GASTROENTEROLOGY 1988;95:1518-22 Scalloped Valvulae Conniventes: An Endoscopic Marker of Celiac Sprue MANSOUR JABBARI, GARY WILD, CARL A. GORESKY, DONALD S. PAL Y, JOHN O. LOUGH, D. PAUL CLELAND, and DOUGLAS G. KINNEAR The McGill University Medical Clinic and the Division of Gastroenterology of the Montreal General Hospital, Montreal, Quebec, Canada The finding, in a patient with celiac sprue, of a characteristic change at endoscopy (scalloping of the valvulae conniventes, evident on close inspection, but forming only a mosaic pattern from a distance) led to an endoscopic survey designed to define its incidence. In a series of 28 sequential patients found to have microscopic changes characteristic of sprue on biopsy, distinctive endoscopic changes were found in 22 (in 6 of 9 with sprue in relapse, and 16 of 19 presenting with initial symptoms). The finding of the distinctive appearance provides an endoscopically recognizable pattern that can be associated with sprue. It also provides the potential for early recognition of the process in patients in whom the diagnosis might otherwise have been delayed due to a lack of substantial evolution of the usually associated symptom complex. Changes in the small bowel mucosa are characteristic of celiac sprue. At the microscopic level these consist primarily of an atrophy of villi accompanied by crypt hyperplasia, most marked in the upper bowel (1). At the clinicallevel this enteropathy is often associated with a full-blown malabsorption syndrome, which improves with a gluten-free diet and relapses on reexposure to gluten (2). Despite the widespread use of fiberoptic endoscopy in the examination of the upper gastrointestinal tract for a variety of symptoms, reports of endoscopic observations in celiac sprue are few (2-5) and, endoscopically, no characteristic change has been described. The overall impression from these observations has been that routine endoscopy is useful in reaching the diagnosis of celiac sprue only if it is used to obtain a mucosal biopsy, or if it is complemented by magnification endoscopy (5), when the changes ordinarily observed on examination of biopsy specimens with a dissecting microscope (loss of villi, formation of surface convolutions, and appearance of a mosaic pattern) can be observed (6). Thus, routine endoscopy has not been expected to be useful in the diagnosis of celiac sprue. Sprue presents in a variety of ways, most severely with massive malabsorption and substantial weight loss, less severely with malabsorption of a particular nutrient dominating the clinical picture or with nonspecific dyspepsia without prominent malabsorption. If changes could be detected at a gross level, diagnostic suspicions would be sharpened at the less symptomatic stage of the disease, as well as when malabsorption is well established. In the present report we describe a characteristic endoscopic abnormality associated with the histopathologic alterations seen in celiac sprue. This observation has proved to be of clinical utility. Materials and Methods Between 1982 and 1987 a diagnosis of celiac sprue based on histopathologic criteria was made in 28 consecutive patients seen in our unit. The clinical presentation of each patient was annotated at the time, and the data were reviewed retrospectively. Duodenoscopy was performed, in each case, with a Olympus end-viewing GIF-type endoscope (Olympus Corporation of America, New Hyde Park, N.Y.), and several endoscopic biopsy specimens were obtained at or below the middle of the second stage of the duodenum. Specimens were fixed in neutral 10% formalin and processed for routine light microscopic examination. One of us (M.J.) had observed, on endoscopic study, that a distinctive scalloping of valvulae conniventes was present in one of his patients with sprue. With this observation, we began a prospective endoscopic search for this lesion in patients known to have sprue in relapse and in patients suspected to have sprue. With experience, it 1988 by the American Gastroenterological Association 0016 5085/88/$3.50
December 1988 ENDOSCOPIC CHARACTERIZATION OF CELIAC SPRUE 1519 also became possible to perceive the presence of this distinctive endoscopic finding in patients who had vague dyspeptic symptoms, but no other complaints suggestive of sprue. The absolute diagnostic criterion utilized was based on histopathologic assessment of biopsy material. The features needed to arrive at a diagnosis were loss of villi with flattening of the mucosal surface, hyperplasia of the crypts, presence of cytologically abnormal surface cells, and in- flammation of the lamina propria. All of the patients responded clinically to the removal of gluten from their diet. To provide a base for assessing whether the severity of the lesion was important to its endoscopic recognition, all biopsy specimens were examined by a pathologist who did not have knowledge of the clinical or endoscopic findings. The degree of villus atrophy and of inflammation were classified as mild, moderate, or severe, and these data were correlated with the endoscopic findings. Figure 1. Endoscopic appearance of the characteristic changes in celiac sprue. Upper left panel: normal duodenal mucosa, exhibiting a smooth and velvety mucosa, with a normal reddish hue. Upper right panel: scalloped valvulae conniventes, viewed en face, from the duodenum of a patient with celiac sprue. Lower left panel: mosaic pattern with visible vasculature, and scalloped valvulae conniventes, seen well only at the edge of a valvula. Lower right panel: mucosal appearance of the duodenum from a patient with sprue in remission. The changes have resolved.
