Objective: To determine the factors influencing the outcome of transethmoidal partial hypophysectomy for suspected prolactinoma and the predictive value of pre and post-operative dynamic PRL function tests.
Design: A retrospective study of patients undergoing surgery for a suspected prolactinoma in Cardiff between 1979 and 1989.
Patients: Eighty-two hyperprolactinaemic patients (75 women, seven men) diagnosed as having a prolactinoma on the basis of dynamic PRL function tests, radiological investigation and exclusion of other causes.
Measurements: TSH and PRL responses to domperidone (10 mg i.v.) and TRH (200 micrograms i.v.) measured preoperatively, 2 months post-operatively, and annually thereafter. CT scan performed preoperatively in 58 patients. Operative findings, including adenoma size, documented in each case.
Results: Forty-two patients (51%) had microadenomas (less than 10 mm), 37 (46%) had macroadenomas and in three no tumour was found at operation. Preoperatively, normal responses of both TSH (incremental rise less than 2.0 mU/l) and PRL (greater than 100% rise) to domperidone were observed in two patients only: both had an abnormal vascular supply to the pituitary rather than an adenoma. Serum PRL was normalized in the early post-operative period (less than 72 h; 'early cure') in 65 patients (79%). The highest early cure rate (96%, n = 26) was in patients with adenomas of 5-9 mm, lower rates being achieved for lesions of 10-19 mm (80%, n = 30), less than 5 mm (63%, n = 19) or greater than or equal to 20 mm (57%, n = 7). The early cure rate was strongly correlated with preoperative PRL, ranging from 100% in patients with PRL less than 1000 mU/l (n = 13) to zero in those with PRL greater than 10,000 mU/l. Dopamine agonist therapy of between 5 weeks and 4 years duration prior to surgery was associated with a significantly reduced early cure rate (60 vs 94%, P less than 0.02) in macroadenoma but not microadenoma patients. Recurrent hyperprolactinaemia during mean follow-up of 51.7 months occurred in eight patients (12%), in five cases within 2 months of surgery and in the others at 26, 48 and 50 months. Recurrence could not be predicted from any preoperative parameter, but a serum PRL greater than 150 mU/l 1-3 days following microadenomectomy was associated with early recurrence and probably indicates failed surgery. An abnormal response of TSH to domperidone was documented 2 months post-operatively in 11/60 patients with normal basal PRL, and preceded all three late recurrences. Of four patients with abnormal responses of both PRL and TSH at this time, two have relapsed to date.
Conclusions: In carefully selected patients, partial hypophysectomy is an acceptable alternative to medical treatment for prolactinoma. Preoperatively, dynamic tests accurately identified those patients whose hyperprolactinaemia was non-adenomatous in origin and, post-operatively, identified a subgroup of patients at increased risk of late recurrence.