Buspirone decreases susceptibility to hypocapnic central sleep apnea in chronic SCI patients

Abstract

Spinal cord injury (SCI) is a risk factor for central sleep apnea (CSA). Previous studies in animal models with SCI have demonstrated a promising recovery in respiratory and phrenic nerve activity post-injury induced by the systemic and local administration of serotonin receptor agonists such as Buspirone and Trazodone. Human trials must be performed to determine whether individuals with SCI respond similarly. We hypothesized that Buspirone and Trazodone would decrease the propensity to hypocapnic CSA during sleep. We studied eight males with chronic SCI and sleep-disordered breathing (SDB) [age: 48.8 ± 14.2 yr; apnea-hypopnea index (AHI): 44.9 ± 23.1] in a single-blind crossover design. For 13 days, participants were randomly assigned either Buspirone (7.5–15 mg twice daily), Trazodone (100 mg), or a placebo followed by a 14-day washout period before crossing over to the other interventions. Study nights included polysomnography and induction of CSA using a noninvasive ventilation protocol. We assessed indexes of SDB, CO₂ reserve, apneic threshold (AT), controller gain (CG), plant gain (PG), and ventilatory parameters. CO₂ reserve was significantly widened on Buspirone (−3.6 ± 0.9 mmHg) compared with both Trazodone (−2.5 ± 1.0 mmHg, P = 0.009) and placebo (−1.8 ± 1.5 mmHg, P < 0.001) but not on Trazodone vs. placebo (P = 0.061). CG was significantly decreased on Buspirone compared with placebo (1.8 ± 0.4 vs. 4.0 ± 2.0 L/(mmHg·min), P = 0.025) but not on Trazodone compared with placebo (2.5 ± 1.1 vs. 4.0 ± 2.0 L/(mmHg·min); P = 0.065). There were no significant differences for PG, AT, or any SDB indexes (AHI, obstructive apnea index, central apnea index, oxygen desaturation index). The administration of Buspirone decreased the susceptibility to induced hypocapnic central apnea by reducing chemosensitivity and increasing CO₂ reserve in chronic SCI patients.

NEW & NOTEWORTHY This research study is novel as it is the first study in a humans that we are aware of that demonstrates the ability of Buspirone to increase CO₂ reserve and hence decrease susceptibility to hypocapnic central apnea in patients with spinal cord injury.

INTRODUCTION

It has been estimated that the global age-standardized prevalence of spinal cord injury (SCI) is 130 (90–170) per 100,000, corresponding to 27 million individuals living with this condition worldwide (15). SCI can result in respiratory complications and is a major known cause of morbidity and mortality in this population (17). Furthermore, cervical SCI is a risk factor for sleep-disordered breathing (SDB), with a prevalence of 27–62% (32).

The treatment of choice for SDB is positive airway pressure (PAP) therapy (16). Although PAP therapy is efficacious in eliminating the majority of respiratory events, adherence is often suboptimal. Patients with SCI are reported to be even less adherent to PAP therapy than able-bodied individuals (33), making alternative therapies greatly desired.

Alternative nonmechanical therapy for SDB remains elusive (2). Serotonin (5-HT) is a modulator of brain stem respiratory circuitry and ventilatory behavior (29). Several 5-HT receptor subtypes act at different loci that mediate excitatory and inhibitory effects on the respiratory drive (30), central CO₂ chemoreception (6, 25, 36, 40), and hypoxia-induced respiratory plasticity (21). In animal models, there is evidence that the administration of the 5-HT_1a receptor agonist Buspirone can stimulate respiration and shift the apneic threshold (AT) to a lower $\text{P}{\text{a}}_{\text{C}\text{O}}{}_{{}_2}$ level (12), reducing ventilatory instability (22, 37). In humans, data suggest that Buspirone can stimulate respiration and that this respiratory stimulation persists during a sleep state (19, 28).

