Int. J. Cancer: 100, 698 –701 (2002) © 2002 Wiley-Liss, Inc.
Publication of the International Union Against Cancer
RISK OF INVASIVE CERVICAL CANCER ASSOCIATED WITH POLYMORPHIC HLA DR/DQ HAPLOTYPES Mehran GHADERI1, Keng-Ling WALLIN2, Fredrik WIKLUND3, Liene Nikitina ZAKE1, Go¨ran HALLMANS4, Per LENNER3, Joakim DILLNER2 and Carani B. SANJEEVI1* 1 Department of Molecular Medicine, Karolinska Institutet, Karolinska Hospital, Stockholm, Sweden 2 Microbiology and Tumor Biology Center, Karolinska Institutet, Stockholm, Sweden 3 Oncology Center, Department of Oncology, UmeåUniversity, Umeå, Sweden 4 Nutritional Research, Department of Public Health and Clinical Medicine, Umeå University, Umeå, Sweden The genes encoding human leukocyte antigens (HLA) have shown to be associated with cervical neoplasia. To obtain reliable data on HLA associations with cervical tumors, the study should be performed within a strictly defined cohort. To investigate the population attributable risk of cervical cancer associated with the HLA class II haplotypes DR15 and DQ6 (DQA1*0102 and DQB1*0602), we performed a nested case-control study of 85 women who developed invasive cervical cancer and 120 healthy women from a population-based cohort of Swedish women. The relative risks of cervical cancer among DR15 and DQ6-positive women were 3.73 [confidence interval (CI): 1.8 –7.4] and 4.33 (CI: 2.1– 8.5), corresponding to population attributable proportions of 27.9% and 30.8%, respectively. A susceptibility locus in the HLA class II region is involved in a substantial fraction of the etiology of cervical cancer. © 2002 Wiley-Liss, Inc. Key words: cervical cancer; CIN; HLA DR-DQ; HPV; MHC
at 4-year intervals. The attendance rate was ⬎80 percent during the follow-up. All cytological diagnoses in this county, both on Papsmears taken in the organized program and on spontaneously taken smears, are made and the smears stored at the Cytology Laboratory, Umeå University Hospital since 1969. Similarly, all histological specimens from the Va¨sterbotten population have been examined and stored at the Pathology laboratory of the same hospital. All cervical cancer patients in this population are treated at the Umeå University Hospital, where data is recorded on clinical stage, histopathology grade, treatment and survival. Study design Eligible women were defined as Va¨sterbotten resident women who had taken at least one cytologically normal cervical smear and had had no operative treatment of the cervix. Linkage between the cytology registry and the Swedish Cancer Registry for the period 1969 to 1995 identified 133 eligible women with invasive cervical cancer (of all stages) diagnosed after the sampling date of a normal smear. Controls were women in the study base who did not develop cervical cancer before the time-point of diagnosis of a corresponding case, matched for age (same calendar year of birth), time of sampling of a normal smear compared to the prediagnostic normal smear of the case and time of sampling of a normal smear taken after diagnosis of cancer in the corresponding case. Since genetic markers are not age or calendar time-dependent, the statistical analyses were in our study performed with pooled data without individual matching. Four case women were excluded because of incorrect entry in the registry. Eleven case women were excluded because of noninvasive cervical neoplasia, leaving 118 cases of invasive cervical
Susceptibility to viral infections is partly controlled by the human leukocyte antigens (HLA). Certain HLA specificities are associated with efficient virus elimination while others are associated with a tendency towards long lasting, persistent virus infections.1–5 Infection with oncogenic human papillomaviruses (HPV), particularly type 16 and 18, is the major cause of cervical intraepithelial neoplasia (CIN) and cervical cancer6 – 8 Although over 90% of cervical cancers harbor HPV genomes,9 only a small fraction of women infected with high-risk papillomaviruses develop cancer, suggesting that other environmental and/or genetic factors contribute to cervical carcinogenesis. Both studies of monozygotic and dizygotic twins10 and family studies11 have found evidence of a genetic component in cervical cancer. Several studies have reported HLA as one genetic factor that is associated with CIN and cervical cancer.12–14 However, a link between HLA polymorphisms and cervical neoplasia has not been consistently reported.15 Varying results could be due to the fact that many studies did not use epidemiologically defined study designs. For valid results, acceptable response rates are required. Two previous studies, a population-based cohort study and a population-based case-control study with high attendance rate, both identified HLA-DQ6 and HLA-DR15 as a susceptibility marker for high grade CIN in Scandinavia.16,17 The aim of our study was to test the hypothesis that HLADR15DQ6 association demonstrated in our earlier CIN study is also true for the invasive cervical cancer study. We performed a populationbased cohort study of the risk of invasive cervical cancer in relation to these HLA haplotypes and to HPV infection.
