Association of gastric adenocarcinoma with the HLA class II gene DQB10301

GASTROENTEROLOGY 1996;111:426–432

Association of Gastric Adenocarcinoma With the HLA Class II Gene DQB1*0301 JEFFREY E. LEE,* ANDREW M. LOWY,* WILLIAM A. THOMPSON,* MEISHENG LU,* PAUL T. LOFLIN,‡ JOHN M. SKIBBER,* DOUGLAS B. EVANS,* STEVEN A. CURLEY,* PAUL F. MANSFIELD,* and JOHN D. REVEILLE§ Departments of *Surgical Oncology and ‡Immunology, University of Texas M. D. Anderson Cancer Center, Houston; and §Division of Molecular Immunology and Rheumatology, University of Texas Health Science Center, Houston, Texas

See editorial on page 523. Background & Aims: The HLA class II gene DQB1*0301 has been linked to several cancers. This study was designed to determine if HLA-DQB1*0301 is present at altered frequency in patients with gastric, colorectal, or pancreatic adenocarcinoma. Methods: Oligotyping for HLA-DQB1*0301 was performed for 159 Caucasian patients with 160 gastrointestinal adenocarcinomas (52 gastric, 62 colorectal, and 46 pancreatic adenocarcinomas) and compared with 200 Caucasian noncancer controls. Patients with gastric adenocarcinoma underwent extended HLA class II region oligotyping. Immunoglobulin G to Helicobacter pylori was detected by enzyme-linked immunosorbent assay. Results: HLADQB1*0301 was more common in patients with gastric adenocarcinoma than controls (54% vs. 27%; Bonferroni-corrected x2 P Å 0.003; odds ratio, 3.2). HLADQB1*0301 was not associated with colorectal or pancreatic adenocarcinoma. No other HLA-DQB1 allele and no HLA-DQA1 or transporter associated with antigen processing 2 (TAP2) allele were present at altered frequency in patients with gastric adenocarcinoma. Serological evidence for H. pylori infection was less frequent in HLA-DQB1*0301–positive patients with gastric adenocarcinoma compared with HLA-DQB1*0301–negative patients (52% vs. 88%; Fisher’s Exact Test; P Å 0.007). Conclusions: HLA-DQB1*0301 is more common in Caucasian patients with gastric adenocarcinoma than noncancer controls. The mechanism linking HLA-DQB1*0301 with gastric adenocarcinoma is not likely through increased susceptiblity to H. pylori infection.

H

LA class II genes are a group of highly polymorphic genes on chromosome 6p that are particularly important in controlling specific immune recognition (see the Appendix for a glossary of terms1). HLA class II antigens are capable of binding tumor peptides,2 – 4 and / 5e10$$0003

07-02-96 18:03:03

gasa

T-cell recognition of the combination of HLA class II and bound tumor antigen may result in either induction of an effective antitumor immune response2,3 or suppression of that immune response.5 In some cases, the HLAbound peptide antigen recognized by T cells is derived from specific oncogene products (e.g., K-ras).4,6,7 Previous investigations have linked specific HLA class II alleles to cancer incidence and progression. Cervical cancer8 – 10 and human T lymphotrophic virus 1 (HTLV1)–associated leukemia-lymphoma11 have been associated with an increased frequency of certain HLA class II genes (including HLA-DQB1*0301) and/or HLA class II antigens (including the HLA-DQB1*0301 antigen product DQ7). Our own studies have shown a strong association between HLA-DQB1*0301 and melanoma; the presence of HLA-DQB1*0301 in patients with melanoma is associated with an advanced stage of disease at diagnosis.12 Furthermore, mapping studies of the HLA class II region of chromosome 6p suggest an etiologic role for the HLA-DQB1 locus in melanoma susceptibility.13 HLA class II antigens, including DQ7, the protein product of HLA-DQB1*0301, have been implicated in the immune response to gastrointestinal adenocarcinomas.6,7 Interestingly, in patients with colorectal cancer, immune response to mutated K-ras peptides by DQ7restricted T cells has been associated with a more favorable stage at presentation.6 The unique sequence that defines HLA-DQB1*0301 includes polymorphisms in the peptide-binding domain, and the allele-specific peptide-binding motif of DQ7 has been defined.14,15 In some autoimmune diseases, allele-specific binding of antigenic peptides may account for associations between HLA class II genes and disease susceptibility.16 Taken together, these data suggest that allele-specific binding of tumor Abbreviations used in this paper: HTLV-1, human T lymphotrophic virus 1; PCR, polymerase chain reaction; TAP2, transporter associated with antigen processing 2. 䉷 1996 by the American Gastroenterological Association 0016-5085/96/$3.00

WBS-Gastro

August 1996

HLA-DQB1*0301 AND GASTRIC CANCER 427

peptides to DQ7 could result in an altered immune response in individuals with or at risk of developing certain cancers. Because immune mechanisms involving HLADQB1*0301 may play an important role in the etiology or progression of some gastrointestinal cancers, we designed this study to determine whether HLADQB1*0301 was present at an altered frequency in patients with gastric, colorectal, or pancreatic adenocarcinoma and, if so, to correlate this with clinicopathologic indicators at presentation.

