Scandinavian Journal of Gastroenterology. 2014; 49: 35–42
ORIGINAL ARTICLE
Scand J Gastroenterol Downloaded from informahealthcare.com by Library of Health Sci-Univ of Il on 12/09/14 For personal use only.
Helicobacter pylori infection is associated with advanced colorectal neoplasia
HAIM SHMUELY1*, EHUD MELZER2*, MICHAL BRAVERMAN2, NOAM DOMNIZ1 & JACOB YAHAV2 1
Department of Internal Medicine D, Kaplan Medical Center, Rehovot and the Faculty of Medicine, Hebrew University, Jerusalem, Israel, and 2Department of Gastroenterology, Kaplan Medical Center, Rehovot and the Faculty of Medicine, Hebrew University, Jerusalem, Israel
Abstract Objective. The aim of this article was to evaluate the prevalence of Helicobacter pylori infection in patients diagnosed with advanced colorectal neoplasia undergoing a colonoscopy compared to patients without neoplasia. Material and methods. This cross-sectional study investigated the association of neoplastic lesions diagnosed on colonoscopy with H. pylori infection in a consecutive series of subjects who had undergone a pancolonoscopy in a single academic medical center. All patients were tested by ELISA and the immunoblot technique for serum anti-H. pylori and CagA protein IgG antibodies. Multivariate analyses were adjusted for potential-relevant confounders, including age, sex, smoking, childhood socioeconomic status, and family history of colorectal cancer. Results. Two hundred and seventy-three patients were included in the study: 75% (84/ 112), diagnosed with neoplastic colorectal lesions and 48% (77/161) without neoplastic lesions, were found to be seropositive for H. pylori infection (p < 0.001). H. pylori infection was found in 66/77 (86 %) patients with advanced neoplasia, 18/35 (51%) patients with nonadvanced neoplasia, and 48% (77/161) patients without neoplasia (p < 0.001). In the adjusted analysis, H. pylori infection was found to be associated with advanced colorectal neoplasia (odds ratio, OR 9.57; 95% CI 4.31–21.2; p < 0.001) and CRC (OR 7.98;95% CI 3.16–20.16; p < 0.001). There was no association in patients who were CagA positive. Conclusion. H. pylori infection is associated with the development of advanced colorectal neoplasia. More studies are needed to confirm our findings.
Key Words: advanced neoplasia, colorectal cancer, colorectal neoplasia, Helicobacter pylori
Introduction Colorectal cancer (CRC) is the third most common cancer in men and the second most common in women worldwide. Annually, nearly 608,000 CRC deaths occur, accounting for 8% of all cancer deaths. CRC has become the fourth most common cause of cancer death [1,2]; however, its etiology is still uncertain. CRC has been shown to be strongly associated with certain hereditary gene mutations, but only 3– 5% of CRCs are due to these mutations alone [3,4]. CRC is a major cause of morbidity and mortality in Western countries resulting in a substantial economic
burden to the population due to surgery, chemotherapy and terminal care costs [5]. Helicobacter pylori, a Gram-negative bacteria, has been found to colonize stomachs of >50% of the world’s population [6]. H. pylori is acquired mostly during childhood [7]. H. pylori, a carcinogen type I for gastric cancer increases by two- to six-fold the risk of gastric cancer compared to the normal noninfected population. Epidemiological studies have demonstrated a greater prevalence of colorectal adenomas and/or adenocarcinomas in patients infected with H. pylori [8,9]. However, this association is controversial in
Correspondence: Haim Shmuely, MD, Department of Internal Medicine D, Kaplan Medical Center, Rehovot and the Faculty of Medicine, Hebrew University, Pasternak Street, POB 1, Rehovot 76100, Jerusalem, Israel. Tel: +972 8 944 1996. Fax: +972 8 944 1866; E-mail: [email protected] *Haim Shmuely and Ehud Melzer contributed equally as first authors.
