- Email: [email protected]
Atomoxetine improves sleepiness and global severity of illness but not the respiratory disturbance index in mild to moderate obstructive sleep apnea with sleepiness
Sleep Medicine 9 (2008) 506–510 www.elsevier.com/locate/sleep
Original Article
Atomoxetine improves sleepiness and global severity of illness but not the respiratory disturbance index in mild to moderate obstructive sleep apnea with sleepiness q R. Bart Sangal *, JoAnne M. Sangal, Krista Thorp Sleep Disorders Institute, 44199 Dequindre, Suite 311, Troy, MI 48085, USA Received 5 December 2006; received in revised form 27 June 2007; accepted 26 July 2007 Available online 27 September 2007
Abstract Background and purpose: Norepinephrine reuptake inhibitors such as protriptyline have been shown to improve sleepiness in sleep apnea, with or without improvement in the respiratory disturbance index (RDI). This study was performed to evaluate whether the selective norepinephrine reuptake inhibitor atomoxetine improves sleepiness, the clinical global impression (CGI) of severity of illness, and the RDI in patients with mild to moderate obstructive sleep apnea with excessive sleepiness. Methods: Patients aged 18–60 years with RDI (including apneas, hypopneas with desaturations and hypopneas with arousals) > 5/h sleep, apnea–hypopnea index (AHI; including apneas, hypopneas with 4% desaturations, but not apneas with arousals) < 15/h sleep, and excessive sleepiness (Epworth Sleepiness Scale [ESS] P 10) received open-label treatment with atomoxetine 40–80 mg HS for 4 weeks, with repeat polysomnography at the end of treatment. Of 20 patients screened, 17 started treatment and 15 completed treatment. Results: ESS improved from 15.3 to 10.5 and CGI improved from 4.3 to 3.1 (both significant at p < 0.01), but there was no significant change in RDI. ESS and CGI improved in a linear fashion across the weeks of treatment. Sleep efficiency and % stage rapid eye movement (REM) sleep were decreased, and % stage 1, awakenings and wake after sleep onset were increased. Conclusions: Atomoxetine improved sleepiness and the CGI in patients with mild to moderate obstructive sleep apnea with sleepiness. However, it did not improve the RDI. Ó 2007 Elsevier B.V. All rights reserved. Keywords: Atomoxetine; Obstructive sleep apnea; Sleepiness; ESS; Norepinephrine reuptake inhibitor; Respiratory disturbance index
1. Introduction Obstructive sleep apnea (OSA) is a significant and common medical disorder. As defined by an apnea– hypopnea index (AHI) > 5/h sleep (including apneas and hypopneas with 4% desaturations, but not hypopneas with arousals), 9% of women and 24% of men in the 30- to 60-year age group have OSA [1], although differq
This study was supported by Eli Lilly & Co. Corresponding author. Tel.: +1 248 879 0707; fax: +1 248 879 2704. E-mail address: [email protected] (R. Bart Sangal). *
1389-9457/$ - see front matter Ó 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.sleep.2007.07.013
ent approaches to defining respiratory disturbances [2], different cut-off indices [3], and whether or not hypersomnolence is required [1] can lead to significant variation in prevalence rates. Mild to moderate OSA, unlike severe OSA, probably has no cardiovascular consequences, but can cause excessive daytime sleepiness. Treatment involves continuous positive airway pressure (CPAP), but compliance is lower in patients with milder sleep apnea [4]. Surgical procedures are available but do not work very well. Compliance issues limit the efficacy of oral mandibular repositioning devices [5]. There is some evidence in the literature [6–9] that biogenic amine reuptake inhibitors, including protriptyline
R. Bart Sangal et al. / Sleep Medicine 9 (2008) 506–510
(a non-selective but predominantly norepinephrine reuptake inhibitor) and fluoxetine (a selective serotonin reuptake inhibitor) improve subjective sleepiness in OSA, with or without an improvement in the respiratory disturbance index (RDI). Brownell et al. [6] treated five patients with severe OSA with 20 mg of the norepinephrine reuptake inhibitor protriptyline in a double-blind crossover study. Four of five patients reported subjective improvement in daytime sleepiness, but there were no significant objective changes in the RDI. Brownell et al. [7] reported that improvement in hypersomnolence persisted long-term, but 5 of 9 patients discontinued protriptyline because of adverse effects (urinary retention, dry mouth, constipation, impotence or loss of interest in sex, or hair loss). Smith et al. [8] studied protriptyline in patients with much more severe OSA. Ten of 12 patients reported subjective improvement in daytime hypersomnolence, again with no significant change in RDI. There was a significant reduction in peak desaturation from 16.2% to 9.2%. Most patients were on 20 mg protriptyline. All patients reported dry mouth, and 5 of 12 patients reported other adverse effects such as constipation, hesitancy while voiding, and difficulty maintaining erection. Hanzel et al. [9] performed a crossover unblinded trial in 12 patients using fluoxetine and protriptyline. Improvement in RDI barely reached significance with each medicine (from 57/h of sleep at baseline to 34/h of sleep with fluoxetine and 33/h of sleep with protriptyline, p = 0.05 for each). All patients tolerated fluoxetine 20 mg, and only 9 of 12 tolerated protriptyline 10 mg. Thus, protriptyline was not well tolerated in these studies. If non-selective norepinephrine reuptake inhibitors such as protriptyline improve sleepiness in OSA, then atomoxetine, a selective norepinephrine reuptake inhibitor shown to be efficacious in the treatment of attention-deficit/hyperactivity disorder [10–13], may also do so, while possibly being better tolerated than protriptyline. The improvement is probably insufficient in comparison to CPAP in patients with severe obstructive sleep apnea/hypopnea syndrome. However, in patients with mild to moderate OSA, being treated because of concomitant excessive sleepiness, improvement in the daytime sleepiness may be quite beneficial, whether or not there is improvement in RDI. Atomoxetine is well-absorbed after oral administration [14]. It is eliminated by oxidative metabolism through the cytochrome P450 2D6 (CYP2D6) enzymatic pathway (to 4-hydroxyatomoxetine) and subsequent glucuronidation. It has a half-life of 5 h, except for 22 h in subjects with reduced CYP2D6 pathway activity (7% of whites and 1–2% of other races). Although 4-hydroxyatomoxetine is pharmacologically active, it circulates in plasma at much lower concentrations (0.1–1% of atomoxetine concentration) so that its actual pharmacological effect is minimal.
