1. Thinking About Sleep in Epilepsy Trials: Interrelationships, Pitfalls, and Novel Opportunities Beth A. Malow, MD, MS Associate Professor of Neurology and Medical Director, Vanderbilt Sleep Disorders Center Director, Vanderbilt Sleep Research Core No conflict of interest. Will be discussing unlabeled use of commercial products.

2. Presentation Goals Identify the causes of disordered sleep in patients with epilepsy (e.g., sleep apnea) Highlight the importance of these causes in designing therapeutic trials Discuss indications and options for measuring sleep and sleepiness in clinical studies Characterize the effects of antiepileptic drugs on sleep, and their use in patients with and without epilepsy Present preliminary data from an NINDS pilot trial on the impact of treating obstructive sleep apnea in epilepsy

3. Impact of Sleep on Epilepsy Sleep disruption/deprivation, either voluntary or involuntary, contribute to impaired daytime functioning and overall health in many individuals. Patients with epilepsy may be more likely to have disturbed sleep. In addition, sleep disruption/deprivation may affect seizure control, or a patient’s ability to tolerate antiepileptic drugs. Daytime alertness and vigilance may also be affected.

4. Epilepsy Affects Sleep Organization On seizure free nights, sleep organization was affected in 80 epilepsy patients as compared to 17 normal controls Decreased sleep efficiency (time asleep/time in bed) Increase in sleep stage shifts Increase in number and duration of awakenings These parameters were most affected in untreated, newly diagnosed patients Treatment with carbamazepine for 1 month improved these parameters Touchon J et al. Sleep organization and epilepsy. In Degen R and Rodin EA, Epilepsy, Sleep, and Sleep Deprivation, 1991.

5. Epilepsy Affects Sleep Organization Temporal lobe epilepsy disrupts sleep organization more than frontal lobe epilepsy and idiopathic generalized epilepsy, even in the absence of seizures Touchon J et al. Sleep organization and epilepsy. In Degen R and Rodin EA, Epilepsy, Sleep, and Sleep Deprivation, 1991. Crespel A et al., Epilepsia 1998;39:150-157. Montplaisir J et al., J Clin Neurophysiol 1987;4:383-388. Seizures themselves have profound effects on sleep architecture, even apart from the resulting arousals and awakenings

6. Adverse Effects of Antiepileptic Drugs Somnolence51/144 (35%) Gingival Hyperplasia34/144 (24%) Nystagmus23/144 (16%) Ataxia16/144 (11%) Irritability10/144 (7%) Vertigo10/144 (7%) AEs were patient-reported and AEDs included PHT, PB, CBZ, VPA Beghi E et al, Epilepsia 27(4):323-330, 1986.

7. Antiepileptic Drugs (AEDs) and Sleep Are AEDs that reduce sleep stage shifts, arousals, and awakenings “better” than those that do not? Might these effects translate into improved daytime alertness, health-related quality of life, or even reduced seizure frequency? Does improving (or not suppressing) REM or slow wave sleep affect cognition? Do these questions transcend the field of epilepsy (e.g., individuals without epilepsy desiring sleep aids)?

8. Antiepileptic Drugs (AEDs) and Sleep Pregabalin received FDA approval for treating for neuropathic pain (and as adjunctive treatment for partial seizures in adults) In one study presented in abstract form, with data drawn from 13 randomized controlled studies of chronic pain syndromes, pregabalin significantly improved sleep components on the Medical Outcomes Study (MOS) Sleep Scale Freeman R et al, Abstract P04.010, AAN 2006 Meeting, San Diego Although not approved for this indication, sleep specialists commonly use pregabalin or gabapentin off-label for treating sleep in pain syndromes, including fibromyalgia We also commonly use pregabalin or gabapentin off-label as a sleep aid, in patients with and without epilepsy

9. Antiepileptic Drugs (AEDs) and Sleep Topiramate has received FDA approval for the prophylaxis of migraine headaches (and for treatment of seizures) Topiramate may also be effective in treating sleep-related eating disorder or night eating syndrome Winkelman JW et al, 2003, Sleep Med 4:243-6. How about considering topiramate for treating obese patients with insomnia, who also have migraine? What about trying topiramate for insomnia in obese patients, even if they don’t have migraine? What constitutes obesity? What constitutes insomnia?

