Clinical trial of home blood pressure monitoring following midodrine administration in hypotensive individuals with spinal cord injury

Abstract

Background

Individuals with spinal cord injury (SCI) above thoracic level-6 (T6) experience impaired descending cortical control of the autonomic nervous system which predisposes them to blood pressure (BP) instability, including includes hypotension, orthostatic hypotension (OH), and autonomic dysreflexia (AD). However, many individuals do not report symptoms of these BP disorders, and because there are few treatment options that have been proven safe and effective for use in the SCI population, most individuals remain untreated.

Objective

The primary aim of this investigation was to determine the effects of midodrine (10 mg) prescribed TID or BID in the home environment, compared to placebo, on 30-day BP, study withdrawals, and symptom reporting associated with OH and AD in hypotensive individuals with SCI.

Design/Methods

Participants were randomly assigned to received midodrine/placebo or placebo/midodrine, with a 2-weeks washout period in between, and both the participants and investigators were blinded to randomization order. Study medication was taken 2 or 3 times/day, depending on their sleep/wake schedule, BP, and any related symptoms were recorded before and 1 h after each dosage and periodically throughout the day.

Results

Nineteen individuals with SCI were recruited; however, 9 withdrew prior to completion of the full protocol. A total of 1892 BP recordings (75 ± 48 recordings/participant/30-day period) were collected in the 19 participants over the two 30-day monitoring periods. Average 30-day systolic BP was significantly increased with midodrine compared to placebo (114 ± 14 vs. 96 ± 11 mmHg, respectively; P = 0.004), and midodrine significantly reduced the number of hypotensive BP recordings compared to placebo (38.7 ± 41.9 vs. 73.3 ± 40.6, respectively; P = 0.01). However, compared to placebo, midodrine increased fluctuations in BP, did not improve symptoms of OH, but did significantly worsen the intensity of symptoms associated with AD (P = 0.03).

Conclusion

Midodrine (10 mg) administered in the home environment effectively increases BP and reduces the incidence of hypotension; however these beneficial effects come at the expense of worsened BP instability and AD symptom intensity.

Introduction

Due to decentralized sympathetic vascular control following spinal cord injury (SCI), many individuals with lesions at or above T6 are hypotensive and experience orthostatic hypotension (OH) upon moving from a supine to a seated position.¹ However, because individuals with chronic SCI remain largely asymptomatic during periods of hypotension and OH, and do not complain and often discount symptoms associated with cerebral hypoperfusion, they are rarely treated. We previously reported that although 40% of Veterans with cervical SCI had hypotensive blood pressure (BP) values entered into the medical record, less than 1% were prescribed an anti-hypotensive agent.² Yet asymptomatic hypotension and OH are associated with adverse effects on cerebral blood flow velocity (CBFv)³ and cognitive function,⁴ and the use of an anti-hypotensive agent has been shown to increase BP and CBFv in the SCI population.^3,5–7 However, treatment to raise BP in hypotensive individuals with SCI is not routinely considered, because there is also a lack of evidence supporting the safe and effective use of anti-hypotensive agents on BP instability (i.e. daily fluctuations in BP) or the risk and incidence of autonomic dysreflexia (AD).

We recently published data describing the effects of single-dose midodrine hydrochloride (10 mg) on seated BP, CBFv, and cognitive performance in 41 hypotensive individuals with SCI.⁸ The results indicate that, compared to placebo, midodrine significantly increased seated systolic BP (SBP) and mean CBFv without beneficial effects on cognitive performance.⁸ While these data provide the strongest evidence to date to support midodrine use in the SCI population, 40% of the participants tested had excessive BP responses to midodrine in a controlled laboratory setting.⁸ Until we understand the effects of midodrine administration on BP in the home environment during routine activities, recommendation for widespread prescription use in the SCI population remains tenuous.

To improve our understanding, we conducted a randomized, placebo-controlled, double-blinded clinical trial to determine the effects of midodrine on BP, adverse events, and symptoms reporting, in hypotensive individuals with SCI over a 30-day period in the home environment. The primary objective was to determine the effects of midodrine (10 mg) prescribed twice daily (BID) or three times daily (TID), compared to identical encapsulated placebo, on 30-day SBP and diastolic BP (DBP). Specifically, we compared the effect of midodrine to placebo on average BP and on the fluctuations (standard deviation) in BP over each 30-day period. We hypothesized that midodrine would significantly increase average 30-day BP but would not significantly affect BP fluctuations as compared to placebo. The secondary objectives were to compare study participant withdrawals, percentages of BP values above and below a pre-defined (see Methods) target range for SBP and DBP and the frequency and intensity of symptoms reported in association with AD and OH following administration of midodrine and placebo.

