Optimized procedures for diagnostic testing for pheochromocytoma and paraganglioma in patients on hemodialysis

Diagnosis of pheochromocytomas and paragangliomas in patients receiving hemodialysis To establish optimal conditions for blood sampling for mass spectrometric measurements of normetanephrine, metanephrine and 3-methoxytyramine in patients on hemodialysis and specic reference intervals for plasma metanephrines under the most optimal sampling conditions. metanephrines.


Abstract Background
Diagnosis of pheochromocytomas and paragangliomas in patients receiving hemodialysis is troublesome.

Aim
To establish optimal conditions for blood sampling for mass spectrometric measurements of normetanephrine, metanephrine and 3-methoxytyramine in patients on hemodialysis and speci c reference intervals for plasma metanephrines under the most optimal sampling conditions.

Methods
Blood was sampled before and near the end of dialysis, including different sampling sites in 170 patients on hemodialysis.

Results
Plasma normetanephrine concentrations were lower (P < 0.0001) and metanephrine concentrations higher (P < 0.0001) in shunt than in venous blood, with no differences for 3-methyxytyramine.
Normetanephrine, metanephrine and 3-methoxytyramine concentrations in shunt and venous blood were lower (P < 0.0001) near the end than before hemodialysis. Upper cut-offs for normetanephrine were 34% lower when the blood was drawn from the shunt and near the end of hemodialysis compared to blood drawn before hemodialysis.

Conclusion
This study establishes optimal sampling conditions using blood from the dialysis shunt near the end of hemodialysis with optimal reference intervals for plasma metanephrines for the diagnosis of pheochromocytomas/paragangliomas among patients on hemodialysis.

Clinical Summary
What is already known about this subject Diagnosis of pheochromocytomas and paragangliomas (PPGLs) in patients receiving hemodialysis is troublesome. We have recently published chronic kidney disease (CKD) speci c cut-offs for plasma metanephrines with the use of liquid chromatography-tandem mass spectrometry in patients with CKD Introduction Pheochromocytomas and paragangliomas (PPGLs) are neuroendocrine tumors derived from the chroma n cells of the adrenal medulla or extra-adrenal chroma n tissue (1). Although rare, these tumors constitute an important endocrine cause of hypertension (2). Current clinical practice guidelines stipulate that biochemical screening for PPGLs should include measurements of either plasma free or urinary fractionated metanephrines (3). Additional measurements of 3-methoxytyramine are useful for identifying occasional tumors that predominantly produce dopamine (4).
In patients with advanced renal insu ciency, the diagnostic work up of PPGLs is troublesome. Similar to patients with PPGLs, many patients with end-stage renal disease (ESRD) on hemodialysis suffer from hypertension with wide swings in blood pressure. Biochemical con rmation or exclusion of PPGLs in such patients is confounded by the effects of impaired renal function on the elimination of catecholamines and their metabolites in urine (5,6). However, as the circulatory clearance of plasma free metanephrines is relatively independent of renal function (7), measurements of these compounds in plasma might be preferred in patients with ESRD. Yet, increases of plasma free metanephrines have been reported in patients with CKD (8-10), possibly re ecting CKD-associated activation of the sympathetic nervous system (11,12).
We have recently published cut-offs for normetanephrine (NMN), metanephrine (MN) and 3methoxytyramine (3-MTY) plasma concentrations with the use of liquid chromatography-tandem mass spectrometry (LC-MS/MS) in patients with CKD stage III, IV from morning peripheral venous blood samples and in patients on HD from morning peripheral venous blood samples, before the beginning of HD (13). However, the in uence of hemodialysis (HD) on the assessment of plasma metanephrines has not been studied to date. Importantly, it is not yet clari ed if modi cations of blood sampling would improve the diagnostic performance of measurements of plasma metanephrines in patients receiving HD. As there is a known arterio-venous gradient for plasma metanephrines in humans, sampling from the easily accessible shunt rather than from a vein might provide bene ts as shunt blood closely matches arterial blood (14,15). In addition, the time point of sampling may have a critical in uence. During HD patients undergo a long period of rest that might be bene cial in terms of minimizing sympatho-adrenal activation (16). On the other hand, the continued ultra ltration has the potential to increase sympathetic activation during treatment (17,18).
The aim of the study was, therefore, to establish optimal procedures for blood sampling to determine plasma concentrations of free metanephrines and 3-methoxytyramine in patients receiving HD.

