We aimed to identify predictors of inadequate response to glucocorticoid (GC) treatment in patients with polymyalgia rheumatica (PMR). We retrospectively studied 32 patients as a derivation cohort and 24 patients as a validation cohort. The patients were divided into two groups according to the response to GC treatment: GC-responders and GC-inadequate responders (GC-IRs). We compared laboratory data and bilateral shoulder ultrasound findings between the groups. Receiver operating characteristic (ROC) analysis was performed to determine the optimal cutoff value of candidate predictors of treatment response; the predictors were examined using multivariate logistic analysis. Gray-scale ultrasound findings of long head of the biceps (LHB) tenosynovitis and subacromial/subdeltoid (SAD) bursitis were scored semiquantitatively (0-3). A total gray-scale score (TGSS) was calculated as the sum of the gray-scale scores. In the derivation cohort, serum lactate dehydrogenase (LDH) levels and TGSS were significantly higher in GC-IRs than in GC-responders. On ROC analysis, the cutoff values of serum LDH levels ≥ 175 IU/ml and TGSS ≥ 5 were found to be the candidate predictors. Multivariate logistic analysis revealed an independent association of both the predictors with inadequate response to GC treatment. In the validation cohort, patients with one or both predictors exhibited a higher incidence of inadequate response to GC treatment. These findings indicate that the severities of LHB tenosynovitis and SAD bursitis evaluated using ultrasound and serum LDH levels are independent predictors of inadequate response to GC treatment in patients with PMR. Treatment adjustment based on prediction model may allow precise treatment of patients with PMR.
Keywords: Bursitis; Glucocorticoids; Polymyalgia rheumatica; Ultrasonography.