Hyperprolactinemia: comparison of thyrotropic-releasing hormone and tomography

A group of 95 women with unexplained hyperprolactinemia (over 20 ng/mL) underwent radiologic examination of the sella turcica with hypocycloidal polytomography (N = 58), computed axial tomography (N = 8), or both (N = 29). All patients also underwent a thyrotropin-releasing hormone (TRH) stimulation test, with serum prolactin (PRL) measurement before and 20 and 30 minutes after a 500-micrograms intravenous bolus of TRH. Their PRL responses were compared with those of two control groups, nine normal women in the follicular phase of the menstrual cycle, and 13 women in the first five months of gestation with pregnancy-related hyperprolactinemia. Both control groups exhibited PRL increases with 95% confidence limits at least 200% above baseline levels. In all, 12 patients from the study group also had a normal PRL response (more than a 200% increase) to TRH, and none of these women had tomographic findings consistent with a pituitary tumor. The remaining 83 women all had diminished or absent PRL increases after TRH administration; 46 (55%) of these patients had radiographic evidence of an adenoma, whereas 37 (45%) had no clear signs of a tumor on either polytomography or computed axial tomography. No patient with a baseline PRL level in excess of 60 ng/mL had a normal PRL response to TRH. The results of the study indicate that 1) in patients with PRL between 20 and 60 ng/mL, a normal TRH test can be relied upon to avoid the expense and radiation of tomography (computed axial tomography or polytomography), 2) there is no benefit to be obtained in performing a TRH test in patients with a baseline PRL level over 60 ng/mL, and 3) about 45% of patients with hyperprolactinemia and an abnormal TRH test have a normal computed tomography or polytomography. These patients may have a small adenoma, and thus warrant closer follow-up than patients with a normal TRH test.