Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Classification of weakly carcinogenic human papillomavirus types: addressing the limits of epidemiology at the borderline

Classification of weakly carcinogenic human papillomavirus types: addressing the limits of... Virtually all cases of cervical cancer are caused by persistent infections with a restricted set of human papillomaviruses (HPV). Some HPV types, like HPV16 and HPV18, are clear and powerful carcinogens. However, the categorization of the most weakly carcinogenic HPV types is extremely challenging. The decisions are important for screening test and vaccine development. This article describes for open discussion an approach recently taken by a World Health Organization International Agency for Research on Cancer (IARC) Monographs Working Group to re-assess the carcinogenicity of different HPV types. Background and rationale carcinogens [3]. The process is described here to promote As a group, human papillomaviruses (HPV) are proven needed debate on what an improved approach might be. human carcinogens. But there are >> 100 HPV genotypes and only a small fraction have any known carcinogenic IARC formally considers the carcinogenicity of exposures potential [1]. Therefore, moving from broad acceptance of to humans. Whereas human carcinogenicity might best be the carcinogenicity of persistent HPV infection to specific considered for some agents like HPV as a continuum of conclusions about individual genotypes requires consid- probabilities without a clear breaking point, IARC classi- eration of each type as an individual agent. Such a type- fies carcinogens categorically as carcinogenic (Group 1), by-type evaluation proves to be very difficult, and probably carcinogenic (Group 2a), possibly carcinogenic stretches epidemiology to its limits because of issues of (Group 2b), not classifiable (Group 3), or probably not confounding and misclassification detailed below. How- carcinogenic (Group 4). There has been very little experi- ever, as described in the accompanying article by Castle mental work on the carcinogenicity of HPV types except [2], it is important for the development of screening tests for HPV16 and HPV18; thus, epidemiologic evidence has and vaccines to judge each HPV type separately; thus, it is been unusually important. Epidemiologic evidence for worth considering how well one actually can decide the carcinogenicity of HPV was originally presented in whether a given HPV type is carcinogenic or not. This arti- IARC Monograph Volume 64 [4], and was extensively cle describes an approach taken by a Working Group of updated in IARC Monograph Volume 90 based on data the World Health Organization International Agency for available as of February 2005 [5]. In February 2009, Vol- Research on Cancer (IARC), when asked to re-assess for ume 100b updated the data once again and it is this latest official purposes which HPV types should be grouped as update addressed here. For the purposes of IARC, epide- Page 1 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 miologic evidence is categorized as sufficient, limited, clade containing a few genetically related species (Figure inadequate or suggesting lack of carcinogenicity. The epi- 1). Epidemiologic data do not support cervical carcino- demiologic data are combined with experimental evi- genicity for other species in the alpha genus or for other dence (in this case lacking for most HPV genotypes) to genera. To save considerable space presenting null evi- arrive at the final Groups 1–4. dence, this section will not include data related to HPV species alpha-1, -2, -3, -4, -8, -10 (other than HPV 6), -13, HPV carcinogenicity has been established most convinc- or 14/15. These species contain HPV types that cause skin ingly for cervical cancer, and this discussion will be lim- or genital warts, minor cytologic atypia, and often no ited to cervical carcinogenicity. To date, no HPV type has apparent disease. been proven to be carcinogenic only at sites other than the cervix. In Monograph 64 in 1995, HPV 16 (alpha-9) and HPV 18 (alpha-7) were classified as cervical carcinogens. HPV 31 and HPV 33 in alpha-9 were categorized as probably car- HPV evolution as the guiding principle of HPV behaviour (the high-risk clade) cinogenic [4]. In 2005, the group of cervical carcinogens HPV behaviour at the cervix is strongly correlated with (Group 1) was expanded to include the following 13 phylogenetic (i.e. evolutionary) categories [6]. All HPV types: alpha-5 genotype HPV 51; alpha-6 genotypes HPV genotypes that are known to be cervical carcinogens 56 and HPV 66; alpha-7 genotypes HPV 18, HPV 39, HPV belong to the alpha genus in an evolutionary branching or HPV6 HPV11 HPV13 HPV74 1, 8, 10, 13 HPV44 HPV55 HPV91 HPV7 Genital Warts HPV40 HPV32 HPV42 And HPV54 HPV52 HPV67 Commensal HPV33 HPV58 HPV16 HPV31 HPV35 HPV34 HPV73 11  5, 6, 7, 9, HPV59 HPV18 HPV45 11 Squamous HPV70 HPV39 HPV68 Evolutionary * HPV85 Cell Carcinoma HPV26 HPV69 Tree HPV51 HPV82 HPV30 (millions of HPV53 Adenocarcino HPV56 HPV66 years) HPV57 HPV2a ma HPV27 HPV71 HPV90 HPV61 HPV72 HPV62 2, 3, 4, 15 HPV81 HPV83 3 HPV89 HPV84 Commensal HPV86 HPV87 HPV28 Infections HPV3 HPV10 HPV29 HPV77 Phylogen Figure 1etic analysis of anogenital HPV types [6] Phylogenetic analysis of anogenital HPV types [6]. Branches determined by 100 bootstrap estimations using each of the methods in the following order: Bayesian credibility value, parsimony bootstrap percentage based on nucleotide alignment, and parsimony bootstrap percentage based on amino acid alignment. All definitely, probably, and possibly carcinogenic HPV types belong to one phylogenetic clade of the alpha genus. Page 2 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 45, and HPV 59; and alpha-9 genotypes HPV 16, HPV 31, tions are present in a large minority of women, although HPV 33, HPV 35, HPV 52, and HPV 58 [5]. they might not be transmitted from the same partner or at the same time. Given the existence of some very power- In the four years between Monograph 90 in 2005 and the fully carcinogenic types, notably HPV 16 and HPV 18, recent update, new evidence further supported that HPV determining which weaker and/or less common types are types in the high-risk clade of the alpha genus HPV type also carcinogenic becomes (for the epidemiologist) an cause virtually all cases of cervical cancer worldwide [7,8]. issue of confounding. None of the traditional approaches In case-control studies, the odds ratios associating cervical to control of confounding are entirely successful. Because cancer and its immediate precursor, CIN3, with HPV DNA HPV 16 causes approximately 50% of cases of cervical positivity for these types in aggregate has consistently cancer, logistic regression and similar approaches will par- exceeded 50. It is persistent infections that are associated simoniously attribute cases associated with both HPV16 with extremely high absolute risk of CIN3 and cancer. In and a less important type to HPV16. HPV 18 is the second cohort studies, women who test negative to this group of most important cervical carcinogen, responsible for HPV types are at extremely low subsequent risk of CIN3, approximately 15–20% of cervical cancer of all histologic cancer, and cancer death for more than 10 years [9-11]. types combined (and a higher fraction of adenocarcino- mas). If a type occurs with either HPV 16 or HPV 18, its Because persistent infection with a restricted group of association with cervical cancer might be confounded by HPV types is a nearly necessary cause of cervical cancer, a either of these powerful carcinogens. For types causing reconsideration of HPV and cervical carcinogenicity based only a very small fraction of cervical cancer, confounding on the new data must decide whether any additional types by any of the more important types is possible. within the high-risk clade are also carcinogenic and whether any types in that clade that were previously cate- Dealing with confounding by exclusion, i.e. examining gorized as carcinogenic should be downgraded. The types the possibility of carcinogenicity of a more minor type in the high-risk clade are listed in Table 1. among cancer