Diastolic blood pressure (DBP) before and after administration of study drug. Each group (placebo, pyridostigmine bromide, pyridostigmine and 2.5 mg of midodrine hydrochloride, pyridostigmine and 5.0 mg of midodrine hydrochloride) averaged for the supine position and standing position. Asterisk indicates P<.05; dagger, P<.01. Error bars represent mean ± SD.
Standing diastolic blood pressure (DBP) (mean ± SD) of all groups with time after administration of placebo or test medication. Error bars represent 95% confidence intervals. Pyridostigmine was administered as pyridostigmine bromide; midodrine as midodrine hydrochloride.
Regression of change in diastolic blood pressure (DBP) with symptom score. The regression equation is as follows: symptom score = 0.524 + 0.0272 × DBP increase (R = 0.34). Pyridostigmine was administered as pyridostigmine bromide.
Regression of increase in standing diastolic blood pressure (DBP) after pyridostigmine bromide administration against orthostatic fall in DBP before the drug was administered.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Singer W, Sandroni P, Opfer-Gehrking TL, et al. Pyridostigmine Treatment Trial in Neurogenic Orthostatic Hypotension. Arch Neurol. 2006;63(4):513–518. doi:10.1001/archneur.63.4.noc50340
Midodrine hydrochloride is the only drug demonstrated in a placebo-controlled treatment trial to improve orthostatic hypotension (OH) but it significantly worsens supine hypertension. By enhancing ganglionic transmission, pyridostigmine bromide can potentially ameliorate OH without worsening supine hypertension.
To evaluate the efficacy of a single 60-mg dose of pyridostigmine bromide, alone or in combination with a subthreshold (2.5 mg) or suprathreshold (5 mg) dose of midodrine hydrochloride, compared with placebo.
We report a double-blind, randomized, 4-way cross-over study of pyridostigmine in the treatment of neurogenic OH. A total of 58 patients with neurogenic OH were enrolled. After 1 day of baseline measurements, patients were given 4 treatments (3 active treatments [60 mg of pyridostigmine bromide; 60 mg of pyridostigmine bromide and 2.5 mg of midodrine hydrochloride; 60 mg of pyridostigmine bromide and 5 mg of midodrine hydrochloride] and a placebo) in random order on successive days. Blood pressure (BP) and heart rate were measured, both supine and standing, immediately before treatment and hourly for 6 hours after the treatment was given.
No significant differences were seen in the supine BP, either systolic (P = .36) or diastolic (P = .85). In contrast, the primary end point of the fall in standing diastolic BP was significantly reduced (P = .02) with treatment. Pairwise comparison showed significant reduction by pyridostigmine alone (BP fall of 27.6 mm Hg vs 34.0 mm Hg with placebo; P = .04) and pyridostigmine and 5 mg of midodrine hydrochloride (BP fall of 27.2 mm Hg vs 34.0 mm Hg with placebo; P = .002). Standing BP improvement significantly regressed with improvement in OH symptoms.
Pyridostigmine significantly improves standing BP in patients with OH without worsening supine hypertension. The greatest effect is on diastolic BP, suggesting that the improvement is due to increased total peripheral resistance.Published online February 13, 2006 (doi:10.1001/archneur.63.4.noc50340).
The baroreflexes are essential in maintaining stable blood pressure (BP) in all positions.1,2When neural pathways of the baroreflex are lesioned by disease such as multiple system atrophy (MSA), pure autonomic failure (PAF), or the autonomic neuropathies, orthostatic hypotension (OH) ensues, typically associated with supine hypertension and a loss of diurnal variation in BP.3,4 The only drug that has been demonstrated in a blind treatment trial to improve OH is midodrine hydrochloride5 but at a price of aggravating supine hypertension in a dose-dependent manner.6 The efferent limb of the baroreflex is a 2-neuron system synapsing at the autonomic ganglion, where acetylcholine is the neurotransmitter. The neurotransmitter is rapidly hydrolyzed by acetylcholinesterase.7 We hypothesized that since traffic through autonomic ganglia is minimal with the patient supine and increases with orthostatic stress, cholinesterase inhibition with pyridostigmine bromide should improve OH without aggravating supine hypertension. Support for this concept was advanced in an open study.7 We reported an improvement in standing BP, especially diastolic BP (which is a better index of systemic vascular resistance than systolic BP), without worsening supine hypertension. We now report a double-blind, randomized, 4-way cross-over study of pyridostigmine in the treatment of neurogenic OH. The primary end point was improvement of standing diastolic BP fall at 1 hour postdrug. We compared the effects of oral placebo, 60 mg of pyridostigmine bromide, 60 mg of pyridostigmine bromide and 2.5 mg of midodrine hydrochloride, and 60 mg of pyridostigmine bromide and 5 mg of midodrine hydrochloride given on sequential days. These treatments were preceded by a run-in study day and a wash-out study day.
