Successful Treatment of Refractory Sprue with Cyclosporine

Celiac sprue is a disorder of absorption that is identified by characteristic histologic abnormalities in the small-bowel mucosa [1, 2]. The clinical and histologic features respond to a gluten-free diet. Patients with refractory sprue do not respond to strict gluten avoidance and have a poor prognosis [3]. A patient with severe sprue who was unresponsive to a gluten-free diet and corticosteroid therapy experienced remission after receiving cyclosporine therapy. Case Report A 45-year-old woman was hospitalized in January 1986 because of diarrhea (5 to 24 stools/d) for 5 months, vomiting for 6 weeks, and a weight loss of 21 kg. She was 1.64 m in height, weighed 95 kg, and had generalized weakness. Results of laboratory tests of blood and urine were unremarkable. Stool examination for pathogens and an alkalinization test for phenolphthalein-containing laxative were negative. The stool volume was 814 to 2014 mL/d. While on a diet containing 100 g of fat, she excreted 9.4 g fat/d. The results of sigmoidoscopic evaluation and a small-bowel roentgenographic series were normal. Esophagogastroduodenoscopic evaluation showed mild duodenal hyperemia, and findings in a duodenal biopsy specimen were typical of sprue (Figure 1, panel A). Treatment included a gluten-free diet and prednisone (initially 60 mg/day). Figure 1. Duodenal biopsy findings in a patient with sprue. Panel A Panel B Panel C Panel D During the next 6 months, she was hospitalized seven times for diarrhea, vomiting, and fluid and electrolyte depletion. A psychiatric evaluation revealed depression that became progressively worse despite drug therapy. Dietitians counseled her several times about a gluten-free diet, and they thought she was adhering to it. Her diarrhea stopped when she fasted. Results of another stool alkalinization test and a fecal assay for Clostridium difficile toxin were negative. The prednisone dose was usually 20 to 30 mg/d. Trials of loperamide and metronidazole were ineffective. She developed a cushingoid appearance, anasarca, cataracts, cutaneous scaling, widespread ecchymoses, gluteal-fold pressure ulcers, and hepatomegaly. A computed tomographic scan of her abdomen showed hepatomegaly with fatty infiltration. She lost an additional 19 kg and became bedridden because of profound weakness and psychomotor retardation. Results of an electroencephalogram showed generalized slow activity. Results of cerebrospinal fluid examination and a computed tomographic scan of her head were normal. In July 1986, laboratory test results were as follows: hemoglobin, 99 g/L; serum albumin, 17 g/L; aspartate aminotransferase, 0.95 ; kat/L (57 IU); total bilirubin, 18 mol/L (1.1 mg/dL); and prothrombin time, 3 seconds prolonged. HLA typing using genomic DNA amplified by polymerase chain reaction and analyzed using molecular probes showed that she was HLA DQB1*0201/DQA1*0201 and DQB1*0604/DQA1* 0102. This correlates with serologic typing DR7/13 and DQ216. Central alimentation via a subclavian vein catheter was started at this time. During the next 6 months, prednisone therapy was tapered. The vomiting and cutaneous abnormalities disappeared, her weakness improved, and her neuropsychiatric status and blood test results became normal. With a low food intake, she still had 4 to 5 loose stools/d, and her diarrhea persisted after norfloxacin therapy. In January 1988, vena cava and bilateral subclavian vein thrombosis led to removal of the catheter and initiation of warfarin therapy. Her diarrhea increased and did not respond to octreotide acetate injections, and she lost another 14 kg by May 1988. Her serum carotene level was 0.3 mol/L [15 g/dL]. While on a diet containing 100 g of fat, her daily excretion of feces was 833 mL to 1299 mL and her daily excretion of fat was 20.8 g to 29.7 g. Findings in a duodenal biopsy specimen did not indicate improvement (Figure 1, panel B). A subclavian vein catheter was again inserted, but its occlusion resulted in resumption of central alimentation via a saphenous vein. She gained 4 kg, but her diarrhea persisted. In July 1988, oral cyclosporine was started at 5 mg/kg body weight per day (330 mg every 24 hours). Rapid improvement occurred. Within 2 weeks she was having only 1 to 2 formed stools/d. Parenteral nutrition and warfarin therapy were stopped. By July 1989, she had gained 15 kg since starting cyclosporine. Cyclosporine levels were 16 to 39 ng/mL (serum) and 128 to 324 ng/mL [whole blood]. The dose was not changed, and she had no drug toxicity. A duodenal biopsy specimen showed marked improvement (Figure 1, panel C), so the cyclosporine therapy was stopped. Her stools remained normal. In July 1992, histologic examination of a duodenal biopsy specimen was normal (Figure 1, panel D). She continued a gluten-free diet and gained an additional 33 kg by January 1993, becoming obese (114 kg). Discussion This patient with refractory sprue had diarrhea, depression, vomiting, nutritional deterioration, and severe enteritis while on a gluten-free diet and prednisone therapy. Her nutritional abnormalities, depression, and vomiting improved with parenteral nutrition and withdrawal of prednisone, but venous thrombosis caused the discontinuation of central alimentation. Because of her desperate condition, cyclosporine therapy was instituted. She had a remarkable clinical and histologic response to this drug. The most common explanation for unresponsive sprue is intentional or inadvertent gluten ingestion [1-3]. This seemed to be an unlikely reason for her failure to improve. The possibility that she was sensitive to nongluten dietary proteins [4] was not investigated. Collagenous sprue, a rare cause of refractory cases [5], was not evident. She lacked the specific DQ2 heterodimer found in most patients with iliac sprue. Some patients with refractory sprue respond to corticosteroid therapy [1-3], but patients tend to relapse when this therapy is discontinued [2]. Most patients who do not respond to corticosteroid therapy progressively deteriorate and usually die without parenteral alimentation [3]. Therefore, spontaneous remission coincidental with the use of cyclosporine is an unlikely explanation for the patient's prompt response after initiation of cyclosporine therapy. A 2-year-old patient with refractory sprue responded to cyclosporine [7], but a review of the literature revealed no reports of cyclosporine use in adults with sprue. Whereas the infant received a combination of cyclosporine and prednisone, my patient received cyclosporine alone. Furthermore, she took a lower dose on a body-weight basis than was given to the child. The baby had a relapse after the drug was stopped, but my patient's histologic improvement continued after the withdrawal of cyclosporine, with therapy limited to a gluten-free diet. Cyclosporine helped my patient, possibly by inhibiting the activation of cytotoxic T cells. Cyclosporine has also benefited patients with severe chronic ulcerative colitis, active Crohn disease, and other disorders attributed to immunologic mechanisms [8-10]. The response to cyclosporine in my patient suggests that the drug might be useful in other cases of refractory sprue.