Pharmacogenetic mechanism of cilostazol‐induced headaches: Splicing‐mediated loss of <i>ABCC5</i> gene function

Abstract Background and Purpose Cilostazol, a phosphodiesterase 3 inhibitor, causes intolerable headaches in over one third of patients, frequently leading to treatment discontinuation. We investigated whether ABCC5 variant rs7636910 (NM_005688.4:c.1146A &gt; G) protects against cilostazol‐induced headaches through altered cyclic nucleotide signalling. Experimental Approach Analysis included 101 healthy Korean volunteers from Phase 1 clinical trials, with independent replication in a multi‐ethnic cohort (n = 369) from the All of Us Research Program. Functional consequences of rs7636910 were evaluated through splicing analysis, expression studies, transport assays and cellular models examining cyclic nucleotide dynamics and vascular responses. Key Results G allele carriers showed a reduced risk of moderate‐to‐severe headaches (OR = 0.29, 95% CI 0.12–0.73), replicated in the multi‐ethnic cohort (OR = 0.26, dominant model). The variant disrupted canonical splicing, causing a 38‐bp deletion, premature termination, and nonsense‐mediated decay, reducing ABCC5 expression by more than 20% in blood and thyroid tissues. Cilostazol was not an MRP5 substrate, and pharmacokinetics were genotype‐independent. In ABCC5 ‐knockdown HCASMCs (Human Coronary Artery Smooth Muscle Cells), baseline cAMP was elevated tonically with paradoxical cGMP accumulation, whereas drug‐induced cAMP fold‐increases remained comparable to controls. Tonic cAMP elevation desensitised PKA, abolishing PDE3 inhibitor‐induced vascular relaxation despite elevated cAMP—a loss recapitulated by PKA inhibition in MRP5‐normal cells. Conclusion and Implications The rs7636910 variant reduces cilostazol‐induced headaches through a signalling‐based mechanism independent of drug pharmacokinetics, suggesting that transporter variants affecting endogenous substrate handling are as clinically important as drug‐metabolising enzyme polymorphisms. This variant represents a candidate pharmacogenetic biomarker for identifying cilostazol‐tolerant patients.