A PILOT STUDY OF HELICOBACTER PYLORI INFECTION AND RISK OF LARYNGOPHARYNGEAL CANCER

Transcription

1 A PILOT STUDY OF HELICOBACTER PYLORI INFECTION AND RISK OF LARYNGOPHARYNGEAL CANCER Zhannat Z. Nurgalieva, MD, 1,2 David Y. Graham, MD, 1,2 Kristina R. Dahlstrom, BS, 3 Qingyi Wei, MD, PhD, 4 Erich M. Sturgis, MD, MPH 3,4 1 Department of Gastroenterology Medicine, Baylor College of Medicine, One Baylor Plaza, Houston, Texas Veterans Administration Medical Center, 2002 Holcombe Boulevard, Houston, Texas Department of Head and Neck Surgery, Unit 441, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX esturgis@mdanderson.org 4 Department of Epidemiology, Unit 189, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, Texas Accepted 4 June 2004 Published online 30 September 2004 in Wiley InterScience ( DOI: /hed Abstract: Background. Squamous cell carcinoma of the laryngopharynx has been linked to laryngopharyngeal reflux disease. Helicobacter pylori corpus gastritis decreases gastric acid secretion and provides some protection against complications of gastroesophageal reflux, including adenocarcinoma of the distal esophagus. The aim of this study was to investigate whether H. pylori infection also protects against laryngopharyngeal carcinoma. Methods. This was a case-control study comparing patients with histologically confirmed, previously untreated laryngeal or pharyngeal squamous cell carcinomas with cancer-free controls selected from a pool of hospital-based cancer-free controls identified during a similar time period. Each subject completed a self-administered questionnaire that elicited information on age, Correspondence to: E. M. Sturgis Contract grant sponsor: SPORE Grant in Head and Neck Cancer [W. K. Hong (P.I.), E. M. Sturgis (Career Development Award)] National Institute of Environmental Health Sciences grant ES (Q. Wei); start-up funds from M. D. Anderson Cancer Center (E. M. Sturgis); K12 CA [R. C. Bast (P.I.), E. M. Sturgis (Faculty Trainee)] and P30 Ca (to M. D. Anderson Cancer Center) B 2004 Wiley Periodicals, Inc. sex, ethnicity, and tobacco and alcohol consumption. The 120 case subjects were frequency matched to 120 control subjects on age (F 5 years), sex, tobacco use, and alcohol use; all subjects were non-hispanic whites. H. pylori and human papillomavirus type 16 (HPV-16) seropositivity was determined by use of an enzyme-linked immunosorbent assay. Results. The serologic assay was unsuccessful in one case subject and nine control subjects; therefore, 119 case subjects and 111 control subjects were included in the analysis. The proportion of subjects with anti H. pylori immunoglobulin G was similar between the two groups (32.8% among cases vs 27.0% among controls; p =.342). Although seropositivity was more common in the patients with laryngeal cancer (39.1%) than in the patients with pharyngeal cancer (28.8%), this difference was neither significant (p =.241) nor associated with a significant risk of laryngeal cancer (adjusted odds ratio, 1.53; 95% confidence interval, ). H. pylori seropositivity was more common among patients who were HPV-16 seronegative (38.2% vs 22.9%; p =.081), and this was particularly true among patients with laryngeal cancer (47.1% vs 18.2%; p =.089). Conclusions. These results do not show that H. pylori infection either protects against or promotes laryngopharyngeal carcinoma. However, segregation analyses suggested that H. pylori may play a role in laryngeal cancers not associated with HPV-16 infection, and further study in this group is warranted. A 2004 Wiley Periodicals, Inc. Head Neck 27: 22 27, Helicobacter pylori and Laryngopharyngeal Cancer HEAD & NECK January 2005

