Fibrosing colonopathy is a newly described complication. Sonographic Evaluation of Bowel Wall Thickness in Patients With Cystic Fibrosis

Transcription

1 J Clin Gastroenterol 2003;37(1): Lippincott Williams & Wilkins, Inc. Sonographic Evaluation of Bowel Wall Thickness in Patients With Cystic Fibrosis Irmela Dialer, MD, Cornelia Hundt, MD, Rose-Marie Bertele-Harms, MD, and Hinrich Karsten Harms, MD Abstract Goals and Background: Fibrosing colonopathy causing severe thickening of the colon wall was recently described in cystic fibrosis (CF). Since it has been suspected that subclinical colon-wall thickening is a common feature in CF patients, bowel-wall thickness was measured in a series of patients and compared with controls. Additionally, possible clinical factors influencing wall thickness were investigated. Study: In 83 CF patients and 31 controls transabdominal ultrasound was performed in the terminal ileum, cecum, ascending and descending colon. The relation of the measured wall thickness to age, sex, pancreatic enzyme intake, and intestinal diseases was analyzed. Results: In controls, mean wall thickness depending on gut region was 1.1 to 1.3 mm (SD ). In CF patients, no evidence of severe thickening or stricture was found, but wall thickness was significantly higher than in controls in all measured regions ( mm; SD ). Patients with meconium ileus and distal intestinal obstruction syndrome had significantly increased thickness of the terminal ileum compared with patients without these complications. Neither the intake of high-strength enzymes nor enzyme dosage, age or sex were associated with wall thickness. Conclusions: We found no evidence that a subclinical stage of fibrosing colonopathy is prevalent among the CF patients. Slightly thickened gut walls in CF can be interpreted as an expression of glandular dysfunction in the CF intestine. Key Words: cystic fibrosis, bowel-wall thickness, fibrosing colonopathy, ultrasonography Fibrosing colonopathy is a newly described complication in patients with cystic fibrosis (CF). It denotes submucosal fibrosis leading to a thickened colon wall with the formation of local or pancolonic stenosis. 1,2 Although its etiology is uncertain, the intake of large doses of pancreatic enzyme supplements appears to be an important risk factor. 3,4 While manifest colonic strictures can be diagnosed by contrast enema, 5 the recognition of gut-wall thickening before the onset of severe complications would be of great importance. High frequency abdominal ultrasound is able to discriminate different layers of the gut wall and its validity has been Submitted August 20, Accepted March 11, From the Universities Kinderklinik und Kinderpoliklinik im Dr. von Haunerschen Kinderspital, Munich, Germany. Address correspondence and reprint requests to Dr. Hinrich Karsten Harms, Univ.-Kinderklinik und Kinderpoliklinik im Dr. von Haunerschen Kinderspital, Lindwurmstr. 4, München, Germany. kharms@kk-i.med.uni-muenchen.de. proven in other intestinal diseases. 6,7 Using this method, colon-wall thickening was demonstrated in a group of CF children. This finding was interpreted as a subclinical form of fibrosing colonopathy because of its association with daily enzyme intake. 8,9 The CF patients attending our care unit on average take considerably smaller doses of pancreatic enzymes than cases studied earlier. Therefore, we investigated if bowel wall thickening is also prevalent in our patients and tried to identify possible factors associated with increased bowelwall thickness. MATERIALS AND METHODS Study Group Eighty-three patients with cystic fibrosis (48 females) with a mean age of 12.8 years (range ) were studied by transabdominal ultrasound. This was approximately one third of all CF patients (259) attended by our care unit. The study group consisted of outpatients attending our unit to have their regular checks. The last 3 patients of every