Measurements of plasma metanephrines by immunoassay versus LC-MS/MS for diagnosis of pheochromocytoma

D Weismann; A Raida; T Deutschbein; A Prejbisz; A Januszewicz; HS Willenberg; M Peitzsch; R Därr; R Klemm; G Manz; M Bidlingmaier; G Eisenhofer; M Fassnacht

doi:10.1055/s-0034-1371996

Experimental and Clinical Endocrinology & Diabetes, Table of Contents

Measurements of plasma metanephrines by immunoassay versus LC-MS/MS for diagnosis of pheochromocytoma

D Weismann ¹, A Raida ¹, T Deutschbein ¹, A Prejbisz ², A Januszewicz ², HS Willenberg ³, M Peitzsch ⁴, R Därr ⁴, R Klemm ⁵, G Manz ⁵, M Bidlingmaier ⁶, G Eisenhofer ⁴, M Fassnacht ⁶

¹Medizinische Klinik I, Universität Würzburg, Endokrinologie, Würzburg, Germany
²Institute of Cardiology, University of Warsaw, Department of Hypertension, Warsaw, Poland
³University of Duesseldorf, Department of Endocrinology and Diabetology, Duesseldorf, Germany
⁴University Hospital Dresden, Institute of Clinical Chemistry and Laboratory Medicine, Dresden, Germany
⁵LDN Labor Diagnostika Nord GmbH und Co KG, Nordhorn, Germany
⁶LMU München, Endocrine Research Unit, München, Germany

Abstract

Aim: To compare enzyme immunoassay (EIA) with liquid chromatographic tandem mass spectrometric (LC-MS/MS) measurements of normetanephrine (NMN) and metanephrine (MN) for diagnosis of pheochromocytoma (PHEO).

Methods: Subjects included 341 patients (174 males) with a mean age of 52 yr (range 13 – 86) tested for PHEO, which was confirmed in 54 patients. Plasma samples were analyzed for NMN, MN and methoxytyramine (MTY) by LC-MS/MS at Dresden and for NMN and MN by immunoassay at Nordhorn.

Results: Plasma concentrations of NMN and MN were measured 44% and 26% lower by the EIA than by LC-MS/MS. For the EIA, and using the upper cut-offs (UC) stipulated by the manufacturer (NMN 180 pg/mL; MN 90 pg/mL), diagnostic sensitivity was 73.6% with a specificity of 98.9%. In contrast, LC-MS/MS measurements of NMN and MN returned a diagnostic sensitivity of 98.1% with a specificity of 99.7%; sensitivity increased to 100% upon inclusion of MTY with minimal loss of specificity (99.3%). Areas under ROC curves for the EIA (0.993) were similar to those for LC-MS/MS, with and without MTY (0.999 and 0.985, resp.), indicating excellent diagnostic performance by both methods and suggesting that the low diagnostic sensitivity by EIA reflected use of inappropriately high UCs. Indeed, diagnostic sensitivity for the EIA was increased to 96.2% with minimal loss in specificity (97.2%) after use of age-adjusted UCs for NMN (defined by the equation UC_NMN= 0.0002169 x [age³] + 56.5) and a UC for MN of 65 pg/mL, as corrected for method-dependent differences from previously defined cut-offs¹.

Conclusions: Diagnostic performance of EIA measurements of NMN and MN can be nearly as good as for LC-MS/MS measurements, but this requires use of appropriately determined UCs. For the EIA we recommend a UC of 65 pg/mL for MN and age-adjusted UCs for NMN ranging from 57 pg/mL for a 5 yr old to a maximum of 116 pg/mL for those over 65 yr.

References:

1. Eisenhofer, G. et al. (2013) Annals of Clinical Biochemistry.