Who should receive oral antiviral therapy for severe acute respiratory syndrome coronavirus 2 infection in the omicron era? Choose wisely

The COVID-19 pandemic, which emerged at the end of 2019, has caused millions of fatalities worldwide. The development and clinical introduction of antiviral drugs directed against the SARS-CoV-2 virus have to be considered as a milestone to combat this pandemic and to treat this infection. Antiviral therapy is most effective when given early in the course of COVID-19 [1,2]. Accordingly, two orally applicable drugs, nirmatrelvir/ritonavir and molnupiravir have demonstrated, in two double-blinded randomized controlled trials, that they significantly reduce the risk of hospitalization or death in patients who are unvaccinated and ambulatory with risk factors for severe infection, when given within the first days after onset of COVID-19 related symptoms [3,4]. The numbers needed to treat (NNT) to avoid one event (hospitalization or death) were between 1:19 for nirmatrelvir/ritonavir and 1:35 for molnupiravir. However, these studies were carried out in 2021 when alpha and delta variants of SARS-CoV-2 were dominant. Since then, the clinical picture of COVID-19 has dramatically changed with the emergence of different Omicron variants resulting in milder courses of the disease and reduced need for hospitalization, intensive care unit (ICU) admission or death (0.3%, 0.1%, 0.03% respectively) than observed for the delta variant (2.8%, 1.0%, 0.5% respectively) [5]. This general reduction in hospitalization because of COVID-19 can also be attributed to increased immune protection in the population based on broad vaccination, previous infection or a combination of both. Of note, the presence of serum antibodies against SARS-CoV-2 resulted in an almost complete disappearance of the protective effect of nirmatrelvir/ritonavir and molnupiravir in the studies mentioned above [3,4].

Thus, the question emerges of who might still benefit from such antiviral treatment in the current situation. To this end, data from the current Omicron wave have been urgently awaited. Arbel et al. [6] report on a retrospective analysis of data extracted from a healthcare database in Israel on the effects of nirmatrelvir/ritonavir treatment on the risk of hospitalization and death from COVID-19 caused by the Omicron variant. The authors found that early treatment with nirmatrelvir/ritonavir resulted in reduced hospitalization and death in patients at risk and aged >65 years as compared with those without therapy, whereas patients aged <65 years had no benefit from the therapy. The latter raises the question of the causes of hospitalization and death because severe infection with lung failure and hyperinflammation syndromes became very infrequent during the Omicron wave owing to the waning pathogenicity of the virus and immune protection by vaccination coverage and/or natural immunity [7]. This is also reflected by the low hospitalization rate reported in this study [6]. Nowadays many patients are hospitalized “with” COVID-19 and another leading diagnosis but not “because of” COVID-19, which makes an important distinction regarding the evaluation of antiviral therapy efficacy. Thus, we need an estimate on the NNT to prevent one severe SARS-CoV-2 infection resulting in hospitalization or death in the current pandemic situation, and we need to know who may specifically benefit from such therapy. This retrospective analysis also indicated that people who were unvaccinated and less frequent in the group receiving nirmatrelvir/ritonavir, were at a 55.79-fold higher risk of being hospitalized with SARS-Cov-2 positivity [6]. This undermines the importance of immune protection by vaccination and/or previous infection for the prevention of severe disease.

Nirmatrelvir/ritonavir has several drug interactions and transient cessation of essential drugs, including f.e. antiarrhythmic or anticoagulant medications, to enable antiviral therapy that may eventually cause serious health problems. Thus, risk-benefit estimation on an individual basis taking into account immune status, vaccination coverage and a combination of risk factors, which could put the patients at a higher risk for hospitalization or death, is needed.

Data on the real-life effects of nirmatrelvir/ritonavir and molnupiravir were reported recently from the United States of America (USA) and Hong Kong [8,9]. In a retrospective analysis of healthcare system data in Massachusetts and New Hampshire, USA, the effect of nirmatrelvir/ritonavir therapy was accessed in nonhospitalized adults aged >50 years with COVID-19 and no contraindication against the use of this antiviral medication [9]. The outcome was hospitalization within 14 days after the initial diagnosis of COVID-19.6036 patients receiving antiviral therapy were compared with 24,286 matched patients not receiving specific COVID-19 therapy. Although patients prescribed nirmatrelvir/ritonavir were significantly older, patients in the control group had a significantly poorer vaccination status and lower prevalence of vaccine boosters. Although the overall incidence of hospitalization was <1%, nirmatrelvir use was associated with a 45% reduction in hospitalization [8]. According to these data, the NNT to avoid one hospitalization is >200, and it is questionable whether or not hospitalization was really due to COVID-19 or, according to the milder course of the infection, “with COVID-19” and another leading diagnosis. Finally, it is not clear to which extent the reduction in hospitalization was at least partly because of better vaccination status in the treatment group. Given the limited beneficial effects of the antiviral drugs in SARS-CoV-2 seropositive individuals [3,4], this would suggest that in patients with effective vaccination and/or natural immunity, the NNT to prevent SARS-CoV-2–induced hospitalization is far higher than 200.

