Is there an explanation for the predominance of Barrett's esophagus and adenocarcinoma in white male patients?
J. Sarosiek, R.W. McCallum (Kansas City)
This question provides one of the greatest challenges in the pathophysiology of the alimentary tract as it relates to gastroesophageal reflux disease (GERD), the most prevalent malaise in the United States and the Western World.
The issue of the role of race and gender in promotion or prevention of gastroesophageal reflux (GER), the development of GERD and its major complications such as Barrett's esophagus (BE) or esophageal adenocarcinoma (AdCa), and in the therapeutic response to various modalities has been greatly overlooked.
There is uncontrovertible evidence that there are racial and gender-related differences within the realm of physiology and pathophysiology of the alimentary tract. Therefore, these topics also require adequate attention regarding the contributory role of race and gender in the GERD, reflux esophagitis (RE), BE and AdCa sequence.
Although the gender ratio may vary in many publications some approximate values are presented in Figure 1. These data indicate that males among the United States population exhibit a preponderance to the development GERD, RE, BE and AdCa and that gender-related factors are constantly acting along this pathogenetic sequence. It is still uncertain if lower esophageal sphincter (LES) basal tone or its rate of relaxation is not different between healthy males and females.
One may not exclude the possibility that males with GERD may exhibit a tendency to compromised LES competence or diminished esophageal acid clearance. The same questions may apply to the issue of the barrier function values of crural diaphragm or gastric emptying. Even if there is no difference between males and females in all mentioned above parameters, the difference could still exist in esophageal mucosal protective mechanisms which are phylogenetically and ontogenetically developed in order to balance aggressive factors of both physiologic and excessive GER.
Figure 1. Gender ratio in a general population of patients with gastroesophageal reflux disease (GERD) in the United States; patients who subsequently develop endoscopic reflux esophagitis (RE) or who never exhibit erosive mucosal changes; in a subpopulation of patients with endoscopic RE who subsequently develop Barrett's esophagus (BE) and RE patients who never exhibit intestinalization of their esophageal mucosa; and a subpopulation of BE patients who develop esophageal adenocarcinoma (AdCa) in the setting of high grade dysplastic (HGD) changes or those BE patients who never develop esophageal AdCa. M = Males; F = Females; NCBE = Non-complicated BE.
The role of gender-related factors in the development of BE and AdCa
In our recent studies we have demonstrated that patients with endoscopic RE exhibit impairment in the rate of secretion of salivary and esophageal epidermal growth factor (EGF) and that this impairment persists after healing of endoscopic changes [1-3]. Patients with endoscopic RE also exhibit impairment in secretion of esophageal glycoconjugates and PGE2 under the impact of luminal HCl/pepsin . All these factors contribute to mucosal protection and to balancing aggressive power of GER. We have demonstrated in our preliminary studies that females have stronger protective mechanisms than males which are pivotal in the maintenance of the integrity of the esophageal mucosa . In addition, the esophageal mucosa in patients with RE and BE exhibits a distinct secretory pattern as compared to the squamous epithelium in patients with RE not accompanied by intestinal metaplasia . These data strongly suggest that although GER is prerequisite of the development of the esophageal mucosal pathology, gender-related quality of mucosal protective mechanisms are significant and could be decisive in determining the development of end point complications.
The role of race-related factors in the development of BE and AdCa
In the United States, the population of patients who develop BE and adenocarcinoma is dominated by Caucasians (C). The ratio between C and African American (AA) males among patients with AdCa is 3.25 and between C and AA females reaches value of 4.0 . Similar trends are reported among the population diagnosed with BE .
Although some factors increasing susceptibility of C population to the development of BE in the setting of GERD and endoscopic RE could be related to the competence of LES and motility-related esophageal acid clearance mechanisms we have recently demonstrated racial differences regarding strength of protective mechanisms operating within the esophageal pre-epithelial barrier . The study was conducted in 11 asymptomatic AA (6 F and 5 M, mean age of 37, range 22-62) and in 20 asymptomatic C volunteers (11 F and 9 M, mean age of 38, range 26-56). Salivary secretions were collected under basal conditions, during mastication, and during intraesophageal mechanical and chemical stimulation. Intraesophageal stimulation was accomplished with an esophageal catheter (Wilson-Cook Med. Inc., NC), which mimics the gastroesophageal reflux scenario. Salivary volume was measured using a sialometer. Salivary bicarbonate and non-bicarbonate were measured by back-titration with TitraLab TIM900 and glycoconjugate (predominantly mucin) by PAS method. Salivary volume was similar in both studied populations. Salivary glycoconjugate output in AA was 2.3-fold and significantly higher than in C (1.09 ± 0.20 vs 0.48 ± 0.13 mg/min, p < 0.05). Salivary glycoconjugate output was also significantly higher (by 79%) in AA during masticatory stimulation (1.79 ± 0.23 vs 1.00 ± 0.23 mg/min, p < 0.05). A similar trend was maintained during stimulation of salivary glycoconjugate with intraesophageal mechanical and chemical stimulation. These data indicate that a significantly higher rate of salivary glycoconjugate (mucin) secretion in basal and stimulated conditions in AA population could be a factor preventing the esophageal injury and complications during GER. Of note, prevalence of hiatal hernia is similar in males and females and in both C and AA populations .
