Abstract
Background: Neurogenic orthostatic hypotension (NOH) usually results from deficient release of the sympathetic neurotransmitter norepinephrine (NE) when the patient stands up. In pure autonomic failure (PAF) and PD with NOH, sympathetic denervation can explain this deficiency, whereas in multiple-system atrophy (MSA), deficient baroreflex regulation of sympathetic traffic to intact terminals probably causes the NOH. From the concept of a unitary sympathoadrenomedullary system, one might predict parallel sympathoneural and adrenomedullary abnormalities in NOH.
Objective: To test the concept of parallel sympathoneural and adrenomedullary abnormalities in NOH by simultaneous measurements of plasma levels of catechols and metanephrines.
Methods: Antecubital venous blood drawn via an indwelling cannula with the subject supine was assayed for catecholamines (NE, epinephrine [EPI]), dihydroxyphenylglycol (DHPG), and metanephrines (normetanephrine [NMN] and metanephrine [MN]) in patients with PAF, PD + NOH, or MSA + NOH. Control subjects had PD lacking NOH or were normal volunteers at least 35 years old. Cardiac sympathetic innervation was assessed by 6-[18F]fluorodopamine PET scanning.
Results: The three NOH groups differed clearly in patterns of plasma catechols and metanephrines. The PAF group had low NE, DHPG, NMN, EPI, and MN levels, the group with MSA + NOH had generally normal levels, and the PD + NOH group low NMN levels and low DHPG levels for given NE levels but normal mean NE, EPI, and MN levels. All patients with PAF or PD + NOH had markedly decreased 6-[18F]fluorodopamine-derived radioactivity throughout the left ventricular myocardium; all patients with MSA + NOH had normal radioactivity.
Conclusions: PAF involves generalized loss of sympathoadrenomedullary cells, MSA + NOH intact sympathoadrenomedullary cells, and PD + NOH intact adrenomedullary cells but organ-selective sympathetic denervation, especially in the heart.
Neurogenic orthostatic hypotension (NOH), which occurs commonly in primary chronic autonomic failure, usually results from deficient release of the sympathetic neurotransmitter norepinephrine (NE) when the patient stands up.1 By consensus, three forms of primary chronic autonomic failure have been recognized: pure autonomic failure (PAF), multiple-system atrophy (MSA) with NOH (formerly called the Shy–Drager syndrome), and NOH in the setting of PD.2 PAF and PD + NOH both involve a postganglionic lesion, as all patients with these diseases have a loss of cardiac sympathetic nerves.3 In marked contrast, patients with MSA + NOH have intact cardiac sympathetic innervation.
This classification schema treats all autonomic failure as the same. As the autonomic nervous system contains at least five components—sympathetic noradrenergic, parasympathetic cholinergic, sympathetic cholinergic, enteric, and adrenomedullary4—treating all forms of autonomic failure as the same might be questioned. Adrenomedullary chromaffin cells do share several features with sympathetic postganglionic neurons. Both cell types express catecholamine synthetic enzymes and cell membrane and vesicular transporters for catecholamines, and both release catecholamines in response to nicotinic cholinergic stimulation. These similarities therefore support the widely accepted concept of a unitary “sympathoadrenal” system that would help maintain homeostasis during emergencies, as proposed by Cannon in the early twentieth century.5 Extending this concept to the pathophysiology of autonomic failure, one would predict that each NOH syndrome would involve parallel abnormalities of sympathoneural and adrenomedullary function. In the current study, we tested this hypothesis by measuring plasma levels of catechols and metanephrines and examining their interrelationships.
Interpreting plasma catechols and metanephrines requires understanding their distinctive sources and their separate meanings in terms of sympathoneural and adrenomedullary function. Figure 1 provides a current overview of major pathways in catecholamine metabolism.6 The main free (unconjugated) endogenous catechols in human plasma are the catecholamines NE and epinephrine (EPI); the catecholamine precursor, 3,4-dihydroxy-l-phenylalanine (dopa); 3,4-dihydroxyphenylglycol (DHPG), which is the main deaminated metabolite of NE; and 3,4-dihydroxyphenylacetic acid, the main deaminated metabolite of dopamine. The main free (unconjugated) metanephrines in human plasma are normetanephrine (NMN), the O-methylated metabolite of NE, and metanephrine (MN), the O-methylated metabolite of EPI.
