Patients

Screening, Enrollment, and Randomization.

Data regarding the disposition and demographic characteristics of the participants, safety, and immunogenicity were obtained from the complete and locked database (date of data extraction, September 27, 2019). Efficacy data were obtained from the locked database after all the participants had completed the 180-day follow-up (date of data extraction, January 30, 2019); these data represented the official efficacy analysis set. The safety population comprised all maternal participants who had undergone randomization and received the respiratory syncytial virus fusion protein nanoparticle vaccine (RSV F vaccine) or placebo and their live-born infants. Data on exclusions are derived from incomplete informed consent documentation, lost or incomplete source documentation, or both. The intention-to-treat efficacy analysis population included all maternal participants and their infants in the safety analysis population for whom at least one respective post-treatment or postpartum efficacy measurement was available for both the mother and the infant, as evidenced by collection of surveillance observations. The maternal participants in the per-protocol efficacy analysis population were those who received the assigned RSV F vaccine or placebo, had at least one post-treatment encounter with trial personnel during which active or passive surveillance (or both) for RSV illness could have been performed, and had no major protocol deviations affecting the primary efficacy end point, as determined and documented by the sponsor before database lock and unblinding. The infant participants in the per-protocol efficacy population were those who were born at 37 weeks or more of gestation, who were born to maternal participants who had undergone randomization and received the assigned vaccine or placebo at least 2 weeks before delivery, and who did not receive prophylactic treatment with palivizumab between the day of birth and day 180 after delivery, had at least one postpartum encounter with trial personnel during which active or passive surveillance (or both) for RSV illness could have been performed, and had no major protocol deviations affecting the primary efficacy end point, as determined and documented by the sponsor before database lock and unblinding. Participants who were excluded from one or more analysis populations may have had more than one of the listed major protocol deviations or exclusionary characteristics. Per-protocol status in an infant required elements of per-protocol performance in the mother.

Baseline Demographic Characteristics and Other Characteristics of the Maternal Participants and Birth and Household Characteristics of their Infants.

Between December 3, 2015, and May 2, 2018, a total of 4636 women were enrolled, of whom 3051 (65.8%) were randomly assigned to receive the RSV F vaccine (Figure 1). Among the 4636 women enrolled, 52.3% were enrolled in South Africa and 23.3% were enrolled in the United States (Table S1). There were 4579 live births. Of the 4636 women, 10 (0.2%) had incomplete consent or other source documentation that could not be recovered (4 in the vaccine group and 6 in the placebo group). These participants and their infants were excluded from all analyses. A total of 4195 infants (91.6%) were included in the per-protocol population, and 4527 infants (98.9%) were included in the intention-to-treat population (Figure 1). The characteristics of the women at baseline and of the infants, as well as gestational age at the time that vaccine or placebo was administered, were similar in the two trial groups (Table 1), including when stratified according country income level (high-income country or low- or middle-income country) (Tables S2 and S3).

Primary and Secondary End Points

Efficacy of Maternal Vaccination against Lower Respiratory Tract Infection in Infants. Kaplan–Meier Survival Plots for the Primary and Secondary Efficacy End Points in the Per-Protocol Population.

Shown are the Kaplan–Meier survival plots for the primary efficacy end point of RSV-associated, medically significant lower respiratory tract infection (Panel A) and the secondary efficacy end points of RSV-associated lower respiratory tract infection with severe hypoxemia (Panel B) and hospitalization for RSV-associated lower respiratory tract infection (Panel C). For each panel, the main figure depicts the percentage of infants in the per-protocol population who survived without the occurrence of the specified end-point event as a function of time from delivery. Because the events under study occurred in 5% of the infant population or less over the first 180 days of life, insets are provided to show the same data on an enlarged y axis.

The percentage of infants who had RSV-associated, medically significant lower respiratory tract infection through 90 days was 1.5% in the vaccine group and 2.4% in the placebo group (estimated vaccine efficacy in the per-protocol analysis, 39.4%; 97.52% confidence interval [CI], −1.0 to 63.7; 95% CI, 5.3 to 61.2) (Table 2). An analysis that was based on the same definition and data but that was performed in the intention-to-treat population yielded an efficacy estimate of 32.2% (95% CI, −4.2 to 55.9) (Table S15). The results of analyses that used expanded data sources in the intention-to-treat population are provided in Table 2. The results for various efficacy end points in infants at 120, 150, and 180 days of life are provided in Table S14. Kaplan–Meier survival curves based on analyses in the per-protocol population are provided in Figure 2.

The percentage of infants in the per-protocol population with RSV-associated lower respiratory tract infection with severe hypoxemia through 90 days was 0.5% in the vaccine group and 1.0% in the placebo group (vaccine efficacy, 48.3%; 95% CI, −8.2 to 75.3; vaccine efficacy in the intention-to-treat population was 44.4% (95% CI, −14.9 to 73.1), as determined with the use of clinical site data only, and 58.8% (95% CI, 31.9 to 75.0), as determined with the use of expanded data (Table 2 and Table S15). The percentage of infants in the per-protocol population who were hospitalized for RSV-associated lower respiratory tract infection through 90 days was 2.1% in the vaccine group and 3.7% in the placebo group (vaccine efficacy, 44.4%; 95% CI, 19.6 to 61.5); the results in the intention-to-treat population were closely similar to those of the per-protocol analysis: 48.1% (95% CI, 26.1 to 63.5), as determined with the use of clinical site data only, and 46.4% (95% CI, 24.7 to 61.9), as determined with the use of expanded data (Table 2 and Table S15). Point estimates of vaccine efficacy through 120, 150, and 180 days of life declined relative to the first 90 days of life with respect to the end point of RSV-associated, medically significant lower respiratory tract infection, but the estimates remained similar throughout with respect to the end points of hospitalization for RSV-associated lower respiratory tract infection and RSV-associated lower respiratory tract infection with severe hypoxemia in both the per-protocol and expanded-data intention-to-treat analyses.

