SAT-069 Prolactin Unveiled - Differentiating Lactotroph Adenoma or Else Through Tumor Metrics

Abstract Disclosure: K. Sanpawithayakul: None. T. Maythapak: None. Background: Hyperprolactinemia associated with pituitary adenomas can arise from lactotroph adenomas (LA) or non-functioning pituitary adenomas (NFPA) due to stalk interruption effect. Typically, serum prolactin levels are directly proportional to prolactinoma size. A serum prolactin threshold of >250 μg/L is clinically indicative of macroprolactinoma; however, relying solely on serum prolactin levels is insufficient for preoperative differentiation between LA and NFPA in some situations. Ratios of serum prolactin to maximal tumor diameter (MD), MD squared, and tumor volume (TV) may offer alternative diagnostic methods, but TV is less frequently utilized in clinical settings compared to MD. Objectives: We aimed to investigate correlation between TV and MD, and diagnostic efficacy of PRL/MD and PRL/TV ratios for distinguishing LA from non-LA. Methods: We conducted a retrospective analysis of 111 patients underwent pituitary tumor resection at a single institution during 2016-2024. Patient characteristics were compared based on presence of LA or non-LA, confirmed through immunohistochemistry. Tumor measurements were obtained via the Picture Archiving and Communication System (PACS), and tumor volumes were calculated using 3D Slicer, a free open-source medical imaging platform. Correlation analysis was performed to assess relationship between MD and TV, and ROC curves were generated to evaluate predictive models and potential enhancement from combined parameters. Results: Average age of patients was 44.34 ± 15.07 years in LA group and 53.67 ± 14.43 years in non-LA group (p=0.002). Average serum prolactin level was significantly higher in LA group (285 μg/L (53.8, 3,975)) compared to non-LA group (36.80 μg/L (23.90, 59.49), p<0.001). Although there was no significant difference in MD and TV between groups, a strong correlation was identified between these two preoperative markers (R=0.9421, p<0.001). Median PRL/MD ratio was higher in the LA group (18.74 μg/(L x mm) (2.25, 112.45) vs 1.35 μg/(L x mm) (0.73, 2.07), p<0.001), as was median PRL/TV ratio (59.6 μg/(L x cm³) (9.95, 286.52) vs 3.84 (1.65, 7.37), p<0.001). Prolactin level threshold of 119 μg/L demonstrated a sensitivity of 0.63, a specificity of 0.97, and an AUC of 80. The PRL/MD ratio cutoff at 2.49 μg/(L x mm²) showed a sensitivity of 0.73, a specificity of 0.87, and an AUC of 80. Meanwhile, the PRL/TV ratio cutoff at 9.87 μg/(L x cm³) resulted in a sensitivity of 0.76, a specificity of 0.86, and an AUC of 81. Notably, addition of PRL/MD or PRL/TV to preoperative serum prolactin did not enhance diagnostic accuracy in logistic regression. Conclusion: Maximal dimension and tumor volume exhibited a strong correlation, indicating potential utility as marker when tumor volume measurements are not available. Presentation: Saturday, July 12, 2025