507
This study was designed to assess the hypothesis that atomoxetine given at bedtime is useful in the treatment of mild to moderate OSA with hypersomnia. The study objectives were to demonstrate that atomoxetine improved the RDI (apneas, hypopneas with 4% desaturations, and hypopneas with arousals/h sleep), the Epworth Sleepiness Scale (ESS) and the clinical global impression (CGI) of the severity of illness. 2. Methods The study was performed in an American Academy of Sleep Medicine (AASM)-accredited sleep disorders center under the supervision of a physician certified by the American Board of Sleep Medicine (ABSM). The study population consisted of patients of the Sleep Disorders Institute, who had been determined to have mild to moderate OSA on clinically necessary polysomnographic testing (with monitoring of central electroencephalogram [EEG], electro-oculogram [EOG], submental electromyogram [EMG], nasal pressure, respiratory effort by piezo-electric thoracic belt, snoring, electrocardiogram [ECG], leg movements and oximetry), as defined by an RDI (apneas plus hypopneas with arousals or 4% desaturation/h sleep) > 5/h sleep and an AHI (apneas plus hypopneas with 4% desaturation/h sleep) < 15/h sleep; and were sleepy during the day, with ESS of 10 or more, without any other primary sleep disorder. All subjects signed an informed consent form approved by an Institutional Review Board (IRB) before any studyrelated procedures were performed. Other inclusion criteria included being aged 18–60 years and normal electrocardiogram (ECG) at baseline. Patients with documented history of bipolar disorder or psychosis or who met diagnostic criteria for major depression were excluded, as were patients with history of seizures and patients taking any central nervous system (CNS)-active or psychotropic medicines, those with drug or alcohol abuse within 3 months of the study, and patients with narrow angle glaucoma. As this was an exploratory open-label study, there was only one treatment group, and all subjects were treated with active study drug. Visit 1 was a screening visit to determine that patients met all inclusion–exclusion criteria. In addition to a review of the full history and a physical examination, and a review of previous polysomnographic findings by the investigator, patients were administered an ECG, urine was collected for drug testing, and urine was collected from women of child-bearing age for pregnancy testing. If patients met all inclusion/exclusion criteria, they were started on atomoxetine 40 mg HS for the first week. They returned for four weekly visits. At each visit, they were asked about adverse events and concomitant medicines, ESS and CGI (on a scale of 1– 7 with 1 being ‘‘Normal, shows no signs of illness’’ and 7 being ‘‘Among the most extremely ill’’) were per-
508
R. Bart Sangal et al. / Sleep Medicine 9 (2008) 506–510
formed, and dose was adjusted as follows: at the end of 1 week, the dose was increased to 60 mg HS if 40 mg was being tolerated without severe adverse effects; at the end of the second week, the dose was increased to 80 mg HS if 60 mg was being tolerated without severe adverse effects; and at the end of the third week dose was maintained or decreased depending on efficacy and tolerability. At the end of the fourth week, the patients returned for a final visit, and an ECG, physical examination, and polysomnography were repeated. Student’s t-tests for paired samples were performed using data from all subjects who completed the study to compare initial and final ESS, CGI and RDI. Student’s t-tests for paired samples were then performed from all completing subjects to compare initial and final AHI, lowest saturation, sleep latency, persistent sleep latency (latency to 10 min of continuous sleep), number of arousals, number of awakenings, wake after persistent sleep onset (WASO), sleep efficiency, and percentages of stages 1, 2, 3–4, and rapid eye movement (REM) sleep. Analysis of variance with repeated measures was also performed separately for ESS and CGI using the five data points for each subject who completed the study. Twenty subjects signed informed consent. Of these, three were screen failures (two because of marihuana in urine and one because the investigator and patient came to the conclusion that CPAP would be a better mode of treatment in view of concomitant ischemic heart disease with the moderate OSA). Of 17 subjects who met inclusion/exclusion criteria, one started the study, was given medicine, but then was lost to followup before any subsequent visit. One discontinued before Visit 2 due to adverse events. Thus, 15 subjects had fol-
low-up data from at least one visit after starting atomoxetine, and all 15 completed the study. 3. Results The 15 subjects with follow-up data available for analysis comprised 11 males and 4 females, aged 43.9 (standard deviation [SD] 9.4) years. Their baseline and final (after 4 weeks of treatment) ESS, CGI, and polysomnographic variables are given in Table 1. As expected, based on the inclusion criteria, ESS was high though RDI was only mildly to moderately abnormal. Mean final dose of atomoxetine was 76.0 mg (SD = 11.2). Subjects had significantly (p < 0.05) greater WASO, awakenings and % stage 1 sleep, and lesser sleep efficiency and % stage REM on atomoxetine. They had significantly lower ESS and CGI (p < 0.01) on atomoxetine. There were no significant differences with treatment in RDI, AHI or lowest saturation. There were also no significant differences in sleep latency, latency to persistent sleep, REM latency, number of arousals, % stage 2, or % stage 3–4. Analysis of variance with repeated measures using the five ESS scores and the five CGI scores revealed significance (p < 0.001) for each in tests of within-subjects effects, and tests of within-subjects contrasts revealed a linear relationship for both ESS (p < 0.001) and CGI (p = 0.001). Adverse events reported by more than one subject included dry mouth (7 subjects, 4 mild and 3 moderate), insomnia (6 subjects, 5 mild and 1 moderate), constipation (5 subjects, 4 mild and 1 moderate), nausea (5 subjects, 4 mild and 1 moderate), nasal congestion (4
Table 1 Baseline and final clinical and polysomnographic measures Baseline Epworth sleepiness scaleb clinical global Impressionb Sleep Latency (min) Latency to persistent sleep (min) Wake after persistent sleep onset (min)a Awakeningsa Arousals Sleep efficiency (%)a Stage 1 (%)a Stage 2 (%) Stage 3–4 (%) Stage REM (%)a Lowest saturation (%) REM latency (min) Apnea–hypopnea index (/h sleep) Respiratory disturbance index (/h sleep) a b
Baseline different from Final, p < 0.05. Baseline different from Final, p < 0.01.