specimens that do not contain a more important type, becomes a problem of misclassification. Conceptual issues in deciding which of the types This main epidemiologic criterion used for classification in the high-risk clade are carcinogenic of an HPV type as a carcinogen, finding the HPV genotype From a virologic perspective, the definitive proof of carci- as a single infection in a cervical scrape or biopsy speci- nogenicity of an HPV type is finding transcriptionally men in a woman with cancer, might sometimes be too lax active HPV in a tumour. HPV is not a "hit and run" carcin- and prone to error. Colposcopic biopsies and cytology ogen, and transcriptional activity is needed for mainte- specimens can be misdirected and fail to obtain the criti- nance of the cancer phenotype. In cervical cancer cell cal cells, while contamination of scrapes and biopsies lines, blocking transcriptional activity by antisense RNA from lower-grade lesions that often surround cancers can leads to apoptosis. This level of evidence is simply lacking detect types other than the causal one. Studies relying on for virtually all HPV types. The vast body of evidence testing of microdissected cervical malignancies will relates simply to finding HPV DNA at the same time of address these issues, but large-scale highly accurate data cervical neoplasia. are not yet available. However, relying on testing of scrapes or biopsies, by Difficulty with control selection adds another level of DNA testing, leads to difficulties. The alpha HPV types complexity in assessing carcinogenicity. Cervical cancer share a common route of transmission and multiple infec- typically follows age infection by decades. HPV transmit- Table 1: HPV Types that Were Considered in Monograph 100B Alpha Species Types Categorized as Definite Carcinogens in Monograph 90 Other Types in Species 551 26 69 82 656 66 30 53 7 18 45 39 59 68 70 85 97 9 16 31 33 35 52 58 67 11 34 73 Page 3 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 ted at young ages usually become undetectable by DNA or ous lesions (CIN3 and the more equivocal CIN2 which RNA assays and no sensitive serologic assay exists to meas- occur in approximately 1% of screened women), which ure HPV exposure. Consequently, odds ratios based on a often used as ethical surrogate endpoints in prospective comparison of HPV prevalence at the time of case diagno- studies and clinical trials; there are now sufficient data for sis to age-matched HPV point prevalence in controls do invasive cancers and it appears that HPV types have differ- not estimate true relative risks. ent potential to progress from CIN2/3 to invasive cervical cancer [14]. The background frequency of cervical HPV There is not much type-specific prospective data on the infection in the general female population was estimated carcinogenicity of individual HPV genotypes. The availa- from a large meta-analysis of HPV genotypes found in ble studies have categorically shown the unique carcino- women with normal cytology [15], as shown in Table 2. genicity of HPV 16 and, to a lesser extent, HPV 18 [9,12]. Although women included in meta-analyses of cervical Khan et al. observed a risk for the remaining women pos- cancer and normal cytology differed by age, region and itive by hc2 after excluding those positive for HPV 16 or HPV testing protocols, it was considered the best method HPV 18 (including an unknown mix of the types 31, 33, to identify a reasonable threshold of confounding and 35, 39, 45, 51, 52, 56, 58, 59 and 68) of only 3.0% (1.9– misclassification for each type. Ancillary analyses examin- 4.2) compared with 0.8% (0.6–1.1) among women who ing the issue of most important types by region were also were HPV negative at baseline. Thus, there is not strong scrutinised. and convincing long-term prospective evidence for indi- Uniqueness of HPV16 and HPV18 vidual HPV types other than HPV16 and HPV18. Comparing the prevalences in women with normal cytol- Finally, the accuracy of detection of HPV genotypes differs ogy to the prevalences for cancers compiled by Smith et al. between the major PCR-based systems used to generate (2007) (Table 2), obvious "case-control" differences can most of the data [13]. Each of the systems shows differen- be seen. The most clearly carcinogenic genotypes, HPV 16 tial sensitivity, and some have exhibited cross-reactivity of and HPV 18 in particular, are more common among can- detection. These detection issues are not critical for evalu- cers and cytologically normal women (and even low- ating the most important HPV types, but make it difficult grade lesions [16]). HPV 18 is especially common in ade- to clarify the role of the most weakly carcinogenic and nocarcinomas [17], as are other members of the alpha 7 least common ones. clade of which HPV18 is a member. The large amount of data regarding HPV 16 and HPV 18 was thought to pro- With these caveats, the cervical carcinogenicity of the HPV vide ample epidemiologic evidence of carcinogenicity types listed above varies in strength in a continuum with- (leading to an overall classification of Group 1). out clear breakpoint, from extremely strong (i.e. HPV 16 and, to a lesser degree, HPV 18) to weak, but still probably The 8 most important carcinogenic HPV types carcinogenic in rare instances (e.g. HPV 68, see below). Including HPV 16 and HPV 18, eight HPV types (alpha-7 Evaluators taking one extreme position could claim that types HPV 18 and 45, alpha-9 types 16, HPV 31, 33, 35, there is reasonable evidence for carcinogenicity of virtu- 52 and 58) are the most common types found in cancers ally all the types in the species listed above, extending fur- in both the IARC meta-analysis [8] and the ICO study [7], ther the list established in Monograph 90. Strict in all regions of the world providing data. Though very interpreters of causal criteria could argue for a return to a often found in non-invasive lesions, these types are all much more limited list. But based on current evidence, no much more common in cancer case specimens than in cut-point between sufficient, limited, and inadequate epi- controls, providing sufficient epidemiologic evidence of demiologic evidence is entirely defensible. carcinogenicity (Group 1). The borderline carcinogens: using HPV 6 as an estimator of Approach taken by the IARC working group Data Sources residual confounding The IARC Working Group chose the following pragmatic To move beyond the most evidently carcinogenic eight approach to creating an imperfect cut-point between suf- HPV types, the Working Group chose an estimator of the ficient, limited, and inadequate epidemiologic evidence percentage of cancers that might contain HPV DNA by for cervical carcinogenicity: Only types in the high-risk accumulated and unknown measurement errors alone. clade were considered because data seemed grossly inade- The group made use of HPV 6 for this estimation. Specifi- quate for all others. The most recent accumulated data on cally, HPV 6, the important and common cause of benign type-specific HPV prevalence in cervical cancers were condyloma acuminata (external genital warts) was con- drawn from a very large single project [7] and from meta- sidered to be a low-risk type, not classified as a cervical analyses performed by IARC ([8] updated by G. Clifford). carcinogen, which only uncommonly is detected in cervi- Excluded from consideration were high-grade precancer- cal cancer specimens. [Of note, it remains possible that Page 4 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 Table 2: Meta-analyses of type-specific HPV DNA prevalence in invasive cervical cancer [15] and women with normal cytology [14,17] Invasive cervical cancer Normal N tested % pos 95% CI N tested % pos 95% CI HPV16 14595 54.4 53.6–55.2 76385 2.6 2.5–2.8 HPV18 14387 15.9 15.3–16.5 76385 0.9 0.8–1.0 HPV33 13827 4.3 4.0–4.6 74141 0.5 0.4–0.5 HPV45 9843 3.7 3.3–4.1 65806 0.4 0.4–0.4 HPV31 11960 3.5 3.2–3.9 74076 0.6 0.6–0.7 HPV58 10157 3.3 2.9–3.6 72877 0.9 0.8–1.0 HPV52 9509 2.5 2.2–2.8 69030 0.9 0.8–1.0 HPV35 9507 1.7 1.5–2.0 74084 0.4 0.3–0.4 HPV59 6972 1.0 0.8–1.3 64901 0.3 0.2–0.3 HPV51 7339 0.7 0.5–0.9 67139 0.6 0.6–0.7 HPV56 7427 0.7 0.5–0.9 68121 0.5 0.5–0.6 HPV39 7078 