The importance of vaccination status in COVID-19 disease outcome was confirmed by a study summarizing real-world data on the use of nirmatrelvir/ritonavir or molnupiravir in Hong Kong during the Omicron BA.2 wave. A retrospective analysis was performed using data extraction from health registers. This study investigated elderly patients hospitalized with mild to moderate symptoms who received antiviral therapy with nirmatrelvir/ritonavir or molnupiravir within 3 days after symptom onset [8]. A comparison was performed with untreated patients using propensity score matching. Treatment with nirmatrelvir reduced in-hospital mortality from 15.9% in matched controls to 8.1% in the treatment group. Surprisingly, only 0.1% of all patients were treated at an ICU and 0.4% vs. 0.9% underwent mechanical ventilation. Patients receiving nirmatrelvir/ritonavir had an all-cause mortality of 3.6% in comparison to 10.3% of matched controls. Intensive care treatment was documented in 0 vs. 0.1% of patients, and 0.7% of both groups received mechanical ventilation [3]. Although the reduction of in-hospital mortality appears to be impressive, several facts in that study are puzzling. First, the causes of death as well as the specific risk factors in patients who died are not clear, and it is questionable to what extent COVID-19 contributed to mortality given the low ICU admission and mechanical ventilation rates. Second, the mortality rates in the two controls are significantly different (15.9% vs. 10.3%) raising questions on the validity of the propensity score matching process. Third, 90% or more patients were unvaccinated, which would be in line with data that patients who were unvaccinated have the greatest benefit from antiviral therapy [3,4]. This data on mortality rates are in contrast to evidence published from USA or Canada where both ICU admission rates and mortality in patients with COVID-19 during the Omicron wave were low [5,7]. However, this study may suggest that patients hospitalized for COVID-19 and specifically those who are unvaccinated may benefit from initiation of antiviral therapy after hospital admission with earlier administration, likewise, being more promising. Of note, a more detailed analysis of the causes of death and underlying comorbidities in that study would be very important to generate knowledge on whether or not antiviral therapy may have beneficial effects in patients with COVID-19 as a secondary diagnose, f.e. by reducing inflammation, hyper-coagulopathy and subsequent thromboembolic events [10].

It has to be emphasized that those studies have not been adjusted by vaccine status and its efficacy and/or seropositivity for SARS-CoV-2 as immune protection has been shown to exert the best effects in terms of risk reduction for severe COVID-19. This is in line with reported but not yet published results on the effects of molnupiravir therapy in vaccinated people or the use of nirmatrelvir/ritonavir in vaccinated subjects with one risk factor for severe disease. In a study involving >25,000 patients in the United Kingdom, the early administration of molnupiravir did not result in a reduction of hospitalization or death in patients at risk who had previously received at least one vaccine dose (0.8% in serum and placebo group) (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4237902).

In the EPIC-SR study, nirmatrelvir/ritonavir therapy did neither result in a significant reduction of the risk of hospitalization or death in fully vaccinated adults with at least one risk factor for progression to severe COVID-19 nor alleviation of symptoms on four consecutive days as compared to placebo. The manufacturer of nirmatrelvir/ritonavir then closed the recruitment of patients owing to the low rate of hospitalization in the current pandemic situation with a less aggressive virus and broad immune protection by vaccination and/or previous infection (https://www.pfizer.com/news/press-release/press-release-detail/pfizer-reports-additional-data-paxlovidtm-supporting). Given the fortunately low incidence of severe infection in the current pandemic surge, these data also question whether the rate of hospitalization or death is still appropriate endpoint for evaluating the clinical efficacy of a specific antiviral medication.

Concerns were also raised regarding the observation of rebound infection following antiviral therapy. This appears to result from re-emergence of the virus, which caused primary infection and not because of the development of viral mutations [11]. This also indicates that some antivirals exert virustatic rather than virucidal effects, a phenomenon that could also promote the emergence of resistance to antiviral therapies.

In summary, we are very pleased to have several antiviral drugs at hand, which have been shown to significantly reduce the risk for a severe course of COVID-19 when caused by the aggressive alpha or delta variants and when given early after symptom onset to patients who were unvaccinated and at risk. However, owing to the emergence of milder omicron variants and broader immune protection based on vaccination and/or previous infection, severe COVID-19 or the need for ICU treatment for lung failure became a rare clinical event. Nowadays, we need to define those people who are still at risk for a severe course of COVID-19 or COVID-19–driven morbidity and mortality and who might benefit from antiviral therapy in the current pandemic situation (Table 1 ). Thus, we need an individual risk-benefit estimation for patients in order to keep the NNT as low as possible and to avoid morbidity due to side effects of unnecessary therapy and hinder the emergence of viral drug resistance. Moreover, data from well-conducted randomized clinical or observational studies providing more specific information on patients who may benefit from therapy are urgently awaited. Specifically, data on the effects of drugs on acute but also long-term risk reduction of most vulnerable but vaccinated/boosted people, specifically patients who are immunocompromised and elderly people with multiple risk factors [12] would be important to guide and optimize therapy with those antiviral drugs.

Table 1.

Indications for eventual use of oral antiviral drugs directed against SARS-CoV-2 Omicron variant

Outpatients without vaccination and higha risk of complicated infection

Patients who are immunocompromised with insufficient immune response to vaccination or infection

Higha risk for severe course in patients who were vaccinated/previously infected due to presence of several risk factors

Patients hospitalized due to COVID-19 and symptom onset <5–7 days.

^aRisk or combinatory risk by presence of multiple risk factors is not well defined in the Omicron era but may include combinations of advanced age (>65a), hypertension, diabetes, obesity, renal insufficiency, chronic cardiac or pulmonary diseases.

Transparency declaration

GW received honoraria for lectures or advisory board participation from Astra-Zeneca, Astro-Pharma, Insmed, Lilly, Menarini, MSD, Pfizer, Shionogi, Takeda and Vifor. This study received no external funding.

Editor: R. Chemaly