The physiologic significance of these findings and their relevance to humans, however, has not previously been studied. In caring for patients with severe autonomic failure, we realized that anemia was not an uncommon occurrence. In exceptional patients, anemia was severe enough that it was thought to contribute to the patient's symptoms and was treated with blood transfusions. No obvious cause for anemia was apparent in these patients. Therefore, we hypothesized that if indeed the sympathetic nervous system modulated erythropoiesis, patients with severe autonomic failure may present with hypoproliferative anemia. We also studied the effect of recombinant erythropoietin therapy in these patients. Preliminary reports of these studies have been presented previously [6, 7].

Methods

Top Methods Results Discussion Author & Article Info References

Patients

We analyzed the clinical and laboratory data of 84 consecutive patients with primary autonomic failure. These patients were referred to Vanderbilt University's Autonomic Dysfunction Center from 1983 to 1992 because of symptomatic orthostatic hypotension and, therefore, represent a selected sample of patients.

Patients were diagnosed as having either multiple system atrophy (the Shy-Drager syndrome) or pure autonomic failure, based on their clinical features. Patients with pure autonomic failure are characterized by isolated involvement of the autonomic nervous system (orthostatic hypotension without compensatory heart rate increase, impaired sinus arrhythmia, lack of blood pressure overshoot during phase IV of the Valsalva maneuver, and impaired pressor response to isometric exercise and the cold pressor test). They also have decreased sweating, impaired gastric emptying, and impotence. Patients with multiple system atrophy, in addition to autonomic impairment, have central neurologic involvement that usually includes, but is not limited to, Parkinson disease. Patients with secondary forms of autonomic failure (for example, diabetes mellitus) were excluded from the study.

Patients with known causes of anemia (for example, iron deficiency, megaloblastic anemia) or concurrent illnesses known to cause the anemia of chronic disease (for example, inflammatory processes) were excluded from evaluation on the basis of clinical findings and laboratory results. All protocols were approved by the institutional review board, and patients gave informed consent.

Measurements

Patients were admitted to Vanderbilt University's Clinical Research Center and received a diet containing 150 mEq of sodium and 60 mEq of potassium per day. All medications were withheld. Autonomic evaluation was done as previously described [8].

Routine laboratory studies and hematologic values were determined in our clinical pathology laboratory. Serum erythropoietin levels were measured in the last 28 patients evaluated, using an enzyme immunoassay (Clinigen; R&D Systems, Inc., Minneapolis, Minnesota) [9]. Reticulocyte counts were corrected for the degree of anemia using the following formula: corrected reticulocyte count = reticulocyte count x (hematocrit/40). Erythrocyte volume was measured with Chromium-51-labeled erythrocytes. Plasma volume was measured with Iodine-131-labeled albumin [10]. Total blood volume was calculated by adding erythrocyte volume plus plasma volume.

Blood samples for catecholamine levels and plasma renin activities were obtained through an indwelling catheter placed in a peripheral vein. Samples were obtained after patients had been lying down overnight (supine position) and after they had been standing for 30 minutes (upright position). Patients were encouraged to remain standing upright throughout this 30-minute period but were allowed to sit down at intervals if symptoms developed. Plasma norepinephrine levels were determined by high-pressure liquid chromatography [11]. Plasma renin activity was measured by the conversion of angiotensinogen to angiotensin I and was expressed as nanograms of angiotensin I produced per millilitre of plasma per hour [12].

Because there is no consensus about what hemoglobin level to use (cutoff value) for diagnosing anemia in elderly persons, we divided our patients arbitrarily into three groups depending on their hemoglobin level: group I, hemoglobin levels less than 120 g/L; group II, hemoglobin levels 120 to 130 g/L; and group III, hemoglobin levels greater than 130 g/L.

Therapeutic Trials

Five patients with anemia (hemoglobin levels < 120 g/L) and isolated autonomic failure were treated with human recombinant erythropoietin (epoetin alfa: Procrit; Ortho Pharmaceuticals, Raritan, New Jersey; or Epogen: Amgen Inc., Thousand Oaks, California). Patients received an initial dose of 25 units/kg body weight subcutaneously, three times a week. Doses were increased by 25 units/kg at 3-week intervals until the hematocrit returned to normal. Patients also received oral iron (300 mg of ferrous sulfate one to three times daily as tolerated), as has been used in other erythropoietin trials [13], and were allowed to continue on their regular medication as long as it was kept constant during erythropoietin therapy. Complete blood cell and reticulocyte counts were done weekly for the first 4 weeks and every 3 weeks thereafter. Erythrocyte mass and plasma volume were measured before treatment and at the time the hematocrit returned to normal.

