Development of a scoring system to predict outcomes of i.v. corticosteroid pulse therapy in rapidly progressive alopecia areata adopting digital image analysis of hair recovery.

Alopecia areata (AA) is a common autoimmune disease manifesting varying degrees of hair loss. Rapidly progressive AA (RP-AA) is a severe subtype of AA and often resistant to skin-directed treatments. i.v. corticosteroid pulse therapy has been applied for RP-AA; however, the treatment outcome can only become evaluable several months after the intervention, discomposing the patients. In this study, we attempted to develop a scoring system to predict treatment outcomes based on statistical correlations between newly identified predictors and the recovery rates calculated by digital image analysis. Thirty RP-AA patients (15 men and 15 women) who underwent pulse therapy and demonstrated total hair loss during the clinical course were included. The percentages of hair regrowth (%HR) at 6 months after the treatment were quantitatively calculated by image analysis software. The correlation between %HR and clinicopathological and immunological variables were statistically assessed. The analysis identified four confirmatory contributors including female sex (P = 0.015), absence of previous AA history (P = 0.02), lower peripheral blood eosinophil count (P = 0.02) and mild to moderate cell infiltration around the hair bulb (P = 0.034), together with a potential contributor, namely absence of atopic dermatitis in their medical history (P = 0.08). The scoring system was developed by double counting confirmatory variables and single counting a potential variable. Importantly, the scores obtained by this system demonstrated significant correlation with %HR (r = 0.61, P < 0.001). The usefulness of this scoring system was further validated by assessing additional 20 cases of RP-AA. When combined with a recently published algorithm for early detection of self-healing subset, the current scoring system may help strategize the therapeutic approach for RP-AA.