Cystic fibrosis (CF) is the most common lethal genetic disease among whites, occurring in 1 of 2500 live births. It is also the most common cause of pancreatic insufficiency (PI) in children. Although CF was previously fatal in childhood, a majority of children now survive to adulthood because of improved care (1). Approximately 85% of the patients with CF have PI with impaired digestion presenting as steatorrhea, fat-soluble vitamin deficiency, and poor nutritional status. Despite adequate pancreatic enzyme replacement therapy (PERT), a majority of patients with CF complain of gastrointestinal symptoms including abdominal pain, steatorrhea, and motility problems. It has been suggested that hitherto poorly understood nonpancreatic intestinal factors are likely contributors (2–4). It is therefore of interest that in the mouse model of CF, the major manifestation is intestinal obstruction leading to perforation consequent to an inflammatory enteropathy (5). In children with CF the histology of duodenal biopsies is normal except for a thick mucous layer (6–8) or mononuclear cell infiltration of the lamina propria (9). Other studies show that the intestinal barrier function is severely compromised in CF (10–14). One suggestion is that such permeability changes inevitably lead to small bowel damage (15). That this may be the case is suggested by findings of increased amounts of inflammatory markers in whole-gut lavage studies (16–18). However, the localization and nature of any such inflammation is uncertain.
The aim of the present study was to search for evidence of inflammation by direct inspection of the mucosa of the entire small intestine. We used the fecal calprotectin test (19) and wireless capsule enteroscopy (WCE) (20) to quantitate and localize intestinal inflammation, respectively, in patients with CF and relate these findings to clinical status.
Patients with CF age 10 years or older, followed in cystic fibrosis centers in Israel, were eligible. The diagnosis of CF was made by a positive sweat test in the presence of chronic lung disease. Mutational analysis for the most common mutations in Israel was performed in all of the patients. PI was determined by a low coefficient of fat absorption and improvement with PERT. All of the patients with PI were receiving PERT at the time of the study. Exclusion criteria included previous intestinal surgery; inability to swallow the capsule; severe neurologic, psychiatric, cardiovascular, or rheumatologic disease; ingestion of nonsteroidal anti-inflammatory drugs (NSAIDs) or other conditions that are known to be associated with intestinal inflammation; and any condition precluding surgery or anesthesia. All of the patients in the study had negative serologic tests for celiac disease and had no clinical, laboratory, or radiological evidence of Crohn disease. Clinical and demographic data were collected from the medical records.
All of the patients enrolled ingested an agile patency capsule (PC) (Given Imaging, Yokneam, Israel) to confirm bowel patency and the absence of strictures before capsule endoscopy. The biodegradable PC is the same size as the capsule endoscope (PillCam, Given Imaging, Yokneam, Israel) but dissolves if retained, such as by a stricture. The PC contains a radiofrequency identification tag. The PC is ingested after an overnight fast. Patency is verified if the PC is excreted intact as verified visually or by the absence of a signal detected by a scanner.
On the day before testing a standard bowel preparation consisting of sodium hydrogen phosphate was given. After an overnight fast the PillCam SB was ingested before 8 AM. Recording was performed as recommended. After 2 to 3 hours a small amount of clear fluid was allowed. Solids could be ingested after 4 to 5 hours. The study was terminated after 8 hours. The studies were interpreted by expert reviewers (S.W., S.N.A.). Previously used criteria published by Maiden et al (21) for video capsule interpretation were used in the present study, as follows.
Category 1: petechiae/“red spot” (demarcated, usually circular, area of crimson mucosa with preservation of villi). Category 2: reddened folds (1 or more valvulae conniventes exhibiting discrete patchy or continuous erythema). Category 3: denuded area (loss of villous architecture without a clear breach of the epithelium; this may or may not be associated with surrounding erythema). Category 4: mucosal break (mucosal erosions and/or ulcers; both represent discrete lesions with central pallor and surrounding hyperemia and loss of villi. These are included together as an ulcer requires, by definition, a degree of penetration [through the muscularis mucosa] and the angle of the image taken by the capsule is often such that it is impossible to evaluate the depth of the lesions). Category 5: stricture. Category 6: presence of blood without visualized lesion. Category 7: other pathological findings, mainly lymphangiectasia and angiodysplasia (dilated, fragile blood vessels, usually the result of arteriovenous malformation), the latter differing from petechiae and red spots on the basis of size. The results were compared with a historical control group of 60 subjects without CF.
