Why have circulating CGRP measurements in migraine not converged? A narrative review with a systematic literature search

Calcitonin gene-related peptide (CGRP) plays a well-established mechanistic and therapeutic role in migraine; however, studies have reported inconsistent circulating CGRP levels, hindering the neuropeptide’s translation into a clinically useful biomarker. Furthermore, recent analytical critiques of commercial assays and a large radioimmunoassay cohort study that did not reproduce the expected elevation of CGRP levels highlighted the need to reassess what current assays measure and whether values obtained using different platforms are comparable. We systematically searched MEDLINE via PubMed and Embase via Embase.com from inception to 14 July 2026 to identify human studies quantifying CGRP levels in plasma or serum in relation to migraine diagnosis, subtype, attack status, clinical characteristics, or treatment responses. Analytical and methodological reports evaluating CGRP immunoassays were also identified through citation searching and supplementary methods. Two reviewers independently screened records and extracted data from the included clinical reports; disagreements were resolved through discussion or consultation with a third reviewer. Owing to the substantial heterogeneity in populations, specimens, assay platforms, and outcome reporting, the findings were synthesized narratively. Across studies, reported circulating CGRP levels differed in direction and magnitude in association with differences in specimen type, pre-analytical processing, assay architecture, antibody epitopes, calibrators, purported isoform specificity, and reporting units. Elevation was reported more often in chronic migraine and ictal sampling, whereas the findings were inconsistent in interictal episodic migraine and were not consistently reproduced across platforms. Even assays that do not discriminate α-CGRP from β-CGRP yielded discordant findings, indicating that “total CGRP” measurements alone do not ensure analytical equivalence. The few direct comparisons between assay platforms also produced discrepancies, with most commercial assays providing limited information regarding antibody epitopes, intact mature CGRP recognition, cross-reactivity, and performance in specific sample matrices (e.g., plasma or serum). Circulating CGRP levels should be regarded as assay-specific immunoreactivity measurements rather than interchangeable estimates of the concentration of a single molecular species. The conflicting literature does not undermine the mechanistic importance of CGRP in migraine. Instead, it shows that mechanistic validity does not necessarily translate into clinical biomarker validity. Progress will require transparent analytical characterization, standardized pre-analytical reporting, direct blinded cross-platform comparisons, and orthogonal confirmation of the molecular species being measured. Not applicable.