Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Randomized Controlled Trial
. 2012 Nov 1;35(11):1529-39.
doi: 10.5665/sleep.2204.

A Randomized, Double-Blind, Placebo-Controlled Study of an Oral, Extended-Release Formulation of Phentermine/Topiramate for the Treatment of Obstructive Sleep Apnea in Obese Adults

Affiliations
Free PMC article
Randomized Controlled Trial

A Randomized, Double-Blind, Placebo-Controlled Study of an Oral, Extended-Release Formulation of Phentermine/Topiramate for the Treatment of Obstructive Sleep Apnea in Obese Adults

David H Winslow et al. Sleep. .
Free PMC article

Abstract

Study objectives: To evaluate safety and efficacy of phentermine 15 mg plus extended-release topiramate 92 mg for treatment of moderate to severe obstructive sleep apnea (OSA) in obese adults.

Design: This phase 2, randomized, double-blind, placebo-controlled study included 2-week screening and 28-week treatment periods. Overnight polysomnography was performed at baseline, Week 8, and Week 28.

Setting: Single-center study conducted from August 2008 to September 2009.

Participants: Forty-five subjects with moderate to severe OSA not receiving positive airway pressure (PAP) treatment with body mass index of 30-40 kg/m(2).

Interventions: Subjects were randomized to receive placebo (n = 23) or phentermine 15 mg plus extended-release topiramate 92 mg (n = 22). Both groups received lifestyle-modification counseling.

Measurements and results: Primary endpoint, change in apnea-hypopnea index (AHI), significantly favored phentermine 15 mg plus extended-release topiramate 92 mg (-31.5 events/h, 95% CI: -40.0, -22.9) over placebo (-16.6 events/h, 95% CI: -25.0, -8.2) at Week 28 (P =0.0084). At Week 28, there was a 10.2% (95% CI: -12.7, -7.6; 10.8 kg, 95% CI: -13.5, -8.0) mean decrease in weight in the phentermine 15 mg plus extended-release topiramate 92 mg group compared with 4.3% (95% CI: -6.6, -2.0; 4.7 kg, 95% CI: -7.2, -2.2) in the placebo group (P = 0.0006) and a positive, significant (P = 0.0003) correlation between percent change in weight and change in AHI. Significant improvements in overnight oxygen saturation and reduction in blood pressure compared with placebo were observed. Phentermine 15 mg plus extended-release topiramate 92 mg was well tolerated with low adverse event rates.

Conclusions: Phentermine 15 mg plus extended-release topiramate 92 mg induced significant weight reductions and concomitant improvements in OSA and related symptoms vs placebo. This suggests weight loss mediated by phentermine 15 mg plus extended-release topiramate 92 mg may be useful in treatment of moderate to severe OSA in obese subjects unable or unwilling to comply with PAP treatment.

Keywords: Obstructive sleep apnea; apnea-hypopnea index; obesity; phentermine; topiramate.

Figures

Figure 1
Figure 1
Study design. *Visits at which overnight polysomnography was performed.
Figure 2
Figure 2
Subject disposition. AE, adverse event.
Figure 3
Figure 3
Change from baseline in AHI. Data represent mean values at baseline, Week 8 (ITT), and Week 28 (ITT-LOCF). Change (Δ) represents LS mean change from baseline. Week 28 with LOCF is defined as the last available measurement during the double-blind treatment period. N values represent the number of subjects with values at both time points. AHI, apnea-hypopnea index; CI, confidence interval; ITT, intent to treat; LOCF, last observation carried forward; LS, least-squares.*P = 0.0009 vs placebo (1-sided 95% CI: -∞, −8.1); P = 0.0084 vs placebo (1-sided 95% CI: -∞, −4.8).
Figure 4
Figure 4
LS mean percent change in body weight from baseline to Week 8 and Week 28 with LOCF. Week 28 with LOCF is defined as the last available measurement during the double-blind treatment period. N values represent the number of subjects with values at both time points. Data represent LS mean change. LOCF, last observation carried forward; LS, least-squares. *P = 0.0003; P = 0.0006 vs placebo.
Figure 5
Figure 5
(A) Changes in secondary overnight PSG endpoints from baseline to Week 28.(B) Changes in blood oxygen saturation from baseline to Week 28. LOCF analysis. Arousal index represents sleep architecture and quality of sleep (mean number of all cause arousals per hour of sleep). Hypopnea index represents the average number of episodes (≥ 10 sec during which airflow is reduced ≥ 30% and oxygen saturation drops by ≥ 4%, with ≥ 90% of the event's duration meeting the amplitude criteria for apnea) per hour of sleep. Apnea index represents the average number of episodes of apnea (drop in peak thermal sensor excursion by ≥ 90% of baseline for ≥ 10 sec, with ≥ 90% of the event's duration meeting the amplitude criteria for apnea) per hour of sleep. Respiratory disturbance index represents the mean number of apneas, hypopneas, and respiratory effort-related arousals observed per hour of sleep. Desaturation index represents the mean number of events per hour of sleep during which blood oxygen saturation falls below 90%. Week 28 with LOCF is defined as the last available measurement during the double-blind treatment period. N values represent the number of subjects with values at both time points. LOCF, last observation carried forward; LS, least-squares.*P = 0.0064; P = 0.0137; P = 0.0140 vs placebo.
Figure 6
Figure 6
Change in sleep quality indices from baseline to Week 28. Week 28 with LOCF is defined as the last available measurement during the double-blind treatment period. N values represent the number of subjects with values at both time points. Data represent LS mean change. ESS, Epworth Daytime Sleepiness Scale; LOCF, last observation carried forward; LS, least-squares; PSQI, Pittsburgh Sleep Quality Index. *P = 0.0419 vs placebo.
Figure 7
Figure 7
Change in SF-36 (quality of life) scores from baseline to Week 28. Week 28 with LOCF is defined as the last available measurement during the double-blind treatment period. N values represent the number of subjects with values at both time points. Data represent LS mean change. LS, least-squares; SF-36, 36-item Short-Form Health Survey. *P = 0.0103 vs placebo.

Similar articles

See all similar articles

Cited by 18 articles

See all "Cited by" articles

Publication types

MeSH terms

Feedback