AIM: To retrospectively analyze the extended follow-up data from a registered randomized controlled trial initiated at our center in 2018 involving low-dose glucocorticoid (GC) treatment for isolated polymyalgia rheumatica (PMR), and to investigate GC tapering and discontinuation after clinical remission, as well as relapse following GC withdrawal. METHODS: A total of 86 treatment-na?ve patients with isolated PMR who were hospitalized at our center between January 2018 and December 2023 and underwent long-term follow-up were included. The median follow-up duration was 19 months [interquartile range (IQR), 10.0-53.0 months]. According to the treatment allocation in the original registered trial, patients were divided into three groups: GC monotherapy (n=25), GC plus methotrexate (MTX, 10 mg/week; n=32), and GC plus Tripterygium wilfordii polyglycosides (TW, 20 mg three times daily; n=29). All patients received an initial prednisone dose of 10 mg/day. Different from the prespecified standardized tapering protocol in the original RCT, GC doses were gradually tapered on an individualized basis according to clinical symptoms and inflammatory markers after achieving clinical remission. Subsequently, all patients received GC monotherapy for maintenance and further tapering.GC tapering, discontinuation, and subsequent clinical outcomes were evaluated, and potential factors associated with relapse after GC withdrawal were explored. RESULTS: All patients in the three treatment groups achieved clinical remission following induction therapy. The median time to remission was 6 months in all three groups, with IQRs of 3.0-7.5, 3.0-7.5, and 3.0-8.8 months in the GC monotherapy, GC plus TW, and GC plus MTX groups, respectively, with no statistically significant difference among groups. Overall, 47 patients (54.7%) discontinued GC therapy. The median interval from clinical remission to complete GC discontinuation was 20.8 months (IQR, 14.9–36.5 months). The GC discontinuation rates were 64.0%, 51.7%, and 50.0% in the GC monotherapy, GC plus TW, and GC plus MTX groups, respectively, with no statistically significant difference (P= 0.532). Among patients who discontinued GC therapy, the relapse rates during follow-up were 0% (0/16), 20.0% (3/15), and 12.5% (2/16), respectively, with no statistically significant difference among groups (P=0.181). Of the 47 patients who discontinued GC therapy, 5 (10.6%) experienced relapse after withdrawal, with a median time to relapse of 10.8 months (range, 6-24 months). Exploratory analysis showed that baseline neutrophil-to-lymphocyte ratio (NLR) was higher in relapsed patients than non-relapsed patients (median, 7.2 vs. 3.3; P=0.034). CONCLUSION: Gradual GC tapering and eventual discontinuation are feasible in some patients with isolated PMR after achieving clinical remission, although relapse may occur following withdrawal. Exploratory findings suggest elevated baseline NLR may correlate with post-withdrawal relapse. Given the limited number of relapse events, its predictive value needs further validation in larger prospective cohorts, and it is not recommended for direct clinical risk stratification at present.