Figure 1. Overview of the metabolic fate of catecho-lamines of the sympathetic nervous system and adrenomedullary hormonal system. COMT = catechol-O-methyltransferase; DHPG = dihydroxyphenylglycol; EPI = epinephrine; MAO = monoamine oxidase; MHPG = methoxyhydroxyphenylglycol; MN = metanephrine; MSA = multiple-system atrophy; NE = norepinephrine; NMN = normetanephrine; SNS = sympathetic nervous system; VMA = vanillylmandelic acid.
As indicated in figure 1, the main determinants of plasma NE are exocytotic release of the neurotransmitter from sympathetic nerves and reuptake of NE via the cell membrane NE transporter (Uptake 1). In contrast, the main determinant of plasma EPI is release of the hormone into the bloodstream, with inactivation mainly by extraneuronal uptake and intracellular metabolism.7 Plasma DHPG derives almost entirely from the actions of monoamine oxidase on NE in the sympathetic axoplasm. Under resting conditions, most of the axoplasmic NE—and therefore most of plasma DHPG—comes from net leakage of NE from storage vesicles, not from reuptake of released NE.8 During sympathetic stimulation, reuptake of released NE via Uptake 1 provides an additional source of plasma DHPG. In contrast, plasma NMN derives partly from NE of sympathoneural origin that undergoes metabolism by catechol-O-methyltransferase (COMT) before it can enter the bloodstream and partly from NE of adrenomedullary origin that undergoes metabolism also by COMT in the cytoplasm of adrenomedullary chromaffin cells, whereas relatively little of plasma NMN is derived from circulating NE.9 Under resting conditions, virtually all of plasma MN is derived from O-methylation of EPI leaking from storage vesicles into the cytoplasm of adrenomedullary cells. During stimulation of adrenomedullary secretion, extraneuronal uptake of EPI provides an additional source of plasma MN. Thus, under resting conditions, most of the metabolism of both NE and EPI takes place in the cells that produce them.
The separate and distinct sources and meanings of plasma levels of catechols and metanephrines afforded an opportunity to test the notion of a single sympathoadrenal pathologic state in NOH syndromes. Prior work has shown that PD + NOH features lower plasma NE levels during supine rest than does PD lacking NOH,10-13⇓⇓⇓ but relationships with plasma EPI levels or metanephrines have not been described. If PAF and PD + NOH involved generalized loss of catecholamine-producing cells, then the patients would have low plasma levels of NE, DHPG, NMN, EPI, and MN because of the lack of NE production in sympathetic nerves and lack of EPI production in adrenomedullary cells. If MSA + NOH involved intact sympathoadrenal function, then the patients would have normal levels of these compounds.
Materials and methods.
The Clinical Research Subpanel of the National Institute of Neurological Disorders and Stroke approved the study protocol. Each subject gave informed, written consent.
Subjects.
The subjects included 10 patients with PAF (7 men, 3 women; mean age 56 ± 5 years), 11 with PD + NOH (6 men, 5 women; mean age 65 ± 3 years), 5 with PD lacking NOH (4 men, 1 woman; mean age 60 ± 3 years), 18 with MSA + NOH (12 men, 6 women; mean age 60 ± 2 years), and 100 normal volunteers at least 35 years old (55 men, 45 women; mean age 47 years).
Orthostatic hypotension was defined by a decrease in systolic blood pressure of at least 20 mm Hg and decrease in diastolic pressure of at least 5 mm Hg between the supine and upright positions. NOH was defined by orthostatic hypotension combined with abnormal responses of beat-to-beat blood pressure during both Phases II-L and IV of the Valsalva maneuver.14
PAF was diagnosed from NOH and absence of clinical signs or symptoms of central neurodegeneration. PD was diagnosed from the classic triad of pill-roll tremor, cogwheel rigidity, and bradykinesia, with obvious improvement by levodopa/carbidopa (n = 5), or from NOH with cogwheel rigidity, expressionless face, hypokinesia, and cardiac sympathetic denervation indicated by 6-[18F]fluorodopamine PET scanning,3 without other signs of central neurodegeneration (n = 6). MSA was diagnosed by NOH and clinical signs and symptoms of progressive central neurodegeneration (parkinsonism, cerebellar, or mixed forms), without obvious improvement by levodopa/carbidopa.
