With all of the talk of the Moderna, Pfizer/BioNTech and AstraZeneca/Oxford vaccines, and recent reports of preliminary data from Novavax and Johnson and Johnson/Janssen vaccine trials, there has not been as much attention paid to the Russian vaccine Sputnik V. This vaccine has pre-orders from 19 countries in the world, with India ordering 100 million doses and various other countries ordering another 100 million, led by Egypt and Nepal with 25 million each, and the UK has signed a deal to commence a study combining the Sputnik and AstraZeneca/Oxford vaccines as a possibly more robust combination. Mexico just provided emergency use approval for Sputnik V.
Controversially, Russia approved Sputnik V on August 11 based on scant data (just antibody biomarker data from 40 participants given both doses, including president Putin's daughter), even before the phase 3 study had commenced. They stated that they would manufacture the vaccine and distribute it in parallel with the phase 3 trial starting late August.
Today, the Lancet published the paper presenting phase 3 study results from this vaccine. This analysis is the third interim analysis after 78 cases in the revised protocol (revised November 5) that specified interim analyses once 20/39/78 cases were observed. A press release on November 24, 2020 presented similar results for the second interim analysis at 39 cases. The data from this paper would have been from not long afterwards, given it reports a median follow-up time of 48 days since first dose, which should be 27 days after second dose.
In spite of this follow-up that is shorter than the other published reports from Moderna and Pfizer/BioNTech, it is useful to evaluate these results.
Sputnik V vaccine
The NY Times published a nice explanation of how the Sputnik V vaccine works. Similar to the mRNA vaccines, the key component to generate SARS-CoV-2 immunity is the genetic instructions for producing the SARS-CoV-2 spike protein, but rather than encasing the mRNA in a lipid sheath, it is spliced into the DNA of a human adenovirus vector that is part of the common cold, and has been genetically engineered so it cannot replicate or invade cells or make people sick. Thus the adenovirus serves as a Trojan horse for the SARS-CoV-2 spike protein which will stimulate neutralizing antibodies for SARS-CoV-2 as a result.
The Johnson and Johnson vaccine uses a similar technology using human adenovirus rAd26 as the vector -- the Sputnik V involves two different inoculations given 21 days apart, with each involving the spike protein spliced into different human adenoviruses, rAd26 for the first dose and rAd5 for the second.
Trial design
Unlike the Moderna and Pfizer/BioNTech phase 3 studies whose full protocols were posted online, I could not find the full Sputnik V phase 3 protocol, but summarize the trial design based on details in the Lancet paper. The study was a randomized, double blind, placebo controlled trial enrolling patients from 25 hospitals and polyclinics in Moscow, Russia. Participants were randomly assigned (3:1) to vaccine or placebo, which was not a saline placebo but involved the same vaccine buffer without the recombinant adenoviruses, with stratification done by age to ensure balance between the arms.
The participants were given PCR and antibody tests at baseline to ensure that they were not infected at baseline, and on the day of the 2nd dose (21 days after 1st dose), a second PCR test was given to look for symptomatic or asymptomatic disease. After second dose, if any of various COVID-19 symptoms were experienced, an additional PCR test was given. The primary endpoint was vaccine efficacy, defined as the percent reduction in symptomatic infection rate after the second dose was given.
Results
In the interim analysis conducted at 78 cases, the primary analysis confirmed that the vaccine had strong efficacy vs. symptomatic infections occurring after the second dose (21 days after first dose), with 16/14,964 vaccinated subjects infected vs. 62/4,902 placebo subjects, for an efficacy of 91.6%. All 20 of the moderate or severe cases occurred in the placebo group, with estimated 100% efficacy in preventing moderate or severe disease (95% confidence interval 94.4% - 100%).
Here is Table 2 in the paper that shows the interim efficacy results.
From the results after dose 1, we can reconstruct estimates of the efficacy after single dose. From days 15-21 after first dose but before second dose, we see that the reported SARS-CoV-2 infections were 14/14,999 in the vaccine group and 17/4,950 in the placebo group, which would correspond to a 72.8% efficacy after single dose. If we look at the first 14 days after first dose, there are 49/14,999 cases in the vaccine group and 17/4,950 in the placebo group, which would correspond to only 4.9% efficacy in the first two weeks after the first shot. This is not surprising, as other vaccine studies have found efficacy does not kick in for 2 weeks or so, plus any undetected infections unknowingly occurring just before vaccination might manifest during that period.
Here is a Kaplan-Meier plot showing how strong the protection was once it kicked in after 2 weeks or so.
For a subset of participants, they also looked at immune response in the blood, and found 98% of the vaccinated were seropositive fo IgG antibodies to SARS-CoV-2, and 95.8% had presence of neutralizing antibodies.
The reported safety data in the study were very positive. Most adverse events were mild, and serious adverse events were balanced between vaccine and placebo arms, with 45/16,427 in the vaccine arm (0.3%) and 23/5,435 (0.4%) in the placebo arm, and none considered to be associated with vaccination. There were a total of 4 deaths, 3 in the vaccine arm and one in the placebo arm, which means that with the 3:1 randomization ratio of vaccine:placebo the death rates in both arms were identical.
Conclusions
While the decision to approve the vaccine and start distributing back in August before the phase 3 trial commenced was, ummmm, controversial, the phase 3 results reported in the Lancet paper are very positive, demonstrating strong efficacy close to that reported for Moderna and Pfizer/BioNTech with some indication of potential single-dose efficacy, showing strong immunological response and solid safety data. This is promising, and provides evidence of vaccine safety and efficacy.
It is disappointing that these results are only based on the third interim analysis at 78 events with a median follow-up of only 27 days after second dose. Given the press release on November 24, 2020 presenting results from the second interim analysis at 39 events with nearly the same total sample size, it is apparent that the data from the third interim analysis were finalized in early December at the latest, maybe sooner, and it is disappointing that more complete data were not presented in this report published nearly two months later. This is understandable given the time to peer review, but it would have been nice to have updated results since by now trial participants would have been followed up for a closer to 90 days than 27 days post second dose by now, which would provide substantially more information on efficacy and safety.
Additionally, the facts that
the Lancet paper acknowledges that the interim analyses at 39 and 78 events was added to the protocol on November 5, 2020, less than 3 weeks before the public report of the 39 event interim analysis and perhaps a month before the 78 event analysis, and
the protocol was apparently not posted online before
raises questions about why that change was made so late and why those event numbers were chosen, and could spark speculation about whether there was any unblinded data available at the time of that decision.
In spite of these quibbles, the data look very good, and if these numbers hold up during population distribution in the various countries that have purchased this vaccine, the Sputnik V will provide another valuable tool in fighting the pandemic.