Real-world use of nirmatrelvir–ritonavir: who benefits?

The oral antiviral combination nirmatrelvir–ritonavir has become a first-line therapy in many countries for non-hospitalised adults with COVID-19. A trial done from July to December, 2021, showed that, when given within 5 days of symptom onset, nirmatrelvir–ritonavir reduced the risk of COVID-19-related hospitalisation compared with placebo among unvaccinated patients at high risk of serious illness.¹ Since nirmatrelvir–ritonavir was authorised in the USA in December, 2021, the landscape of the pandemic has changed: omicron lineage variants have emerged, and there has been widespread vaccination and SARS-CoV-2 infection, lessening the likelihood that patients will progress to severe disease. The number of hospitalisations, admissions to intensive care units, and deaths due to Omicron variants is only a small fraction of that associated with the delta variant (in June–December, 2021).² In these circumstances, whether treatment with nirmatrelvir–ritonavir confers a substantial treatment benefit—particularly when omicron subvariants are the dominant circulating variants and among patients who are vaccinated, or who have been previously infected—has arisen as a key question for clinicians and policy makers.

In The Lancet Infectious Diseases, Joseph A Lewnard³ and colleagues report data for use of nirmatrelvir–ritonavir among outpatients with COVID-19 in the Kaiser Permanente Southern California health-care system between April 8 and Oct 7, 2022, a time when omicron lineages (BA.2, BA.2.12.1, BA.4, and BA.5) were dominant. In this retrospective cohort study, outpatients with a positive PCR test for SARS-COV-2 (their index test) who were dispensed nirmatrelvir–ritonavir (n=7274) were matched with outpatients who also tested positive for SARS-COV-2 but were not given nirmatrelvir–ritonavir (n=126 152). 90 129 (67·5%) of 133 426 patients' index test was within 5 days of symptom onset, and 114 208 (85·6%) had received at least two COVID-19 vaccine doses. The authors used Cox proportional hazard models to calculate the treatment effectiveness of nirmatrelvir–ritonavir in preventing 30-day all-cause hospital admission and death. Treatment effectiveness was 79·6% (95% CI 33·9–93·8) when nirmatrelvir–ritonavir was dispensed within 5 days of symptom onset, but only 53·6% (6·6–7·7) when it was dispensed at any time irrespective of symptom onset. It is important to note the low frequency of 30-day hospitalisation (641 [0·5%]) or death (164 [0·1%]) among untreated patients. Thus, there might be limited potential for absolute treatment benefits in a highly vaccinated population infected with omicron variants of SARS-CoV-2, with a number needed to treat of 100–200 to prevent one hospitalisation or death. The rates of 30-day hospitalisation and death reported by Lewnard and colleagues are notably lower than those reported for a subgroup of vaccinated patients with more than one risk factor for progression in the EPIC-SR trial, which was terminated early because of the low event rate.⁴

Lewnard and colleagues use data from an integrated health system to provide valuable insights into the use of nirmatrelvir–ritonavir in a real-world context. Importantly, these are the first observational data that include timing of symptom onset—a key limitation of previous studies.5, 6 Although the reliability of responses entered at the time of test order could be limited by factors such as recall and ascertainment bias, the clear relationship between duration of symptoms and the effectiveness of nirmatrelvir–ritonavir supports the premise that earlier antiviral treatment is associated with greater clinical benefit, highlighting the need for accessible rapid test-to-treat programmes. Furthermore, Lewnard and colleagues' data show that real-world prescribing differs substantially from that in trial settings and the current authorisation criteria. 1802 (24·8%) patients to whom nirmatrelvir–ritonavir was dispensed had symptoms for more than 5 days or were asymptomatic—populations in which nirmatrelvir–ritonavir has no proven benefit and in which the US Food and Drug Administration has not authorised the treatment's use. The investigators accounted for health care use in the previous year (including outpatient visits and vaccination status for other respiratory infections) in their estimates to attenuate bias related to care-seeking behaviors. However, the requirement of a positive SARS-COV-2 test for study inclusion could have introduced substantial selection bias, because other studies show that up to 80% of patients who receive treatment have missing tests in electronic health records.⁵

How do Lewnard and colleagues' findings affect use of nirmatrelvir–ritonavir? The results of the UK-based PANORAMIC platform trial (ISRCTN30448031), in which nirmatrelvir–ritonavir has reportedly been provided to more than 6000 people, are eagerly awaited. Further randomised trials of nirmatrelvir–ritonavir are unlikely given the question of equipoise and the substantial time and cost of such trials. In the absence of new trial data, several real-world studies5, 6, 7 have suggested that nirmatrelvir–ritonavir is associated with a reduced risk of progression to severe disease across several omicron variants. However, it has become clear that the absolute reduction in risk provided by treatment has decreased substantially, which greatly increases the cost of preventing one hospitalisation. Studies of whether nirmatrelvir–ritonavir affects additional patient-centred outcomes, such as post-COVID-19 condition (also known as long COVID), are planned and could affect the cost-effectiveness of the intervention and whether continued widespread use is merited as the pandemic evolves. Characterisation of patients who most benefit from treatment with nirmatrelvir–ritonavir and studies of when treatment is most effective are needed.

© 2023 Presidencia de la República Mexicana

This online publication has been corrected. The corrected version first appeared at thelancet.com/infection on March 29, 2023

KCM reports grants from the US National Center for Advancing Translational Sciences, the National Institute of Child Health and Human Development, and National Heart, Lung, and Blood Institute. AAG reports investigator-initiated grants from the US National Institutes of Health, Centers for Disease Control and Prevention, and Department of Defense, AbbVie, and Faron Pharmaceuticals. He served on a US National Institutes of Health data and safety monitoring board.