Role of cannulated prolactin test in evaluation of hyperprolactinemia- A retrospective study

Objective: While guidelines propose a single elevated prolactin measurement drawn without excess venipuncture stress as sufficient for diagnosing hyperprolactinemia, this may lead to unnecessary evaluation in the setting of stress-induced hyperprolactinemia. In this study, we aimed to define the role of the cannulated prolactin test in confirmAîng hyperprolactinemia. Methods: We conducted a retrospective review of 757 patients with unexplained hyperprolactinemia who performed a cannulated prolactin test in a community-based referral endocrine clinic between 2000-2015. The prolactin test consisted of "test-baseline"levels taken at rest (T0), and cannulated measurements at 60 and 90 minutes (T60 and T90) without repeated venipuncture. The most recent prolactin level performed prior to the test (referral-prolactin) was collected. Results: Referral-prolactin was available for 621 (82%) patients, of whom 324 (52.2%) normalized at T0. The probability of normoprolactinemia at T0 was 50% if referral-prolactin was 2.0-fold the upper-limit-of-normal (ULN), yet only 5% if referral-prolactin was 5.0-fold the ULN. Of the 359 patients with hyperprolactinemia at T0, prolactin normalized at T60 and/or T90 in 99 (27.6%) patients. The probability of normoprolactinemia was low (<5%) in those with T0 prolactin levels >2.4-fold ULN. Overall, of 757 prolactin tests performed, only 260 (34.3%) patients had persistent hyperprolactinemia. Conclusion: Patients with referral-prolactin levels >5.0-fold the ULN, or a rested-prolactin (T0) >2.4-fold the ULN are unlikely to normalize during the cannulated test and consideration should be made to proceed directly with pituitary imaging. In patients with prolactin levels below these thresholds, the cannulated prolactin test may considAêrably reduce unnecessary investigations, treatment, and cost.