Exploratory End Points

The rate of medically significant lower respiratory tract infections from any cause through the first 90 days of life was 5.5 events per 100 infants in the vaccine group and 7.2 events per 100 in the placebo group (vaccine efficacy in the per-protocol analysis, 23.2%; 95% CI, 1.4 to 40.2) (Table 2). The corresponding rates for lower respiratory tract infection from any cause with severe hypoxemia through 90 days of life were 1.7 and 3.1 events per 100 infants (vaccine efficacy, 46.0%; 95% CI, 18.7 to 64.1), and the rates for lower respiratory tract infection from any cause with hospitalization through 90 days of life were 4.3 and 6.0 events per 100 infants (vaccine efficacy, 27.8%; 95% CI, 4.8 to 45.3). The results of the expanded-data intention-to-treat analysis were similar to those of the per-protocol analysis (Table 2). The effects of vaccine on lower respiratory tract infection from any cause appeared to be durable through 180 days of life.

Estimates of vaccine efficacy against the various end points in both per-protocol and expanded-data intention-to-treat analyses, stratified according to country income level, are provided in Table S16. Efficacy estimates were greater in low- or middle-income countries than in high-income countries in general, whereas the percentages of infants with end-point events were lower in high-income countries and the confidence bounds for vaccine efficacy estimates were therefore wider. Estimates of vaccine efficacy against lower respiratory tract infection according to RSV subtype (A or B) are provided in Table S17.

Estimates of vaccine efficacy against RSV-associated lower respiratory tract infection of any severity and against RSV-associated, medically significant lower respiratory tract infection according to the definition of tachypnea used by the WHO (10 breaths per minute less than in the protocol definition) were 15 to 19% over the first 90 days of life, declining to 12 to 13% over the first 180 days of life (Table S18). There was no clear efficacy against lower respiratory tract infection from any cause when events of any severity were included in the analysis. The incidence of RSV-associated symptomatic respiratory tract infection was similar among the women who received RSV F vaccine (4.9% [148 of 3004]) and among those who received placebo (4.8% [76 of 1569]) through 180 days post partum.

Safety

Safety Profile in Maternal and Infant Participants.

Local injection-site reactions were predominantly mild and were more common among the women who received the vaccine than among those who received placebo (40.7% vs. 9.9%; P<0.001) (Table 3 and Table S4). Fever within 7 days after vaccination occurred in 1.2% of the women who received the active vaccine and in 1.6% of the women who received placebo; the frequency of systemic reactions overall was similar in the two groups of women. No clear between-group differences were observed with regard to the percentages of women who had unsolicited adverse events, including the prespecified adverse events of special interest or adverse delivery outcomes (Table 3 and Section S2.2.4 and Tables S5, S7, and S9).

The overall percentages of infants who had common or serious adverse events or protocol-defined adverse events of special interest were also similar in each trial group (Table 3, and Tables S6, S8, and S10). However, serious adverse events coded as “pneumonia” were less common among the infants in the vaccine group (2.2%) than among those in the placebo group (4.5%) through 364 days (Table 3).

Immunogenicity

Fourteen days after injection of the RSV F vaccine (the timing of peak levels in phase 2), the geometric mean concentration of palivizumab-competitive antibodies was 12.39 times (95% CI, 11.98 to 12.81) as high as it was before injection, and the geometric mean concentration of anti-F IgG was 18.59 times (95% CI, 17.84 to 19.36) as high.7,9 Additional results for these antibodies as well as RSV A and B microneutralization titers are provided in Table S11. Transient decreases in RSV antibody levels in women were observed at the time of delivery; the levels rebounded at day 35 post partum and then declined at day 180 post partum.

The geometric mean concentrations of anti-F IgG and palivizumab-competitive antibodies in cord blood were 9501 ELISA (enzyme-linked immunosorbent assay) units (EU) per milliliter (95% CI, 9224 to 9787) and 136 μg per milliliter (95% CI, 132 to 139), respectively, among the infants in the vaccine group and 752 EU per milliliter (95% CI, 719 to 786) and 15 μg per milliliter (95% CI, 14 to 15), respectively, among the infants in the placebo group. In the vaccine group, the ratio of antibody concentration in cord blood to antibody concentration in the mother at the time of delivery was 1.04 (95% CI, 1.02 to 1.06) for palivizumab-competitive antibodies and 1.17 (95% CI, 1.14 to 1.19) for anti-F IgG. The estimated antibody half-life in the infants in the vaccine group was 49.1 for palivizumab-competitive antibodies and 38.3 for anti-F IgG. The results of analyses of transplacental antibody transfer overall and as stratified according to country income level are provided in Tables S11 and S12.

Source