MON-140 Prolactin to Testosterone Ratio Reduces Pituitary Magnetic Resonance Imaging Expenditures for Hypogonadal Men with Mild Hyperprolactinemia

Abstract

Background: Hyperprolactinemia is a common laboratory finding in men with symptomatic hypogonadism. Persistent elevations in serum prolactin (PRL) are typically evaluated with pituitary magnetic resonance imaging (pitMRI) to assess for structural pathology. However, this practice pattern may result in overutilization of pitMRI and unnecessary healthcare expenditures. Objective: We sought to examine the cost savings associated with utilizing combinations of serum PRL and the prolactin to testosterone ratio (PRL/T) to predict positive findings on pitMRI and obviate the need for unnecessary imaging studies. Methods: A retrospective case-control study was performed. Men <75 presenting with symptomatic hypogonadism and mild hyperprolactinemia (15-55 ng/ml) who obtained pitMRI at a tertiary care center were included. Individuals previously evaluated for pituitary abnormality were excluded, as were those presenting with symptoms strongly concerning for a pituitary abnormality (e.g. visual change, headache). Laboratory studies were abstracted from the electronic medical record and pitMRI results were reviewed. A cost analysis was performed based on the institutional expense of pitMRI and laboratory evaluations. The cost of measuring PRL and T was calculated for all patients. The expense of pitMRI was calculated for those screening positive. Results: 141 men were included in the study. Pituitary lesions were identified in 40/141 men (28%). The total cost of evaluation was calculated at $458,814.Ordering pitMRI when PRL/T >0.10 is 80% sensitive (32/40 lesions captured) and 64% specific (65/101 with normal anatomy excluded). 68/141 are indicated for pitMRI, while 73 patients avoid imaging. Employing this threshold reduces expenses by 46% with cost savings calculated at $212,795. The cost of identifying each lesion was estimated at $7,688.Ordering pitMRI when PRL/T >0.10 or when PRL >25 is 90% sensitive (36/40 lesions captured) and 48% specific (48/101 with normal anatomy excluded). 89/141 are indicated for pitMRI, while 52 patients avoid imaging. Employing this threshold reduces expenses by 33% with cost savings calculated at $151,580. The cost of identifying each lesion was estimated at $8,534.Ordering pitMRI when PRL/T >0.08 or when PRL >25 is 98% sensitive (39/40 lesions captured) and 32% specific (32/101 with normal anatomy excluded). 108/141 are indicated for pitMRI, while 33 patients avoid imaging. Employing this threshold reduces expenses by 21% with cost savings calculated at $96,195. The cost of identifying each lesion was estimated at $9,011. Conclusions: Serum PRL and PRL/T correctly predict the vast majority of pituitary lesions in patients with mild hyperprolactinemia, with screening costs increasing as more sensitive thresholds are employed. Future guidelines should establish a reasonable cutoff for pitMRI to minimize the expense of unnecessary imaging.