The study was performed in the inpatient Mayo Clinic General Clinical Research Center (Rochester, Minn) on a total of 58 patients with neurogenic OH. Inclusion criteria were adult men or nonpregnant women 18 years or older with clinically definite MSA, PAF, autoimmune autonomic neuropathy, diabetic autonomic neuropathy, or otherwise unspecified neurogenic OH. Orthostatic hypotension was defined as a systolic BP reduction of 30 mm Hg or higher or mean BP reduction of 20 mm Hg or higher that occurs within 3 minutes of standing up, criteria that are standard for Mayo and stricter than consensus criteria.8 Neurogenic OH is defined as OH that was due to a structural lesion of baroreflex pathways (afferent, central, or efferent). The primary objective was to evaluate the efficacy of a single 60-mg dose of pyridostigmine bromide, alone or in combination with a subthreshold (2.5 mg) or suprathreshold (5 mg) dose of midodrine hydrochloride, compared with placebo. The placebo and active drugs had identical appearance and all medications were administered orally. The primary end point was improvement of standing diastolic BP fall at 1 hour after drug administration. Secondary end points were (1) influence on systolic BP and supine BP measures; (2) relation of symptoms to change in BP; (3) predictors of a response to treatment; and (4) influence on plasma norepinephrine, epinephrine, and dopamine levels.
All patients underwent the following evaluation: (1) neurological examination; (2) general medical examination; (3) autonomic reflex screen to evaluate cardiovagal, adrenergic, and sudomotor functions9; (4) electrocardiography; and (5) laboratory tests (complete blood cell count with differential, chemistry panel, urinalysis, and pregnancy test in women). Autonomic deficit was corrected for the confounding effects of age and sex after which it was expressed as a total composite autonomic severity score (CASS) (range, 1-10), with subsets of sudomotor (range, 1-3), cardiovagal (range, 1-3), and adrenergic (range, 1-4).10
The study was approved by the Mayo institutional review board. Randomization was performed by our statistician (P.C.O.). The trial medications were prepared at the Mayo pharmacy. All other study personnel remained blinded to treatment assignment throughout the study.
The patient was admitted to the General Clinical Research Center the evening before commencement of the studies. The patients underwent a neurological and general medical examination, electrocardiography, and laboratory tests. On day 1, the pretreatment BP values were measured throughout the day on an identical schedule as done during the double-blind phase. On day 6 (posttreatment phase), these patients underwent a repeat neurological and general medical examination and had repeat electrocardiography and laboratory tests (as before but without a pregnancy test). Each patient was randomized to receive 1 of 4 treatments (days 2-5 inclusive): (1) placebo, (2) 60 mg of pyridostigmine bromide, (3) 60 mg of pyridostigmine bromide and 2.5 mg of midodrine hydrochloride, or (4) 60 mg of pyridostigmine bromide and 5 mg of midodrine hydrochloride. Supine and standing BP, obtained after 1 minute of standing, were measured using a sphygmomanometer (Propac; WA Baum Co Inc, Copague, NY) at baseline (0) and at 1, 2, 3, 4, 5, and 6 hours' posttreatment. A global assessment of orthostatic symptoms was done by the investigator and the patient at the end of the 1-hour point for each leg of the study. The patient's symptoms were rated on a scale of 1 to 5: 1, no improvement; 3, moderate improvement; and 5, excellent improvement.
