- Email: [email protected]
GI distension in severe ulcerative colitis
THE AMERICAN JOURNAL OF GASTROENTEROLOGY © 2002 by Am. Coll. of Gastroenterology Published by Elsevier Science Inc.
Vol. 97, No. 5, 2002 ISSN 0002-9270/02/$22.00 PII S0002-9270(02)04047-9
GI Distension in Severe Ulcerative Colitis Giovanni Latella, M.D., Piero Vernia, M.D., Angelo Viscido, M.D., Giuseppe Frieri, M.D., Giuseppina Cadau, M.D., Andrea Cocco, M.D., Andrea Cossu, M.D., Ernesto Tomei, M.D., and Renzo Caprilli, M.D. Gastroenterology Unit I, University “La Sapienza,” Rome; and Gastroenterology Unit, University of L’Aquila, L’Aquila, Italy
OBJECTIVES: In previous retrospective studies in patients with severe ulcerative colitis (UC), small bowel distension was found to characterize a subgroup of patients at higher risk for both toxic megacolon (TMC) and multiple organ dysfunction syndrome (MODS). In this study we prospectively evaluated the prevalence of GI distension and its relationship to clinical outcome in patients with severe UC. METHODS: Of 109 consecutive inpatients with acute UC (admitted to the GI Unit of the University of Rome during the period 1995–2000), 45 had severe colitis. Routine blood tests and acid-base balance and plain abdominal film evaluations were performed upon admission and repeated every 1–3 days. The gas content of the stomach and small and large intestines was evaluated on plain abdominal films. All patients were submitted to the standard Oxford intensive medical regimen; clinical improvement, occurrence of major complications, need for surgery, and mortality were evaluated. Statistical analysis was carried out using Student’s t, 2, Fisher’s exact, Mann-Whitney, and Wilcoxon rank sum tests, when appropriate. RESULTS: Of 45 patients with severe UC, 24 (53%) had GI distension. Three of these 24 patients had TMC on admission (all underwent surgery and survived), 21 showed increased GI gas content (four developed TMC 1– 4 days after the detection of GI distension and were operated on, two developed MODS and died, and eight did not improve but were submitted to surgery and survived). None of the 21 patients with normal GI gas content had complications; all survived (five did not improve and required surgery). CONCLUSIONS: In severe UC, persistent GI distension characterized a subgroup of patients with poor response to medical therapy and at higher risk for TMC and of need for surgery. The development of MODS was the most important predicting factor for fatal outcome. (Am J Gastroenterol 2002;97:1169 –1175. © 2002 by Am. Coll. of Gastroenterology)
INTRODUCTION The most life-threatening complication of severe ulcerative colitis (UC) is considered to be toxic megacolon (TMC).
This condition is defined as total or segmental nonobstructive hypotonic dilation of the colon, exceeding 6 cm in diameter in the transverse colon on plain abdominal film, with or without signs of systemic toxicity (1–3). Concomitant distension of the small bowel and, less frequently, the stomach on plain abdominal films had been reported in the 1950s, but little attention has been paid to this finding (4 –7). Subsequent studies showed, however, that small bowel distension may precede colonic dilation, and this was termed impending megacolon (8, 9). In a retrospective study, this condition was found in about 50% of patients with severe UC, characterizing a subgroup of patients at high risk for the development of TMC, as 20% of these patients developed colonic dilation 1– 6 days after the detection of small bowel distension (10). The clinical importance of this feature was confirmed later by others (11). Recently, both impending megacolon and TMC were found to be associated with the development of multiple organ dysfunction syndrome (MODS), a condition responsible for the majority of fatal cases in severe UC (12). Taken together, these observations suggest that GI distension is an important clinical feature in severe colitis and may represent a powerful prognostic factor. We have therefore prospectively evaluated the prevalence of GI distension and its relation to clinical outcome in patients with severe UC admitted to our GI unit in the last 5 yr.
MATERIALS AND METHODS The study was prospectively carried out in 109 consecutive patients with UC admitted to the Gastroenterology Unit of the University of Rome “La Sapienza” for an acute attack of UC during the period 1995–2000. Our clinic is a tertiary referral center for inflammatory bowel disease for central/ southern Italy. Of the 109 patients, 45 met the criteria for severe colitis and were included in the study. Diagnosis of colitis was based on clinical, radiological, endoscopic, and histological findings. The extent of colonic lesions was assessed on plain abdominal films according to the criteria of Bartram (13), which consider the extent of fecal residue, evidence of mucosal ulceration, and alteration of haustral pattern. In patients who were operated on, the extent of
1170
Latella et al.
colonic lesions was assessed with surgical specimens. In patients who were not operated on it was confirmed by endoscopy, performed after achieving clinical remission. The disease was classified as pancolitis (lesions extending from the rectum to the cecum), extensive colitis (lesions extending from the rectum to the hepatic flexure), and left colitis (lesions extending from the rectum to the splenic flexure). Severity was defined and classified according to the criteria of Truelove and Witts (14), and MODS according to previously reported criteria (12). Upon admission, routine blood tests, complete evaluation of acid-base balance, and plain abdominal film assessments, in the supine and upright positions, were performed. An endoscopic evaluation of the rectum without air insufflation and with biopsies was performed after the plain abdominal film assessment only in those patients at the first attack of disease. Electrolyte and acid-base balance assays and plain film assessments were repeated every 1–3 days during the acute phase of the disease, depending upon the progress of the patients. A total of 29 clinical and hematological features, studied on the day of admission, were taken into consideration: age and sex of patients; duration and extent of the disease; first attack versus relapse; duration of present attack; bowel actions; bleeding; pulse rate; temperature; erythrocyte sedimentation rate; C-reactive protein (CRP); orosomucoid proteins; white blood cells; red blood cells; Hb; Hct; platelets; total proteins and albumin; serum iron, sodium, potassium, chloride, calcium, and magnesium; and arterial blood pH, bicarbonate, and base excess. The gas content of the stomach and small and large intestines was evaluated on plain abdominal films in the supine position according to criteria described elsewhere (1, 10, 11). Distension of the stomach was defined as increased gas content associated with disappearance of gastric mucosal folds by the consensus of three independent and expert observers. Upon admission patients were classified accordingly into a) patients with normal gastric/small-large bowel gas content, b) patients with increased gastric/small-large bowel gas content (impending megacolon), and c) patients with colonic dilation exceeding 6 cm in diameter in the transverse colon (TMC). All patients were treated according to a standard therapeutic protocol, consisting of bowel rest, total parenteral nutrition (TPN), and i.v. hydrocortisone at a dose of 100 mg q.i.d. Colectomy was promptly performed in TMC patients in those cases of deterioration or failure to improve after 2 days of intensive medical treatment. Patients with severe uncomplicated UC were submtted to surgery if no definite improvement in the clinical condition was achieved within 7–10 days. Outcome of the disease was evaluated, limited to the hospitalization period, as follows: clinical improvement (achievement of remission or reduction of severity to a mild disease); occurrence of major complications, including MODS; need for surgery; and mortality.
AJG – Vol. 97, No. 5, 2002
Figure 1. Plain abdominal radiograph of a patient with total severe UC showing distension of the stomach and colon.
The physicians responsible for clinical decisions were unaware of the study protocol. The plain abdominal radiographs were evaluated by three independent experts who were blind to clinical conditions and outcome of the patients. Statistical analysis was performed using the 2, Fisher’s exact, Student’s t, Mann-Whitney, and Wilcoxon rank sum tests, when appropriate. For all hypotheses tested, two-tailed ps ⬍ 0.05 were considered significant.