Likewise, the 5-HT_2c agonist Trazodone was clinically approved for depression treatment by the US Food and Drug Administration in 1982 and is now one of the most widely prescribed medications as a sleep-promoting agent in the US. Trazodone is a complex drug that acts as a 5-HT_2c antagonist at low doses, but a 5-HT₂c agonist at high doses (18) (as well as an agonist of 5-HT_2a, 5-HT₆, and 5-HT₇ receptor subtypes). Systemically administered Trazodone appears to augment spinal motoneuronal activity presumably through its downstream metabolite m-chlorophenylpiperazine (m-CPP) (18). In both animal and human models, it has been shown to increase deep sleep (41) as well as reduce indexes of SDB (34, 38). Moreover, Trazodone has a unique feature as a hypnotic agent that is particularly helpful in SCI patients who are known to have poor sleep efficiency and frequent arousals. Due to its 5-HT activity, it may also have a stimulating effect on respiratory recovery post-SCI. A recent study demonstrated in able-bodied obstructive sleep apnea (OSA) patients that Trazodone increased the arousal threshold and may stabilize breathing (13). At this time, there are no available data on the effect of Trazodone administration on SDB in an SCI population.

Therefore, we hypothesized that Buspirone and/or Trazodone would reduce the tendency of SCI patients to develop hypocapnic central apnea during sleep.

METHODS

Subjects

The Human Investigation Committee of the Wayne State University and the Veterans Affairs Medical Center approved the experimental protocol. Informed written consent was obtained. The inclusion criteria for SCI subjects were as follows: non-ventilator-dependent participants with chronic SCI (>6 mo postinjury), American Spinal Injury Association (ASIA) grade A, B, C, or D, spanning the spectrum from cervical (C1–C7) to thoracic levels (T1–T6). SCI participants were excluded from the study for any of the following: 1) being <18 yr of age, 2) being pregnant or lactating females, 3) being currently ventilator dependent or with tracheostomy tube in place, 4) having advanced heart failure, peripheral vascular disease, or stroke, 5) having a history of head trauma resulting in neurological symptoms or loss of consciousness, 6) having advanced lung, liver, or chronic kidney disease, 7) being extremely obese, defined for this protocol as BMI > 38 kg/m² (to avoid the effect of morbid obesity on pulmonary mechanics and ventilatory control).

Randomization and Study Design

The study design was a single-blinded, randomized crossover pilot physiological study. Each participant went on Trazodone (100 mg), Buspirone (30 mg), and placebo for 2 wk each and had a night study on day 13 of being on the medication. Each participant had a washout period of at least 2 wk between finishing one medication and beginning a new medication. The order in which participants received the medications was randomized, and the participants were blinded to which medication they were being given. Buspirone was taken twice daily, once in the morning with breakfast and once immediately before bed. Participants were titrated up to 30 mg daily (15 mg) after starting at a dosage of 15 mg daily (7.5 mg) and then increasing the dosage by 5 mg every 3 days until 30 mg total daily dose was achieved. Trazodone and placebo were instructed to be taken immediately before bed.

Measurements

Every participant underwent overnight in-laboratory polysomnography (PSG) using the Comet Polysomnography System (Grass Technologies, Warwick, RI) or the SomnoStar z4 Sleep System (Vyaire Medical, Mettawa, IL). Despite variability in ASIA scores and potential differences in sleeping position, we ensured that all participants remained in a supine position throughout the night and during all measurements. Respiratory events were scored using American Academy of Sleep Medicine (AASM)-recommended criteria (3). Airflow was measured using a pneumotachometer (model 3700A; Hans Rudolph, Shawnee, KS) connected to a tight-fitting nasal mask. Tidal volume (V_T) was obtained by integrating the pneumotachograph flow signal. ECG was measured using a modified lead II electrode placement. Oxygen saturation ($\text{S}{\text{a}}_{{\text{O}}_2}$) was measured using a Datex-Ohmeda Biox 3740 Pulse Oximeter (Louisville, CO) or SomnoStar z4 Sleep System. End-tidal CO₂ (${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$) and end-tidal O₂ (${\text{P}\mathrm{E}\mathrm{T}}_{{\text{O}}_2}$) were measured using the GEMINI O₂ and CO₂ Monitor (CWE, Ardmore, PA). Participants were instructed to abstain from using tobacco products, drinking alcohol, and consuming caffeine on the night of the study.