Final
Mean
SD
Mean
SD
15.3 4.3 19.5 28.5 29.6 10.0 166.7 88.3 16.2 68.3 0.4 15.2 89.7 129.8 6.4 14.2
3.1 0.5 17.2 20.4 18.2 6.0 103.2 8.2 8.8 6.2 1.3 5.8 4.8 69.4 3.8 6.9
10.5 3.1 17.1 30.1 43.1 13.1 150.4 84.5 20.9 70.2 0.7 10.4 88.9 165.3 7.8 13.2
5.2 1.5 14.1 41.4 21.2 8.3 62.9 7.9 12.3 9.2 2.1 6.5 2.7 84.1 10.4 12.2
R. Bart Sangal et al. / Sleep Medicine 9 (2008) 506–510
subjects, 3 mild and 1 moderate), headache (3 subjects, 2 mild and 1 moderate), sore throat (2 subjects, both mild), urinary frequency (2 subjects, 1 mild and 1 moderate), and increased blood pressure (2 subjects, 1 from 125/73 at baseline to a mean of 134/88 over the next four visits, 1 from 126/68 at baseline to a mean of 129/75 over the next 4 visits). There were no significant ECG changes noted. One patient discontinued due to adverse effects (palpitations, headache, nausea and dry mouth) at a dose of 40 mg HS. 4. Discussion Atomoxetine, given at bedtime, significantly decreased ESS and CGI in patients with mild to moderate OSA. However, atomoxetine did not significantly decrease RDI. In other words, patients felt less sleepy, and the clinician perceived global improvement in their symptoms (primarily snoring and sleepiness), but there was no objective evidence of improvement in the frequency of apneic episodes in sleep. This is similar to earlier reports that hypersomnolence improved with protriptyline, but RDI did not [6,8]. It is uncertain what this means. The open-label design is a limitation of this study, and the improvement in ESS and CGI may possibly be a placebo effect on sleepiness and general well-being. On the other hand, previous reports of improvement in sleepiness without improvement in the RDI with another norepinephrine reuptake inhibitor lend credence to this being a real finding. The linear decrease in ESS and CGI over the 4 weeks of treatment also might suggest that this is a real finding, as one might expect a placebo response to be largest initially. We restricted the patient sample to mild to moderate OSA patients. It is possible that this decreased our ability to find an effect of atomoxetine on RDI. Starting out with smaller RDIs by excluding severe OSA patients does restrict the possible decrease in RDI. Larger initial RDIs might have allowed more room for a decrease in RDI. However, given the demonstrated link between severe OSA and cardiovascular morbidity, we could not justify recommending atomoxetine rather than CPAP to severe OSA patients without first demonstrating the effect in mild to moderate OSA patients. We also restricted the patient sample to sleepy patients with higher ESS. It is possible that this yielded an effect of atomoxetine on the ESS that would have been absent if we had included all patients with mild to moderate OSA, without regard to the ESS. However, sleepiness is the reason to treat mild to moderate OSA, and we could not justify subjecting mild to moderate OSA patients without sleepiness to the potential risks of treatment. Interestingly, ESS and CGI decreased despite an increase in WASO, % stage 1 and awakenings, decreased
509
sleep efficiency, and often reported insomnia. Is it worth exchanging an improvement in sleepiness for insomnia? The increase in WASO, % stage 1 and awakenings, and decreased sleep efficiency might suggest a stimulant effect on sleep, decreasing both sleep at night and sleepiness the next day. However, a stimulant effect of a bedtime dose should not last into the next day given the short half-life of atomoxetine. Furthermore, increased WASO, awakenings and % stage 1, with decreased sleep efficiency and the adverse effect of subjective insomnia could just as easily result in increased sleepiness and worse CGI. Therefore, it is not clear that a stimulant effect explains the improvement in ESS and CGI. Atomoxetine given in the morning and evening in children with attention-deficit/hyperactivity disorder decreases awakenings and increases sleep efficiency (predominantly by decreasing sleep latency and latency to persistent sleep onset) [15] arguing against a direct stimulant effect persisting long enough to be evident the next day after a nighttime dose. One way to address the findings reported here is to examine the effects of a morning (or morning and afternoon) dose of atomoxetine on ESS, multiple sleep latency test (MSLT) and/or maintenance of wakefulness test (MWT) in sleepy subjects, such as patients with idiopathic hypersomnia, narcolepsy, mild to moderate OSA, or residual sleepiness in OSA patients treated with CPAP. Perhaps daytime dosing of atomoxetine would improve sleepiness without causing insomnia, since dosing twice a day (the second dose in the evening) is known to decrease awakenings and increase sleep efficiency. One might speculate that inhibiting CNS norepinephrine reuptake in itself improves the sleepiness associated with sleep apnea without affecting the respiratory disturbance index itself. Both methylphenidate [16] (a stimulant) and reboxetine [17,18] (another selective norepinephrine reuptake inhibitor) increase cortical excitability, whereas selective serotonin reuptake inhibitors [19] decrease cortical excitability. The increased cortical excitability with methylphenidate is mediated noradrenergically, not dopaminergically [16], suggesting it is different from a classical stimulant effect. Possibly, therefore, methylphenidate and selective norepinephrine reuptake inhibitors such as reboxetine and atomoxetine may share some cortical effects which may improve sleepiness, separate from methylphenidate’s stimulant effect through dopaminergic pathways. One might also speculate that antidepressants may specifically produce a sense of well-being, even in the absence of depression. Selective norepinephrine reuptake inhibitors such as reboxetine have antidepressant properties [20]. Although atomoxetine has not been shown to have antidepressant efficacy, it may well share an antidepressant effect with reboxetine. Such an increase in the sense of well-being may cause a decreased need for sleep and decreased sleepiness without this
510
R. Bart Sangal et al. / Sleep Medicine 9 (2008) 506–510
effect being in any way related to sleep apnea or its effects. However, this suggestion is not supported by a previous finding of lack of change in mood or well-being scales in healthy subjects given the antidepressant fluoxetine for 6 weeks [21]. It is also possible that sleepiness in mild to moderate OSA is often not really related to the sleep apnea. On an examination of Young et al.’s data [1], approximately 23% of women and 16% of men with AHI> 5/h of sleep self-reported hypersomnolence, and Young et al. attributed this to the sleep apnea in defining a syndrome of sleep apnea with hypersomnolence. However, approximately 11% of women and 3% of men who neither snored nor had AHI> 5/h of sleep also self-reported hypersomnolence. Extrapolating from that, the hypersomnolence of approximately 50% of women and 20% of men with AHI> 5/h of sleep may not be related to sleep apnea. In fact, despite effective treatment of OSA with CPAP, sleepiness may persist and responds to stimulants such as modafinil [22,23], suggesting that the symptom of sleepiness and the polysomnographic finding (or severity) of sleep apnea may be separable and separately treatable. Even in the absence of another primary sleep disorder, sleepiness in a significant subset of patients with mild to moderate sleep apnea may not be related to the sleep apnea or may be separable from the sleep apnea, and norepinephrine reuptake inhibition may somehow improve this sleepiness without any effect on the sleep apnea. It may be important to gather more epidemiological data on sleepiness in the population as well as its association with different definitions and severity levels of OSA.