day were asked to take part in the study. There was no selection in respect to enzyme intake or intestinal disorders. Informed consent was obtained from the parents (in children under age) and patients. The study was approved by the hospital review board. After the sonographic evaluation, data concerning sex, age, weight, pancreatic enzyme supplementation, other medication, abdominal complaints, and previous intestinal disorders were collected retrospectively. Daily doses of pancreatic enzymes were calculated from the type of preparation (high-strength or standard-strength), daily amount of capsules, and kilogram body weight (expressed in PhEur units (U) lipase/kg/d). Apart from 1 child taking Panzytrat (Knoll, Germany), all patients with pancreatic insufficiency were treated with products of 1 brand (Kreon, Solvay Pharma, Germany): 58% were receiving the high-strength formulation (25000 U lipase/ capsule), 39% were on standard-strength Kreon ( U lipase), and 1 child took Kreon granules ( U lipase). Four patients did not take any enzyme supplements. Control Group Additionally, a control group of 31 probands (17 females) with an age structure comparable to the patient s group (mean 13.3 years, range ) was evaluated sonographically. Probands had to be free of abdominal com- 55

2 56 J Clin Gastroenterol, Vol 37, No. 1, 2003 TABLE 1. Correlation of sonographic and histologic layers of the bowel wall Sonographic appearance Histologic correlate 1. echogenic Interface of mucosa with luminal content 2. hypoechoic Mucosa 3. echogenic Submucosa 4. hypoechoic Muscularis propria 5. echogenic Serosa and surrounding tissue plaints and of any previous gastrointestinal illnesses. Most probands were outpatients or children hospitalized for minor surgical problems. To be compared with the older CF patients, also healthy adults were integrated. Ultrasound All ultrasound scans were performed by a radiologist (C.H.) and a clinician (I.D.) who were both present during the examination. Wall thickness measurements were recorded if both observers were in accordance about the correct location of the measuring points. Blinding of both CF patients and control group was not entirely possible because the observers knew some patients, but the observers were blinded to the patient s abdominal complaints and enzyme intake. We used a 7 MHz linear scanner (Acuson L 7384, Mountain View, California) with an axial and lateral resolution of 0.3 mm. During examination the scanner was applied with graded compression, a technique that allows deep penetration into the abdomen and minimizes disturbing gas artifacts. 10 Bowel wall thickness was measured in 4 segments including terminal ileum, cecum, and ascending and descending colon. To ensure correct location of the measurements the different bowel sections had to be identified by the following criteria: The terminal ileum was identified crossing the iliacal vessels between right psoas muscle and abdominal wall. The colon was recognized by absent or reduced peristalsis and typical haustration. The cecum was located lateral of the terminal ileum and ventral of the right psoas. Tracing the colon cranially the ascending colon was measured near the right flexure at the lower hepatic border. Care was taken not to misinterpret stomach as colon. Finally, descending colon was identified near the left kidney and iliacal vessels. The bowel wall was identified in accordance to endosonographic studies of Kimmey et al 11,12 who described a characteristic pattern of 5 layers of different echogenity. Sonographic appearance and corresponding histologic layers are shown in Table 1. Wall thickness was measured in transverse and longitudinal section from the edge of the mucosa to the outer border of the muscularis propria (Fig. 1). Mean values of each area were recorded. If intraluminal gas interfered with the depiction of the posterior wall, it was assumed that the anterior