Apneic Threshold Determination

Noninvasive ventilation protocol.

We used the OmniLab Advanced + Titration System (Philips Respironics, Murrysville, PA) to give positive airway pressure (PAP) in the mask to eliminate obstructive respiratory events. Hyperventilation was then induced by increasing the inspiratory pressure for at least 2 min, resulting in a hypopnea or central apnea. Noninvasive ventilation (NIV) was terminated during expiration to the baseline expiratory PAP for a minimum of 3 min. The hyperventilation trials were incrementally repeated at higher pressure support (1–2 cmH₂O) until a central apnea was obtained. If a central apnea resulted from the NIV trial, the trial was repeated at lower pressure support (1–2 cmH₂O) to identify the nearest ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ to the AT.

Hypercapneia protocol.

For subjects who had spontaneous central apnea during sleep, we mixed 40% CO₂ into a port on the face mask for 2-min trials, starting with a flow rate of 0.5 L/min. If central apneas were not abolished, the CO₂ flow rate was increased by 0.5 L/min in subsequent trials until central apneas were abolished. At least 2 min of recovery time was given between trials. Breaths were excluded for analysis if they followed arousal or if the ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ signal did not adequately plateau.

Data Analysis

Baseline wake and nonrapid eye movement (nREM) sleep ventilation were monitored in each subject. Periods of 10 breaths from wakefulness and stable nREM sleep (no apneas preceding it for 2 min) were measured to assess baseline ventilation (V̇e, V_T, f, T_I, T_E, $\text{P}{\text{E}}_{\text{C}{\text{O}}_2}$, and $\text{S}{\text{a}}_{{\text{O}}_2}$), using the PowerLab acquisition system (model 16SP; AD Instruments, Colorado Springs, CO). SDB was identified if the calculated apnea-hypopnea index (AHI) was greater than or equal to 5 events/h of sleep. Central SDB was defined as AHI greater than or equal to 5 events/h of nREM sleep and a central apnea index (CAI) greater than or equal to 5 central events/h of sleep. As depicted in Fig. 1, after stable nREM sleep was achieved, in each hyperventilation trial, the control period was represented by the average of five breaths immediately preceding the onset of mechanical ventilation. The average ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ value for these five breaths is referred to as the eupneic ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$. The hyperventilation data were the calculated averages of the last five NIV breaths before the ventilator was turned back to the baseline expiratory PAP. The average ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ value for those five breaths that occurred before the induced central apnea is referred to as the AT ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$. The CO₂ reserve was calculated as the difference between the AT ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ and the eupneic ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ values. Apneic Threshold (AT) was defined as the ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ at which the apnea closest to the last hypopnea occurred and was measured at the end of the hyperventilation trial at which the apnea occurred. The CO₂ reserve was defined as Δ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ between control and central apnea onset. If the nadir breath in the recovery period had a maximum flow value of less than 0.05 L/s, the trial was considered to have produced an apnea. Steady-state plant gain (PG) was defined as the ratio of the ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ to V̇e for all breaths during the control segments of each valid trial. The controller gain (CG) was defined as the V̇e of the breaths during the control period of the trial producing the AT divided by the change in ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ between the mechanical ventilation and baseline breaths. Trials were excluded if arousals or respiratory events occurred less than 1 min before the hyperventilation was induced or 1 min before the termination of hyperventilation. Trials were also excluded if they occurred during wake or REM sleep. Individual breaths were excluded from analysis if the ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ signal did not adequately plateau. One participant who had central sleep apnea (CSA) was given CO₂ until the CSA was eliminated. For this individual, the CO₂ reserve was defined as Δ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ between the control period (stable nREM sleep after central apnea is eliminated) and the last five breaths before central apnea occurred. CG could not be calculated for this participant.

Fig. 1.

