Sleep Medicine Pearls 2nd

PATIENT 73 A 55-year-old man with central sleep apnea A 55-year-old man with complaints of frequent awakenings and moderate daytime sleepiness was evaluated by polysomnography. He snored occasionally, and his wife had noted some brief pauses in his breathing during the night. There was no history of leg jerks or symptoms of congestive heart failure. The patient's only medication was an angiotensin-converting enzyme inhibitor for hypertension. Physical Examination: Blood pressure 150/85 mmHg, pulse 80 and regular. HEENT: dependent palate. Cardiac: normal. Chest: clear, no rales. Extremities: no edema. Sleep Study: AHI 30/hr. Respiratory events: obstructive apnea 5%, mixed apnea 15%, central apnea 75%, hypopnea 5%. Figure: A sample tracing is shown below. Question: What treatment do you recommend? 15 sec A 90% 239

Answer: Nasal CPAP is effective treatment in some patients with idiopathic central sleep apnea. Discussion: There is no uniform consensus about the best treatment for patients with idiopathic CSA (central sleep apnea). This group is heterogeneous, and treatment must be individualized. Because idiopathic CSA is relatively rare, no longterm studies of the effectiveness of any treatment have been published. Treatments ofIdiopathic CSA Nasal CPAP Triazolam Acetazolamide Oxygen therapy (selected patients) Triazolam, a benzodiazepine, reduced the frequency of idiopathic CSA in one study, probably by decreasing the number of arousals or the amount of hyperventilation associated with arousal. Obviously, sedatives are contraindicated in the hypercapnic forms of CSA. Nasal CPAP also has been reported to decrease central apnea in patients with idiopathic CSA. The mechanisms by which CPAP works are unknown. Two possibilities are that nasal CPAP slightly increases the sleeping PC02 or prevents upper airway reflexes from initiating apnea. The level of CPAP required to prevent central apnea may exceed the level necessary to prevent obstructive events. Patients who snore or have central apnea mainly in the supine position might be assumed to be the best candidates for CPAP treatment. Of note, one case report found that CP AP helped but bilevel pressure worsened central sleep apnea. The authors hypothesized that CP AP worked by inducing a modest increase in PC02 during sleep. As noted in Patient 72, the addition of dead space or inhaled CO2 has been shown to decrease central apnea in patients with idiopathic CSA. In patients with OSA, central apneas can sometimes appear on CPAP (usually post arousal) or when CPAP induces repetitive arousals from sleep (levels of CPAP presumably too high). Various respiratory stimulants have been tried as treatments for idiopathic CSA, with variable amounts of success. The best evidence is for use of acetazolamide (Diamox), which is a carbonic anhydrase inhibitor. Acetazolamide induces a metabolic acidosis and reduces the pH even if the PC02 also decreases slightly. One study found modest success with a dose of 250 mg given I hour before bedtime: the AHI was reduced by approximately 50% and symptoms improved, although sleep efficiency was not significantly better. Oxygen therapy also has been reported to decrease the amounts of nocturnal desaturation and central apnea in selected patients. In the present patient, apnea was noted to be predominantly central. In the tracing, the small oscillations at A represent movement from cardiac contractions. He snored, and therefore treatment with CPAP was tried. The patient underwent a CP AP titration, and at 10 em HoO the AHI was reduced to 8/hr. Treatment with nasal CPAP resulted in improvement of symptoms. Clinical Pearls I. Idiopathic central sleep apnea is rare, present in 5% or less of patients with sleep apnea. 2. No consensus exists about the best treatment for idiopathic CSA. Treatment must be individualized. 3. Nasal CPAP may be effective in some patients with idiopathic CSA. 4. Alternative treatments include acetazolamide, hypnotics, and oxygen therapy. REFERENCES I. Issa FG. Sullivan CE: Reversal of central sleep apnea using nasal CPAP. Chest 1986; 90: 165-171. 2. Hoffstein V. Slutsky AS: Central sleep apnea reversed by continuous positive airway pressure. Am Rev Respir Dis 1987; 135:1210-1212. 3. Bonnet MH. Dexter JR. Arand DL: The effect of triazolam on arousal and respiration in central sleep apnea patients. Sleep 1990; 13:31-41. 4. DeBacker WA. Verbacken J. Willemen M et al: Central apnea index decreases after prolonged treatment with acetazolamide. Am J Resp Crit Care Med 1995: 151:87-91. 5. Franklin KA. Eriksoon P. Sahlin C. et al: Reversal of central sleep apnea with oxygen. Chest 1997; Ill: 163-169. 6. Hommura F. Nishimura M. Oguri M, et al: Continuous versus bilevel positive airway pressure in a patient with idiopathic central sleep apnea. Am J Respir Crit Care Med 1997; 155: 1482-1485. 240

PATIENT 74 A 70-year-old man with daytime sleepiness and pedal edema A 70-year-old man was evaluated for complaints of waking at night gasping for air and daytime sleepiness over several years. The patient's wife also reported that he snored while supine and occasionally stopped breathing. He was being treated for hypertension and occasional episodes of leg swelling. Physical Examination: Blood pressure 130/80 mmHg, pulse 90, respiratory rate 15. General: no acute distress. Cardiac: S3gallop, PMI laterally displaced. Chest: rales at the lung bases. Extremities: 1+ pedal edema. Sleep Study: AHI 52/hr. Respiratory events: obstructive apnea 0%, mixed apnea 75%, central apnea 15%, hypopnea 10%. Figure: A mixed apnea characteristic of this patient is shown in the tracing below. Question: Is this a typical case of obstructive sleep apnea? Airflow 5a0 1J1Jwv-- vVVlf 2 98% 88% chest abdomen 5sec B l' A 241

Diagnosis: Obstructive (mixed) sleep apnea with underlying Cheyne-Stokes breathing due to congestive heart failure. Discussion: Sleep-disordered breathing is common but often unrecognized in patients with congestive heart failure. At least three forms of sleep apnea exist in these patients: (I) Cheyne-Stokes breathing (CSB) with central apnea, (2) traditional obstructive sleep apnea (OSA), and (3) a combination of CSB and OSA, producing mixed and central apneas. Many clinicians fail to suspect sleep apnea because complaints of poor sleep are attributed to dyspnea secondary to congestive heart failure. Recognition of sleep-disordered breathing is important because treatment not only improves nocturnal oxygenation and sleep quality, but may improve cardiac function. Patients with CSB+OSA can mimic the typical OSA presentation, with complaints of excessive daytime sleepiness and mixed apnea on polysomnography. The presence of CSB may be overlooked until treatment is attempted with nasal CPAP. Once the obstructive component is prevented, the repetitive central apneas of CSB are unmasked. Typical evidence of CSB is a crescendo-decrescendo pattern of ventilation with central apnea or hypopnea at the nadir in ventilatory effort. However, the return of ventilatory effort following central apnea is associated with upper airway closure in some patients. This causes an obstructive portion of apnea following the central part. In some patients, apnea is mixed in the supine position (predisposing to airway closure) and purely central (CSB pattern) in the lateral decubitus position. Some differences in mixed apnea between the two disorders are notable. The mixed apnea ofCSB tends to have a longer central than obstructive component. Also, in typical OSA a number of totally obstructive apneas usually are present. When CSB is secondary to congestive heart failure, the nadir in arterial oxygen saturation is delayed secondary to an increased circulation time. Finally, mixed apnea in patients with typical OSA resolves once upper airway obstruction is prevented. In contrast, patients with CSB have an underlying instability in ventilatory control that causes persistence of CSB. Although neurologic disease can cause CSB, congestive heart failure is by far the most common cause. The present patient had mixed apnea with a long central component and a short obstructive portion (see figure, area A). The small, rapid deflections in the abdominal tracing represent cardiac contractions (B). The most interesting finding was the unusual position of the nadir (88%) in arterial oxygen saturation. At first glance the nadir appears to occur before, rather than after, apnea termination. However, the illustrated nadir is actually the consequence of the preceding apnea (not shown). The severity of congestive heart failure was unappreciated by the physicians taking care of the patient. A subsequent nuclear medicine study (multiple gated acquisition) revealed a left ventricular ejection fraction of 30%. (See Patients 75 and 76 for discussions on the treatment of CSB with and without an obstructive component). Clinical Pearls I. Cheyne-Stokes breathing may present as a sleep apnea syndrome, with mixed and central apneas and daytime sleepiness. 2. A delay in the nadir of the arterial oxygen saturation is a clue that a long circulation time is present. This is evidence of significant cardiac dysfunction. 3. Sleep-disordered breathing in patients with congestive heart failure (obstructive sleep apnea, CSB, or a combination) frequently is unrecognized and may negatively impact patient outcome if not adequately treated. REFERENCES 1. Dowdell WT. lavaheri S. McGinnis W: Cheyne-Stokes respiration presenting as sleep apnea syndrome. Am Rev Respir Dis 1990; 141:871-879. 2. lavaheri S. Parker Tf, Wexler L. et al: Occult sleep-disordered breathing in stable congestive heart failure. Ann Intern Med 1995; 122:487-492. 242

PATIENT 75 A 60-year-old man with severe congestive heart failure and daytime sleepiness A 60-year-old man with known congestive heart failure (left ventricular ejection fraction 30%) was referred for complaints of daytime sleepiness. His wife denied hearing snoring but had noted her husband gasping for air and breathing irregularly during sleep. The patient also noted paroxysmal nocturnal dyspnea and had been admitted several times for exacerbations of congestive heart failure. Physical Examination: Pulse 85 and regular, blood pressure 110170 mmHg. HEENT: normal palate; no jugular venous distention; IS-inch neck circumference. Chest: bilateral rales. Cardiac: S3 gallop, grade 2 holosystolic murmur at the apex. Extremities: 2+ pedal edema. Laboratory Findings: Arterial oxygen saturation (room air): 94%. Sleep Study: AHI 50/hr (70% central apneas, 30% hypopneas), arousal index 40/hr, desaturation index 30/hr « 85%), mean Sa02 at desaturation 85%. Figure: This sample tracing of airflow and arterial oxygen saturation (Sa02) shows multiple central apneas. Vertical arrows mark the timing of arousals. Question: What is the diagnosis? Airflow 20 sec 243

Diagnosis: Central apnea with Cheyne-Stokes breathing secondary to heart failure. Discussion: Cheyne-Stokes breathing (CSB) is defined as a crescendo-decrescendo pattern of breathing with central hypopnea or apnea at the nadir. It is caused by an instability in ventilatory control. The two most common causes of CSB are congestive heart failure (CHF) and neurologic disease (cerebrovascular accidents). The vast majority of CSB is secondary to CHF; the incidence of CSB in patients with severe CHF of any cause is as high as 40-50%. CSB often is unsuspected, as typical patients only exhibit it during sleep. The etiology of CSB in patients with CHF was formerly attributed to delayed feedback of changes in blood gases to the ventilatory controllers (long circulation time), thus producing an "overshoot" in ventilation. During sleep, when the PCO? drops below the apneic threshold, central apnea occurs. One would expect the CHF patients with CSB to have longer circulation times (as a consequence of worse cardiac function). However, when groups of CHF patients with and without CSB were compared, the main difference was not the degree of cardiac dysfunction (left ventricular ejection fraction), but the slightly lower daytime peo z levels in CSB patients. Other studies have shown that patients with CSB have lower PCO? during sleep, higher LV filling pressures, and higher hypercapnic ventilatory drives. Higher ventilatory drives would predispose to an overshoot in ventilation. Higher filling pressures may stimulate pulmonary receptors (stiff lungs), which may also increase ventilation. In some patients, the sleeping PCOz decreases during the night possibly secondary to increasing LV filling (and pulmonary capillary) pressures. In this patient, CSB may appear only in the second or last third of the night. CSB may be associated with frequent arousal from sleep and daytime sleepiness. Unfortunately, clinicians typically assume that these symptoms are secondary to dyspnea associated with CHF. Interestingly, the arousals associated with CSB often occur at the zenith of ventilatory effort (see figure, arrows), rather than at apnea termination. Significant arterial oxygen desaturation is common in CSB patients, despite a normal daytime SaOz' Neither oxygen desaturation nor repetitive activation of the sympathetic nervous system is beneficial to patients already suffering from cardiac dysfunction. It is not surprising that the presence of CSB in patients with CHF signals a worse prognosis. The optimal treatment of CSB associated with CHF is not yet known. Certainly treatment starts with optimization of cardiac function. Other treatments for CSB associated with CHF include posi244 tive airway pressure and oxygen therapy. Theophylline also lowers the amount of CSB; however, treatment may not reduce the number of arousals and there is a concern about inducing arrhythmias with this medication. Thus, this medication is rarely used for CSB and CHF. Nasal CPAP can reduce the amount of CSB acutely in a few patients. However, chronic treatment with CPAP for I or more months is required in most patients. Acutely, an optimal pressure cannot be identified with a CPAP titration in most patients. The usual strategy is to start treatment with 5 ern HzO or a higher pressure if needed to eliminate any component of obstruction. Pressure is then increased to 10-12 em H?O as tolerated over several weeks. If patients are not sleepy, a period of "desensitization" to CPAP during the day may be needed befor CPAP is started. A pressure of 10-12 ern H?O was shown to be effective in several studies. The mechanisms by which CPAP reduces CSB are thought to include a modest increase in PCO}' an increase in oxygen stores (less desaturation), and a long-term improvement in cardiac function. Bilevel pressure can also be tried, although using too high a IPAP-EPAP difference could actually worsen CSB by augmenting a tendency to overshoot. However, in pressure-sensitive patients, CPAP may not be tolerated, and bilevel is an option. Oxygen treatment (2-4 Ipm) by nasal cannula has also been shown to reduce CSB and associated arousals and desaturation. Adaptive servo-ventilation (ASV) is a new, noninvasive ventilatory mode developed with the goal of rapid improvement in CSB (even on the first treatment night). In this method, a low level of EPAP (about 5 cm HzO) is used (sufficient to prevent obstruction), and the level of IPAP varies (4-10 cm H20 ) depending on the previous level of ventilation. The IPAP-EPAP difference increases during low periods of ventilation and decreases during high levels of ventilation. Thus, the amount of ventilatory support "adapts" to stabilize ventilation. The device also uses a back-up rate of IS/min. One study comparing oxygen, nasal CPAP, bilevel pressure (mean pressure 13/5 em HzO) with a back-up rate, and ASV found that all reduced the amount of central apnea, but ASV was the most effective. It reduced the central apnea index from around 35/hr to less than 5/hr. This form of ventilation is promising, but is not yet available for clinical use. Does treatment of CSB with CPAP improve cardiac function or survival? Short-term studies have shown improvement in cardiac function and a reduction in sympathetic tone with CPAP. One