mural thickness was representative. Three patients and 4 probands had to be excluded because of insufficient sonographic conditions. Examples of wall thickness measurements are shown in Figure 2. Analysis Unpaired t test was used to compare CF patients and controls in respect to wall thickness in the terminal ileum and descending colon. In the cecum and ascending colon, Wilcoxon test was used, as the patient s values were not distributed normally in these segments. This test was also FIGURE 1. (A) Wall thickness measurement in the cecum (cross section). (B) The detailed cutout shows 5 layers of different echogenity and defines start and end point of the thickness measurement (white arrow). (* = serosa and surrounding tissue; ** = interface of luminal content and superficial mucosa; M. = M. obliquus internus abdominis)

3 Dialer et al Sonographic Evaluation of Bowel Wall Thickness 57 FIGURE 2. The 4 studied bowel regions (examples of CF patients): (A) Terminal ileum, transverse section; (B) cecum, longitudinal section; (C) ascending colon, transverse section; (D) descending colon, longitudinal section. used to analyze the influence of abdominal complaints to wall thickness. Spearman rank explored a possible correlation between age or patient s enzyme dose and wall thickness. The relation between patient groups with different intestinal complications and wall thickness was analyzed by ANOVA. A possible association of the variables sex and use of highstrength enzymes to wall thickness was explored by the unpaired t test. RESULTS In the control group, mean wall thickness depending on gut region was 1.1 to 1.3 mm (SD ). In CF patients, no evidence of severe thickening was found. Only 3 CF patients (3.6%) had mild colonic thickening over 3 mm (Fig. 3). Nevertheless, wall thickness was significantly higher in the CF group ( mm; SD ; P < 0.001) than in controls in all gut regions (Fig. 4). In both groups there was no significant difference between measurements in different sections (transverse or longitudinal) of the same bowel segment. Twenty-seven CF patients had CF-specific intestinal complications: 9 patients had a history of meconium ileus (MI), 11 had distal intestinal obstruction syndrome (DIOS), and 7 patients had both. These patients with a combined history of MI and DIOS had thicker ileum walls than patients with DIOS (P < 0.05) or patients without these complications (P < 0.01) (Fig. 5). Bowel-wall thickness was not associated to sex, age, abdominal complaints, or administration of high-strength

4 58 J Clin Gastroenterol, Vol 37, No. 1, 2003 FIGURE 3. Differences in wall thickness between controls and CF patients (ascending colon, cross section, L = left hepatic lobe): Normal thickness in a control person (A) and increased wall thickness (3.2 mm) in a patient with CF (B). enzymes, and we did not find a correlation between enzyme dose (expressed in lipase or protease units) and wall thickness. Calculated for lipase, Spearman rank correlation was 0.09 (terminal ileum), 0.10 (cecum), 0.12 (ascending colon) and 0.08 (descending colon). Enzyme requirements of the patients varied widely. Mean lipase intake was 5398 U/kg/d (SD ± 5435, range U/kg/d) with younger age being correlated to higher enzyme intake (P 0.001). Enzyme dosage exceeding 10,000 U/kg/d was necessary in 12 patients (14%), 3 of them needed more than 20,000 U/kg/d. Even in patients without enzyme intake (n 4) bowel-wall diameters exceeded the level of healthy controls. DISCUSSION Today, abdominal ultrasound plays an important role in detecting bowel wall abnormalities, and its reliability has FIGURE 4. Differences in mean bowel-wall thickness between controls and patients with cystic fibrosis. * = P < ; ** = P < controls versus cystic fibrosis. FIGURE 5. Wall thickness in the terminal ileum. Patients without intestinal involvement are compared with patients with meconium ileus (MI) and/or DIOS.