Representative polygraphs from the noninvasive ventilation protocol depicting the difference in CO₂ reserve in a spinal cord injury subject between the placebo (A), Buspirone (B), and Trazodone (C) arms. Eupneic ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ is represented by the top dashed line of the bracket, and the AT ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ is represented by the bottom dashed line of the bracket, and the difference between these two values is the CO₂ reserve. The breaths used to calculate these values are denoted with asterisks. The CO₂ reserve was larger for Buspirone than for placebo and Trazodone. ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$, end-tidal CO₂ partial pressure; AT ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$, end-tidal CO₂ partial pressure at apneic threshold; PMask: mask pressure.

Statistical Analysis

Demographic characteristics were summarized as means (SD) or frequency and percentage scores, as applicable. A repeated-measures analysis of variance (ANOVA) was used to examine the change in CO₂ reserve, CG, and PG as well as SDB and ventilatory parameters between the three treatment groups (Buspirone vs. Trazodone vs. placebo). A Student-Newman-Keuls post hoc test was used for pairwise comparison in the case of statistical significance. In addition, a crossover design was used to investigate the potential influence of sequence (the order of medication administration) and period (the time of medication administration) on the effects of the three medications on CO₂ reserve. The significance level was set at P ≤ 0.05, and all analyses were performed using SigmaStat software (v.12.5; Systat Software, Richmond, CA). A power analysis using the SISA program indicated that the sample size of n = 8 resulted in 80% power to detect a change of 1.5 standard deviations at a 0.05 level of significance (for α double sided = 1.96).

RESULTS

We studied five cervical (C1–C6) and three thoracic (T4–T6) participants, with Table 1 summarizing their baseline demographics. All participants exhibited SDB (defined as AHI ≥ 5 events/h) and seven of the eight participants underwent the NIV protocol for all three arms. The other participant had CSA during the placebo trial and therefore underwent the supplemental CO₂ protocol instead; however he was able to complete the NIV protocol for the Buspirone and Trazodone trials. For this reason, this participant was not included in the analysis for CG. Figure 1 displays representative polygraphs of a participant for the placebo, Trazodone, and Buspirone arms.

Table 1.

Participant demographics at baseline

n	8
Age, yr	47.6 ± 13.8
BMI, kg/m2	26.0 ± 5.7
Sex (M/F)	8/0
NC, cm	41.3 ± 2.2
Level of injury
Cervical	5
Thoracic	3
Time since injury, yr	9.5 ± 8.5
ASIA score
A	4
C	2
D	2
ESS	10.0 ± 6.3
FSS	23.8 ± 18.0
PSQI, n = 7	9.7 ± 3.4
AHI, events/h	36.9 ± 20.1
CAI, events/h	9.2 ± 16.2
ODI, events/h	13.2 ± 17.9

Values are means ± SD. NC, neck circumference; ESS, Epworth sleepiness scale; FSS, fatigue severity scale; PSQI, Pittsburgh sleep quality index; ASIA, American Spinal Injury Association; AHI, apnea-hypopnea index; CAI, central apnea index; ODI, oxygen desaturation index. Note: one subject did not complete the baseline PSQI.

Baseline Ventilation

A number of ventilatory parameters were assessed at baseline during nREM sleep and are displayed in Table 2. None of the parameters assessed were found to be significantly different between placebo, Trazodone, and Buspirone treatments.

Table 2.

Ventilatory parameters of participants after 13 days of placebo vs. Buspirone vs. Trazodone treatment presented as means ± SD

	Placebo	Buspirone	Trazodone
V̇e, L/min	7.8 ± 1.9	8.6 ± 1.5	7.1 ± 1.6
VT, L	0.6 ± 0.1	0.6 ± 0.2	0.6 ± 0.1
f, breaths/min	13.1 ± 3.1	13.8 ± 2.5	12.2 ± 2.6
VMAX Insp, L/s	0.4 ± 0.1	0.4 ± 0.1	0.4 ± 0.1
TI, s	2.0 ± 0.6	2.1 ± 0.5	2.1 ± 0.4
TE, s	2.7 ± 0.7	2.4 ± 0.5	3.0 ± 0.8
TTot, s	4.7 ± 1.1	4.5 ± 1.0	5.1 ± 1.1
TI/TTot	0.4 ± 0.1	0.5 ± 0.0	0.4 ± 0.1
${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$, mmHg	40.1 ± 4.0	41.6 ± 5.4	41.0 ± 5.0
O2Sat, %	96.6 ± 1.7	95.5 ± 1.9	95.6 ± 3.9

Values are means ± SD. V̇e, minute ventilation; V_T, tidal volume; f, respiratory rate; V_{MAX Insp}, maximal inspiratory pressure; T_I, inspiratory duration; T_E, expiratory duration; T_Tot, breath duration; T_I/T_Tot, fractional inspiratory time; ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$, end-tidal CO₂ partial pressure; O₂ Sat, oxyhemoglobin saturation.

