SUN-042 Circulating macroprolactin a cause for discrepancy among hyperprolactinemic patients and is method dependent.

Abstract Disclosure: I.A. Hashim: None. K. Fankhauser: None. E. Abraham: None. Introduction: Circulating human prolactin is known to exhibit molecular heterogeneity the clinical significance of which remains unknown. However, it is sought to impart variable immunoreactivity causing false hyperprolactinemia and contributes to the discrepancy in prolactin results produced by different widely available immunoassay-based methods. This study compared prolactin levels obtained by two different immunoassays in patients suspected of hyperprolactinemia. Methods: De-identified aliquots of leftover plasma samples from patients who had prolactin levels measured using the laboratory automated immunoassay analyzer namely COBAS® (Roche Diagnostics) were stored at -20°C until further analysis. Samples were thawed and prolactin measured suing a different automated immunoassay analyzer namely Alinity® (Abbott Diagnostics). Samples with prolactin levels significantly (>20%) different obtained by the two immunoassay methods were tested for macroprolactin. Samples were diluted 1:1 in 15%(w/v) polyethylene glycol (PEG), centrifuged at 3000xg for 10 minutes, supernatant carefully aspirated, and prolactin levels measured using both automated immunoassays. Results: A total of 107 samples from patients suspected of hyperprolactinemia were entered into the study. Prolactin values obtained using the COBAS® and Alinity® systems ranged from 1.0 to 1808 ng/dL (median 16.7) and from 0.8 to 1471.4 ng/dL (median 15.3) respectively. Although prolactin values obtained by the two methods correlated well r = 0.98 (P<0.005), prolactin values obtained by COBAS® were significantly higher compared to those obtained by Alinity® instrument (P<0.01). Percentage difference in prolactin levels ranged from -90.9 to 149.2 (median -12.1%). 14 samples had prolactin levels > 20% (higher by COBAS®), compared to 6 samples being lower. Samples exhibiting >20% difference in prolactin levels by the two study methods were subjected to macroprolactin analysis. PEG treatment resulted in comparable prolactin values obtained by the two methods. Macroprolactin was detected in 6 samples (5.6%) using the COBAS® automated immunoassay system. Discussion: Prolactin results obtained by different automated immunoassays are often comparable, however, few patients with marked difference between the two systems were attributed to macroprolactin and that results became comparable following polyethylene glycol sample treatment. Presentation: Sunday, July 13, 2025