Biochemical Diagnosis of Pheochromocytoma—Is it Time to Switch to Plasma-Free Metanephrines?

Graeme Eisenhofer

Clinical Neurocardiology Section, National Institute of Neurological Disorders and Stroke, National Institutes of Health Bethesda, Maryland 20892

Address all correspondence and requests for reprints to: Graeme Eisenhofer, Ph.D., Building 10, Room 6N252, National Institutes of Health, Bethesda, Maryland 20892-1620. E-mail: graeme{at}catecholamine.org.

Current recommendations for biochemical diagnosis of pheochromocytoma remain more often based on institutional experience with certain methods than on evidence of carefully designed studies carried out to compare available tests. Experts from some centers emphasize measurements of plasma catecholamines, whereas others advocate urinary measurements. Emerging evidence now indicates that measurements of catecholamines, when carried out alone, are not sufficiently sensitive for reliable diagnosis. In particular, plasma and urinary catecholamines can be normal when testing is carried out between spells or when normotensive and asymptomatic patients with pheochromocytoma are screened for the tumor because of a hereditary predisposition or the incidental finding of an adrenal mass by imaging studies.

Realization that pheochromocytomas may secrete catecholamines intermittently or in low amounts has led to recommendations that diagnosis should entail a combination of tests, and include measurements of metanephrines, the O-methylated metabolites of catecholamines (1, 2). Measurements of urinary metanephrines originally involved spectrophotometric methods where normetanephrine, the O-methylated metabolite of norepinephrine, and metanephrine, the O-methylated metabolite of epinephrine, are measured together in total as a single analyte—hence, the term "urinary total metanephrines." These methods have now been superseded by techniques involving liquid chromatography that allow separation of normetanephrine and metanephrine into fractionated components that can be measured individually—hence, the term "urinary fractionated metanephrines." Most of these latter methods also allow additional measurements of methoxytyramine, the O-methylated metabolite of dopamine.

Measurements of the fractionated metabolites are superior to measurements of total metanephrines in that they allow better detection of tumors that produce predominantly or only one of the three O-methylated metabolites (3, 4). Tests involving measurements of total metanephrines are, therefore, best ignored by clinicians and abandoned by clinical laboratories in favor of modern liquid chromatographic methods.

More recent developments in the biochemical diagnosis of pheochromocytoma include liquid chromatographic measurements of plasma-free metanephrines, the single largest source of which are adrenal chromaffin cells (5). In patients with pheochromocytoma the free metanephrines are produced within tumor cells from catecholamines leaking from storage vesicles continuously and independently of variations in exocytotic catecholamine release. The free metanephrines are either deaminated by monoamine oxidase or sulfate conjugated by a sulfotransferase isoenzyme located mainly in the digestive tract and important for metabolism of both dietary and locally produced monoamines. The sulfate-conjugated metanephrines so produced represent the main form eliminated in the urine. Urinary metanephrines are usually measured after a deconjugation step and, therefore, represent different metabolites from the free metanephrines measured in plasma.

In this issue of JCEM, Sawka et al. (6) report on the results of a study carried out at the Mayo Clinic in which the use of plasma-free metanephrines for diagnosis of pheochromocytoma was compared with that of 24-h urinary total metanephrines and catecholamines. This study follows on the heels of studies from the National Institutes of Health (NIH) (7, 8, 9), where the assay of plasma-free metanephrines was developed, and also from a group in Vienna, Austria, using a similar method (10). The importance of the present report is that it represents an independent study from a major reference laboratory, with extensive experience in urinary assays of total metanephrines and catecholamines, and where the assay of plasma-free metanephrines was established as a service offered on a commercial basis throughout the United States.

The study from the Mayo Clinic confirms the high diagnostic sensitivity of plasma-free metanephrines reported independently by the two other centers (7, 8, 9, 10). Only 1 of 33 patients with pheochromocytoma had normal levels of both normetanephrine and metanephrine, a patient with a dopamine-secreting paraganglioma. Such tumors are extremely rare, and presumably this tumor would have been picked up by the plasma test if measurements had included the O-methylated metabolite of dopamine, methoxytyramine.

