Sleep Medicine Pearls 2nd

Answer: Thyroid replacement and nasal CPAP or an oral appliance. Discussion: Hypothyroidism has been demonstrated to cause or exacerbate obstructive sleep apnea. While some sleep centers routinely obtain thyroid studies in all patients with suspected OSA, a recent study found hypothyroidism in less than 1% of these screened patients. Therefore, routine thyroid studies probably are indicated only ifsymptoms and signs are suggestive of hypothyroidism. Thyroid studies also may be indicated if the subsequent sleep study provides no explanation for sleepiness and fatigue, or if patients with OSA do not respond to treatment. The group of OSA patients at highest risk for coexistent hypothyroidism is older women. While hypothyroidism is present in a low percentage of patients with OSA, a much higher percentage of patients with known hypothyroidism have OSA. One study of patients newly diagnosed with hypothyroidism revealed OSA in 9 of II patients. Treatment of hypothyroidism may dramatically improve OSA. In one study, the mean AHI fell from 78/hr to l2/hr after patients became euthyroid. However, restoration of the euthyroid condition in patients with OSA does not reliably reverse sleep apnea in all patients. Moreover, initiation of even low doses of thyroid replacement in untreated patients with OSA and coronary artery disease has been reported to cause nocturnal angina. Therefore, treatment ofOSA should be begun when thyroid replacement is initiated. After the euthyroid state is attained, a repeat sleep study can determine if continued treatment of OSA (other than thyroid replacement) is required. Some patients do, in fact, experience complete reversal of sleep apnea following adequate treatment of hypothyroidism, even if body weight remains constant. The reason hypothyroidism exacerbates OSA is unclear and possibly multifactorial. Upper airway muscle myopathy, narrowing of the upper airway by mucoprotein deposition in the tongue (macroglossia), and abnormalities in ventilatory control are possible mechanisms. Treatments for snoring or mild OSA include weight loss, avoidance of the supine sleeping position, upper airway surgery (LAUP for snoring, UPPP for snoring or mild OSA), an oral appliance, or nasal CPAP. Although nasal CPAP is rarely used for isolated snoring, it can be quite effective (see figure); increased pressure can eliminate snoring (S). nasal mask pressure 5sec -, S S S S S Increased CPAP In the present patient, the sleep study revealed mild to moderate OSA and heavy snoring, with an overall AHI of 15/hr. An elevated thyroid-stimulating hormone of 18 mIU/L « 6 is normal) and a low free T4 documented primary hypothyroidism. Treatment was begun with low doses of thyroid replacement. The patient did not want surgery or an oral appliance. She underwent a CPAP titration and began nasal CPAP treatment with 5 em H20. The thyroid replacement was gradually increased until the euthyroid state was attained. A repeat sleep study (off nasal CPAP) several months later revealed an AHI < 5/hr and minimal snoring. Nasal CPAP therapy was discontinued, and the patient's snoring remained minimal. Clinical Pearls 1. Hypothyroidism is a predisposing condition for the development of OSA and should be considered in all patients with OSA. 2. Routine thyroid screening may not be cost-effective in all patients with suspected OSA. 3. Restoration of the euthyroid state does not eliminate sleep apnea in all patients with OSA and hypothyroidism. 4. Treatment of both OSA and hypothyroidism should be initiated. A repeat sleep study several months after the euthyroid state is restored determines if continued OSA treatment (other than thyroid replacement) is necessary. 5. Although nasal CPAP is less commonly used for simple snoring or mild OSA, it can be effective. 6. Snoring while using nasal CPAP treatment is a clue that the prescribed pressure should be increased. 189

REFERENCES I. Berry RB, Block AJ: Positive nasal airway pressure eliminates snoring as well as obstructive sleep apnea. Chest 1984: 85: 15-20. 2. Rajagopal KR, Abbrecht PH, Derderian 55, et al: Obstructive sleep apnea in hypothyroidism. Ann Intern Med 1984; 101: 491-494. 3. Rauscher H, Formanek D, Zwick H: Nasal continuous positive airway pressure for non-apneic snoring? Chest 1995; 107:58-61. 4. Winkelman JW, Goldman H, Piscatelli N, et al: Are thyroid function studies necessary in patients with suspected sleep apnea? Sleep 1996; 19:790-793. 190

PATIENT 57 A 50-year-old man with severe hypertension A 50-year-old-man with a history of severe hypertension (previous systolic blood pressure 180-190 mmHg) was admitted to the intensive care unit (lCU) when his physician noted a blood pressure of 2301130 mmHg in the office. At the time of admission the patient was being treated with lisinopril and amlodipine for his hypertension, but he had run out of medication. During the first night in the ICU the patient had periods of obvious obstructive apnea and swings in blood pressure. He adamantly denied symptoms of daytime sleepiness. Physical Examination: Height 5 feet 11 inches, weight 220 pounds. Blood pressure 180/95 mmHg. HEENT: edematous soft palate and uvula; 16-inch neck circumference. Chest: clear. Cardiac: S4 gallop. Extremities: 1+ edema, right arterial line in place. Laboratory Finding: EKG: left ventricular hypertrophy. Figure: Airflow and arterial blood pressure were recorded on a two-channel chart during the night in the ICU. Question: Should this patient be treated for a sleep disorder even though he denies daytime sleepiness? 200 Blood pressure (mm Hg) o 20 sec Airflow 191