RESULTS Of the 45 patients with severe UC, 21 had normal intestinal gas content (46.7%), 21 showed increased intestinal gas content (46.7%), and three had TMC on admission (6.6%). Of 21 patients with increased intestinal gas content, six had only small bowel distension, 14 had both small and large bowel distension, and one had only large bowel distension. Of 14 patients with both small and large bowel distension, four had concomitant abnormal distension of the stomach (Fig. 1) and one had bladder dilation with urinary retention (Fig. 2). These latter features were not observed in the other two groups of patients. No correlation was present between GI gas content and other radiological features of severity. Tables 1 and 2 show the main clinical and laboratory findings in patients with normal GI gas content versus those of patients with GI distension (including the three cases of MTC on admisssion). Only three features differed significantly between the two groups: patients with GI distension
AJG – May, 2002
GI Distension in Severe Ulcerative Colitis
1171
differed between the two groups when data from the three patients with TMC on admission were excluded. All patients in both groups were submitted to TPN. The mean duration of TPN was identical in the two groups (mean values ⫾ SDs ⫽ 8.0 ⫾ 3.2 and 7.3 ⫾ 2.7 days, in patients with normal and increased GI gas content, respectively), and no major complication was observed. No correlation was found between CRP upon admission and the outcome of patients with normal GI gas content. Within patients with increased GI gas content CRP was higher in patients undergoing colectomy than in those not operated on (mean values ⫾ SDs ⫽ 7.2 ⫾ 2.6 and 5.2 ⫾ 1.8 mg/100 ml, p ⬍ 0.047). As far as previous therapy is concerned, 17 of 21 patients with normal GI gas content and 14 of 24 with increased GI gas received oral steroids during the current attack, before admission. The daily dose was identical in the two groups (mean values ⫾ SDs ⫽ 38.5 ⫾ 17.6 and 36.4 ⫾ 12.7 mg of prednisone). Patients with normal GI gas had a slightly longer therapy course than those with GI distension (mean values ⫾ SDs ⫽ 35.2 ⫾ 17.4 and 25.0 ⫾ 16.1 days, p ⬍ 0.048). No drugs affecting GI motility were taken by any patient before admission.
Figure 2. Plain abdominal radiograph of a young woman with total severe UC. Diffuse distension of the small bowel and colon is evident. Urinary bladder distension is also present.
were younger, had lower clinical improvement rates, and needed surgery more frequently than those with normal GI gas content. No clinical and laboratory feature significantly
Outcome of the Attack (Fig. 3) The three patients with TMC on admission underwent surgery (one urgent and two elective operations), and all survived. Of the 21 patients with increased intestinal gas content, four (19%) developed TMC during hospitalization (Fig. 4) despite intensive medical treatment and underwent urgent surgery. One of them, operated on 12 h after the diagnosis of TMC, developed MODS and died. Eight patients in this group did not improve and were submitted to surgery and survived. A further patient, after 5 days of acute severe
Table 1. Characteristics of Patients and Clinical Details Normal GI Gas Content (21 Patients) Age (yr) [mean (SD)] Gender [no. (%)] Males Females Duration of disease (mo) [median (range)] Duration of current attack (days) [median (range)] Clinical course [no. (%)] First attack Relapse Extension of lesions [no. (%)] Pancolitis Extensive colitis Bowel actions 24 h [median (range)] Pulse rate/min [median (range)] Temperature (°C) [median (range)] Clinical improvement [no.(%)] Surgery [no. (%)] Deaths [no. (%)]
42.0 (12.6)
Increased GI Gas Content (24 Patients) 31.7 (11.3)
p 0.007
12 (57.1) 9 (42.9) 30 (10–90) 48 (0.5–240)
10 (41.7) 14 (58.3) 30 (7–90) 21 (0.5–240)
0.300 0.399 0.056
1 (4.8) 20 (95.2)
6 (25.0) 18 (75.0)
0.101
10 (47.6) 11 (52.4) 9 (3–18) 88 (72–110) 37.5 (37–39) 16 (76.2) 5 (23.8) 0
17 (70.8) 7 (29.2) 8 (1–16) 90 (54–120) 38.0 (37–39.5) 8 (33.3) 15 (62.5) 2 (8.3)
0.113 0.471 0.879 0.173 0.006 0.010 0.490
1172
Latella et al.
AJG – Vol. 97, No. 5, 2002
Table 2. Laboratory Findings in Severe UC
ESR (mm/h) CRP (mg/100 ml) Orosomucoid proteins (mg/100 ml) WBCs/mm3 (⫻103) RBCs/mm3 (⫻106) Hb (g/100 ml) Hct (%) PTL/mm3 (⫻103) Total proteins (g/100 ml) Albumin (g/100 ml) Serum iron (g/100 ml) Serum sodium (mEq/L) Serum potassium (mEq/L) Serum chloride (mEq/L) Serum calcium (mg/100 ml) Serum magnesium (mg/100 ml) Arterial blood pH Arterial blood bicarbonate (mmol/L) Arterial blood base excess
Normal GI Gas Content (21 Patients)
Increased GI Gas Content (24 Patients)
p
50 (18–130) 5.8 (0.5–8) 180 (123–320) 9 (4.8–14) 3.7 (2.2–7.9) 11 (6.8–15) 33 (17–45) 356 (45–616) 6 (4.9–8.3) 3.1 (2.1–5.8) 30.5 (8.0–95) 141 (104–148) 4 (3–5.1) 105 (96–122) 8.7 (7.4–10) 1.9 (1.4–2.2) 7.5 (7.4–7.6) 26.4 (23–30) 3.5 (⫺1.3–⫹7)
58 (8–110) 7.0 (0.3–76) 183 (102–430) 10 (2.5–20) 3.8 (2.4–5.1) 11 (7.1–15) 32 (21–44) 361 (92–558) 6 (4.5–7.5) 3.2 (1.4–4.7) 31.6 (13.2–78) 141 (126–146) 4 (2–5.1) 103 (94–112) 8.6 (7.3–10) 2.0 (1.5–2.5) 7.5 (7.4–7.7) 27.3 (22–36) 3.7 (⫺1.7–⫹13)
0.556 0.153 0.622 0.301 0.655 0.481 0.404 0.814 0.963 0.953 0.811 0.922 0.807 0.506 0.541 0.501 0.396 0.099 0.773
Data are shown as medians (ranges). PTL ⫽ platelets; RBC ⫽ red blood cell; WBC ⫽ white blood cell.
disease, developed MODS 12 h after admission and died shortly after before undergoing surgery. In the group with normal intestinal gas, no patient developed TMC or other complications, and all survived. Five of them failed to show any improvement and were submitted to elective surgery. The timing of surgery was identical in the two groups (mean values ⫾ SDs ⫽ 11.5 ⫾ 2.1 and 11.9 ⫾ 11.8 days, in patients with normal and increased GI gas content, respectively). Thus, in the present cohort, only patients with increased intestinal gas upon admission developed TMC during hospitalization. The need for surgery was higher in patients with increased intestinal gas (12/21 [57.1%]) than in those with normal intestinal gas content (5/21 [23.8%]) (p ⬍ 0.05, odds ratio ⫽ 4.267, 95% CI ⫽ 1.134 –16.055). The only two
fatal cases were those of patients with MODS, both showing GI distension on admission. As far as surgical findings in patients undergoing colectomy are concerned, pathological data did not show a significant difference among patients with normal and increased GI gas content. The typical lesions of severe colitis were present in all patients, with the most severe findings usually present in the left colon. Ulcerations were confined to the mucosa and submucosa. Deep and diffuse ulceration extending into the circular muscular layer was observed in the left colon in one of the three patients with TMC on admission, and in the right colon of the patient with increased GI gas content, who first developed TMC and MODS after colectomy. No colonic perforation was found in any patient.
DISCUSSION
Figure 3. Diagram showing the outcome in severe UC patients.