The role of race- and gender-related factors
in the development of squamous cancer
Among the United States population, patients with squamous cell carcinoma are predominantly represented by AA . Therefore, the ratio between AA and C males is 5.6 and between AA and C females 3.8 . Of note, AA males are 3.7 times more likely to suffer from squamous cell carcinoma than females . It is hard to imagine that environmental risk factors for esophageal squamous cell carcinoma are predominantly present among AA population and selectively affect males. Therefore, it seems unlikely that variation in the incidence of esophageal cancer is due solely to variations in the exposure to environmental carcinogens . Individual susceptibility to cancerogenic impact of environmental factors seems to be genetically determined .
Presented data strongly imply that race- and gender-related genetic factors and presumably chromosome Y linked genes are important contributors to the development of esophageal cancers adenocarcinoma or squamous cell cancer. That these factors are representing some still not identified protooncogenes activated by point mutation or tumor suppresser genes inactivated through mutagenicity still remains to be determined. A familial tendency to develop squamous esophageal carcinoma has been observed in some families in Northern China and Northwestern Iran .
Oral cavity, pharynx, larynx and lung cancers, although varied in incidence worldwide, invariably predominate in males [12, 13]. However, in the areas of high esophageal cancer prevalence other cancers such as oral cavity, larynx, lung or stomach are not always more common . This may indicate that esophageal cancerogenesis exhibits its unique pathway combining not only a number of environmental factors but also presumably genetic predisposition.
Esophageal cancer develops in two histologically distinct settings: squamous epithelium giving rise to squamous cancer or columnar epithelium leading to adenocarcinoma. The former may occur throughout the length of the esophagus, whereas, later mostly in the lower third of esophagus or at the gastroesophageal junction .
Gene expression and genetic changes in esophageal cancers
Almost 20 carcinomas of the esophagus with abnormal karyotypes have been reported . In the allotype study of 93 subjects the esophageal squamous cell carcinoma loss of heterozygosity (LOH) at frequencies of at least 30% was observed within chromosome arms 3p&q, 5q, 9p&q, 10p, 13q, 17p&q, 18q, 19q, and 21q . These sites are likely to encode tumor suppressor genes which may play a pivotal role in esophageal canceroprotection . Although some investigators revealed loss of the Y chromosome in some esophageal cancers, the relevance to tumorigenesis of such a solitary sex chromosome aberrations is still debatable .
Esophageal cancer has a remarkable pattern of incidence. In certain restricted localities like Kazakhstan and Northern Iran, the incidence of esophageal cancer is the highest observed in the general population for any type of cancer anywhere in the world comparable to incidence of lung cancer among life-long smokers in London . Also, a sharp decline of frequency occurs between Gurjev in Kazakhstan and regions of Georgia (only a 500 hundred miles apart) 70-fold for men and 230-fold for women, strongly implying the contributing role of genetic factors . Unfortunately, these data do not discriminate esophageal squamous carcinoma from adenocarcinoma, however, in this geographic area squamous carcinoma is predominant .
The incidence of esophageal squamous carcinoma seems to remain steady in the United States and Europe in the last 10-20 years, however, adenocarcinoma is on a rise [7, 10]. Although the role of smoking, alcohol, nutritional deficiency (low fruit and vegetable diet lacking vitamin A, C, riboflavin, and zinc) or carcinogenic factors (substituted hydroxyphenantrenes, mycotoxins or N-nitroso compounds) lacking diet in the development of esophageal squamous carcinoma, little is known regarding predisposing factors for the development of esophageal adenocarcinoma [7, 10].