2 Keywords: Helicobacter pylori; head and neck cancer; nonsmoker; laryngeal cancer; pharyngeal cancer the development of squamous cell carcinoma of the laryngopharynx. Helicobacter pylori is the etiologic agent of a transmissible, infectious disease associated with chronic gastritis, peptic ulcer disease, and gastric cancer. 1 4 The gastric inflammation associated with H. pylori infection may involve the gastric corpus and inhibit parietal cell function, resulting in a reduction of acid secretion. Thus, H. pylori infection has been associated with a reduced risk of gastroesophageal reflux disease and its complications such as Barrett s esophagus and adenocarcinoma of the esophagus. 5 Gastroesophageal reflux accompanied by the retrograde flow of gastric contents containing acid, bile, and pepsin might cause chronic irritations of organs other than esophagus. 6,7 Some have suggested that laryngopharyngeal reflux of gastric contents is associated with laryngopharyngeal cancer (LPCA) Tauber et al 11 suggested that not only does the proximal esophageal reflux of gastric contents directly contact and subsequently injure the laryngopharyngeal mucosa, but acid reflux inside the distal esophagus may also stimulate vagally mediated reflexes that contribute to the disorder. El-Serag et al 8 conducted a retrospective case-control study of 10,140 hospitalized patients and 12,061 outpatients with laryngeal and pharyngeal cancer and 40,560 hospitalized and 48,244 outpatient controls; the study was performed by use of computerized hospital and outpatient databases of the U. S. Department of Veterans Affairs. The major finding of that study was that the risk for laryngeal and pharyngeal cancers is modestly increased by gastroesophageal reflux disease, independent of age, sex, smoking, and alcohol consumption. Others have reported more anecdotal associations between laryngopharyngeal reflux and LPCA. 13,14 Acid-secreting parietal cells are found in the corpus of stomach, and H. pylori infection causes corpus gastritis that inhibits the function of those parietal cells and thus reduces acid secretion. Eradication of H. pylori infection is known to restore the acid secretion associated with such inflammation. 15 Thus, H. pylori infection could have an indirect protective effect against the development of LPCA by inhibiting gastric acid load. This study used a hospital-based casecontrol study design to test the hypothesis that H. pylori infection has a protective effect against MATERIALS AND METHODS Study Population. One hundred twenty patients with histologically confirmed, previously untreated squamous cell carcinoma of the laryngopharynx (ie, the glottic or supraglottic larynx and the oropharynx or hypopharynx) were recruited at our institution from May 1996 to June Cancer-free control subjects were selected from a pool of healthy controls identified during a similar period from an ongoing hospital-based casecontrol study of molecular epidemiology of squamous cell carcinoma of the head and neck. This control population was recruited from cancerfree visitors (friends and unrelated relatives of patients) to our institution. All subjects were non- Hispanic whites. A self-administered questionnaire regarding age, sex, ethnicity, and tobacco and alcohol consumption was completed by each enrolled individual. Ever smokers were defined as those who had smoked more than 100 cigarettes in their lifetimes, and ever drinkers were defined as those who had consumed alcoholic beverages at least once a week for more than 1 year. One hundred twenty case subjects were frequency matched on age (F 5 years), sex, and smoking and alcohol status to 120 control subjects. Subjects were not recruited with a known history of immune compromise (human immunodeficiency virus [HIV] positive, use of steroids in the previous month, or blood transfusion within the prior 6 months). H. pylori Serologic Testing. After informed consent, approved by our institutional review board, each volunteer provided a blood sample. The amount of immunoglobulin G (IgG) antibody to the high-molecular-weight cell-associated proteins of H. pylori was determined by use of a well-characterized enzyme-linked immunosorbent assay (HM-CAP IgG ELISA; Enteric Products, Inc., Stonybrook, NY) according to the manufacturer s instructions. This ELISA measures IgG antibodies to H. pylori infection. Test values were determined by comparison with a standard curve for each assay, which was plotted from negative and low-positive and high-positive controls. The manufacturer-recommended cut points are values <1.8 ELISA units (EU) for negative, 1.8 to 2.2 EU for indeterminate, and >2.2 EU for positive. The results were interpreted Helicobacter pylori and Laryngopharyngeal Cancer HEAD & NECK January