The results of this study on severe UC confirm previous retrospective observations that persistent gastrointestinal distension characterizes a subgroup of patients with poor response to medical therapy and higher risk for the development of TMC (10, 11). These results also confirm that the development of MODS is the most important predicting factor for fatal outcome in severe UC (12), and should be always taken into account in association with other, widely accepted prognostic features, reported elsewhere (15, 16). The pathogenesis of colonic and GI distension in severe colitis is still unclear. Although some data provide possible explanations for the impaired neuromotor function responsible for colonic dilation in widespread colitis, most of them fail to explain the gaseous distension in the uninflamed small bowel and stomach. Lesions in the myenteric plexus and muscle fibers (17–20) and electrolyte/metabolic imbal-
AJG – May, 2002
GI Distension in Severe Ulcerative Colitis
1173
Figure 4. Plain abdominal radiographs of a patient with total severe UC. (Day 1) Gaseous distension of some small bowel loops. (Day 2) Dilation of the colon with a maximum internal diameter exceeding 7 cm in the transverse colon.
ance (21) were considered to contribute to the reduction of colonic motor activity (22–26) and the development of TMC. However, electrolyte imbalance is not invariably present, whereas severe anatomical damage can be excluded in those cases where TMC rapidly subsides. Today the most convincing explanation for TMC seems to be the excessive local production of nitric oxide (NO) and cytokines, and the activation of viscero-visceral reflexes. Clinical and experimental data have been reported suggesting that NO plays a central role in the genesis of colonic distension (27–30). Increased inducible NO synthase (iNOS) activity was detected in inflamed colonic mucosa from patients with UC compared to controls (27–29), with significantly higher levels in TMC then in uncomplicated UC (29). Furthermore, muscle staining for iNOS was observed in TMC but not in uncomplicated colitis. Similarly, in trinitrobenzenesulfonic acid colitis, an experimental model of acute severe colitis associated with colonic dilation, a striking increase in iNOS expression was found, whereas the inhibition of NOS activity by L-nitro-arginine methyl ester, a nonspecific NOS inhibitor, prevented colonic dilation (30). The observation that interleukin 1 (IL-1), tumor necrosis factor-␣, and interferon-␥, which are of prime importance in the inflammatory process of severe colitis, as well as free radicals and lipopolysaccharides are able to induce iNOS expression in human colonic smooth
muscle strips (29 –32) further supports the role of iNOS in the genesis of TMC. The excessive NO generation in the inflamed colon may offer an explanation for distension of the large bowel but not for that of the proximal, uninflamed GI segments. In fact, NOS activity in the dilated, uninflamed colonic segment, in trinitrobenzenesulfonic acid colitis, was of the same order as that of controls (30, 33). Furthermore, it is unlikely that NO exerts its inhibitory effect via the bloodstream because of its extremely short half-life (34). Conversely, this is not the case for cytokines, which are present at high concentrations in the serum both of animals with experimental colitis and of patients with severe UC (35). In these patients, the spilling of cytokines (IL-1, IL-6, IL-8, interferon ␥, tumor necrosis factor ␣) into the bloodstream is responsible for systemic “toxicity” and the so-called systemic inflammatory response syndrome and MODS (12, 36). These cytokines have been demonstrated to suppress neurotransmitter release, which may alter neuromuscular function even at the remote site of inflammation (37– 40). An alternative or additional pathogenetic mechanism that might contribute to GI distension in TMC is the activation of intestinal extrinsic inhibitory reflexes (37, 41). Experimental studies have shown that distension or mucosal irritation of one part of the GI tract can inhibit the motor activity of distant segments through the activation of extrin-
1174
Latella et al.
sic neuronal colon-ileal or ileocolonic inhibitory reflexes (42– 44). Two types of extrinsic reflexes have been identified— namely, the “long reflexes” involving the spinal cord through the dorsal root ganglia and the “short reflexes” involving only the prevertebral ganglia. The activation of visceral afferent fibers occurs through the stimulation of polymodal (chemical, mechanical, etc) sensory receptors by inflammatory mediators and/or visceral distension (44 – 47). The afferent limb of extrinsic inhibitory reflexes is composed of enteric visceral afferents, conveying synaptic inputs to prevertebral ganglion cells, and spinal visceral afferents, sending collateral branch terminals to prevertebral ganglia. It is noteworthy that 90% of the visceral enteric afferent fibers to any of the three prevertebral ganglia (celiac, superior mesenteric, inferior mesenteric) come from the large intestine (48, 49). The interconnections between prevertebral ganglia and the various spinal segments spread the afferent inputs to ganglia not directly reached by afferent fibers, activating an efferent pathway acting on regions of the gut far from affected areas (50). The efferent limb, represented by preganglionic and postganglionic sympathetic nerves, innervates mainly the myenteric plexus (51–53). Its inhibitory effect on GI tone and motility is primarily mediated by noradrenaline-induced inhibition of acetylcholine and release of tachykinins from postganglionic intrinsic cholinergic neurons (54, 55). To summarize, the GI distension characterizing some forms of severe colitis seems to result from the combination of several factors. The increased release of soluble inflammatory mediators (cytokines, reactive oxygen metabolites, eicosanoids, neurotransmitters) in the inflamed colonic mucosa, together with bacterial products, enhances NO generation and activates nerve endings of mucosal afferent fibers. NO generation may lead to muscle relaxation and colonic dilation, whereas the activation of sensory fibers elicits inhibitory reflexes leading to distension of the stomach and small bowel. The distension of small bowel loops might, in turn, aggravate or even trigger the onset of colonic dilation. Finally, the spilling into the systemic bloodstream of both cytokines and bacterial toxins, due to the wide and deep colonic ulcerations that characterize the most severe forms of colitis, may either contribute to GI distension or initiate and perpetuate both systemic inflammatory response syndrome and MODS. The pathogenetic hypothesis proposed for GI distension in severe UC is in keeping with the demonstrated efficacy of steroids, which are potent inhibitors of NO and cytokines, and with the well-known harmful effect of the anticholinergic drugs. It appears, therefore, that distension of the small bowel and/or of the stomach has the same pathophysiological significance as that of the colon. The finding of distension in any segment of the GI tract during the clinical course of colitis should, therefore, be considered as an alarm signal for major complications either local (TMC) or systemic
AJG – Vol. 97, No. 5, 2002
(MODS) and prompt intensive medical treatment. New therapeutic approaches aimed at the selective inhibition of NO production and extrinsic viscero-visceral reflexes should be considered in the future. The timely identification of symptoms and laboratory findings suggesting MODS should prompt admission to an intensive care unit and emergency colectomy. In such instances, surgery should be performed without waiting for the outcome of the traditional 5-day regimen of intensive medical treatment.
ACKNOWLEDGMENT The authors are grateful to Professor Giovanni Corrao, Department of Statistics, University “La Bicocca,” Milan, Italy, for statistical advice. Reprint requests and correspondence: Renzo Caprilli, M.D., Dipartimento di Scienze Cliniche, Cattedra di Gastroenterologia I, Universita` “La Sapienza” di Roma, Policlinico Umberto I, Viale del Policlinico, 00161 Roma, Italy. Received July 17, 2001; accepted Oct. 31, 2001.
