Macroprolactinoma shrinkage during cabergoline treatment is greater in naive patients than in patients pretreated with other dopamine agonists: a prospective study in 110 patients.

To investigate whether previous treatment with bromocriptine (BRC) or quinagolide (CV) impairs a subsequent response to long-term cabergoline (CAB) treatment, we prospectively studied 110 patients with macroprolactinoma. Four groups of patients were considered: 1) naive: 26 untreated patients with a mean serum PRL levels of 1013.4 +/- 277.7 microg/L (+/- SEM; range, 185.5-5611 microg/L); 2) intolerant: 19 patients previously shown to be intolerant of BRC treatment with a mean serum PRL level of 539.4 +/- 172.2 microg/L (range, 174-3564 microg/L); 3) resistant: 37 patients shown to be resistant/hyporesponsive to BRC, CV, or both, with a mean serum PRL level of 602.6 +/- 136.8 microg/L (range, 148-3511 microg/L); and 4) responsive: 28 patients previously treated with BRC or CV for 1-5 yr, achieving normoprolactinemia and restoration of gonadal function, but no longer treated with BRC or CV because of poor compliance or because the drug was not available. After a 15- to 30-day washout period, the serum PRL level was 397 +/- 43.1 microg/L (140-978 microg/L). CAB treatment was given at doses ranging 0.25-3.5 mg weekly for 1 yr to 110 patients, for 2 yr to 104 patients, and for 3 yr to 81 patients. Magnetic resonance imaging was performed before and after 12, 24, and 36 months of CAB treatment to evaluate significant tumor shrinkage (>80\% reduction of pretreatment tumor volume). Among the 26 naive patients, normoprolactinemia was achieved in 21 (80.8\%) after 1-6 months at 0.25-2 mg/week and in 5 patients after 24 months at 0.5-3 mg/week. Tumor volume was reduced from 1431.5 +/- 310.3 to 47.2 +/- 21.5 mm3 (P < 0.0001); average tumor shrinkage was 92.1 +/- 2.9\%; significant tumor shrinkage was observed in 92.3\% of patients, and tumor mass completely disappeared in 16 patients (61.5\%). Among the 19 intolerant patients, normoprolactinemia was achieved in 18 (94.7\%) after 1-6 months of CAB treatment at 0.25-1 mg/week. One patient remained mildly hyperprolactinemic. Tumor volume was reduced from 1925 +/- 423.1 to 842.0 +/- 330.7 mm3 (P < 0.001); average tumor shrinkage was 66.2 +/- 6.4\%; significant tumor shrinkage was obtained in 42.1\% of patients, and tumor mass completely disappeared in 4 patients (21\%). Among the 37 resistant patients, normoprolactinemia was achieved in 19 (51.3\%) after 6-12 months at 1-2 mg/week and in the remaining 18 patients after 18-24 months at 3-3.5 mg/week. Tumor volume was reduced from 1208.0 +/- 173.7 to 471.2 +/- 87.3 mm3 (P < 0.005); average tumor shrinkage was 58.4 +/- 4.9\%; significant tumor shrinkage was obtained in 10 of 33 patients (30.3\%), and in no patient did tumor mass completely disappear. Among the 28 responsive patients, normoprolactinemia was achieved in 23 (82.1\%) after 1-6 months at 1-2 mg/week and in 5 patients after 12 months at 3 mg/week. Tumor volume was reduced from 1351.3 +/- 181.5 to 757.1 +/- 193.6 mm3 (P < 0.01); average tumor shrinkage was 59.2 +/- 6.2\%; significant tumor shrinkage was obtained in 10 of 26 patients (38.4\%), and tumor mass completely disappeared in 4 patients (15.4\%). Nadir PRL levels and percent tumor shrinkage during CAB treatment in naive patients were significantly lower (P < 0.001) and higher (P < 0.001), respectively, than those in the remaining three groups, and the average weekly dose of CAB in resistant patients was significantly higher (P < 0.001) than that in the remaining three groups. A significant association was found between tumor shrinkage and previous treatments (chi2 = 27.1; P < 0.0001). At the multistep correlation analysis, nadir PRL levels were the strongest predictors of tumor shrinkage (r2 = 0.556; P < 0.0001), followed by CAB dose (r2 = 0.577; P < 0.0001). The tolerability was excellent in 105 patients (95.4\%). In conclusion, the prevalence of macroprolactinoma shrinkage after CAB treatment at standard doses for 1-3 yr was higher in naive patients (92.3\%) than in intolerant (42.1\%), resistant (30.3\%), and responsive patients (38.4\%). Thus,