Although measurement of plasma fractionated metanephrines represents the gold standard for diagnosis of pheochromocytoma, the utility of this test is reduced by a high false-positive rate.OBJECTIVE
To evaluate whether follow-up measurement of chromogranin A or urine fractionated metanephrines improves diagnostic accuracy.DESIGN AND INTERVENTION
This was a US-based, single-center, retrospective study of patients who had undergone measurement of plasma fractionated metanephrines to confirm or exclude chromaffinoma. Inclusion criteria included a positive test result, absence of comorbidities known to affect plasma fractionated metanephrine or chromogranin A levels, and availability of a residual blood sample for analysis. Clinical, laboratory and imaging data were reviewed and patients classified as pheochromocytoma-positive or pheochromocytoma-negative. Confirmed cases had definitive histology or a diagnostic imaging result. Plasma and urine fractionated metanephrines were assessed by liquid chromatography-tandem mass spectrometry. Diagnostic cut-off values for these tests were ≥0.5 nmol/l (plasma metanephrine), ≥0.9 nmol/l (plasma normetanephrine), 300 μg/24 h (urine metanephrine), 733 μg/24 h (urine normetanephrine) and 1,000 μg/24 h (total urine metanephrines). Serum chromogranin A was assessed by chemiluminescence and a level >225 ng/ml was considered positive.OUTCOME MEASURES
The main outcome measures were the sensitivity, specificity, positive predictive value (PPV), likelihood ratio (LR), and receiver-operator characteristic area under the curve (ROC AUC) of each diagnostic test.RESULTS
A total of 24,204 plasma fractionated metanephrine tests was performed during the 15-month study period. Of these, 140 patients met the inclusion criteria (median age 62.5 years; 64.2% female). The mean plasma metanephrine level was 0.48 nmol/l, the mean plasma nor metanephrine level was 4.12 nmol/l, and 40 individuals were pheochromocytoma-positive. The PPV of the plasma metanephrine test was 28.6%. Increasing the cut-off values to 1.20 nmol/l for metanephrine and 2.19 nmol/l for normetanephrine decreased the false-positive rate, with a sensitivity of 80% and a specificity of 98%. Chromogranin A levels were elevated in patients receiving proton-pump inhibitors (635 ng/ml versus 174 ng/ml for patients not receiving these drugs; P <0.001). Measurement of chromogranin A had a sensitivity of 87% and a specificity of 73% when these patients were excluded. Increasing the cut-off value to >270 ng/ml increased the specificity to 89%. Combined plasma metanephrine and chromogranin A testing had an ROC AUC of 0.95 and an LR of 7.9. Urine fractionated metanephrines were assessed in 59 patients, 34 of whom were pheochromocytoma-positive. The sensitivity and specificity for this test were 26% and 100% (metanephrine), 88% and 80% (normetanephrine) and 82% and 88% (total metanephrines). Combined plasma and urine fractionated metanephrine testing had a sensitivity of 91%, a specificity of 80%, a ROC AUC of 0.94, and an LR of 4.6.CONCLUSION
Measurement of chromogranin A or urine fractionated metanephrines improved diagnostic accuracy in patients with modest elevations of plasma fractionated metanephrines.