REFERENCES 1. Hywel-Jones J, Chapman M. Definition of megacolon in colitis. Gut 1969;10:562–5. 2. Fazio VW. Toxic megacolon in ulcerative colitis and Crohn’s colitis. Clin Gastroenterol 1980;9:389 – 407. 3. Sheth SG, LaMont JT. Toxic megacolon. Lancet 1998;351: 509 –13. 4. McConnell F, Hanelin J, Robbins LL. Plain film diagnosis of fulminating ulcerative colitis. Radiology 1958;71:674 – 82. 5. Wolf BS, Marshak RH. Toxic segmental dilatation of the colon during the course of fulminating ulcerative colitis: Roentgen findings. Am J Roentgenol Radiat Ther Nucl Med 1959;82:985–95. 6. Rice RP. Plain abdominal film roentgenographic diagnosis of ulcerative diseases of the colon. Am J Roentgenol Radiat Ther Nucl Med 1968;104:544 –50. 7. Jalan KN, Circus W, Cord WI. An experience with ulcerative colitis: Toxic dilatation in 55 cases. Gastroenterology 1969; 57:68 – 82. 8. Caprilli R, Vernia P, Colaneri O, Torsoli A. Blood pH: A test for assessment of severity in proctocolitis. Gut 1976;17:763–9. 9. Vernia P, Colaneri O, Tomei E, Caprilli R. Intestinal gas in ulcerative colitis. Dis Colon Rectum 1979;22:346 –9. 10. Caprilli R, Vernia P, Latella G, Torsoli A. Early recognition of toxic megacolon. J Clin Gastroenterol 1987;9:160 – 4. 11. Chew CN, Nolan DJ, Jewell DP. Small bowel gas in severe ulcerative colitis. Gut 1991;32:1535–7. 12. Caprilli R, Latella G, Vernia P, Frieri G. Multiple organ dysfunction in ulcerative colitis. Am J Gastroenterol 2000;95: 1258 – 62. 13. Bartram CI. Radiology in inflammatory bowel disease. New York: Marcel Dekker, 1983:31– 61. 14. Truelove SC, Witts LJ. Cortisone in ulcerative colitis: Final report on a therapeutical trial. Br Med J 1955;2:1041– 8. 15. Jewell DP, Caprilli R, Mortensen N, et al. Indications and timing of surgery for severe ulcerative colitis. Gastroenterol Int 1991;4:161– 4. 16. Hyde GM, Jewell DP. Review article: The management of
AJG – May, 2002
17. 18. 19. 20. 21. 22. 23. 24. 25. 26.
27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.
severe ulcerative colitis. Aliment Pharmacol Ther 1997;11: 419 –24. Storsten KA, Kernohon JW, Bargan JA. The myenteric plexus in chronic ulcerative colitis. Surg Gynecol Obstet 1953;97: 335– 43. Collins SM, Hurst SM, Main C, et al. Effect of inflammation of enteric nerves cytokine-induced changes in neurotransmitter content and release. Ann N Y Acad Sci 1992;664:415–24. Geboes K, Collins S. Structural abnormalities of the nervous system in Crohn’s disease and ulcerative colitis. Neurogastroenterol Motil 1998;10:189 –202. Roth JLA, Valdes-Dapena A, Stein GN, Bockus HL. Toxic megacolon in ulcerative colitis. Gastroenterology 1959;37: 239 –55. Caprilli R, Vernia P, Colaneri O, Frieri G. Risk factors in toxic megacolon. Dig Dis Sci 1980;25:817–22. Kern F, Almy TP, Abbot FK, Bogdonof MD. The motility of the distal colon in non-specific ulcerative colitis. Gastroenterology 1951;19:492–503. Almy TP. Observations on the pathologic physiology of ulcerative colitis. Gastroenterology 1961;10:299 –306. Snape WJ, Materazzo SA, Cohen S. Abnormal gastrocolonic response in patients with ulcerative colitis. Gut 1980;21: 392– 6. Reddy SN, Bazzochi G, Ghan S, et al. Colonic motility and transit in health and ulcerative colitis. Gastroenterology 1991; 101:1289 –97. Koch TR, Carney JA, Go VLW, Szurszewski JH. Spontaneous contraction and some electrophysiological properties of circular muscle from normal sigmoid colon and ulcerative colitis. Gastroenterology 1988;95:77– 84. Middleton SJ, Shorthouse M, Hunter LO. Increased nitric oxide synthesis in ulcerative colitis. Lancet 1993;341:465– 6. Boughton-Smith NK, Evans SM, Hawkey CJ, et al. Nitric oxide synthase activity in ulcerative colitis and Crohn’s disease. Lancet 1993;342:338 – 40. Mourelle M, Casella F, Guarner F, et al. Induction of nitric oxide synthase in colonic smooth muscle from patients with toxic megacolon. Gastroenterology 1995;109:1497–502. Mourelle M, Villaseca J, Guarner F, et al. Toxic dilatation of colon in rat model of colitis is linked to an inducible form of nitric oxide synthase. Am J Physiol 1996;270:G425–30. Cook HT, Cattel V. Role of nitric oxide in immune-mediated diseases. Clin Sci 1996;91:375– 84. Lundberg JON, Lundberg JM, Alving K, Weitzberg E. Nitric oxide and inflammation: The answer is blowing in the wind. Nat Med 1997;3:30 –1. Taddei G, Aggio A, Dalo S, et al. Circular muscle contractile activity in dilated but non inflamed colonic region in TNBS colitis. Gastroenterology 1999;116:G492. Moncada S, Palmer RMJ, Higgs EA. Nitric oxide: Physiology, pathophysiology and pharmacology. Pharmacol Rev 1991;43: 109 – 42. Fiocchi C. Inflammatory bowel disease. Etiology and pathogenesis. Gastroenterology 1998;115:182–205. Karima R, Matsumoto S, Higashi M, Matsushima K. The mononuclear pathogenesis of endotoxic shock and organ failure. Mol Med Today 1999;5:123–32. Caprilli R, Onori L. Pathogenesis of gastrointestinal distension
GI Distension in Severe Ulcerative Colitis
38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48.
49.
50. 51. 52.
53. 54.
55.
1175
in severe ulcerative colitis: A hypothesis. Gastroenterol Int 1992;5:268 –72. Ruhl A, Hurst S, Collins SM. Synergism between interleukins 1 and 6 on noradrenergic nerves in rat myenteric plexus. Gastroenterology 1994;107:993–1001. Jackobson K, McHugh K, Collins SM. The mechanism of altered neural function in a rat model of colitis. Gastroenterology 1997;112:156 – 62. Sbahidi H, Kilbourn RG. Interleukin-2 therapy in renal cancer induced acute colonic distention. Cancer Metastasis Rev 1998; 17:119 –26. Sengupta JN, Gebhart GF. Gastrointestinal afferent fibers and sensation. In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven, 1994:483–519. Furness JB, Costa M. Adynamic ileus, its pathogenesis and treatment. Med Biol 1974;52:82–9. Szurszewski JH, Miller SM. Physiology of prevertebral ganglia. In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven, 1994:795– 877. Basilisco G, Phillips SF. Ileal distention relaxes the canine colon: A model of megacolon? Gastroenterology 1994;106: 606 –14. Bueno L, Fioramonti J, Delvaux M, Frexinos J. Mediators and pharmacology of visceral sensitivity: From basic to clinical investigations. Gastroenterology 1997;112:1714 – 43. Bueno L, Fioramonti J. Effects of inflammatory mediators on gut sensitivity. Can J Gastroenterol 1999;13(suppl A):42A– 6A. Mayer EA, Gebhart GF. Basic and clinical aspects of visceral hyperalgesia. Gastroenterology 1994;107:271–93. Furness JB, Kuramoto H, Messenger JP. Morphological and chemical identification of neurons that project from the colon to the inferior mesenteric ganglia in the guinea pig. J Auton Nerv Syst 1990;31:203–10. Jalnig W, Morrison JFB. Functional properties of spinal afferents supplying abdominal and pelvic organs, with special emphasis on visceral nociception. In: Cervero F, Morrison JFB, eds. Progress in brain research. Amsterdam: Elsevier, 1986:87–114. Grundy D. Speculations on the structure/function relationship for vagal and splanchnic afferent endings supplying the gastrointestinal tract. J Auton Nerv Syst 1988;22:175– 80. Furness JB, Costa M. The adrenergic innervation of the gastrointestinal tract. Ergeb Physiol 1974;69:1–54. Llewellyn-Smith IJ, Wilson AJ, Furness JB, et al. Ultrastructural identification of noradrenergic axons and their distribution within the enteric plexuses of the guinea pig small intestine. J Neurocytol 1981;10:331–52. Messenger JP, Furness JB. Distribution of enteric nerve cells projecting to the superior and inferior mesenteric ganglia of the guinea pig. Cell Tissue Res 1993;271:333–9. Reilly WM, Hoyle CHV, Burnstock G. Pre- and postjunctional adrenoceptor types in the circular muscle of the guinea-pig caecum. Naunyn Schmiedebergs Arch Pharmacol 1987;336: 37– 47. Willis WE. Anatomy and physiology of descending control of nociceptive responses of dorsal horn neurons: A comprehensive review. Prog Brain Res 1988;77:1–29.