Methods

Participants

This trial was conducted at the James J. Peters Veterans Affair Medical Center (VAMC), Bronx, New York, and at the Kessler Foundation, West Orange, New Jersey. Participants who were enrolled in our previous trial⁸ and did not exhibit excessive increases in BP in response to a single 10 mg dose of midodrine were screened for participation in this trial. Participants with chronic SCI were deemed eligible if they met criteria for hypotension (SBP <110 mmHg for males or <100 mmHg for females)⁹ or OH (fall in SBP ≥ 20 mmHg and/or a fall in DBP ≥ 10 mmHg when moved from supine to the seated position.¹⁰ Participants were deemed ineligible if they met any of the following exclusion criteria: (1) ≥3 self-reported symptomatic episodes of AD per week, (2) current illness or infection, and (3) documented history of hypertension, diabetes, traumatic brain injury or any neurological condition other than SCI. All study procedures were approved by the IRB committees at the James J. Peters VAMC, and the Kessler Foundation, and the study was registered at ClinicalTrials.gov (NCT2307565).

Randomization and dispensing of study medication

Eligible participants were issued a participant ID that was associated with the 30-day randomization scheme as either: midodrine/placebo or placebo/midodrine. The study participants, research staff, and investigators were blinded as to the order of administration during both study arms, which were separated by a two-week washout period. The research pharmacist at the James J Peters VAMC generated the randomization scheme and dispensed the 30-day supply of midodrine/placebo in identical encapsulated tablets. Daily dosing of study medication was individualized based on the sleep/wake cycle of each participant. Prior to dispensing of study medication, the research coordinator met with participants to determine their daily seated/supine resting cycle; participants who reported ≥12 hours/day seated in their wheelchair were dispensed ∼90 ± 9 tablets during each randomization period to allow for TID administration over the 30-day trial; participants who reported ≤12 hours/day seated in their wheelchair were dispensed ∼60 ± 6 tablets for BID administration. Participants were asked to take their study medication in the morning, after transfer from their bed to their wheelchair, and then every four hours (midodrine half-life is 3–4 hours).

Blood pressure monitoring

Participants were asked to closely monitor their BP using an ambulatory home monitor (ABPM, BP791IT, Omron Healthcare, Inc., Lake Forest, IL, or Medtronic Interactive Voice Response System) and to record their BP at least three times a day. Specifically during the trial, participants were instructed to take their BP just before and one-hour after taking the study medication/placebo and during any activity that may have impacted their BP (i.e. routine bowel care, physical exertion, bladder emptying, etc.), and to record any symptom(s) that may have been related to the BP. Participants transmitted their BP in real-time to a secure website, where study investigators were able to see the uploaded BP and could contact the participant if data were missing or if there was any evidence of a hypertensive episode. In addition to recording BP and symptomology, participants were asked to record daily activities (i.e. transfers, bowel care, physical activity, eating, etc.). From these data we calculated the mean and standard deviation of 30-day SBP and DBP, and compared these metrics following placebo or midodrine administration, as our primary objective. Secondarily, we calculated and compared the number and percent of BP recordings above, within and below a pre-defined target range following administration of placebo versus midodrine. The target ranges, i.e. SBP: males: 110–129 mmHg; females: 101–129 mmHg and DBP: 70–89 mmHg, were chosen based on the 1978 W.H.O. definition of hypotension⁹ and recent American Heart Association guidelines of normal SBP <130 mmHg and high/normal DBP < 89 mmHg.¹¹

Adverse events

The number of participant withdrawals and reasons for study withdrawal were compared for placebo versus midodrine administration. The occurrence of BP values that were outside the target range under the condition of placebo versus midodrine administration were recorded as either above the range (SBP ≥ 130 mmHg, DBP ≥ 90 mmHg), based on current hypertension and high blood pressure guidelines^12,13 or below the range (SBP <110 mmHg for males and <100 for females), based on World Health Organization recommendations⁹ and (DBP < 70 mmHg), based on a recent report.¹⁴