Patients and Methods Approval for this research project was obtained from the University of Texas M. D. Anderson Surveillance Committee for Human Subject Research and the Office of the Vice President for Research. Verbal informed consent was obtained from all patients.

Staging and Inclusion Criteria Patients with gastrointestinal adenocarcinoma evaluated in the Surgical Oncology Clinic at the University of Texas M. D. Anderson Cancer Center between January 1994 and September 1995 were considered for inclusion. Race-matching patients and controls for the purposes of HLA typing required the exclusion of non-Caucasian patients. The final study group consisted of 159 Caucasian patients presenting with a total of 160 gastrointestinal adenocarcinomas (52 gastric adenocarcinomas, 62 colorectal adenocarcinomas, and 46 pancreatic adenocarcinomas; a single patient presented with metachronous gastric adenocarcinoma after colorectal adenocarcinoma). The American Joint Committee on Cancer stage at presentation17 of each patient with gastric adenocarcinoma was determined by physical examination, review of histopathology, and laboratory and radiograph examination.

tion oligonucleotides were synthesized by Genosys Biotechnologies (The Woodlands, TX). HLA-DRB1 oligotyping for HLA-DR1, -DR2, -DR3/6, -DR4, -DR7, -DR8/12, -DR9, -DR10, -DR11, DR52a, -DR52b/c, and -DR53 was performed to clarify ambiguous HLA-DQA1 and HLA-DQB1 allelic designations.

TAP2 Typing The presence of transporter associated with antigen processing 2 (TAP2) alleles in genomic DNA from an unselected subset of 32 patients with gastric cancer and 37 controls was determined by PCR–sequence-specific oligonucleotide typing. This technique permits designation of TAP2 alleles by identifying codons present at three dimorphic positions (379 ILE/VAL, 565 ALA/THR, and 665 ALA/THR). The amplification primers, PCR conditions, hybridization oligonucleotides, and screening of hybridization filters were as described previously.13,19 Primers and hybridization oligonucleotides were synthesized by Genosys Biotechnologies. As in previous studies, unequivocal assignment of a TAP2 allele type is not always possible in heterozygotes because genomic DNA is used to type each of the three dimorphic residues individually, and genotyping by PCR–sequence-specific oligonucleotide generates identical patterns for TAP2*0101(A)-TAP2*D and TAP2*C-TAP2*102(E) haplotypes. Therefore, when the oligotyping results for TAP2 alleles were ambiguous, summary allele frequencies were reported as a range.

Helicobacter pylori Serotyping The presence of serum immunoglobulin G antibodies to H. pylori was determined by enzyme-linked immunosorbent assay with the Pyloragen kit (Hycor, Irvine, CA). Samples were tested in duplicate, and the results were averaged. Patients whose initial results were designated equivocal by the test kit instructions were retested.

Clinical Database A patient with gastric carcinoma database of known clinicopathologic prognostic variables was constructed, independent of HLA typing.

HLA Class II Typing HLA-DRB1, HLA-DQA1, and HLA-DQB1 alleles were identified by polymerase chain reaction (PCR) sequencespecific oligonucleotide molecular oligotyping. Briefly, genomic DNA was extracted from peripheral blood leukocytes using phenol-chloroform-isoamyl alcohol. This DNA served as the substrate for PCR amplification of a polymorphic fragment of the first domain exon of HLA-DRB1, HLA-DQA1, or HLADQB1 using locus-specific flanking primers. The sizes of the PCR products were determined by electrophoresis. The amplified DNA was dot blotted, and the blots were hybridized with [g32P]adenosine triphosphate–labeled allele-specific oligonucleotide probes. The amplification primers, PCR conditions, hybridization oligonucleotides, and screening of hybridization filters were as described previously.12,18 Primers and hybridiza-