Diagnosis: Obstructive sleep apnea and nocturnal worsening of systemic hypertension. Discussion: Reversal of excessive daytime sleepiness is not the only reason patients with significant OSA should be treated. Retrospective studies have suggested that when the apnea index is greater than 20/hr, untreated OSA is associated with a decreased survival rate. Further, this decrease is not secondary to links with other disorders, such as obesity and hypertension, because effective treatment of OSA resulted in a normal cumulative 5-year survival. While prospective studies of the impact of OSA and CPAP treatment on life expectancy are needed, it seems likely that untreated, significant OSA does shorten survival. The question is: How? One possibility is that OSA causes or worsens the known morbidity and mortality associated with systemic hypertension. In normal subjects, heart rate and systemic blood pressure fall during NREM sleep. Patients with OSA have increases in heart rate and systemic and pulmonary arterial blood pressure following apnea termination during sleep. Depending on the frequency of apnea during the night, many patients with OSA (with and without daytime hypertension) fail to show a mean fall in blood pressure during the night ("non-dippers"). The etiology of the post-apnea surges in blood pressure is complex, but activation of the autonomic nervous system (increased sympathetic activity) and arousal probably are the major causes. There is conflicting evidence about whether hypoxia also plays a role. Most hypertensive patients without OSA still have a dip in blood pressure during sleep. Conversely, many hypertensive and nonhypertensive patients with OSA are non-dippers (no sleep-associated fall in blood pressure). Studies have suggested that patients with both daytime and nocturnal hypertension (non-dippers) appear to have an increased risk of developing left ventricular hypertrophy. Thus, OSA may worsen the impact of hypertension on the heart and perhaps the peripheral vasculature. Does OSA cause daytime hypertension? A canine model showed that nocturnal hypertension could be induced by periodic acoustic arousal to mimic the periodic arousals of OSA. However, only nocturnal intermittent airway occlusions to simulate OSA resulted in both nocturnal and daytime hypertension. This suggests a role for intermittent hypoxia or airway occlusion as one cause of daytime hypertension. Chronic intermittent hypoxia given to rodents has been shown to increase 192 unstimulated (resting) blood pressure. While some epidemiologic studies have linked daytime hypertension to OSA, others have not. The biggest problem is separating out the coexisting factors of age and obesity. Certainly there is a 50-60% incidence of hypertension in most series of patients with OSA. Peppard and coworkers found a dose-response relationship between the AHI at baseline and the presence of hypertension at a 4-year followup. Other cross-sectional studies have found only a modestly increased risk of hypertension as AHI increases. However. even if OSA does not cause systemic hypertension (only an association), as noted above, it may worsen the consequences. Can treatment of OSA favorably alter the impact of hypertension? Several studies have shown a reduction in nocturnal blood pressure on nasal CPAP in OSA patients with daytime hypertension. In some patients, daytime blood pressure also may improve, although most patients still require treatment of hypertension with medication. Pepperell and coworkers found that the 24-hour mean arterial blood pressure fell by 3.3 mmHg in a group of moderate to severe patients with OSA using nasal CPAP. Both daytime and nighttime blood pressure fell. Based on prospective studies of the effects of hypertension, the authors reasoned that a fall in blood pressure of 3.3 mmHg would be expected to be associated with a stroke risk reduction of 20%, and a coronary artery disease event reduction of about 15%. The patients who used nasal CPAP more than 5 hours per night, those with an AHI > 40/hr, and those on antihypertensive medications had the largest falls in mean blood pressure. Whether nasal CPAP treatment favorably alters the development of left ventricular hypertrophy remains to be determined. In the present case, the chart recorder documented numerous apneas associated with a surge in blood pressure after apnea termination. The patient initially refused a complete sleep study. However, after discussion about the cardiovascular consequences and increased mortality associated with untreated sleep apnea, he agreed. A split-night study showed an AHI of 80/hr. Treatment with nasal CPAP at 12 cm HzO reduced the AHI to 8/hr. After a month of treatment with nasal CPAP, the patient reported an improved energy level. While antihypertensive therapy was still needed, improved control was noted, with systolic blood pressures in the 140-150 mmHg range.

Clinical Pearls I. Reversal of daytime sleepiness is not the only reason patients with significant GSA should be treated. 2. GSA prevents the normal sleep-associated fall in systemic blood pressure. 3. The presence of GSA appears to increase the risk of having systemic hypertension. However, even if GSA is only an association rather than a cause of hypertension, GSA likely worsens the severity or the consequences of hypertension in many patients. 4. Effective treatment of GSA prevents the cyclic nocturnal increases in blood pressure and may improve daytime blood pressure control and/or the long-term consequences of hypertension. REFERENCES J. He J, Kryger MH. Zorick Ff, et at: Mortality and apnea index in obstructive sleep apnea. Chest 1988; 94:9-14. 2. Verdecchia P, Schiallica G. Guerrier M. et al: Circadian blood pressure changes and left ventricular hypertrophy in essential hypertension. Circulation 1990; 81:528-536. 3. Suzuki M. Guilleminault G. Otsuka K. er al: Blood pressure "dipping" and "non-dipping" in obstructive sleep apnea syndrome patients. Sleep 1996; 19:382-387. 4. Brooks D. Horner RL. Kozar LF. et al: Obstructive sleep apnea as a cause of systemic hypertension. Evidence from a canine model. 1 Clin Invest 1997; 99:106-109. 5. Peppard PE. Young T. Palta M. et al: Prospective study of the association between sleep-disordered breathing and hypertension. N Engl1 Med 2000; 342: 1378-1384. 6. Fletcher EC: Invited review: Physiological consequences of intermittent hypoxia: systemic blood pressure. 1 Appl Physiol 2001; 90: 1600-1605. 7. Pepperell1CT. Ramdassingh-Dow S. Crosthwaite N. et al: Ambulatory blood pressure after therapeutic and subtherapeutic nasal continuous positive airway pressure for obstructive sleep apnea: a randomized parallel trial. Lancet 2002;359:204-210. 8. Peker Y. Hedner 1. Norum 1. et al: Increased incidence of cardiovascular disease in middle-aged men with obstructive sleep apnea: A 7-year follow-up. Am 1 Respir Crit Care Med 2002; 166: 159-165. 193