References [1] Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing in middle-aged adults. N Engl J Med 1993;338:1230–5. [2] Redline S, Kapur VK, Sanders MH, Quan SF, Gottlieb DJ, Rapoport DM, et al. Effects of varying approaches for identifying respiratory disturbances on sleep apnea assessment. Am J Respir Crit Care Med 2000;161:369–74. [3] Manser RL, Rochford P, Pierce RJ, Byrnes GB, Campbell DA. Impact of different criteria for defining hypopneas in the apnea– hypopnea index. Chest 2001;120:909–14. [4] Orth M, Kotterba S, Walther JW, Rasche K, Schultze-Werninghaus G, Duchna HW. Long-term compliance of CPAP therapy – update, predictors and interventions. Pneumologie 2006;60:480–4. [5] Rose E, Staats R, Schulte-Monting J, Ridder GJ, Jonas IE. Long term compliance with an oral protrusive appliance in patients with obstructive sleep apnoea. Dtsch Med Wochenschr 2002;127: 1245–9.
[6] Brownell LG, West P, Sweatman P, Acres JC, Kryger MH. Protriptyline in obstructive sleep apnea: a double-blind trial. New Engl J Med 1982;307:1037–42. [7] Brownell LG, Perez-Padilla R, West P, Kryger MH. The role of protriptyline in obstructive sleep apnea. Bull Eur Physiopath Resp 1983;19:621–4. [8] Smith PL, Haponik EF, Allen RP, Bleecker ER. The effects of protriptyline in sleep-disordered breathing. Am Rev Respir Dis 1983;127:8–13. [9] Hanzel DA, Proia NG, Hudgel DW. Response of obstructive sleep apnea to fluoxetine and protriptyline. Chest 1991;100:416–21. [10] Michelson D, Faries D, Wernicke J, Kelsey D, Kendrick K, Sallee FR, et al. Atomoxetine in the treatment of children and adolescents with attention-deficit/hyperactivity disorder: a randomized, placebo-controlled, dose-response study. Pediatrics 2001;108:e83. [11] Spencer T, Heiligenstein JH, Biederman J, Faries DE, Kratochvil CJ, Conners CK, et al. Results from 2 proof-of-concept, placebocontrolled studies of atomoxetine in children with attentiondeficit/hyperactivity disorder. J Clin Psychiatry 2002;63:1140–7. [12] Michelson D, Allen AJ, Busner J, Casat C, Dunn D, Kratochvil C, et al. Once-daily atomoxetine treatment for children and adolescents with attention-deficit/hyperactivity disorder: a randomized, placebo-controlled study. Am J Psychiatry 2002;159:1896–901. [13] Kelsey DK, Sumner CR, Casat CD, Coury DL, Quintana H, Saylor KE, et al. Once-daily atomoxetine treatment for children with attention-deficit/hyperactivity disorder, including an assessment of evening and morning behavior: a double-blind, placebocontrolled trial. Pediatrics 2004;114:e1–8. [14] Sauer JM, Ring BJ, Witcher JW. Clinical pharmacokinetics of atomoxetine. Clin Pharmacokinet 2005;44:571–90. [15] Sangal RB, Owens J, Allen AJ, Sutton V, Schuh K, Kelsey D. Effects of Atomoxetine and Methylphenidate on sleep of children with ADHD. Sleep 2006;29:1573–85. [16] Andrews GD, Lavin A. Methylphenidate increases cortical excitability via activation of alpha-2 noradrenergic receptors. Neuropsychopharmacology 2006;31:594–601. [17] Herwig U, Brauer K, Connemann B, Spitzer M, SchonfeldtLecuona C. Intracortical excitability is modulated by a norepinephrine-reuptake inhibitor as measured with paired-pulse transcranial magnetic stimulation. Psychopharmacology (Berl) 2002;164:228–32. [18] Plewnia C, Hoppe J, Hiemke C, Bartels M, Cohen LG, Gerloff C. Enhancement of human cortico-motoneuronal excitability by the selective norepinephrine reuptake inhibitor reboxetine. Neurosci Lett 2002;27:231–4. [19] Robol E, Fiaschi A, Manganotti P. Effects of citalopram on the excitability of the human motor cortex: a paired magnetic stimulation study. J Neurol Sci 2004;221:41–6. [20] Scates AC, Doraiswamy PM. Reboxetine: a selective norepinephrine reuptake inhibitor for the treatment of depression. Ann Pharmacother 2000;34:1302–12. [21] Yevgenia Gelfin MD, Malka Gorfine MA, Bernard Lerer MD. Effect of clinical doses of fluoxetine on psychological variables in healthy volunteers. Am J Psychiatry 1998;155:290–2. [22] Pack AI, Black JE, Schwartz JR, Matheson JK. Modafinil as adjunct therapy for daytime sleepiness in obstructive sleep apnea. Am J Respir Crit Care Med 2001;164:1675–81. [23] Schwartz JR, Hirshkowitz M, Erman MK, Scmhidt-Nowara W. Modafinil as adjunct therapy for daytime sleepiness in obstructive sleep apnea: a 12-week, open-label study. Chest 2003;124:1292–9.
