Efficacy and Safety of Respiratory Syncytial Virus (RSV) Prefusion F Protein Vaccine (RSVPreF3 OA) in Older Adults Over 2 RSV Seasons

1Respiratory syncytial virus (RSV) is a contagious pathogen that causes seasonal epidemics of acute respiratory illness (ARI). RSV infection does not confer robust long-term immunity, and reinfections are common throughout life. While infection typically causes mild symptoms in adults, older adults and those with certain chronic conditions (like cardiorespiratory diseases or diabetes) are at increased risk of severe lower respiratory tract disease (LRTD), hospitalization, cardiorespiratory complications, and death due to RSV.

2Two vaccines—both based on RSV fusion protein (F) stabilized in its prefusion conformation—have recently been authorized to prevent RSV-LRTD in individuals aged 60 years and older. The ongoing AReSVi-006 trial, which studies 1 of these vaccines (RSVPreF3 OA; Arexvy; GSK), showed that RSVPreF3 OA had an acceptable safety profile and high efficacy against RSV-LRTD (82.6%), severe RSV-LRTD (94.1%), and RSV-ARI (71.7%) over 1 RSV season. Efficacy was observed in different age groups and in participants with co-existing medical conditions.

3The AReSVi-006 trial also investigates RSVPreF3 OA efficacy and safety over multiple RSV seasons. We report the efficacy and safety over 2 seasons of 1 RSVPreF3 OA dose and of a first dose followed by revaccination 1 year later.

4The AReSVi-006 trial is a phase 3, randomized, placebo-controlled trial (NCT04886596) that is ongoing in 17 countries. Adults aged 60 years or older were enrolled (after providing informed consent) between 25 May 2021 and 31 January 2022. Individuals with chronic medical conditions (with/without specific treatment) were eligible if the investigator considered their conditions medically stable. Detailed eligibility criteria were reported previously.

5Results based on data available until April 2022 (end of the first Northern Hemisphere RSV season), including the primary outcome, have been published. Results disclosed here are based on data collected until 31 March 2023, which was considered the end of the second Northern Hemisphere RSV season, because the numbers of reported RSV-ARIs in several Northern Hemisphere countries fell to below the epidemic threshold by that point. The current analyses also include data for Southern Hemisphere participants until the same data lock, covering 1 or more Southern Hemisphere season.

6The protocol is available as an appendix to Papi et al. Information on ethical conduct is included in the Supplementary Methods.

7Pre–RSV season 1, participants were randomized 1:1 to receive 1 intramuscular injection of AS01E-adjuvanted RSVPreF3 OA or placebo. Pre–season 2, RSVPreF3 OA recipients were re-randomized 1:1 to receive a second RSVPreF3 OA dose (RSV_revaccination group) or placebo (RSV_1dose group) (Figure 1). To maintain blinding, participants who had received placebo pre–season 1 received placebo again pre–season 2 (placebo group) (Supplementary Methods). Participants, investigators, and personnel involved in data collection and sample testing were blinded.

8Confirmatory secondary objectives were to demonstrate the efficacy of 1 RSVPreF3 OA dose and of a first dose followed by a revaccination dose 1 year later in preventing RSV-LRTD over 2 RSV seasons post–dose 1. Other (descriptive) secondary objectives were to evaluate the efficacy of these 2 regimens in preventing severe RSV-LRTD, RSV-ARI, and RSV-LRTD by RSV subtype, season, year, participant age at dose 1, presence of co-existing medical conditions, and frailty status, and both regimens’ safety.

9Acute respiratory illness surveillance was done through spontaneous reporting by the participants and scheduled contacts by the site staff with the participants. Whenever a participant experienced 2 or more ARI symptoms/signs lasting 24 hours or longer, an ARI assessment visit was to occur (Supplementary Methods).

