Satomi Haruki

Case of Linear IgA Bullous Dermatosis-involved Ulcerative Colitis

To the Editor: There are a number of reports of linear IgA bullous dermatosis (LABD) associated with preexisting inflammatory bowel disease (IBD), in particular ulcerative colitis (UC).1 We report the case of a patient who developed LABD while receiving treatment for UC. A 28year-old female. UC was diagnosed 1 year ago by colonoscopy (Fig. 1a) and biopsy specimen (Fig. 1b). She was receiving oral mesalazine at 2250 mg daily and prednisolone at 5 mg daily, and her disease activity was well controlled. However, within a few days after she developed a disease relapse, with severe diarrhea and bloody stool, the patient also developed bullae with pruritus on her trunk, gluteal region, and axillary fossa. She was therefore referred to our department. She had noticed clear and tender vesicles and bullae on erosion not associated with erythematous plaques on her trunk and limbs (Fig. 1c). There were no mucosal lesions. Routine laboratory examinations revealed evidence of an acute inflammatory reaction (white blood cell count, 11,000/ L, serum and an elevated erythrocyte sedimentation rate [ESR] of 77 mm/h). An immunological study was performed, including serological tests to determine the presence of autoantibodies, complement levels, rheumatologic markers and immunoglobulin levels, and thyroid tests; however, no significant abnormalities were detected. The skin biopsy specimen showed subepidermal neutrophilrich bullae. Direct immunofluorescence studies showed linear deposition of only IgA on the basement membrane. An indirect immunofluorescence test revealed the presence of circulating anti-basement membrane zone IgA antibodies at a titer of 1:80. Based on these studies, we diagnosed her disease as LABD. She did not receive systemic medications such as vancomycin with subsequent bowel flares, which are sometimes associated with drug-induced linear IgA disease. The intestinal disease activity entered remission spontaneously without any treatments and then, within a few weeks, the eruptions also cleared. No further therapies were required. The patient presented with several episodes of the bullae appearing synchronously with deterioration of the gastrointestinal symptoms and then disappearing within a short time after improvement of the gastrointestinal symptoms. We considered it important to evaluate the pattern of appearance of the skin lesions in relation to the activity of UC as assessed objectively. These usually incorporate the frequency of bowel movements and rectal bleeding, as well as the serum hemoglobin, albumin, and ESR values. The present study used the UC Activity Index (UCAI).2 In our case, the bullae cleared when the UCAI values were below 180, and relapsed when the values exceeded 190 (Fig. 2). Cutaneous diseases have been reported in 10% of patients with UC. Nonspecific eruptions are seen, including urticaria, angioedema, erythema, and purpura.3 In a study of 70 LABD patients selected from a British population, 5 (7.1%) had UC, even though the prevalence of UC in the UK in the general population is only 0.05%, while very few case reports with dermatitis herpetiformis (DH) and UC have been published in the literature.1,4,5 The exact reason for the association between UC and LABD remains unclear. IBDs may induce nonspecific immunoglobulin activation, triggering IgA crossreactive idiotype production against dermoepidermal or epidermal antigens, to produce specific bullous disorders.4,5 Antibodies enter the circulation and are deposited in the skin and mucous membranes because they crossreact with specific peptides in the lamina lucida and sublamina densa regions of the basement membrane. Antibody deposition is thought to stimulate an inflammatory response that damages the basement membrane, leading to mucocutaneous disease.6 Using previously described parameters of activity, we determined the UCAI3 in our patient; the bullae appeared with an increase of the UCAI, and disappeared with a decrease of the UCAI. These findings suggest that the appearance of LABD was related to the activity of UC, with antibodies being produced as the UC activity increased. To the best of our knowledge, this is the first case in which a clear relationship has been documented between LABD and the activity of UC.

Comparison of Diagnostic Accuracies of Various Endoscopic Examination Techniques for Evaluating the Invasion Depth of Colorectal Tumors

This study was designed to assess the clinical value of magnifying endoscopy combined with EUS for estimating the invasion depth of colorectal tumors. We studied 168 colorectal adenomas and carcinomas that were sequentially examined by conventional endoscopy followed by magnifying endoscopy and EUS in the same session to evaluate invasion depth. Endoscopic images obtained by each technique were reassessed by 3 endoscopists to determine whether endoscopic resection (adenoma, mucosal cancer, or submucosal cancer with slight invasion) or colectomy (submucosal cancer with massive invasion or advanced cancer) was indicated. The accuracy of differential diagnosis was compared among the examination techniques. The rate of correct differential diagnosis according to endoscopic examination technique was similar. The proportion of lesions that were difficult to diagnose was significantly higher for EUS (15.5%) than for conventional endoscopy and magnifying endoscopy. Among lesions that could be diagnosed, the rate of correct differential diagnosis was the highest for EUS (89.4%), but did not significantly differ among three endoscopic examination techniques. When it is difficult to evaluate the invasion depth of colorectal tumors on conventional endoscopy alone, the combined use of different examination techniques such as EUS may enhance diagnostic accuracy in some lesions.

RECTAL GRANULAR-CELL TUMOR DIFFICULT TO DISTINGUISH FROM CARCINOID TUMOR

A 60‐year‐old man had a positive fecal occult‐blood test on a medical check‐up. Colonoscopy revealed a yellowish‐white submucosal tumor 8 mm in diameter in the rectum. Endoscopic ultrasonography showed a well‐demarcated mass with a homogeneous, low‐level, internal echo in the second to third layers of the rectal wall. A carcinoid tumor was suspected, and the mass was resected endoscopically. Histopathological examination revealed a granular‐cell tumor. Gastrointestinal granular‐cell tumors rarely arise in the rectum, and the preoperative diagnosis of small lesions is often difficult. In our patient, granular‐cell tumor was difficult to differentially diagnose because the endoscopic and endoscopic ultrasonographic findings closely resembled those of carcinoid tumor. Interestingly, the endoscopic characteristics of the rectal granular‐cell tumor in our patient resembled those of a carcinoid tumor.