Vein recognition has been extensively studied in the past. The topology of the vein patterns allows to adopt simple representations, based on edges and their endings, which are quite similar to the fingerprint minutiae. In this paper, the Minutia Cylinder-Codes (MCC), developed for the comparison of fingerprint minutiae, are adapted to characterize vein minutiae. The algorithm adaptation requires to understand the vein topology and how to consequently define the parameter sets. An optimization problem is defined and solved using a genetic algorithm (GA) approach. The experimental results prove the applicability of MCC on three extended vein data sets of wrist and palm dorsal. The performance of the reference parameter set is improved and can compete with other state-of-the-art algorithms for vein verification.