A 5-mL blood sample was drawn supine and standing for each leg of the study at 0 and 1 hour posttreatment to determine plasma norepinephrine, epinephrine, and dopamine levels. The standing sample was drawn after the patient had been standing for 5 minutes. Patients unable to stand for 5 minutes were sat back down and blood was drawn after a total of 5 minutes of orthostatic stress (standing and sitting). Meals were standardized (low carbohydrate content). Breakfast was consumed at 6 AM and the study started at 8 AM. Lunch was provided immediately after the 5-hour measurements.
All data are expressed as mean ± SD. The primary end point was improvement of standing diastolic BP fall at 1 hour after drug administration. One-way analysis of variance with repeated measures was used to test for an overall difference in treatment effects while accounting for correlation among measurements on the same subject. If a significant overall treatment difference was found, pairwise tests were performed using the paired t test to identify significant differences between individual treatments.
The study included 58 patients with the following diagnoses: MSA, 17 patients; PAF, 15 patients; diabetic autonomic neuropathy, 11 patients; autoimmune autonomic neuropathy, 9 patients; and unspecified neurogenic OH, 6 patients. Mean ± SD age was 59 ± 11 years; 28 patients were female and 30 were male. All patients had an evaluation of cardiovagal, sudomotor, and adrenergic function and had a mean ± SD score of 7.3 ± 2.2 (sudomotor CASS, 2.1 ± 1.1; cardiovagal CASS, 2.0 ± 1.0; adrenergic CASS, 3.2 ± 0.8) on the composite autonomic severity scale (total CASS), indicative of severe generalized autonomic failure (0, no deficit; 10, maximal deficit). Age and sex effects were tested and were not found to be significant and so were not included in the final statistical models. No significant differences were seen in the supine BP measures, either systolic (P = .36) or diastolic (P = .85), indicating that pyridostigmine did not increase supine BP. Following treatment, there was significant evidence of a treatment difference for the primary end point, the standing diastolic BP fall (P = .02) (Figure 1), and a suggestion of a treatment difference in the change in standing systolic BP (P = .06) (Table 1).
Pairwise post hoc tests for the primary end point showed a significant difference between pyridostigmine and 5 mg of midodrine hydrochloride compared with 2 of the other treatments (placebo, P = .002; pyridostigmine and 2.5 mg of midodrine hydrochloride, P = .03) and a nearly significant difference vs the third treatment, pyridostigmine alone (P = .051). The BP fall for the pyridostigmine and 5 mg of midodrine hydrochloride arm was 27.2 mm Hg vs 34.0 mm Hg for placebo. Additionally, there was a significant difference between pyridostigmine alone and placebo (BP fall of 27.6 mm Hg for vs 34.0 mm Hg for placebo; P = .04).
A summary of the BP at each hour during the treatment days is shown in Figure 2 and Table 1. Hour 0 is pretreatment, with hours 1 to 6 being posttreatment. The figure shows the BP profiles over the entire duration of measurements. The lines connect the mean values, with the error bars giving a confidence interval for the mean ± SD. A large increase in BP occurred in all groups between 0 and 1 hour.
To evaluate if the change in BP at 1 hour postdrug was associated with improvement of symptoms, we linearly regressed the change in symptom score at 1 hour to the change in standing BP. A highly significant association was observed (P<.001) (Figure 3), indicating that improvement in BP indeed resulted in improvement in symptoms of orthostatic intolerance. The regression equations are as follows:
1. symptom score = 0.527 + 0.0187 × systolic BP increase (R = 0.36).
2. symptom score = 0.524 + 0.0272 × diastolic BP increase (R = 0.34).
We also evaluated if the severity of OH was predictive of a response to treatment. To do this, we regressed pretreatment orthostatic fall in diastolic BP against the increase in standing diastolic BP after pyridostigmine administration (Figure 4). Postpyridostigmine diastolic BP increase = 0.44 − 0.23 × predrug diastolic BP orthostatic fall (P < .001), indicating that the pressor response was proportional to the severity of OH.
We additionally compared responses in patients with a preganglionic site of lesion (MSA 17 patients) with patients with a postganglionic site of lesion. The site of the lesion is known in the autonomic neuropathies, PAF, and MSA. However, the site is sometimes uncertain in cases of nonspecific neurogenic OH. To handle this issue, we defined postganglionic cases in 2 ways: (1) all cases except MSA and (2) all cases except MSA and nonspecific neurogenic OH.