5 Dialer et al Sonographic Evaluation of Bowel Wall Thickness 59 been proven in many other gastrointestinal disorders. 6,7,13,14 In cystic fibrosis, this method is breaking new ground and a correct assessment of what is bowel wall thickening is difficult: Firstly, because the sonographic appearance of the uncomplicated CF intestine was unknown and secondly, because a clear definition of normal bowel wall thickness at abdominal ultrasound is still lacking. Earlier sonographic studies made under in-vitro conditions 11 or without high-frequency transducers 15 stated a normal colon-wall thickness of 2 to 3 mm. In our study, lower values for controls were found (mean mm). This corresponds to other investigators also using high frequency scanners with a better resolution to mark off wall layers and extramural structures. 9,16 19 Compared with controls, our CF patients had slightly elevated bowel wall diameters with a mean thickness of 1.5 to 1.6 mm. Comparable studies on inflammatory or ischemic bowel disease considered bowel-wall diameters exceeding 3 mm pathologic. 15,20,21 This condition was only seen in 3 of our 83 CF patients, and all 3 were asymptomatic. Therefore, our study does not support the finding that a large number of CF patients present with abnormal colonic thickening. 8 The administration of high-strength enzymes (Kreon ) was not associated with higher bowel-wall thickness in our patients. This corresponds with other studies, 4,16,18,19 yet analyzing a possible relationship to certain products has been difficult because many patients took more than 1 product or switched between preparations. 3 The patients of this study represent 2 fairly homogenous groups as all but 1 were either taking the normal or the high-strength preparation of the same brand, Kreon. This microsphere formulation does not contain methacrylic acid copolymer, a coating substance recently related to fibrosing colonopathy. 22 Several studies following Mc Sweeney s report of a dose-dependent association between colon-wall thickness and the patient s enzyme intake 9 came to divergent results. 16,18,19 In our patients, enzyme dose did not correlate with wall thickness, independent of its expression as protease or lipase. However, average lipase intake was considerably lower in our patients compared with previous studies, 3,4,9 thus an association with very high enzyme doses cannot be excluded by our data. The great majority of patients (86%) were sufficiently treated with doses below 10,000 U/kg/d, the maximum dose recommended by the consensus committee. 2,23 Doses of more than 20,000 U/kg/d, previously common in British and American CF patients, were only required in 3 cases. These results are reassuring, as they indicate that within the recommended guidelines most patients can achieve a clinically adequate nutrient absorption. One third of our patients had a history of neonatal meconium ileus (MI) and/or DIOS, both affecting the ileocecal region. Gut walls in the terminal ileum turned out to be thicker in patients with a history of both MI and DIOS than in patients without intestinal involvement or with DIOS alone. However, these observation should be interpreted cautiously because of the relatively small number of patients in each group. Patients with CF had thicker bowel walls than controls independent of their enzyme intake. Even patients without any enzyme substitution showed increased bowel-wall thickness. These 2 findings lead to the conclusion that this phenomenon is likely to be attributed to the pathophysiologic conditions of the underlying disease. It is known that the genetic defect in chloride transport is also present in the intestinal glands. Studies of the appendix 24 and rectum 25 described characteristic histologic changes in CF consisting in more goblet cells and widely dilated crypts packed with mucus. The increased wall diameter seen on ultrasound could be caused by these mucosal alterations. But unlike histology, abdominal ultrasound is limited in locating the source of wall thickening as the given resolution capacity does not allow single layers to be measured exactly unless they are severely thickened. Although differences in wall thickness between our patients and controls were highly significant, mean differences were small. Furthermore, the influence of bowel motility is difficult to quantify on ultrasound. In the proximal small bowel, wall thickness will be influenced by