Because of possible influences of high circulating levodopa levels on plasma levels of other catechols, all the patients in this study were off levodopa treatment and had normal plasma levels of levodopa (<15 nmol/L) at the time of study. Therapeutic levodopa levels average about 1,000 times this amount.
Caffeine-containing beverages, cigarettes, and alcohol were not allowed for at least 24 hours and acetaminophen for at least 5 days before the testing. Patients were studied while taking their usual other medications, which did not include drugs known to inhibit neuronal uptake of catecholamines.
Antecubital venous blood was drawn through an indwelling catheter, after at least 15 minutes of supine rest. Plasma levels of catechols and metanephrines were assayed by high-pressure liquid chromatography with electrochemical detection, according to methods validated in our laboratory.15,16⇓
To assess cardiac sympathetic innervation, patients underwent thoracic PET scanning after IV injection of 6-[18F]fluorodopamine.3 The scanning results have been published previously but not the associations with plasma metanephrines.3
Mean values for plasma levels of catecholamines and metanephrines in patient and control groups were compared using factorial analyses of variance or two-tailed, independent-means t-tests. Frequency data were examined by calculation of χ2 and relationships across individual patients by linear regression. Mean values were expressed ±1 SEM. A p value of <0.05 defined significance.
Results.
Plasma NE levels varied as a function of diagnosis (F = 9.2, n = 142, p < 0.0001). The PAF group had lower NE levels than did each of the other groups (figure 2), and the group with PD lacking NOH had higher NE levels than did each of the other groups. The PD + NOH and MSA + NOH groups had normal NE levels.
Figure 2. Plasma concentrations of catechols and metanephrines in neurogenic orthostatic hypotension. Yellow = pure autonomic failure; red = PD with neurogenic orthostatic hypotension; black = PD without neurogenic orthostatic hypotension; white = normal; blue = multiple-system atrophy with neurogenic orthostatic hypotension; AHS = adrenomedullary hormonal system; DHPG = dihydroxyphenylglycol; EPI = epinephrine; MN = metanephrine; MSA = multiple-system atrophy; NE = norepinephrine; NMN = normetanephrine; NOH = neurogenic orthostatic hypotension; PAF = pure autonomic failure; SNS = sympathetic nervous system.
Plasma DHPG levels also varied as a function of diagnosis (F = 5.5, n = 117, p = 0.0004). As for NE levels, the PAF group had lower DHPG levels than did each of the other groups.
Plasma NMN levels varied as a function of diagnosis (F = 13.6, n = 140, p < 0.0001), with subnormal mean values in all three groups with NOH.
Plasma EPI levels varied with diagnosis (F = 4.5, p = 0.002), with subnormal levels in PAF and increased levels in MSA + NOH. Plasma MN levels, however, did not vary significantly with diagnosis, although the PAF group had lower levels than did the normal group.
Among normal control subjects, plasma DHPG correlated positively with plasma NE levels (r = 0.62; figure 3, top), the y-intercept value clearly above the origin. In the PAF group, the line of best fit was at about the origin. Patients with PD + NOH had lower DHPG levels than expected for NE levels (data points below the normal line of best fit; χ2 = 7.4, p = 0.007), whereas those with MSA + NOH had normal plasma DHPG for NE levels.
Figure 3. Scatterplots relating plasma levels of dihydroxyphenylglycol (DHPG) (top) and normetanephrine (NMN) (bottom) to norepinephrine (NE) in individual patients. Solid lines are the lines of best fit for the normal control group. Dashed lines show lines of best fit drawn by eye (see Discussion). Same color key as for figure 2.
Plasma NMN also normally correlated positively with plasma NE levels but with more scatter (r = 0.41; see figure 3, bottom). In the PAF group as well as the other NOH patients, the line of best fit was at about the origin. Patients with PD + NOH all had lower NMN levels than expected for NE levels (χ2 = 9.5, p = 0.002), as did 17 of 18 patients with MSA + NOH (χ2 = 14.7, p = 0.0001).
