatment and control groups, although there was a notably higher frequency of symptom rebound among those treated with NPR. Large studies with diverse participants and extended follow-up are needed to better understand the rebound phenomena.

Notes

Financial support. This work was supported by eMed, the National Institute of Allergy and Infectious Diseases (grant number 3U01AI151812-03S2), and National Center for Advancing Translational Sciences (grant number NCATS UL1 TR002550).