PATIENT 58 A 55-year-old man with premature ventricular contractions during sleep A 55-year-old man was undergoing coronary angiography. Before the procedure he was given midazolam (a potent benzodiazepine), and he fell asleep. During this time, heavy snoring and pauses in breathing were noted. The patient was referred for sleep evaluation. His angiogram had shown significant threevessel coronary artery disease. Physical Examination: Blood pressure 135/88 mmHg, pulse 80 and regular. HEENT: large uvula, edematous pharynx; 16-inch neck circumference. Chest: clear. Cardiac: S4 gallop. Extremities: no edema. Sleep Study: AHI 40/hr. Minimum arterial oxygen saturation: 85% during NREM, 75% during REM. Figure: A sample tracing from the sleep study is shown. Question: Are the premature ventricular contractions (PVCs) being caused by obstructive sleep apnea? 194 C 4 - A 1 O 2 -A 1 ROC-A 1 LOC - A 2 chin EMG EKG Airflow Chest Abdomen l"''1' .,.* ,~"..., .• ...,. ,Ill ..... "'....11 '.~ot\N"'''.'' \"1~"~!"~~~1 IlIIf 'l~ ~~~;~~~

Diagnosis: Severe obstructive sleep apnea is present, but frequent and unifocal PYCs are unrelated. Discussion: In normal individuals, the heart rate falls during NREM sleep. This is thought to be due to parasympathetic predominance during sleep. In patients with OSA, the heart rate varies in cycles: slowing with apnea onset, increasing slightly during apnea, and increasing more dramatically in the post-apneic period. These changes are illustrated below in a tracing from another patient. The numbers under the EKG tracing are the instantaneous heart rates. Although these cycles are referred to as bradytachycardia, the heart rate remains between 60 and 100 in many patients. In one series, 25% of OSA patients showed true bradycardia (60 bpm) and tachycardia (> 100 bpm). Bradyarrhythmias including heart block (2nd degree-mobitz types 1 and 2, and 3rd degree) occur in a minority of OSA patients (usually < 10%). Early studies attributed the slowing of heart rate during apnea to increased vagal tone and hypoxia. The slowing was diminished by atropine and supplemental oxygen. The increased vagal tone during apnea is the result of hypoxic stimulation of the carotid body. With resumption of respiration, inflation of the lungs decreases vagal tone, and the hypoxic influences on sympathetic tone are unmasked (tachycardia). More recent studies have not consistently found a reduction in heart rate in the last part of apnea. Instead, they have focused on tachycardia as the primary event at apnea termination, with a subsequent fall in heart rate as sympathetic activity diminishes after the initial burst. These studies suggest that the individual differences in the effect of apnea on heart rate may be secondary to differences in the response of the carotid body to hypoxia. The cycles of heart rate slowing may have little significance, except in cases of significant bradycardia or heart block. However, the periods of tachycardia and elevated blood pressure post-apnea increase myocardial oxygen demand at the same time that hypoxemia exists, predisposing to ischemia and possibly tachyarrhythmias. In normal individuals, sleep usually is a time of reduced tachyarrhythmias and ischemia. Patients with OSA may not enjoy the same protection. While PYCs are not uncommon in patients with OSA, they typically are unrelated to apnea or desaturation. In one series of 400 patients with OSA, PYCs were more frequent during sleep in only 14% of patients. In another study, a clear association between PVC frequency and the severity of nocturnal desaturation was found only when the SaOz was < 60%. Heart rate variability has recently been used as a tool to study the balance of parasympathetic and sympathetic tone in patients with OSA. During wakefulness, OSA patients show less heart rate variability than normal individuals. This is thought to be secondary to an increase in sympathetic tone that is still present during the day. After successful treatment with CPAP, the heart rate variability may increase, which suggests a drop in sympathetic activity. In the present patient, PYCs were noted during the middle of apnea rather than during periods of desaturation and arousal post-apnea. A 24-hour Holter monitor showed that the PVC rate was lower during the nocturnal hours. The patient underwent a nasal CPAP titration and was treated with 10 ern H,O pressure. He reported less daytime sleepiness. -A coronary bypass surgery has been scheduled. C 4 - A 1 O 2 - A 1 ROC-A 1 LOC - A 2 chin EMG EKG Airflow Chest Abdomen ~~~ ~/\..~\fl~Jl~ -...JwI'~ ~+W+AAUUH~~ 85 66 63 67 70 85 92 --------~---- ---------- 80% 195