0.6 0.5–0.9 64521 0.4 0.3–0.4 HPV68 6723 0.5 0.3–0.7 63210 0.3 0.2–0.3 HPV73 5837 0.5 0.3–0.7 44063 0.1 0.1–0.1 HPV66 6664 0.3 0.2–0.5 59774 0.4 0.3–0.4 HPV70 5159 0.2 0.1–0.4 35014 0.3 0.3–0.3 HPV82 5352 0.1 0.1–0.3 42536 0.1 0.0–0.1 HPV6 9911 0.5 0.4–0.7 58370 0.3 0.2–0.3 HPV53 not reported 44,058 0.4 0.4–0.4 HPV26 not reported 44,098 0.0 0.0–0.1 HPV85 not studied in ICC meta-analysis 9,622 0.1 0.1 – 0.1 HPV67 not studied in ICC meta-analysis 18,041 0.0 0.0 – 0.0 HPV34 not studied in ICC meta-analysis 42,588 0.0 0.0 – 0.1 HPV30 not studied in ICC meta-analysis 8,773 0.0 0.0 – 0.1 HPV97 not studied in ICC meta-analysis not studied in normal cytology meta-analysis HPV 6, and other "low risk types" can cause cancer in Table 3: Meta-analysis of type-specific HPV DNA prevalence in extremely rare virus-host circumstances.] When detected, invasive cervical cancer [15], updated with 63 newly published even without detection of a more likely causal type, the studies by Clifford (IARC) Working Group judged that misclassification of some N tested N pos % pos 95% CI kind was a more likely explanation than causality. As given in Table 3, the best IARC estimate of detection of hpv39 13370 172 1.29 1.10 – 1.48 HPV 6 in cervical cancers, not necessarily as a single infec- hpv59 13471 172 1.28 1.09 – 1.47 tion, was judged to be 0.45%, 95% CI 0.35 – 0.56, based hpv51 13057 151 1.16 0.97 – 1.34 on 14,912 cases of cancer ([8] updated by G. Clifford for hpv56 13247 103 0.78 0.63 – 0.93 the Working Group). The Working Group pragmatically hpv68 11982 73 0.61 0.47 – 0.75 made the following rule: An individual HPV type in the hpv73 9939 48 0.48 0.35 – 0.62 high-risk alpha clade (i.e. one with an elevated prior prob- ability of being carcinogenic due to analogy to closely hpv6 14912 68 0.45 0.35 – 0.56 related viral types in the same or closely-related species) was considered to have sufficient epidemiologic evidence hpv53 8140 34 0.42 0.28 – 0.56 of carcinogenicity if its prevalence in cancers was 1) signif- hpv66 12118 47 0.39 0.28 – 0.50 icantly greater than that of HPV 6, and 2) significantly hpv70 10503 35 0.33 0.22 – 0.44 hpv82 9265 25 0.27 0.16 – 0.38 enriched in comparison to the background estimate for hpv26 6111 8 0.13 0.04 – 0.22 the general population, i.e. women with normal cytology. HPV types 30, 34, 67, 85 and 97 were not studied in the meta-analysis By this logic, four more types were categorized as having dataset. sufficient epidemiologic evidence leading to their classifi- Page 5 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 cation as definite (Group 1) carcinogens: alpha-5 type 51 b) HPV 18 is also important, particularly for adenocar- (1.16% of cervical cancer), alpha-6 type 56 (0.78%), and cinoma. alpha-7 HPV types 39 (1.29%) and 59 (1.28%). c) Six additional types in alpha-7 (HPV 45) and alpha- Classification of other types in the high-risk clade 9 (HPV 31, HPV 33, HPV 35, HPV 52, HPV 58) com- The remaining types in the high-risk alpha clade were con- prise the remainder of the eight types that are the most sidered, as a group, to have limited evidence to support important globally, with some regional variation in carcinogenicity. If phylogeny can be taken to predict the etiologic fractions of cancers due to each type. behaviour, it is possible that most of these types can very rarely cause cancer. Indeed, many of the types have been d) There are small and less certain, incremental etio- detected, albeit uncommonly (not significantly greater logic contributions of another group of carcinogenic than HPV 6), in cancers. There are not enough data, even types from alpha-5 (HPV 51), alpha-6 (HPV 56), and after testing of many thousands of specimens, to be sure alpha-7 (HPV 39 and HPV 59). Each causes a few per- which ones are carcinogenic or not. Furthermore, some of cent at most of cervical cancer cases worldwide. There these types, namely 30, 34, 67, 85 and 97 have not yet has not been any large-scale study of regional variabil- been studied in the IARC cervical cancer meta-analysis ity for these uncommon types approach. Nevertheless, within this group, there are two types, alpha-7 type HPV 68 and alpha-11 type HPV 73, for e) There is an unresolved dividing line between the which the data are slightly stronger than for the others HPV types with the weakest evidence judged to be suf- despite methodologic challenges. One of the major PCR- ficient and those judged to have highly suggestive yet based testing methods (SPF10) cannot distinguish these limited data (e.g. alpha-7 HPV 68 is categorized as two types because their amplicons using those primers are probably carcinogenic due to experimental evidence identical [18]. Neither of these two types is optimally while alpha-11 HPV 73 is possible). detected by MY09-MY11 dot blot [19]. Nonetheless, the data supporting the carcinogenicity of HPV 68 and HPV f) The expanded data for HPV66 were re-evaluated and 73 are very suggestive although not sufficient. Ultimately, the evidence was judged to be very limited now that the existence of a cell line whose immortalization is sus- more cases have been studied showing that it is very tained by a subtype of HPV68 (ME180), led the Working rarely found in cancers despite being relatively com- Group to classify HPV 68 as a probable carcinogen mon in the newly collated data on women with nor- (Group 2A). The rest were called possible (Group 2B). mal cytology. HPV 53, also in alpha species 6, shows the same pattern of relative common population prev- Re-classification of HPV 66 alence with extremely rare cases of occurrence alone in Overall, the Working Group approach led to the re-classi- cancer. The Working Group noted that for these types fication from Monograph 90 of HPV 66 to possible in particular, there could be harm to public health if (Group 2B), although the epidemiologic evidence of car- the types are viewed with excessive concern; including cinogenicity was previously judged sufficient (Group 1). these types as carcinogenic in screening assays would In the assembly of much more testing data from cancer decrease the specificity and positive predictive value of cases, HPV 66 has been found so rarely that its percentage the assays with virtually no gain in sensitivity and neg- of detection is less than the relative percentage of detec- ative predictive value [20]. tion among the general population. In the Working Group review of each individual article, HPV 66 was g) There are several types within the high-risk clade found as a single infection in cancers with extreme rarity, that have extremely sparse or no evidence of carcino- well below the threshold of possible confounding and genicity. For some types there are anecdotal but very misclassification. interesting cases that merit pursuit of additional carci- nogenic types. There have been potentially underap- Summary preciated reports of alpha-9 type HPV 70 found as There is no perfect way to categorize a continuum of car- single infections in cancer, but the supportive data are cinogenic potential. The Working Group arrived at the fol- sparse. There are only a few reports of HPV 67 in can- lowing conclusions, with a healthy scepticism concerning cer [21-24], which is intriguing because this is the only the process: known type in the alpha-9 species that is not catego- rized as carcinogenic. For a few types in the high-risk a) Persistent HPV 16 infection is a uniquely powerful clade, no reports of invasive cancers with single-type human carcinogen. infections were found, but isolated reports might exist. Page 6 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 specific implications in cervical neoplasia. Vaccine 2008, h) The possible role of immunosuppression in HPV 26(Suppl 10):K1-16. carcinogenicity was not emphasized in the Working 8. Smith JS, Lindsay L, Hoots B, Keys J, Franceschi S, Winer R, Clifford Group discussions. To cause cancer, an HPV infection GM: Human papillomavirus type distribution in invasive cer- vical cancer and high-grade cervical lesions: a meta-analysis must persist and it is possible that some HPV types are update. Int J Cancer 2007, 121(3):621-32. only weak carcinogens because they persist poorly. For 9. Khan MJ, Castle