Fewer HPV shots, similar protection: what a network meta-analysis found
Pooling the trials, one- and two-dose HPV schedules produced lower antibody titres than three doses but still cut persistent infection — with the usual caveat about surrogate endpoints.
Cutting the human papillomavirus (HPV) vaccine schedule from three doses to two, or even one, buys lower antibody levels but still delivers meaningful protection against the persistent infections that cause cervical cancer, according to a Bayesian network meta-analysis of randomised trials published in August 2026 [s1]. The finding supports what many programmes are already doing — offering fewer doses to reach more girls — while the authors stress that the trials measured immune and infection markers, not cancer itself, so the conclusion carries a real caveat [s1].
The stakes are set by the disease's geography. Cervical cancer was the fifth most commonly diagnosed cancer in women worldwide in 2024, with an estimated 604,000 new cases and around 280,000 deaths, and its burden falls hardest on low- and middle-income countries where vaccine, screening and treatment access is thinnest [s2]. WHO's elimination strategy hinges on a first target of 90% of girls fully vaccinated against HPV by age 15 — a threshold that is far easier to hit if each girl needs one or two shots rather than three [s2].
What the analysis pooled
The reviewers searched five databases through 15 February 2026 for randomised controlled trials of different HPV vaccine dose schedules in females aged 9 to 26, then combined them in a Bayesian random-effects model and ranked schedules by their Surface Under the Cumulative Ranking (SUCRA) scores [s1]. Two endpoints did the work: seroconversion against HPV types 16 and 18, and six-month persistent infection [s1].
On raw immunogenicity, more was more. The nine-valent three-dose regimen produced the highest seroconversion odds versus placebo (odds ratio 62.45, 95% credible interval 15.20–158.30; SUCRA 0.96) — but the authors flag that such enormous effect estimates are inflated by near-zero events in the placebo arms, an artefact of the maths rather than a clean measure of benefit [s1].
Protection against persistent infection is the more decision-relevant outcome, and there the ranking tightened. Three-dose schedules still led (nine-valent SUCRA 0.95; bivalent SUCRA 0.89), but a single bivalent dose retained clinically significant protection against persistent infection (odds ratio 0.32 versus placebo, 95% credible interval 0.07–0.96; SUCRA 0.61) [s1]. In other words, one dose lowered the odds of persistent HPV-16/18 infection by roughly two-thirds in the pooled estimate, even as it generated lower antibody titres than the full course [s1].
Fewer doses also meant fewer side effects. One-dose schedules had the lowest risk of injection-site pain, with safety SUCRA scores of 88.0% for the bivalent single dose and 82.0% for the nine-valent single dose [s1]. And in the subgroup that matters most for national programmes — girls aged 9 to 14 — two-dose regimens looked comparable to the standard three (interaction p = 0.04) [s1].
The caveat the authors put front and centre
This is a synthesis of surrogate endpoints, not a trial with cancer or precancer as the outcome. The authors are explicit that seroconversion and six-month persistent infection are markers, that network meta-analysis is a weak tool for proving one schedule is non-inferior to another, and that long-term clinical-endpoint data are still needed before a single-dose strategy can be called equivalent [s1]. Their bottom line is calibrated to that uncertainty: broad adoption of two-dose schedules in adolescents is well supported, and a single dose is "a viable alternative in resource-constrained settings" to speed cervical-cancer elimination — not a wholesale replacement everywhere [s1].
That framing matters because dose reduction is not a theoretical debate. It is already reshaping how countries deliver the vaccine and how they count coverage against WHO's targets, and it sits alongside the training and screening push that the elimination agenda also depends on. A programme that vaccinates 90% of girls with one dose protects more of them than a programme that reaches 60% with three — provided the one-dose protection holds up over the decades between vaccination and the cancers it is meant to prevent, which is exactly what the surrogate endpoints here cannot yet confirm [s1][s2].
This article is informational and is not medical advice.
Sources
- [s1] "Immunogenicity, efficacy, and safety of different dose schedules of human papillomavirus vaccines: a Bayesian network meta-analysis," Frontiers in Immunology, 6 August 2026. https://doi.org/10.3389/fimmu.2026.1883217
- [s2] World Health Organization, "Cervical cancer" (fact sheet). https://www.who.int/news-room/fact-sheets/detail/cervical-cancer
Sources
- Immunogenicity, efficacy, and safety of different dose schedules of human papillomavirus vaccines: a Bayesian network meta-analysis — Frontiers in Immunology , August 6, 2026
- Cervical cancer (fact sheet) — World Health Organization
More on
The HPV vaccine works, and the cervical cancer data now prove it in real populations
Two national studies settle the question the trials could not: in England, cancer rates fell 87% among those vaccinated at 12-13; in Sweden, the risk was cut by nearly two-thirds.
A free training course, launched on the first World Cervical Cancer Elimination Day
IARC and WHO put five modules on screening and precancer management online. The 90-70-90 targets they support depend on health workers who can act on a positive test.
Can women collect their own cervical screening sample? What the evidence says
For the HPV tests now used in primary screening, a self-collected vaginal sample is about as accurate as one a clinician takes — and mailing kits to under-screened women roughly doubles uptake.
Cervical screening is shifting from the Pap smear to the HPV test
US guidelines now let women aged 30 to 65 screen every five years with an HPV test. One major body would make that the primary test from age 25 — and self-collected samples work about as well.