Vol. 97, No. 5, 2002 ISSN 0002-9270/02/$22.00 PII S0002-9270(02)04047-9
GI Distension in Severe Ulcerative Colitis Giovanni Latella, M.D., Piero Vernia, M.D., Angelo Viscido, M.D., Giuseppe Frieri, M.D., Giuseppina Cadau, M.D., Andrea Cocco, M.D., Andrea Cossu, M.D., Ernesto Tomei, M.D., and Renzo Caprilli, M.D. Gastroenterology Unit I, University “La Sapienza,” Rome; and Gastroenterology Unit, University of L’Aquila, L’Aquila, Italy
OBJECTIVES: In previous retrospective studies in patients with severe ulcerative colitis (UC), small bowel distension was found to characterize a subgroup of patients at higher risk for both toxic megacolon (TMC) and multiple organ dysfunction syndrome (MODS). In this study we prospectively evaluated the prevalence of GI distension and its relationship to clinical outcome in patients with severe UC. METHODS: Of 109 consecutive inpatients with acute UC (admitted to the GI Unit of the University of Rome during the period 1995–2000), 45 had severe colitis. Routine blood tests and acid-base balance and plain abdominal film evaluations were performed upon admission and repeated every 1–3 days. The gas content of the stomach and small and large intestines was evaluated on plain abdominal films. All patients were submitted to the standard Oxford intensive medical regimen; clinical improvement, occurrence of major complications, need for surgery, and mortality were evaluated. Statistical analysis was carried out using Student’s t, 2, Fisher’s exact, Mann-Whitney, and Wilcoxon rank sum tests, when appropriate. RESULTS: Of 45 patients with severe UC, 24 (53%) had GI distension. Three of these 24 patients had TMC on admission (all underwent surgery and survived), 21 showed increased GI gas content (four developed TMC 1– 4 days after the detection of GI distension and were operated on, two developed MODS and died, and eight did not improve but were submitted to surgery and survived). None of the 21 patients with normal GI gas content had complications; all survived (five did not improve and required surgery). CONCLUSIONS: In severe UC, persistent GI distension characterized a subgroup of patients with poor response to medical therapy and at higher risk for TMC and of need for surgery. The development of MODS was the most important predicting factor for fatal outcome. (Am J Gastroenterol 2002;97:1169 –1175. © 2002 by Am. Coll. of Gastroenterology)
INTRODUCTION The most life-threatening complication of severe ulcerative colitis (UC) is considered to be toxic megacolon (TMC).
This condition is defined as total or segmental nonobstructive hypotonic dilation of the colon, exceeding 6 cm in diameter in the transverse colon on plain abdominal film, with or without signs of systemic toxicity (1–3). Concomitant distension of the small bowel and, less frequently, the stomach on plain abdominal films had been reported in the 1950s, but little attention has been paid to this finding (4 –7). Subsequent studies showed, however, that small bowel distension may precede colonic dilation, and this was termed impending megacolon (8, 9). In a retrospective study, this condition was found in about 50% of patients with severe UC, characterizing a subgroup of patients at high risk for the development of TMC, as 20% of these patients developed colonic dilation 1– 6 days after the detection of small bowel distension (10). The clinical importance of this feature was confirmed later by others (11). Recently, both impending megacolon and TMC were found to be associated with the development of multiple organ dysfunction syndrome (MODS), a condition responsible for the majority of fatal cases in severe UC (12). Taken together, these observations suggest that GI distension is an important clinical feature in severe colitis and may represent a powerful prognostic factor. We have therefore prospectively evaluated the prevalence of GI distension and its relation to clinical outcome in patients with severe UC admitted to our GI unit in the last 5 yr.
MATERIALS AND METHODS The study was prospectively carried out in 109 consecutive patients with UC admitted to the Gastroenterology Unit of the University of Rome “La Sapienza” for an acute attack of UC during the period 1995–2000. Our clinic is a tertiary referral center for inflammatory bowel disease for central/ southern Italy. Of the 109 patients, 45 met the criteria for severe colitis and were included in the study. Diagnosis of colitis was based on clinical, radiological, endoscopic, and histological findings. The extent of colonic lesions was assessed on plain abdominal films according to the criteria of Bartram (13), which consider the extent of fecal residue, evidence of mucosal ulceration, and alteration of haustral pattern. In patients who were operated on, the extent of
1170
Latella et al.
colonic lesions was assessed with surgical specimens. In patients who were not operated on it was confirmed by endoscopy, performed after achieving clinical remission. The disease was classified as pancolitis (lesions extending from the rectum to the cecum), extensive colitis (lesions extending from the rectum to the hepatic flexure), and left colitis (lesions extending from the rectum to the splenic flexure). Severity was defined and classified according to the criteria of Truelove and Witts (14), and MODS according to previously reported criteria (12). Upon admission, routine blood tests, complete evaluation of acid-base balance, and plain abdominal film assessments, in the supine and upright positions, were performed. An endoscopic evaluation of the rectum without air insufflation and with biopsies was performed after the plain abdominal film assessment only in those patients at the first attack of disease. Electrolyte and acid-base balance assays and plain film assessments were repeated every 1–3 days during the acute phase of the disease, depending upon the progress of the patients. A total of 29 clinical and hematological features, studied on the day of admission, were taken into consideration: age and sex of patients; duration and extent of the disease; first attack versus relapse; duration of present attack; bowel actions; bleeding; pulse rate; temperature; erythrocyte sedimentation rate; C-reactive protein (CRP); orosomucoid proteins; white blood cells; red blood cells; Hb; Hct; platelets; total proteins and albumin; serum iron, sodium, potassium, chloride, calcium, and magnesium; and arterial blood pH, bicarbonate, and base excess. The gas content of the stomach and small and large intestines was evaluated on plain abdominal films in the supine position according to criteria described elsewhere (1, 10, 11). Distension of the stomach was defined as increased gas content associated with disappearance of gastric mucosal folds by the consensus of three independent and expert observers. Upon admission patients were classified accordingly into a) patients with normal gastric/small-large bowel gas content, b) patients with increased gastric/small-large bowel gas content (impending megacolon), and c) patients with colonic dilation exceeding 6 cm in diameter in the transverse colon (TMC). All patients were treated according to a standard therapeutic protocol, consisting of bowel rest, total parenteral nutrition (TPN), and i.v. hydrocortisone at a dose of 100 mg q.i.d. Colectomy was promptly performed in TMC patients in those cases of deterioration or failure to improve after 2 days of intensive medical treatment. Patients with severe uncomplicated UC were submtted to surgery if no definite improvement in the clinical condition was achieved within 7–10 days. Outcome of the disease was evaluated, limited to the hospitalization period, as follows: clinical improvement (achievement of remission or reduction of severity to a mild disease); occurrence of major complications, including MODS; need for surgery; and mortality.
AJG – Vol. 97, No. 5, 2002
Figure 1. Plain abdominal radiograph of a patient with total severe UC showing distension of the stomach and colon.
The physicians responsible for clinical decisions were unaware of the study protocol. The plain abdominal radiographs were evaluated by three independent experts who were blind to clinical conditions and outcome of the patients. Statistical analysis was performed using the 2, Fisher’s exact, Student’s t, Mann-Whitney, and Wilcoxon rank sum tests, when appropriate. For all hypotheses tested, two-tailed ps ⬍ 0.05 were considered significant.