10Acute respiratory illness was defined as 2 or more respiratory symptoms/signs or 1 or more respiratory and 1 systemic symptom/sign, lasting 24 hours or longer. Lower respiratory tract disease was defined as 2 or more lower respiratory symptoms/signs (including ≥1 lower respiratory sign) or 3 or more lower respiratory symptoms, lasting 24 hours or longer (Supplementary Methods). The presence of RSV-A/RSV-B was confirmed on nasal self-swabs (collected by the participants) or nasal/throat swabs (collected at the ARI visits) by quantitative reverse transcriptase–polymerase chain reaction (qRT-PCR). An external adjudication committee reviewed all qRT-PCR–confirmed RSV-LRTD cases. Only cases adjudicated as RSV-LRTD by this committee were included in the analyses.

11Reactogenicity was evaluated in a subset of participants in selected countries/sites (reactogenicity-immunogenicity cohort) who recorded solicited adverse events (AEs) starting within 4 days and unsolicited AEs within 30 days after each vaccine/placebo dose in paper diaries. Participants not part of this cohort recorded all AEs starting within 30 days after each dose (including reactogenicity events) as unsolicited AEs. Serious AEs (SAEs) and potential immune-mediated diseases (pIMDs) were reported for 6 months after each dose; those considered vaccine/placebo-related by the investigator, fatal SAEs, and AEs leading to withdrawal are reported throughout the trial and presented here until the end of season 2.

12We planned to enroll up to 25 000 participants and include approximately1800 participants in the reactogenicity-immunogenicity cohort and assumed a 20% drop-out rate between seasons (Supplementary Methods). Efficacy over 2 seasons of 1 RSVPreF3 OA dose or a first dose followed by revaccination 1 year later (confirmatory secondary objectives) was analyzed in the modified exposed population (participants who received dose 1 and had no RSV-ARI before day 15 post–dose 1). Efficacy over season 2 was analyzed in the dose 2–modified exposed population (participants who received dose 2 and had no RSV-ARI before day 15 post–dose 2).

13Confirmatory secondary objectives were met if the lower limits of the 2-sided confidence intervals (CIs) around the efficacy estimates were greater than 20%. A Bonferroni adjustment for multiplicity was applied to evaluate these objectives in parallel; therefore, these analyses were conducted using a 1-sided alpha of 1.25% (97.5% CI). No multiplicity adjustments were applied for the other secondary objectives.

14Efficacy against a first occurrence of RSV-LRTD, severe RSV-LRTD, and RSV-ARI over 2 seasons was calculated as 1 minus the relative risk comparing participants receiving 1 RSVPreF3 OA dose or participants receiving the revaccination regimen with participants in the placebo group, using the conditional exact binomial method, based on a Poisson model including season, age, and region as covariates. Season was included as a covariate to account for potential differences between RSV seasons. As studies with other RSV vaccines may not account for these differences, we conducted a post hoc analysis without including season as a covariate.

15Periods-at-risk started on day 15 post–dose 1 (or day 15 post–dose 2 for the season 2–only analysis) and ended at the first event, data lock, or drop-out (Supplementary Methods). We also evaluated how efficacy evolved over time by descriptively comparing efficacy over different follow-up times (1 season, 1 year, mid–season 2 [post hoc], 2 seasons) (Supplementary Methods).

16Dose 2 reactogenicity was analyzed in the dose 2–solicited safety population (participants in the reactogenicity-immunogenicity cohort with solicited safety data available post–dose 2). Safety was analyzed in the exposed population (participants who received dose 1) or the dose 2–exposed population (participants who received dose 2) (Supplementary Methods).

17The exposed population included 24 973 participants, of whom 12 470 received RSVPreF3 OA and 12 503 received placebo pre–season 1. Of these, 19 990 were included in the dose 2–exposed population: 4966 in the RSV_revaccination group, 4991 in the RSV_1dose group, and 10 033 in the placebo group (Figure 2). In total, 4725 (18.9%) participants withdrew before the end of season 2, with similar proportions withdrawn in each group (RSV_revaccination: 19.5%; RSV_1dose: 18.6%; placebo: 18.8%). Baseline characteristics were balanced between groups (Table 1, Supplementary Table 1).