/ 5e10$$0003

07-02-96 18:03:03

gasa

Statistical Analysis HLA-DQA1, HLA-DQB1, and TAP2 allele frequencies were compared between patients with gastrointestinal adenocarcinoma and controls by x2 analysis or by Fisher’s Exact Test, as dictated by sample size. The method of Bonferroni was used to adjust P values for multiple inferences.20

Results HLA-DQB1*0301 was present at increased frequency in patients with gastric adenocarcinoma but not in patients with colorectal adenocarcinoma or pancreatic adenocarcinoma (Table 1). The only HLA-DQB1 allele at altered frequency in patients with gastric adenocarcinoma was HLA-DQB1*0301 (Table 2). Although Bonferroni correction for multiple inferences is not strictly required for the x2 analysis of association between HLADQB1*0301 and gastric adenocarcinoma because this association was specifically sought in the current study, WBS-Gastro

428 LEE ET AL.

GASTROENTEROLOGY Vol. 111, No. 2

Table 1. Presence of HLA-DQB1*0301 in Caucasian Patients With Gastrointestinal Adenocarcinoma and Caucasian Control Individuals

Table 3. TAP2 Allele Frequencies in Caucasian Patients With and Without Gastric Adenocarcinoma Frequency (%)a

Group

n

HLA-DQB1*0301 present (%)

P valuea

Gastric adenocarcinoma Colorectal adenocarcinoma Pancreatic adenocarcinoma Controls

52 62 46 200

54 37 28 27

0.003b,c 0.12 0.86 —

P value determined by x2 analysis. b Odds ratio vs. controls, 3.2. c Bonferroni-corrected for n alleles tested, 15. a

the more conservative Bonferroni-corrected P value is reported for consistency. Because the HLA-DQA1 locus is the closest gene of known function downstream from (or telomeric to) the HLA-DQB1 locus on chromosome 6p21 – 28 and because HLA-DQA1 alleles are known to be in linkage disequilibrium with HLA-DQB1 alleles in Caucasian individuals,29,30 we sought evidence for altered frequency of HLADQA1 alleles in patients with gastric cancer (Table 2). There is a trend towards increased frequency of the HLADQA1*0201 allele in patients with gastric adenocarcinoma compared with controls. However, Bonferroni correction suggests that this difference in frequency may not be significant. Because the TAP2 locus is the closest gene of known

Table 2. HLA Class II Allele Frequencies in Caucasian Patients With and Without Gastric Adenocarcinoma Frequency (%)

Allele

HLA-DQA1 0101 0102 0103 0201 03 0401 0501 HLA-DQB1 0501-0504 0601-0605 0201 0301 0302 0303 0401 0402

Patients Controls (n Å 52) (n Å 200)

P valuea

Bonferroni corrected P valueb

21 30 7 37 20 5 46

32 37 17 20 32 6 36

0.15 0.40 0.09 0.02 0.60 0.73 0.19

— — — 0.14 — — —

20 40 40 54 8 12 0 10

34 52 36 27 17 5 0 7

0.06 0.13 0.60 0.0002 0.11 0.07 — 0.52

— — — 0.003 — — — —

P value determined by x2 analysis. For HLA-DQA1, seven alleles were tested; for HLA-DQB1, 15 alleles were tested.

a b

/ 5e10$$0003

07-02-96 18:03:03

gasa

TAP2 allele 0101 0102 0103 2C 2D 0201 PSF2

(A) (E) (F)

(B)

Patientsb (n Å 37) 78.4 { 5.4 13.5 { 5.4 10.8 15.9 { 5.1 13.2 { 7.8 29.5 { 2.5 5.4

Controlsc (n Å 32) 76.6 { 15.6 { ND 17.2 { 9.4 { 42.2 { 4.7 {

7.8 9.4 7.8 9.4 1.6 1.6

P valued — — — — — 0.60e —

ND, not detected. a Range. b Four ambiguous genotypes. c Six ambiguous genotypes. d P values determined by x2 analysis. e Assumes the maximum difference in frequency to estimate the smallest possible P value.