Grant sponsor: Karolinska Institutet, Stockholm, Sweden, Swedish Medical Research Council; Grant number: k2001-16x-12532-04B; Grant sponsor: Swedish Cancer Society; Grant numbers: 3819, 4330.
METHODS
Received 2 July 2001; Revised 6 February, 3 May 2002; Accepted 3 May 2002
Study base In the Va¨sterbotten County in Northern Sweden [population, 260,472; women, 130,651 (1995)], a population-based invitational cervical screening program started in 1969 with invitations issued
DOI 10.1002/ijc.10551 Published online 3 July 2002 in Wiley InterScience (www.interscience. wiley.com).
Dr. Wallin’s current address is: Cancer Centre Karolinska, CCK, R8:04, Karolinska Hospital, S-171 76 Stockholm, Sweden. E-mail:
[email protected] Dr. Dillner’s current address is: Department of Medical Microbiology, MAS University Hospital, Lund University, Malm¨o, Sweden. E-mail:
[email protected] *Correspondence to: Karolinska Institutet, Karolinska Hospital, CMM, L8;03, S-17176 Stockholm, Sweden. Fax: ⫹46-8-51776179. E-mail:
[email protected]
699
HLA-DQ6 AND INVASIVE CERVICAL CANCER RISK
cancer. For 12 cases the histopathological specimens were missing. The tissue blocks from 2 cases were inadequate for PCR analysis, leaving 104 cases (85 squamous cell carcinomas and 19 adenocarcinomas) with serial samples available. In case of several prediagnostic smears with normal cytology, the slide taken closest to the diagnosis of the cancer was retrieved. In total 85 cases and 120 control women were included in our study of HLA haplotyping. The mean age of case women was 44.2 years (range, 19.1– 74.1) and of the control women 44.1 years (range, 19.5–74.4). Mean length of follow up was 5.6 years for case women (range, 0.5 months–26.2 years) and 6.3 years for control women (range, 0.4 –25.4 years). Ethical approval was obtained from the Institutional Review Board (IRB) of the Umeå University (95-240 and 98-12). DNA extraction DNA was extracted from archival smears and biopsies as previously described.18,19 Knives were changed and an empty paraffin block sectioned in-between each biopsy to prevent cross-contamination. All samples were tested for DNA integrity by PCR using human ribosomal gene S14 primers18,20 that give 150 base pair amplimers. Samples positive for S14 PCR but negative for HPV DNA were alcohol precipitated and the PCR was repeated. PCR-based HLA class II genotyping The polymorphic second exon of the DQA1, DQB1 and DRB1 genes was amplified in a programmable thermal controller with heated lead (PTC-100; MJ Research, Inc., Cambridge, MA) using primers specific for each locus as previously described.21 The 25 l PCR reaction was run in thin-wall PCR tubes containing 100 ng of genomic DNA, 0.5 M of each primer, 200 M of dNTP’s, 1⫻ AmpliTaq Gold PCR buffer (standard MgCl2 concentration) and 2.5 units of the AmpliTaq Gold DNA Polymerase (Applied Biosystems, PE Corporation, Foster City, CA). The PCR mixture was preincubated for 8 min at 95°C to activate the AmpliTaq Gold DNA Polymerase and to denaturate the genomic DNA. The polymorphic second exon was amplified in a 30 cycle PCR reaction with denaturation for 1 min at 96°C and annealing with each specific primer at 62°C for DQA1 and 55°C for DQB1 and DRB1 and primer extension was carried out at 72°C. The PCR products were run in 2% agarose gels and visualised with ethidium bromide (10 g/ml). The amplified products were manually dotted onto nylon membranes (Amersham, Arlington, USA), under denaturing conditions. The membranes were hybridised with sequence specific oligonucleotides (SSOs), where the 3⬘ end was labelled with (␣-32p ) dCTP (NEN Research Products, Boston, MA) and washed under stringent conditions before exposure to X-ray film at ⫺70°C.22 The membranes were stripped of the labelled probe under alkaline conditions and reused for probing with other oligonucleotides. PCR