Clinical and pathological effects of bromocriptine on prolactin-secreting and other pituitary tumors.

Bromocriptine inhibits prolactin secretion and causes size reduction of prolactin-secreting adenomas. The effect of the drug upon pituitary tumors other than prolactinomas is uncertain. The authors report a prospective series of 12 patients with pituitary macroadenomas in whom bromocriptine was administered for 6 weeks prior to transsphenoidal surgery. Five of the patients had computerized tomographic documentation of significant reductions in tumor size (Group A) and six had no change (Group B) during 3 and 6 weeks of bromocriptine administration. One patient who demonstrated size reduction in his tumor was not assigned to either group as he was treated with high-dose dexamethasone concurrently with the bromocriptine. Pathological examination (light and electron microscopy and immunocytochemistry) indicated that all Group A patients harbored tumors with prolactin granules whereas all Group B tumors lacked such granules. Adenoma cells in the responsive tumors were involuted with reduced cytoplasmic, nuclear, and nucleolar areas. Neither widespread cell necrosis, infarction, nor vascular injury was observed. Two of the five Group A patients discontinued bromocriptine prior to completion of the 6-week protocol and had a rapid return of their tumors to pre-treatment size. Although bromocriptine has been reported to cause shrinkage of nonfunctional tumors, there was no radiological evidence of size reduction or pathological changes in the nonfunctional tumors of this series. Interestingly, serum levels of prolactin were modestly elevated (84 and 113 ng/ml) in two of the six Group B patients, an elevation due to stalk compression rather than secretion by adenoma cells. This finding underscores the fact that failure of bromocriptine to reduce pituitary tumor size in the presence of hyperprolactinemia may occur because the tumor is other than a prolactinoma. This is the first moderate-sized group of patients in whom pathological changes in responsive prolactinomas during bromocriptine therapy have been demonstrated. As bromocriptine is not tumoricidal, and thus not curative, there is insufficient evidence to recommend this drug as primary therapy for either prolactin-secreting or nonfunctional macroadenomas, but the drug may have potential as a preoperative adjunct to effect shrinkage of prolactinomas and theoretically, at least, make excision easier and possibly more complete.