10. Obstructive Sleep Apnea  One of the most common sleep disorders, with as many as 24% of men and 9% of women in the general population affected (4% of men and 2% of women with daytime sleepiness) 1  93% of women and 82% of men with moderate to severe sleep apnea are undiagnosed 2  Sleep apnea is a risk factor for a myriad of medical conditions, including hypertension, heart disease, stroke, and impaired glucose tolerance  Sleep apnea also contributes to daytime sleepiness and impairs health-related quality of life  Treated with continuous positive airway pressure (CPAP), which splints airway open 1 Young T, et al, NEJM 1993; 328:1230-5. 2 Young T et al, SLEEP 1997;20:705-706.

11. Epilepsy and OSA: Prior Studies Sleep apnea may coexist with epilepsy; its treatment may improve seizure control, daytime sleepiness, or both  Tracheostomy diminished generalized seizures in 1 patient 1  CPAP or other therapy improved seizure control and daytime alertness in 6 of 7 patients with partial seizures 2  CPAP or positional therapy improved seizure control in 7 of 10 patients with seizures and OSA; 3 had antiepileptic medications optimized 3  CPAP improved seizure control or daytime sleepiness in 7 of 9 adults. 4 Prospective study showed 45% reduction in seizure frequency in 3 adults and 1 child treated with CPAP. 5 1. Wyler AR, Weymuller EA Jr. Ann Neurol. 1981;9(4):403-404. 2. Devinsky O, et al. Neurology. 1994;44(11):2060-2064. 3. Vaughn BV, et al. Seizure. 1996;5(1):73-78. 4. Malow BA, et al. Sleep. 1997;20(12):1105-1110. 5. Malow BA, et al. Sleep Medicine. 2003;4:209-215.

12. How Might OSA Facilitate Seizures? Proposed mechanisms include:  Sleep deprivation resulting from frequent arousals from sleep, increasing neuronal excitability  Frequent arousals or stage shifts into and out of sleep, facilitating sleep-related seizures  Cytokines? (increased by sleep apnea, facilitators of seizures)  Apneas and hypopneas causing seizures  Episodes of hypoxemia  Decreased cardiac output/arrhythmias Less Likely

13. Is OSA More Common in Patients with Epilepsy?  One third of 39 epilepsy surgery candidates had OSA (AHI >5) and 13% had AHIs >20 1  60% of intractable epilepsy patients (>4 seizures/ month) had AHIs of 5 or greater, and 36% had AHIs of 10 or greater 2  Possible reasons OSA may be more common: Sedentary lifestyles? Weight gain from antiepileptic drugs (AEDs)? Effects of AEDs on the upper airway? 1. Malow BA, et al. Neurology. 2000;55(7):1002-1007. 2. Vaughn BV, D'Cruz OF. Epilepsia. 2002;42(suppl 7):43.

14. Presentation Goals Identify the causes of disordered sleep in patients with epilepsy (e.g., sleep apnea) Highlight the importance of these causes in designing therapeutic trials Characterize the effects of antiepileptic drugs on sleep, and their use in patients with and without epilepsy Discuss indications and options for measuring sleep and sleepiness in therapeutic trials Present preliminary data from an NINDS pilot trial on the impact of treating obstructive sleep apnea in epilepsy

15. Measuring Sleep and Related Variables in Therapeutic Trials: Indications and Options Indications include: Screening for sleep disorders in patients with epilepsy being recruited for clinical trials to exclude or at least take these disorders into account in analyses Measuring the impact of AEDs and other therapies on sleep and daytime alertness Options range from questionnaires, to sleep diaries, to actigraphy, to polysomnography and daytime sleep studies.

16. Measuring Sleep and Related Variables in Therapeutic Trials Screening for sleep disorders in patients with epilepsy being recruited for clinical trials Is polysomnography necessary? Are questionnaires sufficient? What constitutes a sleep disorder? In the non-epilepsy population without medical illnesses, loud snoring and daytime sleepiness are highly predictive of OSA Screening for OSA in patients with epilepsy may be more challenging. Patient may lack a bedpartner, seizures may be paramount concern, and daytime sleepiness may be due to seizures or AEDs

17. Can we effectively screen patients with epilepsy for OSA (without PSGs)? We validated a 12-item scale developed for general population, the Sleep Apnea scale of the Sleep Disorders Questionnaire (SA-SDQ) in 125 patients with epilepsy undergoing PSG. 1 SA-SDQ asks questions about age, body mass index, gender, presence of hypertension, loud snoring, witnessed apnea, and other parameters. Positive predictive value: 0.75 in men and 0.8 in women Negative predictive value: 0.65 in men and 0.67 in women 1 Weatherwax, KJ, et al. Sleep Medicine 2003;4: 517-521. Other screening questionnaires have shown similar results. Adding a sleep evaluation may also help.

