Small Intestinal Permeability in Patients With EoE
Small Intestinal Permeability in Patients With EoE
This study shows that small bowel permeability is increased in patients with active eosinophilic oesophagitis, and is normal in patients with eosinophilic oesophagitis in remission. These data suggest that further studies of the role of the small bowel in active eosinophilic oesophagitis deserves further investigation. Eosinophilic oesophagitis is caused to a large degree by exposure to food allergens, and may respond to restriction of allergenic foods or suppression of immune mechanisms associated with tissue injury, as with topical corticosteroids. Although the oesophageal manifestations of this allergic diathesis are well documented in the literature (reviewed in refs 1, 7), there is far less known about the specific interaction of oesophageal epithelium with food antigen, or the mechanisms whereby food antigen gains access to the oesophageal wall to establish the cascade of immunological events that result in the histological manifestations of eosinophilic oesophagitis. As a result, several questions still remain: first, it is unclear what mechanism of antigen exposure establishes the immunological responses; second, it is unknown whether the oesophagus is the sole site of antigen recognition in eosinophilic oesophagitis.
In this study, it is demonstrated that many of the patients with active eosinophilic oesophagitis have increased small bowel permeability when compared to patients with treated disease with evidence of histologic remission and controls. Although one usually associates the finding of increased small bowel permeability in patients with diffuse inflammatory disease of the small bowel, increased permeability is also reported in morphologically normal small bowel in patients with allergic diseases, such as eczema, nasal polyposis, asthma, cow's milk intolerance food allergy, and irritable bowel syndrome with atopy or without atopy. Increased permeability has also been reported with the stress associated with combat training and may vary with age. With increased small bowel permeability, dietary antigens can gain access to the circulation in response to different foods, for example, high fat diet. It is also interesting to note that in severe childhood eczema, as in eosinophilic oesophagitis, an elemental diet may provide effective management raising the question of whether its beneficial effect is on the oesophagus, small bowel or both.
An alternative route to establish antigen exposure might be direct infiltration of antigens in the oesophageal mucosa. This has not been fully investigated, and it is challenging to consider this as the only source of antigen recognition for at least three reasons: first, it is conceptually difficult to understand how food traversing the length of the oesophagus in a few seconds would have sufficient contact time to permeate the predominantly stratified squamous epithelium in the oesophagus that is bereft of any carrier proteins for nutrients. Second, the process of digestion initiated by salivary enzymes (predominantly amylase) is unlikely to result in the release from food and presentation of protein or peptide antigens to the oesophageal mucosa. Third, the normal squamous epithelium has an extremely high transepithelial electrical resistance, permits about 0.01% of probe molecule (fluorescein) flux in vitro, and expresses tight junction proteins such as zonula occludens 1 and different claudins (1, 4, 10, 12 and 17).
It is intriguing to note the observations of Merwat and Spechler on the time course of the introduction and subsequent widespread use of PPIs with the emergence of eosinophilic oesophagitis. They proposed the hypothesis that acid-suppressive medications may predispose to the development of eosinophilic oesophagitis, perhaps through increased upper gastrointestinal permeability documented with the sucrose permeability test. Further studies of oesophageal mucosal permeability in eosinophilic oesophagitis are eagerly awaited.
The physiologic contribution of the gut to the allergic diathesis is unclear, but our observations of increased small bowel permeability in active eosinophilic oesophagitis can modulate primary extraintestinal allergic diseases such as asthma and atopic dermatitis, and partly explains the improvement with dietary manipulation. Whether medications, such as larazotide (which promotes tight junction assembly in epithelial cells) or glutamine (to reduce intestinal permeability) can decrease small bowel antigen permeation in eosinophilic oesophagitis remains to be studied.
One of the interesting aspects of this study was demonstrating that eosinophilic oesophagitis in remission was associated with normal intestinal permeability. A possible mechanism that might explain this observation is the effect of circulating cytokines on intestinal permeability. For example, IL-13 applied to an epithelial monolayer may increase tissue permeability. TNF-α increased endothelial permeability and reduced barrier function in a variety of epithelia. This may correspond to the finding in patients with eosinophilic oesophagitis, where incubation of peripheral blood monocytes with specific allergens may lead to elevations in IL-5 and 13, and a recent study demonstrating that patients with eosinophilic oesophagitis have increased blood levels of several cytokines, including IL-13. Therapeutically, elevation of the tissue IL-13-induced eosinophilic oesophagitis transcriptome is reversed by administration of steroids. It is therefore conceivable that the effect of presumed oesophageal-specific therapy in patients with eosinophilic oesophagitis might be the result of reduction of small bowel permeability as an adjunct site for antigen recognition and amelioration of the disease. It is attractive to think that targeting the small bowel in patients with eosinophilic oesophagitis may be a potential therapeutic target as has been shown in a mouse model of colitis.