Subjects
This study involved prospective analysis of data from 107 patients receiving HD. Patients were enrolled under a multicenter prospective study at six clinical care centers (13). The study was approved by local Ethics-Committees (Ethics Committee of the Technical University Dresden, Germany and Ethics Committee of the University Warsaw, Poland) and all patients provided written informed consents. All methods were performed in accordance with the relevant guidelines and regulations. All patients had an arterio-venous vascular access and were receiving HD for at least one year. Subjects were excluded if they presented with unstable conditions (sepsis or decompensated heart failure) or medication interfering with primary outcome parameters (tricyclic antidepressants, L-DOPA or medication containing sympathomimetic decongestants) (19). Baseline characteristics of the participants are reported in Table   1.

Study design
The main goal of the study was to establish optimal conditions for blood sampling for measurements of plasma free metanephrines and 3-methoxytyramine in patients on HD. For this, two blood samples, one from the shunt and the other from a contralateral antebrachial vein, were collected from all patients after 30 minutes of supine rest before the start of HD and during the last hour of treatment. In a subgroup of 30 patients, ninety minutes after start of HD, the ultra ltration rate was set to zero for 5 minutes and preas well as post-lter blood samples were drawn from the extracorporeal circuit to determine compound extraction. Patients remained in the semi-recumbent position throughout the HD. Duration of HD was approximately four hours.

Laboratory analysis
Measurements of plasma free NMN, MN and MTY concentrations were performed using LC-MS/MS (20). All patients were instructed to fast and refrain from alcohol, nicotine, decaffeinated and caffeinated beverages for 12 hours before the rst sampling. A low amine restricted diet was allowed during dialysis procedures prior to the second dialysis blood sampling, in a semi-recumbent position, whereas catecholamine containing foods were avoided. Blood samples were kept on ice until plasma was separated and stored frozen at -80 o C before analyses.

Statistical analysis
Statistical analyses utilized the JMP statistics software package (SAS Institute Inc, Cary, NC), with comparisons by Wilcoxon´s paired and Mann-Whitney U unpaired tests. Reference intervals for metabolites were established from distributions of the measured variables in populations with CKD receiving HD using non-parametric or parametric approached as indicated by the nature of distributions.
Upper cut-offs of reference intervals were determined by 97.5% percentiles of the distributions of each metabolite. Using the medians of the dialysis ow and the hematocrit, the plasma ow was calculated {=Blood Flow x (1-Ht)}. From the dialysis plasma ow and the difference concentrations of metanephrines before vs after the blood leaves the lter, the dialysis clearance was then calculated {=Plasma ow x [Cmetabolite before lter -Cmetabolite after lter /Cmetabolite before lter ]}.

Patient Characteristics
Hypertension was recorded in 79.4% of patients receiving HD and up to 98.1% received hypertension treatment. Up to 63.5% were treated with ACE Inhibitor/or ARBs and up to 83.5% with ß-blockers. Diabetes was present in 38.3% of patients. The proportion of patients treated with intensi ed insulin treatment (ICT) was 26%. Impaired renal function was mainly due to diabetic and hypertensive kidney disease, followed by glomerulonephritis and polycystic kidney disease (Table 1).

Pre-vs End of Dialysis
Normetanephrine and metanephrine concentrations, both in venous and shunt blood (Fig. 1A, B), were lower near the end than before dialysis (P<0.0001). Similarly, concentrations of 3-methoxytyramine tended to be lower near the end than before dialysis, both in venous (P=0.050) and shunt blood (P=0.062), but failed to reach statistical signi cance (Fig. 1A, B).

Clearance of Metanephrines and 3-Methoxytyramine by the Dialysis Filter
Among patients receiving HD there were lower plasma concentrations of normetanephrine (P<0.0001), metanephrine (P<0.0001) and 3-methoxytyramine (P=0.0014) in blood leaving than entering the dialysis lter (Fig.2). The medians of dialysis clearance for NMN (79.7 mL/min) and MN (76.5 mL/min) were similar. The dialysis clearance of both metanephrines was, therefore, calculated to approximately 78 mL/min (Table 1), representing only 6.5% of the endogenous clearance of plasma metanephrines in patients with ESRD, which is assumed to reach 1200 mL/min (21,22).