(Received 28 May 2013; revised 20 September 2013; accepted 22 September 2013) ISSN 0036-5521 print/ISSN 1502-7708 online 2014 Informa Healthcare DOI: 10.3109/00365521.2013.848468
Scand J Gastroenterol Downloaded from informahealthcare.com by Library of Health Sci-Univ of Il on 12/09/14 For personal use only.
36
H. Shmuely et al.
view of the fact that other researchers have not found an association between adenomas and/or colon adenocarcinomas and H. pylori infection [10,11]. Recently, our group demonstrated that H. pylori CagA infection is associated with CRC compared with the normal population and CagA negative patients [9]. Another Israeli study reported a high prevalence of anti-H. pylori antibodies in CRC patients [12]. In two recent meta-analyses published in 2006 and 2012, the association between H. pylori infection and CRC risk were evaluated and the odds ratios (OR) were 1.4 and 1.6, respectively [13,14]. The important indicators for progression from adenoma to CRC are the pathological characteristics of the adenoma, such as larger size (at least 10 mm in diameter), villous (as opposed to tubular) histology and high-grade (as opposed to low-grade) dysplasia [15]. Advanced adenomas are defined as adenomas with significant villous features (>25%) at least 10 mm in diameter, high-grade dysplasia (HGD), or early invasive cancer [16]. Advanced adenomas assessed by colonoscopy are a more desirable target for screening efficacy than the more uncommon but life-threatening cancer stage or the more common but early, less significant nonadvanced adenoma stage [17]. The objective of this article was to assess the association of advanced colorectal neoplasia defined as cancer or advanced adenoma with H. pylori infection in consecutive patients referred for a colonoscopy in a single university medical center.
Questionnaire Prior to the colonoscopy, subjects were interviewed by a trained staff member using a validated structured questionnaire comprising demographic data and socioeconomic status (current and childhood). Sociodemographic data included age, sex, ethnic origin, place of birth (self, mother), occupation (manual/ nonmanual, other), level of education (higher/lower than eighth grade), crowding in childhood (number of siblings per room in the house – less or more than 1), cigarette smoking (ever/never), history of peptic ulcer, heartburn, history of colonic or gastric cancer, family history of gastric or colonic cancer, constipation, bloody stools, changes in bowel movements, previous colonoscopy, and known colonic polyps or CRC. Serology Specimens of venous blood were obtained on the same day as the colonoscopy. Serum was initially stored at –70 C until tested for H. pylori infection by ELISA (Orion Diagnostica, Espoo, Finland). This method was validated in our laboratory by a pilot study involving patients who had undergone an endoscopy at our hospital, yielding a sensitivity of 94%, and specificity of 90%. Positive and negative predictive values of 100% and 90%, respectively, were observed. IgG antibodies against CagA protein were tested with immunoblot staining using a Western blot kit (Helicoblot 2.0; Genelabs Diagnostic, Singapore). The determinations of CagA seropositivity were in accordance with the manufacturer’s criteria [18,19].
Methods Colonoscopy Participants The study was approved by the Institutional Review Board of the Kaplan Medical Center (KMC). All participants were assigned a study number. Information was transferred to data collection forms to insure confidentiality. Twice weekly, from August 1, 2008 through November 30, 2010, all consecutive subjects aged 25 years or older who presented for an outpatient colonoscopy were asked to participate in the study. The most common indications were rectal bleeding, colonic polyps, and changes in bowel habits. Colonoscopies were performed 5 days a week with no specific preference or predilection of subjects on recruitment days. Those who agreed to participate gave informed consent and were subsequently included into the study cohort. Height and weight were measured by trained nurses. Body mass index (BMI) was calculated as weight in kilograms divided by the square of the height in meters.
