ABSTRACT
Background: Multiple sclerosis (MS) comparative effectiveness research needs to go beyond average treatment effects (ATEs) and post-host subgroup analyses. Objective: This retrospective study assessed overall and patient-specific effects of dimethyl fumarate (DMF) versus teriflunomide (TERI) in patients with relapsing-remitting MS. Methods: A novel precision medicine (PM) scoring approach leverages advanced machine learning methods and adjusts for imbalances in baseline characteristics between patients receiving different treatments. Using the German NeuroTransData registry, we implemented and internally validated different scoring systems to distinguish patient-specific effects of DMF relative to TERI based on annualized relapse rates, time to first relapse, and time to confirmed disease progression. Results: Among 2791 patients, there was superior ATE of DMF versus TERI for the two relapse-related endpoints (p = 0.037 and 0.018). Low to moderate signals of treatment effect heterogeneity were detected according to individualized scores. A MS patient subgroup was identified for whom DMF was more effective than TERI (p = 0.013): older (45 versus 38 years), longer MS duration (110 versus 50 months), not newly diagnosed (74% versus 40%), and no prior glatiramer acetate usage (35% versus 5%). Conclusion: The implemented approach can disentangle prognostic differences from treatment effect heterogeneity and provide unbiased patient-specific profiling of comparative effectiveness based on real-world data.
ABSTRACT
Background: With increasing availability of disease-modifying therapies (DMTs), treatment decisions in relapsing-remitting multiple sclerosis (RRMS) have become complex. Data-driven algorithms based on real-world outcomes may help clinicians optimize control of disease activity in routine praxis. Objectives: We previously introduced the PHREND® (Predictive-Healthcare-with-Real-World-Evidence-for-Neurological-Disorders) algorithm based on data from 2018 and now follow up on its robustness and utility to predict freedom of relapse and 3-months confirmed disability progression (3mCDP) during 1.5 years of clinical practice. Methods: The impact of quarterly data updates on model robustness was investigated based on the model's C-index and credible intervals for coefficients. Model predictions were compared with results from randomized clinical trials (RCTs). Clinical relevance was evaluated by comparing outcomes of patients for whom model recommendations were followed with those choosing other treatments. Results: Model robustness improved with the addition of 1.5 years of data. Comparison with RCTs revealed differences <10% of the model-based predictions in almost all trials. Treatment with the highest-ranked (by PHREND®) or the first-or-second-highest ranked DMT led to significantly fewer relapses (p < 0.001 and p < 0.001, respectively) and 3mCDP events (p = 0.007 and p = 0.035, respectively) compared to non-recommended DMTs. Conclusion: These results further support usefulness of PHREND® in a shared treatment-decision process between physicians and patients.