Inspection of the data in Fig. 2 of the Mayo Clinic report shows that compared with the 1 patient with pheochromocytoma who had normal levels of plasma-free metanephrines, there were 8 of 32 patients with normal urinary outputs of total metanephrines, indicating a sensitivity of only 75% compared with 97% for the test of plasma-free metanephrines. This is close to the sensitivity of 77% for the test of urinary total metanephrines reported in the most recent NIH study (9), supporting the view that tests of urinary total metanephrines are best abandoned in favor of newer tests involving fractionated measurements (3, 4).

Further inspection of the data in Figs. 1 and 2 shows that there were 14 of 31 patients with pheochromocytoma who had normal levels of urinary norepinephrine, compared with 3 of 33 with normal plasma levels of normetanephrine. Moreover, the receiver-operating characteristic curves shown in Fig. 4 illustrate that a single measurement of plasma-free normetanephrine provides superior efficacy for diagnosis of pheochromocytoma than measurements of any of the other analytes. Again, this supports previous findings that measurements of the O-methylated catecholamine metabolites allow more reliable detection of pheochromocytoma than measurements of the parent catecholamines (7, 8, 9, 10).

Although the results of the Mayo Clinic study support the high diagnostic sensitivity of plasma-free metanephrines, enthusiasm for the test is tempered by findings of relatively high numbers of false positive test results. Of 261 patients in whom pheochromocytoma was excluded, 40 had false positive elevations of plasma normetanephrine or metanephrine, indicating a diagnostic specificity of 85%. This compares with only four false positive elevations for the combination of measurements of urinary total metanephrines and catecholamines, indicating a specificity of 98%.

The above differences in diagnostic specificity are at variance with findings of previous studies showing similar specificities of measurements of plasma-free metanephrines and urinary total metanephrines or catecholamines (7, 8, 9, 10). In the most recent NIH study (9), the specificity of measurements of plasma-free metanephrines was 89%, compared with 93% for urinary total metanephrines and 88% for urinary catecholamines. In that study, a large proportion, but not all, of the measurements of urinary total metanephrines and catecholamines were carried out by the Mayo Medical Laboratories under a contract with the NIH. The Mayo Laboratories are well experienced in these urinary assays, and as described in the present article from the Mayo Clinic, particular care was taken to identify assay interferences and exclude these as potential sources of false positive results. Thus, some of the differences in specificities between NIH and Mayo Clinic studies may reflect contribution to the NIH study of urinary test results from outside laboratories in which assay interferences were not as carefully identified as they may have been if all urinary measurements had been carried out by the Mayo Medical Laboratories.

There are, however, other differences in the design and conduct of NIH and Mayo Clinic studies that should be considered. A strength of both studies is that inclusion of patients was based on suspicion of pheochromocytoma due to signs and symptoms, the finding of an adrenal incidentaloma, or because of high risk for the tumor because of previous history or a familial syndrome. This aspect of study design is particularly important for obtaining estimates of diagnostic specificity that are clinically relevant to the patient populations usually tested for the tumor. An associated problem, however, is that the disease under study must be excluded by methods other than the diagnostic tests being compared.

In the Mayo Clinic study, exclusion of pheochromocytoma was based on an alternative diagnosis. The experience at the NIH and at other centers is that a satisfactory alternative diagnosis in patients with paroxysmal hypertension associated with spells of headache, palpitations, or other symptoms is rarely possible (11). Such patients outnumber those with pheochromocytoma, and there is usually no satisfactory explanation or treatment for their signs and symptoms. Moreover, an alternative diagnosis in patients screened for pheochromocytoma because of previous history or a familial syndrome does not satisfactorily exclude the tumor in these high-risk groups.

In the NIH study, exclusion of pheochromocytoma was based on negative results of radiological imaging studies and lack of evidence for the tumor on patient follow-up, on average 2.5 yr after initial testing. However, even these fairly rigorous criteria for exclusion of pheochromocytoma have been criticized (12), and, as stated in the original article (9), it remains possible that some patients with pheochromocytoma may have been missed and incorrectly assigned to the group in which pheochromocytoma was excluded and from which specificity was calculated. This is also a problem with the present study from the Mayo Clinic.