Statistical Analysis

Group differences were assessed by analysis of variance (ANOVA) using NCSS statistical software (NCSS, Kaysville, Utah). Mean differences were determined by paired or unpaired t-tests as appropriate. All hypotheses were two tailed, and the criterion of significance was P < 0.05.

Results

Top Methods Results Discussion Author & Article Info References

Patient Characteristics

The clinical characteristics of the patients are presented in Table 1. As shown by our sample of patients, primary autonomic failure typically occurs in the sixth and seventh decade of life. Patients had dramatic decreases in blood pressure while standing upright. The expected compensatory heart rate increase was virtually absent despite the profound orthostatic hypotension, indicating the severity of their autonomic failure. Plasma norepinephrine levels and plasma renin activity measured in patients lying supine or standing upright were lower than those in normal persons, an inappropriate response considering the magnitude of orthostatic hypotension observed in these patients.

Characterization of Anemia in Autonomic Failure

Patients were divided into three groups depending on their hemoglobin level (Table 1). There were more men in group 3 (P = 0.008 by ANOVA), but no sex differences were apparent in the group with the lowest hemoglobin level (group 1). The only other difference found between groups was in plasma norepinephrine concentrations measured in patients standing upright (P < 0.002 by ANOVA). Differences in plasma renin activity in patients standing upright were of borderline significance (P = 0.09 by ANOVA). No age differences were found among the groups characterized by the severity of the anemia. The diagnoses of pure autonomic failure and multiple system atrophy were evenly distributed within groups 1 and 2, but more patients in group 3 had multiple system atrophy.

Patients with the greatest degree of anemia (group 1) had a mean blood hemoglobin concentration of 108 g/L (range, 87 to 118 g/L) and a hematocrit of 0.33 (range, 0.26 to 0.37). Average values for mean corpuscular hemoglobin and mean corpuscular volume were 29 pg and 89 fL (89 microns³), respectively. The mean serum iron was 16.5 µmol/L (92 µg/dL); the mean total iron binding capacity, 43.3 µmol/L (242 µg/dL); and the mean ferritin level, 184 µg/L. Mean serum vitamin B₁₂ and folate levels were 410 pmol/L (556 pg/mL) and 22.7 nmol/L (10 ng/mL), respectively. The mean corrected reticulocyte count was 0.008. These patients had been treated previously with oral iron, folic acid, and parenteral vitamin B₁₂, with no improvement of the anemia.

We measured total blood volume in 14 patients to determine if peripheral venous hemoglobin adequately reflected erythrocyte mass in patients with primary autonomic failure. The mean hemoglobin level in this subset of patients was 123 g/L (range, 106 to 156 g/L). The mean total erythrocyte volume, corrected for body weight, was lower in patients with autonomic failure (21.4 mL/kg) than predicted values (27.1 mL/kg; P = 0.009). The average percent difference in mean erythrocyte volume ([measured mean erythrocyte volume x 100/predicted mean erythrocyte volume] –100)was –20.5%(range, –49.3% to 2.4%). A relation was found between blood hemoglobin levels and the percent decrease in erythrocyte volume (hemoglobin = percent decrease in mean erythrocyte volume x 0.066 + 13.7; n =14; r = 0.56; P = 0.01). On the other hand, the mean total blood volume in patients with autonomic failure (65.2 mL/kg; 95% CI, 58.7 to 71.7 mL/kg) was not different from values in normal persons (67.0 mL/kg; CI, 64.4 to 69.6 mL/kg) because of a slightly higher plasma volume in patients with autonomic failure (43.8 mL/kg) compared with predicted values (40.2 mL/kg; P = 0.1). These results imply that the decrease in hematocrit observed in patients with autonomic failure was not caused by changes in plasma volume but truly reflected a decreased erythrocyte mass.

We measured serum erythropoietin levels in 27 patients to determine its role in anemia noted in patients with autonomic failure. No correlation was found between serum erythropoietin and blood hemoglobin levels in these patients (serum erythropoietin = 32.5 –1.6 x hemoglobin; r = –0.30;P = 0.13 by linear regression). These results were compared with those in a control group composed of patients with iron deficiency anemia (n = 16) and normal persons (n = 12) in whom the range of hemoglobin levels (104 to 150 g/L) was similar to that found in patients with autonomic failure (105 to 152 g/L). In contrast to the findings in patients with autonomic failure, a correlation was noted between serum erythropoietin and hemoglobin levels in this control group (serum erythropoietin = 112 –7.5 x hemoglobin; r = –0.7;P = 0.003, dashed lines in Figure 1. Seven of 9 patients with autonomic failure who had hemoglobin levels less than 120 g/L fell below the 95% CI estimate of the population mean of serum erythropoietin for this control group without autonomic failure.