Shunt vs Venous Blood Sampling
For blood samples drawn near the end of dialysis, plasma normetanephrine concentrations among patients on HD were signi cantly lower (P=0.005) and metanephrine concentrations higher (P<0.0001) in shunt than in venous blood, with no signi cant difference for 3-methoxytyramine (Fig. 3).

HD-Speci c Reference Intervals under Optimal Sampling Conditions
Due to the high number of outliers involving remarkably high plasma concentrations of 3-MTY in patients with ESRD, HD-speci c cut-off values are recommended only for plasma metanephrines. Compared to already established stage IV/HD speci c reference intervals from the vein before HD (13), upper cut-off levels (97.5% percentiles of reference intervals) for NMN in patients on HD were 34% lower when the blood was drawn from the shunt near the end of HD than from the vein before HD. In contrast, upper cutoff levels for MN were 12% higher when the blood was drawn from the shunt near the end of HD than from the vein before HD (Table 2).

Discussion
The present study establishes that measurements of plasma metanephrines in patients receiving HD are most appropriate using blood drawn from shunt at a time near the end of dialysis. Moreover, this study outlines speci c upper cut-offs of reference intervals for plasma metanephrines under these most optimal conditions in patients on HD that can be expected to minimize false positive results.
The diagnostic work-up of patients suspected of harboring a PPGL among patients on HD is a clinical challenge. Consequently, reliable biochemical tests to exclude or con rm these tumors are crucial. Urinary tests are clearly unreliable for patients on HD (5,6), while for plasma free metanephrines the reported prevalence of false positive results had varied from 85% using high pressure liquid chromatographyecectrochemical detection (HPLC-ED) (12) to approximately 10.2% for NMN and 14% for MN when using LC-MS/MS, con rming the superiority of this method, particularly in terms of relative freedom of analytical interferences (13). However, the remaining relatively high prevalence of false positive results for either elevation of NMN or MN likely re ects chronic activation of the sympathetic nervous system, which characterizes patients with ESRD (8,9). The kidneys are responsible for 14 to 16% of the circulatory clearance of plasma free metanephrines (23), so that some smaller increase in plasma concentrations can also be expected to result from impaired renal function.
The above considerations highlight the need for optimized pre-analytical procedures and speci c upper limits of reference intervals under most optimal sampling conditions. With recognition of the above needs, the main aim of our study was, to establish the most appropriate sampling conditions for measurements of plasma metanephrines and 3-MTY during HD. Due to the high number of outliers involving remarkably high levels of 3-MTY in patients on HD, optimized pre-analytical procedures and speci c upper cut-offs were focused on plasma metanephrines. The lower plasma concentrations of metanephrines, both in venous and shunt blood, near the end rather than before starting HD indicate the former time point as preferable to minimize false positive results. The prolonged recumbency (4 hours HD duration) could provide one explanation (16) while an effect of the dialysis lter to increase clearance of metanephrines likely also contributes to the lowered concentrations towards the end of dialysis. Nevertheless, fractional extractions of metanephrines by the dialysis lter were calculated to contribute to less than 7% of the endogenous clearance of metanephrines in patients on HD.
In addition to showing that the last hour of dialysis is the most appropriate time for blood sampling, the present study also established the shunt as the best sampling site. This conclusion was based on the ndings that plasma concentrations of NMN were lower and those of MN higher in shunt than in venous blood. The latter observation can be explained by extraction of this metabolite during passage from arterial to venous sites resulting in a physiologic arterio-venous concentration gradient (23)(24)(25). NMN is also extracted during passage from arterial to venous sampling sites, but, in contrast to MN, it is also generated by metabolism of locally produced norepinephrine (10). Thus, the dialysis shunt represents the most appropriate sampling site to both minimize false-positive results for normetanephrine and detect any signal for both metabolites from catecholamine-producing tumors.

Conclusion
The present study is important in being the rst to comprehensively address optimal sampling procedures for plasma metanephrines for the diagnosis of pheochromocytomas/paragangliomas among patients on HD and relevant optimal reference intervals that can be expected to minimize false positive results. Our study thereby provides immediate guidance to clinicians in daily routine practice.