study also suggests that CPAP can improve survival in patients with CHF and central sleep apnea (see Fundamentals 16). This may be secondary to reduction in hypoxia, sympathetic tone, and a decrease in afterload. A multicenter trial is underway in Canada to compare conventional care to CPAP in patients with severe CHF (CANPAP study) to determine if CPAP reduces mortality in these patients. The present patient was treated with ongoing nasal CPAP at 10 cm H20 . He reported improved sleep quality, and over the next several weeks noted a dramatic decrease in pedal edema although his medications were unchanged. Clinical Pearls I. The type of central apnea that includes Cheyne-Stokes breathing (CSB) is common and often unsuspected in patients with significant congestive heart failure (CHF). 2. CSB can present with symptoms of daytime sleepiness and disturbed sleep. 3. Adequate treatment of CSB associated with CHF can improve sleep, cardiac function, and perhaps even the long-term prognosis of these patients. 4. The best treatment for patients with CHF and CSB plus central sleep apnea is not known. Nasal CPAP or oxygen are current treatment options. 5. A level of nasal CPAP that effectively reduces CSB acutely often cannot be identified. The usual approach (assuming no OSA is present) is to start with a low level ofCPAP (5 ern H20 ) and increase to 10-12 cm H20 over several weeks as tolerated. REFERENCES I. Hanly PJ. Millar TW, Steljes DO, et al: The effect of oxygen on respiration and sleep in patients with congestive heart failure. Ann Int Med 1989: 111:777-782. 2. Naughton M, Bernard D, Tam A, et al: Role of hyperventilation in the pathogenesis of central sleep apnea in patients with congestive heart failure. Am Rev Respir Dis 1993; 148:330-338. 3. Jahavheri S, Parker TJ, Wexler L, et al: Effect of theophylline on sleep-disordered breathing in heart failure. N Engl J Med 1996; 335:562-567. 4. Teschler H, Dohring J, Wang Y, Berthon-Jones M: Adaptive pressure support servo-ventilation. A novel treatment for CheyneStokes respiration in heart failure. Am J Resp Crit Care Med 2001: 164:614--619. 5. Yan AT, Bradley TD. Liu PP: The role of continuous positive airway pressure in the treatment of congestive heart failure. Chest 2001; 120:1675-1685. 245

PATIENT 76 A 60-year-old man with obstructive sleep apnea and numerous central apneas on CPAP A 60-year-old man with a diagnosis of obstructive sleep apnea (OSA) was referred from another sleep laboratory for a nasal CPAP titration. His previous sleep study showed an AHI of 60/hr. Figure: A tracing of a central apnea at a CPAP level of 12 ern HoO is shown below. Sleep Study-CPAP Trial - CPAP(cm HP) 0 5 7.5 10 12 15 NREM (min) 120 60 60 60 20 30 AHI (lhr) 55 50 50 40 35 30 % obstructive apnea 0 0 0 0 0 0 % mixed apnea 100 50 0 0 0 0 % central apnea 0 0 0 100 100 100 % hypopnea 0 50 100 0 0 0 Desaturations < 85% 40 30 10 0 0 0 Arousal index (/hr) 40 35 20 10 10 10 Question: What is causing the central apnea? Nasal CPAP 12 em H2O Airflow N\) Chest Abdomen Sa02 90% 96% 246

Diagnosis: Cheyne-Stokes breathing presenting as obstructive sleep apnea (mixed apnea). Discussion: When patients with OSA (mixed apnea) and underlying Cheyne-Stokes breathing (CSB) are treated with nasal CPAP, the CSB tends to be unmasked. In typical cases of mixed apnea, CPAP abolishes the obstructive component by preventing upper airway obstruction. However, in patients with OSA plus CSB, the centraL component persists secondary to the underLying eSB. The sudden appearance of repetitive central apnea during a CPAP titration may be a surprise if the underlying CSB was not appreciated on the diagnostic study. Note that central apneas also may appear during upward titration of CPAP if high levels of pressure trigger arousal and hyperventilation (followed by central apnea on return to sleep). However, these central apneas will not be of the Cheyne-Stokes type. Cheyne-Stokes breathing also may not be appreciated when central hypopnea rather than apnea is noted at the nadir in ventilation. If the flow signal is used from the positive-pressure devices, the shape of inspiratory flow is usually rounded during central hypopnea (see Fundamentals 16). The first goal of CPAP titration in these patients is to prevent upper airway obstruction. As discussed in Case 75, further upward titration of CPAP usually does not abolish CSB, but may reduce desaturation and arousal. Acute reduction of CSB may be secondary to increases in PCO, or a reduction in the severity of desaturation. When no level of CPAP abolishes CSB, one approach is to treat with pressure at 10-12 em H20 (or a higher level if needed to prevent upper airway obstruction). Several studies have shown that treatment during sleep with this level of nasal CPAP in patients with congestive heart failure (CHF) produces long-term improvements in cardiac function-by improving oxygenation, decreasing afterload, and decreasing sympathetic stimulation from frequent arousals. Improvements in cardiac function should eventually decrease the amount of CSB. However, one study of similar patients showed no improvement in symptoms of cardiac function when nasal CPAP at a level of 7.5 em H20 was compared with placebo after 2 weeks. This result emphasizes that treatment must be individualized and monitored carefully. It may also take longer than 2 weeks to see an improvement. Long-term effects of CPAP may depend on the preload status of the patient. In the present case, the CPAP trial shows a conversion from mixed apnea to hypopnea and finally to central apnea as the level of CPAP increased. Further upward titration did not eliminate the central apneas, but the number of arousals was decreased, and oxygenation was much improved. Examination showed a crescendo-decrescendo pattern of breathing with central apneas or hypopneas at the nadir, consistent with CSB. An echocardiogram revealed unsuspected cardiac dysfunction. CPAP at 12 em H20 was prescribed, and on a repeat sleep study 2 months later the AHI was reduced to IO/hr. Improved cardiac function also was evident. Clinical Pearls 1. CPAP therapy may uncover Cheyne-Stokes breathing (CSB) in patients with congestive heart failure once the obstructive component has been eliminated. 2. CSB (central apnea) usually persists despite upward titration of CPAP. 3. The goals of CPAP therapy in patients with both upper airway obstruction and CSB secondary to heart failure are to prevent airway obstruction, improve oxygenation, and provide a pressure sufficient to improve cardiac function (10-12 em H20 in most patients). REFERENCES 1. Takasaki Y, Orr D, Popkin J, et al: Effect of nasal continuous positive airway pressure on sleep apnea in congestive heart failure. Am Rev Respir Dis 1989; 140:1578-1584. 2. Dowdell WT, Javaheri S, McGinnis W: Cheyne-Stokes respiration presenting as sleep apnea syndrome. Am Rev Respir Dis 1990; 141:871-879. 3. Bradley TD, Holloway RM, McLaughlin PR, et al: Cardiac output response to continuous positive airway pressure in congestive heart failure. Am Rev Respir Dis 1992; 145:377-382. 4. Davies RJO, Harrington KJ, Ormerod OJM, et al: Nasal continuous positive airway pressure in chronic heart failure with sleep disordered breathing. Am Rev Respir Dis 1993; 147:630-634. 5. Yan AT, Bradley TD, Liu PP: The role of continuous positive airway pressure in the treatment of congestive heart failure. Chest 2001;120: 1675-1685. 247

PATIENT 77 A 12-year-old boy with daytime CO2 retention A 12-year-old boy underwent a sleep study to determine if his noninvasive nocturnal ventilation with bilevel pressure was adequate to prevent severe desaturation or nocturnal hypercapnia. As a young child he had undergone tracheosotomy and home ventilation. Currently the patient was using nasal bilevel ventilatory support only at night. During prior evaluations his respiratory muscle strength and lung function had been normal. Physical Examination: HEENT: healed tracheostomy scar; no tonsillary enlargement. ABG on room air: pH 7.34, PC02 55 mmHg, P02 65 mmHg, HC03 30 mmol/L. Figure: Tracing was obtained during NREM sleep, off bilevel pressure support. Question: What is the diagnosis? ! Arousal J, Awakening airflow !VVVVVVVIN\Arv-IINJV\I\/\/'vv"vwv-"/'/IIVWV'f',t.rVI!W.fJ\fWWJ\NVWN chest abdomen Sa02 248 ( 180 seconds )

Answer: Congenital central hypoventilation syndrome. Discussion: Congenital central hypoventilation syndrome (CCHS), also known as idiopathic CCHS, was formerly called Ondine' curse. CCHS is a congenital form of severe central hypoventilation of undetermined etiology. The hypoventilation is not on the basis of neuromuscular disease or lung disease: the defect is in the central metabolic control of breathing. The voluntary control of breathing is intact, and the peripheral chemoreceptors have normal responses to hypercapnic and hypoxia. However, there is an absence or severe decrease in the hypoxic or hypercapnic ventilatory responses. The defect is believed to be in the central integration of chemoreceptor information. Most patients with CCHS present at birth with apnea or erratic spontaneous breathing. They typically have cyanosis. Other patients present soon after birth with pulmonary hypertension or respiratory failure. The hypoventilation is worse during sleep, but also usually is present during wakefulness. The additional drive to breath during wakefulness (wakeful drive) is the reason most patients hypoventilate less during wake. Some children may require 24-hour ventilatory support. There is a higher than expected frequency of neuroblastoma and Hirschsprung's disease with CHHS; other causes of central hypoventilation syndrome are listed in the table. Late-onset CCHS is a rare syndrome in which patients present at ages 2-4 and have associated hypothalamic abnormalities. Sleep studies in CCHS typically show a reduction in tidal volume and breathing rate, but frank central apnea is not common. Unlike other breathing disorders, some patients with CHHS have worse breathing problems during NREM than REM sleep. This may be explained by the fact that metabolic control of respiration occurs during NREM sleep. Ventilation during REM sleep is less related to hypoxic and hypercapnic stimuli. Patients with CCHS will arouse from high levels of hypoxic hypercapnia. Central Hypoventilation Syndromes in Children Primary Congenital CHS Late-onset CHS Secondary Obesity hypoventilation Associated with brainstem lesions* *Amold-Chiari type I or II, achondroplasia with stenosis of the foremen magnum. hypoxic encephalopathy, trauma, meningoencephalitis, poliomyelitis, tumor In the present case the tracing shows central hypoventilation and hypopnea during sleep. Note the decrease in Sa02• The patient was placed back on bilevel pressure of 16/9 via a nasal mask (see figure below), On this level ofsupport the patient did well during NREM and REM sleep. This is an example of a milder case of CCHS in which the patient does not require ventilatory support during the day. Bllevel19/6 em H20 bilevelnow vVY'{IlV\Yo/YY'N'{'{'lvY'IV'vYNvYN-..'Y'V''JY'fV'VY'{VVV'N\'Y'NYY''/VV'{V'V\ chest abdomen 5002 98% ( 180 seconds ) 249

Clinical Pearls I. Congenital central hypoventilation syndrome (CCHS) is secondary to a defect in the metabolic control of ventilation. The hypoventilation is not due to lung disease, upper airway abnormality, or neuromuscular weakness. 2. Sleep studies of CCHS show periods of central hypopnea and hypoventilation (rare central apnea). 3. Noninvasive pressure support (bilevel) with or without oxygen during sleep may suffice in some cases. Others need 24-hour ventilatory support via tracheostomy. 4. In some patients with CCHS, the hypoventilation is more severe in NREM than REM sleep. REFERENCES I. American Thoracic Society: Idiopathic congenital central hypventilation syndrome: Diagnosis and management. Am J Resp Crit Care Med 1999;160:368-373. 2. Marcus CL: Sleep-disordered breathing in children. Am J Resp Crit Care Med 2001; 164:16-30. 250

PATIENT 78 A 75-year-old man with a history of polio A 75-year-old man was seen for complaints of daytime somnolence and pedal edema. At age 30 he had a severe case of poliomyelitis with weakness in the arms and legs. After partial recovery, he enjoyed good health until the last few years. About 2 years ago, unexplained pedal edema began to occur. An arterial blood gas test (room air) at that time revealed: pH 7.36, PCOz 50 mmHg, pOz 65 mmHg, HCO) 25 mmol/L. The patient was started on nocturnal oxygen, at 2 L/min by nasal cannula at night, and his pedal edema improved. No sleep study was performed at that time. The patient's wife reported that he rarely snored or gasped for air at night. Over the last few months, the patient had become increasingly somnolent during the day, and his ankles began to swell. Physical Examination: Vital signs: unremarkable. HEENT: edentulous, otherwise normal. Chest: clear to ausculation and percussion. Cardiac: no murmurs. Abdomen: normal. Extremities: 2 + pedal edema. Neurologic: mild-to-moderate symmetric reduction in strength in the arms and legs; gag reflex intact. Laboratory Findings: ABG (room air): pH 7.34, PCOz 60 mmHg, POz 55 mmHg, HCO) 27 mmol/L. Question: What evaluation and treatment would you suggest? 251