Physiological Outcomes

Figure 2A displays the CO₂ reserve on placebo, Trazodone, and Buspirone. CO₂ reserve was widened significantly on Buspirone compared with placebo (−3.6 ± 0.9 vs. −1.8 ± 1.5 mmHg, respectively, P < 0.001), and Buspirone compared with Trazodone (−3.6 ± 0.9 vs. −2.5 ± 1.0 mmHg, respectively, P = 0.009) but not on Trazadone compared with placebo (−2.5 ± 1.0 vs. −1.8 ± 1.5 mmHg, respectively, P = 0.061). Using generalized least squares analyses as performed by Procedure Mixed in SAS, the findings for the crossover design indicated that there were significant differences in mean CO₂ reserve between Buspirone and Trazodone treatments (P = 0.015) and between Buspirone and placebo treatments (P = 0.003). However, the sequence and period effects were not significant. Figure 2B shows a summary of AT ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ between the different interventions. There were no significant changes in AT ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ for Buspirone (37.3 ± 3.0 mmHg) or Trazodone (36.7 ± 3.7 mmHg) compared with placebo [37.2 ± 4.7 mmHg, P = not significant (NS)]. In addition, for eupneic CO₂ there was no significant difference between placebo (39.32 ± 4.76 mmHg) and either Trazodone (39.31 ± 3.12 mmHg, P = NS) or Buspirone (40.96 ± 2.55 mmHg, P = 0.084). Figure 2C shows that Buspirone significantly decreased CG compared with placebo (1.8 ± 0.4 vs. 4.0 ± 2.0 L/(mmHg·min) respectively, P = 0.025) but not Trazodone compared with placebo (2.5 ± 1.1 vs. 4.0 ± 2.0 L/(mmHg·min) respectively; P = NS). As illustrated in Fig. 2D, PG was not significantly different for either Buspirone (5.6 ± 1.1 mmHg·min/L, P = NS) or Trazodone (6.5 ± 2.0 mmHg·min/L, P = NS) compared with placebo (5.1 ± 1.7 mmHg·min/L).

Fig. 2.

Effect of Buspirone and Trazodone on CO₂ reserve (n = 8; A), ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$ at apneic threshold (AT ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$, n = 8; B), controller gain (CG; n = 7; C), and plant gain (PG; n = 8; D). AT ${\text{P}\mathrm{E}\mathrm{T}}_{\text{C}{\text{O}}_2}$, end-tidal CO₂ partial pressure at apneic threshold. *P < 0.05 vs. placebo using repeated-measures ANOVA.

Clinical Outcomes

A summary of the indices of SDB between each intervention arm is listed in Table 3. AHI was not significantly different for either Buspirone (48.3 ± 12.2 events/hour; P = NS) or Trazodone (44.8 ± 24.3 events/hour; P = NS) compared with placebo (44.9 ± 23.1 events/hour) (Table 3). There were also no significant differences between any of the interventions for the central apnea index (CAI), obstructive apnea index (OAI), or oxygen desaturation index (ODI) (Table 3).

Table 3.

Sleep-disordered breathing parameters of participants after 13 days of placebo vs. Buspirone vs. Trazodone treatment

	Buspirone	Placebo	Trazodone
AHI, events/h	48.3 ± 21.2	44.9 ± 23.1	44.8 ± 24.3
OAI, apneas/h	12.6 ± 17.0	15.6 ± 21.4	8.8 ± 7.6
CAI, apneas/h	0.9 (0.1–2.5)	0.8 (0.0–3.6)	0.4 (0.2–1.1)
ODI, desaturations/h	17.8 ± 16.1	12.9 ± 13.7	20.0 ± 30.6

Values are presented as means ± SD. AHI, apnea-hypopnea index; OAI, obstructive apnea index; CAI, central apnea index; ODI, oxygen desaturation index.