Subjects were pretreated with either polyethylene glycol-electrolyte solution or sodium picosulphate and sedated with intravenous dolestine and midazolam just prior to the endoscopy. All subjects signed an informed consent form. All included subjects underwent a pancolonoscopy (CF-Q165L, Olympus Optical Co., Tokyo, Japan). Identification of polypoid lesions by colonoscopy was followed by a polypectomy. Unresectable lesions were biopsied. Specimens were fixated in a 4% formaldehyde solution, embedded in paraffin and stained with hematoxilin-eosin (H&E). Specimens were reviewed by two experienced KMC pathologists. Adenomas were classified as tubular, tubulovillous or villous, each with either low-grade dysplasia (LGD), HGD, or malignancy (CRC). Location and size of all lesions were indicated. If the colonoscopic examination was incomplete, the patient was asked to return for a second attempt.
Scand J Gastroenterol Downloaded from informahealthcare.com by Library of Health Sci-Univ of Il on 12/09/14 For personal use only.
H. pylori & advanced colorectal neoplasms If a complete examination was performed within 6 months following the first attempt, a complete examination was reported and the combined results of the two examinations were included in the analysis. If the second examination was also incomplete, the results were excluded from the analysis. Since H. pylori infection is acquired mostly during childhood and adolescence [7] with almost no new infections in adulthood [20], we searched the participants’ medical files for previous colonoscopies and included positive findings of neoplastic lesions instead of the present negative findings, since colonoscopy after a previous polypectomy can be negative. Thus, using only present colonoscopy findings without taking into account previous findings, can lead to underestimation of the association. The age of the patients was adjusted accordingly. If no neoplastic lesions were diagnosed on previous colonoscopies, we included the results of the present colonoscopy in order to include the longest period of time from the acquisition of H. pylori. Definitions The findings on colonoscopy were categorized on the basis of the most advanced lesion found [21]. Advanced adenoma was defined as an adenoma of at least 10 mm in diameter, high-grade dysplasia, villous or tubulovillous histologic characteristics, or any combination thereof [22]. Cancer was defined as the invasion of malignant cells beyond the muscularis mucosa. Advanced neoplasia was defined as cancer or advanced adenoma. Nonadvanced colorectal neoplasia was defined as lesions with tubular histology 0.2 on univariate analysis were included in the multivariate analysis. Logistic regression was used for this purpose. OR with a 95% confidence interval (CI) were reported. Statistical analysis was performed using the SPSS version 19. p-Value < 0.05 was considered statistically significant. Results Of the 273 subjects who had undergone a colonoscopy, 139 (51%) were men and 134 (49%) women. The median age was 64.5 (IQR, 74-55). No neoplastic lesions were found in 161 patients (59%); 112 subjects (41%) had neoplastic lesions. Fifty-two percent of patients with neoplastic lesions (58/112) underwent previous colonoscopies: 42/77(55%) with advanced neoplasia, and 16/35(46 %) with nonadvanced neoplasia. Table I contains a stratification of participants by absence or presence of colorectal neoplasms on colonoscopy. Participants diagnosed with neoplastic colorectal lesions were older and included more men than those without neoplasia (p = 0.012, p = 0.014, accordingly). H. pylori infection was significantly more frequent among patients with neoplasms than without (p < 0.001). The two groups did not differ in BMI, smoking history, ethnicity, socioeconomic status and crowding index at childhood. In the crude analyses, CagA seropositivity in patients with
38
H. Shmuely et al.
Table I. Demographic, socioeconomic, H. pylori status and presence or absence of colorectal neoplasms found on colonoscopy in all study groups. Colonoscopy findings
Scand J Gastroenterol Downloaded from informahealthcare.com by Library of Health Sci-Univ of Il on 12/09/14 For personal use only.
No Neoplasia Helicobacter pylori, n (%) Male, n (%) Age (years), mean (SD) BMI, mean (SD) Family history of GI neoplasm, n (%) Smoker, n (%) Current Past Origin, n (%) Ashkenazi Sephardic Occupation father, n (%) Manual worker Clerk Education father
ORIGINAL ARTICLE
Scand J Gastroenterol Downloaded from informahealthcare.com by Library of Health Sci-Univ of Il on 12/09/14 For personal use only.
