When undetected, pheochromocytoma can maim or kill. Therefore, infrequent as the disease may be, accurate, safe, and feasible screening to find these tumors remains an important goal for the primary care physician, the clinical researcher, and the subspecialist alike. This crucial need is addressed by Lenders and colleagues in this issue [1]. Excess catecholamines released into the circulation by a pheochromocytoma cause the clinical manifestations of the disease. Increased plasma concentrations of norepinephrines or epinephrines or both are found in most but not all patients with these neoplasms. However, high plasma concentrations of catecholamines may be found in other disorders, and the limited sensitivity and specificity of measurement of plasma catecholamines for diagnosis of pheochromocytoma has been well established [2, 3].

Enzymatic metabolism of catecholamines proceeds both by the o-methylation pathway (yielding normetanephrine and metanephrine) and by less specific monoamine oxidation; the product is vanillylmandelic acid. Urinary excretion of vanillylmandelic acid is often elevated in patients with pheochromocytoma, but the sensitivity and specificity of the assay for measuring this also fall short of perfection. Although measurement only of the "neuron-specific" product of monoamine oxidation, dihydroxyphenylglycol, was shown in one study to be highly accurate for pheochromocytoma [4], this assay has not been widely used and is not well characterized in other forms of secondary hypertension or cardiac failure. The o-methylated metanephrines are primarily conjugated by sulfation and circulate in plasma before being excreted in urine, where they are easily measured.

Measurement of metanephrine excretion for either 24 hours or as a timed excretion rate is perhaps the most widely used chemical screening test for pheochromocytoma; its sensitivity ranges from 80% to 85%, and its specificity is in excess of 90% [5]. Lenders and colleagues [1] studied 52 patients with pheochromocytoma and compared the results with those from 67 normal persons, 51 patients with essential hypertension, 23 patients with other forms of secondary hypertension, and 50 patients with congestive heart failure or angina pectoris. Using a highly accurate assay developed by their colleagues, Lenders and coworkers determined plasma concentrations of conjugated and nonconjugated (free) normetanephrine and metanephrine during standardized, unstressed conditions. They compared their results with available measurements of plasma catecholamine concentrations or urine metanephrine excretion rates. The comparisons show the superior accuracy of the assays for combined plasma free metanephrine. In all patients with pheochromocytomas, either plasma normetanephrine or metanephrine concentrations were elevated. Yet, the specificity of metanephrine assays was also high (range, 97% to 100%) and was superior to the specificity of measurements of plasma catecholamines. When compared by receiver-operating characteristic curves, the high predictive accuracy for the assay of plasma metanephrines seems readily apparent.

Close inspection of Lenders and colleagues' results, however, somewhat tempers my optimism. A few false-positive results did occur, most often in patients with other forms of secondary hypertension or cardiac disease rather than uncomplicated essential hypertension. These findings suggest that excluding pheochromocytoma may be more difficult when either cardiac disease or other forms of secondary hypertension are present. Another warning comes in a report from France [6], which suggested that tests for total serum metanephrine concentration may be accurate in patients with normal renal function but that false-positive results may occur in those with chronic renal disease. The unconjugated (free) compounds were not measured in this study. However, it is the complex patient whose clinical picture is uncertain because of cardiac or renal disease for whom an accurate diagnosis is most important. Elevated plasma concentrations of catecholamines are clearly related to advanced congestive heart failure [7]. However, some patients with pheochromocytoma present with a severe cardiomyopathy [8]. Biochemical methods alone might not define the diagnosis accurately enough, and imaging studies in addition to localization will be necessary to establish the diagnosis. Admittedly, these complex cases are unusual problems. What should be done in the more ordinary situations in which pheochromocytoma is likely to be considered?

In Lenders and colleagues' study, a substantial fraction of patients with proven pheochromocytoma had plasma metanephrine concentrations within the normal range. Although only one proven tumor resulted in a normal normetanephrine level, several tumors resulted in minimal elevations. A few false-negative measurements will probably be found in these patients eventually. A suspect group for investigation in this regard consists of asymptomatic members of kindreds with type 2 multiple endocrine neoplasia. Accurate early detection of pheochromocytoma in these persons remains less than ideal.

In recent years, physicians have shifted away from "complete workups" for secondary hypertension for most of their patients with high blood pressure and have used clinical suspicion to make their choices in accordance with guidelines provided by consensus committees [9]. Their attention has perhaps shifted to patients with symptoms of episodic hypertension or paroxysmal events suggestive of the syndrome. Patients with obvious anxiety or panic-like reactions have been identified as a special target group [10, 11]. If this is indeed a widespread practice, it seems reasonable to know the characteristics of the new test, that is, the plasma normetanephrine and metanephrine concentrations in patients with anxiety states or the panic disorder syndrome. Will the test prove accurate, or will false-positive results occur? How often will such patients need additional studies, such as clonidine suppression, glucagon stimulation [12], or imaging? What will prove to be cost-effective in this regard?

As a final consideration, measurement of plasma metanephrines by standardized conditions in a research setting may well differ from the results obtained when the test is transferred to more usual clinical settings and local or regional reference laboratories. Many tests, developed under rigid quality control in research laboratories, have successfully been disseminated for widespread application. Yet, each new and possibly better test must be viewed separately before one concludes that it performs as expected.

Detecting pheochromocytoma is not like searching for a needle in a haystack. In 1995, the latter is a simple task; just use a strong magnet. In contrast, the quest for pheochromocytoma is like the search for the needle in a sewing box: There are so many look-alikes that any one test must have perfect accuracy. Without this perfectly accurate single test, appropriate combinations of biochemical tests can substantially shift the odds for detection [13]. Additional necessary studies will determine whether measurement of the plasma metanephrines adds diagnostic information to currently available strategies for detection of pheochromocytoma. Nonetheless, the goal of such research is certainly worth the effort and cost because there is a compelling reason to find each pheochromocytoma before it causes a devastating effect.