This study has several potential pitfalls. First, the testing for small bowel permeability was not performed in the same patients before and after treatment. On the other hand, the highly statistical significance seen between active disease and disease in remission with such a limited number of patients studied reinforces the validity of our findings, and we also noted the similar results in patients in remission and healthy controls. Second, we observed relatively wide variations in absolute values of lactulose and mannitol excretion in our patients. This has been noted in prior studies from our laboratory and is likely accounted for by variations in gut transit of the sugars, hence, the reliance on the ratio for measurement. A third potential limitation was the presence of coincident cause for increased small bowel permeability. In this study, great effort was made to exclude diseases that would affect small bowel permeability, through a normal small bowel biopsy, and no symptoms on a detailed questionnaire to suggest small bowel disease. Although other atopic diseases may be associated with increased small bowel permeability, there was an equal distribution of extraoesophageal allergic disorders in both patient groups, so this would not explain the differences in results. Fourth, the finding of normalised small bowel permeability in the patients with eosinophilic oesophagitis in remission could conceivably result from small bowel exposure to the topical steroid. We believe this is unlikely, since we administered approximately 10 mL of a diluted budesonide formulation (which has little systemic absorption), and do not perceive this could treat the length of small bowel that is reflected by the 0–2 h urine collections of the two sugars. Furthermore, one of our patients in remission was effectively treated with exclusion diet only, and this patient had a normal urine lactulose to mannitol excretion ratio. Fifth, one might postulate that increased oesophageal permeability might explain the abnormal lactulose to mannitol ratio. Indeed, recent work by Sherrill and colleagues has demonstrated that there is increased esophageal permeability in patients with eosinophilic esophagitis. We feel, this is unlikely to explain most of our results, however, given the short exposure time of the oesophagus to a liquid bolus and the measurement of peaks and total for urinary sugar excretion most commonly after the first 30 min of sugar ingestion, far longer than would be associated with rapid oesophageal absorption.
In conclusion, this study demonstrates that patients with eosinophilic oesophagitis have increased small bowel permeability when compared to patients with eosinophilic oesophagitis in remission who have normal small bowel permeability, similar to that of healthy controls. These data support the need for further studies to replicate the current findings, and to embark on approaches that normalise small bowel permeability for treating eosinophilic oesophagitis.
Discussion
This study shows that small bowel permeability is increased in patients with active eosinophilic oesophagitis, and is normal in patients with eosinophilic oesophagitis in remission. These data suggest that further studies of the role of the small bowel in active eosinophilic oesophagitis deserves further investigation. Eosinophilic oesophagitis is caused to a large degree by exposure to food allergens, and may respond to restriction of allergenic foods or suppression of immune mechanisms associated with tissue injury, as with topical corticosteroids. Although the oesophageal manifestations of this allergic diathesis are well documented in the literature (reviewed in refs 1, 7), there is far less known about the specific interaction of oesophageal epithelium with food antigen, or the mechanisms whereby food antigen gains access to the oesophageal wall to establish the cascade of immunological events that result in the histological manifestations of eosinophilic oesophagitis. As a result, several questions still remain: first, it is unclear what mechanism of antigen exposure establishes the immunological responses; second, it is unknown whether the oesophagus is the sole site of antigen recognition in eosinophilic oesophagitis.
In this study, it is demonstrated that many of the patients with active eosinophilic oesophagitis have increased small bowel permeability when compared to patients with treated disease with evidence of histologic remission and controls. Although one usually associates the finding of increased small bowel permeability in patients with diffuse inflammatory disease of the small bowel, increased permeability is also reported in morphologically normal small bowel in patients with allergic diseases, such as eczema, nasal polyposis, asthma, cow's milk intolerance food allergy, and irritable bowel syndrome with atopy or without atopy. Increased permeability has also been reported with the stress associated with combat training and may vary with age. With increased small bowel permeability, dietary antigens can gain access to the circulation in response to different foods, for example, high fat diet. It is also interesting to note that in severe childhood eczema, as in eosinophilic oesophagitis, an elemental diet may provide effective management raising the question of whether its beneficial effect is on the oesophagus, small bowel or both.
An alternative route to establish antigen exposure might be direct infiltration of antigens in the oesophageal mucosa. This has not been fully investigated, and it is challenging to consider this as the only source of antigen recognition for at least three reasons: first, it is conceptually difficult to understand how food traversing the length of the oesophagus in a few seconds would have sufficient contact time to permeate the predominantly stratified squamous epithelium in the oesophagus that is bereft of any carrier proteins for nutrients. Second, the process of digestion initiated by salivary enzymes (predominantly amylase) is unlikely to result in the release from food and presentation of protein or peptide antigens to the oesophageal mucosa. Third, the normal squamous epithelium has an extremely high transepithelial electrical resistance, permits about 0.01% of probe molecule (fluorescein) flux in vitro, and expresses tight junction proteins such as zonula occludens 1 and different claudins (1, 4, 10, 12 and 17).