Helicobacter pylori infection is associated with advanced colorectal neoplasia
HAIM SHMUELY1*, EHUD MELZER2*, MICHAL BRAVERMAN2, NOAM DOMNIZ1 & JACOB YAHAV2 1
Department of Internal Medicine D, Kaplan Medical Center, Rehovot and the Faculty of Medicine, Hebrew University, Jerusalem, Israel, and 2Department of Gastroenterology, Kaplan Medical Center, Rehovot and the Faculty of Medicine, Hebrew University, Jerusalem, Israel
Abstract Objective. The aim of this article was to evaluate the prevalence of Helicobacter pylori infection in patients diagnosed with advanced colorectal neoplasia undergoing a colonoscopy compared to patients without neoplasia. Material and methods. This cross-sectional study investigated the association of neoplastic lesions diagnosed on colonoscopy with H. pylori infection in a consecutive series of subjects who had undergone a pancolonoscopy in a single academic medical center. All patients were tested by ELISA and the immunoblot technique for serum anti-H. pylori and CagA protein IgG antibodies. Multivariate analyses were adjusted for potential-relevant confounders, including age, sex, smoking, childhood socioeconomic status, and family history of colorectal cancer. Results. Two hundred and seventy-three patients were included in the study: 75% (84/ 112), diagnosed with neoplastic colorectal lesions and 48% (77/161) without neoplastic lesions, were found to be seropositive for H. pylori infection (p < 0.001). H. pylori infection was found in 66/77 (86 %) patients with advanced neoplasia, 18/35 (51%) patients with nonadvanced neoplasia, and 48% (77/161) patients without neoplasia (p < 0.001). In the adjusted analysis, H. pylori infection was found to be associated with advanced colorectal neoplasia (odds ratio, OR 9.57; 95% CI 4.31–21.2; p < 0.001) and CRC (OR 7.98;95% CI 3.16–20.16; p < 0.001). There was no association in patients who were CagA positive. Conclusion. H. pylori infection is associated with the development of advanced colorectal neoplasia. More studies are needed to confirm our findings.
Key Words: advanced neoplasia, colorectal cancer, colorectal neoplasia, Helicobacter pylori
Introduction Colorectal cancer (CRC) is the third most common cancer in men and the second most common in women worldwide. Annually, nearly 608,000 CRC deaths occur, accounting for 8% of all cancer deaths. CRC has become the fourth most common cause of cancer death [1,2]; however, its etiology is still uncertain. CRC has been shown to be strongly associated with certain hereditary gene mutations, but only 3– 5% of CRCs are due to these mutations alone [3,4]. CRC is a major cause of morbidity and mortality in Western countries resulting in a substantial economic
burden to the population due to surgery, chemotherapy and terminal care costs [5]. Helicobacter pylori, a Gram-negative bacteria, has been found to colonize stomachs of >50% of the world’s population [6]. H. pylori is acquired mostly during childhood [7]. H. pylori, a carcinogen type I for gastric cancer increases by two- to six-fold the risk of gastric cancer compared to the normal noninfected population. Epidemiological studies have demonstrated a greater prevalence of colorectal adenomas and/or adenocarcinomas in patients infected with H. pylori [8,9]. However, this association is controversial in
Correspondence: Haim Shmuely, MD, Department of Internal Medicine D, Kaplan Medical Center, Rehovot and the Faculty of Medicine, Hebrew University, Pasternak Street, POB 1, Rehovot 76100, Jerusalem, Israel. Tel: +972 8 944 1996. Fax: +972 8 944 1866; E-mail: [email protected] *Haim Shmuely and Ehud Melzer contributed equally as first authors.