Symptoms monitoring

Each week, participants were asked questions related to the symptoms of AD and OH to capture both the frequency and intensity of symptoms over the preceding 7-day period. The symptoms included headache, pounding in the ears, tingling or itching of the scalp, goosebumps, chills, blurry vision, and dizziness. Additionally, participants were asked questions related to known adverse effects of midodrine including constipation, muscle spasms, urinary urge or increased lower abdominal pressure. The participants were asked: “how often did you experience this symptom?” which were rated on the following scale: 0-never, 1-occasionally, 2-sometimes, 3-often, or 4-always, and then: “on a scale of 0–100, how intense was the symptom?” 0 = not at all, 100 = very intense. The frequency and intensity of symptoms reported over each 30-day period were compared weekly under the conditions of midodrine versus placebo.

Statistical analyses

To assess the effects of drug treatment on SBP and DBP, multiple pressure values across both study arms, in each participant (number of pressure recordings ranged from 2 to 179), were analyzed for all 19 study participants using mixed modeling. For each variable, the random effect was “subject” and the fixed effect was “condition” (midodrine or placebo). To assess 30-day BP fluctuations, the absolute deviation of each BP value for each participant was calculated relative to that participant’s mean BP over each condition. The data were transformed into ranks; the ranks were then used as the dependent variable in the mixed model analysis. As above the random effect was “subject” and the fixed effect was “condition”.

For the symptoms data, the AD and OH intensity scores were converted to numerical rankings; the resulting transformed data were analyzed using mixed models with the random factor = “subject” and the fixed factor = “condition”. For the frequency of AD and OH symptoms data, the count data were analyzed using mixed modeling with the Poisson distribution. Similarly, for the adverse events data, the counts of SBP and DBP recordings above and below the target range, adjusted for the total number of recordings, were analyzed using mixed modeling with the Poisson distribution. To compare study withdrawals occurring under placebo versus midodrine, the events were compared using the nonparametric Cochran’s Q test.

All continuous data are presented as mean ± standard deviation and significance was set at an alpha of <0.05. The mixed modeling statistical calculations for the continuous variables were performed using R software (v 3.6.2), and the number of incidents of symptoms reporting for AD and OH was analyzed with Poisson mixed modeling using Jamovi software (v1.2.9). The dropout data were analyzed using a custom Cochran’s Q program written with LabVIEW programming software.

Results

Characteristics of the participants

A total of 19 individuals with SCI were consented to participate; the demographic characteristics are presented (Table 1). Six participants were female, all were chronically injured (1–45 years post-SCI) with cervical (C4-T1), mostly motor complete (79%) lesions. Self-reported levels of physical activity suggested that most (85%) engaged in at least 1 day of exercise/week at a moderate intensity and that they were otherwise healthy. During the first 30-day period, 8 participants were randomly assigned to receive midodrine and 11 participants received placebo; there were no differences in demographic characteristics by randomization order.

Table 1.

Participant characteristics.

ID	Order	ARM_1 days	ARM_2 days	Sex	Age years	Level	AIS	DOI years
M001	MP	30	30	Female	40	C5-6	A	24
M002	PM	30	30	Male	48	C5-6	A	32
M008	MP	30	30	Male	53	C4	C	4
M009	PM	30	30	Male	27	C5-6	B	3
M010	PM	30	30	Female	34	C4-5	B	2
M012	PM	30	30	Female	38	C5	B	16
M014	MP	30	30	Male	25	C4-5	B	2
M015	PM	30	30	Male	37	C7	A	18
M016	PM	30	29	Male	50	C5	A	13
M017	PM	30	30	Male	52	C7	C	28
M003	MP	0	0	Male	46	T1	B	19
M004	MP	1	0	Male	30	C4-5	B	14
M005	PM	30	1	Female	62	C6	A	45
M006	MP	4	0	Male	56	C6-7	C	28
M007	PM	30	1	Male	34	C6	B	10
M011	MP	0	0	Female	32	C5	A	1
M013	PM	28	0	Male	41	C4-6	A	23
M018	MP	0	0	Female	47	C6-7	C	3
M019	PM	0	0	Male	37	C6-7	B	12

Note: Order: MP = midodrine/placebo; PM = placebo/midodrine; ARM_1 days and Arm_2 days = number of days of participation in the treatment and control arms.