function upstream of (or centromeric to) the HLA-DQB1 locus on chromosome 6p,21 – 28 we investigated whether TAP2 alleles were present at an altered frequency in the unselected subset described in Patients and Methods (Table 3). No altered frequency of any TAP2 allele was detected in patients with gastric cancer, and the frequencies of TAP2 alleles we found in the Caucasian control population correspond with those reported for a larger Caucasian population.31 Because of previously noted associations between cancer stage at presentation and HLA-DQB1*0301 status in patients with melanoma12 as well as a reported association of the HLA-DQB1*0301 antigen DQ7 with favorable Dukes’ stage in patients with colorectal cancer,6 we sought an association between HLA-DQB1*0301 status and clinicopathologic indicators in patients with gastric adenocarcinoma. As shown in Table 4, there were no differences between HLA-DQB1*0301–positive and –negative patients with gastric adenocarcinoma with regard to age at presentation, sex ratio, primary anatomic tumor site (distal vs. proximal), TNM stage at presentation, or histological grade of the primary tumor. Because our follow-up of this series of patients with gastric adenocarcinoma is still relatively short (approximately 12 months), any attempt to associate the presence of the HLA-DQB1*0301 gene with prognosis in patients with gastric adenocarcinoma is premature. Therefore, no analysis of disease-free survival or overall survival with regard to HLA-DQB1*0301 status in patients with gastric adenocarcinoma has yet been attempted. Because H. pylori has been proposed as an etiologic agent in the development of gastric adenocarcinoma31–36 and because an association has been reported between WBS-Gastro

August 1996

HLA-DQB1*0301 AND GASTRIC CANCER 429

Table 4. Clinical Characteristics and HLA-DQB1*0301 Status at Presentation of 52 Caucasian Patients With Gastric Adenocarcinoma Characteristic

HLA-DQB1*0301 –positive patients (n Å 28)

HLA-DQB1*0301 –negative patients (n Å 24)

P valuea

Mean age (yr ) Sex ratio (M/F) Tumor site (% distal) Tumor stage (% T3–T4) Nodes stage (% N1–N2) Metastasis stage (% M1) Grade (% poorly differentiated) H. pylori immunoglobulin G antibodies present (%)

62 1.5/1 61 71 53 29 68 52

62 1.7/1 65 63 74 25 67 88

0.87 0.90 0.74 0.58 0.18 0.77 0.93 0.007

a

P value determined as follows: age by two-tailed Student’s t test and sex ratio; tumor site; tumor, nodes, and metastasis stages; grade; and H. pylori status by x2 analysis.

HLA-DQA1 alleles and H. pylori–associated benign gastroduodenal disease,37 we sought evidence for an association between the presence of HLA-DQ alleles and serological evidence of H. pylori infection in patients with gastric adenocarcinoma. HLA-DQB1*0301–positive patients with gastric cancer were less likely to have serological evidence of H. pylori infection than patients lacking the allele (Table 4). Contrary to the findings of others in benign gastroduodenal disease,37 we found no association between any HLA-DQA1 allele and H. pylori seropositivity, and there was no association between any other HLADQB1 allele and H. pylori seropositivity (data not shown).

Discussion As shown in Table 1, HLA-DQB1*0301 was present at increased frequency in Caucasian patients with gastric adenocarcinoma but not in those patients with colorectal or pancreatic adenocarcinoma. This suggests that HLA-DQB1*0301 may be important in the etiology of gastric adenocarcinoma. Furthermore, none of the remaining common HLA-DQB1 alleles occurred at markedly altered frequency in the patients with gastric adenocarcinoma (Table 2). Although there may be other secondary alleles present at altered frequency in patients with gastric adenocarcinoma that would be shown in a larger series, the association of HLA-DQB1*0301 with gastric cancer is likely to be the strongest of the alleles we investigated. Two distinct potential mechanisms could be responsible for the association between HLA-DQB1*0301 and cervical cancer, HTLV-1 leukemia-lymphoma, melanoma, and gastric adenocarcinoma: linkage disequilibrium or HLA-DQB1*0301–mediated alterations in antigen-specific recognition. Linkage disequilibrium would require strong linkage between HLA-DQB1*0301 and a putative etiologic gene nearby on chromosome 6p. The HLA-DQA1 and TAP2 loci are the closest genes of known function telomeric and centromeric to HLA/ 5e10$$0003

