Endocrine Abstracts

Introduction: Surgical treatment is indicated in patients with dopamine agonist-resistant macroprolactinoma. However, the resection of the tumor is usually partial because of fibrotic changes under long-term dopamine agonist therapy. Thus, early identification and management of patients at high risk of dopamine agonist resistance is crucial to a complete resection outcome and a better remission. The aim of this study was to develop a score predicting dopamine agonist resistance in patients with macroprolactinoma.

Methods: This was a single-center, retrospective study including 52 patients with macroprolactinoma. All enrolled patients were treated with dopamine agonists (cabergoline or bromocriptine). Resistance to dopamine agonists was defined as failure to achieve normoprolactinemia and/or 30% tumor shrinkage. Univariate and multivariate logistic regression analyses were performed to assess predictive factors of dopamine agonist resistance.

Results: Dopamine agonist resistance was observed in 16 patients (31%). Its predictive factors were age at diagnosis <30 years (Odds Ratio (OR):5, 95%-confidence interval (CI):1.39-18.41, P=0.011), baseline prolactin level > 7000 µg/l (OR:4.861, 95%-CI:1.145-20.632, P=0.035), invasive adenoma (OR:10.263, 95%-CI:1.203-87.550, P=0.018), presence of hemorrhagic or necrotic components (OR=28.5, 95%-CI:3.273-248.175, P=10-3) and tumor size > 40 mm (OR=7, 95%-CI:1.717-28.545, P=0.004). In multivariate analysis, the presence of hemorrhagic or necrotic components was the only independent predictor of resistance. We developed a new score to predict dopamine agonist resistance based on five items including age <30 years (5 points), prolactin level >7000 (4 points), invasive adenoma (10 points), tumor size > 40 mm (7 points) and hemorrhagic or necrotic components (28 points). The total score was 54 points. The mean total score was 45.2 ± 8.4 in patients with dopamine agonist resistance and 17.5 ± 17.3 in patients responding to treatment (P=10^-3). This score had an area under the ROC curve of 0.902 (95%-CI: 0.816-0.987, P<10^-3). A total score > 39 points predicted the risk of resistance with a sensitivity of 87% and a specificity of 88%.

Conclusion: We developed in this study a new score for the prediction of dopamine agonist resistance in patients with macroprolactinoma. This new tool would have a decisive impact on surgery timing for better management of macroprolactinomas.