Predictors of time to relapse following ustekinumab withdrawal in patients with psoriasis who had responded to therapy: An eight-year multicenter study

Hsien-Yi Chiu, Rosaline Chung-Yee Hui, Tsen-Fang Tsai, Yang-Ching Chen, Nien-Feng Chang Liao, Po-Hua Chen, Po-Ju Lai, Ting-Shun Wang, Yu-Huei Huang

Research output: Contribution to journalArticlepeer-review

Abstract

BACKGROUND: Data on predictors and time to relapse in psoriasis patients discontinuing therapy in a real world setting are scarce.

OBJECTIVE: To investigate predictors of relapse after withdrawal of ustekinumab in psoriasis patients.

METHOD: This study screened 500 psoriasis patients who received ustekinumab (669 treatment episodes (TEs)) between 2011 and 2018. Overall, 202 patients (304 TEs), who had responded to therapy and were withdrawn from ustekinumab treatment, were included.

RESULTS: The cumulative probabilities of relapse-free at 6 months, 12 months, 18 months, 24 months and 36 months of withdrawal from ustekinumab treatment was 49.3%, 12.6%, 5.3%, 4.7% and 1.6%, respectively. Multivariate regression analyses with a generalized estimating equation showed that after adjustments, biologics-naive, the maximum Psoriasis Area and Severity Index (PASI) improvement on ustekinumab, time to achieve PASI-50 after initiation of ustekinumab, family history of psoriasis, chronic kidney disease, and immunosuppressants use while off ustekinumab were significant predictors of time to relapse following ustekinumab discontinuation.

LIMITATION: Non-randomized allocation of duration of treatment and follow-up.

CONCLUSION: Given high rates of relapse, withdrawal of ustekinumab from patients with well-controlled psoriasis cannot be recommended.

Original languageEnglish
JournalJournal of the American Academy of Dermatology
DOIs
Publication statusE-pub ahead of print - Jan 28 2019

Fingerprint Dive into the research topics of 'Predictors of time to relapse following ustekinumab withdrawal in patients with psoriasis who had responded to therapy: An eight-year multicenter study'. Together they form a unique fingerprint.

Cite this