This study shows that ileitis in UC patients undergoing RPC is not rare. It represents a distinct disease-specific subset of patients who are more likely to be younger, treatment-resistant, progress to fulminant stage, and require surgery early. They are also at a higher risk for the subsequent development of pouch-itis after RPC.
Ileitis was originally underreported in our series. In the eight patients in whom BWI was overlooked originally, cuts from the most proximal ileal margin were obtained and stained as per protocol but were not reported on.
A possible bias in the incidence of BWI might have been introduced by the fact that this study group is composed exclusively of patients who underwent RPC. This group of patients represents at most 20 percent to 30 percent of all patients with UC. Nevertheless, the incidence of BWI (22 percent) in this series is substantiated by similar figures reported in other studies.6,13–15
As the histologic examination was limited to the proctocolectomy specimens and only a small segment of the terminal ileum was available; it was not possible to assess the exact extent of ileal involvement. In autopsy studies the extent of ileal inflammation varied from 3 to 45 cm.1,16 For the same reason we do not know whether this involvement occurred early in the course of the colitis or whether it was a terminal event in the downhill progress of the disease.
We found no case of BWI without concomitant pancolitis, and the morphology of BWI was milder but similar to that of colitis. Patients with BWI required surgical intervention earlier than patients without BWI evidenced by a shorter duration of colitis before surgery; many of them progressed to or presented as fulminant colitis and required urgent colectomy. Whether the presence of BWI had influenced the natural course of UC or its response to therapy is unclear.
Patients with BWI are significantly more likely to develop pouchitis after RPC. The multivariate analyses confirmed that this association is independent of the extent of colitis. Therefore, the strong association of BWI with pouchitis cannot be attributed to the accompanying pancolitis, and the risk potential of BWI was presumably assigned to that of pancolitis in a number of patients in earlier studies which did not involve detailed assessment of the terminal ileum.
This finding is in contrast to that of a study by Gustavsson et al.,13 who examined the relation of radiographic and clinical presence of BWI with subsequent pouchitis. The authors concluded that the presence of BWI did not predispose the patient to later development of pouchitis. This result is at variance with our study; the difference could be explained in part by the present study’s use of blinded and strict histologic criteria to diagnose or exclude BWI and clinical, endoscopic, and histologic criteria to diagnose pouch-itis, as opposed to radiologic or clinical criteria for ileitis and clinical diagnosis of pouchitis with no histologic confirmation. Also, the follow-up of their study was much shorter than ours.
In support of our findings, Schmidt et al.6 compared the histopathology of the preoperative terminal ileum with pouch biopsies in the absence of clinical pouchitis and found that the presence of active inflammation, high eosinophil counts, and villous blunting in terminal ileal resection specimens were associated with significantly higher pouch biopsy active inflammation scores.
We found a strong association between the presence of BWI and PSC in UC patients. The multivariate analyses did not demonstrate an independent association between PSC and pouchitis. This is likely a result of the small number of patients with PSC in our study. PSC has previously been reported as an independent risk factor for pouchitis in larger studies,17 and recent findings also substantiate the link between PSC and BWI.18
Defects have been detected in the excretion and enterohepatic circulation of bile acids in patients with PSC.19 Bile acids are absorbed in the terminal ileum and may therefore play a role in the pathogenesis of BWI. Alternatively, the inflammatory process in the mucosa of patients with BWI may influence the enterohepatic circulations of bile acids, increasing the risk of PSC. It has been speculated that alterations in the metabolism of bile acids are involved in the pathogenesis of pouchitis.20
As a means of explaining the simultaneous presence of BWI, PSC, and pouchitis, one might speculate that some immunologic, metabolic, or genetic defect is the common underlying cause for all three pathologic changes.
Patients with previous BWI have marked degrees of villous atrophy (type C adaptive response). Setti-Carraro et al.12 reported that patients with type C adaptive response could be identified on histologic criteria within weeks of closure of ileostomy and were those exclusively at risk of chronic pouchitis. Merret et al.21 reported a linear correlation between the degree of villous atrophy and pouch mucosa permeability. Increased pouch mucosa permeability may lead to a reduction in barrier function and has been hypothesized in the pathogenesis of both pouchitis and extraintestinal manifestations of inflammatory bowel disease.21–23
The current treatment of pouchitis has mainly been empirical and directed at presumed bacterial or inflammatory mechanisms. The mainstays of therapy include antibiotics or anti-inflammatory agents. The results of this study, however, indicate that further research into the specific etiology and pathogenesis of this unique form of inflammation in UC could lead to more focused therapy of pouchitis. Patients with BWI should have detailed counseling on their higher risk of developing pouchitis, its effect on their quality of life, and the potential need for long-term therapy and the risk of failure. They should also undergo more regular postoperative surveillance and may be considered for early and more focused therapeutic strategies. It would be useful to investigate whether starting prophylactic therapy in patients with BWI before or immediately after pouch construction would prevent or decrease the incidence of pouchitis.
Classically the inflammation associated with UC is thought to involve only the colon and the presence of ileal inflammation is unexplained by the conventional understanding of the disease. Furthermore, if and when inflammation is noted in the small intestine, it is not thought to have any clinical or pathologic significance.13
In this study the presence of BWI did not correlate strongly with the severity of colitis, was associated with type C adaptive response, showed a strong association with PSC, and was an independent risk factor for the subsequent development of pouchitis. Furthermore, Heuschen et al.14 reported that the presence of BWI is strongly associated with colorectal carcinoma, and multiple tumor growth was significantly more frequent in patients with pancolitis and BWI than in patients with pancolitis without BWI in their series. Reports also have suggested that BWI has a premalignant potential, and several cases of de novo pouch, ileostomy, and proximal small bowel carcinomas in patients with BWI were reported.24–28 The ileum of one of our patients with BWI underwent dysplastic changes similar to that of the corresponding colon.
The findings of this study, supported by those studies mentioned above, suggest that the presence of ileitis in UC represents a primary involvement of the terminal ileum by the same mucosal inflammation as in the colon, refuting the “backwash” theory and suggesting that ileitis in UC is not necessarily a benign phenomenon.
It is unclear whether modification of the surgical technique, such as resecting more ileum in patients with terminal ileum inflammation, delaying the pouch procedure, or formation of the pouch with immediate prophylactic therapy, in such patients would affect the subsequent postoperative outcome. In our series we have taken biopsies above the pouch, at the top, middle, and bottom, and anterior and posterior of the pouch. The final topographic analysis of these data may help us understand if there is such a gradient of tissue susceptibility. Otherwise, a well-designed clinical trial will have to answer these questions.
Furthermore, by selecting a higher-risk group of UC patients, this study has implications for preoperative counseling, more intensive postoperative surveillance, and early and more specific preventative and therapeutic strategies aimed at a more targeted population.
The true incidence, natural history, and clinical significance of ileitis in patients with UC can be determined only if a routine and detailed microscopic characterization of the terminal ileum is performed in every patient with UC from presentation. Colonoscopy and retrograde ileoscopy is achievable in 85 percent to 90 percent of the patients.29–31 New modalities like MRI with enteroclysis,32 wireless capsule endoscopy,33 virtual enteroscopy,34 or confocal colonoscopy35 may also be useful. Identification of a molecular immunogenetic or immunophenotypic marker for this topographic variant of UC would be helpful.