DISCUSSION

Summary of Findings

Our study demonstrated the following novel findings. 1) Administration of Buspirone was associated with the widening of the CO₂ reserve compared with both placebo and Trazodone and decreased controller gain compared with placebo in patients with chronic SCI. 2) Neither Buspirone nor Trazodone reduced SDB severity indexes, including AHI, CAI, OAI, or ODI. There were also no significant differences in PG or AT between the groups.

Physiological Response to Buspirone Administration

Our finding that daily intake of Buspirone was able to significantly decrease controller gain seems to be in line with those of the preliminary data presented by Borrelli and colleagues (4, 5), who observed a significant reduction in chemoreceptor sensitivity following Buspirone intake. In contrast, however, they found significant improvements in markers of SDB such as reduced AHI, CAI, and ODI, whereas we did not. This discrepancy is likely due to either a difference in the population studied or a difference in initial CAI severity. Specifically, they studied a cohort of individuals with systolic heart failure and a mean baseline CAI of 12.5 events/h. Our population consisted of SCI patients with a baseline CAI of 2.8 events/h. At this time, there have been no other assessments of these drugs in an SCI cohort, constraining us to compare our observations with other clinical populations. We excluded individuals with advanced heart failure, peripheral vascular disease, or stroke but not individuals with less severe cardiopulmonary comorbidities. It has been long since established that SCI confers a greater risk of developing comorbidities such as cardiovascular disease compared with an ambulatory population (43), with prevalence ranging from 30% to more than 50% in complete SCI compared with the 5–10% in an age-matched able-bodied population (23). To ensure a more representative sample of SCI patients, we did not exclude those with lesser cardiopulmonary morbidities, which may have influenced our findings. In addition, the greater CAI severity that the population of Borrelli and colleagues displayed may have allowed the reduced CG to manifest more readily in SDB indexes during sleep. The reduction of 2.8 events/h to 1.5 events/h with Buspirone that we observed was unlikely to be of sufficient magnitude to be detected with our statistical analysis, especially considering the small sample size and inherent variability of SDB measurements. Our study was primarily powered to detect changes in CO₂ reserve rather than indexes of SDB; so with a greater sample size there is a possibility that these alterations to CO₂ reserve and CG may have resulted in observable changes to SDB. Another difference in the study design was the size of the dose prescribed, with Borelli and colleagues giving 45 mg compared with the 30 mg that we titrated up to (4). Previous investigations of Buspirone on breathing in humans have used doses of 10 mg [5 mg twice daily (1)], 30 mg [10 mg thrice daily (9)] and 45 mg [15 mg thrice daily (4, 5)]. To mitigate adverse side-effects in our SCI population, we titrated up to a moderate 30-mg dose. At this time, it is still unclear on the optimal dose required to most effectively alleviate SDB, but the recent data from Borrelli and colleagues suggests that 45 mg may be more effective than the 30 mg we provided.

It has been shown that individuals with OSA exhibit a higher chemoreflex sensitivity (CG) and CO₂ reserve and that these factors contribute to an increased likelihood of developing central SDB (31). Our laboratory has observed that individuals suffering from cervical SCI have a reduced CO₂ reserve as well as increased PG compared with an able-bodied population (32). During sleep, the hypocapnic apneic threshold is unmasked, and the size of the CO₂ reserve can play a significant role in the development of breathing instability (7). If the CO₂ reserve is narrowed, it means that a smaller perturbation in Pco₂ levels can cause a central apnea. The combined effect of enlarging the CO₂ reserve and reducing CG is advantageous for reducing the likelihood of a CSA from occurring.