Answer: Sleep study with titration of bilevel positive airway pressure. Discussion: Patients with a distant history of poliomyelitis may experience a worsening of the muscle weakness later in life (post-polio syndrome). Respiratory muscle weakness can result in hypoventilation during the day, with a worsening during sleep. The associated desaturation during sleep may result in findings consistent with cor pulmonale. During NREM sleep, periods of central apnea, hypopnea, obstructive apnea, or regular breathing with hypoventilation may occur. During REM sleep, more profound arterial oxygen desaturation usually is noted. Transcutaneous PC02 monitoring reveals an increase in PC02 during NREM sleep and a further increase during REM sleep. The treatment of this syndrome must be individualized. Milder cases may respond to oxygen administration. As hypoventilation worsens, some degree of ventilatory support, such as bilevel pressure via nasal mask, is indicated. The level of expiratory positive airway pressure (EPAP) is titrated to maintain upper airway patency. The level of inspiratory positi ve airway pressure (IPAP) is adjusted to provide a level of pressure support (pressure support = IPAP - EPAP). The level of pressure support is adjusted until the spontaneous tidal volume is in a normal range (400-500 cc). Oxygen can be added if de saturation persists despite pressure support of ventilation. Negative-pressure ventilation via body wrap or cuirass also has been used successfully in some patients. One potential problem is the development of upper airway obstruction because upper airway muscle activity is not coordinated with the negativepressure breaths. If patients do not improve adequately with the aforementioned approaches, the next step is positivepressure, volume-cycled ventilation via nasal mask. Mouth leaks may require chin straps or a full face mask. In very severe cases, tracheostomy and volume ventilation can be employed. Even then, many patients may only require ventilatory support at night. Nocturnal ventilation can improve the daytime PC02 by preventing hypoxic depression of ventilatory drive and resting the respiratory muscles. In the present patient, a sleep study revealed desaturation to 80% without discrete hypopneas during NREM sleep. During REM sleep, central apneas and arterial oxygen desaturation to 50% were noted. During the second part of the night, bilevel pressure was titrated to a level of 15/5 ern H20, and oxygen at 2 Llmin was added to prevent desaturation. Using this approach, nocturnal desaturation was prevented. After I week of treatment the patient felt more awake, and his daytime PC02 decreased to 50 mmHg. He was followed carefully so that the amount of pressure support could be increased or volume-cycled ventilation via mask instituted if a progressive rise in PC02 was noted. Clinical Pearls I. Respiratory failure can develop many years after the initial infection in patients with a history of poliomyelitis. 2. Patients with neuromuscular disorders may experience severe nocturnal hypoventilation (with or without obstructive or central sleep apnea) and symptoms of cor pulmonale and daytime sleepiness. 3. Nocturnal ventilatory assistance with bilevel pressure or volume-cycled ventilation via nasal or full face mask can prevent nocturnal hypoventilation and allow many patients to function during the day without ventilatory support. REFERENCES 1. Bach JR. Alba MS: Management of chronic alveolar hypoventilation by nasal ventilation. Chest 1990; 97:52-57. 2. Steljes DG. Kryger MH. Kirk BW, Millar TW: Sleep in postpolio syndrome. Chest 1990; 98: 133-140. 3. Waldhorn RE: Nocturnal nasal intermittent positive pressure ventilation with bi-Ievel positive airway pressure (BIPAP) in respiratory failure. Chest 1992; 10I:516-521. 4. Meyer TI, Hill NS: Noninvasive positive-pressure ventilation to treat respiratory failure. Ann Intern Med 1994; 120:760-770. 5. Claman DM, Piper AM, Sanders MH, et al: Nocturnal noninvasive positive pressure ventilatory assistance. Chest 1996; 110: 1581-1588. 252

FUNDAMENTALS OF SLEEP MEDICINE 17 The Restless Leg Syndrome and Periodic Leg Movements in Sleep Restless Leg Syndrome (RLS). RLS is characterized by paresthesias (abnormal sensations) and dysesthesias (uncomfortable sensations) in the limbs that compel the person to move to relieve the sensation and that are exacerbated by rest. The symptoms occur primarily in the evening or at night. In 1995, the International RLS Study Group published the primary and associated features of the syndrome (see table). The primary features are required to make the diagnosis of the RLS. The associated features are not required to make the diagnosis, but are frequently present and support the diagnosis. International RLS Study Group Criteria for Diagnosis ofRLS PRIMARY FEATURES ASSOCIATED FEATURES • Unpleasant limb sensations: desire to move the limbs usually associated with paresthesias/ dysesthesias (abnormal/unpleasant sensations) • Motor restlessness: patient is compelled to move • Symptoms precipitated by rest and relieved by activity: symptoms are worse or exclusively present at rest (i.e., sitting or lying) with at least partial and temporary relief by activity • Symptoms worse in the evening or at night • Sleep disturbance and consequences: difficulty initiating or maintaining sleep; less commonly, excessive daytime sleepiness • Invo!unatary movements during wake or sleep (PLMS) • Normal neurologic exam in primary RLS; in secondary forms, possible evidence of neuropathy • Clinical course: onset any age, usually chronic and progressive, remissions may occur, can be exacerbated by or exclusively during pregnancy • Family history: sometimes present; suggestive of autosomal dominant pattern The unpleasant sensations associated with RLS have been described as a creepy or crawling feeling, burning, bone ache, pulling, electrical current, throbbing legs, "heeby jeebies," or worms crawling under the skin. Sometimes sensations are absent, and there is only an irresistible urge to move ("Elvis legs"). Other patients report that the legs just move without an associated urge to move. About 20% of patients report the unusual sensations to be painful. Of note, about 20-30% also report similar sensations in the arms (usually in more severely affected patients). One study found that 10% of adults experienced RLS symptoms either often or very often. The prevalence of RLS symptoms increases with age and may be as high as 10-25% in adults over age 65 (depending on whether patients with only occasional symptoms are included). The differential diagnosis of RLS is discussed in Patient 80. Periodic Leg Movements (PLMs). PLMs are repetitive, stereotypic dorsiflexions of the big toe with fanning of the small toes, accompanied by flexion of the ankles, knees and thighs that recur at intervals of 253

5-90 seconds with a duration of 0.5-5 seconds. The leg movements resemble the triple flexion or Babinski response. At least four movements must be counted per episode to meet the polysomnographic criteria for PLMs. Some clinicians recommend counting all leg movements whether or not they meet this criteria and whether or not they occur in wakefulness or sleep. Technical information on the recording of leg movements during sleep is discussed in Patient 79. Periodic leg movements in sleep (PLMS) occur most commonly in stages 1 and 2, but can also occur in stages 3,4, and-less commonly-during REM sleep. Periodic leg movements should not be confused with hypnic jerks (sleep starts), which are whole-body jerks at sleep onset. Patients with frequent leg movements may remember awakenings, but rarely are aware of the leg movements themselves. Bed partners usually report that patients jerk or kick during the night. The PLM index is the number of periodic leg movements per hour of sleep. It is said that a PLM index of > 5/hr is abnormal, but this cutoff is arbitrary and not based on scientific data. These movements are present in 5-6% of all adults and 30-86% of adults over 60 years of age. PLMS may be an asymptomatic finding, especially in older patients. The term periodic leg movement disorder (PLMD) is used to identify the syndrome of leg movements + symptoms (insomnia or excessive daytime sleepiness). The disorder is diagnosed when these symptoms are present in association with an abnormal PLM index and/or PLM arousal index. This assumes that no other disorder can explain the symptoms. While PLMD is thought to be the etiology in about 10-12% of patients seen in sleep centers for insomnia complaints, only 2-3% of patients presenting with excessive daytime sleepiness are thought to have PLMD as the major cause of their sleepiness. International Classification ofSleep Disorders Criteria for PLMS Severity SEVERITY PLM INDEX (lHR) PLM AROUSAL INDEX (lHR) Mild Moderate Severe 5-24 25-49 50 Not specified Not specified > 25 /hr While the terms "restless leg syndrome" and "periodic limb/leg movements in sleep" are often used interchangeably, such usage is inaccurate. RLS is a clinical diagnosis, and PLMS is usually a polysomnographic diagnosis. It is estimated that 70-90% of patients with RLS will have PLMS on a sleep study. In contrast, only 30% or less of patients with PLMS have RLS. There seems to be little doubt that RLS is a distinct clinical entity with significant morbidity, deserving treatment. However, whether isolated PLMD (no RLS) is really a clinical syndrome that warrants treatment has been called into question (see reference 4 and Patient 80). Causes of RLS/PLMS. RLS is often divided into primary RLS (no other disease present) and secondary RLS (secondary to an indentifiable cause). The cause of primary RLS is not known. There may be an abnormality in iron transport into the CNS or in use of iron as it relates to dopaminergic neurons. Examples of secondary RLS include end-stage renal failure (with or without dialysis), pregnancy, iron-deficiency (with or without anemia), and certain drugs. It is prudent to check serum iron, total iron-binding capacity, and ferritin in patients with RLS. RLS associated with renal failure is not helped by dialysis, but is cured by a renal transplant. The RLS of pregnancy commonly vanishes or improves with delivery. The RLS of iron deficiency may improve with iron supplementation. A number of medications can cause or worsen RLS, including selective serotonin reuptake inhibitors. However, the occasional patient with RLS improves on SSRIs. Bupropion is an antidepressant that increases dopamine and may be an alternative in patients with the onset or worsening of RLS on other antidepressants. If the onset of RLS can be linked to the start of medication, try a switch to an alternate medication. The causes of PLMS are the same as RLS. In addition, PLMs can be seen during CPAP titration for OSA and are also seen in patients with narcolepsy, OSA, and the REM behavior disorder. PLMs have also been noted upon withdrawal from anticonvulsants, barbiturates, and hypnotics. For a detailed discussion of the treatments of RLS and PLMS, see Patients 82 and 83. 254

Classification ofRestless Leg Syndrome Primary Secondary Iron deficiency Renal failure Pregnancy Drugs (caffeine, tricyclic antidepressants, serotonin reuptake inhibitors, dopamine blockers [compazine, metaclopramide]) Causes and Associations ofPLMS Any cause of RLS Withdrawal of anticonvulsants, barbituates, hypnotics Associated with narcolepsy, GSA, CPAP titration Key Points 1. The restless leg syndrome (RLS) is characterized by symptoms that occur during wakefulness and that meet certain criteria. 2. Periodic limb movements in sleep (PLMS) is diagnosed when the PLM index is > 5/hr. For a leg movement to be counted as a PLM, it must occur in a sequence of a least four leg movements separated by 5 to 90 seconds. 3. The PLM Disorder (PLMD) is defined as PLMS + symptoms of insomnia or daytime sleepiness. Some have questioned whether PLMD is a real clinical entity. PLMs occur in many aysmptomatic individuals. 4. About 70-90% of patients with RLS have PLMS on a given night. 5. Many patients with PLMS do not have RLS. REFERENCES I. American Sleep Disorders Association: International Classification of Sleep Disorders: Diagnostic and Coding Manual. Lawrence. KS. Allen Press. 1990. pp 65-71. 2. Ancoli-Israel S. Kripke DF. Klauber MR. et al: Periodic leg movements in sleep in community-dwelling elderly. Sleep 1991; 14:496-500. 3. Waters AS: Toward a better definition of the restless leg syndrome. The International Restless Leg Syndrome Study Group. Mov Disord 1995; 10:634-632 4. Phillips B. Young T. Finn L. et al: Epidemiology of restless leg symptoms in adults. Arch Intern Med 2000; 160:2137-2141. 5. Mahowald MW: Assessment of periodic leg movements is not an essential component of an overnight sleep study. Am J Resp Crit Care Med 2001; 164:1340-1341. 255

----. -----~,.------1.,."""",-- ""'" --~-~ tPATIENT 79 A 50-year-old man with snoring and leg kicks A 50-year-old man with a history of snoring and leg kicks during sleep underwent a sleep study. The tracings shown below were obtained. Question 1: Note the right and left leg EMGs on separate channels (Fig. I). How many PLMs are present in this 60-second tracing? Question 2: Note the 180-second segment with repetitive obstructive apneas (Fig. 2). The right and left leg EMGs are displayed on a single channel. How many PLMs are shown? C4-A1 ~"'~~~*",Mf,\J~-WJ~~~ 02-A1Nfl.t.~I,.,M.J¥-Lf'lf'\Al"'''f~V.~AtM''W.f'~'t¥'N1 ROC-A1'~IfYI~~NJw.(~~r..Ilk'iv0rrf\\~ LOC-A2 ~'I"rI'o/~~II.Af'~ chin EMG ---.--------.-----...;.--......, R Leg EMG L Leg EMG 5 sec FIGURE I 256 C4-Al 02-Al ROC-A1 LOC-A2 chin EMG airflow chest abdomen R. L Leg EMG - - - - _._.. . . -- ---. .... ~. """. ( ) 30 seconds FIGURE 2

Answers: In Figure I, four PLMs are shown. In the first pair of leg movements, the left leg movement starts more than 5 seconds after the onset of the right leg movement, and is therefore considered a separate movement. In Figure 2, none of the leg movements are considered PLMs because they all are associated with the termination of obstructive apnea. Discussion: The International Classification of Sleep Disorders states that the diagnosis of periodic limb movement disorder (PLMD) requires symptoms (excessive daytime sleepiness/insomnia) + PLMs noted on a sleep study. Other sleep disorders may be present, but cannot account for the leg movements. There should also be no evidence of a medical or psychiatric disorder that could account for the complaint. As discussed in Patient 84, the presence of leg movements in patients with OSA has a questionable impact on sleepiness in most patients unless the restless leg syndrome is present. Some have suggested that PLMD is not a true sleep disorder, but simply a polysomnographic finding. The diagnosis of periodic leg movements in sleep (PLMS) requires monitoring of leg EMGs. Surface electrodes usually are placed over the anterior tibialis muscle (anterior lateral calf). Movements may occur in one or both legs; therefore, monitoring of both legs is suggested. This dual monitoring can be performed using a single polygraph channel (Fig. 2) or in separate tracings (Fig. I). Usually two electrodes are placed on each anterior tibialis muscle about 2-4 ern apart. Then each leg EMG can be recorded in a bipolar manner as the voltage between the two electrodes. If the legs are displayed on a single channel, the voltage difference between one electrode on each leg is measured. The other electrodes are held in reserve in case one of those monitored fails. A burst of anterior tibialis muscle activity with a duration between onset and resolution of 0.5 to 5 seconds, with an amplitude of at least 25% of the bursts recorded during patient biocalibration (voluntary toe dorsiflexion), is scored as a leg movement (LM). A leg movement is considered part of a PLM sequence if it belongs to a group of four or more leg movements separated by more than 5 and less than 90 seconds (most separations are 20-40 seconds). This defining separation is from the onset of one leg movement until the onset of the second. In contrast, the inter-LM interval is from LM offset to the next LM onset. Leg movements following arousals or those associated with apnea termination are not counted as PLMs in most laboratories. Others also do not count LMs occurring during crescendo snoring, but those occurring in the middle of an apnea are counted in some centers. In the two-channel method of display, simultaneous movements in both legs are counted as one movement, unless the start of the second movement is more than 5 seconds after the start of the first (some use 5 seconds after the offset of the initial LM). Of note, an American Sleep Disorders Association (ASDA) task force actually recommended that all LMs be counted, and those that are part of a PLM sequence, associated with respiratory events, or occuring during wake be tabulated separately. However, most sleep laboratories simply count only those LMs that occur during sleep, are part of a PLM sequence, are not felt to be associated with respiratory events, and do not follow an arousal. A leg movement must be preceded by at least 10 seconds of sleep to be considered a PLM in sleep. To be scored as a PLM with arousal, an arousal must occur simultaneously with a LM or follow it by less than 3 seconds. The total number of PLMs, the number of PLMs with arousal, and the PLM index and PLM arousal index are usually determined. The PLM index is the number of periodic leg movements per hour of sleep, and the PLM arousal index is the number of PLMs associated with arousal per hour of sleep respectively. The PLM arousal index has been thought to be a better indicator of the impact of PLMs on sleep than the PLM index. This said, many leg movements are associated with EEG changes that do not meet ASDA arousal criteria. For example, K complexes with or without alpha bursts are frequently seen. Some of these can be associated with changes in airflow, heart rate, or blood pressure in the absence of cortical arousal. Of note, the PLM arousal index has not been shown to correlate with objective measures of sleepiness. The International Classification of Sleep Disorders offers a grading system for severity of PLMS: a PLM index < 5 is considered normal; 5-24 is mild; 25-49 is moderate; and 2= 50/hr is severe. A PLM arousal index > 25/hr is also considered severe. However, these cutoffs are entirely arbitrary and are not based on any outcomes data. Normative data on the PLM and PLM arousal index are not available. One must always use clinical correlation (presence or absence of symptoms) in determining the importance of PLMS. In the present case, Figure 1 shows three pairs of right and left leg movements. However, in the first pair the left leg movement starts 5 seconds after the right and is considered a separate movement. The other pairs are each considered to represent one leg movement. As noted above, some centers require 5 seconds from offset to onset of the next LM to con257