We evaluated if there was a differential treatment effect in individuals with preganglionic vs postganglionic sites of lesion. To test this, a 2-way analysis of variance model with repeated measures was used to assess a treatment × condition interaction. A significant interaction would suggest a stronger treatment effect in 1 group. Treatment effects were examined in both the preganglionic and postganglionic groups and were found to be similar. Using both definitions of preganglionic lesion, no significant treatment × condition interactions were seen in any of the BP measures (all P>.35).
Plasma norepinephrine level was measured both supine and standing immediately before and 1 hour after treatment. The primary measure of interest was the change from pretreatment to posttreatment levels in the amount of norepinephrine on standing. The individual norepinephrine measures were highly skewed, but the within-subject differences were well behaved. Plasma norepinephrine values were available for 38 of the 58 subjects in the trial. However, 12 subjects had at least 1 collection that failed, resulting in between 29 and 35 subjects available for each treatment comparison. Table 2 gives the norepinephrine levels and pretreatment to posttreatment changes overall and by treatment.
No significant treatment differences were found in supine (P = .99), standing (P = .35), or the difference between standing and supine (P = .39). Because no overall differences were found, no pairwise statistical comparisons were made.
The main findings of this study are that pyridostigmine alone or with low-dose (5 mg) midodrine hydrochloride will improve orthostatic BP in patients with neurogenic OH without aggravating supine hypertension, confirming the main finding of the smaller open study.7 The increase in BP is significantly related to improvement in symptoms, suggesting that the drug, especially with low-dose midodrine, will alleviate symptoms of orthostatic intolerance.
Until the present study, to our knowledge, the only drug that has been demonstrated in a blinded study to improve OH is midodrine.5 Unfortunately midodrine and other adrenergic agonists dose-dependently increase supine hypertension.6 This is a major problem because BP varies greatly throughout the day in patients with neurogenic OH, whose baroreflexes are in abeyance.11 Pressures are often low on awakening, decrease significantly following a meal, and rise in the afternoon and evening.12 Some of these changes are evident in the 6 hours of BP recordings of this study.
Baroreflex-mediated efferent activity through autonomic ganglia is proportional to changes in BP. With the patient supine, BP changes are minimal and traffic through sympathetic ganglia is relatively low. Sympathetic activity and therefore ganglionic transmission greatly increases with orthostatic stress. On this background, a cholinesterase inhibitor should have the characteristic of a “smart drug” in the management of OH.13 Pyridostigmine would augment baroreflex-mediated increases in systemic resistance proportional to the magnitude of orthostatic stress. Hence, it should reduce OH (by enhancing sympathetic activity) without supine hypertension.The drug is known to increase baroreflex sensitivity in patients and the rat without autonomic failure.14,15 Our open study showing improvement of standing BP without aggravating supine hypertension appeared to support this hypothesis.7
Improvement in standing BP was achieved without a statistically significant increase in postdrug norepinephrine levels. This lack of a measurable increase was not surprising given that our patients with neurogenic OH overall had a modest increment in norepinephrine level in response to standing up, reflecting the severe autonomic failure due to a lesion of the preganglionic (MSA) or postganglionic (diabetic autonomic neuropathy, PAF, autoimmune autonomic neuropathy) axon. In addition, norepinephrine level is an insensitive index of adrenergic activity. Only 20% of the neurotransmitter spills over into plasma; 80% is taken back up by the axon and metabolized.16,17 It is quite possible that there is a sufficient increase in muscle sympathetic nerve activity and norepinephrine, which is released and bound by adrenergic receptors, to improve BP without measurable increase of spillover. Finally, plasma norepinephrine change may not reflect a change in secretion because there are simultaneous changes in clearance in patients with autonomic failure.18
The orthostatic pressor response was less robust in the blinded study than in the open trial. One confounding effect of this more ambitious blinded study is the timing of the 1-hour postdrug BP recordings. Because the study design required BP recordings over 6 hours, we planned patient food intake at times that were thought to minimize the effects of a meal. It is clear from the large and highly significant BP increase between 0 and 1 hour (P<.001) in all treatment arms that there is reflection of recovery from the hypotensive effects of breakfast. An additional hour delay in commencing the study might have minimized this effect.
