what is the evidence for the HPV vaccine in cancer reduction?

The short version: HPV vaccination likely reduces precancerous lesions caused by the vaccine-type strains, but the evidence that it has reduced actual cancer incidence is thin, indirect, and contradicted by rising cervical cancer rates in several countries.

The strongest evidence for benefit comes from Australia, where HPV 16 and 18 prevalence dropped from about 26 in 100 women before vaccination to about 5 in 100 after ten years of the program [7]. Genital wart cases also dropped sharply. These are surrogate endpoints — viral prevalence and pre-cancerous lesions, not cancer itself. Cervical cancer takes decades to develop, so the first vaccinated cohorts are only now reaching the ages where cancer appears. The evidence from actual cancer registries is less encouraging.

A 2025 Canadian study found that cervical cancer incidence in women aged 35 to 54 has been rising since about 2001, with rates in 2022 higher than they were in 1992 [1]. The largest rise was in women aged 35 to 44, the group that should have been first protected by the vaccine. A 2026 U.S. study showed that while overall cervical cancer incidence declined slowly from 2000 to 2021, the decline slowed to nearly zero after 2011 — and advanced-stage (stage IV) diagnoses rose steadily at about 2.6 in 100 per year [4]. The proportion of cases diagnosed at stage IV went from about 10 in 100 to about 15.5 in 100. These are not the patterns you would expect from a vaccine that prevents cancer.

There is also a documented mechanism that could inflate apparent vaccine benefit: a 2012 modeling study showed that when you vaccinate against the most common HPV types, the rarer types can "unmask" — filling the ecological niche and producing an apparent 3 to 10 in 100 increase in cancer from non-vaccine types [3]. This is not speculation; it is a well-characterized phenomenon in population dynamics. The same study notes that the vaccine does not cover all oncogenic HPV types [2, 5].

The trials that established the vaccine were funded by the manufacturer, used precancerous lesions (CIN2/3) as the endpoint rather than cancer, and the longest follow-up is still short relative to the natural history of cervical cancer [6]. No trial compared vaccinated women to an unvaccinated control group over the decades needed to see cancer rates. The safety data for long-term effects — immune imprinting, narrowed antibody repertoire, autoimmune activation — was never collected in a design that could detect it.

My call: HPV vaccination likely reduces pre-cancerous lesions caused by specific high-risk HPV types, but the evidence that it reduces actual cervical cancer incidence is not yet there. The cancer registry data from several countries shows rates rising in the age groups that should be protected. The most informative study — vaccinated versus unvaccinated over 30 to 40 years — was never run. Confidence: low that the vaccine reduces cancer incidence as claimed, moderate that it reduces pre-cancerous lesions for vaccine-type strains.