Real-world optimization of cutoff values for enzyme immunoassay-based plasma-free fractionated metanephrines for the diagnosis of pheochromocytoma and paraganglioma: a multicenter retrospective study.

Pheochromocytoma and paraganglioma (PPGL) require sensitive biochemical screening to prevent delayed or missed diagnoses. In Japan, plasma-free metanephrines (pMN/pNMN) measured using enzyme immunoassay have been covered by national insurance since 2019; however, the real-world diagnostic performance of the currently applied cutoff values has not been evaluated. We retrospectively identified consecutive adults who underwent pMN/pNMN testing in routine practice at five endocrinology departments between 2019 and 2025, of whom 817 were included in the final analysis. Patients were classified post hoc as having PPGL (n = 107) or serving as controls (n = 710) based on pathological findings, functional imaging, and clinical follow-up. Diagnostic performance at the current cutoff values (pMN >130 pg/mL or pNMN >506 pg/mL) was evaluated by sensitivity and specificity, compared with spot and 24-hour urinary metanephrine measurements using receiver operating characteristic (ROC) analyses, and optimized cutoff combinations were determined using the Youden index. The current cutoff values showed high specificity (99.9%) but limited sensitivity (82.2%), particularly in patients with abdominal paraganglioma (76.0%). Plasma and urinary metanephrine measurements demonstrated comparable performance in the ROC analyses. False-negative cases based on the current cutoff values had significantly smaller tumors than true-positive cases (p = 0.016). Urinary biochemistry and functional imaging offered crucial complementary diagnostic value. Our optimized thresholds (pMN >120 pg/mL or pNMN >290 pg/mL) significantly enhanced sensitivity (92.5%) while maintaining high specificity (99.2%), consistent with postoperative biochemical remission. These findings suggest that current pMN/pNMN cutoff values may limit diagnostic sensitivity, necessitating optimized thresholds and a multimodal approach in accordance with the PPGL guidelines. Registration Number: UMIN000060658.