sider LMs to be separate events. Using this criterion, only three PLMs would be counted in this patient. In Figure 2, all leg movements are associated with the termination of apnea. Hence, most centers would not score them as PLMs. Note that in this patient, the leg EMG shows a much more brisk change at apnea termination than the chin EMG. This is a frequent finding in some patients. Clinical Pearls I. The rules for scoring leg movements and periodic leg movements vary between sleep laboratories. 2. Simultaneous leg movements in right and left legs are counted as one movement. 3. Inorder to be considered a periodic leg movement in sleep, a leg movement must occur during sleep in a sequence of four or more movements, each separated by more than 5 and less than 90 seconds (time between leg movement onsets). Leg movements that follow an arousal are not counted. 4. The PLM index and PLM arousal index are computed by dividing the number of PLMs or the number of PLMs associated with arousal (arousal simultaneous or following the LM by < 3 seconds) by the hours of sleep. 5. It has been said that a PLM index of 5 or less is normal, but this cutoff is arbitrary. Many asymptomtic individuals have a higher PLM index. 6. The PLM arousal index may be a better indicator of the impact of PLMS on sleep quality, but a normal range for the PLM arousal index has not been established. REFERENCES 1. Coleman RM: Periodic movements in sleep (nocturnal myoclonus) and the restless leg syndrome. In Guilleminault C (ed): Sleep and Waking Disorders: Indications and Techniques. Boston, Butterworth Publishers, 1932, pp 265-295. 2. Diagnostic Classification Steering Committee, Thorpy MJ, Chair: International Classification of Sleep Disorders. Rochester, Minnesota, American Sleep Disorders Association, 1990. 3. The ASDA Task Force: Recording and scoring leg movements. Sleep 1993; 16:749-759. 4. Walters AS: Assessment of periodic leg movements is an essential component of an overnight sleep study. Am J Respir Crit Care Med 2001; 164: 1339-1340. 5. Marshall B. Davila DG: Monitoring limb movements during sleep. In Lee-Chiong TL, Sateia MJ, Caraskadon MA (eds): Sleep Medicine. Philadelphia, Hanley & Belfus, 2002. 258

PATIENT 80 A 56-year-old man with crawling sensations in his legs at bedtime A 56-year-old man reported an inability to fall asleep at night secondary to a crawling sensation in both legs. These unpleasant sensations frequently started an hour before bedtime when he was seated or recumbent in bed. The discomfort was relieved by moving his legs or walking, but it returned as soon as he became inactive. After great difficulty falling asleep, the patient experienced frequent and prolonged awakenings during the night. His wife reported that he kicked and jerked during sleep, but did not snore. Because of these problems, the patient was quite fatigued. An evaluation by the patient's primary care physician revealed no evidence of anemia, iron or folate deficiency, or renal failure. Physical Examination: Neurologic: intact position and vibration sense; sensation to pin and touch normal; deep tendon reflexes normal and symmetric. Laboratory Finding: Hct 40%. Questions: What is the diagnosis? Should a sleep study be ordered? 259

Diagnosis: Restless leg syndrome. A sleep study is not needed unless other sleep pathology (e.g., sleep apnea) is suspected. Discussion: The diagnosis of the restless leg syndrome (RLS) is based primarily on clinical history from the patient and bed partner. The essential elements of RLS (see Fundamentals 17) should be present. The differential diagnosis of RLS includes paresthesia due to neuropathy, neuroleptic akathisia, claudication, and the painful legs and moving toes syndrome (see table). In RLS, symptoms are worse in the evening and at night, exacerbated by inactivity and the supine position, and temporarily improved by activity. The neurological exam as well as EMG and nerve conduction studies are normal in most patients with RLS (unless the patient also has a coexistent neuropathy). Paresthesias or dysesthesias from a neuropathy are usually not relieved by moving or walking, and generally are not worse at night. EMG and nerve conduction studies are usually abnormal. Neuroleptic akathisia is characerized by repetitive restless movementsuch as body rocking, marching in place, or movement of the extremities. Movements may not be worse at night or when lying down. There is always a history of prior or current neuroleptic use (phenothiazines). The movements are associated with inner restlessness, but not dysesthesia/paresthesia. The painful legs and moving toes syndrome is characterized by semi-continuous toe movements not necessarily influenced by activity. The pain is often not worse at night, EMG studies are abnormal, and MRI imaging may reveal evidence of nerve root compression. While many patients focus on the specific symptoms of RLS, others complain of difficulty falling asleep or maintaining sleep (insomnia). Most patients with RLS have periodic leg movements in sleep (PLMS). However, a history of leg jerks at night does not eliminate the possibility that sleep apnea is also present. Remember that PLMS and obstructive sleep apnea (OSA) frequently coexist, and that OSA patients can also complain of insomnia. Therefore, a sleep study is indicated if the patient being evaluated for RLS snores, or if daytime sleepiness persists after adequate treatment of RLS. Otherwise, a sleep study is not needed to diagnose RLS. The etiology of RLS is unknown. The condition has been associated with iron deficiency (less commonly, folate or B12 deficiency), vascular insufficiency, uremic neuropathy, and caffeine abuse. Some ofthe described associations are based on relatively few anedoctal reports. When sleep monitoring is performed in patients with RLS, there are quasi-periodic movements of the legs during wakefulness, and the sleep latency is prolonged. After sleep onset, PLMs are noted in 70-90% of patients. In the present patient, clear-cut symptoms of RLS were present, and there was no reason to suspect sleep apnea (no daytime sleepiness or snoring). Therefore, a sleep study was not ordered. The patient was started on pramipexole .125 rng at 9 PM (usual bedtime was II PM). The dose was increased to .25 mg after 1 week. The patient reported nearly complete relief of his symptoms. (For a detailed discussion of the treatment of RLS and PLMS, see Patients 82 and 83.) WORSE IN IMPROVED DYSESTHESIA/ THE EVENING/ By NEUROLEPTIC PARESTHESIA NIGHT MOVEMENT STUDIES USE RLS Usually Always Temporarily Neuro exam, EMG No often normal Neuropathy Usually Sometimes No Decreased sensa- No tion, abnormal EMG/nerve conduction Neuroleptic No Not usually Not usually Variable Yes akathisia Claudication Pain Not usually Walking may Decreased No worsen pulses Painful toes, Pain Not usually No Abnormal EMG, No moving leg MRI shows syndrome lumbosacral radiculopathy 260

Clinical Pearls I. The diagnosis of RLS usually can be made by history and physical examination. 2. Iron-deficiency with or without anemia should be excluded in patients with RLS. 3. Most patients with RLS have PLMS during sleep. 4. A sleep study should be ordered if snoring or if daytime sleepiness is present in a patient with a clinical diagnosis of RLS. Sleep apnea frequently coexists with PLMS. REFERENCES I. Diagnostic Classification Steering Committee. Thorpy MJ, Chair: International Classification of Sleep Disorders. Rochester, Minnesota, American Sleep Disorders Association, 1990. 2. Walters AS, Hening W, Rubinstein M, et al: A clinical and polysomnographic comparison of neuroleptic-induced akathisia and the idiopathic restless leg syndrome. Sleep 1991; 14:339-345. 3. Waters AS: Toward a better definition of the restless leg syndrome. The International Restless Leg Syndrome Study Group. Mov Disord 1995; 10:634-632. 261

PATIENT 81 A 40-year-old man who kicks in his sleep A 40-year-old man was referred because his wife complained that he kicked in his sleep and constantly disturbed her. The patient remembered awakening several times each night, but never noticed any discomfort at those times. He admitted that at bedtime he did have an irresistible urge to move his legs. However, this delayed his sleep only rarely. The patient's wife said he regularly snored at night, but this did not bother her. The snoring was worse when the patient complained of nasal congestion. His Epworth sleepiness scale was 10/24 (normal). Physical Examination: HEENT: normal except for a mildly dependent palate and long uvula. Sleep Study: Total sleep time 360 minutes. Mild snoring, AHI 5/hr, AHI during REM l5/hr. Lowest sao,92%. Figure: There were 240 events similar to the two identified (A and B) in the tracing below. Twenty percent of these events were associated with arousal. Note that fewer PLMs occurred in the lateral sleeping position. Question: What is the diagnosis? C3-A2 Ol-A2 ROC-Al LOC-A2 EKG chin EMG Snore airflow chest abdomen R,L Legs Sa02 262

Diagnosis: Periodic leg movements in sleep (PLMS) associated with snoring airflow limitation. Discussion: Periodic limb (leg) movements in sleep (PLMS), also known as nocturnal myoclonus, consists of stereotypic periodic leg (arm) movements during sleep that mayor may not be associated with arousals. (See Fundamental 17 and Patient 79 for a detailed description of these movements.) The conventional wisdom once was that if the movements were associated with a sufficient frequency of arousals, the combination could result in either daytime sleepiness or insomnia (periodic limb movement disorder [PLMDJ). Recently, however, the idea that isolated PLMD (no RLS) is really a sleep disorder requiring treatment has been challenged. First, PLMS is very common in elderly patients and could simply be a polysomnographic finding. Second, the PLM arousal index does not correlate with measures of excessive sleepiness. Even in patients with insomnia and PLMs, it is not clear that the PLMs are actually the cause of the syndrome. Third, PLMs are common in patients with obstructive sleep apnea and narcolepsy. In one study of patients with sleep-disordered breathing, a higher PLM arousal index actually was associated with less daytime sleepiness. Anotherstudy of patients with the upper airway resistance syndrome found an association between respiratory effort-related arousals and PLMs. This suggests that PLMS could be a manifestation of upper airway narrowing in some patients. Despite the controversy about the diagnosis of PLMD, there is probably a subgroup of patients with PLMS in whom treatment of the leg movements does improve sleep. While leg movements may be associated with EEG changes consistent with cortical arousal, often other, more subtle changes are present-such as a K complex with alpha waves (K-alpha). Sometimes only an "autonomic" arousal is present, with an increase in heart rate or blood pressure. One study of the arousals associated with PLMs showed that the arousals actually followed the PLMs in less than 25% of the cases (other arousals preceded or were coincident with leg movements). Some have suggested that the PLMs and arousals are simply both manifestations ofperiodic arousals that mayor may not be associated with dopamine dysfunction. In patients with heavy snoring and airflow limitation as well as PLMs, it is often difficult to know if these are two separate entities, or if airflow limitation (increased respiratory effort) is causing the PLMs. In some patients with mild OSA and frequent PLMs, treatment with nasal CPAP abolishes the PLMs-suggesting that the sleep-disordered breathing was causing the PLMs. In others, PLMs may persist once the upper airway is stabilized. In such cases, first ensure that respiratory effortrelated arousals are abolished (increase in CPAP as needed). Then consider pharmacological treatment of PLMs-but only if the patient is still symptomatic after adequate treatment of OSA. Some have suggested that PLMs are simply a peripheral manifestation of periodically occurring central arousals, rather than the cause of the arousals. PLMs certainly can disturb the sleep of bedmates. It could be argued that PLMs without RLS could be treated to improve the sleep quality of the bedmate. While the area will likely remain controversial, it is probably prudent not to assume that an elevated PLM index explains symptoms of daytime sleepiness until you have carefully excluded other causes. For example, PLMs are common in patients with narcolepsy. Additionally, you should probably be conservative in treatment of PLMs in the absence of RLS. In the present patient with PLMS, the tracing is typical in that it shows PLMs associated with airflow limitation and snoring. The second PLM is followed by a K complex in the EEG tracings and a slight improvement in airflow, although cortical arousal is not evident. The PLM index was 40/hr (240/6), and the PLM arousal index was (8/hr). The patient was diagnosed as having mild OSA (REM related) and frequent PLMs, with a mildly increased PLM arousal index. He was absolutely asymptomatic and felt he had no problems. The couple was given three options: (l) sleep in separate beds, (2) trial of treatments for mild OSA, (3) trial of treatments for PLMS. Sleeping in separate beds was unacceptable and the patient was reluctant to try dopaminergic agents. He underwent a weight loss program (IO-pound weight loss), and his nasal congestion was treated with nasal steroids. On this treatment regimen the patient snored less and seemed to kick much less at night. He and his wife were satisfied. However, improvement in the PLM index was not documented by a repeat sleep study because of financial considerations. 263

Clinical Pearls I. Periodic leg movements in sleep are common in asymptomatic elderly patients and patients with narcolepsy and obstructive sleep apnea. Be cautious in ascribing complaints of daytime sleepiness or insomnia to PLMS. 2. In the absence of RLS, asymptomatic PLMS probably do not need treatment, unless the sleep of the bedmate is disturbed and other measures (separate beds) are not acceptable to the couple. 3. In some patients, PLMs may be a marker for airflow limitation/snoring and/or respiratory effort-related arousals. Treatment ofsnoring/UARS/mild GSA may be the best way to decrease PLMs in these individuals. REFERENCES I. Mendelson WB: Are periodic leg movements associated with clinical sleep disturbance? Sleep 1996; 19:219-223. 2. Exner EN, Collop NA: The association of upper airway resistance with periodic limb movement. Sleep 2000: 24: 188-192. 3. Karadeniz D, Ondze B. Besset A. Billiard M: EEG arousals and awakenings in relation with periodic leg movements during sleep. J Sleep Res 2000; 9:273-277. 4. Montplasir J, Michaud M, Denesle R, Gosseline A: Periodic leg movements are not more prevalent in insomnia of hypersomnia. but are specifically associated with sleep disorders involving a dopaminergic impairment. Sleep Medicine 2000; I: 163-167. 5. Mahowald MW: Assessment of periodic leg movements is not an essential component of an overnight sleep study. Am J Resp Crit Care Med 200 I; 164: 1340-1341. 6. Chervin RD: Periodic leg movements and sleepienss in patients evaluated for sleep-disordered breathing. Am J Respir Crit Care Med 2001; 164:1454-1458. 264