3 Table 1. Distribution of demographic and exposure variables among the cases and controls. Case subjects (N = 119) Control subjects (N = 111) Variable No. (%) No. (%) p value* Age, y.941 V > Sex.938 Male Female Smoking status.052 Never smokers Ever smokers Alcohol status.356 Never drinkers Ever drinkers according to the manufacturer s instructions. Human papillomavirus type-16 (HPV-16) serologic status was determined by use of another ELISA as previously described. 16 Statistical Analyses. The sample size was chosen to identify a 15% lower H. pylori seropositivity incidence in case subjects with a power of 0.8 (alpha, 0.05), assuming a cancer-free control subject seropositivity rate of 30%. Chi-square analyses were performed to determine the difference in the distribution of each demographic and exposure variable between case and control subjects. The odds ratios (ORs) and their 95% confidence intervals (CIs) for H. pylori seropositivity were calculated by logistic regression analyses with multivariate adjustment for confounders. Chi-square analyses were also performed to determine the difference in the distribution of seropositivity prevalence between tumor site, stage, grade, and HPV-16 serologic status. All the statistical analyses and tests were two-sided and were performed with Statistical Analysis System software (Version 8; SAS Institute Inc., Cary, NC) or STATA 7.0 software (Stata Corporation, College Station, TX). RESULTS One hundred nineteen case subjects and 111 control subjects were included in the analysis. One case subject and nine control subjects had indeterminate results when tested for presence of H. pylori specific IgG and were excluded from the analysis. The mean ages were 55.5 years for the case subjects (median, 55; range, 20 86) and 55.5 for the control subjects (median, 55; range, 25 82). There were no significant differences between the case and control subjects for age, sex, alcohol use, or smoking status (Table 1). However, there was a trend for cases being more likely current or former smokers than controls, suggesting that the matching was imperfect, so multivariate adjustment was used in later risk estimate analyses. Thirty-nine (32.8%) of the 119 case subjects and 30 (27.0%) of the 111 controls were seropositive for H. pylori, but the difference was not statistically significant (p =.342) (Table 2). After multivariate adjustment for age, sex, alcohol status, and smoking status, there was no statistical evidence that H. pylori seropositivity was protective for LPCA (adjusted OR, 1.27; 95% CI, ). When the cases were segregated by tumor site (larynx [n = 46] and pharynx [n = 73]), Table 2. H. pylori serologic status and risk estimates. Cases Controls H. pylori serologic status No. (%) No. (%) p value* Crude OR (95% CI) Adjustedy OR (95% CI) Total H. pylori negative 80 (67.2) 81 (73.0) H. pylori positive 39 (32.8) 30 (27.0) 1.32 ( ) 1.27 ( ) Pharyngeal cancer cases H. pylori negative 52 (71.2) 81 (73.0) H. pylori positive 21 (28.8) 30 (27.0) 1.09 ( ) 1.09 ( ) Laryngeal cancer cases H. pylori negative 28 (60.9) 81 (73.0) H. pylori positive 18 (39.1) 30 (27.0) 1.74 ( ) 1.53 ( ) yadjusted for age, sex, alcohol status, and smoking status. 24 Helicobacter pylori and Laryngopharyngeal Cancer HEAD & NECK January 2005

4 Table 3. Segregation analyses of H. pylori serologic status and risk estimates. Case subjects Control subjects Variable Total no. H. pylori +, no. (%) Total no. H. pylori +, no. (%) p value* Crude OR (95% CI) Age V (22.6) 56 7 (12.5) ( ) > (43.9) (41.8) ( ) Sex Male (28.8) (23.0) ( ) Female (37.7) (32.0) ( ) Smoking status Never smokers (25.5) (22.4) ( ) Ever smokers (37.5) (32.1) ( ) Alcohol status Never drinkers 33 8 (24.2) (29.7) ( ) Ever drinkers (36.0) (25.7) ( ) the pharyngeal cancer patients had an H. pylori seropositivity (28.8%) similar to that of the cancerfree controls (27.0%), whereas the laryngeal cancer patients had a higher rate (39.1%). However, this increased risk of laryngeal