Original Article
Atomoxetine improves sleepiness and global severity of illness but not the respiratory disturbance index in mild to moderate obstructive sleep apnea with sleepiness q R. Bart Sangal *, JoAnne M. Sangal, Krista Thorp Sleep Disorders Institute, 44199 Dequindre, Suite 311, Troy, MI 48085, USA Received 5 December 2006; received in revised form 27 June 2007; accepted 26 July 2007 Available online 27 September 2007
Abstract Background and purpose: Norepinephrine reuptake inhibitors such as protriptyline have been shown to improve sleepiness in sleep apnea, with or without improvement in the respiratory disturbance index (RDI). This study was performed to evaluate whether the selective norepinephrine reuptake inhibitor atomoxetine improves sleepiness, the clinical global impression (CGI) of severity of illness, and the RDI in patients with mild to moderate obstructive sleep apnea with excessive sleepiness. Methods: Patients aged 18–60 years with RDI (including apneas, hypopneas with desaturations and hypopneas with arousals) > 5/h sleep, apnea–hypopnea index (AHI; including apneas, hypopneas with 4% desaturations, but not apneas with arousals) < 15/h sleep, and excessive sleepiness (Epworth Sleepiness Scale [ESS] P 10) received open-label treatment with atomoxetine 40–80 mg HS for 4 weeks, with repeat polysomnography at the end of treatment. Of 20 patients screened, 17 started treatment and 15 completed treatment. Results: ESS improved from 15.3 to 10.5 and CGI improved from 4.3 to 3.1 (both significant at p < 0.01), but there was no significant change in RDI. ESS and CGI improved in a linear fashion across the weeks of treatment. Sleep efficiency and % stage rapid eye movement (REM) sleep were decreased, and % stage 1, awakenings and wake after sleep onset were increased. Conclusions: Atomoxetine improved sleepiness and the CGI in patients with mild to moderate obstructive sleep apnea with sleepiness. However, it did not improve the RDI. Ó 2007 Elsevier B.V. All rights reserved. Keywords: Atomoxetine; Obstructive sleep apnea; Sleepiness; ESS; Norepinephrine reuptake inhibitor; Respiratory disturbance index
1. Introduction Obstructive sleep apnea (OSA) is a significant and common medical disorder. As defined by an apnea– hypopnea index (AHI) > 5/h sleep (including apneas and hypopneas with 4% desaturations, but not hypopneas with arousals), 9% of women and 24% of men in the 30- to 60-year age group have OSA [1], although differq
This study was supported by Eli Lilly & Co. Corresponding author. Tel.: +1 248 879 0707; fax: +1 248 879 2704. E-mail address: [email protected] (R. Bart Sangal). *
1389-9457/$ - see front matter Ó 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.sleep.2007.07.013
ent approaches to defining respiratory disturbances [2], different cut-off indices [3], and whether or not hypersomnolence is required [1] can lead to significant variation in prevalence rates. Mild to moderate OSA, unlike severe OSA, probably has no cardiovascular consequences, but can cause excessive daytime sleepiness. Treatment involves continuous positive airway pressure (CPAP), but compliance is lower in patients with milder sleep apnea [4]. Surgical procedures are available but do not work very well. Compliance issues limit the efficacy of oral mandibular repositioning devices [5]. There is some evidence in the literature [6–9] that biogenic amine reuptake inhibitors, including protriptyline
R. Bart Sangal et al. / Sleep Medicine 9 (2008) 506–510
(a non-selective but predominantly norepinephrine reuptake inhibitor) and fluoxetine (a selective serotonin reuptake inhibitor) improve subjective sleepiness in OSA, with or without an improvement in the respiratory disturbance index (RDI). Brownell et al. [6] treated five patients with severe OSA with 20 mg of the norepinephrine reuptake inhibitor protriptyline in a double-blind crossover study. Four of five patients reported subjective improvement in daytime sleepiness, but there were no significant objective changes in the RDI. Brownell et al. [7] reported that improvement in hypersomnolence persisted long-term, but 5 of 9 patients discontinued protriptyline because of adverse effects (urinary retention, dry mouth, constipation, impotence or loss of interest in sex, or hair loss). Smith et al. [8] studied protriptyline in patients with much more severe OSA. Ten of 12 patients reported subjective improvement in daytime hypersomnolence, again with no significant change in RDI. There was a significant reduction in peak desaturation from 16.2% to 9.2%. Most patients were on 20 mg protriptyline. All patients reported dry mouth, and 5 of 12 patients reported other adverse effects such as constipation, hesitancy while voiding, and difficulty maintaining erection. Hanzel et al. [9] performed a crossover unblinded trial in 12 patients using fluoxetine and protriptyline. Improvement in RDI barely reached significance with each medicine (from 57/h of sleep at baseline to 34/h of sleep with fluoxetine and 33/h of sleep with protriptyline, p = 0.05 for each). All patients tolerated fluoxetine 20 mg, and only 9 of 12 tolerated protriptyline 10 mg. Thus, protriptyline was not well tolerated in these studies. If non-selective norepinephrine reuptake inhibitors such as protriptyline improve sleepiness in OSA, then atomoxetine, a selective norepinephrine reuptake inhibitor shown to be efficacious in the treatment of attention-deficit/hyperactivity disorder [10–13], may also do so, while possibly being better tolerated than protriptyline. The improvement is probably insufficient in comparison to CPAP in patients with severe obstructive sleep apnea/hypopnea syndrome. However, in patients with mild to moderate OSA, being treated because of concomitant excessive sleepiness, improvement in the daytime sleepiness may be quite beneficial, whether or not there is improvement in RDI. Atomoxetine is well-absorbed after oral administration [14]. It is eliminated by oxidative metabolism through the cytochrome P450 2D6 (CYP2D6) enzymatic pathway (to 4-hydroxyatomoxetine) and subsequent glucuronidation. It has a half-life of 5 h, except for 22 h in subjects with reduced CYP2D6 pathway activity (7% of whites and 1–2% of other races). Although 4-hydroxyatomoxetine is pharmacologically active, it circulates in plasma at much lower concentrations (0.1–1% of atomoxetine concentration) so that its actual pharmacological effect is minimal.