Blood pressure monitoring

For the average 30-day BP analyses, the residuals from the mixed model deviated from normality (visual examination of the Q-Q plots); as such, the data were converted to log scores for subsequent mixed model analysis. Midodrine significantly increased average 30-day SBP (mean increase ∼12.6 mmHg, P < 0.0001) and DBP (mean increase = 7.7 mmHg, P < 0.0001) compared to placebo (Figure 1). For the BP fluctuations analyses, the effect of condition was significant for SBP (P < 0.0001; mean difference in ranks = 284.0, 95% CI = 213.2–354.7), indicating significantly wider fluctuations over the 30-day observation following midodrine as compared to placebo. For DBP fluctuations, the condition effect was also significant (P < 0.0001, mean difference in ranks = 192.1, 95% CI = 113.7–270.5), indicating significantly wider fluctuations in DBP with midodrine compared to placebo.

Figure 1.

The total recorded SBP [a] and DBP [b] values over the 30-day observation under placebo (open circles) and midodrine (gray squares) administration. The median of each BP distribution is represented by the horizontal line. *** P < 0.0001 placebo versus midodrine.

Adverse events reporting

Nine participants withdrew prior to completion of the study (Table 2). All withdrawals occurred while these participants were randomized to midodrine, and the differences in withdrawals compared to placebo were significant (Cochran’s Q = 7, P = 0.0082). Since there were no withdrawals under placebo condition, we could not compare differences in reporting of adverse events. Five of the first 7 participants withdrew due to high BP values (160/75; 141/93; 153/108; 136/85; 170/120 mmHg), with symptoms of goosebumps reporting. As such, the dosing regimen of the placebo/midodrine administration was titrated up as follows: 3-days at 2.5 mg; 3-days at 5.0 mg, and the remaining 24-days at 10 mg.

Table 2.

Participant withdrawals.

ID	Outcome	Reason	Related
M001	Completed
M002	Completed
M008	Completed
M009	Completed
M010	Completed
M012	Completed
M014	Completed
M015	Completed
M016	Completed
M017	Completed
M003	Withdrew	Inconvenient	Unrelated
M004	Withdrew	High BP	Related
M005	Withdrew	High BP	Related
M006	Withdrew	Constipation	Possibly
M007	Withdrew	High BP	Related
M011	Withdrew	Inconvenient	Unrelated
M013	Withdrew	Inconvenient	Unrelated
M018	Withdrew	Inconvenient	Unrelated
M019	Withdrew	Inconvenient	Unrelated

Differences between placebo and midodrine for the number of BP recordings below, within, and above the target range, expressed as a percentage of total recordings, are presented for each participant (Figure 2). The counts of BP recordings that remained below the target range were significantly reduced with midodrine for both SBP (∼1.7 times lower, P = 0.001) and DBP (∼1.5 times lower, P < 0.001) compared to placebo. The counts of BP recordings within the pre-defined target range were significantly increased with midodrine for both SBP (∼2-times higher, P < 0.001) and DBP (∼1.5-times higher, P < 0.001) compared to placebo. The counts of BP recordings above the target range were significantly increased with midodrine for SBP (∼2 times higher, P = 0.005) and DBP (∼3.4 times higher, P < 0.001) compared to placebo.

Figure 2.

The percent of recorded BP values below [a, b] within [c, d] and above [e, f] the target range for SBP (left column) and DBP (right column) under conditions over the 30-day observation period under the placebo (left) and midodrine (right) administration. ** P < 0.01; *** P < 0.001 indicate differences in total counts between placebo and midodrine from the mixed models with the Poisson distribution.

