Associated factors with poor treatment response to initial glucocorticoid therapy in patients with adult-onset Still’s disease

Abstract

Background

High-dose glucocorticoids (GC) are first-line treatment for adult-onset Still’s disease (AOSD); however, some of the patients remain refractory to initial GC therapy, or rapidly relapse. The aim of this study was to identify prognostic factors for poor treatment response to initial GC therapy for AOSD.

Methods

Data on newly diagnosed AOSD patients were extracted from our database (n=71, mean age 51.6 years). The primary outcome was a poor treatment outcome at 4 weeks, which was defined as failure to achieve remission or relapse after achieving remission within 4 weeks, followed by administration of two or more rounds of GC pulse therapy or of any other immunosuppressive drugs.

Results

The initial mean dose ± standard deviation of prednisolone was 0.82 ± 0.23 mg/kg/day, and 34 (47.3%) patients received GC pulse therapy at week 0. Twenty-nine of 71 patients exhibited a poor treatment outcome at 4 weeks (40.8%). The second round of GC pulse therapy or immunosuppressive drugs was added in 17 or 24 of the 29 patients, respectively. These patients had higher baseline white blood cell (WBC) counts, serum ferritin levels, systemic feature score based on clinical symptoms (modified systemic feature score, mSFS), more hemophagocytic syndrome (HPS) over the 4 weeks, and the higher severity score based on modified Pouchot score or severity index of the Japanese Ministry of Health, Labour and Welfare, than the remaining 42 patients. Multivariable logistic regression model identified baseline WBC count as a prognostic factor for poor outcome (odds ratio per 1000/μl increment: 1.12, 95% CI 1.04–1.29), while thrombocytopenia, hyperferritinemia, and mSFS at baseline did not achieve statistical significance. Receiver-operating characteristic curve analysis showed that the optimal cut-off for WBC count was 13,050/μl. The Kaplan-Meier method showed the cumulative rate of poor treatment outcome to be 60.0% in patients with WBC ≥13,050/μl and 23.5% in those with WBC <13,050/μl.

Conclusions

A higher WBC count but not thrombocytopenia, hyperferritinemia, and mSFS at baseline was a significant prognostic factor for poor treatment outcome at week 4 in this retrospective cohort of AOSD patients. Our findings provide important information for determining the initial treatment strategy of newly-diagnosed AOSD.