Barium Studies in Patients with Candida Esophagitis: Pseudoulcerations Simulating Viral Esophag itis

Transcription

1 X/94/ American Roentgen Ray Society Seth N. Glick1 Received January 1 9, 1994; accepted after ravision March 15, Department of Diagnostic Radiology, Hahnamann University, Broad and Vine Sts., Mail Stop 206, Philadelphia, PA Address corraspondance to S. N. Glick. Barium Studies in Patients with Candida Esophagitis: Pseudoulcerations Simulating Viral Esophag itis OBJECTIVE. The purpose of this report is to describe an abnormality identified on barium studies in patients with Candida esophagitis that simulates discrete or aphthous ulceration. This finding may be misinterpreted as suggesting viral esophagitis. PATIENTS AND RESULTS. Between and 1993, four patients with endoscopically confirmed Candida esophagitis and no evidence of associated viral infection were treated at Hahnemann University. These patients were selected for study because double-contrast barium studies showed an aphthoid complex consisting of a superficial punctate barium collection surrounded by a thin radiolucent rim. Most of the complexes were 2-3 mm in diameter. No typical plaques were present, and the lesions were separated by normal mucosa. In two cases, only one or two lesions were observed on initial esophagograms because of limited distension. Additional esophagograms with additional distension showed more lesions. In two cases, the results of the barium studies were misinterpreted as evidence of viral esophagitis. CONCLUSION. Multiple esophageal lesions that resemble aphthous ulcers are a subtle radiographic manifestation of Candida esophagitis and may be responsible for either missed diagnoses or misinterpretation as viral esophagitis. Features suggestive of pseudoulceration include uniformity of size and relatively sharp margination of the filling defect, and punctate rather than stellate collections of barium. AJR 1 994;163: The characteristic appearance of Candida esophagitis on barium studies is that of diffuse discrete mucosal plaques that, in advanced cases, may become confluent to form a cobblestone or shaggy esophagus [1-8]. Esophageal ulceration is uncommon and suggests the presence of viral esophagitis [9]. I describe four cases of Candida esophagitis in which barium esophagognams showed multiple barium collections that simulated aphthous ulceration and could be misdiagnosed as viral esophagitis. Patients and Results The four patients, three women and one man years old (mean, 43 years), were treated between 1989 and Two had odynophagia of recent onset, one had dysphagia, and one had epigastnic pain and burning. Two of the patients were undergoing chemotherapy for leukemia. The other two had no discemnible underlying condition that would predispose to infection with Candida. All the patients had endoscopy within 1 week of the barium study. In each case, endoscopy showed multiple white plaques without evidence of erosion on ulceration. Cultures of material obtained by endoscopic brushings were positive for Candida albicans, and examination of biopsy specimens obtained via endoscopy did not show viral inclusion bodies. Antifungal therapy was effective in all four cases. A review of the radiology records for the interval when these cases were diagnosed indicated 11 additional cases of Candida esophagitis.

