Treating When You Cannot Trust Prolactin Levels: Pitfalls in Laboratory Measurements

Serum prolactin levels are typically lower than 200 nanograms per milliliter (ng/ml) in patients with microprolactinomas. When levels are discordant to imaging, the presence of macroprolactin must be considered. However, there may be other forms of prolactin that may not be measurable, and there is the potential for immune complexed prolactin to dissociate, posing a challenge to treatment goals based on laboratory measurements.

Case Description :

A 30-year-old female presented to her primary physician with a one year history of secondary amenorrhea. Workup was notable for prolactin 275 ng/ml, normal TSH, and negative pregnancy testing. MRI showed a four millimeter (mm) hypoenhancing lesion in the inferior pituitary. She started on cabergoline 0.25 micrograms (mcg) biweekly. Unfortunately, she stopped her medication and did not follow-up. Two years later she presented to endocrinology with seven months of secondary amenorrhea, prolactin 355 ng/ml and new MRI showing a stable four mm microadenoma. No other secondary cause for hyperprolactinemia was found. Given that the prolactin was out-of-proportion to her small adenoma, further testing was done. Repeat prolactin was 587 ng/ml with monomeric prolactin 271ng/ml (confirmed by dilution) and 54% macroprolactin. Repeat testing for macroprolactin with polyethylene glycol precipitation (Roche cobas e immunoassay) showed total prolactin 361ng/ml, unprecipitated 332ng/ml, and only 11% macroprolactin. The patient was initiated on cabergoline 0.25mcg biweekly. Two months later, she had resumption of menses. Repeat prolactin was 42 ng/ml, monomeric prolactin 41 ng/ml (confirmed by dilution) with 2% macroprolactin.

Discussion :

Here we present a case of a four mm pituitary adenoma with disproportionately high prolactin. Initially the elevation was partially explained by a macroprolactin component, but subsequent lab testing showed a low macroprolactin percent. With such variable lab measurements, the decision was made to treat based on clinical response, to resumption of menses. Interestingly, both prolactin and macroprolactin components were dramatically reduced with low dose cabergoline treatment. We postulate that the presence of other forms of unmeasured prolactin (possibly dimeric or “big” prolactin) were present, or that there was variable dissociation of prolactin from IgG (i.e., immune complexed macroprolactin or “big-big” prolactin) leading to such inconsistency with initial lab studies. When pre-treatment prolactin measurements are widely variable, treatment should ultimately be managed clinically.