pronounced peristalsis. Because measurements were located distally in this study, where peristalsis is reported to be greatly reduced or absent, 26 it was assumed that peristaltic waving did not have a significant influence on the ultrasound measurements. Despite the limitations mentioned above transabdominal ultrasound seems to be a very helpful diagnostic tool in distinguishing between a common situation in CF slightly elevated bowel wall diameters and severely thickened bowel walls requiring further investigation. REFERENCES 1. Smyth RL, van Velzen D, Smyth AR, et al. Strictures of the ascending colon in cystic fibrosis and high-strength pancreatic enzymes. Lancet. 1994;343: Borowitz DJ, Grand RJ, Durie PR, The Consensus Committee. A consensus statement on the use of pancreatic enzyme supplements for patients with cystic fibrosis in the context of fibrosing colonopathy. J Pediatr. 1995;127: Smyth RL, Ashby D, O Hea U, et al. Fibrosing colonopathy in cystic fibrosis: results of a case-control study. Lancet. 1995;346: FitzSimmons SC, Burkhart GA, Borowitz D, et al. High-Dose Pancreatic- Enzyme Supplements and Fibrosing Colonopathy in Children With Cystic Fibrosis. N Engl J Med. 1997;336: Crisci KL, Greenberg SB, Wolfson BJ, et al. Contrast enema findings of fibrosing colonopathy. Pediatr Radiol. 1997;27: Hollerbach S, Geißler A, Schiegl H, et al. The accuracy of abdominal ultrasound in the assessment of bowel disorders. Scand J Gastroenterol. 1998;33: Maconi G, Parente F, Bollani S, et al. Abdominal Ultrasound in the Assessment of Extent and Activity of Crohn s Disease: Clinical Significance and Implication of Bowel Wall Thickening. Am J Gastroenterol. 1996;91: Oades PJ, Rosenthal M, Mac Sweeney EJ, et al. Subclinical colonic thickening. J R Soc Med. 1996;89(suppl 27): Mac Sweeney EJ, Oades PJ, Buchdahl R, et al. Relation of thickening of colon wall to pancreatic-enzyme treatment in cystic fibrosis. Lancet. 1995;345:

6 60 J Clin Gastroenterol, Vol 37, No. 1, Puylaert JB. Acute Appendicitis: US Evaluation Using Graded Compression. Radiology. 1986;158: Kimmey MB, Martin RW, Haggitt RC, et al. Histologic Correlates of Gastrointestinal Ultrasound Images. Gastroenterology. 1989;90: Wilson SR. The Gastrointestinal tract. In: Rumack C, Wilson SR, Charboneau JW, eds. Diagnostic Ultrasound. 2nd ed. New York: Mosby; 1997: Lee HC, Yeh HJ, Leu YJ. Intussusception: The Sonographic Diagnosis and Its Clinical Value. J Pediatr Gastroenterol Nutr. 1989;8: Siegel MJ, Friedland JA, Hildebolt CF. Bowel Wall Thickening in Children: Differentiation with US. Radiology. 1997;203: Hata J, Haruma K, Fujimura J, et al. Ultrasonographic Evaluation of the Bowel Wall in Inflammatory Bowel Disease: Comparison of In Vivo and In Vitro Studies. Abdom Imaging. 1994;19: Haber HP, Benda N, Fitzke G, et al. Colonic wall thickness measured by ultrasound: striking differences in patients with cystic fibrosis versus healthy controls. Gut. 1997;40: Pohl M, Krackhardt B, Posselt HG, et al. Ultrasound Studies of the Intestinal Wall in Patients with Cystic Fibrosis. J Pediatr Gastroenterol Nutr. 1997;25: Ramsden WH, Moya EF, Littlewood JM. Colonic wall thickness, pancreatic enzyme dose and type of preparation in cystic fibrosis. Arch Dis Child. 1998; 79: Connett GJ, Lucas JS, Atchley JTM, et al. Colonic wall thickening is related to age and not dose of high strength pancreatin microspheres in children with cystic fibrosis. Eur J Gastroenterol Hepatol. 1999;11: Solvig J, Ekberg O, Lindgren S, et al. Ultrasound examination of the small bowel: comparison with enteroclysis in patients with Crohn disease. Abdom Imaging. 1995;20: Teefey SA, Roarke MC, Brink JA, et al. Bowel wall thickening: differentiation of inflammation from ischemia with color doppler and duplex US. Radiology. 1996;198: Van Velzen D, Ball LM, Dezfulian AR, et al. Comparative and experimental pathology of fibrosing colonopathy. Postgrad Med J. 1996;72(suppl 2):S39 S Littlewood JM. Implications of the Committee on Safety of Medicines IU Lipase/kg/day recommendation for use of pancreatic enzymes in cystic fibrosis (Commentary). Arch Dis Child. 1996;74: Andersen D. Pathology of cystic fibrosis. Ann N Y Acad Sci. 1962;93: Parkins R, Ruben C, Eidelman S, et al. The diagnosis of cystic fibrosis by rectal suction biopsy. Lancet. 1963;2: Kawahara H, Hirai K, Aoki T. Evaluation of colonic motility with extrasomatic ultrasonography. J Jpn Soc Colo-Proctol. 1999;52:91 97.