CGRP, VIP and PACAP are important in migraine pathophysiology, but the diagnostic value of their plasmatic determination is controversial.

We conducted a prospective study in two headache units in Spain. Patients with episodic (EM) and chronic migraine (CM) according to ICH3Beta classification were included. Migraine impact was assessed with MIDAS and HIT-6. Plasmatic CGRP, VIP and PACAP were determined by ELISA. We stablished 3 subgroups according to treatment: botulinum toxin (BT), any oral preventative (AOP) and non-treated (NT). Statistical analysis:χ² for categorical variables, parametric or non-parametric tests according to normality (alpha = 0.05) for quantitative variables, and correction with Bonferroni.

A total of 128 patients (86 CM, middle age 44.8±11.2; 42 EM, 49.2±12.2) were recruited. In the univariate analysis, CGRP, VIP and PACAP were higher in CM vs EM (p<0.001). The multinomial regression model could only correctly predict 60.2% of CM, 68.8% of EM and 15.5% of healthy controls. Only VIP (B=0.009; p=0.012) and PACAP (B=0.003, p=0.002) predicted CM diagnosis, but not EM. CGRP did not predict CM or EM. Individually, CGRP was the best to classify CM (65%) and PACAP for EM (88.5%). Neuropeptide levels correlated with MIDAS. Only PACAP correlated with total headache days (p=0.002). Neuropeptide levels were higher in patients under BT. CGRP and VIP did not differentiate AOP from NT, but PACAP did (higher in AOP than in NT; p <0.001). In the sensitivity(Se) – specificity(Sp) analysis, Sp was low for all of them as CM biomarkers (VIP:Se=0.77, Sp=0.43; PACAP:Se=1, Sp= 0.53 ; CGRP:Se=0.60, Sp=0.445).

In real practice, determination of plasmatic CGRP, VIP and PACAP in migraine patients have limited utility for clinical diagnosis.