A recent study published through the Department of Veterans Affairs has suggested that Paxlovid treatment within 5 days of COVID-19 symptoms can reduce the risk of long COVID (news story link: https://www.militarytimes.com/news/coronavirus/2022/11/07/va-study-finds-paxlovid-may-help-prevent-long-covid-19-problems/).  Researchers examined a total of 56,340 participants of whom 9,217 were treated with Paxlovid and 47,123 who were treated with supportive care (i.e. no antivirals or monoclonal antibodies within 30 days of infection). Study dates were from March 1st to June 30th 2022 (paper link: https://www.medrxiv.org/content/10.1101/2022.11.03.22281783v1.full.pdf). 

The authors define long COVID as “Post-Acute Sequelae of SARS-CoV-2” or PASC and, in the study, identified those individuals who suffered from at least one extended outcome from a total of 12 pre-specified outcomes beginning 30 days after infection (and, importantly, did not display this outcome prior to COVID-19 infection). The same primary author had characterized PASC to include pulmonary, cardiovascular, hematologic, GI, kidney, mental health, musculoskeletal, and neurologic disorders as well as fatigue (link: https://www.nature.com/articles/s41586-021-03553-9). 

So looking at the study, it was interesting to see that about 16% of the cohort (all outpatient as Paxlovid is approved for non-hospital use) was prescribed Paxlovid. As a comparison, in our practice, 44 of 166 patients diagnosed in 2022 have been prescribed Paxlovid (26.5%). The mean age in the VA study was 65 years (65.06 for those prescribed Paxlovid and 65.07 for those who received supportive care). One limitation of the VA study, however, is that it was predominately (87.6%) male perhaps limiting conclusions about Paxlovid efficacy among women (this is, of course, a reflection of the VA population as a whole and not a specific critique of this study in particular). An additional oddity in the data was that 25% of the VA population studied in this effort was unvaccinated which is slightly higher than the 20% of the US population that has not received any COVID vaccination whatsoever.

Additionally, Paxlovid was given in the majority of occasions (80% of the time) to individuals who had completed a primary series of vaccination or had a primary series plus a booster. Only 16% of the time was the treatment given to an unvaccinated individual which, of interest, is the only cohort in whom Paxlovid has been shown to have significant clinical benefit. As an aside, a relative risk reduction of about 50% was seen in terms of likelihood of hospitalization or death among those who had received the COVID-19 vaccine, but this did not attain statistical significance (link: https://www.pfizer.com/news/press-release/press-release-detail/pfizer-reports-additional-data-paxlovidtm-supporting). This intersection between vaccination and treatment also extended to influenza vaccination where 71% of those treated with Paxlovid had received a seasonal flu vaccination as opposed to 60.2% of those in the control group.

On the other hand, it did appear that Paxlovid was given appropriately to those with more chronic clinical conditions that would leave them at increased risk of poor outcome.  Those in the treatment group were heavier (BMI: 30.94 versus 30.55; as a quick aside the mean BMI of all participants was 30.61 and obesity is defined as a BMI above 30), more frequently had a cancer diagnosis (15.0% versus 12.9%), diabetes (33.9% versus 29.8%; again noting that the prevalence of diabetes was 30.5% in the entire cohort), cardiovascular disease (30.9% versus 29.1%) and elevated cholesterol (80.9% versus 71.4%; noting that 73% of the entire study group had high cholesterol).

In the VA study, Paxlovid treatment was associated with a reduced frequency of long COVID (PASC) with 9.43 per 100 individuals reporting long COVID symptoms at 90 days in the control group as compared to 7.11 among those treated with Paxlovid – a 26% reduction in risk. Despite being less well at baseline, those treated with Paxlovid also had reduced rates of hospitalization (2.57 events per 100 patients as compared to 3.66 in the control group) and fewer deaths (0.30 deaths per 100 patients treated versus 0.58 in the control group), also at 90 days.  As a comparison, mortality rates for those hospitalized with COVID-19 during Delta were 15.1 per 100 and during Omicron, 4.9 per 100. Numbers reported in this VA study were during Omicron as well.

The authors take their analyses a couple of steps further, which is helpful to better elucidate the precise role of Paxlovid beyond a statement that it ‘reduces Long COVID.’  They tested the association between Paxlovid and the risk of PASC according to the number of baseline risk factors for progression to severe acute COVID-19 illness. While Paxlovid was associated with reduced risk of PASC in people across different risk factors strata, no pattern emerged.  Among those with fewer than 2 risk factors, Paxlovid reduced the risk of long COVID by 33%. The same magnitude of effect was seen among patients with 3-4 risk factors but dropped slightly for those with 5 or more to 30%.

We know from Pzifer’s clinical trial studies that the greatest magnitude of protective effect in terms of reduction of hospitalization and death rates was seen among those who were unvaccinated. A similar pattern emerged in this study when it came to long COVID. While Paxlovid reduced the risk of PASC/Long COVID among those vaccinated and boosted by 21%, it lowered the risk by 32% among the unvaccinated.

Lastly, Paxlovid was associated with reduced risk of PASC in people with primary SARS-CoV-2 infection and in people with reinfection to an identical degree, both with a 25% reduction in Long COVID risk.

To date, Paxlovid has been a ‘no brainer’ treatment among those who are unvaccinated, given strong clinical evidence of efficacy in reducing hospitalization and death. Of course, the vaccination and boosters have also shown strong efficacy in reducing COVID-19 hospitalization and death but that is a separate discussion.  This current VA study suggests that among their cohort (average age of 65 years, 75% Caucasian, 20% African American, 87% male and 42% former smokers), there may be an additional benefit – namely in the reduction of Long COVID. As with the general indications for prescribing Paxlovid in general, it seems that among an older, male population with comorbidities including prior tobacco, that Paxlovid is efficacious. It would be helpful to see additional studies in other populations, particularly comprised of a greater proportion of women.