PE, Lorincz AT, Wacholder S, Sherman M, Scott DR, Rush BB, Glass AG, Schiffman M: The elevated 10-year risk of example, the carcinogenicity of alpha-5 type HPV 26 cervical precancer and cancer in women with human papil- has been supported by a recent report of multiple peri- lomavirus (HPV) type 16, or 18, and the possible utility of ungual cancers in an immunosuppressed individual, type-specific HPV testing in clinical practice. J Natl Cancer Inst 2005, 97(14):1072-9. containing high viral loads and active transcription of 10. Sankaranarayanan R, Nene BM, Shastri SS, Jayant K, Muwonge R, HPV 26 alone [25]. HPV 26 is an uncommon type; Budukh AM, Hingmire S, Malvi SG, Thorat R, Kothari A, Chinoy R, perhaps the immunosuppression in this individual Kelkar R, Kane S, Desai S, Keskar VR, Rajeshwarkar R, Panse N, Din- shaw KA: HPV screening for cervical cancer in rural India. N was a significant contributor to carcinogenesis. As one Engl J Med 2009, 360(14):1385-94. avenue of research, there should be more studies of 11. Schiffman M, Wacholder S: From India to the world – a better way to prevent cervical cancer. N Engl J Med 2009, whether HPV types in invasive cancers in HIV-infected 360(14):1453-5. individuals differ in type from the types found in 12. Wheeler CM, Hunt WC, Schiffman M, Castle PE, Atypical Squamous immunocompetent individuals [25]. Cells of Undetermined Significance/Low-Grade Squamous Intraepi- thelial Lesions Triage Study Group: Human papillomavirus geno- types and the cumulative 2-year risk of cervical precancer. J Conclusion Infect Dis 2006, 194(9):1291-9. 13. Gravitt PE, Coutlée F, Iftner T, Sellors JW, Quint WG, Wheeler CM: When epidemiology serves as the science informing pub- New technologies in cervical cancer screening. Vaccine 2008, lic health policy, its limitations must be acknowledged. 26(Suppl 10):K42-52. Weak causal associations with one HPV type are extremely 14. Clifford GM, Smith JS, Aguado T, Franceschi S: Comparison of HPV type distribution in high-grade cervical lesions and cervical hard to prove in the presence of powerful confounding by cancer: a meta-analysis. Br J Cancer 2003, 89(1):101-5. strong carcinogens like HPV16. A coming generation of 15. de Sanjosé S, Diaz M, Castellsagué X, Clifford G, Bruni L, Muñoz N, intensive molecular studies of microdissected cancers, Bosch FX: Worldwide prevalence and genotype distribution of cervical human papillomavirus DNA in women with nor- with examination of transcriptional activity, may clarify mal cytology: a meta-analysis. Lancet Infect Dis 2007, 7(7):453-9. the borderline between carcinogenic and non-carcino- 16. Clifford GM, Rana RK, Franceschi S, Smith JS, Gough G, Pimenta JM: Human papillomavirus genotype distribution in low-grade genic HPV types. For public health purposes, the main cervical lesions: comparison by geographic region and with agents that merit consideration in screening tests and vac- cervical cancer. Cancer Epidemiol Biomarkers Prev 2005, cines have already been identified. 14(5):1157-64. 17. Clifford G, Franceschi S: Members of the human papillomavirus type 18, family (alpha-7 species) share a common association Acknowledgements with adenocarcinoma of the cervix. Int J Cancer 2008, A special thank for comments and contributions from the IARC Working 122(7):1684-5. 18. Molijn A, Kleter B, Quint W, van Doorn LJ: Molecular diagnosis of Group and the IARC Secretariat (in particular to Veronique Bouvard, human papillomavirus (HPV) infections. J Clin Virol 2005, Responsible Officer for Volume 100B and Vincent Cogliano, Head of the 32(Suppl 1):S43-51. Monograph Program). Dr FM Buonaguro, has been recipient of Merck and 19. Handisurya A, Rieger A, Bankier A, Koller A, Salat A, Stingl G, Kirn- GSK support for Conference organization, and is a 2008 Member of the bauer R: Human papillomavirus type 26, infection causing multiple invasive squamous cell carcinomas of the fingernails GSK-sponsored PanEuropean HPV Immunology Group. in an AIDS patient under highly active antiretroviral ther- apy. Br J Dermatol 2007, 157(4):788-94. Epub 2007, Jul 19. Erratum References in: Br J Dermatol. 2007, Dec;157(6):1304. 1. Schiffman M, Castle PE, Jeronimo J, Rodriguez AC, Wacholder S: 20. Schiffman M, Khan MJ, Solomon D, Herrero R, Wacholder S, Hild- Human papillomavirus and cervical cancer. Lancet 2007, esheim A, Rodriguez AC, Bratti MC, Wheeler CM, Burk RD, PEG 370(9590):890-907. Group; ALTS Group: A study of the impact of adding HPV 2. Castle P: The Evolving Definition of Carcinogenic Human types to cervical cancer screening and triage tests. J Natl Can- Papillomavirus Infectious. Agents and Cancer 2009 in press. cer Inst 2005, 97(2):147-50. 3. Bouvard V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, Ben- 21. Gudleviciene Z, Didziapetriene J, Ramael M, Uleckiene S, Valuckas KP: brahim-Tallaa L, Guha N, Freeman C, Galichet L, Cogliano V, WHO Human papillomavirus and p53 polymorphism in Lithuanian International Agency for Research on Cancer Monograph Working cervical cancer patients. Gynecol Oncol 2006, 102(3):530-3. Group: A review of human carcinogens – Part B: biological 22. Wentzensen N, Schiffman M, Dunn ST, Zuna RE, Walker J, Allen RA, agents. Lancet Oncol 2009, 10(4):321-2. Zhang R, Sherman ME, Wacholder S, Jeronimo J, Gold MA, Wang SS: 4. IARC Monographs: Human Papillomaviruses. In IARC Monographs Grading the severity of cervical neoplasia based on com- on the Evaluation of Carcinogenic Risks to Human Volume 64. IARC, Lyon, bined histopathology, cytopathology, and HPV genotype dis- France . tribution among 1,700 women referred to colposcopy in 5. Cogliano V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, Oklahoma. Int J Cancer 2009, 124(4):964-9. WHO International Agency for Research on Cancer: Carcinogenic- 23. Gargiulo F, De Francesco MA, Schreiber C, Ciravolo G, Salinaro F, ity of human papillomaviruses. Lancet Oncol 2005, 6(4):204. Valloncini B, Manca N: Prevalence and distribution of single and 6. Schiffman M, Herrero R, Desalle R, Hildesheim A, Wacholder S, Rod- multiple HPV infections in cytologically abnormal cervical riguez AC, Bratti MC, Sherman ME, Morales J, Guillen D, Alfaro M, samples from Italian women. Virus Res 2007, 125(2):176-82. Hutchinson M, Wright TC, Solomon D, Chen Z, Schussler J, Castle 24. Andersson S, Mints M, Sällström J, Wilander E: The relative distri- PE, Burk RD: The carcinogenicity of human papillomavirus bution of oncogenic types of human papillomavirus in types reflects viral evolution. Virology 2005, 337(1):76-84. benign, pre-malignant and malignant cervical biopsies. A 7. Bosch FX, Burchell AN, Schiffman M, Giuliano AR, de Sanjose S, Bruni study with human papillomavirus deoxyribonucleic acid L, Tortolero-Luna G, Kjaer SK, Muñoz N: Epidemiology and nat- sequence analysis. Cancer Detect Prev 2005, 29(1):37-41. ural history of human papillomavirus infections and type- Page 7 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 25. De Vuyst H, Gichangi P, Estambale B, Njuguna E, Franceschi S, Tem- merman M: Human papillomavirus types in women with inva- sive cervical carcinoma by HIV status in Kenya. Int J Cancer 2008, 122(1):244-6. Publish with Bio Med Central and every scientist can read your work free of charge "BioMed Central will be the most significant development for disseminating the results of biomedical researc h in our lifetime." Sir Paul Nurse, Cancer Research UK Your research papers will be: available free of charge to the entire biomedical community peer reviewed and published immediately upon acceptance cited in PubMed and archived on PubMed Central yours — you keep the copyright BioMedcentral Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp Page 8 of 8 (page number not for citation purposes) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Infectious Agents and Cancer Springer Journals

Classification of weakly carcinogenic human papillomavirus types: addressing the limits of epidemiology at the borderline

Loading next page...
 
/lp/springer-journals/classification-of-weakly-carcinogenic-human-papillomavirus-types-djFzPq44EA

References (26)

Publisher
Springer Journals
Copyright
Copyright © 2009 by Schiffman et al; licensee BioMed Central Ltd.