Systematic administration of serotonergic receptor agonists such as Buspirone has consistently been shown to increase ventilation and/or reduce chemosensitivity to ultimately reduce ventilatory instability (22, 37). In animal models, this has been shown in healthy cats (12) and rats (20) as well as in rats with SCI (37). Furthermore, it has been shown that Buspirone can significantly alter hypercapnic ventilatory responsiveness but not hypoxic ventilatory responsiveness in mice (42). In humans, Buspirone has been shown to increase ventilation in a patient with Rett syndrome (1) as well as in an individual with brain stem infarct (9). At present, the underlying mechanisms and anatomic sites of action are still unclear (27). The presynaptic 5-HT_1a autoreceptors in the medullary raphe may play a role, appearing to be responsive in sleep (24). Another region of interest is the pre-Bötzinger complex and ventral respiratory column; however, recent data suggest that the stimulatory effect on ventilation caused by 5-HT_1a serotonin agonists may not be produced in these regions (27). Regardless of the exact location, the primary mechanisms through which Buspirone can modulate breathing instability seems to be its ability to blunt chemosensitivity (and reduce loop gain) and increase ventilation.

Physiological Response to Trazodone Administration

We did not detect any significant differences between the Trazodone and placebo interventions in any of the physiological, ventilatory, or clinical outcome measures in patients with chronic SCI. There is some evidence that Trazodone can increase the arousal threshold as well as alleviate the indexes of SDB in an OSA population (34). For example, in both animals (38) and humans (34), it has been shown to significantly reduce the number of respiratory events that occurred during nREM sleep. Although the mechanisms are not completely understood, it has been suggested that Trazodone may enhance upper airway dilatory activity or improve stability of sleep (i.e., fewer arousals during sleep, leading to fewer CO₂ fluctuations and thus less instability) (8, 38, 39). Serotonergic control of breathing is a complex and nuanced system, with many receptor subtypes (18). It has been shown that Trazodone can act as an antagonist to these receptors at low doses and an agonist (as well as an agonist to a multitude of other 5-HT receptor subtypes) at high doses (18). Low doses (25–100 mg) are typically used in the sleep field to elicit hypnotic effects (via blocking of the 5-HT_2a, histamine H1, and α1-adrenergic receptors) (14, 35), whereas higher doses of 150–600 mg may be required for antidepressant effects to manifest (via simultaneous blocking of 5-HT_2a and serotonin reuptake transporters) (35). Furthermore, the majority of previous investigations into the efficacy of Trazodone for improving indexes of sleep have used a 100-mg dose (8, 13, 34). We opted to use a 100-mg dose not only to be consistent with the previous literature but also to mitigate the risk of side effects, considering the vulnerable population being assessed. Our study design does not allow us to determine how differing doses of Trazodone may affect sleep in an SCI population; however, future investigations should seek to evaluate this factor. At this time, we do not have sufficient data to fully explain how this drug may be interacting with and modulating the mechanisms involved with breathing control. Furthermore, there do not seem to be any studies assessing the effectiveness of Trazodone on CSA, so comparisons cannot be made. The results of this study do not allow us to conclude that Trazodone can reduce the propensity to develop CSA or be effective in managing SDB in an SCI population.

Clinical Implications

Our findings imply that Buspirone may reduce the susceptibility to developing hypocapnic central apnea and hence may be a viable treatment for central SDB in patients with chronic SCI. There is currently no single medication that is considered the “drug of choice” for treating CSA (26), and medications such as sedative agents or acetazolamide may have undesirable side effects or be inappropriate for chronic use. Buspirone, on the other hand, has been shown to have a good side effect profile, shows no evidence of misuse/abuse potential, shows no interference with mental acuity, and is well tolerated with long-term therapy (>3 mo) (11). Further research is required before Buspirone can be recommended as a viable alternative to PAP treatment; however, our data show that it may have the potential to reduce central events by improving factors involved with breathing stability as well as by widening the CO₂ reserve.