18The modified exposed population included 24 967 participants (RSV_revaccination: 6242; RSV_1dose: 6227; placebo: 12 498). The median efficacy follow-up from day 15 post–dose 1 until the end of RSV season 2 was 17.8 months. During this follow-up, 50 adjudicated RSV-LRTD cases were accrued among participants who received RSVPreF3 OA (10 in the RSVPreF3 OA group during season 1 and 20 cases each in RSV_revaccination and RSV_1dose during season 2), and 139 were accrued in the placebo group (47 in season 1, 92 in season 2). In both seasons, RSV-B was the dominant subtype: 40 of 57 (70.2%) RSV-LRTD cases in season 1 and 83 of 132 (62.9%) in season 2 were RSV-B positive.

19The efficacy of 1 RSVPreF3 OA dose given pre–season 1 in preventing RSV-LRTD over 2 full seasons was 67.2% (97.5% CI: 48.2–80.0%; confirmatory objective met) (Table 2). A sensitivity analysis showed similar efficacy against RSV-LRTD without viral coinfections (67.0%; 97.5% CI: 42.9–82.0%). Cumulative incidence curves confirmed sustained efficacy over 2 full seasons (Figure 3A). When calculating efficacy without accounting for potential interseason differences (model without season as a covariate), the efficacy over 2 seasons was 74.5% (Table 2).

20The efficacy over 2 seasons was 78.8% against severe RSV-LRTD and 52.7% against RSV-ARI. Efficacy was observed against medically attended RSV-LRTD and RSV-ARI (Supplementary Table 2) but could not be evaluated against RSV-related hospitalizations, because only 1 participant in the RSV_1dose group and 5 in the placebo group were hospitalized for RSV-related respiratory disease.

21Sustained efficacy over 2 seasons was observed against RSV-LRTD caused by RSV-A (80.5%) and RSV-B (59.7%), among participants 60–69 years old (65.4%) and 70–79 years old (74.9%), those with 1 or more co-existing condition of interest (66.7%), and pre-frail participants (73.3%) (Table 2). Too few RSV-LRTD cases were reported among participants aged 80 years and older and frail participants to conclude on efficacy in these subgroups.

22Although efficacy against RSV-LRTD remained high during 2 seasons, estimates tended to decline with increasing follow-up postvaccination: 82.6% over 1 season (median follow-up, 6.7 months), 78.9% over 1 year, 77.3% until mid–season 2 (median follow-up, 13.9 months), and 67.2% over 2 full seasons (median follow-up, 17.8 months). Similar trends were observed for severe RSV-LRTD and RSV-ARI (Table 3).

23When considering season 2 only, the efficacy of 1 RSVPreF3 OA dose given pre–season 1 was 56.1% (95% CI: 28.2–74.4%) against RSV-LRTD, 64.2% (95% CI: 6.2–89.2%) against severe RSV-LRTD, and 40.6% (95% CI: 19.0–57.0%) against RSV-ARI over a median follow-up of 6.3 months during season 2 (Supplementary Table 3).

24The efficacy of a first RSVPreF3 OA dose given pre–season 1 and revaccination given pre–season 2 in preventing a first occurrence of RSV-LRTD over 2 seasons post–dose 1 was 67.1% (97.5% CI: 48.1–80.0%; confirmatory objective met) (Table 4). Cumulative incidence curves confirmed the efficacy of the revaccination regimen over 2 seasons (Figure 3B).

25The efficacy of the revaccination regimen over 2 seasons was 78.8% against severe RSV-LRTD and 60.3% against RSV-ARI. None of the participants who received RSVPreF3 OA revaccination were hospitalized due to RSV-related respiratory disease.

26The revaccination regimen was efficacious over 2 seasons against RSV-LRTD caused by RSV-A (55.9%) and RSV-B (72.1%) among participants aged 60–69 years old (71.6%) and 70–79 years old (66.2%), those with 1 or more co-existing condition of interest (75.1%), and pre-frail participants (77.3%) (Table 4). We could not conclude on efficacy in participants aged 80 years and older and frail participants due to too few cases.

27The efficacy of the revaccination regimen over season 2 only was 55.9% (95% CI: 27.9–74.3%) against RSV-LRTD, 64.1% (95% CI: 5.9–89.2%) against severe RSV-LRTD, and 55.8% (95% CI: 37.5–69.5%) against RSV-ARI (median follow-up, 6.3 months during season 2) (Supplementary Table 3).