PATIENT 82 A 50-year-old man who kicks during sleep A 50-year-old man was evaluated because he kicked all night in bed, disturbing his wife's sleep. He described a feeling of "pins and needles" in his legs in the evening that sometimes made falling asleep difficult. He constantly moved his legs while trying to fall asleep. His wife reported that he snored occasionally. The patient denied symptoms of daytime sleepiness, and his Epworth Sleepiness Scale score was 9/24 (normal). There was no history of recent weight gain or cataplexy. The patient remembered awakening several times each night but never noticed any discomfort at those times. He was treated with c1onazepam I mg qhs, but this made him sleepy in the morning, and his wife reported that there was no improvement in his leg kicking during sleep. Physical Examination: Normal. Sleep Study: Total sleep time 360 minutes. AHI lO/hr. Figure: There were 240 events similar to the two identified (A and B) in the tracing below during 6 hours of sleep. Only 5% of the events were associated with arousal. Question: What treatment would you recommend? EKG R,l leg EMG 265

Answer: Dopamine agonist for the restless leg syndrome and periodic leg movements in sleep. Discussion: Many of the same drugs used to treat restless leg syndrome (RLS) are used to treat patients with the periodic leg movement disorder (see table). When treating patients with RLS (with or without periodic leg movements [PLMs]), the goal is to relieve symptoms and improve sleep quality. In patients with isolated leg movements in sleep (PLMS) but no RLS, the goal is to imin the morning. A controlled-release form of the medication can be used to avoid these problems. A more significant problem is the development of augmentation, i.e., symptoms of the RLS have earlier onset and spread to other areas of the body. Augmentation is more common when RLS is present (versus isolated PLMS) and with a dose in excess of 300 mg of levodopa. This problem is hanTreatment Options for PLMD and RLS PLMD RLS Mild or intermittent Moderate Severe Carbidopa/levodopa Benzodiazepines Carbidopa/levodopa Dopamine agonists Benzodiazepines Dopamine agonists Benzodiazepines Narcotics Nonpharmacological* Carbidopa/levodopa Benzodiazepines Dopamine agonists Anticonvulsants Benzodiazepines Narcotics Dopamine agonists Benzodiazepines Narcotics Anticonvulsants Combination/rotating agents PLMD = periodic leg movement disorder *Avoid alcohol. caffeine: do leg stretching, exercise; take warm baths prove sleep. The leg movements are not felt to be harmful except for disturbing the sleep of bedpartners. Dopaminergic medications (dopamine precursors and dopamine agonists) are probably the most effective and widely used treatments for PLMS and RLS. They reduce the frequency of PLMs, decrease RLS symptoms, and improve sleep quality. Carbidopa/levodopa (Sinemet and others) works because L-DOPA, a dopamine precursor, enters the central nervous system and is converted to dopamine. Carbidopa, a decarboyxlase inhibitor, does not enter the central nervous system. It prevents peripheral conversion of DOPA to dopamine, thereby decreasing peripheral side effects and increasing availability of L-DOPA in the brain. A typical dose of carbidopa/levodopa is 25/100 to 100/300 taken at bedtime and repeated during the night if needed. The medicine is started at a low dose of 25/100 and increased gradually. Compared to dopamine agonists, CD/LD requires only a short time to onset of action, but has a short half life. This agent is very useful for for rapid treatment of symptoms such as may occur on long airplane or car trips. However, the short half life of CD/LD often results in recurrence of symptoms during the night and a rebound in symptoms 266 died by a switch to dopamine agonists. In general, chronic CD/LD treatment of RLS/PLMS does not lead to the dyskinesias seen in Parkinson's disease (possibly because the daily dose is much lower). Pergolide (Permax) is an effective ergotamine dopamine agonist, but it is often associated with significant nausea and nasal congestion as well as constipation and orthostatic hypotension. These side effects can be controlled with dromperidone 10-30 mg a day (available in Canada but not the U.S.), which is a dopamine receptor blocker that does not cross the blood-brain barrier. Pergolide is less commonly used today. Pramipexole and ropinirole, non-ergotamine dopamine agonists, are available and effective. Their popularity is increasing because they are associated with much less nausea than ergotamine dopamine agonists (bromocriptine and pergolide); have a longer half life than CD/LD (no repeat dosing during the night); and are less likely to lead to augmentation. They are started at low doses (e.g., 0.125 mg of pramipexole or 0.25 mg of ropinirole) and titrated upward slowly to minimize side effects such as nausea and postural hypotension. Because the time to peak level is longer than CD/LD, it is important to give these agents 1-2 hours before bedtime (or symptom onset). Side effects include nau-

sea, postural hypotension, and daytime sleepiness. Leg edema has also been reported with pramipexole. If augmentation occurs with pramipexole, it can usually be handled by splitting the dose and giving a portion ofthe medication earlier in the evening (e.g., 0.125 mg at 6 PM and 9 PM). The dose of pramipexole must be reduced in renal insufficiency. Benzodiazepines have variable effectiveness in RLS and PLMs. Some studies reported that they improved sleep in patients with PLMS, but not the frequency of leg movements. Others showed an improvement in PLM frequency. Benzodiazepines should be used with caution-if at all-in patients with OSA. Clonazepam is the most commonly used benzodiazepine for RLS/PLMS. Unfortunately, it has a long half-life and can cause early morning grogginess. Other benzodiazepines such as triazolam and temazepam also improve sleep in patients with RLS. Clonazepam should be started low (0.125--0.25 mg) and titrated up slowly to minimize early morning grogginess. Giving the drug 1-2 hours before bedtime may decrease morning grogginess and give adequate blood levels at sleep onset. Narcotics (oxycodone, propoxyphene, and others) are also effective in RLS, but are usually reserved for severe cases. Two problems are the potential for abuse and the development of tolerance. One longterm study of narcotics in RLS did not find these potential factors to be major problems. However, these drugs might worsen sleep apnea, and some evidence for this problem was found in the study. Anticonvulsants (carbamazepine and gabapentin) are said to be effective in RLS, although less so than dopaminergic agents. There is not sufficient documentation of efficacy in PLMS to recommend their use ifPLMS occur without RLS symptoms. Anticonvulsants typically are used in RLS patients who have accompanying neuropathic pain, or in patients who cannot tolerate the other agents. Gabapentin can cause daytime fatigue and somnolence. Carbamazepine can cause nausea, dizziness, and, rarely, aplastic anemia and agranulocytosis. In patients with very severe RLS, combinations of agents (dopaminergic, benzodiazepines, narcotics) are frequently required. Some clinicians combine opiates and dopamine agents so that a lower dose of each agent is effective, resulting in fewer side effects. Another approach is to use different medications on alternate weeks. In the present case, although RLS symptoms were not a problem for the patient, reducing the PLMS was essential to keep his wife in the same bed. He was started on pramipexole 0.125 mg 2 hours before sleep, and this was increased to 0.25 mg I week later. His wife reported a great reduction in leg kicking. The patient did not experience daytime sleepiness on this treatment. Comparison ofPharmacologic Treatments for RLS and PLMS TIME To PEAK LEVEL (MIN) HALFLiFE (HRS) MODE OF ELIMINATION INITIAL DOSE DOSING ADJUSTTIMING OF MENT DoSE (MIN) (INCREASED BEFORE EVERY HS X DAYS) MAXIMUM DOSE 251100 mg 30-60 min 3-7 days Dopamine precursors CD/LD (Sinemet) CD/LDCR (Sinemet CR) 30 120 1.5-3 6-8 hepatic hepatic 25/100- 120 min 50/200 mg 3-7 days 300 mg of LD* 300 mg of LD* Dopamine agonists Bromocriptine (Parlodel) Pergolide (Permax) Pramipexole (Mirapex) Ropinirole (Requip) 45-60 3-4 hepatic 2.5 mg 45-60 min 3 days 20 mg 1/2 to I 60 27 renal 0.05 mg 60 min 2-3 days 1-2 mg one at supper or hs 120 8-12 renal .125 mg 120 min 2-3 days 1.5 mg 60-120 about 6 hepatic .25 mg 60-120 2-3 days 3 mg min (continued) 267

Comparison ofPharmacologic Treatments for RLS and PLMS (Continued) DOSING ADJUSTTIME To TIMING OF MENT PEAK HALF- MODE DOSE(MIN) (INCREASED LEVEL LiFE OF EUM- INITIAL BEFORE EVERY MAXIMUM (MIN) (HRS) INATION DOSE HS X DAYS) DOSE Benzodiazepines Clonazepam 60-120 19-39 hepatic .125- 60-120 3-5 days 4mg (Klonopin) .25 mg min Anticonvulsants Carbamazepine 4-5 hrs variable hepatic 200 at hs 3-6 days 400 mg (Tegretol) 12-17 hrs with repeated dosing Gabapentin variable 5-7 hrs renal 100 at hs 3-6 days 1200 mg (Neurotonin) (100-400) CD/LD = carbidopa/levodopa, CR = controlled release. HS = hour of sleep(bedtime) *Toavoid augmentation Clinical Pearls I. Treatment of PLMS with clonazepam improves sleep quality (reduces arousals), but often does not reduce the frequency of leg movements. 2. Early morning somnolence, a common side effect of clonazeparn, may limit the usefulness of this medication. 3. Dopaminergic agents reduce the number of PLMs and improve sleep quality. Many clinicians consider these medications the treatment of choice. 4. Carbidopa/levodopa has a rapid onset of action, but a short duration. It may cause augmentation of symptoms in high doses. 5. The dopamine agonists pramipexole and ropinerol are less likely to cause nausea than pergolide and bromocriptine. 6. Pramipexole and ropinerol have a longer duration of action than regular preparations of carbidopa/levodopa. However, these dopamine agonists should be given 1.5 to 2 hours before bedtime (or symptom onset). REFERENCES I. Mitler, MM. Browman CPo Menn SJ.etal:Nocturnal myoclonus: Treatment efficacy ofclonazepam andtemazepam. Sleep1986; 9:385-392. 2. Peled R.Lavie P:Double-blind evaluation of clonazepam on periodic legmovements insleep. J Neurol Neurosurg Psych 1987; 50: 1679-1681. 3. Doghramji K.Browman CPo Gaddy JR.et al:Triazolam diminishes daytime sleepiness andsleep fragmentation inpatients with periodic legmovements in sleep. J ClinPsychopharmacol 1991; 11:284-290. 4. Practice parameters for the treatment of the restless legs syndrome and periodic limb movement disorder. Sleep 1999; 22:961-968. 5. Henning W.Allen RA. Early C. etal:Thetreatment ofrestless legssyndrome andperiodic limb movement disorder. Sleep1999; 22:970-999. 6. Montplaisir J. Nicolas A. Denesle R. et al: Restless leg syndrome improved by pramipexole: A double-blind randomized trial. Neurology 1999; 52:938-943. 7. OndoW: Ropinirole forrestless legssyndrome. Mov Disord 1999; 14:138-140. 8. Happe S. Klosch G. Saletu B. et al: Treatment of idiopathic restless leg syndrome with gabapentin. Neurology 200 1; 57:1717-1719. 9. Walters AS.Winkelmann J.Trenkwalder C. et al: Long-term follow-up on restless legs syndrome patients treated with opioids. Mov Disord 2001; 16:1105-1109. 268

PATIENT 83 A 72-year-old man with worsening symptoms during treatment for restless legs A 72-year-old man was diagnosed as having the restless leg syndrome (RLS) on the basis of a history of crawling sensation in both legs. These unpleasant sensations frequently started 30 minutes before bedtime when he was seated or recumbent in bed. The discomfort usually was relieved by moving his legs or walking, but the sensations returned once he became inactive. The patient was started on one-half pill of carbidopa/levodopa 25/100 mg, and the dose was slowly increased to three pills 90 minutes before bedtime over several weeks. This treatment initially resulted in good control of his symptoms. However, he began to notice severe RLS symptoms in the early evening that involved his arms as well as his legs. Physical Examination: Unremarkable. Neurologic: no involuntary movements, normal muscle tone, sensation intact. Question: What treatment do you recommend? 269

Answer: Discontinue carbidopa/levodopa and try a dopamine agonist. Discussion: Dopamine precursors (carbidopa/ levodopa [CD/LD]) and dopamine agonists (see table in Patient 82) are the most widely used agents for RLS and PLMS. CD/LD is short acting and can result in a rebound in symptoms in the early morning. The dose can either be repeated during the night or a continuous release form of the medication can be used. However, treatment of RLS/PLMS with CD/LD eventually results in some augmentation of symptoms in up to 80% of patients with RLS and in 30% of patients with PLMS and no RLS. The augmentation effect describes: (I) earlier onset of the usual bedtime symptoms (early evening or late afternoon), (2) increased severity of symptoms, and (3) spread ofsymptoms to the arms. Augmentation is more likely to occur at doses of more than 300 mg of levodopa. It may temporarily improve with an increase in dosage, but the symptoms eventually occur again. Augmentation usually responds to discontinuing the medication and switching to another agent. The usual approach is to wean CD/LD and switch to a dopamine agonist. Some clinicians wean CD/LD while starting the dopamine agonist (overlap), although this may not be necessary. Despite the problem of augmentation, CD/LD can be effective in mild or intermittent RLS. It has a more rapid onset of action, although a shorter half life, than dopamine agonists. Patients who get RLS symptoms only on long plane or car trips might find this medication useful. The dopamine agonist pramipexole (Mirapex) is effective for both RLS and PLMs, and is less likely to cause augmentation than CD/LD. Side effects include nausea (less than pergolide), postural hypotension, somnolence, headache, and (rarely) lower extremity edema. If augmentation occurs, it often can be handled by starting the medicine earlier in the evening. Side effects can be minimized by starting with a low dose (0.125 mg about 2 hours before bedtime or before typical symptom onset). The dose can be increased by 0.125 mg every 2-3 days to a maximum of 1.5 mg. Some patients find splitting the medication (e.g., one at 6 PM, one at 8 PM) effective if nausea or augmentation is a problem. Pramipexole is excreted by the kidneys and the dose should be decreased in patients with renal failure. Also, use the lowest dose possible. Ropinirole (Requip) has a slightly shorter time to peak level and half-life than pramipexole, and is cleared by hepatic metabolism. The usual starting dose is 0.25 mg 1-2 hours before bedtime or symptom onset, with dose increases every 2-3 days to a maximum of 3 mg. Side effects include nausea, postural hypotension, and sleepiness. In the present patient, carbidopa/levodopa was discontinued, and the early evening symptoms as well as the spread to his arms gradually ceased. The patient was begun on pramipexole 0.125 mg at 8 PM (typical symptoms started at 10 PM). The dose was increased over 1 week to to 0.25 mg at 8 PM, with resolution of symptoms. Clinical Pearls 1. Augmentation of RLS symptoms can occur while on treatment with carbidopa/ levodopa, but usually responds to a discontinuation of medication and a switch to a dopamine agonist. 2. Augmentation is more likely in patients with RLS than with PLMS alone. The lowest effective dose of carbidopa/levodopa should be used to decrease the risk of augmentation. 3. In patients with moderate to severe RLS, it may be prudent to begin treatment with a dopamine agonist rather than CD/LD, to avoid augmentation. 4. Alternative treatments for RLS include other dopamine agonists, opiates, anticonvulsants, and benzodiazepines. 5. Augmentation can occur with dopamine agonists, but is less likely than with dopamine precursors. 270