07-02-96 18:03:03

gasa

DQB1, respectively.21 – 28 Each of these loci codes for a distinct immunologic product; HLA-DQB1 is highly polymorphic and codes for the b subunit of HLA-DQ, HLA-DQA1 is less polymorphic and codes for the a subunit, and TAP2 is moderately polymorphic and codes for antigens that transport peptides across the endoplasmic reticulum for loading onto HLA class I molecules. HLA-DQA1 alleles are in linkage disequilibrium with HLA-DQB1 alleles.29,30 Although controversial, TAP2 alleles also may be in linkage disequilibrium with HLADQB1 alleles.38 – 40 In this study, as observed with melanoma,12,13 neither HLA-DQA1 alleles nor TAP2 alleles were strongly associated with gastric cancer. Although not all HLA haplotypes have been examined thoroughly, the relatively small (approximately 400 kilobases) fragment of chromosome 6 containing HLA-DQB1, defined by TAP2 upstream and HLA-DRB1 downstream, is not known to contain other functional genes. Several loci within this region, such as HLA-DOB, -DQB2, -DQB3, and -DQA2, are either pseudogenes or encode molecules of unknown function. Although we cannot exclude completely a role for other nearby and as of yet uncharacterized genes in gastric cancer etiology, these data taken together suggest that HLA-DQB1*0301 may be etiologic in gastric cancer. As for the immunologic mechanism, because T-cell recognition of the combination of HLA-DQ and bound tumor antigen may result in altered antitumor immune responses,6 it is reasonable to speculate that allele-specific binding of peptide antigens to the HLA-DQB1*0301 antigen product DQ7 could result in altered susceptibility to gastric cancer. One reasonable argument may be that HLA-DQB1*0301 results in altered susceptibility to H. pylori infection or altered susceptibility to cancer progression after H. pylori infection because evidence of prior infection with H. pylori is associated with later development of gastric adenocarcinoma32,33; HLA-DQB1 seems to control susceptibility to a variety of chronic WBS-Gastro

430 LEE ET AL.

GASTROENTEROLOGY Vol. 111, No. 2

infectious diseases, including schistosomiasis,41 leprosy,42 malaria,43 and hepatitis B virus44; altered susceptibility to chronic infection may play a role in the association of HLA-DQB1 alleles with cervical cancer45 – 47 and/or HTLV-1 leukemia-lymphoma11; and certain HLA-DQA1 alleles are found at altered frequency in patients with H. pylori–associated benign gastroduodenal disease.37 However, in our study, we found a decreased rate of H. pylori seropositivity in patients with gastric cancer carrying the HLA-DQB1*0301 allele. This suggests that the mechanism responsible for the association of HLA-DQB1*0301 with gastric cancer is not through increased susceptibility to H. pylori infection. Rather, our results suggest that HLA-DQB1*0301–positive patients with gastric cancer may be at a decreased risk for infection with H. pylori, that HLA-DQB1*0301–positive individuals infected with H. pylori may be less likely to develop gastric cancer, that HLA-DQB1*0301–positive patients with gastric cancer may be less likely to develop an immunoglobulin G response to H. pylori, or that HLA-DQB1*0301 and H. pylori infection may represent alternative mechanisms for the development of gastric cancer.

In summary, our preliminary findings show that patients with gastric adenocarcinoma are more likely than noncancer controls to carry the immune recognition gene HLADQB1*0301. The association of HLA-DQB1*0301 with gastric cancer, cervical cancer, HTLV-1 leukemia-lymphoma, and melanoma suggests that this gene may represent a genomic marker of susceptibility to certain cancers. The decreased H. pylori seropositivity rate in HLA-DQB1*0301–positive patients with gastric adenocarcinoma implies that, although both HLA-DQB1*0301 and H. pylori infection could play a role in the etiology of gastric cancer, HLA-DQB1*0301 does not seem to influence gastric cancer susceptibility via increased risk of H. pylori infection. Future investigations should seek to define potential associations between HLA-DQ alleles and the presence or absence of H. pylori infection in patients with benign peptic ulcer disease37 and mucosa-associated lymphatic tissue lymphoma,48,49 as well as gastric adenocarcinoma. Identification of the mechanism associating HLADQB1*0301 with gastric cancer could ultimately help target individuals most likely to benefit from cancer screening and prevention programs and could suggest novel therapeutic strategies for cancer immunoprevention.