analysis (HPV) The HPV consensus primers MY09 and MY1123 and GP5⫹ and GP6⫹24 were used in nested single-tube PCR assay.25,26 A nonnested PCR was also performed with the GP5⫹/GP6⫹ primers that amplify 150 bp products, a similar size as generated by the S14 PCR. Both PCR systems utilized 1.5 mM MgCl2 and 0.4% BSA in a 50 l buffer mix. The amplimers were run on 2% agarose gels and stained with 10 g/ml EtBr at a concentration of 1 l per 10 ml of agarose gel (Stock 10g/ml). The nested PCR was able to detect one copy of HPV16 L1 plasmid when we mixed with DNA extracted from HPV negative smear.26 The limit of detection for SiHa DNA in each PCR run was consistently 1.0 fg or less. For direct automated sequencing of the PCR products, the amplimers were purified with the QIA quick PCR Purification Kit (QIAGEN, Hilden, Germany). Another round of PCR utilizing the GP5⫹ and GP6⫹ fragment as template with 3.0 pmol of GP6⫹ primer was performed with the recommended amount of Big Dye Terminators (ABI Prism™ Big Dye™ Termi-
nator Cycle Sequencing Ready Reaction Kit, (PE Biosystems, Foster City, CA). Sequencing was performed using an ABI 310 sequencer (PE Biosystems). The PCR analyses were performed blindly (coded). The samples were arranged in a manner ensuring that cases and controls were analyzed in the same analytic runs to avoid the risk of assay performance variation (the most obvious risk being that of contamination) systematically biasing relationships of HLA or HPV status to case-control status. Statistical calculations Odds Ratios (OR) and their confidence intervals were estimated by logistic regression. A 2-sided p-value of ⬍0.05 was considered to be significant. In case of small values, Fisher’s exact test was performed to estimate the statistical significance using the approximation of Woolf.27,28 p values were corrected (pc) for the number of comparisons made, which was n ⫽ 9 for DQA1, n ⫽ 14 for DQB1, n ⫽ 14 for DRB1 and n ⫽ 28 for DQA1-DQB1 haplotype. RESULTS
Evaluation of the primary hypothesis The 2 HLA haplotypes previously found to be associated with CIN (DR15 and DQ6) were also found to be clearly associated with invasive carcinoma of the cervix DR15 [Odds Ratio (OR) ⫽ 3.7, CI ⫽ 1.8 –7.4, p ⫽ 0.0002, pc⬍0.05) and DQ6 (DQA1*0102 - DQB1*0602) (OR ⫽ 4.3, CI ⫽ 2.1– 8.5, p ⬍ 0.0001, pc⬍0.05). Both alleles in the DQ6 haplotype, DQA1*0102 and DQB1*0602, were associated with invasive cervical cancer (OR 4.9, CI ⫽ 2.6 –9.1, p ⬍ 0.0001 and OR ⫽ 4.3, CI ⫽ 2.1– 8.5, p ⬍ 0.0001, respectively, Tables 1– 4). The estimates of the population attributable proportion (PAP) for cervical cancer for HLA-DQ6 and HLA-DR15 were 30.8% and 27.9%, respectively (data not shown). Correlation of HLA alleles and HPV infection There was a tendency for over representation of HLA-DQ6 allele frequency (DQA1*0102 - DQB1*0602) among the patients with persistent6 HPV16 positivity (data not shown) both for the prediagnostic smear and biopsy compared to patients who were HLA-DQ6 negative (p value not significant). Other alleles Two alleles that were not part of the predefined study hypothesis, the DRB1 allele 07 and the DQB1 allele 0201, exhibited marginally significant negative associations with cervical cancer [OR ⫽ 0.4, CI ⫽ 0.2– 0.9, p ⫽ 0.0476, corrected p-value (pc) ⫽ 0.5712] and (OR ⫽ 0.5, CI ⫽ 02–1.0, p ⫽ 0.0628, pc ⫽ 0.942), respectively (Tables 2,3). DISCUSSION
The aim of our study was to investigate the association of HLA class II alleles (mainly DR15-DQ6) in invasive cervical cancer
TABLE I – FREQUENCY OF HLA-DQA1 ALLELES IN PATIENTS AND CONTROLS DQA1
Patients (n ⫽ 85)
Controls (n ⫽ 120)
0101 0102 0103 0201 0301 0501 0401 0601 0104
21 (25%) 46 (54%) 14 (16%) 12 (14%) 39 (46%) 19 (22%) 1 2 0
27 (23%) 23 (19%)1 32 (27%) 27 (23%) 52 (43%) 37 (31%) 1 0 1
1
OR ⫽ 4.92, 95% CI ⫽ 2.64 –9.18, p ⬍ 0.0001, pc ⬍ 0.05.