REFERENCES I. Monteplaisir J, Lapierre 0, Warnes H, et al: The treatment of the restless leg syndrome with or without periodic leg movements in sleep. Sleep 1992; 15:391-395. 2. Becker PM. Jamieson AO, Brown WD: Dopaminergic agents in restless leg syndrome and periodic leg movements of sleep: Response and complications of extended treatment in 49 cases. Sleep 1993; 16:713-716. 3. Walters AS, Wagner ML. Hening WA, et al: Successful treatment of the idiopathic restless leg syndrome in a randomized doubleblind trial of oxycodone versus placebo. Sleep 1993; 16:327-332. 4. Allen RP, Earley CJ: Augmentation of the restless leg syndrome with carbidopallevodopa. Sleep 1996; 19:205-213. 5. Earley CJ, Yaffee JB, Allen RP: Randomized, double-blind. placebo-controlled trial of pergoIide in restless legs syndrome. Neurology 1998; 51:1599-1602. 6. Montplaisir J, Nicolas A, Denesle R, et al: Restless leg syndrome improved by pramipexole: A double-blind randomized trial. Neurology 1999; 52:938-943. 7. Walters AS, Winkelmann J, Trenkwalder C, et al: Long-term follow-up on restless legs syndrome patients treated with opioids. Mov Disord 2001; 16:1105-1109. 271

PATIENT 84 A 58-year-old man with sleep apnea and leg jerks during sleep A 58-year-old man was diagnosed as having severe obstructive sleep apnea (GSA) on an initial study. He then underwent a nasal CPAP titration, after which he remarked that he had had the "best night of sleep in years." However, there was a drastic change in his PLM index on CPAP. Physical Examination: Vital signs: normal. HEENT: large tongue, dependent palate; 16-inch neck circumference. Chest: clear. Cardiac: normal. Extremities: no edema. Neurologic: sensation in extremities intact. Figure: A sample tracing is shown below. The airflow is flow signal from the CPAP machine in the sleep laboratory Sleep Studies» AHI PLM index PLM-arousal index Arousal index Diagnostic 66/hr IO/hr 5/hr 50/hr CPAP Titration (12 em H20) IO/hr 60/hr IO/hr 15/hr ,', All indices are the number of events per hour of sleep. Question: What is your diagnosis? C3-A2 01-A2 ROC-A1 LOC-A2 EKG chin EMG airflow chest abdomen R,L Legs 5002 272

Diagnosis: Periodic leg movements associated with nasal CPAP treatment of OSA. Discussion: A significant increase in periodic limb (leg) movements in sleep (PLMS) has been reported in patients with OSA following the initiation of nasal CPAP. The etiology of this change is unknown, but several possibilities have been suggested. One is that the severe fragmentation of sleep by apnea (pre-CPAP) did not allow manifestation of the PLMS, and treatment with CPAP unmasked the PLMS by allowing continuous sleep. Another is that the rebound in sleep following initial CPAP treatment results in less spontaneous patient movement, and the stasis or pressure on the nerves due to immobility leads to PLMS. Many patients spend more time supine when treated with CPAP than in the untreated state. Whatever the etiology, the clinician must decide whether to treat the PLMS or follow the patient's symptoms after nasal CPAP therapy is initiated. In some patients on nasal CPAP, the large increase in PLMS is transient. In others, the PLM-arousal index is low, and the impact on sleep is insignificant. In such cases, observation is adequate, and if symptoms of daytime sleepiness resolve, no additional treatment is necessary. However, if daytime sleepiness persists or returns after treatment of OSA is initiated, PLMS could be one cause of treatment failure. Note that the mere presence of PLMS should not exclude consideration of other causes of persistent daytime sleepiness such as inadequate pressure, poor compliance, inadequate sleep, and narcolepsy (which often coexists with PLMS). The patient's bedpartner should be questioned about the frequency of body movements while the patient is on nasal CPAP. Repeat sleep monitoring to determine the PLM-arousal index (with the patient on nasal CPAP) may be needed for clarification and to document that the level of CPAP is adequate. If treatment of PLMS is indicated, special treatment consideration is necessary in patients with PLMS and significant OSA. Benzodiazepines, a common treatment for PLMS, have the potential to worsen sleep apnea. The risk probably is small if the patient is on an adequate amount of nasal CPAP. However, the effect of benzodiazepines on the efficacy of nasal CPAP has not been specifically studied. In addition, there is always the possibility that the medication will be taken on nights when CPAP is not used. Therefore, dopaminergic treatment of PLMS with carbidopa/levodopa or dopamine agonists is the logical choice. Certainly, benzodiazepines are absolutely contraindicated in patients with hypoventilation and CO2 retention. In the present case, the PLM-arousal index was modest, and the patient noted a marked improvement in daytime sleepiness almost immediately after starting nasal CPAP. His wife reported that he did not kick at night and was "as cool as a cucumber." Good control of his symptoms persisted on nasal CPAP therapy, and no treatment for the PLMS was initiated. Clinical Pearls I. A large increase in PLMS can follow initiation of nasal CPAP therapy in OSA patients. 2. In many patients with OSA and PLMS, treatment of the OSA alone can result in a complete resolution of symptoms. PLMS treatment is usually not needed unless the restless legs syndrome is present. 3. When adequate treatment of OSA (nasal CPAP) fails to abolish symptoms of daytime sleepiness, PLMS may be one cause of persistent sleepiness. However, the mere presence of PLMS should not discourage the clinician from excluding other causes of persistent daytime sleepiness. 4. Benzodiazepines can worsen OSA. Dopaminergic agents should be used when treating PLMS or the restless legs syndrome in patients with sleep apnea. REFERENCES I. Fry JM. Diphillip MA. Pressman MR: Periodic leg movements in sleep following treatment of obstructive sleep apnea with nasal CPAP. Chest 1989; 96:89-91. 2. Yamashiro Y. Kryger MH: Acute effect of nasal CPAP on periodic limb movements associated with breathing disorders during sleep. Sleep 1994; 17:172-175. 3. Berry RB. Kouchi K. Bower J, et al: Effect oftriazolam in obstructive sleep apnea. Am J Resp Crit Care Med 1995; 151:450-454. 4. Guilleminault C. Phillip P: Tiredness and somnolence despite initial treatment of obstructive sleep apnea syndrome. Sleep 1996; 19:5 117-S 122. 5. Chervin RD: Periodic leg movements and sleepienss in patients evaluated for sleep-disordered breathing. Am J Respir Crit Care Med 2001; 164:1454-1458. 273

FUNDAMENTALS OF SLEEP MEDICINE 18 Narcolepsy Narcolepsy is a neurological disorder characterized by excessive daytime sleepiness (EDS) and symptoms related to the abnormal regulation of REM sleep (see table). Hallmarks of the disorder are a short REM latency and inappropriate intrusion of REM sleep physiology into wakefulness: cataplexy (loss of muscle tone during periods of high emotion), hypnagogic hallucinations (dreaming at sleep onset), and sleep paralysis (loss of muscle tone at sleep onset or on awakening). Some patients also develop episodes of automatic behavior (up to 30 minutes of semi-purposeful behavior with amnesia for interval) during periods of reduced alertness. The short periods of decreased vigilance may be experienced by the patient as a "blackout" and be mistaken for a seizure. Tetrad ofNarcolepsy SYMPTOMS ApPROXIMATE PREVALENCE Excessive sleepiness Cataplexy Hypnagogic hallucinations Sleep paralysis 100% 70% 66% 60% The prevalence of the disorder is 0.03-0.05% in the general population. The most common age of onset is adolescence, but a second peak occurs near 40 years of age. The disorder can also occur in children. About 10% of cases start before age 10 and 5% after age 50. Usually daytime sleepiness is the first symptom, followed in months to years by other symptoms. Cataplexy is the only symptom that is specific for narcolepsy, but it is present in only about 70% of cases. Patients with primary narcolepsy have a normal neurological examination. However, secondary or "symptomatic" narcolepsy can occur in patients with head trauma, stroke, multiple sclerosis, brain tumors, neurodegenerative disorders, and CNS infections. Narcolepsy was found to be associated with the presence of the human leukocyte antigens (HLA) DR2 and DQ6 in the Japanese population. About 95% of Caucasians with narcolepsy also are HLA-DR2 positive. However, the incidence of HLA-DR2 in African-American narcoleptics is lower. The antigen DQB I*0602 is the most sensitive marker for narcolepsy across all ethnic groups. Unfortunately, antigen testing is not that useful in making a diagnosis of narcolepsy because (I) most individuals positive for the antigen do not have narcolepsy, (2) antigen-negative cases of narcolepsy have been reported, and (3) the group of narcoleptics without cataplexy are less likely to be antigen positive. The latter group provides the greatest difficulty in making a diagnosis. The current thinking is that having certain genes may predispose a person to the development of narcolepsy. In 1998, two peptides secreted by the hypothalmus and similar to secretin were identified (hypocretin [HCRT] 1 and 2). Another group identified two peptides that bound to two G protein--coupled receptors on the hypothalamus. These proteins stimulated food intake and were called orexin A and B ("orexin" is Greek for "appetite"). Ultimately it was found that the hypocretin and oxrexin peptides were identical. The hypocretin-secreting cells are found only in the hypothalamus. They project to the hypothalamus but also to many other brain areas. Two major pathways have been identifed to the cortex and two to the brainstem. The descending pathways are to areas known to affect the sleep-wake cycle. One especially important projection is to the the locus ceruleus (LC). This is the location of norepinephrine-secreting neurons that are important in maintaining wakefulness. 274

HCRT I PEPTIDE HCRT 2 PEPTIDE Binding Affinities HCRT I RECEPTOR High Low HCRT 2 RECEPTOR High High Hypocretin knockout mice were found to have narcoleptic-like behavior (decreased HCRT production). Dogs with canine narcolepsy-cataplexy were found to have a mutation of genetic coding for the hypocretin-2 receptor. Thus, disorders that decrease either production of HCRT or abnormalities of the HCRT receptor could cause narcolepsy. These findings suggested an important genetic influence on narcolepsy. However, the inheritance of human narcolepsy is more complex that canine narcolepsy. For example, when looking at identical twins, when narcolepsy involves one of the pair, the other exhibits narcolepsy in only 25-30% of the cases. Thus, a combination of genetic susceptibility and some environmental trigger or infectious/autoimmune disorder may ultimately determine if narcolepsy develops. The link between the HCRT system and human narcolepsy has been further supported by the findings that seven of nine patients with narcolepsy had low HCRT-I levels in the cerebrospinal fluid. Another study of narcoleptic brains showed an absence of hypocretin-secreting neurons in the hypothalamus. Nishino et al have hypothesized that human narcolepsy involves a disruption of hypocretin neurotransmission. This could invol ve HCRT secretion, the number of HCRT receptors, or HCRT receptor function (for example an abnormal receptor or blocking antibodies). Physiologically, the loss of hypocretin stimulation of the locus ceruleus may be very important for development of the manifestations of narcolepsy. The LC cells are active during wakefulness but cease functioning during REM sleep or prior to and during cataplexy. Key Points I. Narcolepsy is a disorder associated with daytime sleepiness the intrusion of REM sleep physiology into wakefulness. 2. Narcolepsy is associated with certain HLA markers, but genetics alone cannot explain why some patients develop the disorder. 3. Dysfunction of the hypocretin system appears to playa key role in the development of narcolepsy. 4. Secondary narcolepsy can be associated with head trauma, stroke, multiple sclerosis, brain tumors, neurodegenerative disorders, and CNS infections. 5. Cataplexy, the only symptom specific for narcolepsy, is present in only 70% of patients. REFERENCES I. Mignot E, Hayduk R, Black J et al: HLA DQB I*0602 is associated with cataplexy in 509 narcoleptic patients. Sleep 1997; 20: 1012-1020. 2. de Lecea L, Kilduff TS. Peyron C. et al: The hypocretins: Hypothalamus-specific peptides with neuroexcitatory activity. Proc Natl Acad Sci USA 1998; 95:322-327. 3. Sakurai T, Amemiya A, Ishii M, et al: Orexins and orexin receptors: A family of hypothalamic neuropeptides and G protein-coupled receptors that regulate feeding behavior. Cell 1998; 92:573-585. 4. Nishino S. Ripley B, Overeem S, et al: Hypocretin (orexin) deficiency in human narcolepsy. Lancet 2000; 355:39-40. 5. Thannickal TC, Moore RY, Nienhuis R, et al: Reduced number of hypocretin neurons in human narcolespy. Neuron 2000; 27:469-474. 6. Brooks SN, Mignot E: Narcolepsy and idiopathic hypersomnia. In Lee-Chiong TL, Carskadon MA (eds): Sleep Medicine. Philadelphia, Hanley and Belfus. 2002, pp 193-20 l. 275

PATIENT 85 A 30-year-old man with daytime sleepiness and episodes of weakness A 30-year-old man was evaluated for excessive daytime sleepiness of 5-year duration. There was no history of snoring or observed apnea. The patient recalled feeling weak in the knees when he laughed or was embarrassed. The patient's wife reported that sometimes he kicked the covers at night. Rarely, the patient felt he could not move for awhile as he was falling asleep at night. Physical Examination: Normal. Sleep Study Time in bed Total sleep time Sleep period time (SPT) Sleep efficiency % Sleep latency REM latency Arousal index AHI 480 min (414--455) 350 min (400--443) 425 min (405--451) 73 (95-99) 10 min (2-10 min) 2.5 min (70-100) 20/hr l/hr « 5) Sleep Stages Stage Wake Stage 1 Stage 2 Stages 3 and 4 Stage REM PLM index PLM-arousal index %SPT 18 (0-3) 13 (2-9) 49 (50-64) 5 (7-18) 15 (20-27) 30/hr 10/hr ( ) = normal values for age. PLM = periodic leg movement Question: What is the likely diagnosis? 276