Symptoms reporting

The frequency and intensity of symptoms reporting association with AD and OH are presented (Figure 3). The effect of condition was not statistically significant for the frequency of AD (P = 0.14); however, the effect of condition was significant for the intensity of AD symptom reporting (P = 0.03; mean difference = −6.3 ranks for placebo versus midodrine, 95% CI for the differences in ranks of AD symptoms = −12.0 to −0.50), suggesting that the intensity of symptoms associated with AD were heightened with midodrine administration compared to placebo. We should note that the two participants who reported the highest intensity of AD symptoms (Figure 3(b)) had BP responses to midodrine that were in the hypertensive range and, as such, these participants were withdrawn from the study. Neither the frequency (P = 0.50) nor the intensity (P = 0.95) of symptoms reported in association with OH differed significantly comparing the midodrine to the placebo condition, suggesting that midodrine had no effect on OH symptoms reporting compared to placebo.

Figure 3.

The frequency and intensity of symptoms reported that were associated with AD [a, b] and OH [c, d] over each 30-day period under conditions of placebo and midodrine. * P < 0.03; indicates differences in the intensity of AD with placebo compared to midodrine administration from the mixed model analysis.

Discussion

The use of anti-hypotensive agents to treat persistently low BP and OH is not common in the SCI population which is due, in part, to the lack of safe and effective treatment options.¹⁵ As such, we set out to better define the effects of the most commonly prescribed anti-hypotensive agent on BP, adverse events and symptoms reporting in a hypotensive cohort of participants with SCI. The findings suggest that while midodrine 10 mg administered either BID or TID significantly increased average 30-day BP, and reduced the number of BP values recorded below our target range, midodrine significantly increased the number of BP recordings above our target range, widened the 30-day fluctuations in BP, and increased the intensity of AD symptoms reported compared to placebo. Therefore, before considering long-term clinical prescription of midodrine in patients with SCI, close ambulatory monitoring of BP and recording of symptoms during the activities of daily living in the home environment is warranted.

Although it is not advisable to neglect to treat hypotension and OH in persons with SCI simply because the patients do not complain of symptoms or report the adverse impact that BP instability has on their cognitive capabilities, mood, and daily activities, the lack of clinical attention to prescribe an anti-hypotensive agent is all too common. Diverse cognitive deficits are widely reported in the SCI population^16–18 in association with concomitant traumatic brain injury (TBI),¹⁹ pre-morbid conditions,²⁰ and impaired cardiovascular autonomic control.²¹ Further, our previous findings indicate poorer performance on cognitive tasks of memory and attention processing in asymptomatic hypotensive individuals with SCI compared to a normotensive SCI cohort matched for the incidence of TBI.⁴ Changes in BP accounted for 20% of the variance in test scores on a cognitive task of attention and processing speed²² and increases in BP with midodrine improved cognitive performance in hypotensive individuals with SCI.³ Moreover, recent evidence points to an association between cognitive deficits and increased self-reported levels of anxiety and depression in persons with SCI,²³ and we found significantly increased self-reported symptoms of depression in hypotensive compared to normotensive individuals with SCI.⁴ More worrisome, is that patients with SCI limit their daily activities due, in part, to hypotension and OH.²⁴ Thus, there is a clinical imperative to identify safe and effective treatments for hypotensive conditions in the SCI population that can be more widely prescribed with a greater degree of comfort and confidence.

Treatment of hypotension in the SCI population must consider not only increasing, but also stabilizing BP. Documented evidence shows that SBP can swing by more than 140 mmHg in a 24-hour period, and that these swings can occur upwards of 50 times per day in individuals with SCI at or above T8 due to bouts of AD and OH.²⁵ Although BP instability is more common in individuals with lesions above T6, we reported significant fluctuations over a 30-day home monitoring observation in individuals with low thoracic injuries.²⁶ Frequent, dramatic swings in BP in association with AD and OH can place a significant burden on the cerebral vasculature, and may be responsible for the increased stroke risk documented in the SCI population.²⁷ Moreover, chronic BP instability challenges the cerebral circulation, impeding autoregulation and leading to impaired neurovascular coupling and cognitive deficits.^28,29 Our finding of significantly increased fluctuations in BP over a 30-day period following midodrine administration compared to placebo is troublesome and warrants confirmation in a larger clinical trial.