700
GHADERI ET AL. TABLE II – FREQUENCY OF HLA-DQB1 ALLELES IN PATIENTS AND CONTROLS
DQB1
Patients (n ⫽ 85)
Controls (n ⫽ 120)
0501 0503 0602 0603 0604 0605 0301 0302 0303 0201 0402 0502 0504 0601
18 (21%) 5 (6%) 34 (41%) 12 (14%) 10 (12%) 2 (2%) 20 (24%) 24 (28%) 10 (12%) 19 (22%) 1 1 1 0
24 (20%) 5 (4%) 16 (14%)1 27 (23%) 6 (5%) 2 (2%) 31 (26%) 33 (28%) 16 (13%) 42 (35%) 1 1 0 1
1
OR ⫽ 4.33, 95% CI⫽2.19 – 8.57, p ⬍ 0.0001, pc ⬍ 0.05. TABLE III – FREQUENCY OF HLA-DRB1 ALLELES IN PATIENTS AND CONTROLS
DRB1
Patients (n ⫽ 84)
Controls (n ⫽ 113)
18 (21%) 14 (17%) 31 (37%) 10 (12%) 8 (10%) 5 (6%) 18 (21%) 9 (11%) 32 (38%) 1 0 3 0 2
20 (18%) 25 (21%) 33 (28%) 28 (25%) 7 (6%) 7 (6%) 27 (24% 12 (11%) 16 (14%)1 1 0 1 1 2
01 03 04 07 09 11 13 14 15 16 02 08 10 12
1 OR ⫽ 0.87, 95%CI ⫽ 0.44 –1.71, 2) OR ⫽ 3.73, 95% CI ⫽ 1.87–7.42, p ⫽ 0.0002, pc ⬍ 0.05.
TABLE IV – HLA-DQ6 (DQA1*0102-DQB1*0602) FREQUENCY IN PATIENTS AND CONTROLS1
1
Patients (n ⫽ 85)
Controls (n ⫽ 120)
34 (40%)
16 (13%)
OR ⫽ 4.33, 95% CI ⫽ 2.19 – 8.57, p ⬍ 0.0001, pc ⬍ 0.05.
