Recently, pooled IARC studies1, 2 and meta-analyses3, 4 have showed that at least 13 human papillomavirus (HPV) types, including types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68, are commonly associated with invasive cervical cancer (ICC). Based on HPV type prevalence data from these studies, current HPV vaccines against only HPV16/18 are considered to prevent a majority (>70%) of cervical cancer worldwide.5, 6 However, HPV data in East Asia have not been fully evaluated in previous pooled and meta-analyses. The present study thus focused on HPV type prevalence in Japan. We performed a meta-analysis of published data to obtain the representative results, and investigated HPV type prevalence and type-specific risks for cervical carcinogenesis in Japan. Furthermore, obtained data were compared with those in China, Korea and other regions.
Source articles presenting HPV prevalence data among Japanese women were identified from National Library of Medicine (PubMed). For the meta-analysis, the following inclusion criteria were considered: (i) Studies were published between 1995 and 2005. (ii) Studies had to use PCR-based assays to identify at least 16 strains of HPV6, 11, 16, 18, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59 and 68. (iii) Studies had to include at least 20 HPV-positive women with squamous intraepithelial lesion (SIL), cervical intraepithelial neoplasia (CIN) or ICC. When data or data subsets from an identical study had been published in more than one article, only the publication with the largest sample size was included. However, data from different studies conducted by the same study group were included. Overall, a total of 14 Japanese studies were identified for the present study.7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 For some articles, additional type-specific data were obtained from the authors.11, 12, 20 HPV type prevalence data were collected separately for squamous cell carcinoma (SCC) and for adeno- and adenosquamous carcinoma (ADC). Where histological data were not reported, ICC cases were classified as unspecified carcinoma (UC). On the basis of the pooled analysis of IARC studies,2 HPV types were separated into 2 groups. High-risk HPVs considered as carcinogenic or probably carcinogenic included HPV16, 18, 26, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68, 73 and 82; all other HPV types were classified as low-risk types. To evaluate HPV genotype-specific risks for progression from LSIL (low grade SIL) to HSIL (high grade SIL) or more, prevalence ratios [(HSIL + ICC):LSIL] and 95% confidence intervals were calculated after adjusting for study area, specimen used for HPV DNA testing, and PCR primers. Logistic regression model was used for statistical adjustment, and the analysis was carried out using JMP 6.0J statistics package (SAS Institute, Cary, NC, USA). The p-values obtained in all tests were considered significant at <0.05.
This analysis included 7,262 Japanese women (4,941 normal cytology, 475 LSIL, 720 HSIL and 1,126 ICC) from 14 Japanese HPV studies. Consensus primers for the L1, E6/E7 or E7-LCR region or GP5+/GP6+ primers were employed in these studies. In this meta-analysis, HPV prevalence was 10.2% in women with normal cytology, 79.4% in LSIL, 89.0% in HSIL and 87.4% in ICC. HPV positivity was significantly lower in ADC (69.8%) than in SCC (90.7%). Since overall and type-specific HPV prevalence in UC was very similar to that in SCC, UC cases were combined with SCC cases for comparison of HPV type-specific prevalence by histological type.
HPV type distributions by cervical disease were summarized in Table I. In ICC, the most common HPV types were, in order of decreasing prevalence, HPV16 (44.8%), 18 (14.0%), 52 (7.0%), 58 (6.7%), 33 (6.3%), 31 (5.1%), 35 (2.3%), 51 (1.0%) and 56 (0.9%). HPV16 was the predominant type (49.2%) in SCC, although HPV18 was the most prevalent (58.2%) in ADC. On the basis of the comparison of HPV type distributions between LSIL and HSIL/ICC, the type-specific risks for progression from LSIL to HSIL/ICC were analyzed (Table I). HPV16, 18 and 33 conferred a significantly higher risk. Seven types of HPV16/18/31/33/35/52/58 (the prevalence ratio; 1.86, 95% CI 1.63–2.15) presented a higher risk of progression than the other high-risk types (0.24, 95% CI 0.17–0.33) and low-risk types (0.18, 95% CI 0.09–0.37).
|HPV types||Normal cytology(n = 503)||LSIL (n = 377)||HSIL(n = 641)||SCC1 (n = 917)||ADC (n = 67)||ICC (n = 984)||Prevalence ratios2 (HSIL + ICC):LSIL|
Among Japanese women with ICC, HPV16 and 18 were the most common types, which was consistent with the results from other regions. A pooled analysis demonstrated that HPV16 and 18 were associated with 73.5% of ICC in Southeast Asia, 76.9% in Northern Africa and 71.5% in Europe/North America, and that HPV52 and 58 were detected in 6.1% of ICC cases in Southeast Asia, 1.5% in Northern Africa and 1.1% in Europe/North America.2 In Japan, however, HPV16 and 18 were less frequently identified (58.8%), and HPV52 and 58 were more common (13.7%). In China, although HPV16 and 18 were predominant (87.1% of all ICC cases from 5 areas), a multicenter study showed that HPV52 and 58 were locally common in Hong Kong (11.8%) and Guangzhou (southern China, 8.3%).21 Other small studies also reported that HPV52 and 58 were highly prevalent in southern areas of China including Taiwan (19.2%, 19/94), Shanghai (48%, 15/31) and Jiangxi (20.0%, 8/40),22, 23, 24 while rarely detected in central China (3.3%, 1/30).25 A recent Korean study showed that HPV16 and 18 were detected in only 58.4% of women with ICC, and HPV52 and 58 were detected in 9.7% of ICC cases.26 In addition, HPV31 and 33 were more common in Japan (11.5%) and Korea (12.5%),26 compared with that in Southeast Asia (3.8%), Northern Africa (7.4%) and Europe/North America (4.5%). By contrast, HPV45 was rarely associated with ICC in Japan (0.3%) and Korea (1.4%),26 although common in Southeast Asia (7.9%), Northern Africa (5.6%) and Europe/North America (9.0%).2 Since HPV vaccines confer only type-specific immunity, HPV type prevalence in ICC cases has important implications for HPV vaccine strategy. These observations suggest that adding HPV31/33/52/58 vaccines to current HPV16/18 vaccines may be necessary for cervical cancer prevention in East Asia.
The comparison of HPV type distributions between LSIL and HSIL/ICC revealed that 7 types of HPV16/18/31/33/35/52/58 present a higher progression risk than the other high-risk types in Japan. Interestingly, the highest risk of HPV16 and 33 was consistent to results from a recent Dutch study.27 Since our finding suggests that LSIL associated with HPV16/18/31/33/35/52/58 may preferentially progress to HSIL/ICC, distinguishing these 7 types may be very helpful in the management of Japanese women with LSIL. To confirm significant cancer risks for women with these types, a large-scaled cohort study is under way in Japan.
We confirmed that HPV16 and 18 are less frequently identified and HPV31, 33, 52 and 58 are more common in ICC cases in Japan compared with Southeast Asia, Northern Africa, Europe and North America, with HPV16/18/31/33/35/52/58 carrying a higher risk of progression from LSIL to HSIL/ICC in Japan. With regard to HPV type prevalence, Japan appears to share similarity with China and Korea. Geographic variations in HPV type distributions suggest that a different strategy for HPV screening and vaccination will be necessary in Japan, probably in East Asia.