Fecal calprotectin was measured in 30 patients and 30 age-matched controls. Subjects provided stool samples, and these were frozen (at −20°C) within 12 hours and stored for later analysis. After thawing, 200-mg aliquots were processed for quantitative measurements by a sensitive and specific commercial enzyme-linked immunoassay (Calprest, Eurospital, Italy) as previously described (22). The upper normal limit for fecal calprotectin concentration is 50 μg/g stool).
Statistical analysis was performed by analysis of variance, Pearson correlation coefficient, and χ2 test between patient groups. Written informed consent was obtained from all of the patients or their parents. The present study was approved by the Helsinki Committee of the Hadassah Medical Center, Jerusalem, Israel; the Israel Ministry of Health; and the Research and Publications Committee of the Children's Hospital of Wisconsin, Milwaukee, WI.
Forty-two patients entered the study but 1 failed to pass the PC after 36 hours, so he did not undergo WCE. Forty-one patients with CF (27 males) aged 11 to 56 years (mean 22 ± 11 years) underwent WCE. Twenty-eight had PI, and 13 were pancreatic sufficient (PS). Demographic details, symptoms and the results of genotyping, WCE, and fecal calprotectin measurements are shown in Table 1. The most common genotypes found were W1282X/ΔF508 (n = 9), G542X/ΔF508 (n = 7), and W1282X/W1282X (n = 5). Nine Arabic subjects had unknown mutations. Fifteen subjects reported abdominal pain (PI, 10; PS, 5). Five subjects with PI complained of diarrhea (Table 1).
The findings on WCE showed varied pathological findings in the jejunum and ileum. Diffuse or localized small bowel lesions including villous blunting, edema, erythema, denuded mucosa, and mucosal breaks (erosions or ulcers) were observed throughout the jejunum and ileum in 26 of 41 (63%) patients: 20 (71%) of the 28 patients with PI, and 6 (46%) of the 13 patients who were PS. There was no correlation between specific CFTR mutations and WCE findings. Representative images are shown in Figure 1. In 1 patient an ileoileal intussusception was observed, which spontaneously reverted. No Maiden category 5 to 7 were observed.
Table 2 summarizes the Maiden classification in these patients. Patients with PI on the whole had more significant mucosal changes, with many patients showing areas of edema, mucosal erosions, and discrete ulcerations. Interestingly, some patients who were PS exhibited similar mucosal lesions, thus ruling out PERT or the coating of the enzymes as the cause.
There were no abnormalities more severe than category 1 in the control group (n = 60) apart from petechiae (18%), lymphangiectasia (5%), and angiodysplasia (5%).
Stool for fecal calprotectin was obtained from 30 subjects. It was normal (23.2 μg/g) in all of the 9 subjects who were PS and was elevated in the subjects with PI (257.7 μg/g), indicating intestinal inflammation only in the patients with PI. Three patients with PI had normal values. All of the control patients had values within the normal range, 10 to 50 μg/g.
Twenty-five patients had body mass index >20 and 16 had <20 (Table 1). Eight of the former group (32%) and 9 of the latter group (56%) had a Maiden score of 4 (P < 0.124). This reveals a possible trend between poor nutritional status and more severe findings on capsule endoscopy. We did not find any trend between fecal calprotectin and nutritional status.
The present study demonstrates a new observation, a high prevalence of small bowel injury in patients with CF, both patients with PI and those who were PS. The macroscopic appearance of the small intestine may be an integral part of the CF phenotype because it does not relate to the degree of pancreatic disease.