1520 JABBARI ET AL. GASTROENTEROLOGY Vol. 95, No.6 Results Clinical Manifestations Celiac sprue was confirmed by endoscopic duodenal biopsy in the 9 patients with previously established sprue. These patients had undergone duodenoscopy with biopsy for investigation of either a recurrence of malabsorptive symptoms or dyspepsia, and were suspected of having relapsed because of deviation from their gluten-free diet. The 19 patients not previously known to have celiac disease underwent duodenoscopy for investigation of symptoms related to iron-deficiency anemia (n = 6), chronic diarrhea (n = 7), malabsorption (n = 4), or dyspepsia (n = 2). In each patient, endoscopic biopsy specimens demonstrating the characteristic changes of sprue were obtained. Endoscopic Findings The endoscopic appearance of the normal duodenal mucosa is shown in the upper left panel of Figure 1. In 6 of the 9 patients with previously known celiac sprue, the endoscopic findings were classified as abnormal. The mucosa was atrophic and had a mottled appearance with patchy areas of pale mucosa alternating with more erythematous mucosa; the pale areas had a pronounced mosaic appearance. The underlying blood vessels appeared prominent in some regions. The mucosa in these areas was not friable and petechiae were absent. In each of these cases, the valvulae conniventes exhibited what we have come to regard as the distinctive endoscopic lesion of sprue. The valvulae had a scalloped appearance when viewed en face through the endoscope (Figure 1, upper right panel). When viewed from a distance, a mosaic appearance becomes more prominent, and scalloping of the valvula is evident only at its luminal edge (Figure 1, lower left panel). The scalloped appearance is best brought into view when the bowel is distended with air. All 6 of the patients previously known to have sprue and who exhibited the characteristic finding had had an exacerbation of diarrhea or malabsorption at the time of examination. In the remaining 3 patients with known celiac sprue, the characteristic endoscopic findings were not noted. These 3 patients had presented with dyspeptic symptoms only, and although moderate villus atrophy was observed on biopsy, only mild inflammatory change was noted in each instance. Follow-up duodenoscopy in 3 of the endoscopically positive patients in the group with sprue in relapse demonstrated resolution of almost all the previously noted endoscopic features after 6 mo on a gluten-free diet (Figure 1, lower right panel). There is a suggestion of a mosaic pattern, but it is no longer clearly present and the scalloping has disappeared. The marker endoscopic features found in the patients with previously known celiac sprue were also seen in 16 of the other 19 patients. These included, in each instance, a scalloped configuration of the valvulae conniventes. In 3 of these 16 patients the duodenal mucosa was initially assessed as normal, and then judged to be abnormal on repeat endoscopy carried out because of persistence of symptoms. In 3 of 19 patients not previously known to have celiac sprue, who exhibited duodenal mucosa judged to be endoscopically normal but who had positive findings on biopsy, the evaluation was being carried out because of anemia. In these 3 patients, the degree of villus atrophy on biopsy was mild in 2 and severe in 1, and the degree of inflammation was mild in 2 and moderate in 1. Cumulation of the data coming from the histopathologic analysis suggests that the degree of inflammation was an important determinant in the genesis of the endoscopically recognizable lesion, and that there may be a threshold phenomenon for endoscopic recognition. In 5 of the 6 endoscopically negative instances, the degree of inflammation was mild and, in the remaining case, it was moderate. For the group with endoscopically recognizable lesions, 6 were mild, 13 were moderate, and 1 was severe. The characteristic endoscopic lesion has not yet been observed in a case where biopsy has shown a normal mucosa. We discovered that it is important to use an endoscope with a number