18. Measuring Daytime Alertness Several sleepiness scales exist, including the Epworth Sleepiness Scale The Epworth Sleepiness Scale asks questions about how likely the individual is to doze in a variety of situations (eg., watching TV; sitting and talking to someone) The Multiple Sleep Latency Test (MSLT) provides for several naps, spaced out throughout the day, and measures sleep latency and REM latency using sleep recordings. The Maintenance of Wakefulness Test (MWT) is similar to the MSLT but instructs the individual to stay awake, rather than fall asleep. A variety of psychomotor vigilance tasks exist as well Subjective and objective tests may not correlate and are viewed by some as complementary

19. NINDS Pilot Clinical Trial We carried out an NINDS-supported pilot clinical trial to work out critical design issues prior to embarking on a definitive Phase III Randomized Clinical Trial that will answer the question: Does treatment of coexisting obstructive sleep apnea in epilepsy patients improve seizure frequency, daytime sleepiness, and health- related quality of life? Sites: Vanderbilt, Michigan, Cleveland Clinic, UNC-Chapel Hill

20. Aim 1: Can we effectively identify patients with epilepsy who have OSA prior to PSG?  Of 43 patients with epilepsy meeting inclusion criteria for likely OSA (by SA-SDQ score and clinical impression), 35 had OSA on polysomnography, defined by an apnea-hypopnea index of 5 or greater.  This translates into an 81% true positive rate or only one in 5 screen failures  As we only tested patients with suspected OSA, the proportion of patients with epilepsy and OSA who we missed in the screening is unknown

21. Aim 2: Is more than one night of PSG needed to determine if an epilepsy patient has OSA?  Sleep research is affected by a first night effect, with the first night of sleep in a laboratory affected by the novelty of the laboratory, wires, etc.  We compared night 1 and 2 of PSG in our cohort  Other than decreased total sleep time and REM sleep on night 1, sleep parameters, including sleep latency, efficiency, stages, arousals, AHI and oxygen desaturation were comparable.  Using our AHI inclusion criterion of 5 or greater, the first PSG identified all but one patient with OSA – night 1, AHI 3.0, night 2, 5.8.  One night of PSG should be sufficient for diagnosing OSA in patients with epilepsy

22. Aim 3: Will subjects use CPAP?  Differences in CPAP adherence in the therapeutic and sham CPAP groups were not statistically significant  Percent of nights with CPAP device usage for our sample as a whole was 71.8%, with the average number of hours of CPAP usage per night on nights used being 4.4 hours.  Older subjects (ages 45 or older) used their CPAP for a greater number of hours/night than younger subjects (5.4 hours vs. 3.4 hours; p=0.01).  Subjects, Coordinators, and Principal Investigators could not tell whether therapeutic or sham CPAP was used (kappa values < 0.4)

23. What are important outcome measures for patients with epilepsy and OSA?  Is seizure frequency the most important outcome?  Other outcome measures to consider are: Daytime sleepiness Insomnia Health-related quality of life Ability to tolerate AEDs if OSA treated Improvement in obesity Minimizing cardiovascular/cerebrovascular consequences of OSA, especially in older adults with epilepsy

24. What would a definitive trial of treating OSA in epilepsy accomplish? Heighten the awareness of identifying and treating OSA in epilepsy among practitioners, in that the vast majority of patients with OSA are still not diagnosed and treated, even by neurologists Epilepsy patients may be more at risk for OSA, and have more to gain from treatment Epilepsy patients tend to put seizures in the forefront, and may not be attuned to possibility that OSA may be contributing to daytime sleepiness, AED intolerance, and impaired quality of life Patients with epilepsy may be more likely to use their CPAP than those in the general population if they perceive a benefit in seizures or daytime sleepiness

25. Conclusions Sleep is disturbed in patients with epilepsy due to a variety of causes, including seizures, AEDs, and sleep disorders Various options exist for measuring sleep and related parameters in patients with epilepsy, although further research will be needed to determine optimal use of these measurement tools In designing therapeutic trials in epilepsy, consideration of sleep disturbances presents challenges, but also opportunities to improve seizure control and other aspects of quality of life