507
This study was designed to assess the hypothesis that atomoxetine given at bedtime is useful in the treatment of mild to moderate OSA with hypersomnia. The study objectives were to demonstrate that atomoxetine improved the RDI (apneas, hypopneas with 4% desaturations, and hypopneas with arousals/h sleep), the Epworth Sleepiness Scale (ESS) and the clinical global impression (CGI) of the severity of illness. 2. Methods The study was performed in an American Academy of Sleep Medicine (AASM)-accredited sleep disorders center under the supervision of a physician certified by the American Board of Sleep Medicine (ABSM). The study population consisted of patients of the Sleep Disorders Institute, who had been determined to have mild to moderate OSA on clinically necessary polysomnographic testing (with monitoring of central electroencephalogram [EEG], electro-oculogram [EOG], submental electromyogram [EMG], nasal pressure, respiratory effort by piezo-electric thoracic belt, snoring, electrocardiogram [ECG], leg movements and oximetry), as defined by an RDI (apneas plus hypopneas with arousals or 4% desaturation/h sleep) > 5/h sleep and an AHI (apneas plus hypopneas with 4% desaturation/h sleep) < 15/h sleep; and were sleepy during the day, with ESS of 10 or more, without any other primary sleep disorder. All subjects signed an informed consent form approved by an Institutional Review Board (IRB) before any studyrelated procedures were performed. Other inclusion criteria included being aged 18–60 years and normal electrocardiogram (ECG) at baseline. Patients with documented history of bipolar disorder or psychosis or who met diagnostic criteria for major depression were excluded, as were patients with history of seizures and patients taking any central nervous system (CNS)-active or psychotropic medicines, those with drug or alcohol abuse within 3 months of the study, and patients with narrow angle glaucoma. As this was an exploratory open-label study, there was only one treatment group, and all subjects were treated with active study drug. Visit 1 was a screening visit to determine that patients met all inclusion–exclusion criteria. In addition to a review of the full history and a physical examination, and a review of previous polysomnographic findings by the investigator, patients were administered an ECG, urine was collected for drug testing, and urine was collected from women of child-bearing age for pregnancy testing. If patients met all inclusion/exclusion criteria, they were started on atomoxetine 40 mg HS for the first week. They returned for four weekly visits. At each visit, they were asked about adverse events and concomitant medicines, ESS and CGI (on a scale of 1– 7 with 1 being ‘‘Normal, shows no signs of illness’’ and 7 being ‘‘Among the most extremely ill’’) were per-
508
R. Bart Sangal et al. / Sleep Medicine 9 (2008) 506–510
formed, and dose was adjusted as follows: at the end of 1 week, the dose was increased to 60 mg HS if 40 mg was being tolerated without severe adverse effects; at the end of the second week, the dose was increased to 80 mg HS if 60 mg was being tolerated without severe adverse effects; and at the end of the third week dose was maintained or decreased depending on efficacy and tolerability. At the end of the fourth week, the patients returned for a final visit, and an ECG, physical examination, and polysomnography were repeated. Student’s t-tests for paired samples were performed using data from all subjects who completed the study to compare initial and final ESS, CGI and RDI. Student’s t-tests for paired samples were then performed from all completing subjects to compare initial and final AHI, lowest saturation, sleep latency, persistent sleep latency (latency to 10 min of continuous sleep), number of arousals, number of awakenings, wake after persistent sleep onset (WASO), sleep efficiency, and percentages of stages 1, 2, 3–4, and rapid eye movement (REM) sleep. Analysis of variance with repeated measures was also performed separately for ESS and CGI using the five data points for each subject who completed the study. Twenty subjects signed informed consent. Of these, three were screen failures (two because of marihuana in urine and one because the investigator and patient came to the conclusion that CPAP would be a better mode of treatment in view of concomitant ischemic heart disease with the moderate OSA). Of 17 subjects who met inclusion/exclusion criteria, one started the study, was given medicine, but then was lost to followup before any subsequent visit. One discontinued before Visit 2 due to adverse events. Thus, 15 subjects had fol-
low-up data from at least one visit after starting atomoxetine, and all 15 completed the study. 3. Results The 15 subjects with follow-up data available for analysis comprised 11 males and 4 females, aged 43.9 (standard deviation [SD] 9.4) years. Their baseline and final (after 4 weeks of treatment) ESS, CGI, and polysomnographic variables are given in Table 1. As expected, based on the inclusion criteria, ESS was high though RDI was only mildly to moderately abnormal. Mean final dose of atomoxetine was 76.0 mg (SD = 11.2). Subjects had significantly (p < 0.05) greater WASO, awakenings and % stage 1 sleep, and lesser sleep efficiency and % stage REM on atomoxetine. They had significantly lower ESS and CGI (p < 0.01) on atomoxetine. There were no significant differences with treatment in RDI, AHI or lowest saturation. There were also no significant differences in sleep latency, latency to persistent sleep, REM latency, number of arousals, % stage 2, or % stage 3–4. Analysis of variance with repeated measures using the five ESS scores and the five CGI scores revealed significance (p < 0.001) for each in tests of within-subjects effects, and tests of within-subjects contrasts revealed a linear relationship for both ESS (p < 0.001) and CGI (p = 0.001). Adverse events reported by more than one subject included dry mouth (7 subjects, 4 mild and 3 moderate), insomnia (6 subjects, 5 mild and 1 moderate), constipation (5 subjects, 4 mild and 1 moderate), nausea (5 subjects, 4 mild and 1 moderate), nasal congestion (4
Table 1 Baseline and final clinical and polysomnographic measures Baseline Epworth sleepiness scaleb clinical global Impressionb Sleep Latency (min) Latency to persistent sleep (min) Wake after persistent sleep onset (min)a Awakeningsa Arousals Sleep efficiency (%)a Stage 1 (%)a Stage 2 (%) Stage 3–4 (%) Stage REM (%)a Lowest saturation (%) REM latency (min) Apnea–hypopnea index (/h sleep) Respiratory disturbance index (/h sleep) a b
Baseline different from Final, p < 0.05. Baseline different from Final, p < 0.01.