View larger version (16K): [in this window] [in a new window]   Figure 1. Relation between hemoglobin concentrations and serum immunoreactive erythropoietin concentrations in 27 patients with autonomic failure. Individual values are shown by solid circles (\#9679;). No linear association was noted between serum erythropoietin (EPO) and hemoglobin levels in patients with autonomic failure (r = –0.3;P = 0.13). For comparison, the linear regression and 95% confidence intervals from a control group composed of patients with iron deficiency anemia and normal persons are shown by dashed lines (EPO = 112 –7.5 x hemoglobin; r = –0.7;P = 0.003).

Relation between Anemia and Autonomic Failure

Because of the magnitude of orthostatic hypotension observed in these patients, the upright posture constitutes a considerable stimulus for sympathetic activation. Plasma norepinephrine levels measured in patients standing upright, therefore, probably reflect the maximal degree of sympathetic activity these patients can achieve and can be used as an indicator of the severity of autonomic failure. As shown in Figure 2, plasma norepinephrine levels in patients standing upright were different between groups (P = 0.002 by ANOVA); the group with the lowest blood hemoglobin level (an indicator of the severity of anemia) had the lowest plasma norepinephrine level (measured in patients standing upright), arguably an indication of the severity of autonomic failure.

View larger version (27K): [in this window] [in a new window]   Figure 2. Plasma norepinephrine levels in patients standing upright according to severity of anemia. Open bars indicate patients with pure autonomic failure (PAF), hatched bars indicate those with multiple system atrophy (MSA), and closed bars indicate all patients. Blood hemoglobin levels were less than 12 g/dL in group 1, were 12 to 13 g/dL in group 2, and were greater than 13 g/dL in group 3. Plasma norepinephrine (NE) levels were measured in patients standing upright for 30 minutes. A difference was noted in plasma norepinephrine levels among groups (P = 0.002 by ANOVA). To convert norepinephrine levels to SI units (nmol/L), divide by 169.2.

It must be noted that patients with multiple system atrophy are known to have greater plasma norepinephrine levels than patients with pure autonomic failure [11]. This was also confirmed in our sample of patients. For these reasons, each group was stratified according to diagnosis (Figure 2). The differences in norepinephrine levels (measured in patients standing upright) between hemoglobin groups persisted in both forms of the disease.

Effect of Treatment with Erythropoietin

Therapy with recombinant erythropoietin reversed anemia in all patients treated (Table 2). Mean hemoglobin levels increased from 108 to 133 g/L, and the mean hematocrit increased from 0.33 to 0.41. Only one patient required an increase in the initial dose, to 50 units/kg three times per week. No adverse side effects were observed. A tendency was noted for supine blood pressure to increase during erythropoietin therapy. Our study was not designed to determine if erythropoietin treatment had any beneficial effect on orthostatic symptoms. Nonetheless, an improvement in blood pressures measured in patients standing upright was observed during erythropoietin therapy (Table 2).

Discussion

Top Methods Results Discussion Author & Article Info References

Our results indicate that hypoproliferative anemia is common in patients uniformly characterized with severe primary autonomic failure and is unrelated to iron deficiency, megaloblastic diseases, or chronic inflammatory processes. If the World Health Organization definition of anemia is followed (hemoglobin <120 g/L for women and <130 g/L for men), then 38% of our sample of patients was anemic. Hemoglobin levels of less than 120 g/L were found in 26% of patients. Direct measurements of erythrocyte mass and total blood volume showed that the decreased hemoglobin levels were not caused by changes in plasma volume but were a reliable indicator of decreased erythrocyte production. We speculate that the lack of sympathetic stimulation results in decreased erythropoietin production, which contributes to the development of anemia in patients with autonomic failure. We review data that support this hypothesis and discuss the limitations inherent in this study.

Plasma norepinephrine levels measured in patients standing upright, arguably an indicator of the severity of autonomic failure, correlated with hemoglobin levels, an indicator of the severity of the anemia; the lower the plasma norepinephrine level for patients standing upright, the lower the hemoglobin level. It appears, therefore, that the development of anemia depends on the lack of sympathetic stimulation. However, from these results alone we cannot conclude a direct causal relation between sympathetic deprivation and anemia. It is possible, for example, that the underlying mechanisms of primary autonomic failure, which are unknown, may directly suppress erythropoietin production. Sympathetic failure could also produce anemia indirectly, for example, by decreasing the basal metabolic rate or by decreasing tissue demand for oxygen or both. However, anemia was observed in patients with pure autonomic failure and multiple system atrophy, conditions with presumably different pathophysiologic mechanisms. Anemia in patients with autonomic failure might just be a consequence of aging. It remains controversial, however, whether aging itself can produce anemia in otherwise healthy persons [14]. Likewise, the erythropoietin response to anemia does not seem to be diminished by aging [15].

