LAUSR

Fig 1. ROC curve for bullous pemphigoid 180 (BP180) autoantibody diagnostic test. Serum BP180 autoantibody was measured in 173 dermatologic in-patients seen during March 2016-October 2017 by using the BP180 NC16A ELISA (MBL, Nagoya, Japan). The performance of this test, measured by the area under the curve, was 0.866. When maximizing the Youden index (J 1⁄4 sensitivity 1 specificity e 1), the optimum cutoff value was set at 27.2. ROC, Receiver operative characteristic. To the Editor:Detection of serumbullous pemphigoid 180 (BP180) autoantibody by enzyme-linked immunosorbent assay (ELISA) has proven a practical and reliable diagnostic test for bullous pemphigoid (BP). However, in multiple studies, elevated serum BP180 autoantibody has been reported in non-BP dermatologic conditions, highlighting the necessity to reevaluate the presence of BP180 autoantibody in non-BP dermatologic conditions and establish an optimized cutoff value to better differentiate genuine BP from BP-like conditions. To achieve these goals, we performed a crosssectional study including all the in-patients who were tested for serum BP180 autoantibody during March 2016-October 2017 in a tertiary dermatologic center in Shanghai, China. The serum level of BP180 autoantibody was analyzed by using the commercially available BP180 NC16A ELISA (MBL, Nagoya, Japan), for which a result[9 U/mL was defined as abnormal. Receiver operating characteristic (ROC) curve analysis, a plot of the test sensitivity versus 1 especificity (Fig 1), was used to generate paired sensitivity and specificity values on the basis of the input BP180 autoantibody titers through SPSS software (version 21.0, IBM, Armonk, NY). The optimum cutoff value to differentiate BP from non-BP patients was calculated on the basis of maximizing the Youden index (J 1⁄4 sensitivity 1 specificity e 1). A total of 173 in-patients were included, consisting of 26 patients with BP and 147 patients with other dermatologic conditions for whom a BP diagnosis was suspected but later excluded. On the basis of a comprehensive assessment of clinical, histologic, and immunologic findings, we found 14.3% (21/147) of patients with non-BP dermatologic conditions showed abnormal BP180 autoantibodies (Table I). SerumBP180 autoantibodieswere found in 7.69% (4/52) of eczema patients, 12% (3/25) of pemphigus patients, 4.34% (1/23) of erythema multiform patients, 7.14% (1/14) of prurigo patients, and 44.4% (4/9) of Stevens-Johnson syndrome. In these diseases, the primary inflammation around the basement membrane zone might lead to the exposure or release of BP180 antigen and subsequent hormonal immune responses, as hypothesized by former researchers. The titers of BP180 autoantibodies in non-BP patients were significantly lower than those of