In patients with BE, an increase in expression of TGFa or EGF/TGFa receptor has been demonstrated . This increase was parallel to progression from nonspecialized to specialized intestinal epithelium and was even more profound in epithelium exhibiting dysplasia and adenocarcinoma . No such dynamics in EGF expression was observed in the same specimens. This indicates that metaplastic epithelium within Barrett's mucosa produces its own mitosis stimulating molecules which may lead to uncontrolled growth in an autocrine fashion. These changes in growth factors and its receptor expression was accompanied by an increase in proliferating cell nuclear antigen (PCNA), Ki-67 antigen, ornithine decarboxylase activity (ODC) and aneuploidy indicating enhanced cell proliferation in patients with BE augmented also by p53 gene mutation [17, 18]. Changes in nuclear oncogene c-myc, signal transduction oncogene ras, tumor suppressor gene rb, or fibroblast growth factor-related oncogenes hst-1 and int-2 in primary esophageal cancer or developed esophageal cancer cell lines have also been reported .
In our own study, patients with endoscopic RE exhibited diminished salivary and/or esophageal EGF secretion persisting even after healing of endoscopic esophagitis suggesting that low EGF secretion could be the predisposing factor to the development of erosive RE in patients with GERD [1-3]. In addition, in our preliminary, data we have also demonstrated that patients with BE which developed in the setting of GERD also exhibited impairment in EGF secretion . This impairment in the rate of secretion of esophageal EGF could be attributed to lower expression of EGF gene at its peptide level within esophageal submucosal mucous glands or due to a lower number of submucosal mucous glands. A variation in the number of submucosal mucous glands in the human esophagus from as low as 62 to as high as 741 glandular lobules has been reported by Goetsch .
A similar phenomenon (diminished expression or smaller size of salivary glands) could also lead to a decline of salivary EGF in both populations. Only approximately one third of patients exhibited low EGF in both salivary and esophageal secretion, the remaining showed a decline in salivary or esophageal secretion only. The clinical relevance of this phenomenon still remains to be established.
As it has been recently demonstrated by immunohistochemistry, a steady increase in the content of TGFa (but not EGF) and EGF/TGFa receptor has been demonstrated starting from a fundic type of gastric epithelium going through specialized intestinal metaplasia, dysplasia and finally adenocarcinoma . Therefore, metaplastic epithelium is acquiring its own source of peptide growth factor, predominantly TGFa, which binds to the same ERGF receptor. It may serve as a source of high proliferative activity in metaplastic epithelium which may promote the development of dysplastic changes and, perhaps, adenocarcinoma.
According to some sources, 60 to 90% of human cancers are related primarily to environmental factors and smoking-induced lung cancer is considered a prototype of such cancers . However, still some families of smokers are more susceptible than others, pointing towards the significance of genotype even in etiology of cancers with such obvious contribution of environmental carcinogens as lung cancer .
These data indicate that both within C and AA populations there must be a gene(s) which increases susceptibility to the development of esophageal adenocarcinoma and squamous cell carcinoma which is linked to the chromosome Y. There are a number of genes linked to the development of alimentary tract cancers. The best known for colon cancer are p53, K-Ras, APC, DCC . The search for other genes continues and the pathogenetic sequence leading to esophageal cancer development will gradually unravel.
2. Rourk RM, Namiot Z, Sarosiek J, Yu Z, McCallum RW. Impairment of salivary epidermal growth factor secretory response to esophageal mechanical and chemical stimulation in patients with reflux esophagitis. Am J Gastroenterol 1994;89:237-244.
3. Edmunds MC, Namiot Z, Sarosiek J, Rourk RM, Yu Z, McCallum RW. Esophageal epidermal growth factor impairment persists depite healing of endoscopic changes in patients with reflux esophagitis. Gastroenterology 1994;106:A73.
5. Namiot Z, Sarosiek J, Edmunds MC, McCallum RW. Gender-related salivary and esophageal epidermal growth factor secretory patterns: are they predictive factors of the development of esophageal disease? Gastroenterology 1994;106:A20.
9. Marcinkiewicz M, Scheurich J, Goldin G, Sarosiek J, McCallum RW. Salivary glycoconjugate (mucin): is it the answer to the lower prevalence of gastroesophageal reflux disease in African-Americans? Am J Gastroenterol 1996;91:1893.
11. Thompson JJ. Esophageal cancer and the premalignant changes of esophageal diseases. In: Cohen S, Soloway RD, eds.Diseases of the esophagus. New York, Edinburgh, London, Melbourne: Churchill Livingstone, 1982:239-276.