Clinical Pearls I. pves in patients with OSA usually are unrelated to sleep apnea unless the SaO~ is less than 60%. - 2. The most common cardiac rhythm during sleep in patients with OSA is a periodic slowing and speeding of the sinus rate. Although the rhythm is sometimes called bradytachcardia, the heart rate often remains at 60-100 bpm in many patients. The heart rate slows at apnea onset and increases following apnea termination. 3. Untreated OSA patients have decreased heart rate variability that increases after effective treatment. REFERENCES I. Zwillich C, Devlin T, White D, et al: Bradycardia during sleep apnea. Characteristics and mechanisms. J Clin Invest 1982; 69: 1286-1292. 2. Guilleminault C, Connoly SJ, Winkle RA: Cardiac arrhythmia and conduction disturbances during sleep in 400 patients with sleep apnea syndrome. Am J Cardiol 1983; 52:490--494. 3. Shepard JW, Jr. Garrison MW, Grither DA, et al: Relationship of ventricular ectopy to oxyhemoglobin desaturation in patients with obstructive sleep apnea. Chest 1985; 88:335-340. 4. Becker H, Brandenburg U, Peter JH, et al: Reversal of sinus arrest and atrioventricular conduction block in sleep apnea during nasal continuous positive airway pressure. Am J Resp Crit Care Med 1995; 151:215-218. 5. Sato F, Nishimura M, Sinano H, et al: Heart rate during obstructive sleep apnea depends on individual hypoxic chemosensitivity of the carotid body. Circulation 1997;96:274-281. 6. Leung RST, Bradley TD: Sleep Apnea and Cardiovascular Disease. State of the Art. Am J Resp Crit Care Med 200 I; 164: 2147-2165. 7. Khoo MC, Belozeroff V, Berry RB, Sassoon CS: Cardiac autonomic control in obstructive sleep apnea: Effects of long-term CPAP therapy. Am J Respir Crit Care Med 2001; 164:807-812. 196

PATIENT 59 A 30-year-old pregnant woman with onset ofsnoring A 30-year-old woman in the third trimester of her first pregnancy was noted by her husband (a physician) to snore heavily during the night. This occurred although she spent nearly all of the night sleeping in the lateral decubitus position. The patient had gained about 25 pounds during the pregnancy. During regular visits with her obstetrician, all fetal monitoring indicated a healthy pregnancy. There was no history of snoring prior to the pregnancy. Because the patient had been complaining of fatigue and was taking frequent naps, her husband was concerned that she might have obstructive sleep apnea (GSA). He had not heard any pauses in breathing during sleep. The patient denied falling asleep while watching television or reading during the day. Physical Examination: General: healthy, gravid appearance. HEENT: moderately congested nasal mucosa; edematous palate and uvula; 15-inch neck circumference. Extremities: trace edema. Question: Should a sleep study be performed? 197

Answer: A sleep study is unnecessary for snoring associated with pregnancy. Discussion: Pregnancy is associated with a number of physiologic changes that affect respiration during wakefulness and sleep. In the first trimester, increased sleepiness and total sleep time and decreased stages 3, 4, and REM sleep are common. In the second trimester sleep normalizes. In the third trimester, sleep commonly is disturbed again, secondary to frequent urination, backache, fetal movement, leg cramps, and heartburn. The restless leg syndrome can appear or become worse during pregnancy. A high level of progesterone (a respiratory stimulant) in the third trimester is associated with a lowering of the arterial PCO,. Growing abdominal girth results in an upward displacement of the diaphragm. In addition, edema develops in the nasal passages and pharynx. These last two changes result in snoring in up to 30% of all pregnant women. Sleep Problems During Pregnancy Daytime sleepiness Insomnia Snoring Restless leg syndrome Narcolepsy Obstructive sleep apnea Possible worsening of pre-eclampsia by snoring/upper airway resistance syndrome Although snoring is common in pregnant women, overt OSA is uncommon. A few cases of severe OSA have been reported. However, this condition may be underdiagnosed. Some pregnant patients with OSA continued to have sleep apnea after delivery; thus, pregnancy probably worsened but did not cause sleep apnea in these patients. One recent study suggested that airflow limitation can worsen blood pressure during pregnancy in patients with preeclampsia in the absence of overt apnea. 198 Thus, indications for sleep monitoring when snoring is present might include witnessed apnea, severe hypertension, previous pregnancy with fetal growth retardation, and severe hypersomnia or insomnia. If sleep apnea is present, therapeutic options are somewhat limited. Severe cases probably should be treated with nasal CPAP. Close monitoring of both fetus and patient is essential. There is some evidence to suggest that severe OSA in the mother causes retardation of infant fetal growth, but this has not been determined conclusively. If narcolepsy is present during pregnancy, the recommended stimulant treatment is pemoline, which is a category B drug. Category B means no controlled human studies have been published but animal studies show no risk to fetus, or animal studies show an adverse effect on the fetus but wellcontrolled studies in pregnant women failed to demonstrate a risk to the fetus. Of note, pemoline has been associated with liver failure and is no longer considered a first-line medication in nonpregnant patients with narcolepsy. Most of the medications used for the restless leg syndrome are category C or D, and physicians generally recommend only conservative management, such as iron and folate supplements. Diphenhyramine and zolpidem, category B drugs, address insomnia, but little experience is available. Therefore, most would recommend using no medications if possible, especially in the first trimester. In the present patient, the absence of respiratory pauses and symptoms of daytime sleepiness suggested that simple snoring was present. Regular obstetric care showed no evidence of fetal compromise. Therefore, the patient and husband were reassured and informed that if snoring persisted after delivery or if symptoms of daytime sleepiness were noted, then a sleep study would be performed.