Pharmacological treatment of neurogenic OH is problematic. Fludrocortisone acetate has been reported to be beneficial19,20 but aggravates supine hypertension. α-Adrenergic agonists, including midodrine, aggravate supine hypertension and carry the risk of intracerebral hemorrhage.21-23 One such agent, phenylpropanolamine, has been withdrawn because of this complication. On this background, pyridostigmine is a welcome addition. Its dose-effect properties need to be established. If its benefits are similar to that at another synapse, the neuromuscular junction, then dose titration to much higher doses may result in greater efficacy. Even the current level of benefit offers promise as starting treatment, with additional benefits gained by combination with the smallest doses of midodrine needed to achieve improved orthostatic function.
Some support for the value of pyridostigmine derives from a follow-up study by Sandroni et al.24 Long-term use of pyridostigmine was not part of our protocol, but because this is an approved drug, the majority of patients chose to continue the use of the drug. Twenty (69%) of 29 patients available for follow-up were still taking pyridostigmine (mean ± SD treatment duration, 19.5 ± 8.9 months); 5 of 20 were receiving pyridostigmine monotherapy. Seventeen (85%) of 20 were extremely satisfied with the medication and rated their orthostatic symptoms as moderately to markedly improved. Ten patients reported an increased energy level. Among the 9 patients who had stopped pyridostigmine treatment, 6 found the drug not helpful while 3 had unacceptable adverse effects. Two of 9 had marked disease progression since starting pyridostigmine treatment and reported that the drug was initially helpful but was no longer efficacious with disease progression.
No specific characteristic (either diagnosis or site of the lesion) predicted the response to pyridostigmine treatment, although patients with more severe OH may respond better than those with mild OH. Although the mean improvement in standing BP is modest, symptomatic improvement in some individuals was dramatic. To illustrate, 1 patient in the study is employed in heavy physical labor. Taking pyridostigmine alone, he is asymptomatic and has been able to forego treatment with midodrine.
The effect of the drug is similarly efficacious for preganglionic and postganglionic disorders with a sufficient representation of the major categories of causes of neurogenic OH in our cohort. In conclusion, pyridostigmine provides the physician with an alternative therapeutic approach that minimizes the biggest problem with pressor agents in patients with impaired baroreflexes. Although the effect is modest, a small increase in standing BP may suffice to alleviate symptoms in patients with OH, who typically have an expanded cerebral autoregulation.25 Finally, combination of pyridostigmine with low-dose midodrine could provide a more potent and more sustained pressor response without aggravating supine hypertension.
Correspondence: Phillip A. Low, MD, Mayo Medical Center, 200 First St SW, Rochester, MN 55905 (firstname.lastname@example.org).
Published Online: February 13, 2006 (doi:10.1001/archneur.63.4.noc50340).
Accepted for Publication: December 20, 2005.
Author Contributions:Study concept and design: Singer, Sandroni, Opfer-Gehrking, O’Brien, and Low. Acquisition of data: Singer, Sandroni, Opfer-Gehrking, Suarez, Klein, Hines, and Low. Analysis and interpretation of data: Singer, O’Brien, Slezak, and Low. Drafting of the manuscript: Singer, Opfer-Gehrking, O’Brien, and Low. Critical revision of the manuscript for important intellectual content: Sandroni, Suarez, Klein, Hines, O’Brien, Slezak, and Low. Statistical analysis: O’Brien and Slezak. Obtained funding: Low. Administrative, technical, and material support: Sandroni, Opfer-Gehrking, Hines, and Low. Study supervision: Opfer-Gehrking, Suarez, Klein, and Low.
Funding/Support: This work was supported in part by grants NS 32352, NS 39722, NS 44233, and NS 43364 from the National Institutes of Health, Bethesda, Md; grant MO1 RR00585 from the Mayo General Clinical Research Center, Rochester, Minn; and Mayo Funds.
Previous Presentation: Results of this study were presented at the 129th Annual Meeting of the American Neurological Association; October 5, 2004; Toronto, Ontario.