A cancer is being vaccinated away. Down 87 percent.

Researchers in England found that women who received the HPV vaccine at 12 or 13 had 87 percent less cervical cancer; Scotland later reported zero cases among fully vaccinated women.

A virus found inside a tumour

In 1983 Dürst, Gissmann, Ikenberg and zur Hausen cloned a papillomavirus DNA from a biopsy of invasive cervical cancer and proposed to call it HPV 16. Using it as a probe, they found matching sequences in 11 of 18 cervical cancer samples from German patients, 61.1 percent, and in 8 of 23 samples from Kenya and Brazil. The virus was rare in benign genital warts. The data, they wrote, indicated that HPV 16 "prevails in malignant tumors."

Sixteen years later Walboomers and colleagues reanalyzed the cases from a worldwide survey that had found HPV in 93 percent of cervical cancers, and asked whether the remaining 7 percent were truly negative. Most were not. After excluding inadequate specimens, the worldwide prevalence of HPV in cervical carcinoma came to 99.7 percent, which the authors called the highest attributable fraction ever reported for a specific cause of any major human cancer. A cancer with a single necessary cause is a cancer that can, in principle, be prevented by blocking that cause.

A programme begins in England

England introduced HPV immunisation on September 1, 2008, using the bivalent vaccine Cervarix. Routine vaccination was offered to girls aged 12 to 13, with a catch-up programme for those aged 14 to 18 running from 2008 to 2010. Because cervical cancer typically develops decades after infection, the first vaccinated cohorts would not reach the ages at which the disease is usually diagnosed for many years. By 2021 the oldest of them were approaching 30, and a register-based test became possible.

13.7 million years of follow-up

Falcaro, Sasieni and colleagues published that test in The Lancet in November 2021. They extracted data from the national cancer registry on January 26, 2021, covering diagnoses of cervical cancer and grade 3 cervical intraepithelial neoplasia, the immediate precursor to cancer, from January 1, 2006 to June 30, 2019, in women aged 20 to 64 resident in England. They compared three vaccinated cohorts, defined by the school year in which the vaccine was offered, with earlier cohorts that had not been eligible, using an age-period-cohort model adjusted for changes in screening policy. The analysis drew on 13.7 million person-years of follow-up among women aged 20 to under 30.

The relative reduction in cervical cancer was 34 percent in women offered the vaccine at 16 to 18, 62 percent in those offered it at 14 to 16, and 87 percent in those offered it at 12 to 13, with a 95 percent confidence interval of 72 to 94 percent. For CIN3 the reductions were 39, 75 and 97 percent in the same order. The authors estimated that by June 30, 2019 there had been 448 fewer cervical cancers and 17,235 fewer cases of CIN3 than expected in the vaccinated cohorts. They concluded that the programme had "successfully almost eliminated cervical cancer in women born since Sept 1, 1995."

The gradient by age matters. The vaccine does not treat an existing infection. Offered at 12 or 13, before most girls have been exposed to the virus, it prevented the great majority of later cancers. Offered at 16 to 18, when a share of recipients had already been infected, it prevented about a third.

Sweden, a year earlier

The English result was not the first. In 2020 Lei and colleagues reported in the New England Journal of Medicine on an open population of 1,672,983 Swedish girls and women aged 10 to 30, followed from 2006 through 2017 using national registers. Cervical cancer was diagnosed in 19 women who had received the quadrivalent HPV vaccine and 538 who had not: a cumulative incidence of 47 per 100,000 among the vaccinated against 94 among the unvaccinated. After adjustment for age, calendar year, county and parental education, income, country of birth and disease history, the incidence rate ratio was 0.37 overall, 0.12 among women vaccinated before 17, and 0.47 among those vaccinated between 17 and 30.

Scotland and the word zero

The number that this story's title rests on came from Scotland in January 2024. Palmer, Cuschieri and colleagues linked the Scottish cervical screening system to the cancer registry, immunisation and deprivation records for women born between January 1, 1988 and June 5, 1996, with data extracted in July 2020. Among women immunised with the bivalent vaccine at 12 or 13 years of age, no cases of invasive cervical cancer were recorded, irrespective of whether they had received one, two or three doses.

Zero applied to that group and not to every vaccinated woman. Women vaccinated at 14 to 22 with three doses had an incidence of 3.2 per 100,000 person-years, against 8.4 among all unvaccinated women. The study also measured inequality. Unadjusted incidence was 10.1 per 100,000 in the most deprived fifth of Scotland against 3.9 in the least deprived, and the fall after three doses was largest in the most deprived areas, from 13.1 to 2.29 per 100,000. The vaccine narrowed a gap that screening alone had not closed.

The global count

In 2020, according to the GLOBOCAN estimates analysed by Singh and colleagues, there were 604,127 new cases of cervical cancer and 341,831 deaths worldwide. Age-standardised incidence ranged from 2.2 per 100,000 women-years in Iraq to 84.6 in Eswatini; mortality ranged from 1.0 in Switzerland to 55.7 in Eswatini. Incidence was three times higher, and mortality six times higher, in countries with low human development than in those with very high. The World Health Organization has defined elimination as a public health problem as fewer than 4 new cases per 100,000 women per year, and set targets for 2030 of 90 percent of girls vaccinated by 15, 70 percent of women screened by 35 and again by 45, and 90 percent of women with precancer or cancer treated.

What the story does and does not show

The English, Swedish and Scottish studies are observational register analyses, not randomized trials. They compared cohorts offered the vaccine with cohorts that were not, and adjusted for the confounders that could be measured; the 87 percent figure carries a confidence interval from 72 to 94 percent. The vaccinated women in all three countries were still young, mostly under 30, at an age when cervical cancer is uncommon even without vaccination, so the absolute number of cancers prevented so far, 448 in England by mid-2019, is small against the number expected over those women's lifetimes. Scotland's zero describes women vaccinated at 12 or 13 and born after 1988, followed to 2020; it is not a count of all vaccinated women, and it cannot yet say what will happen as those women age into their forties and fifties.

What the data do show is consistent across three countries, two vaccines and three sets of investigators: the earlier the vaccine was given, the larger the fall in invasive cancer, and in the youngest cohorts the fall was close to complete. The 341,831 deaths in 2020 occurred overwhelmingly in countries where the vaccine has not yet reached most girls. The path to elimination exists in the registers of the countries that vaccinated first.

Eighty-seven percent fewer cases. Zero in Scotland.