Clinical Pearls 1. Snoring during pregnancy (especially in the last trimester) is common. 2. Development of overt sleep apnea during pregnancy is not common. 3. Indications for a sleep study in a pregnant woman who snores include witnessed apnea, severe hypertension, and previous or current unexplained pregnancy with fetal growth retardation. 4. If patients with OSA become pregnant, potential harm to the developing fetus is possible, and a sleep evaluation is warranted. 5. Limited data suggests that nasal CPAP is the OSA treatment of choice during pregnancy. Close fetal monitoring is essential. 6. The restless legs syndrome may develop or worsen during pregnancy. REFERENCES 1. Charbonneau M, Falcone T, Cosio, MG, et al: Obstructive sleep apnea during pregnancy. Am Rev Respir Dis 1991; 144:461--463. 2. Feinsilver SH. Hertz G: Respiration during sleep in pregnancy. Clin Chest Med 1992; 13:637-644. 3. Loube DI, Poceta JS, Morales MC, et al: Self-reported snoring during pregnancy: Association with fetal outcome. Chest 1996; 109:885-889. 4. Edwards N, Blyton DM, Kirjavainen T, et al: Nasal continuous positive airway pressure reduces sleep induced blood pressure increments in preeclampsia. Am J Respir Crit Care Med 2000; 162:252-257. 5. Santiago JR, Nolledo MS, Kinzler W, Santiago TV: Sleep and sleep disorders in pregnancy. Ann Intern Med 2001;134:396--408 199

PATIENT 60 A 45-year-old man with snoring and hypercapnia A 45-year-old obese man was evaluated for severe, bilateral pedal edema of I-year duration. The patient had smoked one pack of cigarettes per day for 10 years, but he denied a history of cough or wheezing. There was no history of hypertension, chest pain, or myocardial infarction. The patient's wife reported that he snored heavily. However, the patient denied excessive daytime sleepiness. Physical Examination: Height 5 feet 9 inches, weight 275 pounds. Blood pressure 130/85 mmHg, pulse 80 and regular. Neck: short, 18-inch circumference. Chest: clear to auscultation. Cardiac: distant heart sounds, no murmurs. Extremities: 3+ pedal edema. Laboratory Findings: Spirometry: FEVI 3.0 L (77% of predicted), FVC 3.4 L (70% of predicted), FEV/FVC 0.88. Arterial blood gas (room air): pH 7.35, PC02 52 mmHg, P02 55 mmHg, HC03 33 mmol/L. Chest radiograph: borderline cardiomegaly. Sleep Study: AHI 66/hr. Minimum oxygen saturation 40%. Number of desaturations to < 85%: 300. Questions: What is the diagnosis? What treatment will reduce the level of hypercapnia? 200

Diagnosis: Obesity hypoventilation syndrome. Nasal CPAP or tracheostomy reduces hypercapnia. Discussion: The diagnosis of obesity hypoventilation syndrome (OHS) requires that the patient be obese and hypoventilate for reasons other than lung disease or neuromuscular weakness. Most patients with OHS have obstructive sleep apnea (OSA). An occasional patient experiences worsening of daytime hypoventilation during sleep without discrete apneas or hypopneas. Patients with OHS sometimes are called Pickwickian. This term is best avoided because some physicians use it to refer to all patients with OSA. The combination of obesity, snoring, and unexplained CO2 retention always suggests the possibility of OHS. The absence of a history of severe daytime sleepiness does not rule out this diagnosis. Many patients underestimate the severity of their daytime sleepiness. Note that only about 15% of patients with OSA have significant daytime CO2 retention, and these patients form two groups: those with OHS and those with overlap syndrome (OSA + chronic obstructive pulmonary disease). These groups tend to have especially severe nocturnal oxygen desaturation and evidence of cor pulmonale. The etiology of CO2 retention in OHS is multifactorial. Patients with OHS have reduced ventilatory responses to hypercapnia and hypoxia. In addition, they have a lower chest wall compliance (increased work of breathing) than patients with a similar amount of obesity without hypoventilation. After adequate therapy of the OSA with tracheostomy or nasal CPAP, the hypercapnic ventilatory response changes. There is a parallel shift of the ventilation versus PCOz curve to the left, reflecting a higher ventilation at a given PCOz (slope unchanged). This type of alteration in the hypercapnic ventilatory response is probably due to prevention of nocturnal worsening ofCO2 retention and the associated retention of HC03. Removal of the depressant effects of chronic hypoxia on ventilatory drive also may be a factor. In any case, daytime PCOz usually decreases after treatment. Adequate treatment of patients with OHS usually requires nasal CPAP or tracheostomy. In cases of severe daytime hypoxia, daytime oxygen and the addition of oxygen to nasal CPAP at night may be required until clinical improvement occurs. Although diuretic therapy often is prescribed, the cornerstone of treatment for cor pulmonale is relief of hypoxia (and the associated pulmonary arterial vasoconstriction). Medroxyprogesterone (Provera), a respiratory stimulant that takes several days to reach maximal effect, has been used to treat patients with OHS. Treatment usually improves the level of daytime CO, retention, nocturnal oxygenation, and signs of cor pulmonale. However, this agent does not reduce the AHI nor improve symptoms of daytime sleepiness. Side effects of medroxyprogesterone include decreased libido (decreased testosterone levels), alopecia, and hyperglycemia. For these reasons, it is no longer the treatment of choice in these patients. In the present patient, the spirometric results (mild restrictive pattern) made COz retention secondary to lung disease unlikely. The presence of obesity, severe sleep apnea, and high daytime PC02 is consistent with OHS. After several weeks of treatment with 14 em H20 of nasal CPAP, the daytime PC02 fell to 45 mmHg. Clinical Pearls I. Unexplained CO2 retention, obesity, and OSA suggest obesity hypoventilation syndrome (OHS). 2. Patients with OHS may present with signs of cor pulmonale rather than major complaints of excessive daytime sleepiness. 3. Adequate treatment of OSA frequently reduces the level of daytime CO2 retention and improves the nocturnal arterial oxygen saturation and cor pulmonale. 4. Medroxyprogesterone reduces the daytime PC02 level in patients with OHS, but does not improve the AHI. Positive airway pressure is the treatment of choice for patients with OHS. REFERENCES I. Sullivan CE, Berthon-Jones M, Issa FG: Remission of severe obesity-hypoventilation syndrome after short-term treatment during sleep with nasal continuous positive airway pressure. Am Rev Respir Dis 1983; 128:177-81. 2. Rajagopal KR, Abbrecht PH, Jabbari B: Effects of medroxyprogesterone acetate in obstructive sleep apnea. Chest 1986; 90:815-821. 3. Rapoport OM, Garay SM, Epstein H, et al: Hypercapnia in the obstructive sleep apnea syndrome. Chest 1986; 89:627-635. 4. Berthon-Jones M, Sullivan CE: Time course of change in ventilatory response to C02 with long-term CPAP therapy for obstructive sleep apnea. Am Rev Respir Dis 1987; 135:144-147. 201