Subject
Biomedicine; Cancer Research; Infectious Diseases; Oncology
eISSN
1750-9378
DOI
10.1186/1750-9378-4-8
pmid
19486508
Publisher site
See Article on Publisher Site

Abstract

Virtually all cases of cervical cancer are caused by persistent infections with a restricted set of human papillomaviruses (HPV). Some HPV types, like HPV16 and HPV18, are clear and powerful carcinogens. However, the categorization of the most weakly carcinogenic HPV types is extremely challenging. The decisions are important for screening test and vaccine development. This article describes for open discussion an approach recently taken by a World Health Organization International Agency for Research on Cancer (IARC) Monographs Working Group to re-assess the carcinogenicity of different HPV types. Background and rationale carcinogens [3]. The process is described here to promote As a group, human papillomaviruses (HPV) are proven needed debate on what an improved approach might be. human carcinogens. But there are >> 100 HPV genotypes and only a small fraction have any known carcinogenic IARC formally considers the carcinogenicity of exposures potential [1]. Therefore, moving from broad acceptance of to humans. Whereas human carcinogenicity might best be the carcinogenicity of persistent HPV infection to specific considered for some agents like HPV as a continuum of conclusions about individual genotypes requires consid- probabilities without a clear breaking point, IARC classi- eration of each type as an individual agent. Such a type- fies carcinogens categorically as carcinogenic (Group 1), by-type evaluation proves to be very difficult, and probably carcinogenic (Group 2a), possibly carcinogenic stretches epidemiology to its limits because of issues of (Group 2b), not classifiable (Group 3), or probably not confounding and misclassification detailed below. How- carcinogenic (Group 4). There has been very little experi- ever, as described in the accompanying article by Castle mental work on the carcinogenicity of HPV types except [2], it is important for the development of screening tests for HPV16 and HPV18; thus, epidemiologic evidence has and vaccines to judge each HPV type separately; thus, it is been unusually important. Epidemiologic evidence for worth considering how well one actually can decide the carcinogenicity of HPV was originally presented in whether a given HPV type is carcinogenic or not. This arti- IARC Monograph Volume 64 [4], and was extensively cle describes an approach taken by a Working Group of updated in IARC Monograph Volume 90 based on data the World Health Organization International Agency for available as of February 2005 [5]. In February 2009, Vol- Research on Cancer (IARC), when asked to re-assess for ume 100b updated the data once again and it is this latest official purposes which HPV types should be grouped as update addressed here. For the purposes of IARC, epide- Page 1 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 miologic evidence is categorized as sufficient, limited, clade containing a few genetically related species (Figure inadequate or suggesting lack of carcinogenicity. The epi- 1). Epidemiologic data do not support cervical carcino- demiologic data are combined with experimental evi- genicity for other species in the alpha genus or for other dence (in this case lacking for most HPV genotypes) to genera. To save considerable space presenting null evi- arrive at the final Groups 1–4. dence, this section will not include data related to HPV species alpha-1, -2, -3, -4, -8, -10 (other than HPV 6), -13, HPV carcinogenicity has been established most convinc- or 14/15. These species contain HPV types that cause skin ingly for cervical cancer, and this discussion will be lim- or genital warts, minor cytologic atypia, and often no ited to cervical carcinogenicity. To date, no HPV type has apparent disease. been proven to be carcinogenic only at sites other than the cervix. In Monograph 64 in 1995, HPV 16 (alpha-9) and HPV 18 (alpha-7) were classified as cervical carcinogens. HPV 31 and HPV 33 in alpha-9 were categorized as probably car- HPV evolution as the guiding principle of HPV behaviour (the high-risk clade) cinogenic [4]. In 2005, the group of cervical carcinogens HPV behaviour at the cervix is strongly correlated with (Group 1) was expanded to include the following 13 phylogenetic (i.e. evolutionary) categories [6]. All HPV types: alpha-5 genotype HPV 51; alpha-6 genotypes HPV genotypes that are known to be cervical carcinogens 56 and HPV 66; alpha-7 genotypes HPV 18, HPV 39, HPV belong to the alpha genus in an evolutionary branching or HPV6 HPV11 HPV13 HPV74 1, 8, 10, 13 HPV44 HPV55 HPV91 HPV7 Genital Warts HPV40 HPV32 HPV42 And HPV54 HPV52 HPV67 Commensal HPV33 HPV58 HPV16 HPV31 HPV35 HPV34 HPV73 11  5, 6, 7, 9, HPV59 HPV18 HPV45 11 Squamous HPV70 HPV39 HPV68 Evolutionary * HPV85 Cell Carcinoma HPV26 HPV69 Tree HPV51 HPV82 HPV30 (millions of HPV53 Adenocarcino HPV56 HPV66 years) HPV57 HPV2a ma HPV27 HPV71 HPV90 HPV61 HPV72 HPV62 2, 3, 4, 15 HPV81 HPV83 3 HPV89 HPV84 Commensal HPV86 HPV87 HPV28 Infections HPV3 HPV10 HPV29 HPV77 Phylogen Figure 1etic analysis of anogenital HPV types [6] Phylogenetic analysis of anogenital HPV types [6]. Branches determined by 100 bootstrap estimations using each of the methods in the following order: Bayesian credibility value, parsimony bootstrap percentage based on nucleotide alignment, and parsimony bootstrap percentage based on amino acid alignment. All definitely, probably, and possibly carcinogenic HPV types belong to one phylogenetic clade of the alpha genus. Page 2 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 45, and HPV 59; and alpha-9 genotypes HPV 16, HPV 31, tions are present in a large minority of women, although HPV 33, HPV 35, HPV 52, and HPV 58 [5]. they might not be transmitted from the same partner or at the same time. Given the existence of some very power- In the four years between Monograph 90 in 2005 and the fully carcinogenic types, notably HPV 16 and HPV 18, recent update, new evidence further supported that HPV determining which weaker and/or less common types are types in the high-risk clade of the alpha genus HPV type also carcinogenic becomes (for the epidemiologist) an cause virtually all cases of cervical cancer worldwide [7,8]. issue of confounding. None of the traditional approaches In case-control studies, the odds ratios associating cervical to control of confounding are entirely successful. Because cancer and its immediate precursor, CIN3, with HPV DNA HPV 16 causes approximately 50% of cases of cervical positivity for these types in aggregate has consistently cancer, logistic regression and similar approaches will par- exceeded 50. It is persistent infections that are associated simoniously attribute cases associated with both HPV16 with extremely high absolute risk of CIN3 and cancer. In and a less important type to HPV16. HPV 18 is the second cohort studies, women who test negative to this group of most important cervical carcinogen, responsible for HPV types are at extremely low subsequent risk of CIN3, approximately 15–20% of cervical cancer of all histologic cancer, and cancer death for more than 10 years [9-11]. types combined (and a higher fraction of adenocarcino- mas). If a type occurs with either HPV 16 or HPV 18, its Because persistent infection with a restricted group of association with cervical cancer might be confounded by HPV types is a nearly necessary cause of cervical cancer, a either of these powerful carcinogens. For types causing reconsideration of HPV and cervical carcinogenicity based only a very small fraction of cervical cancer, confounding on the new data must decide whether any additional types by any of the more important types is possible. within the high-risk clade are also carcinogenic and whether any types in that clade that were previously cate- Dealing with confounding by exclusion, i.e. examining gorized as carcinogenic should be downgraded. The types the possibility of carcinogenicity of a more minor type in the high-risk clade are listed in Table 1. among cancer