Diagnosis: Narcolepsy. Discussion: Narcolepsy is a disorder of unknown etiology that causes excessive daytime sleepiness and usually is associated with cataplexy as well as other phenomena linked to REM sleep. A relatively young age of onset (10-30 years old) is typical; in 70-80% of cases, symptoms start before age 25. Interestingly, 5% of cases start after age 50. The history of cataplexy (episodes of weakness preceded by high emotion such as laughter, surprise, or embarrassment) is strong evidence for narcolepsy. This symptom is the only member of the classic tetrad of narcoleptic symptoms (sleep attacks, cataplexy, hypnagogic hallucinations, and sleep paralysis) that is pathognomonic for a narcolepsy. Cataplexy is present in about 70% of patients with narcolepsy. The entire tetrad is present in only 10-15% of patients. Sleep paralysis is characterized by an inability to move while still awake at sleep onset (hypnagogic) or, less commonly, on awakening (hypnopompic). This symptom can occur in normal individuals, especially after periods of sleep deprivation. Hypnagogic hallucinations ("awake dreaming") are vivid sensory sensations occurring while awake at sleep onset (termed hypnopompic just after awakening). The sensations may include visual imagery (a person or animal in the room) or auditory hallucinations. Sleep attacks are sudden periods of irresistable daytime sleepiness. While this is a classic symptom of narcolepsy, many patients complain of sleepiness throughout the day. Sleep attacks also can be a symptom of other disorders, such as sleep apnea. Daytime naps are said to be more refreshing in narcolepsy than OSA. However, not all studies have documented this difference. The onset ofsleep attacks may precede cataplexy by several years (rarely, by as many as 40 years), making the diagnosis of narcolepsy more difficult. Additionally, even when present, episodes of cataplexy may be uncommon and subtle, so that obtaining an unequivocal history of cataplexy is not a simple task. Some characteristics of cataplexy may help the clinician differentiate this symptom from other episodes of a vague weak feeling or passing out. Consciousness is always maintained during attacks of cataplexy. The episodes rarely last more than afew minutes (not hours). The weakness of cataplexy is symmetric. although weakness may involve only the muscles of the neck or face (head bobbing forward, jaw dropping, or a facial droop). Other patients report buckling knees and falling. Episodes ofcataplexy can be terminated by hypnagogic hallucinations and then sleep. The frequency of cataplexy is highly variable, from daily to a few times per year. Polysomnography in narcolepsy usually reveals sleep fragmentation and often PLMS. The sleep latency is usually short, but problems with sleep maintenance are common. A short REM latency (time from sleep onset to the first REM sleep) of 20 minutes or less (termed sleep-onset REM), is the characteristic finding on polysomnography, although it is not always present (about 40-50% of the time). Again, this symptom can occur with other disorders, such as sleep apnea, depression, withdrawal ofREMsuppressing medication, and prior REM deprivation of any cause. A diagnosis of narcolepsy can be made by history in a patient with daytime sleepiness (daily lapses into sleep for at least 3 months) and unequivocal cataplexy. A nocturnal sleep study and multiple sleep latency test (MSLT) are still recommended for confirmation and to rule out other sleep disorders. If cataplexy is not present, the diagnosis of narcolepsy depends on a nocturnal polysomnogram to rule out other causes of daytime sleepiness (e.g., sleep apnea, PLMS), and demonstration of either a short nocturnal REM latency « 20 minutes) orby MSLT performed on the following day-a mean sleep latency < 5 minutes and two or more naps with REM sleep. An additional important requirement is that the physician exlude other factors (e.g., recent medication changes, poor sleep) that could explain the findings. (See Patient 86 for a detailed discussion of the use of the MSLT in the diagnosis of narcolepsy.) Of note, an MSLT showing a short sleep latency but fewer than two REM periods does not rule out narcolepsy. Patients with narcolepsy and cataplexy have only a 70-80% chance of having a diagnostic MSLT on a given day. In addition, patients with obstructive sleep apnea can have two or more REM onsets on an MSLT. Some patients with narcolepsy have an MSLT with two REM onsets and a slightly longer mean sleep latency (5-8 minutes). The sleep latency of patients with narcolepsy does tend to be shorter than patients with OSA (see figure, next page), although there is considerable overlap. The importance of a short nocturnal REM latency in making the diagnosis of narcolepsy is often overlooked. While present only 40-50% of the time, Aldrich et al found a REM latency < 20 minutes to have as high a positive predictive value for narcolepsy as an MSLT with a mean sleep latency < 5 minutes and two REM onsets. Again, you must rule out other explanations, such as OSA, for a short nocturnal REM latency. However, in that study only I% of patients with OSA had a nocturnal REM latency < 20 minutes. The present patient noted a young age of onset and had symptoms consistent with cataplexy and 277

sleep paralysis. The sleep study showed a short REM latency, a low sleep efficiency, reduced slow wave sleep, and PLMS. PLMS is not uncommon in narcoleptics and can disturb sleep enough to contribute to the symptoms of daytime sleepiness. However, PLMS does not typically result in a short REM latency. In the present patient, a PLM arousal index of lO/hr, while mildly disturbing sleep, is unlikely to be responsible for the severe symptoms. All of these findings are highly suggestive of narcolepsy. 4 8 12 16 20 Mean sleep latency (min) o -(]) .0 E :J Z en 140 (]) .- 120 "0 .2 100 en 80 60 40 20 O~LJII...fII-fIL.fIIL..fII-fIILJIL1--4'o4-i..j...l.4.-4-I-.f....4.-4-L.4-4-..l....\-I-l o - Narcolepsy c:::::J Sleep-related breathing disorder Mean Sleep Latency in Patients with Narcolepsy and Sleep-related Breathing Disorders Clinical Pearls 1. Cataplexy is the only pathognomonic symptom of narcolepsy. Unfortunately, sleep attacks may precede symptoms of cataplexy by several (rarely, many) years, and not all patients with narcolepsy have cataplexy. 2. Attacks of cataplexy are characterized by maintenance of consciousness and rarely last more than a few minutes. Weakness is symmetric, but may involve only the neck or facial muscles. 3. In the absence of unequivocal cataplexy, the diagnosis of narcolepsy depends on a sleep study to rule out other explanations for daytime sleepiness and demonstration of a short REM latency on the nocturnal sleep study and/or a diagnostic MSLT. 4. Neither a short nocturnal REM latency nor an MSLT meeting criteria for narcolepsy are very sensitive for the diagnosis. A short nocturnal REM latency is present only 40-50% ofthe time in patients with confirmed narcolepsy. An initial MSLT meeting criteria for narcolepsy is present only 70-80% of the time in patients with narcolepsy and cataplexy. 5. A short nocturnal REM latency and MSLT meeting diagnostic criteria for narcolepsy can occur in other disorders-especially OSA. REFERENCES I. Mosko SS. Shampain OS. Sassin JF: Nocturnal REM latency and sleep disturbance in narcolepsy. Sleep 1984; 7: 115-125. 2. Aldrich MS. Chervin RD. Malow BA: Value of the multiple sleep latency test for the diagnosis of narcolepsy. Sleep 1997; 20:620-629. 3. Thorpy M: Current concepts in the etiology. diagnosis and treatment of narcolepsy. Sleep Medicine 200 I; 2:5-17. 4. Guilleminault C. Anagnos A: Narcolepsy. In Kryger MH. Roth T. Dement WC (eds): Principles and Practice of Sleep Medicine. Philadelphia. WB Saunders. 2000. pp 676-686. 278

PATIENT 86 A 23-year-old man with daytime sleepiness, but no symptoms of cataplexy A 23-year-old man complained of severe daytime sleepiness present for at least 2 years. If the patient was able to take a nap, he usually awoke feeling refreshed. There was no history of cataplexy or hypnagogic hallucinations, and the patient denied snoring. However, he did remember a few episodes of sleep paralysis during which he was aware of having awakened but could not move for a few minutes. The patient provided a sleep log (diary) documenting at least 7 hours of sleep each night. A sleep study and multiple sleep latency test (MSLT) were ordered. Physical Examination: Normal. Sleep Study Time in bed Total sleep time Sleep period time (SPT) Sleep efficiency % Sleep latency REM latency Arousal index 440 min (430-454) 390 min (405-434) 425 min (410-439) 88 (91-99) 10 min (3-26 min) 40 min (78-99) 20/hr Sleep Stages Stage Wake Stage 1 Stage 2 Stages 3 and 4 Stage REM AHI PLM index %SPT 8 (0-1) 13 (3-6) 49 (40-51) 15 (16-26) 15 (22-34) O/hr« 5) O/hr ( ) = normal values for age, PLM = periodic leg movement MSLT Sleep Latency (min) Nap 1 3.5 Nap 2 2.0 Nap 3 3.5 Nap 4 1.0 Nap 5 4.0 ---------_.---------- - -------------. Mean 2.8 (min) Question: What is the cause of the patient's daytime sleepiness? REM Latency (min) 5 3 None 3 None 3 of 5 naps with REM 279

Diagnosis: Narcolepsy, on the basis of the MSLT. Discussion: The MSLT can help support the diagnosis of narcolepsy, but the characteristic findings are neither absolutely sensitive nor specific for this disorder (see table). The findings must be analyzed with the results of the previous nocturnal polysornnogram, the clinical history, and the patient's recent medication and sleep history in mind. The usual MSLT criteria for narcolepsy include a mean sleep latency < 5 minutes, documenting severe daytime sleepiness, and REM sleep present in two or more of five naps. However, narcoleptic subjects occasionally have two or more REM onsets and a slightly longer mean sleep latency (5-10 minutes). Only about 60-80% of patients with classic narcolepsy (excessive daytime sleepiness plus cataplexy) have a positive MSLT on anyone day. Therefore, a negative MSLT does not rule out the possibility of narcolepsy. In such cases, a history of unequivocal cataplexy still allows the diagnosis of narcolepsy. In cases where the polysomnogram does not suggest another cause for daytime sleepiness, and the MSLT documents a short sleep latency but insufficient REM periods, there are three major possibilities: narcolepsy, idiopathic hypersomnolence, and mild OSA/upper airway resistance syndrome. This scenario assumes that insufficient sleep (during the sleep study or in the prior week) and acute withdrawal of stimulants are not responsible for the short sleep latency on the MSLT. In a study by Aldrich et al (see reference 3), a diagnosis of narcolepsy was made in patients with no cataplexy by repeating the MSLT. Unfortunately, this may not always be possible for financial considerations. Additionally, an MSLT meeting criteria for narcolepsy is not specific for this diagnosis. REM onsets are seen in other sleep pathology, such as obstructive sleep apnea. While a much lower percentage of patients with OSA have an MSLT meeting criteria for narcolepsy, these patients make up the majority of patients studied in sleep laboratories. This is why the preceding nocturnal polysomonography is absolutely essential to rule out sleep apnea or sleep disturbance (decreased REM sleep) that may alter the REM latency. It is recommended that sleep apnea be treated before performing an MSLT (see Patients 90 and 91). A sleep diary should reveal if recent sleep deprivation caused a false positive MSLT. A drug history (and often a urine drug screen) helps clarify if medication or medication withdrawal altered the MSLT results. All drugs affecting sleepiness or REM sleep should be withdrawn for 2-3 weeks before testing, if possible. In the present case, the patient reported sleep attacks and sleep paralysis, but neither of these symptoms is specific for narcolepsy. The sleep study showed a slightly shortened REM latency, but not as short as is typical in narcolepsy. Neither sleep apnea nor PLMS was recorded, and the amount of sleep (and REM sleep) was adequate. Thus, the sleep study did not indicate a reason for daytime sleepiness or REM onsets. The MSLT met criteria for narcolepsy, as the patient exhibited severe sleepiness (sleep latency < 5 minutes), and three of five naps included REM sleep. The sleep log and medication history showed no reasons to suspect another cause for these findings. Thus, a diagnosis of narcolepsy was well-supported, despite an absence of cataplexy. Some have suggested that narcolepsy should be subdivided into a classic syndrome with cataplexy and a syndrome without cataplexy (excessive daytime sleepiness and positive MSLT). However, at the present time both are treated in the same manner. MSLT Findings in Narcolepsy and Sleep-Disordered Breathing NARCOLEPSY NARCOLEPSY WITH WITHOUT CATAPLEXY CATAPLEXY SLEEPDISORDERED BREATHING Polysomnograph: SOREM period 33% 24% 1% MSLT: A: 2+ SOREM periods 74% 91% 7% B: Mean sleep latency 87% 81% 39% < 5 minutes Both A and B 67% 75% 4% SOREM = sleep-onset REM Data from Aldrich MS, Chervin RD, Malow BA: Value of the multiple sleep latency test for the diagnosis of narcolepsy. Sleep 1997; 20:620-629. 280

Clinical Pearls 1. MSLT testing can help support a diagnosis of narcolepsy. However, a negati ve test does not eliminate the possibility that narcolepsy is present. 2. A positive MSLT is not specific for narcolepsy and must be interpreted in light of information from the prior nocturnal polysomnogram, a medication history, and the recent pattern and amount of sleep. 3. The MSLT criteria for narcolepsy are a mean sleep latency < 5 minutes and REM sleep present in two or more of five naps. REFERENCES I. American Sleep Disorders Association: The clinical use of the multiple sleep latency test. Sleep 1992; 15:268-276. 2. Bassetti C. Aldrich MS: Narcolepsy. Neurol Clin 1996; 14:545-569. 3. Aldrich MS. Chervin RD. Malow BA: Value of the multiple sleep latency test for the diagnosis of narcolepsy. Sleep 1997; 20:620-629. 4. Chervin RD. Aldrich MS: Sleep-onset REM periods during multiple sleep latency testis in patients evaluated for sleep apnea. Am J Resp Crit Care Med 2000; 161:426-431. 281

PATIENT 87 A 25-year-old man with narcolepsy who is still sleepy on medication A 25-year-old man was diagnosed with narcolepsy on the basis of symptoms of excessive daytime sleepiness and cataplexy. A polysomnogram revealed no apnea or periodic leg movements in sleep. A multiple sleep latency test showed a mean sleep latency of 3 minutes and REM sleep in two of five naps. The patient was started on methylphenidate (Ritalin) 10 mg tid, but still had severe daytime sleepiness in the early afternoon. He filled out a sleep diary which showed that some nights he only slept for 6 hours due to late bedtimes. He also admitted to taking an additional pill at night to allow him to study into the latenight hours. Physical Examination: Normal. Question: What treatment would you recommend? 282