cancer was not significant (adjusted OR, 1.53; 95% CI, ) (Table 2). Further analyses showed that H. pylori seropositivity within demographic and exposure subgroups had no protective effect on risk of LPCA (Table 3). The proportion of H. pylori seropositivity was higher for older, female, and ever-smoking case and control subjects, but none of these differences were statistically significant (Table 3). Ever-drinking case subjects but not control subjects had higher H. pylori seropositivity rates than never-drinking case and control subjects, respectively (Table 3). To explore any impact of disease status on seropositivity rates, we performed subgroup analyses among the cases by tumor site, stage, or differentiation and found no statistically significant differences in the incidence of H. pylori seropositivity (Table 4). Of the patients with pharyngeal cancer, the three with hypopharyngeal tumors and the 70 with oropharyngeal tumors had similar H. pylori seropositivity rates (33.3% and 28.6%, respectively; p =.858). Of the patients with laryngeal cancers, 13 (43.3%) of 30 with supraglottic tumors were seropositive for H. pylori compared with five (31.3%) of 16 with glottic tumors ( p =.424). In addition to tobacco and alcohol, HPV-16 is perhaps the best-studied risk factor for oropharyngeal cancer, and HPV-16 seropositivity has been well studied as a marker for such risk. Consequently, we compared the H. pylori seropositivity rates among cases for those with and without evidence of prior HPV-16 infection. A borderline significant trend was found for an elevated H. pylori seropositivity rate in patients who were seronegative for HPV-16 ( p =.081) (Table 4). Among the patients with pharyngeal cancer, there was no difference in the frequency of H. pylori seropositivity for the 37 with and the 34 without HPV-16 seropositivity (24.3% Table 4. Segregation analyses of H. pylori serologic status by case factors. H. pylori + Variable Total no. No. (%) p value* Tumor site.241 Pharynx (28.8) Larynx (39.1) Tumor classification or (29.6) 3 or (39.5) Nodal status.932 Negative (33.3) Positive (32.5) Overall stage.386 I or II (40.0) III or IV (30.9) Tumor differentiationy.747 Well or moderate (27.9) Poorz (30.9) HPV-16 serologic status.081 Positive (22.9) Negative (38.2) ydifferentiation status not recorded for 21 patients. zincludes nine patients with moderately poor differentiated tumors. HPV-16 serologic status unavailable for three patients. Helicobacter pylori and Laryngopharyngeal Cancer HEAD & NECK January

5 and 29.4%, respectively; p =.629). Among the patients with laryngeal cancer, there was a borderline significant trend for more frequent H. pylori seropositivity in the 34 without HPV-16 seropositivity compared with the 11 with HPV- 16 seropositivity (47.1% vs 18.2%, respectively, p =.089). DISCUSSION H. pylori has been detected in saliva and dental plaques, 20 suggesting that H. pylori infection in the upper aerodigestive tract can disrupt the mucosal barriers, impair immune barriers, and permit direct contact of the laryngopharynx with known carcinogens such as tobacco and alcohol. 21 In contrast, because H. pylori can ameliorate gastroesophageal reflux disease, it is possible that H. pylori infection protects against development of LPCA, which has been associated with laryngopharyngeal reflux of gastric contents. 22,23 Three previous studies have investigated this question. A case-control study conducted in Ankara, Turkey, suggested an association between H. pylori infection and laryngeal cancer. 21 In a study of 26 patients with laryngeal carcinoma and 32 hospital-based cancer-free controls, H. pylori infection was found in 19 patients (73.1%) but only in 13 control subjects (40.6%). All subjects were male smokers, but the sample was small, and no adjustments for alcohol use or age were performed. In a previous case-control study from the United States, no association between H. pylori infection and squamous cell carcinoma of the oral cavity, oro/hypopharynx, or larynx was identified. 