specimens that do not contain a more important type, becomes a problem of misclassification. Conceptual issues in deciding which of the types This main epidemiologic criterion used for classification in the high-risk clade are carcinogenic of an HPV type as a carcinogen, finding the HPV genotype From a virologic perspective, the definitive proof of carci- as a single infection in a cervical scrape or biopsy speci- nogenicity of an HPV type is finding transcriptionally men in a woman with cancer, might sometimes be too lax active HPV in a tumour. HPV is not a "hit and run" carcin- and prone to error. Colposcopic biopsies and cytology ogen, and transcriptional activity is needed for mainte- specimens can be misdirected and fail to obtain the criti- nance of the cancer phenotype. In cervical cancer cell cal cells, while contamination of scrapes and biopsies lines, blocking transcriptional activity by antisense RNA from lower-grade lesions that often surround cancers can leads to apoptosis. This level of evidence is simply lacking detect types other than the causal one. Studies relying on for virtually all HPV types. The vast body of evidence testing of microdissected cervical malignancies will relates simply to finding HPV DNA at the same time of address these issues, but large-scale highly accurate data cervical neoplasia. are not yet available. However, relying on testing of scrapes or biopsies, by Difficulty with control selection adds another level of DNA testing, leads to difficulties. The alpha HPV types complexity in assessing carcinogenicity. Cervical cancer share a common route of transmission and multiple infec- typically follows age infection by decades. HPV transmit- Table 1: HPV Types that Were Considered in Monograph 100B Alpha Species Types Categorized as Definite Carcinogens in Monograph 90 Other Types in Species 551 26 69 82 656 66 30 53 7 18 45 39 59 68 70 85 97 9 16 31 33 35 52 58 67 11 34 73 Page 3 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 ted at young ages usually become undetectable by DNA or ous lesions (CIN3 and the more equivocal CIN2 which RNA assays and no sensitive serologic assay exists to meas- occur in approximately 1% of screened women), which ure HPV exposure. Consequently, odds ratios based on a often used as ethical surrogate endpoints in prospective comparison of HPV prevalence at the time of case diagno- studies and clinical trials; there are now sufficient data for sis to age-matched HPV point prevalence in controls do invasive cancers and it appears that HPV types have differ- not estimate true relative risks. ent potential to progress from CIN2/3 to invasive cervical cancer [14]. The background frequency of cervical HPV There is not much type-specific prospective data on the infection in the general female population was estimated carcinogenicity of individual HPV genotypes. The availa- from a large meta-analysis of HPV genotypes found in ble studies have categorically shown the unique carcino- women with normal cytology [15], as shown in Table 2. genicity of HPV 16 and, to a lesser extent, HPV 18 [9,12]. Although women included in meta-analyses of cervical Khan et al. observed a risk for the remaining women pos- cancer and normal cytology differed by age, region and itive by hc2 after excluding those positive for HPV 16 or HPV testing protocols, it was considered the best method HPV 18 (including an unknown mix of the types 31, 33, to identify a reasonable threshold of confounding and 35, 39, 45, 51, 52, 56, 58, 59 and 68) of only 3.0% (1.9– misclassification for each type. Ancillary analyses examin- 4.2) compared with 0.8% (0.6–1.1) among women who ing the issue of most important types by region were also were HPV negative at baseline. Thus, there is not strong scrutinised. and convincing long-term prospective evidence for indi- Uniqueness of HPV16 and HPV18 vidual HPV types other than HPV16 and HPV18. Comparing the prevalences in women with normal cytol- Finally, the accuracy of detection of HPV genotypes differs ogy to the prevalences for cancers compiled by Smith et al. between the major PCR-based systems used to generate (2007) (Table 2), obvious "case-control" differences can most of the data [13]. Each of the systems shows differen- be seen. The most clearly carcinogenic genotypes, HPV 16 tial sensitivity, and some have exhibited cross-reactivity of and HPV 18 in particular, are more common among can- detection. These detection issues are not critical for evalu- cers and cytologically normal women (and even low- ating the most important HPV types, but make it difficult grade lesions [16]). HPV 18 is especially common in ade- to clarify the role of the most weakly carcinogenic and nocarcinomas [17], as are other members of the alpha 7 least common ones. clade of which HPV18 is a member. The large amount of data regarding HPV 16 and HPV 18 was thought to pro- With these caveats, the cervical carcinogenicity of the HPV vide ample epidemiologic evidence of carcinogenicity types listed above varies in strength in a continuum with- (leading to an overall classification of Group 1). out clear breakpoint, from extremely strong (i.e. HPV 16 and, to a lesser degree, HPV 18) to weak, but still probably The 8 most important carcinogenic HPV types carcinogenic in rare instances (e.g. HPV 68, see below). Including HPV 16 and HPV 18, eight HPV types (alpha-7 Evaluators taking one extreme position could claim that types HPV 18 and 45, alpha-9 types 16, HPV 31, 33, 35, there is reasonable evidence for carcinogenicity of virtu- 52 and 58) are the most common types found in cancers ally all the types in the species listed above, extending fur- in both the IARC meta-analysis [8] and the ICO study [7], ther the list established in Monograph 90. Strict in all regions of the world providing data. Though very interpreters of causal criteria could argue for a return to a often found in non-invasive lesions, these types are all much more limited list. But based on current evidence, no much more common in cancer case specimens than in cut-point between sufficient, limited, and inadequate epi- controls, providing sufficient epidemiologic evidence of demiologic evidence is entirely defensible. carcinogenicity (Group 1). The borderline carcinogens: using HPV 6 as an estimator of Approach taken by the IARC working group Data Sources residual confounding The IARC Working Group chose the following pragmatic To move beyond the most evidently carcinogenic eight approach to creating an imperfect cut-point between suf- HPV types, the Working Group chose an estimator of the ficient, limited, and inadequate epidemiologic evidence percentage of cancers that might contain HPV DNA by for cervical carcinogenicity: Only types in the high-risk accumulated and unknown measurement errors alone. clade were considered because data seemed grossly inade- The group made use of HPV 6 for this estimation. Specifi- quate for all others. The most recent accumulated data on cally, HPV 6, the important and common cause of benign type-specific HPV prevalence in cervical cancers were condyloma acuminata (external genital warts) was con- drawn from a very large single project [7] and from meta- sidered to be a low-risk type, not classified as a cervical analyses performed by IARC ([8] updated by G. Clifford). carcinogen, which only uncommonly is detected in cervi- Excluded from consideration were high-grade precancer- cal cancer specimens. [Of note, it remains possible that Page 4 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 Table 2: Meta-analyses of type-specific HPV DNA prevalence in invasive cervical cancer [15] and women with normal cytology [14,17] Invasive cervical cancer Normal N tested % pos 95% CI N tested % pos 95% CI HPV16 14595 54.4 53.6–55.2 76385 2.6 2.5–2.8 HPV18 14387 15.9 15.3–16.5 76385 0.9 0.8–1.0 HPV33 13827 4.3 4.0–4.6 74141 0.5 0.4–0.5 HPV45 9843 3.7 3.3–4.1 65806 0.4 0.4–0.4 HPV31 11960 3.5 3.2–3.9 74076 0.6 0.6–0.7 HPV58 10157 3.3 2.9–3.6 72877 0.9 0.8–1.0 HPV52 9509 2.5 2.2–2.8 69030 0.9 0.8–1.0 HPV35 9507 1.7 1.5–2.0 74084 0.4 0.3–0.4 HPV59 6972 1.0 0.8–1.3 64901 0.3 0.2–0.3 HPV51 