Nine of the participants withdrew, 8 while randomized to midodrine, and although the reason for study withdrawal was unrelated to the protocol in 5 of the 9 participants; four of the withdrawals were probably (e.g. high BP) or possibly (e.g. constipation) related to the study medication. It should also be noted that the unrelated study withdrawals were due to inconvenience, because involvement in a clinical trial for upwards of 3-months places a significant burden of time on the participant. Additionally, although average 30-day BP remained below the target range (SBP: <110 mmHg for males and <100 mmHg for females or DBP: <70 mmHg) in 60% of the participants following midodrine administration, the observation of Stage 1 or 2 hypertension (SBP: >130 mmHg or DBP: >80 mmHg)^12,13 was significantly increased following midodrine administration compared to placebo. Although these hypertensive BP thresholds were not determined in the SCI population, the long-term implication is not known, and more data are needed to determine the safety and efficacy of midodrine for long-term use in the SCI population.

Although most hypotensive individuals with SCI are asymptomatic, understanding the impact of midodrine treatment on common symptoms of orthostatic intolerance i.e. dizziness, blurred vision, is an important component of this placebo-controlled study design. Although a dissociation between BP changes during OH and symptoms has been reported,³⁰ it is important to note recent findings that indicate an association between OH symptoms burden and cognitive deficits in hypotensive individuals with SCI.¹⁶ That said, midodrine did not lessen reporting of the frequency or intensity of OH symptoms compared to placebo in our cohort, which may have been due to the relatively few reported symptoms under placebo conditions. On the other hand, midodrine did significantly worsen the self-reported intensity of symptoms reporting associated with AD. This is an important finding because, even though we did not capture the instigating events that caused the AD symptoms, participants were blinded during the questioning and reported between 30 and 200% more intense symptoms while randomized to midodrine compared to placebo. Development of a valid and sensitive tool that can be used to identify the impact of BP disorders on the health and lives of individuals with SCI remains a clinical imperative, given the general lack of symptoms reported, and resulting diminished diagnosis and treatment of this adverse hemodynamic condition.^2,31

Limitations

The small sample size of 19 participants with SCI, and the similarity in injury characteristics among those tested, limits extrapolation of the findings to the wider SCI population. Monitoring of BP was through periodic brachial auscultation, by patients or their caregivers, at potentially different times relative to when the participant took the study medication/placebo, which might have led to the under-reporting of BP instability and the percentages of recordings below, above and within our target range. We do not report the ability of midodrine, compared to placebo, to normalize BP because there is limited guidance as to the definition of normal BP. It would be beneficial to the development of treatment algorithms for hypotension and OH to have a generally accepted target normotensive range; in the current study, we used published guidelines for high^12,13 and low BP⁹ to define the upper and lower bounds of a target range, respectively. We tested the efficacy of midodrine 10 mg administrated twice or three times a day but recognize the need to test other doses and dosing regimen, given the relatively high incidence of BP both above the upper threshold and below the lower threshold of our target range. While midodrine did not reduce the frequency or intensity of OH symptoms reported, participants were only asked questions pertaining to dizziness and blurry vision, and the use of a validated survey, such as the Autonomic Dysfunction Following SCI^25,32 may have improved our ability to discern a beneficial effect. Finally, there may have been selection bias in our study sample, since we recruited participants from a prior study who did not have hypertensive responses to a single dose of midodrine administered in the laboratory setting.

Conclusion

Treatment of hypotension and OH should be considered a high priority in the SCI population given the mounting evidence to support adverse qualitative and quantitative effects. However, the paucity of safe and effective anti-hypotensive pharmacologic agents available, limits widespread clinical intervention, thereby diminishing cognitive, emotional, and physical health and vitality for individuals living with SCI. Our findings, that describe the use of midodrine, compared to placebo, over a 30-day observation period in the home setting offer some evidence of efficacy, but this beneficial effect was observed to be at the expense of increased BP instability and heightened AD symptoms intensity. Because this study was performed in a relatively small cohort of hypotensive individuals with SCI, this is far from the final word on treatment options for hypotension and OH in persons with SCI. More work should be performed to investigate the effects of varied doses of midodrine on BP, cerebral blood flow, cognitive function and quality of life. Finally, a concerted effort should be made to greatly increase the armamentarium of safe and effective treatment options for hypotension and OH for use in the SCI population.

Funding Statement

This work was funded by the Craig H. Neilsen Foundation grant #284196; U.S. Department of Veterans Affairs, Veterans Health Administration, Rehabilitation Research & Development Service SPiRE award grant #D1382-P.

Conflict of interest

No conflicts of interest, financial or otherwise, are declared by the author(s).