Eighty-five percent of the patients with CF have PI and maldigestion/malabsorption. It is clear that the common belief that pancreatic enzyme dose can be titrated to control the maldigestion/malabsorption symptoms is incorrect (2). Hence, despite an apparent optimal dose of PERT, most patients have an incomplete response with significant residual malabsorptive symptoms. The finding of an enteropathy in a substantial number of patients with CF suggests that small bowel inflammation may contribute significantly to the malabsorption in these patients. The severity of the lesions may be related to poor nutritional status.
Previous studies have demonstrated intestinal inflammation in both the mouse model of CF and in children with CF. In the CF mouse model Norkina et al (5) showed that genes associated with inflammation are upregulated. Bensalem et al (23) have shown in the CF mouse that the protein Annexin 1, which negatively modulates leukocyte activation in response to inflammation, is downregulated, which may be responsible for the proinflammatory state. Similarly, in studies in children with CF increased inflammatory proteins were measured in gut lavage fluid (16–18).
There is no universally accepted scoring system for interpreting WCE studies. Some previously reported scoring systems are not appropriate for our survey of patients with CF. For example, the 4 parameters in the scoring index proposed by Kornbluth et al (24) are edema, nodularity, ulcer, and stenosis, and the 3 parameters in the index proposed by Gralnek et al (25) are villous edema, ulcer, and stenosis. In our study stenosis was an exclusion criterion, and ulcers were not an expected finding. Thus, we used a comprehensive scoring system that is suitable for research but perhaps less so in a busy clinical setting. We acknowledge that this classification, like the others, has not been validated sufficiently.
Some studies using healthy volunteers have described significant baseline abnormalities. Maiden et al (21) have pointed out that healthy volunteers may have category 1 lesions, lymphangiectasia, and angiodysplasia, and that denuded areas, villous blunting, mucosal breaks, and so on are indicators of injury. The abnormalities in the patients with CF were indeed in category 2 to 4, which are not a feature in healthy subjects. The prevalence and type of the lesions are similar to that seen in subjects and patients taking short- and long-term NSAIDs, respectively (26,27). Furthermore, in the absence of clinical information these changes can easily be interpreted as being suggestive of Crohn disease.
Fecal calprotectin was elevated in 85% subjects with PI and normal in all of the patients who were PS. From our study it cannot be determined whether the inflammation, as defined by an elevated calprotectin, is a function of the PI state or is due to PERT because it is known that high doses of PERT cause fibrosing colonopathy, an inflammatory condition (28). The inflammation was, in many cases, comparable to that seen in patients with active Crohn disease (fecal calprotectin >400) (29). However, there have been no definitive studies reported on the diagnostic value of fecal calprotectin in CF. A single measurement of fecal calprotectin may not be sufficiently accurate to rule out inflammation, and other markers may be required. Furthermore, irrespective of the inflammatory activity there is a high prevalence of morphologic injury in the small intestine of patients with CF. Epithelial injury may contribute to fat malabsorption in humans. At present, we do not know whether the novel findings in the present study are because of an enteropathy intrinsically related to “CF enteropathy,” chronic enzyme usage, dysmotility, bacterial overgrowth, or secondary to another unidentified abnormality present in patients with CF. The possibility that there is a CF enteropathy is suggested by the findings that PS is associated with morphological small bowel changes, although they may not be “inflammatory” in nature because the calprotectin is normal.
The site of the intestinal inflammation in humans, as shown by whole-gut lavage studies, has until now remained uncertain. The pathogenesis of the damage seen on WCE in the patients with CF is uncertain. However, in an extensive review of the literature assessing the interaction between the intestinal barrier function, luminal aggressors, and intestinal inflammation, it has been hypothesized that most small bowel disorders have a common final pathogenic pathway (15). Accordingly, most small bowel diseases are characterized by similar changes (quantitatively) in intestinal permeability and inflammation. It is suggested that the inflammatory response is consequent to commensal bacteria gaining access to the mucosa because of the breach in the barrier function, regardless of whether this was caused by direct damage (eg, alcohol, NSAIDs), luminal aggressors (eg, enteric infections, bacterial overgrowth), or impaired mucosal defense (immune deficiency syndromes). Other possible mechanisms include that the reduced intestinal chloride transport could lead to an alteration of the intestinal mucous layer, causing damage directly or secondarily from small intestinal bacterial overgrowth, and the polypharmacy that these patients ingest may have an effect on the intestinal mucosa. The small bowel injury in patients with CF could therefore have been predicted (15), but the precise pathogenesis requires additional study. Nevertheless, our WCE findings are similar to those described in other small bowel diseases, whereas the inflammatory component (as reflected by fecal calprotectin levels) changes related to the degree of pancreatic impairment.