patients that had previously been found to be associated with CIN16,17 in women from the Va¨ sterbotten area of Northern Sweden. We designed a case-control study nested within a populationbased cohort: a population for which all cervical smears taken had been stored. Our study design minimizes the risk for selection biases, since samples from all eligible cases and controls in the study base are available for analysis. The only information loss encountered was due to the samples that were not adequate for PCR. The generalizability of the findings to the general population is adequate, since the population-based attendance rate for cervical screening in this population is ⬎ 80%. Therefore, we were even able to estimate the population attributable proportion of DQ6/ DR15 in the etiology of cervical cancer. These study design characteristics are similar to those of our previous cohort study that also disclosed DQ6/DR15 as risk factors for CIN. However, the present study advances the previous knowledge in 2 aspects: i) invasive cancer was used as the endpoint. Many CIN lesions do not progress to cancer and may spontaneously regress. ii) The study had a predefined hypothesis to evaluate only the alleles that were associated with disease in our previous
cohort study, thus limiting the problem of multiple hypothesis testing. Based on our results, we conclude that a genetic risk factor that accounts for as much as 30% of invasive cancer in this population is located at the MHC class II locus close to the DQ6/DR15 alleles. It is difficult to ascertain if DR15 or DQ6 mediates the association with cervical cancer because they are in tight linkage disequilibrium with each other. Several previous studies reported that an association with DR15/ DQ6 became stronger when analyses were restricted to HPV type 16 positive cases.13,16,17 In our study, only a tendency for over representation of DQ6 allele frequencies among the patients with persistent HPV 16 positivity compared to patients who were DQ6 negative was found. A recent case-control study of the Brazilian population has also indicated an association of HLA-DR15 with cervical cancer.29 The HLA-DRB1 *07 allele, which was not part of the predefined study hypothesis, exhibited an insignificant increase among the patients. DRB1 *7 was associated with HPV 16 susceptibility in a study that employed Dutch women in which HLAB*44 also was reported to be increased in patients with worse clinical outcomes during follow-up. Interestingly, HLA-B*44 is in linkage with DRB1*07 as an extended haplotype with DQB1*0201.30 The association of specific HLA alleles with susceptibility to infectious and autoimmune diseases is probably attributable to a direct involvement of the HLA molecules as an antigen presenter or possibly due to a close gene in linkage. The human MHC is an immunologic gene cluster, containing genes of known and less known function.31,32 Apart from the HLA class I and class II genes coding for HLA molecules involved in antigen presentation, this genetic region contains e.g., genes coding for TNF-␣, TNF-, the complement system, Heat Shock Proteins (HSP), peptide transporter (TAP) and the MHC Class I Chain related proteins (MIC-A and B). An association of HLA with diseases does not necessarily reflect the direct involvement of this molecule in the disease process. Since many genes are in linkage disequiblirium with each other on the human MHC, it is possible that the association is due a closely linked gene. The present study indicates that class II genes are particularly important in cervical carcinogenesis. In fact, more than 70% of cervical cancers express HLA class II antigens, despite the fact that squamous cervical epithelium is HLA class II negative.33 Up-regulation of HLA class II expression in the cervical tumors may improve the ability of the immune response to destroy the virus-infected cells. However, the critical costimulatory molecules such as CD28 ligand and CD 80 (B7) are not expressed in cervical tumors and thus MHC class II expression in the absence of co-stimulatory molecules may instead result in tolerance. Cervical carcinoma cells regularly over-express the virally encoded oncoproteins E6 and E7.7 These oncogenic proteins are able to form complexes with the cell regulatory proteins p53 and retinoblastoma (Rb) tumor-suppressors and disrupt their cell cycle regulatory functions.34 The multifunctional p53 protein is also able to enhance the expression of the transporter associated with antigen processing (TAP1) molecule.35 TAP molecules are involved in loading of processed antigen peptides into the HLA class I molecules. Inactivating of p53 by the E6 protein is therefore an indirect mechanism for HPV 16 to inhibit loading of HPV16 peptides into HLA class I molecules. Albeit HLA class I alleles are commonly down regulated in cervical cancers,36 there is little evidence to associate specific alleles with cervical cancer risk. Vaccination strategies targeting MHC class I- presented CTL epitopes encoded by E6 and E7 proteins have been highly successful in murine models37 but have not been successful in humans. The immunologic mechanisms responsible for clearance of HPV infection are not defined, but observations of
HLA-DQ6 AND INVASIVE CERVICAL CANCER RISK
immunodeficient subjects have implicated cellular immunity. The present study suggests that the mechanisms responsible for control of HPV infection may primarily involve MHC class II-mediated responses. In conclusion, cervical carcinoma is associated with specific HLA class II alleles and haplotypes. A further elucidation of the mechanisms responsible for this association may provide
701
useful information for design of preventive or therapeutic vaccines. ACKNOWLEDGEMENTS
This work was supported by grants from Karolinska Institutet, Stockholm, Sweden, Swedish Medical Research Council to CBS (project number k2001-16x-12532-04B).
REFERENCES
1. 2.
3.
4. 5. 6.
7. 8. 9. 10. 11. 12. 13.
14. 15.
16.
17.
18. 19. 20.
