Sympathoadrenal activity is decreased during overnight rest. This study assessed whether urinary-free normetanephrine, metanephrine and methoxytyramine in overnight/first-morning urine collections might offer an alternative to measurements in 24-h collections or plasma for diagnosis of pheochromocytoma and paraganglioma (PPGL).
Design and methods
Prospective multicenter cross-sectional diagnostic study involving 706 patients tested for PPGL, in whom tumors were confirmed in 79 and excluded in 627 after follow-up. Another 335 age- and sex-matched volunteers were included for reference purposes. Catecholamines and their free O-methylated metabolites were measured in 24-h collections divided according to waking and sleeping hours and normalized to creatinine. Plasma metabolites from blood sampled after supine rest were measured for comparison.
Urinary outputs of norepinephrine, normetanephrine, epinephrine and metanephrine in the reference population were respectively 50 (48–52)%, 35 (32–37)%, 76 (74–78)% and 15 (12–17)% lower following overnight than daytime collections. Patients in whom PPGLs were excluded showed 28 (26–30)% and 6 (3–9)% day-to-night falls in normetanephrine and metanephrine, while patients with PPGLs showed no significant day-to-night falls in metabolites. Urinary methoxytyramine was consistently unchanged from day to night. According to receiver-operating characteristic curves, diagnostic accuracy of metabolite measurements in overnight/first-morning urine samples did not differ from measurements in 24-h urine collections, but was lower for both than for plasma. Using optimized reference intervals, diagnostic specificity was higher for overnight than daytime collections at similar sensitivities.
Measurements of urinary-free catecholamine metabolites in first-morning/overnight urine collections offer an alternative for diagnosis of PPGL to 24-h collections but remain less accurate than plasma measurements.
Supplemental table 1. Differences in sensitivity and specificity with 95% confidence intervals and P-values for comparisons of the diagnostic tests in table 3