2 350 GLICK AJR:163, August 1994 A B Fig year-old woman with dysphagia and no predisposing medical condition. Candida plaques were identified at endoscopy and confirmed by culture of biopsy samples. A, Double-contrast esophagograrn shows multiple superficial collections of barium with surrounding radiolucent rims (arrows) in mid esophagus. B, Esophagogram obtained with flow technique (i.e., exposure immediately after barium has coated the mucosal surface) shows a larger Icsion with a central collection of barium (arrow). On barium studies, the typical lesion appeared as a punctate collection of barium surrounded by a madiolucent rim resembling an aphthous ulcer (Fig. 1A). The overall size of this complex was approximately 2-3 mm in diameter, but a few were larger, and one was 1 cm in diameter (Fig. 1 B). The width of the lucent rim with respect to the collection was vanable and in some instances was barely visible (Fig. 2). No classic monilial plaques were observed. The lesions were mostly in the middle third of the esophagus; a few isolated lesions were present more proximally on distally. The lesions were multiple but not numerous, and normal mucosa was evident between them in three of the four cases. In three of the four patients, as few as two or three lesions could be detected on initial esophagognams (Figs. 3 and 4). This was mostly a result of underdistension, as patients with odynophagia have difficulty ingesting sufficient quantities of barium to optimally distend the esophagus. In these cases, esophagograms obtained with additional distension showed further lesions. Two of the cases were diagnosed initially as viral esophagitis. The other two were diagnosed correctly after the experience with the first two. A review of the 11 other cases of Candida esophagitis revealed three cases in which similar pseudoulcenations on central collections of barium on larger plaques were present in association with the characteristic radiologic pattern. Analysis of the four cases described here as well as the other three cases showed several findings suggestive of pseudoulcenation, including uniformity of size and relatively sharp mangination of the filling defect and punctate barium collections rather than the stellate ulcers described in viral esophagitis. Discussion In the clinical management of patients at risk for opportunistic infection, distinguishing cases of Candida esophagitis from cases of viral infections, such as those caused by herpes simplex virus or cytomegalovimus, has important therapeutic implications. Candida esophagitis can be reliably detected on double-contrast barium studies. Most cases have a characteristic appearance, ranging from scattered discrete plaques to a florid confluence producing a shaggy contour [1-8]. The entire esophagus is usually involved. On the other hand, viral esophagitis is distinguished by the presence of discrete ulcers, usually stellate, surrounded by nadiolucent mounds on an otherwise normal mucosal background [8-i 0]. Atypical findings occasionally occur with Candida and include segmental disease, stricture, fistula, and polypoid lesions that may appear as discrete (1-4 cm) lobulated polyps or large bulky intraluminal masses [2, 11-16]. Ulceration has also been reported [1-4, 11, 17]; however, the true prevalence of this morphologic alteration is unclean. In some instances, such a described finding may have not been endoscopically confirmed. An appearance suggestive of ulceration may be produced on full-column esophago- Fig year-old woman with leukemia and odynophagia. Doubie-contrast esophagogram shows a few small superficial barium collections (arrows) distributed along thoracic esophagus. Radiolucent rims are barely detectable. Candida plaques were identified at endoscopy and confirmed by culture of biopsy samples.

3 AJR:163, August 1994 BARIUM STUDIES OF CANDIDA ESOPHAGITIS 351 A B A B Fig year-old man with leukemia and odynophagia. Candida plaques were identified at endoscopy and confirmed by culture of biopsy samples. A, Double-contrast esophagogram shows isolated aphthoid lesion (arrow) in mid esophagus. Esophagus is not optimally distended. B, Double-contrast esophagogram obtained with increased distension shows multiple punctate barium collections along entire length of esophagus. Collections are subtle and could be easily overlooked. grams by barium projecting between plaques along the contour, by barium undermining large plaques and tracking parallel to the lumen, or by intramural pseudodiverticuli, which are often associated with esophageal candidiasis [18]. In one review [2] of the radiologic findings of esophageal candidiasis, abnormalities similar to those in the four cases in this report were described as ulceration. In addition, Candida esophagitis is frequently present in conjunction with viral esophagitis [19, 20]. Thus, it is difficult to determine which organism produced the ulceration, particularly if tests of samples obtained at endoscopy do not indicate the presence of the virus. Patients with HIV ulceration and Candida esophagitis may sometimes fit this scenario [21]. One report [1 7] described two cases of large discrete ulceration as a Fig year-old woman with epigastric pain and burning but no predisposing factors. Cand!da plaques were identified at endoscopy and confirmed by culture of biopsy samples. A, Double-contrast esophagogram obtained with less than optimal distension shows two tiny aphthoid lesions (arrows) in mid esophagus. B, Double-contrast esophagogram obtained with increased distension shows additional similar abnormalities (arrows). new radiographic manifestation of esophageal moniliasis [19]. Although the patients had proved candidiasis, and antifungal therapy was effective, the possibility of associated ulceration due to HIV cannot be excluded. Indeed, C. albicans has a tendency to colonize gastrointestinal ulcers and, in some cases, may be a saprophytic bystander rather than an invasive organism. The pathophysiologic mechanism that accounts for these pseudoulcerations is unclear. The endoscopic appearance was typical for Candida, although no attempt was made to correlate with the radiographic observations. Perhaps some plaques due to Candida have central depressions that can collect barium. Alternatively, the center of the plaques may be somewhat raised and flat and may become coated with