RESULTS Of the 45 patients with severe UC, 21 had normal intestinal gas content (46.7%), 21 showed increased intestinal gas content (46.7%), and three had TMC on admission (6.6%). Of 21 patients with increased intestinal gas content, six had only small bowel distension, 14 had both small and large bowel distension, and one had only large bowel distension. Of 14 patients with both small and large bowel distension, four had concomitant abnormal distension of the stomach (Fig. 1) and one had bladder dilation with urinary retention (Fig. 2). These latter features were not observed in the other two groups of patients. No correlation was present between GI gas content and other radiological features of severity. Tables 1 and 2 show the main clinical and laboratory findings in patients with normal GI gas content versus those of patients with GI distension (including the three cases of MTC on admisssion). Only three features differed significantly between the two groups: patients with GI distension
AJG – May, 2002
GI Distension in Severe Ulcerative Colitis
1171
differed between the two groups when data from the three patients with TMC on admission were excluded. All patients in both groups were submitted to TPN. The mean duration of TPN was identical in the two groups (mean values ⫾ SDs ⫽ 8.0 ⫾ 3.2 and 7.3 ⫾ 2.7 days, in patients with normal and increased GI gas content, respectively), and no major complication was observed. No correlation was found between CRP upon admission and the outcome of patients with normal GI gas content. Within patients with increased GI gas content CRP was higher in patients undergoing colectomy than in those not operated on (mean values ⫾ SDs ⫽ 7.2 ⫾ 2.6 and 5.2 ⫾ 1.8 mg/100 ml, p ⬍ 0.047). As far as previous therapy is concerned, 17 of 21 patients with normal GI gas content and 14 of 24 with increased GI gas received oral steroids during the current attack, before admission. The daily dose was identical in the two groups (mean values ⫾ SDs ⫽ 38.5 ⫾ 17.6 and 36.4 ⫾ 12.7 mg of prednisone). Patients with normal GI gas had a slightly longer therapy course than those with GI distension (mean values ⫾ SDs ⫽ 35.2 ⫾ 17.4 and 25.0 ⫾ 16.1 days, p ⬍ 0.048). No drugs affecting GI motility were taken by any patient before admission.
Figure 2. Plain abdominal radiograph of a young woman with total severe UC. Diffuse distension of the small bowel and colon is evident. Urinary bladder distension is also present.
were younger, had lower clinical improvement rates, and needed surgery more frequently than those with normal GI gas content. No clinical and laboratory feature significantly
Outcome of the Attack (Fig. 3) The three patients with TMC on admission underwent surgery (one urgent and two elective operations), and all survived. Of the 21 patients with increased intestinal gas content, four (19%) developed TMC during hospitalization (Fig. 4) despite intensive medical treatment and underwent urgent surgery. One of them, operated on 12 h after the diagnosis of TMC, developed MODS and died. Eight patients in this group did not improve and were submitted to surgery and survived. A further patient, after 5 days of acute severe
Table 1. Characteristics of Patients and Clinical Details Normal GI Gas Content (21 Patients) Age (yr) [mean (SD)] Gender [no. (%)] Males Females Duration of disease (mo) [median (range)] Duration of current attack (days) [median (range)] Clinical course [no. (%)] First attack Relapse Extension of lesions [no. (%)] Pancolitis Extensive colitis Bowel actions 24 h [median (range)] Pulse rate/min [median (range)] Temperature (°C) [median (range)] Clinical improvement [no.(%)] Surgery [no. (%)] Deaths [no. (%)]
42.0 (12.6)
Increased GI Gas Content (24 Patients) 31.7 (11.3)
p 0.007
12 (57.1) 9 (42.9) 30 (10–90) 48 (0.5–240)
10 (41.7) 14 (58.3) 30 (7–90) 21 (0.5–240)
0.300 0.399 0.056
1 (4.8) 20 (95.2)
6 (25.0) 18 (75.0)
0.101
10 (47.6) 11 (52.4) 9 (3–18) 88 (72–110) 37.5 (37–39) 16 (76.2) 5 (23.8) 0
17 (70.8) 7 (29.2) 8 (1–16) 90 (54–120) 38.0 (37–39.5) 8 (33.3) 15 (62.5) 2 (8.3)
0.113 0.471 0.879 0.173 0.006 0.010 0.490
1172
Latella et al.
AJG – Vol. 97, No. 5, 2002
Table 2. Laboratory Findings in Severe UC
ESR (mm/h) CRP (mg/100 ml) Orosomucoid proteins (mg/100 ml) WBCs/mm3 (⫻103) RBCs/mm3 (⫻106) Hb (g/100 ml) Hct (%) PTL/mm3 (⫻103) Total proteins (g/100 ml) Albumin (g/100 ml) Serum iron (g/100 ml) Serum sodium (mEq/L) Serum potassium (mEq/L) Serum chloride (mEq/L) Serum calcium (mg/100 ml) Serum magnesium (mg/100 ml) Arterial blood pH Arterial blood bicarbonate (mmol/L) Arterial blood base excess
Normal GI Gas Content (21 Patients)
Increased GI Gas Content (24 Patients)
p
50 (18–130) 5.8 (0.5–8) 180 (123–320) 9 (4.8–14) 3.7 (2.2–7.9) 11 (6.8–15) 33 (17–45) 356 (45–616) 6 (4.9–8.3) 3.1 (2.1–5.8) 30.5 (8.0–95) 141 (104–148) 4 (3–5.1) 105 (96–122) 8.7 (7.4–10) 1.9 (1.4–2.2) 7.5 (7.4–7.6) 26.4 (23–30) 3.5 (⫺1.3–⫹7)
58 (8–110) 7.0 (0.3–76) 183 (102–430) 10 (2.5–20) 3.8 (2.4–5.1) 11 (7.1–15) 32 (21–44) 361 (92–558) 6 (4.5–7.5) 3.2 (1.4–4.7) 31.6 (13.2–78) 141 (126–146) 4 (2–5.1) 103 (94–112) 8.6 (7.3–10) 2.0 (1.5–2.5) 7.5 (7.4–7.7) 27.3 (22–36) 3.7 (⫺1.7–⫹13)
0.556 0.153 0.622 0.301 0.655 0.481 0.404 0.814 0.963 0.953 0.811 0.922 0.807 0.506 0.541 0.501 0.396 0.099 0.773
Data are shown as medians (ranges). PTL ⫽ platelets; RBC ⫽ red blood cell; WBC ⫽ white blood cell.
disease, developed MODS 12 h after admission and died shortly after before undergoing surgery. In the group with normal intestinal gas, no patient developed TMC or other complications, and all survived. Five of them failed to show any improvement and were submitted to elective surgery. The timing of surgery was identical in the two groups (mean values ⫾ SDs ⫽ 11.5 ⫾ 2.1 and 11.9 ⫾ 11.8 days, in patients with normal and increased GI gas content, respectively). Thus, in the present cohort, only patients with increased intestinal gas upon admission developed TMC during hospitalization. The need for surgery was higher in patients with increased intestinal gas (12/21 [57.1%]) than in those with normal intestinal gas content (5/21 [23.8%]) (p ⬍ 0.05, odds ratio ⫽ 4.267, 95% CI ⫽ 1.134 –16.055). The only two
fatal cases were those of patients with MODS, both showing GI distension on admission. As far as surgical findings in patients undergoing colectomy are concerned, pathological data did not show a significant difference among patients with normal and increased GI gas content. The typical lesions of severe colitis were present in all patients, with the most severe findings usually present in the left colon. Ulcerations were confined to the mucosa and submucosa. Deep and diffuse ulceration extending into the circular muscular layer was observed in the left colon in one of the three patients with TMC on admission, and in the right colon of the patient with increased GI gas content, who first developed TMC and MODS after colectomy. No colonic perforation was found in any patient.
DISCUSSION
Figure 3. Diagram showing the outcome in severe UC patients.