7339 0.7 0.5–0.9 67139 0.6 0.6–0.7 HPV56 7427 0.7 0.5–0.9 68121 0.5 0.5–0.6 HPV39 7078 0.6 0.5–0.9 64521 0.4 0.3–0.4 HPV68 6723 0.5 0.3–0.7 63210 0.3 0.2–0.3 HPV73 5837 0.5 0.3–0.7 44063 0.1 0.1–0.1 HPV66 6664 0.3 0.2–0.5 59774 0.4 0.3–0.4 HPV70 5159 0.2 0.1–0.4 35014 0.3 0.3–0.3 HPV82 5352 0.1 0.1–0.3 42536 0.1 0.0–0.1 HPV6 9911 0.5 0.4–0.7 58370 0.3 0.2–0.3 HPV53 not reported 44,058 0.4 0.4–0.4 HPV26 not reported 44,098 0.0 0.0–0.1 HPV85 not studied in ICC meta-analysis 9,622 0.1 0.1 – 0.1 HPV67 not studied in ICC meta-analysis 18,041 0.0 0.0 – 0.0 HPV34 not studied in ICC meta-analysis 42,588 0.0 0.0 – 0.1 HPV30 not studied in ICC meta-analysis 8,773 0.0 0.0 – 0.1 HPV97 not studied in ICC meta-analysis not studied in normal cytology meta-analysis HPV 6, and other "low risk types" can cause cancer in Table 3: Meta-analysis of type-specific HPV DNA prevalence in extremely rare virus-host circumstances.] When detected, invasive cervical cancer [15], updated with 63 newly published even without detection of a more likely causal type, the studies by Clifford (IARC) Working Group judged that misclassification of some N tested N pos % pos 95% CI kind was a more likely explanation than causality. As given in Table 3, the best IARC estimate of detection of hpv39 13370 172 1.29 1.10 – 1.48 HPV 6 in cervical cancers, not necessarily as a single infec- hpv59 13471 172 1.28 1.09 – 1.47 tion, was judged to be 0.45%, 95% CI 0.35 – 0.56, based hpv51 13057 151 1.16 0.97 – 1.34 on 14,912 cases of cancer ([8] updated by G. Clifford for hpv56 13247 103 0.78 0.63 – 0.93 the Working Group). The Working Group pragmatically hpv68 11982 73 0.61 0.47 – 0.75 made the following rule: An individual HPV type in the hpv73 9939 48 0.48 0.35 – 0.62 high-risk alpha clade (i.e. one with an elevated prior prob- ability of being carcinogenic due to analogy to closely hpv6 14912 68 0.45 0.35 – 0.56 related viral types in the same or closely-related species) was considered to have sufficient epidemiologic evidence hpv53 8140 34 0.42 0.28 – 0.56 of carcinogenicity if its prevalence in cancers was 1) signif- hpv66 12118 47 0.39 0.28 – 0.50 icantly greater than that of HPV 6, and 2) significantly hpv70 10503 35 0.33 0.22 – 0.44 hpv82 9265 25 0.27 0.16 – 0.38 enriched in comparison to the background estimate for hpv26 6111 8 0.13 0.04 – 0.22 the general population, i.e. women with normal cytology. HPV types 30, 34, 67, 85 and 97 were not studied in the meta-analysis By this logic, four more types were categorized as having dataset. sufficient epidemiologic evidence leading to their classifi- Page 5 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 cation as definite (Group 1) carcinogens: alpha-5 type 51 b) HPV 18 is also important, particularly for adenocar- (1.16% of cervical cancer), alpha-6 type 56 (0.78%), and cinoma. alpha-7 HPV types 39 (1.29%) and 59 (1.28%). c) Six additional types in alpha-7 (HPV 45) and alpha- Classification of other types in the high-risk clade 9 (HPV 31, HPV 33, HPV 35, HPV 52, HPV 58) com- The remaining types in the high-risk alpha clade were con- prise the remainder of the eight types that are the most sidered, as a group, to have limited evidence to support important globally, with some regional variation in carcinogenicity. If phylogeny can be taken to predict the etiologic fractions of cancers due to each type. behaviour, it is possible that most of these types can very rarely cause cancer. Indeed, many of the types have been d) There are small and less certain, incremental etio- detected, albeit uncommonly (not significantly greater logic contributions of another group of carcinogenic than HPV 6), in cancers. There are not enough data, even types from alpha-5 (HPV 51), alpha-6 (HPV 56), and after testing of many thousands of specimens, to be sure alpha-7 (HPV 39 and HPV 59). Each causes a few per- which ones are carcinogenic or not. Furthermore, some of cent at most of cervical cancer cases worldwide. There these types, namely 30, 34, 67, 85 and 97 have not yet has not been any large-scale study of regional variabil- been studied in the IARC cervical cancer meta-analysis ity for these uncommon types approach. Nevertheless, within this group, there are two types, alpha-7 type HPV 68 and alpha-11 type HPV 73, for e) There is an unresolved dividing line between the which the data are slightly stronger than for the others HPV types with the weakest evidence judged to be suf- despite methodologic challenges. One of the major PCR- ficient and those judged to have highly suggestive yet based testing methods (SPF10) cannot distinguish these limited data (e.g. alpha-7 HPV 68 is categorized as two types because their amplicons using those primers are probably carcinogenic due to experimental evidence identical [18]. Neither of these two types is optimally while alpha-11 HPV 73 is possible). detected by MY09-MY11 dot blot [19]. Nonetheless, the data supporting the carcinogenicity of HPV 68 and HPV f) The expanded data for HPV66 were re-evaluated and 73 are very suggestive although not sufficient. Ultimately, the evidence was judged to be very limited now that the existence of a cell line whose immortalization is sus- more cases have been studied showing that it is very tained by a subtype of HPV68 (ME180), led the Working rarely found in cancers despite being relatively com- Group to classify HPV 68 as a probable carcinogen mon in the newly collated data on women with nor- (Group 2A). The rest were called possible (Group 2B). mal cytology. HPV 53, also in alpha species 6, shows the same pattern of relative common population prev- Re-classification of HPV 66 alence with extremely rare cases of occurrence alone in Overall, the Working Group approach led to the re-classi- cancer. The Working Group noted that for these types fication from Monograph 90 of HPV 66 to possible in particular, there could be harm to public health if (Group 2B), although the epidemiologic evidence of car- the types are viewed with excessive concern; including cinogenicity was previously judged sufficient (Group 1). these types as carcinogenic in screening assays would In the assembly of much more testing data from cancer decrease the specificity and positive predictive value of cases, HPV 66 has been found so rarely that its percentage the assays with virtually no gain in sensitivity and neg- of detection is less than the relative percentage of detec- ative predictive value [20]. tion among the general population. In the Working Group review of each individual article, HPV 66 was g) There are several types within the high-risk clade found as a single infection in cancers with extreme rarity, that have extremely sparse or no evidence of carcino- well below the threshold of possible confounding and genicity. For some types there are anecdotal but very misclassification. interesting cases that merit pursuit of additional carci- nogenic types. There have been potentially underap- Summary preciated reports of alpha-9 type HPV 70 found as There is no perfect way to categorize a continuum of car- single infections in cancer, but the supportive data are cinogenic potential. The Working Group arrived at the fol- sparse. There are only a few reports of HPV 67 in can- lowing conclusions, with a healthy scepticism concerning cer [21-24], which is intriguing because this is the only the process: known type in the alpha-9 species that is not catego- rized as carcinogenic. For a few types in the high-risk a) Persistent HPV 16 infection is a uniquely powerful clade, no reports of invasive cancers with single-type human carcinogen. infections were found, but isolated reports might exist. Page 6 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 specific implications in cervical neoplasia. Vaccine 2008, h) The possible role of immunosuppression in HPV 26(Suppl 10):K1-16. carcinogenicity was not emphasized in the Working 8. Smith JS, Lindsay L, Hoots B, Keys J, Franceschi S, Winer R, Clifford Group discussions. To cause cancer, an HPV infection GM: Human papillomavirus type distribution in invasive cer- vical cancer and high-grade cervical lesions: a meta-analysis