Predicting chronic spontaneous urticaria symptom return after omalizumab treatment discontinuation: exploratory analysis

Ferrer, Marta; Giménez Arnau, Anna Maria; Saldana, Diego; Janssens, Nico; Balp, Maria Magdalena; Khalil, Sam; Risson, Valéry

doi:http://dx.doi.org/10.1016/j.jaip.2018.04.003

Predicting chronic spontaneous urticaria symptom return after omalizumab treatment discontinuation: exploratory analysis

License

Citation

Ferrer M, Giménez-Arnau A, Saldana D, Janssens N, Balp MM, Khalil S et al. Predicting chronic spontaneous urticaria symptom return after omalizumab treatment discontinuation: exploratory analysis. J Allergy Clin Immunol Pract. 2018 Jul - Aug;6(4):1191-7. DOI: 10.1016/j.jaip.2018.04.003

Abstract

BACKGROUND: Omalizumab is highly effective in controlling chronic spontaneous urticaria (CSU) symptoms; however, patients can experience symptom return on treatment discontinuation. Pivotal clinical trials have identified 2 categories of patients who experience symptom return: rapid and slow. OBJECTIVE: The objective of this study was to identify potential predictors of the speed of symptom return after stopping omalizumab treatment. METHODS: Phase III randomized controlled trial (RCT) data from ASTERIA I (n = 319; 6 × 4 weekly injections of omalizumab 75, 150, 300 mg or placebo; NCT01287117) and ASTERIA II (n = 323; 3 × 4 weekly injections of omalizumab 75, 150, 300 mg, or placebo; NCT01292473) were pooled to identify predictors of symptom return after stopping omalizumab treatment (16-week follow-up). The least absolute shrinkage and selection operator regularization regression model was used to select predictive variables, and relapse probability was represented using heatmap visualizations. Model accuracy was tested using data from the GLACIAL phase III RCT (n = 336; 6 × 4 weekly injections of omalizumab 300 mg or placebo; NCT0126493). RESULTS: Of 746 variables assessed, 2 were selected by the model as predictors of symptom return: baseline urticaria activity score over 7 days (UAS7) and early area above the curve (AAC; determined by plotting the UAS7 scores across time points). Results suggest that high baseline UAS7 and low UAS7 AAC (slow decrease of symptoms) indicate a higher probability of rapid symptom return than low baseline UAS7 and high UAS7 AAC. CONCLUSIONS: These results suggest that the probability of rapid symptom return in patients with CSU who discontinue treatment with omalizumab can be estimated based on baseline UAS7 and early treatment response.