The role of bacteria in the inflammation is further evident from the efficacy of probiotics in reducing both the intestinal inflammation and pulmonary exacerbations (30,31). A possible explanation of our findings is that the pancreas may play a more important role in maintaining low bacterial counts in the small bowel than previously recognized. Alternatively, the pancreatic supplementation in the patients with PI may be damaging to the small bowel.
In summary, the present proof-of-concept study suggests that there is a condition compatible with a “CF-bowel.” The additional contribution of chronic enzyme usage, dysmotility, bacterial overgrowth, or another unidentified abnormality may play an additive role. CF predisposes to inflammatory changes not only in the respiratory system but also in the gastrointestinal tract. Larger studies undoubtedly need to be performed including patients with exocrine PI receiving PERT without CF, such as those with other forms of chronic pancreatitis. Most important, our findings raise the possibility that if further research confirms and explains the mechanisms of these findings, there is an opportunity for targeted treatment of the enteropathy with the potential for improving the quality of life of these patients.
2. Borowitz D, Durie PR, Clarke LL, et al
. Gastrointestinal outcomes and confounders in cystic fibrosis
. J Pediatr Gastroenterol Nutr 2005; 41:273–285.
3. Negila JP, FitzSimmons SC, Maisonneuve P, et al
. Risk of cancer among patients with cystic fibrosis
. Cystic Fibrosis
and Cancer Study Group. N Engl J Med 1995; 332:494–499.
4. Maisonneuve P, FitzSimmons SC, Neglia JP, Campbell PW 3rd, Lowenfels AB. Cancer risk in non transplanted and transplanted cystic fibrosis
patient: a 10 year study. J Natl Cancer Inst 2003; 95:381–387.
5. Norkina O, Kaur S, Zeimer D, De Lisle RC. Inflammation of cystic fibrosis
mouse small intestine. Am J Physiol Gastrointest Liver Physiol 2004; 286:G1032–G1041.
6. Freye HB, Kurtz S, Spock A, Capp MP. Light and electron microscopic examination of the small bowel of children with cystic fibrosis
. J Pediatr 1964; 64:575–579.
7. Thomadis TS, Arey JB. The intestinal lesions in cystic fibrosis
of the pancreas. J Pediatr 1963; 63:16–21.
8. Eggermont E. The role of the small intestine in cystic fibrosis
patients. Acta Paediatr Scand Suppl 1985; 317:16–21.
9. Raia V, Maiuri L, de Ritis G, et al
. Evidence of chronic inflammation in morphologically normal small intestine of cystic fibrosis
patients. Pediatr Res 2000; 47:344–350.
10. Hallberg K, Grzegorczyk A, Larson G, Strandvik B. Intestinal permeability in cystic fibrosis
in relation to genotype. J Pediatr Gastroenterol Nutr 1997; 25:290–295.
11. Van Elburg RM, Uil JJ, van Aalderen WM, Mulder CJ, Heymans HS. Intestinal permeability in exocrine pancreatic insufficiency due to cystic fibrosis
or chronic pancreatitis. Pediatr Res 1996; 39:985–991.
12. Escobar H, Perdomo M, Vasconez F, et al
. Intestinal permeability to 51Cr-EDTA and orocecal transit time in cystic fibrosis
. J Pediatr Gastroenterol Nutr 1992; 14:204–207.
13. Leclercq-Foucart J, Forget PP, Van Cutsem JL. Lactulose-rhamnose Intestinal permeability in children with cystic fibrosis
. J Pediatr Gastroenterol Nutr 1987; 6:66–70.
