Endocrine cells in the oxyntic mucosa of the stomach in patients with irritable bowel syndrome.

AIM To study the different endocrine cell types in the oxyntic mucosa of patients with irritable bowel syndrome (IBS). METHODS Seventy-six patients with IBS were included in the study (62 females and 14 males; mean age 32 years, range 18-55 years), of which 40 also fulfilled the Rome III criteria for functional dyspepsia (FDP). Of the entire IBS cohort, 26 had diarrhea as the predominant symptom (IBS-D), 21 had a mixture of diarrhea and constipation (IBS-M), and 29 had constipation as the predominant symptom (IBS-C). Forty-three age and sex-matched healthy volunteers without any gastrointestinal complaints served as controls. The patients were asked to complete the Birmingham IBS symptom questionnaire. Both the patients and controls underwent a standard gastroscopy, during which three biopsy samples were taken from the corpus. Sections from these biopsy samples were immunostained using the avidin-biotin complex (ABC) method, for ghrelin, serotonin, somatostatin and histamine. The densities of these cell types and immunoreactivity intensities were quantified using computerized image analysis with Olympus cellSens imaging software (version 1.7). RESULTS The densities of the ghrelin cells in the control, IBS-total, IBS-D, IBS-M and IBS-C groups were 389 (320, 771), 359 (130, 966), 966 (529, 1154), 358 (120, 966) and 126 (0, 262) cells/mm(2), respectively. There was a significant difference between the tested groups (P < 0.0001). Dunn's multiple comparison test showed that the ghrelin cell density was significantly higher in IBS-D and lower in IBS-C than in the controls (P = 0.03 and 0.0008, respectively). The ghrelin cell density in patients with both IBS and FDP was 489 (130, 966), and in those with IBS only 490 (130, 956). There was no statistical significant difference between these 2 groups of patients (P = 0.9). The immunoreactivity intensity did not differ between any of the groups (P = 0.6). The diarrhea score of the Birmingham IBS symptom questionnaire was significantly positively correlated with ghrelin cell density (r = 0.65; P < 0.0001) and significantly inversely correlated with that of constipation (r = 90.69; P < 0.0001). The densities of the serotonin cells were 63 (51, 82), 51 (25, 115), 120 (69, 128), 74 (46, 123) and 40 (0, 46) cells/mm(2) in the control, IBS-total, IBS-D, IBS-M and IBS-C groups, respectively. A statistically significant difference was found between the tested groups (P < 0.0001). Posttest revealed that serotonin cell density was significantly higher in IBS-D and lower in IBS-C than in controls (P = 0.02 and 0.004, respectively), but did not differ in the IBS-total and IBS-M groups from that in controls (P = 0.5 and 0.4, respectively). The serotonin cell density in patients with both IBS and FDP was 62 (25, 115) and in those with IBS only 65 (25, 123). There was no statistically significant difference between these 2 groups of patients (P = 1). The immunoreactivity intensity of serotonin did not differ significantly between any of the groups (P = 0.0.9). The serotonin cell density was significantly positively correlated with the diarrhea score of the Birmingham IBS symptom questionnaire (r = 0.56; P < 0.0001) and significantly inversely correlated with that of constipation (r = 0.51; P < 0.0001). The densities of the somatostatin cells were 97 (72, 126), 72 (0, 206), 29 (0, 80), 46 (0, 103) and 206 (194, 314) cells/mm(2) in the control, IBS-total, IBS-D, IBS-M and IBS-C groups, respectively (Figures 7 and 8). There was a statistically significant difference between the controls and the IBS subgroups (P < 0.0001). The density of somatostatin cells was significantly lower in the IBS-D and IBS-M groups but higher in IBS-C patients than in the controls (P < 0.01, P = 0.02, and P = 0.0008, respectively). The somatostatin cell density in patients with both IBS and FDP was 86 (0-194), and in those with IBS only 110 (0-206). There was no statistically significant difference between these 2 groups of patients (P = 0.6). There was no significant difference in somatostatin immunoreactivity intensity between the controls. The diarrhea score of the Birmingham IBS symptom questionnaire was inversely correlated with somatostatin cell density (r = 0.38; P = 0.0007) and was positively correlated with that of constipation (r = 0.64; P < 0.0001). CONCLUSION The finding of abnormal endocrine cells in the oxyntic mucosa shows that the endocrine cell disturbances in IBS are not restricted to the intestine. Furthermore, it appears that ghrelin, serotonin and somatostatin in the oxyntic mucosa of the stomach may play an important role in the changing stool habits in IBS through their effects on intestinal motility.