Appendix: Glossary Term

Definition

HLA antigen

Locus HLA allele

Linkage Linkage disequilibriium TAP

Also referred to as HLA specificity. The HLA antigen is the HLA protein as expressed on the cell surface: a heterodimer composed (in the case of HLA-DQ and other HLA class II antigens) of an a chain and a b chain. HLA antigens are commonly detected by serological typing and are designated by terms such as DQ7 or HLA-DR1. HLA antigens bind (present) short linear peptides (peptide antigens) for recognition by T cells through the formation of a trimolecular complex between the HLA antigen, the peptide antigen, and the antigen-specific T-cell receptor of the T cell. The formation of this trimolecular complex is the fundamental antigen-specific signaling event in cellular immunity. A site on a chromosome occupied by a specific gene. One of the potentially many forms of the HLA gene. Because this term refers to differences in the DNA sequence of the gene, different alleles (or forms) of an HLA gene are usually detected by examination of the DNA sequence. Because DNA typing examines the alleles at a given locus, the name of the locus precedes the designation of the specific allele (with the two terms separated by an asterisk); for example, HLA-DQB1*0301 refers to the 0301 allele of the HLA-DQB1 locus. One HLA-DQ antigen is encoded by at least one HLA-DQB1 allele (coding for the b chain) in conjunction with the product of the HLA-DQA1 locus (coding for the a chain). Two genes that are physically spaced close together on the same chromosome. Alleles of two linked genes that tend to be passed from generation to generation en bloc and, thus, occur together more frequently than would be expected based on the frequencies of the individual genes. The transporter associated with antigen processing (TAP) genes (TAP1 and TAP2) are moderately polymorphic genes that are located within the HLA class II region of chromosome 6p. They code for the TAP1 and TAP2 proteins that are responsible for transporting peptides across the endoplasmic reticulum for loading onto HLA class I molecules.

References 1. McDaniel DO, Alarcon GS, Pratt PW, Reveille JD. Most AfricanAmerican patients with rheumatoid arthritis do not have the rheumatoid antigenic determinant (epitope). Ann Intern Med 1995; 123:181–187. 2. Topalian SL, Rivoltini L, Mancini M, Markus NR, Robbins PF, Kawakami Y, Rosenberg SA. Human CD4/ T cells specifically recognize a shared melanoma-associated antigen encoded by the tyrosinase gene. Proc Natl Acad Sci USA 1994;91:9461– 9465.

/ 5e10$$0003

07-02-96 18:03:03

gasa

3. Takahashi T, Chapman PB, Yang SY, Hara I, Vijayasaradhi S, Houghton AN. Reactivity of autologous CD4/ T lymphocytes against human melanoma. J Immunol 1995;154:772–779. 4. Peace DJ, Chen W, Nelson H, Cheever MA. T cell recognition of transforming proteins encoded by mutated ras proto-oncogenes. J Immunol 1991;146:2059–2065. 5. Becker JC, Brabletz T, Czerny C, Termeer C, Brocker EB. Tumor escape mechanisms from immunosurveillance: induction of unresponsiveness in a specific MHC-restricted CD4/ human T cell clone by the autologous MHC class II/ melanoma. Int Immunol 1993;5:1501–1508.