of optical fibers contributing to the image, at least equivalent to that in the Olympus Q10 scope (~31,OOO). With endoscopes of poorer resolution, the characteristic lesion was no longer clearly discernible. Discussion Few studies have been reported concerning the endoscopic appearance of the proximal duodenal mucosa in celiac sprue. In the earliest report, which describes a patient with celiac sprue and a gastrojejunostomy, the authors observed a diffusely erythematous mucosa without ulceration or friability (3). No other mucosal abnormalities were noted, and the authors concluded that it was unlikely that a diagnosis of celiac sprue could be made based solely on evidence obtained at duodenoscopy (3). McCarthy and Stevens (4) examined the first portion of the duodenum endoscopically in treated and untreated patients with celiac disease in whom the diagnosis had been made using conventional suction biopsy of the proximal jejunum. Using indigocarmine dye scattering, they noted severe atrophy of the
December 1988 ENDOSCOPIC CHARACTERIZATION OF CELIAC SPRUE 1521 mucosa in the duodenal cap and a mosaic pattern in 7 of 11 untreated celiac patients. No mention was made of the appearance of the valvulae conniventes. Biopsy specimens taken from the duodenal cap at the time of endoscopy showed villus atrophy but specimens were too superficial to include crypts. In the study of Gillberg and Ahren (5), 19 patients with suspected, and later documented, celiac disease underwent suction capsule biopsy and duodenoscopy with biopsy. A close concordance was observed in the histopathology of the biopsy specimens obtained from the two methods. With supplementary magnification to reduce the focal length of the viewing lens, a mucosal pattern of villus atrophy was observed in 6 of 8 patients in whom it was utilized. The authors concluded that endoscopy with additional magnification may provide a useful adjunct in the diagnosis of celiac sprue. Recent studies have confirmed the usefulness of duodenoscopy in obtaining biopsy specimens for the diagnosis of small bowel disorders including celiac sprue (7,8), but have not commented on the appearance of the small intestinal mucosa at the time of endoscopy. The diagnosis of early celiac sprue may be difficult to establish and often requires a high index of suspicion, as the presenting clinical features may be subtle (9). This may relate in part to initially limited involvement of only the proximal small bowel (10), and the functional reserve of the ileum, which can compensate for the malabsorption occurring in the proximal small bowel (11). The spectrum of problems in our patients serves to emphasize the range of variation in the clinical presentations of celiac sprue. Of the 19 patients not previously known to have abnormalities of the small bowel mucosa, only 11 (58%) had the clinical features of diarrhea and malabsorption. The remainder presented with unexplained anemia or with dyspepsia. In the patients with previously known celiac sprue, an approximately similar proportion, 4 of 9 (44%) presented with an exacerbation of diarrhea or malabsorption. Dyspepsia was the presenting complaint in a similar proportion of celiac patients in the two groups. One patient presented with dermatitis herpetiformis in addition to dyspeptic symptoms. As all patients in the present study had an abnormal endoscopically obtained biopsy specimen, it was of interest to relate this finding to the appearance of the duodenal mucosa at the time of endoscopy. Characteristic endoscopic mucosal abnormalities were noted in two-thirds of the patients previously known to have celiac sprue. The duodenal mucosa often had a characteristic mosaic appearance, in agreement with the previously reported findings of McCarthy and Stevens (4). In addition, we observed in the present study a consistent scalloping of the valvulae conniventes. That this feature is not present in our entire patient group appears to relate, in part, to the severity of the proximal small bowel inflammation, the lesion being more likely to be seen if the inflammation