(Received 28 May 2013; revised 20 September 2013; accepted 22 September 2013) ISSN 0036-5521 print/ISSN 1502-7708 online 2014 Informa Healthcare DOI: 10.3109/00365521.2013.848468
Scand J Gastroenterol Downloaded from informahealthcare.com by Library of Health Sci-Univ of Il on 12/09/14 For personal use only.
36
H. Shmuely et al.
view of the fact that other researchers have not found an association between adenomas and/or colon adenocarcinomas and H. pylori infection [10,11]. Recently, our group demonstrated that H. pylori CagA infection is associated with CRC compared with the normal population and CagA negative patients [9]. Another Israeli study reported a high prevalence of anti-H. pylori antibodies in CRC patients [12]. In two recent meta-analyses published in 2006 and 2012, the association between H. pylori infection and CRC risk were evaluated and the odds ratios (OR) were 1.4 and 1.6, respectively [13,14]. The important indicators for progression from adenoma to CRC are the pathological characteristics of the adenoma, such as larger size (at least 10 mm in diameter), villous (as opposed to tubular) histology and high-grade (as opposed to low-grade) dysplasia [15]. Advanced adenomas are defined as adenomas with significant villous features (>25%) at least 10 mm in diameter, high-grade dysplasia (HGD), or early invasive cancer [16]. Advanced adenomas assessed by colonoscopy are a more desirable target for screening efficacy than the more uncommon but life-threatening cancer stage or the more common but early, less significant nonadvanced adenoma stage [17]. The objective of this article was to assess the association of advanced colorectal neoplasia defined as cancer or advanced adenoma with H. pylori infection in consecutive patients referred for a colonoscopy in a single university medical center.
Questionnaire Prior to the colonoscopy, subjects were interviewed by a trained staff member using a validated structured questionnaire comprising demographic data and socioeconomic status (current and childhood). Sociodemographic data included age, sex, ethnic origin, place of birth (self, mother), occupation (manual/ nonmanual, other), level of education (higher/lower than eighth grade), crowding in childhood (number of siblings per room in the house – less or more than 1), cigarette smoking (ever/never), history of peptic ulcer, heartburn, history of colonic or gastric cancer, family history of gastric or colonic cancer, constipation, bloody stools, changes in bowel movements, previous colonoscopy, and known colonic polyps or CRC. Serology Specimens of venous blood were obtained on the same day as the colonoscopy. Serum was initially stored at –70 C until tested for H. pylori infection by ELISA (Orion Diagnostica, Espoo, Finland). This method was validated in our laboratory by a pilot study involving patients who had undergone an endoscopy at our hospital, yielding a sensitivity of 94%, and specificity of 90%. Positive and negative predictive values of 100% and 90%, respectively, were observed. IgG antibodies against CagA protein were tested with immunoblot staining using a Western blot kit (Helicoblot 2.0; Genelabs Diagnostic, Singapore). The determinations of CagA seropositivity were in accordance with the manufacturer’s criteria [18,19].
Methods Colonoscopy Participants The study was approved by the Institutional Review Board of the Kaplan Medical Center (KMC). All participants were assigned a study number. Information was transferred to data collection forms to insure confidentiality. Twice weekly, from August 1, 2008 through November 30, 2010, all consecutive subjects aged 25 years or older who presented for an outpatient colonoscopy were asked to participate in the study. The most common indications were rectal bleeding, colonic polyps, and changes in bowel habits. Colonoscopies were performed 5 days a week with no specific preference or predilection of subjects on recruitment days. Those who agreed to participate gave informed consent and were subsequently included into the study cohort. Height and weight were measured by trained nurses. Body mass index (BMI) was calculated as weight in kilograms divided by the square of the height in meters.