WBS-Gastro

August 1996

HLA-DQB1*0301 AND GASTRIC CANCER 431

6. Fossum B, Breivik J, Melig GI, Gedde-Dahl T, Hansen T, Knutsen I, Rognum TO, Thorsby E, Gaudernack G. A K-ras 13Gly r Asp mutation is recognized by HLA-DQ7 restricted T cells in a patient with colorectal cancer: modifying effect of DQ7 on established cancers harboring this mutation? Int J Cancer 1994;58:506– 511. 7. Qin H, Chen W, Takahashi M, Disis ML, Byrd DR, McCahill L, Bertram KA, Fenton RG, Peace DJ, Cheever MA. CD4/ T-cell immunity to mutated ras protein in pancreatic and colon cancer patients. Cancer Res 1995;55:2984–2987. 8. Wank R, Thomssen C. High risk of squamous cell carcinoma of the cervix for women with HLA-DQw3. Nature 1991;352:723– 725. 9. Wank R, Schendel DJ. HLA antigens and cervical carcinoma (lett). Nature 1992;356:22–23. 10. Helland A, Borresen AL, Kaern J, Ronningen KS, Thorsby E. HLA antigens and cervical cancer (letter). Nature 1992;356:23. 11. Uno H, Kawano K, Matsuoka H, Tsuda K. HLA and adult T cell leukaemia: HLA-linked genes controlling susceptibility to human T cell leukaemia virus type I. Clin Exp Immunol 1988;71:211– 216. 12. Lee JE, Reveille JD, Ross MI, Platsoucas CD. HLA-DQB1*0301 association with increased cutaneous melanoma risk. Int J Cancer 1994;59:510–513. 13. Lee JE, Loflin PT, Laud PR, Lu M, Reveille JD, Lawlor DA. The human leukocyte antigen TAP2 gene defines the centromeric limit of melanoma susceptibility on chromosome 6p. Tissue Antigens 1996;47:117–121. 14. Sidney J, Oseroff C, del Guercio M-F, Southwood S, Krieger JI, Ishioka GY, Sakaguchi K, Appella E, Sette A. Definition of a DQ3.1-specific binding motif. J Immunol 1994;152:4516–4525. 15. Falk K, Rotzschke O, Stevanovic S, Jung G, Rammensee H-G. Pool sequencing of natural HLA-DR, DQ and DP ligands reveals detailed peptide motifs, constraints of processing, and general rules. Immunogenetics 1994;39:230–242. 16. Wucherpfennig KW, Strominger JL. Selective binding of self peptides to disease-associated major histocompatibility complex (MHC) molecules: a mechanism for MHC-linked susceptibility to human autoimmune diseases. J Exp Med 1995;181:1597– 1601. 17. American Joint Committee on Cancer. Stomach. In: Manual for staging of cancer. 3rd ed. Philadelphia: Lippincott, 1992:63– 67. 18. Reveille JD, Durban E, MacLeod-St. Clair MJ, Goldstein R, Moreda R, Altman RD, Arnett FC. Association of amino acid sequences in the HLA-DQB1 first domain with the antitopoisomerase I–autoantibody response in scleroderma (progressive systemic sclerosis). J Clin Invest 1992;90:973–980. 19. Colonna M, Bresnahan M, Bahram S, Strominger JL, Spies T. Allelic variants of the human putative peptide transporter associated with antigen processing. Proc Natl Acad Sci USA 1992;89: 3932–3936. 20. Dunn OJ. Multiple comparisons among means. J Am Stat Assoc 1961;56:52–64. 21. Okada K, Boss JM, Prentice H, Spies T, Mengler R, Auffrey C, Lillie J, Grossberger D, Strominger JL. Gene organization of DC and DX subregions of the human major histocompatibility complex. Proc Natl Acad Sci USA 1985;82:3410–3414. 22. Tonnelle C, DeMars R, Long EO. DOB: a new b chain gene in HLA-D with a distinct regulation of expression. EMBO J 1985;4: 2839–2847. 23. Jonsson AK, Hyldig-Nielsen JJ, Servenius B, Larhammer D, Andersson G, Jorgensen F, Peterson PA, Rask L. Class II genes of the human major histocompatibility complex: comparison of the DQ and DX a and b genes. J Biol Chem 1987;262:8767–8777. 24. Ando A, Kawai J, Maeda M, Tsuji K, Trowsdale J, Inoko H. Map-

/ 5e10$$0003

07-02-96 18:03:03

gasa

25.

26.

27. 28. 29.

30.

31.

32.

33.

34.

35.

36. 37.

38.

39.

40.

41.

42.

43.