Almarri A, Batchelor JR. HLA and hepatitis B infection [see comments]. Lancet 1994;344:1194 –5. Lu SJ, Day NE, Degos L, Lepage V, Wang PC, Chan SH, Simons M, McKnight B, Easton D, Zeng Y, et al. Linkage of a nasopharyngeal carcinoma susceptibility locus to the HLA region. Nature 1990;346: 470 –1. Thursz MR, Kwiatkowski D, Allsopp CE, Greenwood BM, Thomas HC, Hill AV. Association between an MHC class II allele and clearance of hepatitis B virus in the Gambia [see comments]. N Engl J Med 1995;332:1065–9. Oza AM, Tonks S, Lim J, Fleetwood MA, Lister TA, Bodmer JG. A clinical and epidemiological study of human leukocyte antigen-DPB alleles in Hodgkin’s disease. Cancer Res 1994;54:5101–5. Burt RD, Vaughan TL, Nisperos B, Swanson M, Berwick M. A protective association between the HLA-A2 antigen and nasopharyngeal carcinoma in US Caucasians. Int J Cancer 1994;56:465–7. Wallin KL, Wiklund F, Angstrom T, Bergman F, Stendahl U, Wadell G, Hallmans G, Dillner J. Type-specific persistence of human papillomavirus DNA before the development of invasive cervical cancer. N Engl J Med 1999;341:1633– 8. zur Hausen H. Papillomavirus infections: a major cause of human cancers. Biochim Biophys Acta 1996;1288:F55–78. Schiffman MH, Brinton LA. The epidemiology of cervical carcinogenesis. Cancer 1995;76:1888 –901. Lowy DR, Kirnbauer R, Schiller JT. Genital human papillomavirus infection. Proc Natl Acad Sci USA 1994;91:2436 – 40. Ahlbom A, Lichtenstein P, Malmstrom H, Feychting M, Hemminki K, Pedersen NL. Cancer in twins: genetic and nongenetic familial risk factors. J Natl Cancer Inst 1997;89:287–93. Magnusson PK, Sparen P, Gyllensten UB. Genetic link to cervical tumours [letter]. Nature 1999;400:29 –30. Odunsi K, Terry G, Ho L, Bell J, Cuzick J, Ganesan TS. Association between HLA DQB1 * 03 and cervical intra-epithelial neoplasia. Mol Med 1995;1:161–71. Apple RJ, Becker TM, Wheeler CM, Erlich HA. Comparison of human leukocyte antigen DR-DQ disease associations found with cervical dysplasia and invasive cervical carcinoma. J Natl Cancer Inst 1995;87:427–36. Wank R, Thomssen C. High risk of squamous cell carcinoma of the cervix for women with HLA- DQw3 [see comments]. Nature 1991; 352:723–5. Glew SS, Duggan-Keen M, Ghosh AK, Ivinson A, Sinnott P, Davidson J, Dyer PA, Stern PL. Lack of association of HLA polymorphisms with human papillomavirus-related cervical cancer. Hum Immunol 1993;37:157– 64. Helland A, Olsen AO, Gjoen K, Akselsen HE, Sauer T, Magnus P, Borresen-Dale AL, Ronningen KS. An increased risk of cervical intra-epithelial neoplasia grade II-III among human papillomavirus positive patients with the HLA-DQA1*0102- DQB1*0602 haplotype: a population-based case-control study of Norwegian women. Int J Cancer 1998;76:19 –24. Sanjeevi CB, Hjelmstrom P, Hallmans G, Wiklund F, Lenner P, Angstrom T, Dillner J, Lernmark A. Different HLA-DR-DQ haplotypes are associated with cervical intraepithelial neoplasia among human papillomavirus type-16 seropositive and seronegative Swedish women. Int J Cancer 1996;68:409 –14. Chua KL, Hjerpe A. Polymerase chain reaction analysis of human papillomavirus in archival cervical cytologic smears. Anal Quant Cytol Histol 1995;17:221–9. Chua KL, Hjerpe A. Persistence of human papillomavirus (HPV) infections preceding cervical carcinoma. Cancer 1996;77:121–7. Rhoads DD, Dixit A, Roufa DJ. Primary structure of human ribosomal protein S14 and the gene that encodes it. Mol Cell Biol 1986;6:2774 – 83.