Our data suggest that the lack of an adequate erythropoietin response contributes to anemia in patients with autonomic failure. Low levels of erythropoietin might be the primary cause of anemia in these patients. Absolute erythropoietin levels, however, were higher than those required to maintain a normal hemoglobin level in healthy persons. Thus, even though erythropoietin levels were inappropriately low for the degree of anemia in patients with autonomic failure, it is difficult to determine if they were low enough to be the sole cause of their anemia. It remains possible that other as-yet-undetermined factors could contribute to anemia in these patients. The characteristics of the anemia of autonomic failure—mild to moderate in severity, normochromic, and normocytic—are similar to those found in patients with anemia who have chronic diseases associated with inflammatory processes, infections, or neoplasias. Controversy exists about the pathophysiologic mechanism of these disorders, but an increased production of cytokines appears to be involved, which inhibits erythropoiesis and erythropoietin production [16-18]. Whether cytokines contribute to anemia in autonomic failure needs to be determined.

Previous studies in animals have shown a direct sympathetic stimulation of erythropoietin production by the kidney, and this mechanism may provide an explanation to our findings. Patients with autonomic failure also have low levels of plasma renin activity Table 1, renin being another hormone produced by renal cells under sympathetic modulation. A decreased number of renin-containing cells was noted in the kidneys of these patients, and it has been speculated that the loss of sympathetic innervation deprives these cells of some trophic factor [19]. On the other hand, acute sympathetic renal denervation does not appear to deplete erythropoietin-producing cells in animals [20]. It remains to be determined whether chronic denervation affects erythropoietin production in animals, as our clinical observations suggest.

Even though the precise mechanism of anemia in these patients with extreme autonomic failure is not completely elucidated, our results, taken together, support the hypothesis that the autonomic nervous system provides a stimulatory influence on erythropoiesis in humans. They also raise the possibility that anemia may contribute to these patients' symptoms. Whereas normal persons are generally not symptomatic with these levels of hemoglobin, it is possible that, in the setting of severe orthostatic hypotension, this degree of anemia could contribute to symptoms in patients with autonomic failure. Two of our patients received blood transfusions before being referred to our service and reported subjective improvement of symptoms. This prompted us to attempt therapy with erythropoietin, which effectively reversed anemia in the five patients treated. The doses of erythropoietin that were effective in patients with autonomic failure were smaller or comparable to those used in patients with renal failure [21] and were lower than those used in patients with the anemia of chronic disease, such as rheumatoid arthritis [22], the acquired immunodeficiency syndrome [23], or cancer [24].

Our study was not designed to determine if correction of the anemia would improve orthostatic symptoms. However, an improvement in blood pressure measured in patients standing upright was apparent during therapy with erythropoietin. Recent studies by Hoeldtke and Streeten [25] also suggest that erythropoietin treatment improves orthostatic hypotension in patients who have other types of autonomic neuropathy including diabetes mellitus. Symptomatic improvement could occur because of an increase in erythrocyte mass and central blood volume, as well as by improvement of tissue oxygen delivery caused by correction of the anemia. Additionally, erythropoietin therapy has been associated with increased activity of the sympathetic nervous system and with increased sensitivity to the pressor effects of angiotensin II [26]. Although the severity of the anemia alone would not justify treatment with recombinant erythropoietin in most patients with autonomic failure, the potential beneficial effect of erythropoietin for patients with autonomic failure merits further study. In patients with chronic renal failure, worsening of hypertension is a recognized side effect of erythropoietin therapy, but the mechanisms of this phenomenon remain uncertain. Further study of patients with autonomic failure may provide additional insight into the pathogenesis of erythropoietin-induced hypertension.

Mild anemia is a frequent occurrence in patients with severe autonomic failure. These patients have a low reticulocyte response and their anemia is associated with a blunted erythropoietin response. Our findings may be relevant to other conditions characterized by milder forms of sympathetic deprivation accompanied by decreased erythrocyte production, such as diabetes mellitus, prolonged bedrest, and space flight [27]. Recombinant erythropoietin reversed anemia in five patients with primary autonomic failure. Further study is needed to determine if erythropoietin provides a therapeutic approach for this condition.