ping and nucleotide sequence of a new HLA class II light chain gene, DQB3. Immunogenetics 1989;30:243–249. Ragoussis J, Monaco A, Mockridge I, Kendall E, Campbell RD, Trowsdale J. Cloning of the HLA class II region in yeast artificial chromosomes. Proc Natl Acad Sci USA 1991;88:3753–3757. Kozono H, Bronson SK, Taillon-Miller P, Moorti MK, Jamry I, Chaplin DD. Molecular linkage of the HLA-DR, HLA-DQ, and HLADO genes in yeast artificial chromosomes. Genomics 1991;11: 577–586. Nepom GT, Erlich H. MHC class II molecules and autoimmunity. Ann Rev Immunol 1991;9:493–525. Campbell RD, Trowsdale T. Map of the human MHC. Immunol Today 1993;14:349–352. Begovich AB, McClure GR, Suraj VC, Helmuth RC, Fildes N, Bugawan TL, Erlich HA, Klitz W. Polymorphism, recombination, and linkage disequilibrium within the HLA class II region. J Immunol 1992;148:249–258. ˜a MA, Gao X, Moraes ME, Moraes JR, Salatiel I, Fernandez-Vin Miller S, Tsai J, Sun Y, An J, Laysisse Z, Gazit E, Brautbar C, Stastny P. Alleles at four HLA class II loci determined by oligonucleotide hybridization and their associations in five ethnic groups. Immunogenetics 1991;34:299–312. Powis SH, Tonks S, Mockridge I, Kelly AP, Bodmer JG, Trowsdale J. Alleles and haplotypes of the MHC-encoded ABC transporters TAP1 and TAP2. Immunogenetics 1993;37:373–380. Parsonnet J, Friedman GD, Vandersteen DP, Chang Y, Vogelman JH, Orentreich N, Sibley RK. Helicobacter pylori infection and the risk of gastric carcinoma. N Engl J Med 1991;325:1127–1131. Nomura A, Stemmermann GN, Chyou PH, Kato I, Perez-Perez GI, Blasser MJ. Helicobacter pylori infection and gastric carcinoma among Japanese Americans in Hawaii. N Engl J Med 1991;325: 1132–1136. The Eurogast Study Group. An international association between Helicobacter pylori infection and gastric cancer. Lancet 1993; 341:1359–1362. Hansson LE, Engstrand L, Nyren O, Evans DJ, Lindgren A, Bergstrom R, Andersson B, Athlin L, Bendtsen O, Tracz P. Helicobacter pylori infection: independent risk indicator of gastric adenocarcinoma. Gastroenterology 1993;105:1098–1103. Nightingale TE, Gruber J. Helicobacter and human cancer. J Natl Cancer Inst 1994;86:1505–1509. Azuma T, Konishi J, Tanaka Y, Hirai M, Ito S, Kato T, Kohli Y. Contribution of HLA-DQA gene to host’s response against Helicobacter pylori. Lancet 1994;343:542–543. van Endert PM, Lopez MT, Patel SD, Monaco JJ, McDevitt HO. Genomic polymorphism, recombination, and linkage disequilibrium in human major histocompatibility complex–encoded antigen-processing genes. Proc Natl Acad Sci USA 1992;89:11594– 11597. Caillat-Zucman S, Bertin E, Timsit J, Boitard C, Assan R, Bach J. Protection from insulin-dependent diabetes mellitus is linked to a peptide transporter gene. Eur J Immunol 1993;23:1784–1788. Carrington M, Colonna M, Spies T, Stephens JC, Mann DL. Haplotypic variation of the transporter associated with antigen processing (TAP) genes and their extension of HLA class II region haplotypes. Immunogenetics 1993;37:266–273. Hirayama K, Matsushita S, Kikuchci I, Iuchi M, Ohta N, Sasazuki T. HLA-DQ is epistatic in controlling the immune response to schistosomal antigen in humans. Nature 1987;327:426–430. Salgame P, Convit J, Bloom BR. Immunological suppression by human CD8/ T cells is receptor dependent and HLA-DQ restricted. Proc Natl Acad Sci USA 1991;88:2598–2602. Riley EM, Morris-Jones S, Taylor-Robinson AW, Holder AA. Lymphoproliferative responses to a merozoite surface antigen of Plasmodium falciparum: preliminary evidence for seasonal activation of CD8//HLA-DQ–restricted suppressor cells. Clin Exp Immunol 1993;94:64–67.

WBS-Gastro

432 LEE ET AL.

GASTROENTEROLOGY Vol. 111, No. 2

44. Sera HM, Crimi C, Sette A, Celis E. Fine restriction analysis and inhibition of antigen recognition in HLA-DQ–restricted T cells by major histocompatibility complex blockers and T cell receptor antagonists. Eur J Immunol 1993;23:2967–2991. 45. Apple RJ, Erlich HA, Klitz W, Manos MM, Becker TM, Wheeler CM. HLA DR-DQ associations with cervical carcinoma show papillomavirus-type specificity. Nat Genet 1994;6:157–162. 46. Mehal WZ, Lo W-MD, Herrington CS, Evans MF, Papodopoulos MC, Odunsi K, Ganesan TS, McGee JO’D, Bell JI, Fleming KA. Role of human papillomavirus in determining the HLA associated risk of cervical carcinogenesis. J Clin Pathol 1994;47:1077– 1081. 47. 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;87:427–436. 48. Witherspoon AC, Ortiz-Hidalgo C, Ralzon MR, Isaacson PG. Heli-

/ 5e10$$0003

07-02-96 18:03:03

gasa

cobacter pylori –associated gastritis and primary B-cell gastric lymphoma. Lancet 1991;338:1175–1176. 49. Bayerdo¨orffer E, Neubauer A, Rudolph B, Thiede C, Lehn N, Eidt S, Stolte M. Regression of primary gastric lymphoma of mucosaassociated lymphoid tissue type after cure of Helicobacter pylori infection. Lancet 1995;345:1591–1594. Received January 11, 1996. Accepted April 12, 1996. Address requests for reprints to: Jeffrey E. Lee, M.D., Department of Surgical Oncology, University of Texas M. D. Anderson Cancer Center, Box 106, 1515 Holcombe Boulevard, Houston, Texas 77030. Fax: (713) 792-0722. Supported by the University of Texas M. D. Anderson Cancer Center Core Grant CA16672 from the National Cancer Institute and a grant from the Gillson Longenbaugh Foundation. The authors thank Elizabeth A. Dove and Jeraldine Fojtik for their work on this manuscript.

WBS-Gastro