It is intriguing to note the observations of Merwat and Spechler on the time course of the introduction and subsequent widespread use of PPIs with the emergence of eosinophilic oesophagitis. They proposed the hypothesis that acid-suppressive medications may predispose to the development of eosinophilic oesophagitis, perhaps through increased upper gastrointestinal permeability documented with the sucrose permeability test. Further studies of oesophageal mucosal permeability in eosinophilic oesophagitis are eagerly awaited.
The physiologic contribution of the gut to the allergic diathesis is unclear, but our observations of increased small bowel permeability in active eosinophilic oesophagitis can modulate primary extraintestinal allergic diseases such as asthma and atopic dermatitis, and partly explains the improvement with dietary manipulation. Whether medications, such as larazotide (which promotes tight junction assembly in epithelial cells) or glutamine (to reduce intestinal permeability) can decrease small bowel antigen permeation in eosinophilic oesophagitis remains to be studied.
One of the interesting aspects of this study was demonstrating that eosinophilic oesophagitis in remission was associated with normal intestinal permeability. A possible mechanism that might explain this observation is the effect of circulating cytokines on intestinal permeability. For example, IL-13 applied to an epithelial monolayer may increase tissue permeability. TNF-α increased endothelial permeability and reduced barrier function in a variety of epithelia. This may correspond to the finding in patients with eosinophilic oesophagitis, where incubation of peripheral blood monocytes with specific allergens may lead to elevations in IL-5 and 13, and a recent study demonstrating that patients with eosinophilic oesophagitis have increased blood levels of several cytokines, including IL-13. Therapeutically, elevation of the tissue IL-13-induced eosinophilic oesophagitis transcriptome is reversed by administration of steroids. It is therefore conceivable that the effect of presumed oesophageal-specific therapy in patients with eosinophilic oesophagitis might be the result of reduction of small bowel permeability as an adjunct site for antigen recognition and amelioration of the disease. It is attractive to think that targeting the small bowel in patients with eosinophilic oesophagitis may be a potential therapeutic target as has been shown in a mouse model of colitis.
This study has several potential pitfalls. First, the testing for small bowel permeability was not performed in the same patients before and after treatment. On the other hand, the highly statistical significance seen between active disease and disease in remission with such a limited number of patients studied reinforces the validity of our findings, and we also noted the similar results in patients in remission and healthy controls. Second, we observed relatively wide variations in absolute values of lactulose and mannitol excretion in our patients. This has been noted in prior studies from our laboratory and is likely accounted for by variations in gut transit of the sugars, hence, the reliance on the ratio for measurement. A third potential limitation was the presence of coincident cause for increased small bowel permeability. In this study, great effort was made to exclude diseases that would affect small bowel permeability, through a normal small bowel biopsy, and no symptoms on a detailed questionnaire to suggest small bowel disease. Although other atopic diseases may be associated with increased small bowel permeability, there was an equal distribution of extraoesophageal allergic disorders in both patient groups, so this would not explain the differences in results. Fourth, the finding of normalised small bowel permeability in the patients with eosinophilic oesophagitis in remission could conceivably result from small bowel exposure to the topical steroid. We believe this is unlikely, since we administered approximately 10 mL of a diluted budesonide formulation (which has little systemic absorption), and do not perceive this could treat the length of small bowel that is reflected by the 0–2 h urine collections of the two sugars. Furthermore, one of our patients in remission was effectively treated with exclusion diet only, and this patient had a normal urine lactulose to mannitol excretion ratio. Fifth, one might postulate that increased oesophageal permeability might explain the abnormal lactulose to mannitol ratio. Indeed, recent work by Sherrill and colleagues has demonstrated that there is increased esophageal permeability in patients with eosinophilic esophagitis. We feel, this is unlikely to explain most of our results, however, given the short exposure time of the oesophagus to a liquid bolus and the measurement of peaks and total for urinary sugar excretion most commonly after the first 30 min of sugar ingestion, far longer than would be associated with rapid oesophageal absorption.
In conclusion, this study demonstrates that patients with eosinophilic oesophagitis have increased small bowel permeability when compared to patients with eosinophilic oesophagitis in remission who have normal small bowel permeability, similar to that of healthy controls. These data support the need for further studies to replicate the current findings, and to embark on approaches that normalise small bowel permeability for treating eosinophilic oesophagitis.
