Human papillomavirus types in glandular lesions of the cervix: A meta‐analysis of published studies

Dear Editor, Glandular preinvasive lesions of the cervix (i.e., atypical glandular cells, AGC and adenocarcinoma in situ, AIS) are more likely to be missed in cytological cervical screening than their squamous-cell equivalents, owing to their location in the endocervical canal and difficulties in their cytological and colposcopical assessment.1, 2 Hence, in well-screened populations, although the incidence of cervical cancer has been decreasing over the last three decades, the ratio of adeno/adenosquamous (ADC) to squamous cell carcinoma (SCC) has been rising.1, 3 Screening using high-risk (HR) HPV rather than cytology can improve prevention of ADC.4 ADC: adeno/adenosquamous cell carcinoma; AGC: atypical glandular cells; AIS: adenocarcinoma in situ; ASCUS: atypical squamous cells of undetermined significance; CIN: cervical intra-epithelial neoplasia; HPV: human papillomavirus; HR: high-risk; HSIL: high-grade squamous intraepithelial lesions; LSIL: low-grade squamous intraepithelial lesions; SCC: squamous cell carcinoma Human papillomavirus (HPV) 18 is more frequently detected in ADC than in SCC and this tendency may also apply to other types (e.g., HPV45) that belong to the same alpha-7 species as HPV18.5 However, in contrast to the extensive information about HPV types across the complete spectrum of squamous cervical lesions,6 little is known about the pattern of HPV types in glandular preinvasive lesions, and how it compares to that in ADC. With the aim to improve our understanding of the carcinogenic process for individual HR types from infection of glandular tissue to adenocarcinoma, we performed a systematic literature review of HPV type-specific prevalence data in preinvasive glandular neoplastic lesions, and compared to similar data on women with normal cytology and ADC reported in a recently published meta-analysis.6 Methods have been reported previously.6 In brief, Medline was used to search for publications from January 1990 to November 2011 using combinations of the MeSH terms: "cervical cancer," "cervical intraepithelial neoplasia," "HPV," "human," "female" and "polymerase chain reaction." Eligible studies met the following criteria: (i) use of broad-spectrum consensus PCR-assays based on the primers MY09/11, PGMY09/11, GP5+/6+, SPF10, SPF1/GP6 or L1C1/L1C2, and (ii) reporting of overall and type-specific HPV prevalence, by strata of cyto- and/or histo-pathological cervical diagnoses. Overall HPV prevalence is reported as a percentage of all women tested by consensus PCR. Type-specific HPV positivity is presented among HPV-positive women only, for each of the 13 HR HPV types judged to be carcinogenic or probably carcinogenic (i.e., HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68),1 in the order of their frequency of detection in ADC.7 Type-specific positivity includes that contributed by single and multiple HPV infections. Standard errors of type-specific positivity estimates were computed assuming the nonindependence of cases within the same study using cluster-correlated robust variance estimates.8 We identified 12 studies reporting relevant data on AGC and/or AIS from Europe (n = 4, Belgium, Italy and two studies from The Netherlands), North America (n = 5, all from the United States), Eastern Asia (n = 1, Thailand), Western Asia (n = 1, Kuwait) and South/Central America (n = 1, Brazil). As data on both AGC and AIS were available only for Europe and North America, comparisons of HPV-type distribution with that in women with normal cytology and ADC are restricted to studies from these two regions. Table 1a summarizes the prevalence of any HPV type, and, among HPV-positive women, of 13 HR HPV types, in normal cytology and in different glandular lesions. HPV infection was detected in 10.5, 57.5, 99.1 and 84.6% of women with normal cytology, AGC, AIS and ADC, respectively. Among HPV-positive women, HPV16 positivity was similar in normal cytology (23.4%) and AGC (23.4%) but it increased to 48.7% in AIS and 45.7% in ADC (Table 1a). HPV18 positivity increased slightly between normal cytology (8.9%) and AGC (15.6%), but rose to 51.8% in AIS and 43.9% in ADC. HPV45 positivity was approximately 7% in all glandular lesions. ADC:normal cytology ratio was, therefore, higher for HPV18 (4.93) than HPV16 (1.95) and HPV45 (1.24). All HR HPV types other than HPV16, 18 and 45 showed steep declines in prevalence between AGC and AIS. There were no substantial changes in prevalence between AIS and ADC for any HR HPV type. Table 1b is an adaptation of the findings from Guan et al.6 restricted to lesions of cervical squamous-cell epithelium from Europe and North America. It highlights the much less frequent detection of HPV18 in high-grade lesions of the squamous-cell epithelium (i.e., high-grade intraepithelial squamous lesions, HSIL and cervical intra-epithelial neoplasia, CIN, grades 2 and 3) compared to that in high-grade lesions of the glandular epithelium (i.e., AIS). In conclusion, HPV16 and 18 were highly enriched in ADC and AIS compared to lower grades of diagnosis, confirming their higher potential to cause ADC than other HR-HPV types. The HPV-type distribution was similar in HPV-positive AIS and ADC (of which more than 90% were positive for HPV16 and/or 18). Our present findings support, therefore, the role of AIS as a true ADC precursor and suggest that findings from vaccination trials on prevention of AIS2 can be expected to hold for ADC. The HPV-type distribution in AGC, on the other hand, resembled that in normal cytology, suggesting that, like its squamous-cell equivalent low-grade intraepithelial lesion and CIN1, these lesions are more of a proxy of HPV infection than cancer precursors. The greater importance of HPV18 in ADC in comparison to SCC has been well described.7 However, we were able to show that this is also the case for AIS, which is in contrast with the relatively rare detection of HPV18 in HSIL or CIN2/3 in comparison to SCC. Restriction to European and North-American studies was unavoidable, but it probably allowed better quality and consistency in cytological and histological classification. It did not overcome, however, the paradox of lower HPV positivity in ADC (84.6%) than in AIS (99.1%) and SCC (92.0%). Some misclassification of extra-uterine adenocarcinoma as ADC9 and the existence of some rare varieties of ADC that are not related to HPV infection10 cannot be ruled out. Yours sincerely, The authors thank Dr. Silvia de Sanjosé and Dr. Laia Bruni (Institut Catalan d'Oncologia, Barcelona, Spain), for helping to provide information on HPV prevalence in women with normal cytology. Peng Guan* , Gary M. Clifford*, Silvia Franceschi*, * International Agency for Research on Cancer, Lyon, France, China Medical University, Shenyang, People's Republic of China.