28In the dose 2–solicited safety population, 63.6% of participants in the RSV_revaccination group reported solicited AEs within 4 days after the second RSVPreF3 OA dose, versus 26.4% and 22.0% after the placebo doses in the RSV_1dose and placebo groups, respectively (Table 5). Pain and fatigue were the most frequently reported solicited administration-site and systemic AEs (Figure 4). Most events were mild or moderate and resolved within 2–3 days. Solicited AE rates after RSVPreF3 OA dose 2 were similar to those post–dose 1 (Figure 4).

29In the dose 2–exposed population, 28.5% (RSV_revaccination), 15.8% (RSV_1dose), and 14.9% (placebo) of participants reported unsolicited AEs within 30 days post–dose 2 (Table 5). The higher AE rate in the RSV_revaccination group is primarily due to reactogenicity events, reported as unsolicited AEs by participants who were not included in the dose 2–solicited safety population.

30Both SAEs and pIMDs were reported at similar rates in the 3 groups (Table 5). The most common system organ classes were cardiac disorders and infections for SAEs (Supplementary Table 4) and musculoskeletal/connective tissue disorders and skin/subcutaneous tissue disorders for pIMDs (Supplementary Table 5). No cases of acute disseminated encephalomyelitis, Guillain-Barré syndrome, or other demyelinating disorders were reported.

31A numerical imbalance was noted in atrial fibrillation within 30 days post–dose 1 (higher rates in vaccine than placebo recipients) but not within 30 days post–dose 2. Within 6 months postvaccination, a numerical imbalance in atrial fibrillation reported as an SAE was seen post–dose 2 but not post–dose 1 (Supplementary Results). In-depth review of atrial fibrillation cases did not support a vaccine-related effect; the data more plausibly reflect a typical picture of paroxysms of atrial fibrillation in a small number of high-risk individuals, at a rate no higher than expected in this population.

32From dose 2 until the end of season 2, 20 (0.4%) participants in the RSV_ revaccination group, 26 (0.5%) in the RSV_1dose group, and 41 (0.4%) in the placebo group died, with general disorders and cardiac disorders as most common system organ classes (Supplementary Table 6).

33During the entire follow-up period from dose 1 until the end of season 2, 297 participants died: 77 (1.2%) in the RSV_revaccination group, 71 (1.1%) in the RSV_1dose group, and 149 (1.2%) in the placebo group. Four fatal SAEs (group allocation still blinded) were considered by the investigator as possibly vaccine/placebo-related; 3 of these (pulmonary embolism 148 days post–dose 1, cardiopulmonary failure 31 days post–dose 1, and unknown cause 326 days post–dose 1) occurred between doses 1 and 2 (see); 1 occurred post–dose 2 (myocardial ischemia 37 days post–dose 2). Because pre-existing and concurrent medical conditions were present in these participants, the sponsor considered alternative explanations for these deaths plausible.

34This trial demonstrated that 1 RSVPreF3 OA dose was efficacious in preventing RSV-related disease in participants aged60 years and older over 2 full RSV seasons, with efficacy estimates of 67.2% against RSV-LRTD, 78.8% against severe RSV-LRTD, and 52.7% against RSV-ARI. These estimates were similar to those in participants who received RSVPreF3 OA revaccination pre–season 2 (67.1% against RSV-LRTD, 78.8% against severe RSV-LRTD, and 60.3% against RSV-ARI over 2 seasons post–dose 1). Hence, RSVPreF3 OA revaccination given 1 year post–dose 1 did not seem to provide additional efficacy benefit in the overall study population. Findings are summarized in plain language in Figure 5. Sustained efficacy of 1 RSVPreF3 OA dose over 2 seasons was observed with advanced age, among pre-frail participants, and those with underlying conditions of interest.