Final
Mean
SD
Mean
SD
15.3 4.3 19.5 28.5 29.6 10.0 166.7 88.3 16.2 68.3 0.4 15.2 89.7 129.8 6.4 14.2
3.1 0.5 17.2 20.4 18.2 6.0 103.2 8.2 8.8 6.2 1.3 5.8 4.8 69.4 3.8 6.9
10.5 3.1 17.1 30.1 43.1 13.1 150.4 84.5 20.9 70.2 0.7 10.4 88.9 165.3 7.8 13.2
5.2 1.5 14.1 41.4 21.2 8.3 62.9 7.9 12.3 9.2 2.1 6.5 2.7 84.1 10.4 12.2
R. Bart Sangal et al. / Sleep Medicine 9 (2008) 506–510
subjects, 3 mild and 1 moderate), headache (3 subjects, 2 mild and 1 moderate), sore throat (2 subjects, both mild), urinary frequency (2 subjects, 1 mild and 1 moderate), and increased blood pressure (2 subjects, 1 from 125/73 at baseline to a mean of 134/88 over the next four visits, 1 from 126/68 at baseline to a mean of 129/75 over the next 4 visits). There were no significant ECG changes noted. One patient discontinued due to adverse effects (palpitations, headache, nausea and dry mouth) at a dose of 40 mg HS. 4. Discussion Atomoxetine, given at bedtime, significantly decreased ESS and CGI in patients with mild to moderate OSA. However, atomoxetine did not significantly decrease RDI. In other words, patients felt less sleepy, and the clinician perceived global improvement in their symptoms (primarily snoring and sleepiness), but there was no objective evidence of improvement in the frequency of apneic episodes in sleep. This is similar to earlier reports that hypersomnolence improved with protriptyline, but RDI did not [6,8]. It is uncertain what this means. The open-label design is a limitation of this study, and the improvement in ESS and CGI may possibly be a placebo effect on sleepiness and general well-being. On the other hand, previous reports of improvement in sleepiness without improvement in the RDI with another norepinephrine reuptake inhibitor lend credence to this being a real finding. The linear decrease in ESS and CGI over the 4 weeks of treatment also might suggest that this is a real finding, as one might expect a placebo response to be largest initially. We restricted the patient sample to mild to moderate OSA patients. It is possible that this decreased our ability to find an effect of atomoxetine on RDI. Starting out with smaller RDIs by excluding severe OSA patients does restrict the possible decrease in RDI. Larger initial RDIs might have allowed more room for a decrease in RDI. However, given the demonstrated link between severe OSA and cardiovascular morbidity, we could not justify recommending atomoxetine rather than CPAP to severe OSA patients without first demonstrating the effect in mild to moderate OSA patients. We also restricted the patient sample to sleepy patients with higher ESS. It is possible that this yielded an effect of atomoxetine on the ESS that would have been absent if we had included all patients with mild to moderate OSA, without regard to the ESS. However, sleepiness is the reason to treat mild to moderate OSA, and we could not justify subjecting mild to moderate OSA patients without sleepiness to the potential risks of treatment. Interestingly, ESS and CGI decreased despite an increase in WASO, % stage 1 and awakenings, decreased
509
sleep efficiency, and often reported insomnia. Is it worth exchanging an improvement in sleepiness for insomnia? The increase in WASO, % stage 1 and awakenings, and decreased sleep efficiency might suggest a stimulant effect on sleep, decreasing both sleep at night and sleepiness the next day. However, a stimulant effect of a bedtime dose should not last into the next day given the short half-life of atomoxetine. Furthermore, increased WASO, awakenings and % stage 1, with decreased sleep efficiency and the adverse effect of subjective insomnia could just as easily result in increased sleepiness and worse CGI. Therefore, it is not clear that a stimulant effect explains the improvement in ESS and CGI. Atomoxetine given in the morning and evening in children with attention-deficit/hyperactivity disorder decreases awakenings and increases sleep efficiency (predominantly by decreasing sleep latency and latency to persistent sleep onset) [15] arguing against a direct stimulant effect persisting long enough to be evident the next day after a nighttime dose. One way to address the findings reported here is to examine the effects of a morning (or morning and afternoon) dose of atomoxetine on ESS, multiple sleep latency test (MSLT) and/or maintenance of wakefulness test (MWT) in sleepy subjects, such as patients with idiopathic hypersomnia, narcolepsy, mild to moderate OSA, or residual sleepiness in OSA patients treated with CPAP. Perhaps daytime dosing of atomoxetine would improve sleepiness without causing insomnia, since dosing twice a day (the second dose in the evening) is known to decrease awakenings and increase sleep efficiency. One might speculate that inhibiting CNS norepinephrine reuptake in itself improves the sleepiness associated with sleep apnea without affecting the respiratory disturbance index itself. Both methylphenidate [16] (a stimulant) and reboxetine [17,18] (another selective norepinephrine reuptake inhibitor) increase cortical excitability, whereas selective serotonin reuptake inhibitors [19] decrease cortical excitability. The increased cortical excitability with methylphenidate is mediated noradrenergically, not dopaminergically [16], suggesting it is different from a classical stimulant effect. Possibly, therefore, methylphenidate and selective norepinephrine reuptake inhibitors such as reboxetine and atomoxetine may share some cortical effects which may improve sleepiness, separate from methylphenidate’s stimulant effect through dopaminergic pathways. One might also speculate that antidepressants may specifically produce a sense of well-being, even in the absence of depression. Selective norepinephrine reuptake inhibitors such as reboxetine have antidepressant properties [20]. Although atomoxetine has not been shown to have antidepressant efficacy, it may well share an antidepressant effect with reboxetine. Such an increase in the sense of well-being may cause a decreased need for sleep and decreased sleepiness without this
510
R. Bart Sangal et al. / Sleep Medicine 9 (2008) 506–510
effect being in any way related to sleep apnea or its effects. However, this suggestion is not supported by a previous finding of lack of change in mood or well-being scales in healthy subjects given the antidepressant fluoxetine for 6 weeks [21]. It is also possible that sleepiness in mild to moderate OSA is often not really related to the sleep apnea. On an examination of Young et al.’s data [1], approximately 23% of women and 16% of men with AHI> 5/h of sleep self-reported hypersomnolence, and Young et al. attributed this to the sleep apnea in defining a syndrome of sleep apnea with hypersomnolence. However, approximately 11% of women and 3% of men who neither snored nor had AHI> 5/h of sleep also self-reported hypersomnolence. Extrapolating from that, the hypersomnolence of approximately 50% of women and 20% of men with AHI> 5/h of sleep may not be related to sleep apnea. In fact, despite effective treatment of OSA with CPAP, sleepiness may persist and responds to stimulants such as modafinil [22,23], suggesting that the symptom of sleepiness and the polysomnographic finding (or severity) of sleep apnea may be separable and separately treatable. Even in the absence of another primary sleep disorder, sleepiness in a significant subset of patients with mild to moderate sleep apnea may not be related to the sleep apnea or may be separable from the sleep apnea, and norepinephrine reuptake inhibition may somehow improve this sleepiness without any effect on the sleep apnea. It may be important to gather more epidemiological data on sleepiness in the population as well as its association with different definitions and severity levels of OSA.
