Precisely Wrong? Urinary Fractionated Metanephrines and Peer-Based Laboratory Proficiency Testing

To the Editor:

Measurements of urinary fractionated metanephrines (normetanephrine and metanephrine), the O-methylated metabolites of norepinephrine and epinephrine, provide a sensitive test for diagnosis of pheochromocytoma. We present evidence suggesting inaccurate metanephrine and normetanephrine calibration by US laboratories.

Metanephrines are present in urine mainly as sulfate- and glucuronide-conjugated metabolites produced from free metanephrines by the actions of conjugating enzymes (1). An acid hydrolysis step is usually performed to liberate the free metanephrines from the conjugated metabolites. This step minimizes requirements for high analytical sensitivity, simplifying subsequent measurement.

Urinary fractionated metanephrines are usually measured by HPLC with electrochemical detection (HPLC-EC). Gas chromatography with mass spectrometry (GC-MS) and liquid chromatography with tandem mass spectrometry (LC-MS/MS) are more recent alternatives offering high sample throughput and improved analytical specificity (2)(3).

Clinical laboratories in the US participate in proficiency testing programs, primarily aimed at determining agreement in diagnostic test results among laboratories. Identical survey samples are distributed to participating laboratories, with results stratified according to testing methodologies. Laboratories that report results deviating by more than a certain amount from the mean of their peer groups fail the proficiency test.

Until the spring of 2004, the only source of commercially available urine metanephrine calibrators in the US was Bio-Rad. As with the survey samples used for proficiency testing, these calibrators are supplied as lyophilized urine, which must be reconstituted and taken through the same sample-processing steps as patient samples. One laboratory not using these commercial calibrators (Mayo Clinic Endocrine Laboratory) reported significantly higher results for urinary fractionated metanephrines than all other participating laboratories during three successive proficiency surveys in 2003/2004. Mayo laboratory calibrators are prepared in house from normetanephrine and metanephrine supplied by Sigma-Aldrich Chemical Co. The Mayo laboratory suspected that differences between their in-house calibrator and the commercial calibrator used by other laboratories might have been responsible for their successive failures to pass proficiency testing. Comparisons of the two calibrators by LC-MS/MS indicated 33% higher values for normetanephrine and 40% higher values for metanephrine in the commercial calibrator than in the in-house calibrator (Table 1¹ ). Values in the commercial calibrator were only slightly higher with than without the acid hydrolysis step, indicating that most of the metanephrines were in the free rather than the conjugated form.

To resolve the discrepancy, two independent laboratories (one at the NIH and the other at ARUP) were asked by those overseeing the proficiency program to investigate the problem. The calibrators used by the NIH laboratory (also prepared in house from normetanephrine and metanephrine supplied by Sigma-Aldrich) showed close agreement by HPLC-EC with those supplied by the Mayo laboratory. Values for the commercial calibrator were 36% higher than assigned values for normetanephrine and 50% higher than assigned values for metanephrine (Table 1¹ ). Again, values were only slightly higher with an acid hydrolysis step, confirming that most of the metanephrines in the commercial calibrator were in the free form. The ARUP laboratory routinely used the Bio-Rad calibrator for GC-MS measurements of urinary fractionated metanephrines. When compared with freshly prepared calibration solutions from Sigma-Aldrich, the concentrations of metanephrines in the Bio-Rad calibrator were also calculated to be significantly higher than assigned values (Table 1¹ ). Possible degradation of free metanephrines during the acid hydrolysis step was determined to be insignificant (<3%).

The three laboratories used different methods of analysis (HPLC-EC, LS-MS/MS, and GC-MS), indicating that the observed differences were unrelated to methodology or instrumentation. However, all three laboratories used the same source for their in-house-prepared calibrators. Thus, an impure starting material from Sigma-Aldrich could explain the discrepancy in assigned and measured values for the calibrators from Bio-Rad. To examine this possibility, samples of the Sigma-supplied normetanephrine and metanephrine were tested by elemental analysis. The results (Table in the Data Supplement that accompanies the online version of this letter at http://www.clinchem.org/content/vol51/issue2/) confirmed that the purity of the Sigma-supplied material was >98%, as stipulated by the manufacturer. This analysis and the highly consistent findings from the three independent laboratories indicate that the values assigned to the Bio-Rad calibrator are indeed 24–33% lower than their real values.

If most of the US laboratories participating in the proficiency testing program have been using the Bio-Rad calibrator, then this would explain the differences in results of proficiency testing between the Mayo laboratory and the other laboratories participating in the program. It is important to note that this program is peer-based, not accuracy-based. Although peer-based programs provide an adequate system for assessing agreement among members of a peer group, they do not guarantee accuracy of results.

Finally, Bio-Rad has informed one of us (A.L.R.) that their calibrator is for the Bio-Rad HPLC assay and cannot be assumed to work with other methods. If this indeed is the official position of Bio-Rad, then those laboratories using their calibrator, but not their assay reagents, should consider changing to an accurate calibrator prepared in house or provided by an alternative commercial source.