Plasma MN and NMN correlated positively among normal control subjects (r = 0.35; figure 4, top). Patients with PAF had relatively low levels of both metabolites, whereas patients with PD + NOH or MSA + NOH had normal values (with substantial interindividual variability). Plasma MN was unrelated to plasma EPI, although, as noted above, PAF patients had low levels of both compounds (see figure 4, bottom).
Figure 4. Scatterplots relating plasma levels of metanephrine (MN) to normetanephrine (top) and epinephrine (EPI) (bottom) in individual patients. Solid line is the line of best fit for the normal control group. Same color key as for figure 2.
All patients with PAF or PD + NOH had markedly decreased septal myocardial concentrations of 6-[18F]fluorodopamine-derived radioactivity (means 3,825 ± 610 [n = 9] and 2,773 ± 279 [n = 10] nCi-kg/mL-mCi, normal 8,780 ± 417 nCi-kg/mL-mCi; p < 0.0001); all patients with MSA + NOH had normal radioactivity (mean 10,058 ± 345 [n = 18] nCi-kg/mL-mCi). Patients with PD lacking NOH had an intermediate mean value (4,980 ± 465 nCi-kg/mL-mCi; n = 25).
Discussion.
It is generally accepted that among patients with primary chronic autonomic failure, those with PAF have a loss of sympathetic noradrenergic innervation, whereas those with MSA have intact sympathetic noradrenergic innervation. Thus, during supine rest, PAF patients have low plasma levels of NE,1 DHPG,17 and methoxyhydroxyphenylglycol,18 whereas MSA patients have normal levels. The current results confirmed and extended these findings, as PAF patients had clearly lower plasma levels not only of NE and DHPG but also of NMN, MN, and EPI than did patients with MSA + NOH. Considering the sources of these compounds in the bloodstream, as shown in figure 1, the findings point to decreased release and turnover of both NE and EPI in PAF.
The findings in PAF and MSA + NOH therefore agreed with the concept of a unitary sympathoadrenal system, with either loss (PAF) or dysregulation (MSA + NOH) of catecholamine-producing cells in the sympathetic nervous system and adrenal medulla. Generalized sympathetic denervation in PAF can also explain why in this group, but not in MSA + NOH, the line of best fit for the relationship between plasma DHPG and NE levels passed through the origin.
The findings in the PD + NOH group, however, were inconsistent with the notion of parallel abnormalities in the sympathoneural and adrenomedullary systems, as explained below.
NOH occurs fairly commonly in PD.19 In PD, NOH has been ascribed to sympathetic denervation, because all patients with PD + NOH have sympathetic neurocirculatory failure, cardiac sympathetic denervation, and subnormal orthostatic increments in plasma NE levels; and as a group, they have significantly lower plasma NE levels than do PD patients lacking NOH.13 In the current study, as plasma levels not only of NE but also of NMN were lower in PD + NOH than in PD without NOH and patients with PD + NOH had lower levels of both DHPG and NMN than expected for their NE levels, the results supported the notion of partial loss of sympathetic noradrenergic innervation in PD + NOH.
Nevertheless, some of these findings were inconsistent with generalized loss of peripheral catecholamine-producing cells in PD + NOH. Two of the most glaring were that plasma levels of NE, the sympathetic neurotransmitter, and of DHPG, the main neuronal metabolite of NE, were normal and also clearly exceeded those in PAF. As levels of DHPG were decreased for given NE levels in this group, plasma NE levels were probably maintained by concurrently attenuated release and reuptake of NE, as one would expect from partial or heterogeneous denervation.
