There are studies that are bound to succeed, and those that are likely to fail. A failed trial is nothing to be proud of – and in this case, it was a gamble against overwhelming odds. We feel the Gilead marketing department was pushing their luck big time with this trial.[1]
Maybe we should be glad the trial failed; otherwise, we would have another Tamiflu situation, 1 day improvement in symptom relief at best, being touted as clinically important. For obeldesivir, an oral version of remdesivir, we would have expected a remdesivir-like study, trying to mimick past successes with a drug that has a different PK and more favorable metabolic features but the same MoA.
Not here. The OAKTREE study was more or less an early intervention trial with the assumption that early neutralization of Covid-19 might somehow be of clinical benefit, be it by reducing the duration of symptomatic disease or by reducing the frequency of other bad outcomes, such as hospitalization and death. This rationale is not wrong, but one would presumably first test this hypothesis in a population at high risk for disease progression, as in the elderly, immunocompromised, or patients with underlying diseases who cannot deal well with Covid-19 infection. Once proven effective in this population, it would make sense to take the next step…
Well, the opposite was done here, namely, obeldesivir was tested in a population at very low risk of progression, in newly infected young adults (median age = 41 years!) without immune compromise or comorbidities. Most had a history of prior infection or vaccination. This is not the population for which remdesivir had shown benefit in the past and been approved.
However, this premise was not altogether wrong, as this table showed; earlier intervention was associated with earlier resolution of symptoms. In this regard, obeldesivir was clearly superior to placebo. Just not enough to make it clinically relevant.
If this were a timed exposure study of volunteers to the virus, with a fixed nasal inoculum at time zero, followed by drug administration after 8, 24, 48, and 72 hrs, we might have seen a treatment effect on symptom relief, and maybe even a dose response. However, this would be a PoC trial, not reflecting the real world situation. The way the study was designed though, in the presence of preexisting immunity, the envelope was pushed too far.
Of course, Gilead knew that and wisely embarked on another Phase 3 tree trial which they called BIRCH. This one was in patients with risk factors for severe Covid-19 infections, a more appropriate target population; the endpoints were COVID-19–related hospitalization or all-cause death. Too bad this study had to be stopped prematurely, because actual hospitalization rates were much lower than anticipated, almost zero in either arm. Hence, efficacy for obeldesivir could not be established.
“Ultimately, the mild clinical course and rapid recovery seen in this study, which is representative of the current COVID-19 landscape in non-hospitalised adults without risk factors for progression to severe disease might have impeded demonstration of a significant and clinically meaningful treatment benefit with regard to time to symptom alleviation or resolution”1
Unfortunately, neither BIRCH nor OAKTREE achieved its stated objectives. According to the authors, it may have been vaccination that protected patients, obliterating any significant contribution of the antiviral drug.[2] We agree with this assessment wholeheartedly.
Just as a reminder: Remdesivir was approved in the early days of the Covid-19 epidemic based on data generated in non-immune, unvaccinated patients. It was never considered to be a particularly potent anti-Covid drug; it still only plays a secondary role in the latest Covid treatment guidelines (see below). We are not so sure it deserves to be on the list at all, as we are now in the post-vaccine era and dealing with patients that have at least partial immunity.
Of course, the same holds true for molnupiravir, another nucleoside drug approved based on study results in unvaccinated patients
Gilead is clearly a company committed to nucs and nuc research; having developed remdesivir in-house, there was a strong drive to push development of an oral version. We believe that obeldesivir would not have failed in the early Covid days; it would have received Emergency Use Authorization (EUA) and been approved.
Where does all this leave obeldesivir? A recent Gilead press release informs us that obeldesivir studies in RSV infection have been stopped for lack of clinically relevant benefit. It seems obeldesivir is on its way out.
Maybe not surprisingly, Gilead presentations at ID Week last month were not centered on obeldesivir but rather on lenacapavir, their great new HIV drug, a powerful long-acting therapeutic both for prevention and treatment. Lenacapavir is a game changer for HIV, just as Sovaldi was for HCV. It is also a product of Gilead research but not a nucleoside, sorry to say!
Actually, we find it quite amazing that the nucleoside class of drugs is still being worked on, still producing NCEs worthy of development. AZT was a nucleoside making it into the clinic as the first anti-HIV, developed by BMS. John Martin recognized the value of tenofovir, a nucleotide, and started his own company, Gilead, as he left BMS. Gilead then bought sofosbuvir, a nucleoside drug from Pharmasset, which for the first time was able to cure HCV infection. Gilead has done very well managing its nucleoside franchise by combining drugs in the HIV and HCV portfolios to further improve viral load suppression and clinical efficacy.
Lenacapavir will add a new non-nuc chapter.
REFERENCES
[1] Ogbuago O. Efficacy and safety of obeldesivir in low-risk, nonhospitalised patients with COVID-19 (OAKTREE): a phase 3, randomised, double-blind, placebo-controlled study. Lancet Infect Dis 2025, 25: 1282
[2] Steichu-Cercel A. Efficacy and Safety of Obeldesivir in High-Risk Nonhospitalized Patients With Coronavirus Disease 2019. (BIRCH): A Phase 3, Randomized, Double-Blind, Placebo-Controlled Study. https://doi.org/10.1093/cid/ciaf406