PATIENT 61 A 55-year-old man with hypercapnic respiratory failure A 55-year-old man was admitted to the intensive care unit (ICU) with hypercapnic respiratory failure. The patient's wife reported that he snored heavily, had apneic episodes at night, and had been sleepy during the day for several years. Prior to admission, he had become increasingly somnolent and his ankles had swollen. There was no history of chest pain or fever. A previous pulmonary function test revealed only mild restrictive ventilatory dysfunction: FEV I 2.56 L (70% of predicted), FVC 3.33 L (72% of predicted), FEV/FVC 0.77. Physical Examination: Height 5 feet 9 inches, weight 250 pounds. HEENT: large tongue, dependent palate; IS-inch neck circumference. Chest: a few rales at the bases, no wheezes. Cardiac: distant heart sounds. Abdomen: very obese. Extremities: 3+ pedal edema. Neurologic: easily arousable, but very sleepy. Laboratory Findings: Chest radiograph: enlarged pulmonary arteries, no evidence of pulmonary edema. EKG: sinus rhythm, right axis deviation. Echocardiogram: normal left ventricular function, dilated right ventricle, increased estimated pulmonary arterial pressure (40 mmHg). Arterial blood gas (room air): pH 7.24, PC02 70 mmHg, P02 45 mmHg, HC03 30 mmol/L. Question: What is the cause of the patient's respiratory failure? 202

Diagnosis: Obesity hypoventilation syndrome with acute worsening of chronic respiratory failure. Discussion: Patients with the obesity hypoventilation syndrome (OHS) or the overlap syndrome (OSA + chronic obstructive pulmonary disease [COPD]) sometimes present with hypercapnic respiratory failure and hypoxemia. There usually is evidence of a chronic component and a history of slow deterioration with increasing somnolence and evidence of right heart failure. The diagnosis should be suspected in any obese, hypercapnic patient or in any hypercapnic patient with COPD who has an FEY I > I L. With COPD alone, hypercapnia is unusual until the FEY I falls below I L (40% of predicted). It is important to recognize the existence of OSA in patients with OHS or overlap syndrome because adequate treatment of the sleep apnea (usually tracheostomy or nasal CPAP) results in a reduction of daytime PC02 and an improvement in oxygenation and cor pulmonale. When sleep monitoring equipment is available in the ICU, the presence of sleep apnea can be precisely documented in stable patients. Sometimes simple observation confirms the diagnosis, but more often empiric treatment is begun, and a confirmatory sleep study is obtained after the patient's condition improves. There are several therapeutic approaches. If the patient is somnolent or the pH low on admission, nasal bilevel pressure with oxygen can be started immediately. If the patient is alert and the respiratory acidosis is well compensated, controlled oxygen therapy can be employed during the day and empiric treatment with nasal bilevel pressure or CPAP plus oxygen can be used during sleep. The level of positive pressure is titrated until obstructive apnea is prevented andlor the arterial blood gases are stabilized. One approach might be to start with a bilevel pressure of 8/4 and then titrate upward as needed. If patients have difficulty keeping their mouth shut, a full face mask can be used. Elevation of the head of the bed will minimize the amount of positive pressure required. The addition of heated humidity may help if patients have mouth leak. Of note, if significant leak is present, the high flow provided by the machine may reduce the effective FiO, supplied by the addition of oxygen. The Fi02 is adjusted to keep the Sa02 > 90-92%. It is not unusual to require 4-6 L per minute of oxygen or more to the positive pressure device tubing. When using bilevel pressure, remember that the level of end-expiratory positive airway pressure (EPAP) is titrated to prevent upper airway closure. When the patient is awake, a low level of EPAP usually suffices (3-5 em H,O). During sleep, the level can be increased until airway obstruction is prevented. The inspiratory positive airway pressure (IPAP) level is titrated above the EPAP level to prevent hypopnea and to assist ventilation. The IPAPEPAP difference is the level of pressure support. As in any chronically hypoventilating patient, the immediate goal of noninvasive ventilation is to stabilize the pH rather than normalize the PC02. Another alternative is to use positive-pressure volume-cycled ventilation (assist control mode) via nasal or full-face mask. The ventilator must be leak-tolerant; chin strips may be needed with a nasal mask to reduce mouth leak. Positive end-expiratory pressure is added to prevent upper airway closure. The Fi02 (inspired oxygen concentration) is increased to maintain adequate oxygenation. While noninvasive mask CPAP, bilevel, or volume ventilation will suffice in many patients, obtunded or rapidly deteriorating patients are best treated with endotracheal intubation and mechanical ventilation. Remember that there is considerable risk in treating patients with impaired conciousness with mask positive-pressure ventilation (aspiration, bloating). Clinical judgment is needed to determine if an acutely ill patient can be handled with mask ventilatory support. Close observation and monitoring is essential, which almost always requires an ICU setting. Acute respiratory acidosis even with a chronic component should not be treated on a general medical floor. Patients presenting with respiratory failure secondary to the OHS usually have signs of cor pulmonale (right heart failure). Treatment with diuretics may be employed; however, the main treatment of this problem is prevention of hypoxemia and the associated pulmonary arterial vasoconstriction. In the present patient, the restrictive pattern on spirometry was not severe enough to account for CO2 retention. The history was highly suggestive of OHS. The patient was alert enough to attempt nasal ventilation, and he was started on bilevel pressure via nasal mask at an IPAP/EPAP of 10/3 cm Hp, with the addition of oxygen titrated to keep the Sa02 above 90%. IPAP/EPAP was increased to ISIS ern H20 over the first hour. The PC02 stabilized at 65 mmHg with the Sa02 above 90%. Treatment included diuretics. Over the next 3 days, edema decreased, and the PC02 gradually improved to 50 mmHg; supplemental oxygen was no longer required during the day. A partial-night sleep study revealed an AHI of 80/hr and demonstrated that bile vel pressure of 17/12 was needed to maintain upper airway patency during sleep when the patient was supine. Supplemental oxygen at I Llmin was required to maintain an oxygen saturation> 90%. The patient eventually was discharged on this treatment. When seen in clinic 3 weeks later, the daytime PC02 had improved to 45 mmHg. 203