Answer: Longer sleep time, improved sleep hygiene, and increased medication-especially before the most sleepy periods of the day. Discussion: The treatment of narcolepsy can be divided into treatment of daytime sleepiness and treatment of cataplexy/hypnagogic hallucinations. The treatment of daytime sleepiness begins with evaluation of the polysomnogram to determine if arousals from PLMS or sleep apnea could be worsening daytime sleepiness. Next, sleep hygiene and the amount of sleep must be optimized. Regular bedtime and an adequate sleep period are essential. Any sleep disturbance magnifies symptoms of narcolepsy. Those patients who find a short nap restorative may benefit from regularly scheduled naps during the day. A number of stimulant medications, including indirect sympathomimetics, are available to treat the daytime sleepiness of narcolepsy by increasing the synaptic availability of norepinepherine and dopamine (see table). Adequate control of daytime sleepiness can be attained in about 60-80% of patients. Methamphetamine, dextroamphetamine, and methylphenidate appear to be the more efficacious medications. Methylphenidate is less expensive, has a shorter half-life, and is the preferred stimulant for moderate-to-severe narcolepsy. Milder cases may be treated with pemoline, which is not a schedule II medication and often is better tolerated than the other stimulants. It has a long half-life, and the once-a-day dosing may also increase compliance. However, pemoline is no longer a first-line agent because it has been associated with severe liver damage. There are several potential problems with the use of stimulant medications. First, tolerance may develop, requiring escalating doses and leading to ineffectiveness at the highest dose. In some patients, effectiveness can be restored by a "drug holiday"- no medications for several days. Unfortunately, severe sleepiness may occur during that time. Second, the medications can increase blood pressure, although this effect is not usual in normotensive patients. Third, side effects ofstimulants include nervousness, irritability, headache, decreased appetite, and insomnia. Thus, they should not be taken near bedtime, especially methamphetamine and dextroamphetamine, both of which have a relatively long half-life. Fourth, attacks of paranoia or hallucinations have been reported with amphetamines, but major psychiatric side effects are rare in the absence of underlying psychiatric disorders. The peak action of dextroamphetamine, methamphetamine, and methylphenidate is 1-3 hours from ingestion, so the medications should be taken at least I hour before the time of desired effectiveness. If a sleep attack has begun before medication is taken, then a nap may be the best treatment in some patients. Also note that methylphenidate has a much shorter half-life than the amphetamines and must be taken several times a day (bid to tid). However, methylphenidate appears less likely to produce side effects than the amphetamine drugs. All of these medications are schedule II drugs. Modafinil is a new wake-enchancing nonstimulant medication that is now available in the United States. Its mechanism of action is not known, but the action is not specific to patients with narcolepsy. It has been shown to be effective at decreasing subjective and objective measures of sleepiness in narcolepsy in double-blind placebo-controlled studies. There is no evidence that tolerance develops or that the drug impairs sleep quality. It has a number of advantages including once daily dosing, low abuse potential, and the fact that it is not a schedule II medication. While generally felt to have fewer side effects that methylphenidate and other stimulants, this has never been proven in a randomized comparison. In addition, modafinil has not been shown to be more effective than stimulants. The general impression of most clinicians is that it may be less effective than methylphenidate. The drug is fairly expensive (thirty 200-mg tablets cost about $150). Unlike the indirect sympathomimetics, withdrawal of modafinil does not result in a rebound of REM and slow wave sleep (see Patient 89 for details). Another alternative that can be tried in patients not tolerating stimulants is the irreversible MAO type B inhibitor selegiline. At doses of 10--40 mg/day, this drug has been shown to improve narcoleptic symptoms. Unfortunately, at doses over 20 mg/day it loses its MAO inhibitor selectivity, and a low tyramine diet is indicated to avoid the risk of hypertensive reactions. The drug also has anticataplectic activity in addition to its alerting ability. Patients who experience intolerable side effects with other agents may benefit from this medication as long as they are willing to adhere to a low tyramine diet. In the present case, the problem was not side effects but persistent sleepiness. The dose of methylphenidate was changed to 10 mg every morning, 20 mg before lunch, and 10 mg at 4 PM, with improvement in early afternoon symptoms. As an alternative, on nonworking days the patient was encouraged to take an early-afternoon nap. He also was encouraged to get 7-8 hours of sleep each night and to avoid any stimulant medication after supper. 283

Medications Used to Treat Excessive Daytime Sleepiness MAXIMUM BRAND DOSE SELECTED DRUG NAME DOSE (DAILY) HALF-LIFE SIDE EFFECTS Pemoline Cylert 18.75-37.5 mg qd 150 mg 12 hrs Hepatitis. liver (18.75.37.5 mg (adults) failure tabs) Methylphenidate" Ritalin 5-10 mg bid or tid 100 mg 2-4 hrs Nervousness. trem- (5, 10,20 mg tabs) ulousness, headache, palpitations Dextroamphetamine" Dexedrine 5 mg qd to bid 60mg 10-30 hrs Nervousness, tremDextrostat (5, 10 mg tabs) ulousness, headOthers ache. palpitations Methamphetamine* Desoxyn 5 mgqd 60 mg 12-34 hrs Nervousness, trem- (5, 10 mg tabs) ulousness, headache, palpitations Selegiline** Eldepryl 10-20 mg 20mg 9-14 hrs Nausea. dizziness, (5-10 mg bid) confusion, dry mouth Modafinil' Provigil 200 or 400 mg qd 400 mg 10-12 hrs Headache, drug (100,200 mg tabs) interactions, nervousness *Schedule II medication **Low tyramine diet (MAOI) 'Only modafinil isFDA-approved asa narcolepsy treatment. Clinical Pearls I. With proper dose titration, stimulant medication can control sypmptoms of daytime sleepiness in 60-80% of narcoleptic patients. 2. Non-pharmacological measures such as good sleep hygiene, adequate sleep, and daytime naps are an essential part of treatment of daytime sleepiness in patients with narcolepsy. 3. It is essential to plan dosing according to the time profile of symptoms and to avoid medication-induced insomnia or sleep disturbance. 4. The largest doses should be taken 1-2 hours before the periods of maximum sleepiness. 5. If patients fail to respond to treatment, the coexistence of other sleep disorders such as obstructive sleep apnea should be considered. 6. Modafinil may be better tolerated than stimulants in some patients with narcolepsy. REFERENCES I. Mitler M. Aldrich MS, Koob GF, et al: ASDA standards of practice: Narcolepsy and its treatment with stimulants. Sleep 1994; 17:352-371. 2. Mitler M, Harsch J, Hirshkowitz M, et al: Long-term efficacy and safety of modafinil forthe treatment of excessive daytime sleepiness associated with narcolepsy. Sleep 1994; 17:352-371. 3. Bassetti C,Aldrich MS: Narcolepsy. Neurol Clin 1996; 14:545-569. 4. U.S. Modafinil inNarcolepsy Study Group: Randomized trial of modafinil for the treatment of pathological somnolence innar- colepsy. Ann Neurol 1998; 43:88-97. 5. U.S. Modafinil inNarcolepsy Study Group: Randomized trial of modafinil asa treatment forthe excessive daytime somnolence of narcolepsy. Neurology 2000; 53:1166-1175. 6. Littner M, Johnson SF, McCall MV, etal,Standard of Practice Committee ofthe AASM: Practice parameters forthe treatment of narcolepsy: An update 2000. Sleep 200 I; 24:451--466. 284

PATIENT 88 A 25-year-old man with frequent episodes of cataplexy A 25-year-old man who was diagnosed with narcolepsy experienced almost daily episodes of cataplexy that commonly were associated with laughter or surprise. At these times, the patient felt weakness in his legs; a few times he almost fell to the ground. The episodes varied in frequency, but were increased after periods of irregular sleep. In addition to cataplexy, the patient also was troubled by hypnagogic hallucinations, which consisted of seeing a stranger in the room as he was falling asleep. While he realized the imagery was not real, the patient sometimes felt quite anxious about the episodes. He was treated with imipramine (a tricyclic antidepressant), with decreases in the frequency of the episodes of cataplexy and hypnagogic hallucinations. However, at a dose that controlled his cataplexy, he found the medication intolerable secondary to side effects (especially dry mouth). When he abruptly stopped the medicines, he noted a significant increase in the episodes of cataplexy. Physical Examination: Normal. Question: What treatment would you recommend? 285

Answer: A trial of fluoxetine or other selective serotonin reuptake inhibitor. Discussion: Cataplexy is the only manifestation of narcolepsy that is virtually pathognomonic for this disorder. This symptom is characterized by sudden bilateral loss of muscle tone at moments of emotion (e.g., surprise, laughter, anger, embarrassment). The severity varies from minor facial drooping with eye closure, slurred speech, and a sagging jaw to loss of postural muscle tone and falling. The loss of muscle tone is not always instantaneous; it may progress over a few minutes. Duration of a cataplectic episode usually is about I minute, but it can last up to 20 minutes in the rare patient. During cataplexy, consciousness is preserved. However, the patient can enter a period of sleep and dream during the attacks, or can have hallucinations. The frequency of cataplectic episodes varies widely between patients with narcolepsy. In general, patients with the most severe cataplectic episodes tend to have them frequently, and they are very disturbing. In some patients, cataplexy is infrequent and requires no treatment. Cataplexy typically is treated more successfully than daytime sleepiness. Hypnagogic hallucinations are images occurring while the patient is still awake at sleep onset or on awakening (hypnopompic). They usually are visual, but may include or feature exclusively other senses, such as smell and hearing. Images may be simple or complex and bizarre. Patients typically know the hallucinations are not real, but are still quite frightened. The most common hypnagogic hallucination is a stranger or animal in the room. The usual treatment for cataplexy is low doses of tricyclic antidepressants. Protriptyline (5-30 mg daily) is a non sedative medication, but it is commonly associated with anticholinergic side effects. Urinary retention is a typical side effect in older patients. Imipramine (Tofranil) (25-200 mg daily) and clomipramine (75-125 mg daily) are other effective tricyclics. Successful treatment of cataplexy with serotonin reuptake inhibitors in their usual antidepressant doses has also been reported. For example, fluoxetine (Prozac), a selective serotonin reuptake inhibitor (20-80 mg daily) was effective. However, fluoxetine can cause sleep disturbance, thereby worsening daytime sleepiness. Venlafaxine, a reuptake inhibitor of both serotonin and norepinephrine, has also been reported to be effective treatment of cataplexy. All of these medications tend to suppress REM sleep, but the exact mechanism by which they suppress cataplexy is not clear. Recently, carbamazepine (Tegretol) was reported to decrease cataplexy in a patient who did not respond to the traditional medications. In that study, a dose of 200 mg twice daily was effective. The medications used to 286 treat cataplexy also suppress hypnagogic hallucinations and sleep paralysis; however, treatment of these two symptoms usually is not required. Abrupt withdrawal of medications used to treat cataplexy can result in an exacerbation of this symptom. A form of continuous cataplectic attacks (status cataplecticus) has been reported after abrupt cessation of treatment. The use of alpha-I adrenergic blockers (e.g., prazosin) also has been reported to worsen cataplexy. Therefore, these medications should be avoided in patients with narcolepsy. Guilleminault and coworkers recently reported that patients switched from amphetamines to modafinil sometimes had an increase in cataplexy. Evidently, amphetamines have some activity against cataplexy. The addition of venlafaxine to modafinil worked well in these patients. Gamma hydroxybutyrate (GHB), also know as sodium oxybate, has been shown to decrease cataplexy and increase sleep continuity in patients with narcolepsy. The drug does not acutely decrease daytime sleepiness. However, by improving nocturnal sleep, alertness may improve with chronic use of the medication. Unfortunately, the fact that this drug is abused (the "date-rape drug") has delayed approval for use in the United States. The drug has a short halflife and is usually taken on retiring and 2.5-4 hours later. Nightly total doses of 3, 6, and 9 grams were used in a large randomized trial. Nausea, headache, dizziness, and enuresis were the most common side effects. The abuse of GHB has been associated with a number of important CNS adverse clinical events, including seizure, respiratory depression, and profound decreases in level of consciousness, with instances of coma and death. Even at recommended doses, use has been associated with confusion. Orphan Medical has received U.S. FDA permission to market a liquid oral form of GHB (Xyrem). The drug is a Schedule III controlled substance. In addition, its distribution is governed by the FDA's subpart H regulations. To comply with these regulations, the company has developed a rigorous system that makes Xyrem available to patients from a single, specialty pharmacy. Both physicians and patients must receive an education program from the company before obtaining Xyrem. The company plans to release the medication in the last part of 2002. The indication will be treatment of cataplexy, but current studies are underway to evaluate the drug's long-term effects on daytime sleepiness. In the current patient, treatment was begun with f1uoxetine 20 mg daily. On this medication the frequency of cataplexy was significantly reduced. The patient tolerated the medication well and was satisfied with the treatment.

Clinical Pearls 1. Specific treatment for cataplexy may not be required in all patients with narcolepsy. 2. An increase in the frequency of cataplectic episodes can occur after withdrawal of medications treating this symptom or after initiation of treatment of other disorders with alpha-l adrenergic receptor antagonists (e.g., prazosin). 3. While the traditional treatment of cataplexy has been low doses of tricyclic antidepressants, treatment with serotonin reuptake inhibitors (fluoxetine) in the usual antidepressant dosage also has proven effective. 4. Carbamazepine can be efficacious when cataplexy does not respond to more conventional treatments. 5. Sodium oxybate may soon be released and will provide an alternative treatment for cataplexy with the additional benefits of improving nocturnal sleep. REFERENCES I. Aldrich M, Rogers AE: Exacerbation of human cataplexy by prazosin. Sleep 1989; 12:254-256. 2. Frey J. Darbonne C: Fluoxetine suppresses human cataplexy. Neurology 1994; 44:707-709. 3. Guilleminault C, Gelb M: Clinical aspects and features of cataplexy. Adv Neurol 1995; 67:65-77. 4. Vaughn BV, D'Cruz OF: Carbamazepine as a treatment for cataplexy. Sleep 1996; 19:101-103. 5. Guilleminault C. Aftab FA, Karadeniz D, et al: Problems associated with switch to modafinil-A novel alerting agent in narcolepsy. Eur J Neurol 2000: 7:381-384. 6. U.S. Xyrem Multicenter Study Group: A randomized, double blind, placebo-controlled multicenter trail comparing the effects of three doses of orally-administered sodium oxybate with placebo for treatment of narcolepsy. Sleep 2002;25:42-49. 287

PATIENT 89 A 24-year-old man with narcolepsy who was irritable and lost weight on methylphenidate A 24-year-old man was diagnosed as having narcolepsy on the basis of a sleep study showing no evidence of sleep apnea and an MSLT showing a short mean sleep latency (3.5 minutes) and three REM onsets in five naps. His original Epworth score was 18/24, which was reduced to 13/24 on methylphenidate 20 mg tid. Lower doses were not as effective. On this dose of medication he became quite irritable and had many fights with his wife. He also noted a dramatic decrease in his appetite and lost 15 pounds even though he was quite thin before starting medication. Question: What treatment would you recommend? 288