The ARTISTIC cohort is one of the few population-based cohorts with long follow-up for HPV associated cancers.
The ARTISTIC trial recruited 24,510 women aged 20-64 undergoing routine cervical screening in Manchester in 2001-2004. Women underwent LBC and HC2 testing over 3 screening rounds. The women have been flagged for cancer incidence and mortality through national registrations with up to 14 years follow-up. Standard Incidence Ratios (SIR, compared to national cancer incidence rates) and rate ratios (RR) using Poisson regression were calculated.
As expected, the incidence rate of cervical carcinoma-in-situ (CIS) was slightly higher than the national rate in this screened cohort (n=388, SIR=1.2, 95% CI: 1.1-1.3), and the incidence of cervical cancer was lower (n=23, SIR=0.6, 95% CI: 0.4-0.9). The ratio of CIS to cervical cancers decreased sharply with age. The ratio was 130:1 in women aged under 30 years (259 CIS:2 cancers), 12:1 in women aged 35-39 (55 CIS:10 cancers), 6:1 in women aged 40-49 (55 CIS:10 cancers), 3:1 in women aged 50-59 (10 CIS:4 cancers) and 1:1 in women aged over 60 (2 CIS:2 cancers). Among women who developed cervical cancer, all tests were HC2 negative in 2 of 17 women aged under 50 years and 3 of 6 women aged over 50 at diagnosis.
Higher rates of CIS and cervical cancer were seen in those HR-HPV positive at entry, particularly those who were infected with HPV16. There were few vulva, vaginal and anal cancers in the cohort (n=5, 1 & 2 respectively), but the rates were elevated amongst those with HPV16 infection (RR=19.3, 95%CI: 3.3-113.3, p=0.001). There was no excess of head & neck cancers among those HPV positive at baseline.
Despite small numbers of cancers in this screened cohort, there was an elevated risk of vulva, vaginal and anal cancers associated with cervical HPV16 infection. We found no association between cervical HPV infection and head and neck cancer incidence. CIS became much rarer in women aged over 40 and cervical cancer was as common as CIS in women aged over 60 years in this screened cohort. The failure to detect HR-HPV in cancers particularly in older women suggests that either a more sensitive HPV test or improved sampling would be appropriate in women aged over 50.