Clinical Pearls 1. Patients with OHS or the overlap syndrome may present with a mixture of acute and chronic ventilatory failure. 2. Conservative therapy with oxygen and nasal CPAP, bilevel pressure. or volumecycled positive pressure ventilation (by mask) may avoid the need for intubation. If mouth leaks are a problem. try a full-face mask. Heated humidity may also improve tolerance to treatment if mouth leaks are present. 3. During sleep, the level of nasal CPAP or bilevel pressure can be titrated to prevent upper airway closure and hypopnea and desaturation. A formal sleep study and pressure titration can be performed once the patient's condition stabilizes. REFERENCES I. Sullivan CEoBerthon-Jones M. Issa FG: Remission of severe obesity-hypoventilation syndrome after short-term treatment during sleep with nasal continuous positive airway pressure. Am Rev Respir Dis 1983; 128:177-181. 2. Shivaram U. Cash ME. Beal A: Nasal continuous positive airway pressure in decompensated hypercapnic respiratory failure as a complication of sleep apnea. Chest 1993; 104:770-774. 3. Piper AJ. Sullivan CE: Effects of short-term NIPPV in the treatment of patients with severe obstructive sleep apnea and hypercapnia. Chest 1994; 105:434-440. 204

PATIENT 62 A 57-year-old man with severe obstructive sleep apnea treated with oxygen An obese 57-year-old man was admitted to the intensive care unit twice in the same year for hypercapnic respiratory failure and congestive heart failure. Mechanical ventilation was required on one occasion, and on the other he was treated with bilevel pressure and oxygen by nasal mask. A sleep study performed after one admission showed an AHI of 80/hr and severe arterial oxygen desaturation. The patient responded to nasal CPAP, but refused treatment with this therapy. He was treated with nocturnal oxygen at 3 Umin. However, 3 months after his last admission, the patient was again admitted with hypercapnic respiratory failure and a weight gain of 30 pounds over 2 months. His wife had problems waking him up during the week prior to admission. Physical Examination: Blood pressure 160/90 mmHg, pulse 88. Height 5 feet 10 inches, weight 300 pounds. HEENT: massive tongue; 18-inch neck circumference. Chest: rales at bases. Cardiac: S4 gallop. Extremities: massive edema to upper thigh. Laboratory Findings: Arterial blood gas: pH 7.30, PC02 80 mmHg, P02 55 mmHg, HC03 38.4 mmollL on 4 Umin oxygen. Thyroid studies: normal. Spirometry: FEV I 2.5 L (66% predicted), FVC 3.0 L (62% predicted), FEV /FVC 0.83. Question: What long-term treatment do you recommend? 205

Answer: Tracheostomy and oxygen are recommended for this patient with obesity hypoventilation syndrome. Discussion: Oxygen is not the treatment of choice in patients with OSA, because while oxygen may decrease the severity of nocturnal desaturation, it has only a minor impact (slight decrease) on the frequency of apnea and therefore the severity of daytime sleepiness. In severe cases, nocturnal desaturation may persist despite oxygen administration. In addition, oxygen induces modest increases in apnea duration. Interestingly, oxygen tends to decrease central and mixed apneas and increase obstructive apneas. In patients with cor pulmonale secondary to OSA who refuse more effective therapy, oxygen sometimes can result in clinical improvement. In patients with the overlap syndrome (OSA + COPD), oxygen can considerably worsen nocturnal hypercapnia. However, even in these patients, oxygen treatment may be the only alternative if nasal CPAP or other effective treatment for OSA is refused. Tracheostomy, although rarely used today, still has a place in the treatment of patients with very severe OSA. When such patients refuse other treatment, tracheostomy is preferable to an early death from repeated bouts of severe cor pulmonale and respiratory failure. In some patients with OHS or the overlap syndrome, nocturnal desaturation persists during sleep even if upper airway obstruction is abolished. These patients have low baseline POz values in the supine position and may hypoventilate during sleep even with a patent upper airway. In such patients, treatment with both tracheostomy (or nasal CPAP) and oxygen may be necessary, at least initially. The level of hypercapnia and oxygenation may subsequently improve with adequate treatment of the cor pulmonale and OSA. Thus, not all patients require long-term oxygen therapy. After the third visit in the ICU in I year, the present patient was advised to have a tracheostomy or face possible early death. He underwent the procedure and was treated with nocturnal oxygen after a nocturnal oximetry study showed persistent desaturation post-tracheostomy. Over the subsequent month, the patient's daytime PCO z fell to 45 mmHg, and his edema improved tremendously. Two years have passed since his last bout of respiratory failure. Clinical Pearls 1. Oxygen treatment in patients with OSA may improve nocturnal desaturation, but does not significantly decrease the AHI or the severity of daytime sleepiness. Therefore, it should be used as a treatment of last resort. 2. Patients with severe obesity, OHS, or the overlap syndrome may still have nocturnal desaturation even if upper airway patency is restored. (Supplemental oxygen is added to the trachostomy or nasal positive airway pressure tubing). 3. Tracheostomy still is a valid treatment for patients with severe, life-threatening OSA who refuse other effective therapy. REFERENCES I. Smith PL. Haponik EF, Bleecker ER: The effects of oxygen in patients with sleep apnea. Am Rev Respir Dis 1984; 130:958-963. 2. Fletcher EC, Brown DL: Nocturnal oxyhemoglobin desaturation following tracheostomy for obstructive sleep apnea. Am J Med 1985; 79:35-42. 3. Gold AR, Schwartz AR, Bleecker ER, et al: The effect of chronic nocturnal oxygen administration upon sleep apnea. Am Rev Respir Dis 1986; 134:925-929. 4. Fletcher EC, Munafo DA: Role of nocturnal oxygen therapy in obstructive sleep apnea. Chest 1990; 98: 1497-1504. 206

