The investigational tuberculosis vaccine VPM1002 did not demonstrate noninferiority to the standard BCG vaccine for preventing Mycobacterium tuberculosis infection in newborns, according to results from the phase 3 priMe trial (NCT04351685).1
The trial, conducted at 10 main sites and 4 satellite sites across 5 sub-Saharan African countries, enrolled 6950 infants and is among the largest tuberculosis vaccine trials conducted in infants to date.1 The effort was led by a team of investigators including Videlis Nduba, MD, PhD, an epidemiologist and center director at the Kenya Medical Research Institute and Center for Respiratory Disease Research (KEMRI/CRDR), in Nairobi and Siaya, Kenya.
What Infectious Disease Specialists Need to Know
Did VPM1002 meet its primary end point in the priMe trial?
No. VPM1002 did not demonstrate noninferiority to BCG; the hazard ratio for QFT conversion was 1.23, with a 95% CI upper bound of 1.53, exceeding the prespecified non-inferiority margin of 1.25.
Why is this result difficult to interpret?
The trial accrued only 334 of a planned 632 tuberculosis infection events, and the study authors noted discordance between the QFT surrogate marker and confirmed tuberculosis, along with a higher rate of household tuberculosis exposure in the VPM1002 group.
Were there safety differences between VPM1002 and BCG?
No. Adverse event profiles, including serious adverse events and deaths, were similar between groups, and no vaccine-related serious adverse events were reported in either arm.
VPM1002 is a recombinant BCG vaccine that showed manageable toxicity and immunogenicity in earlier phase 1 trials in adults and phase 2 trials in South African newborns.1,2 Healthy newborns aged 0 to 14 days with a birthweight of at least 2.3 kg were randomly assigned 1:1 to receive a single 0.05 mL intradermal dose of VPM1002 or BCG, stratified by maternal HIV status.1 Enrollment ran from November 9, 2020, to June 21, 2022; 6940 infants were ultimately randomized, including 720 born to mothers living with HIV and 6220 who were HIV-unexposed, and 6897 infants received vaccination.1
Phase 3 priMe trial in nearly 7000 African newborns found higher rates of QFT conversion with the investigational vaccine
The primary end point was noninferiority of VPM1002 versus BCG for prevention of M tuberculosis infection, defined by incident QuantiFERON-TB Gold Plus (QFT) conversion, an interferon-γ release assay, with noninferiority requiring the upper bound of the 95% CI for the hazard ratio to be less than 1.25.1 Follow-up was extended from 36 months to 48 months after the trial saw a lower than expected event rate, and the trial was ultimately terminated early in October 2024 for the same reason.1
Over a median follow-up of 35 months, QFT conversion occurred in 184 of 3471 infants, or 5.3%, in the VPM1002 group compared with 150 of 3469 infants, or 4.3%, in the BCG group.1 The resulting HR was 1.23 (95% CI, 0.99-1.53), meaning VPM1002 was not shown to be noninferior to BCG, since the upper bound of the confidence interval exceeded the prespecified 1.25 margin.1 Adverse event (AE) profiles, including serious AEs and deaths, were similar between groups, and no vaccine-related serious AEs were reported in either arm.1
Trial Design Complicates Interpretation of the Result
The study authors cited several factors that complicate interpretation of the findings.1 Only 334 of a planned 632 M tuberculosis infection events accrued despite the extended follow-up period, leaving the trial without sufficient statistical certainty to definitively compare the 2 vaccines, a setback for a pipeline that the WHO has said needs new vaccines to meet global TB reduction targets.1 The authors also noted discordance between the QFT surrogate marker and confirmed tuberculosis end points, along with an approximately 33.0% higher rate of household tuberculosis exposure in the VPM1002 group compared with the BCG group.1
Additionally, when household tuberculosis exposure was included as a covariate in the statistical model, the proportional hazards assumption was violated, meaning the effect of exposure on QFT conversion was not constant over time and could not be reliably estimated without penalized modeling.1 The study authors said these findings highlight broader challenges in using interferon-γ release assay-defined infection end points in infant vaccine trials, beyond the specific result for VPM1002.
VPM1002 has been in clinical development since being licensed from Max Planck Institute researchers in the early 2000s and later sublicensed to the Serum Institute of India for manufacturing.1 The priMe trial, funded by the European and Developing Countries Clinical Trials Partnership 2 program and co-funded by the European Union, the German Aerospace Center, and the European Investment Bank, is also registered in the Pan African Clinical Trials Registry (PACTR202007868402718) and is now complete.1
BCG remains the only tuberculosis vaccine currently in widespread use, and provides variable protection against pulmonary tuberculosis in adolescents and adults despite reducing risk of severe disease in young children, a gap other next-generation TB vaccine candidates are also racing to address.
References
Nduba V, Mathebula M, Nankabirwa V, et al; priMe consortium. Comparison of VPM1002 with BCG in the prevention of tuberculosis in newborn infants: a multicentre, double-blind, randomised, phase 3, non-inferiority trial. Lancet Infect Dis. Published online August 19, 2026. doi:10.1016/S1473-3099(26)00374-9
Cotton MF, Madhi SA, Luabeya AK, et al. Safety and immunogenicity of VPM1002 versus BCG in South African newborn babies: a randomised, phase 2 non-inferiority double-blind controlled trial. Lancet Infect Dis. 2022;22(10):1472-1483. doi:10.1016/S1473-3099(22)00222-5