Subjects were pretreated with either polyethylene glycol-electrolyte solution or sodium picosulphate and sedated with intravenous dolestine and midazolam just prior to the endoscopy. All subjects signed an informed consent form. All included subjects underwent a pancolonoscopy (CF-Q165L, Olympus Optical Co., Tokyo, Japan). Identification of polypoid lesions by colonoscopy was followed by a polypectomy. Unresectable lesions were biopsied. Specimens were fixated in a 4% formaldehyde solution, embedded in paraffin and stained with hematoxilin-eosin (H&E). Specimens were reviewed by two experienced KMC pathologists. Adenomas were classified as tubular, tubulovillous or villous, each with either low-grade dysplasia (LGD), HGD, or malignancy (CRC). Location and size of all lesions were indicated. If the colonoscopic examination was incomplete, the patient was asked to return for a second attempt.
Scand J Gastroenterol Downloaded from informahealthcare.com by Library of Health Sci-Univ of Il on 12/09/14 For personal use only.
H. pylori & advanced colorectal neoplasms If a complete examination was performed within 6 months following the first attempt, a complete examination was reported and the combined results of the two examinations were included in the analysis. If the second examination was also incomplete, the results were excluded from the analysis. Since H. pylori infection is acquired mostly during childhood and adolescence [7] with almost no new infections in adulthood [20], we searched the participants’ medical files for previous colonoscopies and included positive findings of neoplastic lesions instead of the present negative findings, since colonoscopy after a previous polypectomy can be negative. Thus, using only present colonoscopy findings without taking into account previous findings, can lead to underestimation of the association. The age of the patients was adjusted accordingly. If no neoplastic lesions were diagnosed on previous colonoscopies, we included the results of the present colonoscopy in order to include the longest period of time from the acquisition of H. pylori. Definitions The findings on colonoscopy were categorized on the basis of the most advanced lesion found [21]. Advanced adenoma was defined as an adenoma of at least 10 mm in diameter, high-grade dysplasia, villous or tubulovillous histologic characteristics, or any combination thereof [22]. Cancer was defined as the invasion of malignant cells beyond the muscularis mucosa. Advanced neoplasia was defined as cancer or advanced adenoma. Nonadvanced colorectal neoplasia was defined as lesions with tubular histology 0.2 on univariate analysis were included in the multivariate analysis. Logistic regression was used for this purpose. OR with a 95% confidence interval (CI) were reported. Statistical analysis was performed using the SPSS version 19. p-Value < 0.05 was considered statistically significant. Results Of the 273 subjects who had undergone a colonoscopy, 139 (51%) were men and 134 (49%) women. The median age was 64.5 (IQR, 74-55). No neoplastic lesions were found in 161 patients (59%); 112 subjects (41%) had neoplastic lesions. Fifty-two percent of patients with neoplastic lesions (58/112) underwent previous colonoscopies: 42/77(55%) with advanced neoplasia, and 16/35(46 %) with nonadvanced neoplasia. Table I contains a stratification of participants by absence or presence of colorectal neoplasms on colonoscopy. Participants diagnosed with neoplastic colorectal lesions were older and included more men than those without neoplasia (p = 0.012, p = 0.014, accordingly). H. pylori infection was significantly more frequent among patients with neoplasms than without (p < 0.001). The two groups did not differ in BMI, smoking history, ethnicity, socioeconomic status and crowding index at childhood. In the crude analyses, CagA seropositivity in patients with
38
H. Shmuely et al.
Table I. Demographic, socioeconomic, H. pylori status and presence or absence of colorectal neoplasms found on colonoscopy in all study groups. Colonoscopy findings
Scand J Gastroenterol Downloaded from informahealthcare.com by Library of Health Sci-Univ of Il on 12/09/14 For personal use only.
No Neoplasia Helicobacter pylori, n (%) Male, n (%) Age (years), mean (SD) BMI, mean (SD) Family history of GI neoplasm, n (%) Smoker, n (%) Current Past Origin, n (%) Ashkenazi Sephardic Occupation father, n (%) Manual worker Clerk Education father