35Our prespecified efficacy analysis over 2 RSV seasons was based on a model including season as a covariate to account for differences between seasons in RSV-LRTD incidence rates, efficacy, follow-up time, or group sizes. Studies with other RSV vaccines may not account for interseason differences. Therefore, we performed a post hoc analysis using the model without season as a covariate. In this post hoc analysis, efficacy is driven by the season with the largest person-years of follow-up (ie, season 1). Hence, efficacy estimates tended to be higher than in our prespecified analysis.

36Although no immunological correlate of protection against RSV disease has been established, the perceived lack of additional efficacy benefit of revaccination given 1 year post–dose 1 is in line with results from immunogenicity studies showing that a second dose of RSVPreF3 OA or other RSV prefusion F–based vaccines given 1 year post–dose 1 induced a humoral immune response, but post–dose 2 RSV neutralizing titers or immunoglobulin G concentrations did not reach post–dose 1 levels. Similarly, RSVPreF3 OA revaccination given 18 months after 2 initial RSVPreF3 OA doses boosted RSV neutralizing titers, but post-revaccination titers were lower than post–dose 2 titers. The limited booster effect of revaccination given 12–18 months after initial vaccination might relate to the high levels of RSV-specific antibodies persisting up to that point, which may also explain the sustained efficacy over 2 seasons. Further research is needed to evaluate whether efficacy is maintained over more than 2 seasons and if extending the interval between initial vaccination and revaccination may enhance boosting. Longer intervals between vaccine doses have been associated with stronger immune responses and improved protective efficacy for other vaccines.

37Epidemiological differences between the 2 RSV seasons in our trial may have contributed to the lower efficacy over 2 seasons than over the first season. Overall, in our study, RSV-B was the dominant subtype in both seasons, although both subtypes were shown to co-circulate in various proportions worldwide depending on the region during these seasons. During season 2, there were more RSV circulation and RSV disease in the community than during season 1, presumably because public health measures to reduce severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spread had been lifted in many countries. Other respiratory viruses also circulated more during season 2 than season 1 in some regions. Despite this higher circulation and a larger proportion of viral coinfections during season 2 in our trial (Supplementary Results), a sensitivity analysis showed that efficacy was not impacted by the presence of viral coinfections.

38The high efficacy of RSVPreF3 OA observed over 1 year (78.9% against RSV-LRTD) suggests that RSVPreF3 OA may be administered several months before the start of the RSV season without losing substantial clinical benefit throughout the season. This, together with results of studies showing acceptable reactogenicity/safety and no clinically relevant immunological interference of co-administration with influenza vaccines, supports flexible administration of RSVPreF3 OA, either concomitantly with other vaccines (eg, influenza) at the start of the season or separately ahead of the season.

39The reactogenicity/safety profile of the RSVPreF3 OA revaccination dose was comparable to that of the first dose, with mostly mild-to-moderate and transient adverse reactions. Safety follow-up for approximately 18 months post–dose 1 confirmed the vaccine's acceptable safety profile.

40As described previously, a limitation of our trial was the relatively low number of frail participants and participants aged 80 years and older and the low number of RSV-LRTD cases in these subgroups at high risk of severe RSV disease. Conclusions on efficacy in these subgroups might be drawn as more cases accumulate during the trial. Results from another RSVPreF3 OA phase 3 study (NCT04732871) showed RSV neutralizing titers at 12 months postvaccination in individuals aged 80 years and older similar to those in individuals aged 60–69 and 70–79 years. The number of RSV-related hospitalizations was low in all arms; therefore, longer follow-up and data from larger populations are needed to inform the efficacy against RSV-related hospitalizations. Further, the comparison of efficacy between the 2 seasons is complicated by the sizes of the RSV_revaccination and RSV_1dose groups at season 2, which were half that of the RSVPreF3 OA group at season 1.

41Our trial showed that 1 dose of RSVPreF3 OA had an acceptable safety profile and was efficacious against RSV-LRTD during at least 2 RSV seasons in individuals aged 60 years and older, including those with an increased risk of severe RSV disease. Revaccination 1 year post–dose 1 was well tolerated but did not increase efficacy over 2 seasons combined or over season 2 alone. Further research is needed to determine the optimal timing of revaccination.

42Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.