The results of this study on severe UC confirm previous retrospective observations that persistent gastrointestinal distension characterizes a subgroup of patients with poor response to medical therapy and higher risk for the development of TMC (10, 11). These results also confirm that the development of MODS is the most important predicting factor for fatal outcome in severe UC (12), and should be always taken into account in association with other, widely accepted prognostic features, reported elsewhere (15, 16). The pathogenesis of colonic and GI distension in severe colitis is still unclear. Although some data provide possible explanations for the impaired neuromotor function responsible for colonic dilation in widespread colitis, most of them fail to explain the gaseous distension in the uninflamed small bowel and stomach. Lesions in the myenteric plexus and muscle fibers (17–20) and electrolyte/metabolic imbal-
AJG – May, 2002
GI Distension in Severe Ulcerative Colitis
1173
Figure 4. Plain abdominal radiographs of a patient with total severe UC. (Day 1) Gaseous distension of some small bowel loops. (Day 2) Dilation of the colon with a maximum internal diameter exceeding 7 cm in the transverse colon.
ance (21) were considered to contribute to the reduction of colonic motor activity (22–26) and the development of TMC. However, electrolyte imbalance is not invariably present, whereas severe anatomical damage can be excluded in those cases where TMC rapidly subsides. Today the most convincing explanation for TMC seems to be the excessive local production of nitric oxide (NO) and cytokines, and the activation of viscero-visceral reflexes. Clinical and experimental data have been reported suggesting that NO plays a central role in the genesis of colonic distension (27–30). Increased inducible NO synthase (iNOS) activity was detected in inflamed colonic mucosa from patients with UC compared to controls (27–29), with significantly higher levels in TMC then in uncomplicated UC (29). Furthermore, muscle staining for iNOS was observed in TMC but not in uncomplicated colitis. Similarly, in trinitrobenzenesulfonic acid colitis, an experimental model of acute severe colitis associated with colonic dilation, a striking increase in iNOS expression was found, whereas the inhibition of NOS activity by L-nitro-arginine methyl ester, a nonspecific NOS inhibitor, prevented colonic dilation (30). The observation that interleukin 1 (IL-1), tumor necrosis factor-␣, and interferon-␥, which are of prime importance in the inflammatory process of severe colitis, as well as free radicals and lipopolysaccharides are able to induce iNOS expression in human colonic smooth
muscle strips (29 –32) further supports the role of iNOS in the genesis of TMC. The excessive NO generation in the inflamed colon may offer an explanation for distension of the large bowel but not for that of the proximal, uninflamed GI segments. In fact, NOS activity in the dilated, uninflamed colonic segment, in trinitrobenzenesulfonic acid colitis, was of the same order as that of controls (30, 33). Furthermore, it is unlikely that NO exerts its inhibitory effect via the bloodstream because of its extremely short half-life (34). Conversely, this is not the case for cytokines, which are present at high concentrations in the serum both of animals with experimental colitis and of patients with severe UC (35). In these patients, the spilling of cytokines (IL-1, IL-6, IL-8, interferon ␥, tumor necrosis factor ␣) into the bloodstream is responsible for systemic “toxicity” and the so-called systemic inflammatory response syndrome and MODS (12, 36). These cytokines have been demonstrated to suppress neurotransmitter release, which may alter neuromuscular function even at the remote site of inflammation (37– 40). An alternative or additional pathogenetic mechanism that might contribute to GI distension in TMC is the activation of intestinal extrinsic inhibitory reflexes (37, 41). Experimental studies have shown that distension or mucosal irritation of one part of the GI tract can inhibit the motor activity of distant segments through the activation of extrin-
1174
Latella et al.
sic neuronal colon-ileal or ileocolonic inhibitory reflexes (42– 44). Two types of extrinsic reflexes have been identified— namely, the “long reflexes” involving the spinal cord through the dorsal root ganglia and the “short reflexes” involving only the prevertebral ganglia. The activation of visceral afferent fibers occurs through the stimulation of polymodal (chemical, mechanical, etc) sensory receptors by inflammatory mediators and/or visceral distension (44 – 47). The afferent limb of extrinsic inhibitory reflexes is composed of enteric visceral afferents, conveying synaptic inputs to prevertebral ganglion cells, and spinal visceral afferents, sending collateral branch terminals to prevertebral ganglia. It is noteworthy that 90% of the visceral enteric afferent fibers to any of the three prevertebral ganglia (celiac, superior mesenteric, inferior mesenteric) come from the large intestine (48, 49). The interconnections between prevertebral ganglia and the various spinal segments spread the afferent inputs to ganglia not directly reached by afferent fibers, activating an efferent pathway acting on regions of the gut far from affected areas (50). The efferent limb, represented by preganglionic and postganglionic sympathetic nerves, innervates mainly the myenteric plexus (51–53). Its inhibitory effect on GI tone and motility is primarily mediated by noradrenaline-induced inhibition of acetylcholine and release of tachykinins from postganglionic intrinsic cholinergic neurons (54, 55). To summarize, the GI distension characterizing some forms of severe colitis seems to result from the combination of several factors. The increased release of soluble inflammatory mediators (cytokines, reactive oxygen metabolites, eicosanoids, neurotransmitters) in the inflamed colonic mucosa, together with bacterial products, enhances NO generation and activates nerve endings of mucosal afferent fibers. NO generation may lead to muscle relaxation and colonic dilation, whereas the activation of sensory fibers elicits inhibitory reflexes leading to distension of the stomach and small bowel. The distension of small bowel loops might, in turn, aggravate or even trigger the onset of colonic dilation. Finally, the spilling into the systemic bloodstream of both cytokines and bacterial toxins, due to the wide and deep colonic ulcerations that characterize the most severe forms of colitis, may either contribute to GI distension or initiate and perpetuate both systemic inflammatory response syndrome and MODS. The pathogenetic hypothesis proposed for GI distension in severe UC is in keeping with the demonstrated efficacy of steroids, which are potent inhibitors of NO and cytokines, and with the well-known harmful effect of the anticholinergic drugs. It appears, therefore, that distension of the small bowel and/or of the stomach has the same pathophysiological significance as that of the colon. The finding of distension in any segment of the GI tract during the clinical course of colitis should, therefore, be considered as an alarm signal for major complications either local (TMC) or systemic
AJG – Vol. 97, No. 5, 2002
(MODS) and prompt intensive medical treatment. New therapeutic approaches aimed at the selective inhibition of NO production and extrinsic viscero-visceral reflexes should be considered in the future. The timely identification of symptoms and laboratory findings suggesting MODS should prompt admission to an intensive care unit and emergency colectomy. In such instances, surgery should be performed without waiting for the outcome of the traditional 5-day regimen of intensive medical treatment.
ACKNOWLEDGMENT The authors are grateful to Professor Giovanni Corrao, Department of Statistics, University “La Bicocca,” Milan, Italy, for statistical advice. Reprint requests and correspondence: Renzo Caprilli, M.D., Dipartimento di Scienze Cliniche, Cattedra di Gastroenterologia I, Universita` “La Sapienza” di Roma, Policlinico Umberto I, Viale del Policlinico, 00161 Roma, Italy. Received July 17, 2001; accepted Oct. 31, 2001.
REFERENCES 1. Hywel-Jones J, Chapman M. Definition of megacolon in colitis. Gut 1969;10:562–5. 2. Fazio VW. Toxic megacolon in ulcerative colitis and Crohn’s colitis. Clin Gastroenterol 1980;9:389 – 407. 3. Sheth SG, LaMont JT. Toxic megacolon. Lancet 1998;351: 509 –13. 4. McConnell F, Hanelin J, Robbins LL. Plain film diagnosis of fulminating ulcerative colitis. Radiology 1958;71:674 – 82. 5. Wolf BS, Marshak RH. Toxic segmental dilatation of the colon during the course of fulminating ulcerative colitis: Roentgen findings. Am J Roentgenol Radiat Ther Nucl Med 1959;82:985–95. 6. Rice RP. Plain abdominal film roentgenographic diagnosis of ulcerative diseases of the colon. Am J Roentgenol Radiat Ther Nucl Med 1968;104:544 –50. 7. Jalan KN, Circus W, Cord WI. An experience with ulcerative colitis: Toxic dilatation in 55 cases. Gastroenterology 1969; 57:68 – 82. 8. Caprilli R, Vernia P, Colaneri O, Torsoli A. Blood pH: A test for assessment of severity in proctocolitis. Gut 1976;17:763–9. 9. Vernia P, Colaneri O, Tomei E, Caprilli R. Intestinal gas in ulcerative colitis. Dis Colon Rectum 1979;22:346 –9. 10. Caprilli R, Vernia P, Latella G, Torsoli A. Early recognition of toxic megacolon. J Clin Gastroenterol 1987;9:160 – 4. 11. Chew CN, Nolan DJ, Jewell DP. Small bowel gas in severe ulcerative colitis. Gut 1991;32:1535–7. 12. Caprilli R, Latella G, Vernia P, Frieri G. Multiple organ dysfunction in ulcerative colitis. Am J Gastroenterol 2000;95: 1258 – 62. 13. Bartram CI. Radiology in inflammatory bowel disease. New York: Marcel Dekker, 1983:31– 61. 14. Truelove SC, Witts LJ. Cortisone in ulcerative colitis: Final report on a therapeutical trial. Br Med J 1955;2:1041– 8. 15. Jewell DP, Caprilli R, Mortensen N, et al. Indications and timing of surgery for severe ulcerative colitis. Gastroenterol Int 1991;4:161– 4. 16. Hyde GM, Jewell DP. Review article: The management of
AJG – May, 2002
17. 18. 19. 20. 21. 22. 23. 24. 25. 26.