24 Although the cases and controls were matched on age and sex, the sample also was small (21 cases and 21 hospital-based cancerfree control subjects), and no control or adjustment for tobacco or alcohol use was performed. A group of investigators from the United Kingdom demonstrated an association between H. pylori infection and benign laryngeal disorders, 25 as well as premalignant and malignant head and neck neoplasms. 26 They detected serum anti H. pylori antibodies more often in the experimental group (63%; n = 61) than in the matched control group (40.7%; n = 187). 25,26 Our study is the largest case-control study to date examining the role of H. pylori infection in LPCA, and we included adjustment for potential confounders. This is also the first study to compare the site-specific (larynx vs pharynx) effects of H. pylori infection. We found no association between H. pylori and squamous cell carcinoma of the laryngopharynx. However, subgroup analysis (Table 2) showed that H. pylori infection was associated with a more than 1.5 times higher risk for laryngeal cancer compared with H. pylori negativity. The 95% CI for this association was not statistically significant but spans the range from a moderate decrease in risk to a more than threefold increased risk. Interestingly, we also found that H. pylori infection may be of greater importance in laryngeal cancers not associated with HPV-16 infection, but this needs to be substantiated in larger studies. Our findings contradict our initial hypothesis that H. pylori has a protective role in LPCA. However, others have suggested that alkaline reflux is involved in LPCA. Galli et al 12 found that six (28.5%) of 21 patients diagnosed with LPCA had evidence of alkaline reflux to the laryngopharynx. They also found LPCA or laryngopharyngeal premalignancy in six (15%) of 40 gastrectomized patients with evidence of biliary or alkaline reflux. 12 Furthermore, our study has several limitations. First, although the study sample size was chosen to detect a 15% lower incidence of H. pylori infection with a power of 0.8, it was underpowered to detect smaller differences in H. pylori seropositivity or to detect a higher seropositivity incidence in the cases, and our findings may have been due to chance. For example, to have detected a 12% higher incidence of H. pylori seropositivity in laryngeal cancer patients with a cancer-free control incidence of 27% and a power of 0.8 (alpha, 0.05), we would have needed 206 patients with laryngeal cancer and 206 cancer-free controls. Second, neither patients nor controls had ph probe documentation of or characterization of reflux or lack thereof. This is a potential confounder not controlled or adjusted for. In addition, despite not recruiting subjects with known immune compromise, it is possible that a group of patients was recruited with unknown antibody response problems and subsequently lower seropositivity rates. However, despite hypothesizing lower seropositivity rates would be found in patients, we found seropositivity rates in patients were similar to that of controls. Furthermore, we found no evidence of advanced disease affecting seropositivity rates (Table 4), and our rates of seropositivity to HPV-16 are consistent with reports in the literature Finally, the case-control study design does not allow temporal assessment as cohort studies do. It is possible that the de- 26 Helicobacter pylori and Laryngopharyngeal Cancer HEAD & NECK January 2005

6 velopment of laryngeal cancer creates a susceptible state for H. pylori infection, so that infection is opportunistic rather than carcinogenic. However, H. pylori seropositivity was not associated with pharyngeal cancer. Clearly, further studies of H. pylori s role in laryngeal cancer, with larger samples, objective characterization of laryngopharyngeal reflux, and serologic assessment of HPV-16 infection are needed. Acknowledgments. We thank Ms. Margaret Lung and Ms. Leonel A. Fairly for assistance with recruiting patients and database entry, Ms. Deanna Thomas for manuscript preparation, and Dr. Maureen E. Goode for manuscript editing. REFERENCES 1. Graham DY. Helicobacter pylori infection is the primary cause of gastric cancer. J Gastroenterol 2000;35(Suppl 