Predictors of Response to Fremanezumab in Migraine Patients with at Least Three Previous Preventive Failures: Post Hoc Analysis of a Prospective, Multicenter, Real-World Greek Registry

Objective: To define, in a real-world population of patients with high-frequency episodic (HFEM) or chronic migraine (CM), the predictive role of socio-demographic or phenotypic profiling of responders to fremanezumab. Patients and methods: Two-hundred and four adult fremanezumab-treated patients with either HFEM or CM, who failed to at least three preventive treatments, provided data at baseline on several individual socio-demographic and phenotypic variables. These variables were analyzed for their ability to independently predict the response (50-74% response rates) or super-response (>= 75% response rates) to fremanezumab. Patients were followed from 3-18 months of fremanezumab exposure. Results: The main finding to emerge from univariate analyses was that three baseline socio-demographic/clinical variables, i.e., age group 41-70 years (p = 0.02); female gender (p = 0.03); patients with HFEM (p = 0.001), and three clinical phenotypic variables, i.e., strict unilateral pain (p = 0.05); pain in the ophthalmic trigeminal branch (p = 0.04); and the "imploding" quality of pain (p = 0.05), were significantly related to fremanezumab response. However, in multivariate analysis, only HFEM (p = 0.02), the presence of strict unilateral (p = 0.03), and pain location in the ophthalmic trigeminal branch (p = 0.036) were independently associated with good fremanezumab response. Allodynia (p = 0.04) was the only clinical predictive variable of super-responsiveness to fremanezumab. Conclusions: A precise phenotypic profiling with identification of pain characteristics consistent with peripheral and/or central sensitization might reliably predict the responsiveness to fremanezumab in migraine prophylaxis.