must persist and it is possible that some HPV types are update. Int J Cancer 2007, 121(3):621-32. only weak carcinogens because they persist poorly. For 9. Khan MJ, Castle PE, Lorincz AT, Wacholder S, Sherman M, Scott DR, Rush BB, Glass AG, Schiffman M: The elevated 10-year risk of example, the carcinogenicity of alpha-5 type HPV 26 cervical precancer and cancer in women with human papil- has been supported by a recent report of multiple peri- lomavirus (HPV) type 16, or 18, and the possible utility of ungual cancers in an immunosuppressed individual, type-specific HPV testing in clinical practice. J Natl Cancer Inst 2005, 97(14):1072-9. containing high viral loads and active transcription of 10. Sankaranarayanan R, Nene BM, Shastri SS, Jayant K, Muwonge R, HPV 26 alone [25]. HPV 26 is an uncommon type; Budukh AM, Hingmire S, Malvi SG, Thorat R, Kothari A, Chinoy R, perhaps the immunosuppression in this individual Kelkar R, Kane S, Desai S, Keskar VR, Rajeshwarkar R, Panse N, Din- shaw KA: HPV screening for cervical cancer in rural India. N was a significant contributor to carcinogenesis. As one Engl J Med 2009, 360(14):1385-94. avenue of research, there should be more studies of 11. Schiffman M, Wacholder S: From India to the world – a better way to prevent cervical cancer. N Engl J Med 2009, whether HPV types in invasive cancers in HIV-infected 360(14):1453-5. individuals differ in type from the types found in 12. Wheeler CM, Hunt WC, Schiffman M, Castle PE, Atypical Squamous immunocompetent individuals [25]. Cells of Undetermined Significance/Low-Grade Squamous Intraepi- thelial Lesions Triage Study Group: Human papillomavirus geno- types and the cumulative 2-year risk of cervical precancer. J Conclusion Infect Dis 2006, 194(9):1291-9. 13. Gravitt PE, Coutlée F, Iftner T, Sellors JW, Quint WG, Wheeler CM: When epidemiology serves as the science informing pub- New technologies in cervical cancer screening. Vaccine 2008, lic health policy, its limitations must be acknowledged. 26(Suppl 10):K42-52. Weak causal associations with one HPV type are extremely 14. Clifford GM, Smith JS, Aguado T, Franceschi S: Comparison of HPV type distribution in high-grade cervical lesions and cervical hard to prove in the presence of powerful confounding by cancer: a meta-analysis. Br J Cancer 2003, 89(1):101-5. strong carcinogens like HPV16. A coming generation of 15. de Sanjosé S, Diaz M, Castellsagué X, Clifford G, Bruni L, Muñoz N, intensive molecular studies of microdissected cancers, Bosch FX: Worldwide prevalence and genotype distribution of cervical human papillomavirus DNA in women with nor- with examination of transcriptional activity, may clarify mal cytology: a meta-analysis. Lancet Infect Dis 2007, 7(7):453-9. the borderline between carcinogenic and non-carcino- 16. Clifford GM, Rana RK, Franceschi S, Smith JS, Gough G, Pimenta JM: Human papillomavirus genotype distribution in low-grade genic HPV types. For public health purposes, the main cervical lesions: comparison by geographic region and with agents that merit consideration in screening tests and vac- cervical cancer. Cancer Epidemiol Biomarkers Prev 2005, cines have already been identified. 14(5):1157-64. 17. Clifford G, Franceschi S: Members of the human papillomavirus type 18, family (alpha-7 species) share a common association Acknowledgements with adenocarcinoma of the cervix. Int J Cancer 2008, A special thank for comments and contributions from the IARC Working 122(7):1684-5. 18. Molijn A, Kleter B, Quint W, van Doorn LJ: Molecular diagnosis of Group and the IARC Secretariat (in particular to Veronique Bouvard, human papillomavirus (HPV) infections. J Clin Virol 2005, Responsible Officer for Volume 100B and Vincent Cogliano, Head of the 32(Suppl 1):S43-51. Monograph Program). Dr FM Buonaguro, has been recipient of Merck and 19. Handisurya A, Rieger A, Bankier A, Koller A, Salat A, Stingl G, Kirn- GSK support for Conference organization, and is a 2008 Member of the bauer R: Human papillomavirus type 26, infection causing multiple invasive squamous cell carcinomas of the fingernails GSK-sponsored PanEuropean HPV Immunology Group. in an AIDS patient under highly active antiretroviral ther- apy. Br J Dermatol 2007, 157(4):788-94. Epub 2007, Jul 19. Erratum References in: Br J Dermatol. 2007, Dec;157(6):1304. 1. Schiffman M, Castle PE, Jeronimo J, Rodriguez AC, Wacholder S: 20. Schiffman M, Khan MJ, Solomon D, Herrero R, Wacholder S, Hild- Human papillomavirus and cervical cancer. Lancet 2007, esheim A, Rodriguez AC, Bratti MC, Wheeler CM, Burk RD, PEG 370(9590):890-907. Group; ALTS Group: A study of the impact of adding HPV 2. Castle P: The Evolving Definition of Carcinogenic Human types to cervical cancer screening and triage tests. J Natl Can- Papillomavirus Infectious. Agents and Cancer 2009 in press. cer Inst 2005, 97(2):147-50. 3. Bouvard V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, Ben- 21. Gudleviciene Z, Didziapetriene J, Ramael M, Uleckiene S, Valuckas KP: brahim-Tallaa L, Guha N, Freeman C, Galichet L, Cogliano V, WHO Human papillomavirus and p53 polymorphism in Lithuanian International Agency for Research on Cancer Monograph Working cervical cancer patients. Gynecol Oncol 2006, 102(3):530-3. Group: A review of human carcinogens – Part B: biological 22. Wentzensen N, Schiffman M, Dunn ST, Zuna RE, Walker J, Allen RA, agents. Lancet Oncol 2009, 10(4):321-2. Zhang R, Sherman ME, Wacholder S, Jeronimo J, Gold MA, Wang SS: 4. IARC Monographs: Human Papillomaviruses. In IARC Monographs Grading the severity of cervical neoplasia based on com- on the Evaluation of Carcinogenic Risks to Human Volume 64. IARC, Lyon, bined histopathology, cytopathology, and HPV genotype dis- France . tribution among 1,700 women referred to colposcopy in 5. Cogliano V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, Oklahoma. Int J Cancer 2009, 124(4):964-9. WHO International Agency for Research on Cancer: Carcinogenic- 23. Gargiulo F, De Francesco MA, Schreiber C, Ciravolo G, Salinaro F, ity of human papillomaviruses. Lancet Oncol 2005, 6(4):204. Valloncini B, Manca N: Prevalence and distribution of single and 6. Schiffman M, Herrero R, Desalle R, Hildesheim A, Wacholder S, Rod- multiple HPV infections in cytologically abnormal cervical riguez AC, Bratti MC, Sherman ME, Morales J, Guillen D, Alfaro M, samples from Italian women. Virus Res 2007, 125(2):176-82. Hutchinson M, Wright TC, Solomon D, Chen Z, Schussler J, Castle 24. Andersson S, Mints M, Sällström J, Wilander E: The relative distri- PE, Burk RD: The carcinogenicity of human papillomavirus bution of oncogenic types of human papillomavirus in types reflects viral evolution. Virology 2005, 337(1):76-84. benign, pre-malignant and malignant cervical biopsies. A 7. Bosch FX, Burchell AN, Schiffman M, Giuliano AR, de Sanjose S, Bruni study with human papillomavirus deoxyribonucleic acid L, Tortolero-Luna G, Kjaer SK, Muñoz N: Epidemiology and nat- sequence analysis. Cancer Detect Prev 2005, 29(1):37-41. ural history of human papillomavirus infections and type- Page 7 of 8 (page number not for citation purposes) Infectious Agents and Cancer 2009, 4:8 http://www.infectagentscancer.com/content/4/1/8 25. De Vuyst H, Gichangi P, Estambale B, Njuguna E, Franceschi S, Tem- merman M: Human papillomavirus types in women with inva- sive cervical carcinoma by HIV status in Kenya. Int J Cancer 2008, 122(1):244-6. Publish with Bio Med Central and every scientist can read your work free of charge "BioMed Central will be the most significant development for disseminating the results of biomedical researc h in our lifetime." Sir Paul Nurse, Cancer Research UK Your research papers will be: available free of charge to the entire biomedical community peer reviewed and published immediately upon acceptance cited in PubMed and archived on PubMed Central yours — you keep the copyright BioMedcentral Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp Page 8 of 8 (page number not for citation purposes)

Journal

Infectious Agents and CancerSpringer Journals

Published: Jun 1, 2009

There are no references for this article.