References [1] Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing in middle-aged adults. N Engl J Med 1993;338:1230–5. [2] Redline S, Kapur VK, Sanders MH, Quan SF, Gottlieb DJ, Rapoport DM, et al. Effects of varying approaches for identifying respiratory disturbances on sleep apnea assessment. Am J Respir Crit Care Med 2000;161:369–74. [3] Manser RL, Rochford P, Pierce RJ, Byrnes GB, Campbell DA. Impact of different criteria for defining hypopneas in the apnea– hypopnea index. Chest 2001;120:909–14. [4] Orth M, Kotterba S, Walther JW, Rasche K, Schultze-Werninghaus G, Duchna HW. Long-term compliance of CPAP therapy – update, predictors and interventions. Pneumologie 2006;60:480–4. [5] Rose E, Staats R, Schulte-Monting J, Ridder GJ, Jonas IE. Long term compliance with an oral protrusive appliance in patients with obstructive sleep apnoea. Dtsch Med Wochenschr 2002;127: 1245–9.
[6] Brownell LG, West P, Sweatman P, Acres JC, Kryger MH. Protriptyline in obstructive sleep apnea: a double-blind trial. New Engl J Med 1982;307:1037–42. [7] Brownell LG, Perez-Padilla R, West P, Kryger MH. The role of protriptyline in obstructive sleep apnea. Bull Eur Physiopath Resp 1983;19:621–4. [8] Smith PL, Haponik EF, Allen RP, Bleecker ER. The effects of protriptyline in sleep-disordered breathing. Am Rev Respir Dis 1983;127:8–13. [9] Hanzel DA, Proia NG, Hudgel DW. Response of obstructive sleep apnea to fluoxetine and protriptyline. Chest 1991;100:416–21. [10] Michelson D, Faries D, Wernicke J, Kelsey D, Kendrick K, Sallee FR, et al. Atomoxetine in the treatment of children and adolescents with attention-deficit/hyperactivity disorder: a randomized, placebo-controlled, dose-response study. Pediatrics 2001;108:e83. [11] Spencer T, Heiligenstein JH, Biederman J, Faries DE, Kratochvil CJ, Conners CK, et al. Results from 2 proof-of-concept, placebocontrolled studies of atomoxetine in children with attentiondeficit/hyperactivity disorder. J Clin Psychiatry 2002;63:1140–7. [12] Michelson D, Allen AJ, Busner J, Casat C, Dunn D, Kratochvil C, et al. Once-daily atomoxetine treatment for children and adolescents with attention-deficit/hyperactivity disorder: a randomized, placebo-controlled study. Am J Psychiatry 2002;159:1896–901. [13] Kelsey DK, Sumner CR, Casat CD, Coury DL, Quintana H, Saylor KE, et al. Once-daily atomoxetine treatment for children with attention-deficit/hyperactivity disorder, including an assessment of evening and morning behavior: a double-blind, placebocontrolled trial. Pediatrics 2004;114:e1–8. [14] Sauer JM, Ring BJ, Witcher JW. Clinical pharmacokinetics of atomoxetine. Clin Pharmacokinet 2005;44:571–90. [15] Sangal RB, Owens J, Allen AJ, Sutton V, Schuh K, Kelsey D. Effects of Atomoxetine and Methylphenidate on sleep of children with ADHD. Sleep 2006;29:1573–85. [16] Andrews GD, Lavin A. Methylphenidate increases cortical excitability via activation of alpha-2 noradrenergic receptors. Neuropsychopharmacology 2006;31:594–601. [17] Herwig U, Brauer K, Connemann B, Spitzer M, SchonfeldtLecuona C. Intracortical excitability is modulated by a norepinephrine-reuptake inhibitor as measured with paired-pulse transcranial magnetic stimulation. Psychopharmacology (Berl) 2002;164:228–32. [18] Plewnia C, Hoppe J, Hiemke C, Bartels M, Cohen LG, Gerloff C. Enhancement of human cortico-motoneuronal excitability by the selective norepinephrine reuptake inhibitor reboxetine. Neurosci Lett 2002;27:231–4. [19] Robol E, Fiaschi A, Manganotti P. Effects of citalopram on the excitability of the human motor cortex: a paired magnetic stimulation study. J Neurol Sci 2004;221:41–6. [20] Scates AC, Doraiswamy PM. Reboxetine: a selective norepinephrine reuptake inhibitor for the treatment of depression. Ann Pharmacother 2000;34:1302–12. [21] Yevgenia Gelfin MD, Malka Gorfine MA, Bernard Lerer MD. Effect of clinical doses of fluoxetine on psychological variables in healthy volunteers. Am J Psychiatry 1998;155:290–2. [22] Pack AI, Black JE, Schwartz JR, Matheson JK. Modafinil as adjunct therapy for daytime sleepiness in obstructive sleep apnea. Am J Respir Crit Care Med 2001;164:1675–81. [23] Schwartz JR, Hirshkowitz M, Erman MK, Scmhidt-Nowara W. Modafinil as adjunct therapy for daytime sleepiness in obstructive sleep apnea: a 12-week, open-label study. Chest 2003;124:1292–9.