Cervical determinants of anal HPV infection and high-grade anal lesions in women: a collaborative pooled analysis

Background: Cervical cancer screening might contribute to the prevention of anal cancer in women. We aimed to investigate if routine cervical cancer screening results-namely high-risk human papillomavirus (HPV) infection and cytohistopathology-predict anal HPV16 infection, anal high-grade squamous intraepithelial lesions (HSIL) and, hence, anal cancer.

Methods: We did a systematic review of MEDLINE, Embase, and the Cochrane library for studies of cervical determinants of anal HPV and HSIL published up to Aug 31, 2018. We centrally reanalysed individual-level data from 13 427 women with paired cervical and anal samples from 36 studies. We compared anal high-risk HPV prevalence by HIV status, cervical high-risk HPV, cervical cytohistopathology, age, and their combinations, using prevalence ratios (PR) and 95% CIs. Among 3255 women with anal cytohistopathology results, PRs were similarly calculated for all anal HSIL and HPV16-positive anal HSIL.

Findings: Cervical and anal HPV infections were highly correlated. In HIV-negative women, anal HPV16 prevalence was 41% (447/1097) in cervical HPV16-positive versus 2% (214/8663) in cervical HPV16-negative women (PR 16·5, 95% CI 14·2-19·2, p<0·0001); these values were 46% (125/273) versus 11% (272/2588) in HIV-positive women (4·4, 3·7-5·3, p<0·0001). Anal HPV16 was also associated with cervical cytohistopathology, with a prevalence of 44% [101/228] for cervical cancer in HIV-negative women (PR vs normal cytology 14·1, 11·1-17·9, p<0·0001). Anal HSIL was associated with cervical high-risk HPV, both in HIV-negative women (from 2% [11/527] in cervical high-risk HPV-negative women up to 24% [33/138] in cervical HPV16-positive women; PR 12·9, 95% CI 6·7-24·8, p<0·0001) and HIV-positive women (from 8% [84/1094] to 17% [31/186]; 2·3, 1·6-3·4, p<0·0001). Anal HSIL was also associated with cervical cytohistopathology, both in HIV-negative women (from 1% [5/498] in normal cytology up to 22% [59/273] in cervical HSIL; PR 23·1, 9·4-57·0, p<0·0001) and HIV-positive women (from 7% [105/1421] to 25% [25/101]; 3·6, 2·5-5·3, p<0·0001). Prevalence of HPV16-positive anal HSIL was 23-25% in cervical HPV16-positive women older than 45 years (5/20 in HIV-negative women, 12/52 in HIV-positive women).

Interpretation: HPV-based cervical cancer screening programmes might help to stratify anal cancer risk, irrespective of HIV status. For targeted secondary anal cancer prevention in high-risk groups, HIV-negative women with cervical HPV16, especially those older than 45 years, have a similar anal cancer risk profile to that of HIV-positive women.