was moderate. With the advent of newer and easier techniques of small bowel enteroscopy (12), it would be of interest to verify if similar changes occur in the valvulae conniventes of more distal regions of the intestine. At present such an approach appears to have been used only for the diagnosis of small bowel mucosal and vascular abnormalities other than celiac sprue (13). Of the 19 patients who had no previously known small bowel pathology, 16 (84%) had demonstrable duodenal mucosal abnormalities at endoscopy. It is noteworthy that all of the patients who presented clinically with malabsorption and diarrhea were found to have mucosal changes at endoscopy. Considering all patient groups, 22 of the total of 28 patients (79%) with biopsy changes characteristic of sprue were found to have endoscopic changes at the level of the duodenal mucosa. In view of the integral role that endoscopy plays in evaluation of symptoms referable to the upper gastrointestinal tract and in the workup of anemia, the mucosal changes reported here would appear to be of great value. In this study, the scalloped configuration of the valvulae conniventes has served as a consistently recognizable feature marking underlying mucosal change. It appears that this is a novel and useful endoscopic marker for celiac sprue; the presence of the accompanying characteristic histopathologic changes should, of course, always be confirmed by biopsy. The corollary to this observation is that, when the characteristic endoscopic finding is observed unexpectedly, biopsy evaluation of the mucosa should also be carried out. By doing so, the diagnosis of sprue will be reached earlier than otherwise would be the case. References 1. Perera DR, Weinstein WM. Rubin CE. Symposium on pathology of the gastrointestinal tract. II. Small intestinal biopsy. Hum PathoI1975;6:157-217. 2. Booth CC. Definition of adult coeliac disease. In: Hekkens WT, Th JM, Pena AS, eds. Coeliac disease, Proceedings of the Second International Coeliac Symposium. Leiden: Steinfert Kroese, 1974:17-22. 3. Schwob D, Kleinman MS, Turner MD. Endoscopic appearance of the jejunum in a patient with gluten-sensitive enteropathy and a gastrojejunostomy. Am J Dig Dis 1972;17:430-4. 4. McCarthy CF, Stevens FM. The endoscopic demonstration of coeliac disease. Endoscopy 1976;8:177-80. 5. Gillberg R, Ahren C. Coeliac disease diagnosed by means of duodenoscopy and endoscopic duodenal biopsy. Scand J Gastroenterol 1977;12:911-6. 6. Holmes R, Hourihane DO'B, Booth CC. Dissecting microscope
1522 JABBARI ET AL. GASTROENTEROLOGY Vol. 95, No.6 appearances of jejunal biopsy; specimens from patients with "idiopathic steatorrhea." Lancet 1961;i:81-3. 7. Achkar E, Carey WD, Petras R, Sivak MV, Revta R. Comparison of suction capsule and endoscopic biopsy of small bowel mucosa. Gastrointest Endosc 1986;32:278-82. 8. Barkin JS, Schonfelt W, Thomsen S, Mantan HD, Rogers AI. Enteroscopy and small bowel biopsy-an improved technique for the diagnosis of small bowel disease. Gastrointest Endosc 1985;31:215-7. 9. Falchuk ZM. Gluten sensitive enteropathy. Clin Gastroenterol 1983;12:475-94. 10. MacDonald WC, Brandborg LL, Fick AL. Studies of celiac sprue. IV. The response of the whole length of the small bowel to a gluten-free diet. Gastroenterology 1964;47:573-9. 11. Silk DBA, Kumar PJ, Webb JPW. Ileal function in patients with untreated adult coeliac disease. Gut 1975;16:261-7. 12. Lewis BS, Waye JD. Total small bowel enteroscopy. Gastrointest Endosc 1987;33:435-8. 13. Tada M, Kawai K. Small bowel endoscopy. Scand J Gastroenterol 1984;19(Suppl 102):39-52. Received February 29, 1988. Accepted July 8, 1988. Address requests for reprints to: Dr. Carl A. Goresky, University Medical Clinic, Montreal General Hospital, 1650 Cedar Avenue, Montreal, Quebec, H3G 1A4 Canada. This work was supported in part by the Fast Foundation. The authors thank Olga Clarke and Lila Burgess for the superb manner in which they have run their endoscopy facility, providing background for this study, and Nancy Firth and Nella Astorino for typing this manuscript.