14. Leclercq-Foucart J, Forget P, Sodoyez-Goffaux F, Zappitelli A. Intestinal permeability to [51Cr]EDTA in children with cystic fibrosis
. J Pediatr Gastroenterol Nutr 1986; 5:384–387.
15. Bjarnason I, Takeuchi K, Bjarnason A, et al
. The G.U.T. of gut. Scand J Gastroenterol 2004; 39:807–815.
16. Smyth RL, Croft NM, O'Hea U, et al
. Intestinal inflammation in cystic fibrosis
. Arch Dis Child 2000; 82:394–399.
17. Croft NM, Marshall TG, Ferguson A. Gut inflammation in children with cystic fibrosis
on high dose enzyme supplements. Lancet 1995; 346:1265–1267.
18. Croft NM, Marshall TG, Ferguson A. Direct assessment of gastrointestinal inflammation and mucosal immunity in children with cystic fibrosis
. Postgrad Med J 1996; 72(Suppl 2):S32–S36.
19. Canani RB, Terrin G, Rapacciuolo L, et al
. Faecal calprotectin
as reliable non-invasive marker to assess the severity of mucosal inflammation in children with inflammatory bowel disease. Dig Liver Dis 2008; 40:547–553.
20. Sidhue R, McAlindon ME, Sanders DS, Thomson M. Capsule endoscopy
in the evaluation of gastrointestinal disease. Curr Opin Pediatr 2007; 29:586–590.
21. Maiden L, Thjodleifsson B, Seigal A, et al
. Long-term effects of nonsteroidal anti-inflammatory drugs and cyclooxygenase-2 selective agents on the small bowel: a cross-sectional capsule enteroscopy study. Clin Gastroenterol Hepatol 2007; 5:1040–1045.
22. Johne B, Kronborg O, Ton H, Kristinsson J, Fuglerud P. A new fecal calprotectin
test for colorectal neoplasia. Clinical results and comparison with previous method. Scand J Gastroenterol 2001; 36:291–296.
23. Bensalem N, Ventura AP, Vallee B, et al
. Down-regulation of the anti-inflammatory protein Annexin-1 in cystic fibrosis
knock-out mice and patients. Mol Cell Proteomics 2005; 4:1591–1601.
24. Kornbluth A, Colombel JF, Leighton JA, Loftus E. ICCE consensus for inflammatory bowel disease. Endoscopy 2005; 37:1051–1054.
25. Gralnek IM, Defranchis R, Seidman E, et al
. Development of a capsule endoscopy
scoring index for small bowel mucosal inflammatory change. Aliment Pharmacol Ther 2008; 27:146–154.
26. Goldstein JL, Eisen GM, Lewis B, et al
. Video capsule endoscopy
to prospectively assess small bowel injury with celecoxib, naproxen plus omeprazole and placebo. Clin Gastroenterol Hepatol 2005; 3:133–141.
27. Goldstein JL, Eisen GM, Lewis B, et al
. Small bowel mucosal injury is reduced in healthy subjects treated with celecoxib compared with ibuprophen plus omeprazole as assessed by video capsule endoscopy
. Aliment Pharmacol Therap 2007; 25:1211–1222.
28. FitzSimmons SC, Burkhart GA, Norowitch D, et al
. High-dose pancreatic-enzyme supplements and fibrosing colonopathy in children with cystic fibrosis
. N Engl J Med 1997; 336:1283–1289.
29. Tibble JA, Sigthorsson G, Foster R, Forgacs I, Bjarnason I. Use of surrogate markers of inflammation and Rome criteria to distinguish organic from nonorganic intestinal disease. Gastroenterology 2002; 123:450–460.
30. Bruzzese E, Raia V, Gaudiello G, et al
. Intestinal inflammation is a frequent feature of cystic fibrosis
and is reduced by probiotic administration. Aliment Pharmacol Ther 2004; 20:813–819.
31. Bruzzese E, Raia V, Spagnuolo MI, et al
. Effect of Lactobacillus GG supplementation on pulmonary exacerbations in patients with cystic fibrosis
: a pilot study. Clin Nutr 2007; 26:322–328.