4 352 GLICK AJR:163, August 1994 barium as the agent passes across them. The possibility of the central portion having different adherence properties with regard to the barium cannot be excluded. Nevertheless, radiologists should be aware of this manifestation of esophageal candidiasis, particularly when it is not associated with characteristic plaques, so that the proper diagnosis can be made and unnecessary endoscopy, which might be necommended because of the suggestion of viral esophagitis, can be avoided. REFERENCES 1. Levine MS, Macones AJ, Laufer I. Candida esophagitis: accuracy of radiographic diagnosis. Radiology 1985;154: Roberts L, Gibbons, Gibbons G, Rice RP, Thompson WM. Adult esophageal candidiasis: a radiographic spectrum. RadioGraphics 1987;7: Lawicki AM, Moore JP. Esophageal moniliasis: a review of common and lass frequent characteristics. AJR 1 975; 125: Athey PA, Goldstein HM, Dodd GD. Radiologic spectrum of opportunistic infections ofthe upper gastrointestinal tract. AJR 1977;129: Guyer PB, Brunton FJ, Rooke HW. candidiasis of the oesophagus. Br J Radiol : Gonzalez G. Esophageal moniliasis. AJR 1971;113: Goldberg HI, Dodds wj. cobblestone esophagus due to monilial infection. AJR 1968:104: Levine MS. Woldenbarg A, Herlinger H, Laufer I. Opportunistic esophagitis in AIDS: radiographic diagnosis. Radiology 1 987; 165: Levine MS. Radiology of esophagitis: a pattem approach. Radiology 1991;179: Levine MS, Loavnar LA, Saul SH, Rubasin SE, Harlinger H, Laufar I. Herpes esophagitis: sensitivity of double-contrast esophagography. AJR 1988;151 : Farman J, Tavitian A, Rosenthal LE, Schwartz GE, Raufman JP. Focal esophageal candidiasis in acquired immunodaficiency syndrome (AIDS). Gastrointest Radioll986;11 : Ott DJ, Gelfand Dw. Esophageal stricture secondary to candidiasis. Gastrointest Radioll978;2: Kelvin FM, Clark WM, Thompson WM, Hauch T. chronic oesophageal stricture due to moniliasis. BrJ Radioll978;31: Rohrmann CA, Kidd R. Chronic mucocutaneous candidiasis: radiologic abnormalities in the esophagus. AJR 1 978; 130: Ho CS, Cullen JB, Gray RR. An unusual manifestation of esophageal moniliasis. Radiology 1 977; 123: Kowal LE, Goodman LR, Teplick 5K, Glick SN. Multiple esophago-pulmonary communications of infectious origin. Am J Gastroenteroll983;78: Bier SJ, Keller RJ, Krivisky BA, Liftin AJ. Esophageal moniliasis: a new radiographic presentation. Am J Gastroenterollg85;80: Troupin RN. Intramural esophageal diverticulosis and moniliasis. AJR 1968;104: Bonacini M, Young T, Lame L. The causes of esophageal symptoms in human immunodeficiancy virus infection. Arch Intern Med 1991;151: Wilcox CM. Esophageal disease in the acquired immunodeficiancy syndrome: etiology, diagnosis, and management. Am J Med 1992;92: Levine MS, Loercher G, Katzka DA, Herlingar H, Rubasin SE, Laufar I. Giant, human immunodaficiancy virus-related ulcers in the esophagus. Radiology :

5 This article has been cited by: 1. C. Mel Wilcox. Esophagitis in the Immunocompromised Host [Crossref] 2. Shmuel Shoham, Marcio Nucci, Thomas J. Walsh. Mucocutaneous and Deeply Invasive Candidiasis [Crossref] 3.. Candida Esophagitis [Crossref] 4. Marc S. Levine. Infectious Esophagitis [Crossref]