12): Sepulveda AR, Graham DY. Role of Helicobacter pylori in gastric carcinogenesis. Gastroenterol Clin North Am 2002; 31: El-Omar EM, Oien K, Murray LS, et al. Increased prevalence of precancerous changes in relatives of gastric cancer patients: critical role of Helicobacter pylori. Gastroenterology 2000;118: Uemura N, Okamoto S, Yamamoto S, et al. Helicobacter pylori infection and the development of gastric cancer. N Engl J Med 2001;345: Graham DY, Yamaoka Y. H. pylori and caga: relationships with gastric cancer, duodenal ulcer, and reflux esophagitis and its complications. Helicobacter 1998;3: Olson NR. Aerodigestive malignancy and gastroesophageal reflux disease. Am J Med 1997;103:97S 99S. 7. Paterson WG. Extraesophageal complications of gastroesophageal reflux disease. Can J Gastroenterol 1997; 11(Suppl B):45B 50B. 8. El-Serag HB, Hepworth EJ, Lee P, Sonnenberg A. Gastroesophageal reflux disease is a risk factor for laryngeal and pharyngeal cancer. Am J Gastroenterol 2001;96: Biacabe B, Gleich LL, Laccourreye O, Hartl DM, Bouchoucha M, Brasnu D. Silent gastroesophageal reflux disease in patients with pharyngolaryngeal cancer: further results. Head Neck 1998;20: Ward PH, Hanson DG. Reflux as an etiological factor of carcinoma of the laryngopharynx. Laryngoscope 1988;98: Tauber S, Gross M, Issing WJ. Association of laryngopharyngeal symptoms with gastroesophageal reflux disease. Laryngoscope 2002;112: Galli J, Cammarota G, Calo L, et al. The role of acid and alkaline reflux in laryngeal squamous cell carcinoma. Laryngoscope 2002;112: Koufman JA. The otolaryngologic manifestations of gastroesophageal reflux disease (GERD): a clinical investigation of 225 patients using ambulatory 24-hour ph monitoring and an experimental investigation of the role of acid and pepsin in the development of laryngeal injury. Laryngoscope 1991;101(4 Pt 2 Suppl 53): Copper MP, Smit CF, Stanojcic LD, Devriese PP, Schouwenburg PF, Mathus-Vliegen LM. High incidence of laryngopharyngeal reflux in patients with head and neck cancer. Laryngoscope 2000;110: Graham DY, Dixon MF. Acid secretion, Helicobacter pylori infection and peptic ulcer disease. In: Graham DY, Genta RM, Dixon MF, editors. Gastritis. Philadelphia: Lippincott Williams and Wilkins; 1999, p Dahlstrom KR, Adler-Storthz K, Etzel CJ, et al. Human papillomavirus type 16 infection and squamous cell carcinoma of the head and neck in never smokers: a matched pair analysis. Clin Cancer Res 2003;9: Schwartz SM, Daling JR, Doody DR, et al. Oral cancer risk in relation to sexual history and evidence of HPV infection. J Natl Cancer Inst 1998;90: Mork J, Lie AK, Glattre E, et al. Human papillomavirus infection as a risk factor for squamous-cell carcinoma of the head and neck. N Engl J Med 2001;344: Herrero R, Castellsague X, Pawlita M, et al. Human papillomavirus and oral cancer: The International Agency for Research on Cancer Multicenter Study. J Natl Cancer Inst 2003;95: Graham DY. Public health issues relating to Helicobacter pylori infection and global eradication. In: Graham DY, Genta RM, Dixon MF, editors. Gastritis. Philadelphia: Lippincott Williams and Wilkins; p Aygenc E, Selcuk A, Celikkanat S, Ozbek C, Ozdem C. The role of Helicobacter pylori infection in the cause of squamous cell carcinoma of the larynx. Otolaryngol Head Neck Surg 2001;125: Koufman J, Sataloff RT, Toohill R. Laryngopharyngeal reflux: consensus conference report. J Voice 1996;10: Freije JE, Beatty TW, Campbell BH, Woodson BT, Schultz CJ, Toohill RJ. Carcinoma of the larynx in patients with gastroesophageal reflux. Am J Otolaryngol 1996;17: Grandis JR, Perez-Perez GI, Yu VL, Johnson JT, Blaser MJ. Lack of serologic evidence for Helicobacter pylori infection in head and neck cancer. Head Neck 1997;19: Rubin JS, Benjamin E, Prior A, Lavy J, Ratcliffe P. The prevalence of Helicobacter pylori infection in benign laryngeal disorders. J Voice 2002;16: Rubin JS, Benjamin E, Prior A, Lavy J. The prevalence of Helicobacter pylori infection in malignant and premalignant conditions of the head and neck. J Laryngol Otol 2003;117: Helicobacter pylori and Laryngopharyngeal Cancer HEAD & NECK January