Limitations

Our study had some limitations, the most noteworthy of which was the small sample size. However, we utilized a crossover design to attempt to mitigate the impact of this. This study was not powered to detect changes in parameters such as indexes of SDB, so we cannot yet be sure if the differences we observed in CO₂ reserve and CG would alter SDB indexes. This study was designed as a novel physiological pilot study to determine whether there is a sufficient physiological rationale to warrant a full-scale clinical trial in the future focused specifically on clinical outcomes as a primary outcome. In addition to the small sample, we encountered some difficulties in obtaining a sufficiently homogenous population due to variations in injury severity (n = 4 ASIA A, n = 2 ASIA C, n = 2 ASIA D) and level (n = 5 cervical, n = 3 thoracic). Due to physiological differences caused by these injury variations, our ability to detect changes in ventilatory parameters or SDB indexes may have been obscured. However, despite the variation, the thoracic and cervical groups manifested similarly under the current doses of Buspirone and Trazodone, as depicted in Fig. 2. Future studies focusing on comparison between different levels of SCI could allow for a better assessment of the clinical significance these medications may confer. Furthermore, the time since SCI occurrence was variable within our cohort (median 8.5 yr, range 1–27 yr). Guidelines for conducting clinical trials within SCI patients (10) indicate that the most dramatic improvements occur over the first 9 mo, although a completely stable baseline may not be achieved until 18 mo after SCI (dependent on severity and level of the injury). Considering that SCI had occurred at least 18 mo prior in all but one participant (1 yr since injury), we believe that our sample displayed an adequate functional baseline. Another consideration is that only male participants completed this study, meaning that the conclusions of this research must be interpreted carefully and can be applied only to that demographic at this time. Additionally, some participants had difficultly tolerating the instrumentation involved with undergoing polysomnography, which might have affected natural sleep and thus the applicability of these findings to normal sleep. All trials were performed during stable nREM sleep using incremental increases in inspiratory pressure to induce the central apnea which usually occurred later in the night after achieving adequate ventilation. Therefore, it is unlikely that adaptation to instrumentation may have resulted in differences in trials during the same night. Another consideration was participant compliance for taking the prescribed medications. To attempt to ensure compliance, we called each participant three times per week to remind them to take the medications and also asked them to bring the pill bottle containing the medication to the laboratory on the night of the test (enabling us to count any remaining pills). Despite taking these measures and asking them to self-report intake, they may have forgotten and/or misplaced the medication. Additionally, it is worth noting that the Buspirone arm involved the participants taking the medication twice daily while progressively increasing the dose, whereas the placebo and Trazodone arms involved the participants simply taking the medication once at bedtime. This difference in administration of medications in these arms resulted from differences in recommended dosing regimens between Buspirone and Trazodone. It is unlikely that the differences in doses between the arms had any significant impact on the final results. Finally, it is worth noting that most participants recruited into this study exhibited a higher degree of obstructive sleep apnea than central sleep apnea. This could explain the observation that these medications did not produce a change in AHI, as we would expect the medications to affect only the central component of SDB. It would be worthwhile to perform studies on SCI participants with a greater central component of SDB to see if this would influence AHI.

Conclusion

In conclusion, our data showed that Buspirone was able to increase the CO₂ reserve and to reduce chemosensitivity in a population of chronic SCI individuals. The combination of improving these factors is a decreased susceptibility to developing central sleep apneas. Future studies utilizing a larger sample size and a population with a greater central apnea component should be performed to assess the effectiveness of Buspirone in decreasing indexes of SDB in patients with SCI.

GRANTS

This study was funded by the US Department of Veterans Affairs (VA) Merit Review No. 1I01CX001040 and R01HL130552 and Career Development Award No. 1IK2CX000547 from the Clinical Science Research and Development Service of the VA Office of Research and Development.

DISCLAIMERS

The opinions expressed in this article reflect those of the authors and do not necessarily represent official views of the Department of Veterans Affairs. This was not an industry-supported study.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

M.S.B. and A.S. conceived and designed research; S.M. and E.K. performed experiments; S.M., S.V., E.K., B.A., and H.Y. analyzed data; S.M. M.S.B. and A.S. interpreted results of experiments; prepared figures; S.M., J.L.P., M.S.B., and A.S. drafted manuscript; S.M., J.L.P., S.V., E.K., H.Y., M.S.B., and A.S. edited and revised manuscript; M.S.B. and A.S. approved final version of manuscript.