27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.
severe ulcerative colitis. Aliment Pharmacol Ther 1997;11: 419 –24. Storsten KA, Kernohon JW, Bargan JA. The myenteric plexus in chronic ulcerative colitis. Surg Gynecol Obstet 1953;97: 335– 43. Collins SM, Hurst SM, Main C, et al. Effect of inflammation of enteric nerves cytokine-induced changes in neurotransmitter content and release. Ann N Y Acad Sci 1992;664:415–24. Geboes K, Collins S. Structural abnormalities of the nervous system in Crohn’s disease and ulcerative colitis. Neurogastroenterol Motil 1998;10:189 –202. Roth JLA, Valdes-Dapena A, Stein GN, Bockus HL. Toxic megacolon in ulcerative colitis. Gastroenterology 1959;37: 239 –55. Caprilli R, Vernia P, Colaneri O, Frieri G. Risk factors in toxic megacolon. Dig Dis Sci 1980;25:817–22. Kern F, Almy TP, Abbot FK, Bogdonof MD. The motility of the distal colon in non-specific ulcerative colitis. Gastroenterology 1951;19:492–503. Almy TP. Observations on the pathologic physiology of ulcerative colitis. Gastroenterology 1961;10:299 –306. Snape WJ, Materazzo SA, Cohen S. Abnormal gastrocolonic response in patients with ulcerative colitis. Gut 1980;21: 392– 6. Reddy SN, Bazzochi G, Ghan S, et al. Colonic motility and transit in health and ulcerative colitis. Gastroenterology 1991; 101:1289 –97. Koch TR, Carney JA, Go VLW, Szurszewski JH. Spontaneous contraction and some electrophysiological properties of circular muscle from normal sigmoid colon and ulcerative colitis. Gastroenterology 1988;95:77– 84. Middleton SJ, Shorthouse M, Hunter LO. Increased nitric oxide synthesis in ulcerative colitis. Lancet 1993;341:465– 6. Boughton-Smith NK, Evans SM, Hawkey CJ, et al. Nitric oxide synthase activity in ulcerative colitis and Crohn’s disease. Lancet 1993;342:338 – 40. Mourelle M, Casella F, Guarner F, et al. Induction of nitric oxide synthase in colonic smooth muscle from patients with toxic megacolon. Gastroenterology 1995;109:1497–502. Mourelle M, Villaseca J, Guarner F, et al. Toxic dilatation of colon in rat model of colitis is linked to an inducible form of nitric oxide synthase. Am J Physiol 1996;270:G425–30. Cook HT, Cattel V. Role of nitric oxide in immune-mediated diseases. Clin Sci 1996;91:375– 84. Lundberg JON, Lundberg JM, Alving K, Weitzberg E. Nitric oxide and inflammation: The answer is blowing in the wind. Nat Med 1997;3:30 –1. Taddei G, Aggio A, Dalo S, et al. Circular muscle contractile activity in dilated but non inflamed colonic region in TNBS colitis. Gastroenterology 1999;116:G492. Moncada S, Palmer RMJ, Higgs EA. Nitric oxide: Physiology, pathophysiology and pharmacology. Pharmacol Rev 1991;43: 109 – 42. Fiocchi C. Inflammatory bowel disease. Etiology and pathogenesis. Gastroenterology 1998;115:182–205. Karima R, Matsumoto S, Higashi M, Matsushima K. The mononuclear pathogenesis of endotoxic shock and organ failure. Mol Med Today 1999;5:123–32. Caprilli R, Onori L. Pathogenesis of gastrointestinal distension
GI Distension in Severe Ulcerative Colitis
38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48.
49.
50. 51. 52.
53. 54.
55.
1175
in severe ulcerative colitis: A hypothesis. Gastroenterol Int 1992;5:268 –72. Ruhl A, Hurst S, Collins SM. Synergism between interleukins 1 and 6 on noradrenergic nerves in rat myenteric plexus. Gastroenterology 1994;107:993–1001. Jackobson K, McHugh K, Collins SM. The mechanism of altered neural function in a rat model of colitis. Gastroenterology 1997;112:156 – 62. Sbahidi H, Kilbourn RG. Interleukin-2 therapy in renal cancer induced acute colonic distention. Cancer Metastasis Rev 1998; 17:119 –26. Sengupta JN, Gebhart GF. Gastrointestinal afferent fibers and sensation. In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven, 1994:483–519. Furness JB, Costa M. Adynamic ileus, its pathogenesis and treatment. Med Biol 1974;52:82–9. Szurszewski JH, Miller SM. Physiology of prevertebral ganglia. In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven, 1994:795– 877. Basilisco G, Phillips SF. Ileal distention relaxes the canine colon: A model of megacolon? Gastroenterology 1994;106: 606 –14. Bueno L, Fioramonti J, Delvaux M, Frexinos J. Mediators and pharmacology of visceral sensitivity: From basic to clinical investigations. Gastroenterology 1997;112:1714 – 43. Bueno L, Fioramonti J. Effects of inflammatory mediators on gut sensitivity. Can J Gastroenterol 1999;13(suppl A):42A– 6A. Mayer EA, Gebhart GF. Basic and clinical aspects of visceral hyperalgesia. Gastroenterology 1994;107:271–93. Furness JB, Kuramoto H, Messenger JP. Morphological and chemical identification of neurons that project from the colon to the inferior mesenteric ganglia in the guinea pig. J Auton Nerv Syst 1990;31:203–10. Jalnig W, Morrison JFB. Functional properties of spinal afferents supplying abdominal and pelvic organs, with special emphasis on visceral nociception. In: Cervero F, Morrison JFB, eds. Progress in brain research. Amsterdam: Elsevier, 1986:87–114. Grundy D. Speculations on the structure/function relationship for vagal and splanchnic afferent endings supplying the gastrointestinal tract. J Auton Nerv Syst 1988;22:175– 80. Furness JB, Costa M. The adrenergic innervation of the gastrointestinal tract. Ergeb Physiol 1974;69:1–54. Llewellyn-Smith IJ, Wilson AJ, Furness JB, et al. Ultrastructural identification of noradrenergic axons and their distribution within the enteric plexuses of the guinea pig small intestine. J Neurocytol 1981;10:331–52. Messenger JP, Furness JB. Distribution of enteric nerve cells projecting to the superior and inferior mesenteric ganglia of the guinea pig. Cell Tissue Res 1993;271:333–9. Reilly WM, Hoyle CHV, Burnstock G. Pre- and postjunctional adrenoceptor types in the circular muscle of the guinea-pig caecum. Naunyn Schmiedebergs Arch Pharmacol 1987;336: 37– 47. Willis WE. Anatomy and physiology of descending control of nociceptive responses of dorsal horn neurons: A comprehensive review. Prog Brain Res 1988;77:1–29.