Another potential cause for the differences in diagnostic specificities between NIH and Mayo Clinic studies centers on differences in blood sampling procedures. Experience with plasma assays at the NIH has shown that it is best to draw blood samples after an overnight fast. This results in cleaner chromatograms and fewer problems with interfering substances that may not always be identified by inspection of chromatographic results. Also, although plasma-free metanephrines are less responsive than catecholamines to sympathoadrenal activation, they nevertheless show similar directional responses as the parent amines to different stressors (13). The free metanephrines are also cleared rapidly from the circulation, have short plasma half-lives, and, therefore, respond rapidly to changes in sympathetic outflow, such as occur with changes in posture. Recent studies in eight normotensive and hypertensive volunteers at the NIH have shown significant (P = 0.006) and consistent increases of plasma normetanephrine concentrations within 5 min after a change from the supine to the upright position (Eisenhofer et al., unpublished observations). The percentage increase in plasma normetanephrine was less than that in plasma norepinephrine (27% vs. 130%). Nevertheless, any increase in plasma normetanephrine levels is likely to increase the chance of a false positive result.

Similar to recommendations for plasma catecholamines, blood samples collected for plasma-free metanephrines should be collected following an overnight fast and after at least 15 min rest in the supine position. The Mayo Medical Laboratories may not be obtaining blood samples for measurements of plasma-free metanephrines with a mind to the above precautions. Increased background noise in measurements of plasma-free metanephrines when these precautions are not followed can be expected to result in increased numbers of false positive results.

Despite the above limitations, the results from the Mayo Clinic are encouraging. Concentrations of free metanephrines in plasma are more than 3 orders of magnitude lower than those of metanephrines commonly measured in urine. Assays of plasma-free metanephrines are, therefore, not as easy to establish and run on a routine basis as conventional assays of urinary metanephrines and catecholamines. Even at the NIH, transfer of the technology from the research to the routine laboratory environment was not without some difficulty, despite the presence of highly competent and skilled technical staff. The investigators at the Mayo Medical Laboratories are to be commended for successfully taking a technically difficult assay and showing that it can be used in a routine laboratory environment to test for pheochromocytoma. There remains a continuing need for improvements in assay technology that will increase the robustness of the method and advance its portability to other routine laboratories that lack the same level of technical expertise available at the NIH and Mayo Clinic.

As tests of plasma-free metanephrines become more widely available, it will become increasingly necessary to weigh the available evidence and determine whether it is time to switch to the newer method of diagnosis. In their article, Sawka et al. (6) conclude that measurements of plasma-free metanephrines may be the biochemical test of choice in patients at high risk for pheochromocytoma, such as those with a familial syndrome. However, because of less than ideal diagnostic specificity, they indicate hesitancy in extending routine use of the test to the more common setting of sporadic pheochromocytoma. They suggest that in this setting combined measurements of 24-h urinary total metanephrines and catecholamines may be preferable due to a lower likelihood of false positive results than with measurements of plasma-free metanephrines.

Before accepting the above conclusion, it should be considered that there are increasingly, and quite appropriately, fewer centers where spectrophotometric measurements of urinary total metanephrines are available. At the Mayo Medical Laboratories, for instance, recent implementation of a new method for measurements of urinary fractionated metanephrines (14) will likely lead to a phasing out of measurements of urinary total metanephrines. As shown in the NIH study (9), measurements of urinary fractionated metanephrines provide a more sensitive test for diagnosis of pheochromocytoma than measurements of urinary total metanephrines, but suffer from poor specificity. In particular, for patients tested for sporadic pheochromocytoma, rates of false positive results for urinary fractionated metanephrines considerably exceeded those for plasma-free metanephrines (55% vs. 18%). Combining measurements of urinary fractionated metanephrines and catecholamines in this patient population can only be expected to further increase rates of false positive results.

False positive results are likely to remain a problem with any reasonably sensitive biochemical test used for diagnosis of pheochromocytoma. In particular, a certain percentage of false positive results must be expected when reference intervals are established using the 95% confidence intervals of values from a reference population. Development of strategies for distinguishing true positive from false positive results is required to streamline the diagnostic decision-making process and reduce costs of follow-up. In the meantime, it should be considered that a missed diagnosis due to a false negative result in a patient with pheochromocytoma could have serious consequences. An appropriately sensitive biochemical test, therefore, represents the first choice for diagnosis of pheochromocytoma.

