The Prolactin per Unit Tumor Volume Ratio Accurately Distinguishes Prolactinomas From Secondary Hyperprolactinemia due to Stalk Effect

Abstract

ObjectiveThe prolactin levels alone are insufficient to distinguish between some cases of prolactinomas and stalk effect. We aimed to formally characterize the relationship between serum prolactin and prolactinoma volume, determine a cutoff for prolactin/mm3 that accurately distinguishes prolactinomas from stalk effect, and validate this cutoff in a cohort selected to include ambiguous prolactin values ranging from 50 to 150 ng/mL. MethodsWe used the Research Patient Data Registry and transsphenoidal surgery database in our institution to retrospectively identify adult patients with clinically nonfunctioning (NF) tumors (primary analysis, n = 279; validation cohort, n = 10) and prolactinomas (primary analysis, n = 94; validation cohort, n = 18). Solid tumor volumes were measured by Visage 7 software, and cystic foci within tumors were excluded. ResultsProlactin levels were significantly correlated with prolactinoma volume (r2 = 0.801) but were not a relevant predictor of NF tumor size (r2 = 0.015). The prolactin/mm3 values did not overlap between NF tumors (median, 0.016; interquartile range, 0.009-0.028) and prolactinomas (median, 0.551; interquartile range, 0.265-0.845) (P < .0001). A cutoff of 0.065 ng/mL)/mm3 correctly discriminated between prolactinomas and NF tumors in all 401 patients in the primary analysis and validation cohort. ConclusionThe prolactin/volume ratio correctly distinguished all prolactinomas from stalk effect in this study, including a validation cohort specifically chosen for potential ambiguity. To our knowledge, this study is the first formal volumetric analysis of prolactin secretion in pituitary adenomas, and our results suggest that the measurement of prolactin/mm3 is a valuable tool to better characterize challenging cases of primary tumoral secretion versus secondary hyperprolactinemia due to stalk effect.