If patients with PD + NOH had normal entry of NE into the bloodstream, why would they have low plasma NMN levels for their NE levels? Clinicians often treat NOH with the synthetic mineralocorticoid fludrocortisone, and steroids can inhibit Uptake 2, which would decrease delivery of NE from the bloodstream to COMT in extraneuronal cells20; and COMT inhibitors now constitute a common adjunctive treatment of PD.21 In the current series, however, all three patients with PD + NOH who were not taking fludrocortisone or a COMT inhibitor still had low plasma NMN levels (0.15 nmol/L compared with a normal mean value of 0.30 nmol/L). The y-intercept value for the line of best fit relating plasma NMN to plasma NE was clearly above the origin in normal control subjects but close to the origin in patients with PD + NOH. This finding indicates that, whatever the source was of the y-intercept above the origin, this source was missing in the patients with PD + NOH. There are two such sources possible. One would be NE metabolized in adrenomedullary cells before secretion into the venous drainage. The other would be NE released from sympathetic nerves and taken up by extraneuronal cells before the NE entered the bloodstream. The first explanation does not suffice because it cannot account for decreased NMN levels yet normal MN levels in PD + NOH. We therefore favor the second explanation, recognizing, however, that for this to suffice, the pathogenic mechanism of sympathetic denervation would have to spare the adrenal medulla. Sympathetic denervation does appear to decrease Uptake 2 or COMT activity,22,23⇓ but whether decreased nerve traffic to intact terminals does so remains unknown.
In contrast with neurochemical evidence for partial or heterogeneous loss of sympathetic noradrenergic nerves in PD + NOH, there was no evidence for loss of adrenomedullary cells, as plasma levels of both EPI and MN were normal. This pattern did not fit with the concept either of generalized loss of catecholamine-producing cells or of a unitary sympathoadrenal lesion in PD + NOH.
In PD + NOH, plasma NE and EPI levels were approximately normal, whereas in PAF, those levels were clearly decreased, despite similarly extensive loss of cardiac sympathetic innervation in both diseases, as indicated by universally low myocardial concentrations of 6-[18F]fluorodopamine-derived radioactivity in both groups. The sympathetic noradrenergic lesion in PD + NOH therefore appears to be heterogeneous among innervated organs and is especially prominent in the heart, an inference in agreement with results of several sympathetic neuroimaging studies.24-26⇓⇓ Future research should address the bases of organ-selective loss of sympathetic noradrenergic nerves, with relative sparing of adrenomedullary cells, in PD.
The current report involved only subjects studied during supine rest. In primary chronic autonomic failure syndromes, abnormalities of sympathetic neuronal or adrenomedullary hormonal function are generally more apparent during exposure to stressors such as orthostasis or insulin-induced hypoglycemia than during supine rest.1,27⇓ In PAF, MSA + NOH, and PD + NOH, responses of plasma levels of catechols and metanephrines to stressors have not been studied.
The observed group differences in plasma levels of catechols and metanephrines in the current study lead us to predict differential responses to treatment. Because of sympathetic denervation and consequent denervation supersensitivity, patients with PAF and generalized sympathetic denervation should have improved orthostatic tolerance during treatment with the direct-acting α-adrenoceptor agonist midodrine or with l-threo-3,4-dihydroxyphenylserine. As conversion of l-threo-3,4-dihydroxyphenylserine to NE depends on decarboxylation catalyzed by l-aromatic amino acid decarboxylase and carbidopa effectively inhibits the decarboxylase, patients with PD + NOH treated with levodopa/carbidopa should not respond to l-threo-3,4-dihydroxyphenylserine. Finally, patients with MSA + NOH should have large pressor responses, patients with PD + NOH intermediate responses, and patients with PAF attenuated responses to drugs that release NE from sympathetic nerves (e.g., some sympathomimetic amines, yohimbine).
Considered in isolation, the findings depicted in figure 3 would raise the possibility of a bilinear process describing the normal relationship between plasma DHPG and NE levels, by which the low levels of both compounds in PAF might reflect decreased exocytotic release of NE as an alternative to sympathetic denervation. Previous reports, however, have demonstrated that when exocytotic release of NE is markedly decreased, such as by ganglion blockade or clonidine administration, the plasma DHPG level remains above the origin, and the slope of the relationship between plasma DHPG and NE remains about the same.8 Therefore, the current finding of a y-intercept value for plasma DHPG near or at the origin in PAF supports the notion of sympathetic denervation.
- Received September 12, 2002.
- Accepted January 9, 2003.
References
Disputes & Debates: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.