Acknowledgments

Thanks are extended to Karel Pacak, M.D., Ph.D., David Goldstein, M.D., Ph.D., Yehonatan Sharabi, M.D., and Patti Sullivan, B.S. for helpful discussions and contributions to presented data.

Received December 5, 2002.

Accepted December 12, 2002.

References

  1. Young Jr WF 1997 Pheochromocytoma: issues in diagnosis and treatment. Compr Ther 23:319–326[Medline]
  2. Bravo EL 2002 What is the best diagnostic approach when pheochromocytoma is suspected? Cleve Clin J Med 69:257–258[Medline]
  3. Rosano TG, Swift TA, Hayes LW 1991 Advances in catecholamine and metabolite measurements for diagnosis of pheochromocytoma. Clin Chem 37:1854–1867[Abstract]
  4. Gardet V, Gatta B, Simonnet G, Tabarin A, Chene G, Ducassou D, Corcuff JB 2001 Lessons from an unpleasant surprise: a biochemical strategy for the diagnosis of pheochromocytoma. J Hypertens 19:1029–1035[CrossRef][Medline]
  5. Eisenhofer G, Huynh TT, Hiroi M, Pacak K 2001 Understanding catecholamine metabolism as a guide to the biochemical diagnosis of pheochromocytoma. Rev Endocrinol Metab Disord 2:297–311
  6. Sawka AM, Jaeschke R, Singh RJ, Young Jr WF 2003 A comparison of biochemical tests for pheochromocytoma: measurement of fractionated plasma metanephrines compared with the combination of 24-hour urinary metanephrines and catecholamines. J Clin Endocrinol Metab 88:553–558[Abstract/Free Full Text]
  7. Lenders JW, Keiser HR, Goldstein DS, Willemsen JJ, Friberg P, Jacobs MC, Kloppenborg PW, Thien T, Eisenhofer G 1995 Plasma metanephrines in the diagnosis of pheocromocytoma. Ann Intern Med 123:101–109[Abstract/Free Full Text]
  8. Eisenhofer G, Lenders JW, Linehan WM, Walther MM, Goldstein DS, Keiser HR 1999 Plasma normetanephrine and metanephrine for detecting pheochromocytoma in von Hippel-Lindau disease and multiple endocrine neoplasia type 2. N Engl J Med 340:1872–1879[Abstract/Free Full Text]
  9. Lenders JW, Pacak K, Walther MM, Linehan WM, Mannelli M, Friberg P, Keiser HR, Goldstein DS, Eisenhofer G 2002 Biochemical diagnosis of pheochromocytoma: which test is best? JAMA 287:1427–1434[Abstract/Free Full Text]
  10. Raber W, Raffesberg W, Bischof M, Scheuba C, Niederle B, Gasic S, Waldhausl W, Roden M 2000 Diagnostic efficacy of unconjugated plasma metanephrines for the detection of pheochromocytoma. Arch Intern Med 160:2957–2963[Abstract/Free Full Text]
  11. Mann SJ 1999 Severe paroxysmal hypertension (pseudopheochromocytoma): understanding the cause and treatment. Arch Intern Med 159:670–674[Abstract/Free Full Text]
  12. Neumann HP 2002 Imaging vs. biochemical testing for pheochromocytoma. JAMA 288:314–315[Free Full Text]
  13. Eisenhofer G, Friberg P, Pacak K, Goldstein DS, Murphy DL, Tsigos C, Quyyumi AA, Brunner HG, Lenders JW 1995 Plasma metadrenalines: do they provide useful information about sympatho-adrenal function and catecholamine metabolism? Clin Sci 88:533–542[Medline]
  14. Taylor RL, Singh RJ 2002 Validation of liquid chromatography-tandem mass spectrometry method for analysis of urinary conjugated metanephrine and normetanephrine for screening of pheochromocytoma. Clin Chem 48:533–539[Abstract/Free Full Text]