21. Erlich HA, Bugawan BL. HLA class II gene polymorphisms: DNA typing, evolution and relationship to disease susceptibility. In: Erlich HA, ed. PCR technology: principles and applications for DNA amplification. New York: Stocktorn Press, 1989. 193–208. 22. Sanjeevi CB, Seshiah V, Moller E, Olerup O. Different genetic backgrounds for malnutrition-related diabetes and type 1 (insulindependent) diabetes mellitus in south Indians. Diabetologia 1992;35: 283– 6. 23. Manos MM, Ting Y, Wright DK, Lewis AJ, Broker TR, Wolinsky SM. Use of polymerase chain reaction for detection of genital human papillomaviruses. In: Furth M, Greaves M, eds. Molecular diagnostics of human cancer. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, 1989. vol. 7, 209 –14. 24. de Roda Husman AM, Walboomers JM, van den Brule AJ, Meijer CJ, Snijders PJ. The use of general primers GP5 and GP6 elongated at their 3⬘ ends with adjacent highly conserved sequences improves human papillomavirus detection by PCR. J Gen Virol 1995;76:1057– 62. 25. Evander M, Edlund K, Boden E, Gustafsson A, Jonsson M, Karlsson R, Rylander E, Wadell G. Comparison of a one-step and a two-step polymerase chain reaction with degenerate general primers in a population-based study of human papillomavirus infection in young Swedish women. J Clin Microbiol 1992;30:987–92. 26. Chua KL, Wiklund F, Lenner P, Angstrom T, Hallmans G, Bergman F, Sapp M, Schiller J, Wadell G, Hjerpe A, Dillner J. A prospective study on the risk of cervical intra-epithelial neoplasia among healthy subjects with serum antibodies to HPV compared with HPV DNA in cervical smears. Int J Cancer 1996;68:54 –9. 27. Miettinen O. Estimability and estimation in case-referent studies. Am J Epidemiol 1976;103:226 –35. 28. Woolf B. On estimating the relation between blood group and disease. Am Hum Genet 1955:251–253. 29. Maciag PC, Schlecht NF, Souza PS, Franco EL, Villa LL, Petzl-Erler ML. Major histocompatibility complex class II polymorphisms and risk of cervical cancer and human papillomavirus infection in Brazilian women. Cancer Epidemiol Biomarkers Prev 2000;9:1183–91. 30. Bontkes HJ, van Duin M, de Gruijl TD, Duggan-Keen MF, Walboomers JM, Stukart MJ, Verheijen RH, Helmerhorst TJ, Meijer CJ, Scheper RJ, Stevens FR, Dyer PA, et al. HPV 16 infection and progression of cervical intra-epithelial neoplasia: analysis of HLA polymorphism and HPV 16 E6 sequence variants. Int J Cancer 1998; 78:166 –71. 31. Campbell RD, Trowsdale J. Map of the human MHC. Immunol Today 1993;14:349 –52. 32. Trowsdale J, Powis SH. The MHC: relationship between linkage and function. Curr Opin Genet Dev 1992;2:492–7. 33. Glew SS, Connor ME, Snijders PJ, Stanbridge CM, Buckley CH, Walboomers JM, Meijer CJ, Stern PL. HLA expression in pre-invasive cervical neoplasia in relation to human papilloma virus infection. Eur J Cancer 1993;14:1963–70. 34. Farthing AJ, Vousden KH. Functions of human papillomavirus E6 and E7 oncoproteins. Trends Microbiol 1994;2:170 – 4. 35. Zhu K, Wang J, Zhu J, Jiang J, Shou J, Chen X. p53 induces TAP1 and enhances the transport of MHC class I peptides. Oncogene 1999; 18:7740 –7. 36. Keating PJ, Cromme FV, Duggan-Keen M, Snijders PJ, Walboomers JM, Hunter RD, Dyer PA, Stern PL. Frequency of down-regulation of individual HLA-A and -B alleles in cervical carcinomas in relation to TAP-1 expression. Br J Cancer 1995;72:405–11. 37. Meneguzzi G, Cerni C, Kieny MP, Lathe R. Immunization against human papillomavirus type 16 tumor cells with recombinant vaccinia viruses expressing E6 and E7. Virology 1991;181:62–9.