PATIENT 63 A 5-year-old child with behavior problems. A 5-year-old boy had developed behavior problems over the last 6 months. He often refused to go the bed at night and would not obey his kindergarten teacher. He bullied his fellow students and stole their toys, In addition, he seemed hyperactive and was unable to pay attention during class activities.. The child's parents reported that he had restless sleep and sometimes snored. There was no history of daytime sleepiness. The child had been evaluated by his pediatrician, who felt that his tonsils were only mildly enlarged. The patient was diagnosed as having the attention deficit/hyperactivity disorder and was being treated with methylphenidate, This resulted in minimal improvement in behavior. The pediatrician ordered a sleep study, Physical Examination: Normal weight and development. HEENT: mild to moderate tonsillar enlargement. Resting awake Sa02 96%, end-tidal PC02 36 mmHg. Figure: The tracing below was noted on the sleep study. The end-tidal PC02 tracing is not a capnogram, but the average of recent peak end-tidal PC02 values (see Patient 64 for an example of a capnogram). Question: What is causing this child's behavior problems? C4-Al C3-A2 02-Al 01 -A2 """'--.;..........~/""v........ ROC-A1 """W-w"'~.J+V" ,,.,H'..,,.......,,.,.,........ft'-ovo-......,, ........"I'"'l'........VO"'-........ovJ,J~.N'I'''"''''-l-..". '"""'\..1'1""""0"\, LOC-A2 chin EMG ~~~~'":""":'~~-:-":""'"':'~~..,....--:-~-:-~--:--~~~~--r"~:--:-:- EKG airflow (therm) nasal pressure chest abdomen 5002 PetC02 (value) '10 '10 1. 'I' '10 'II '!Z '13 '13 13 '13 'It 'It'll '11 H 90 " '1 '1 '11 '11 '10 '10 'II '11 '10 '10 11 ;( :t X x ~~~.~u~~n~~~~~~~~~~3~~~~~~~U~ 207

Answer: Obstructive sleep apnea/hypopnea syndrome. Discussion: Obstructive sleep apnea in children usually does not present with complaints of daytime sleepiness. In fact, complaints are often of inattentive, hyperactive, and aggressive behavior. Some children who have received a diagnosis of attention-deficit/hyperactivity disorder (ADHD) are later found to have OSA. In fact, some patients evaluated in sleep centers have already been placed on methylphenidate for ADHD! While tonsillar enlargement is the etiology in most cases of pediatric OSA, the severity ofthe disease does not correlate with tonsillar size. The relative size and and structure of the other components of the upper airway as well as upper airway muscle activity are important for determing the effect of a given degree of tonsillar hypertrophy. Many children with large tonsils have no problems, while others with only moderate tonsillar enlargement have severe disease. Many children with sleep-disordered breathing have discrete apneas or hypopneas only during REM sleep. During NREM, there are often long periods of obstructive hypoventilation secondary to high upper airway resistance. When airflow is monitored with thermistors, obstructive hypoventilation may only be detected by a low oxygen saturation or a high end-tidal PCO). However, nasal pressure monitoring often clearly-shows a severe pattern of airflow limitation (see figure). The present patient exhibited symptoms of hyperactivity rather than daytime sleepiness. Although snoring was not prominent, long episodes of obstructive hypoventilation occurred. Note that the thermistor really does not look abnormal. However, the nasal pressure signal clearly shows a pattern of severe airflow limitation. If nasal pressure was not used, the only clues for hypoventilation would be the low Sa02, paradoxical motion of the chest and abdomen, and high end-tidal PC02. The end-tidal CO2 signal is not the usual capnograph, but is the last measured end-tidal PC02 value. The child was referred for tonsillectomy and adenoidectomy. His sleep and behaviors were noticeably improved only a few weeks after surgery, and he was weaned off methylphenidate. Important Differences Between Adult and Pediatric GSA CHILDREN ADULTS Clinical Findings Peak age Sex ratio Etiology Excessive daytime sleepiness Neurobehavioral Sleep Study Findings Sleep architecture Obstruction Cortical arousal Usual Treatment Preschoolers (4-6) Equal Adenotonsillar hypertrophy Uncommon Hyperactivity, developmental delay Normal Obstructive hypoventilation or cyclic obstruction < 50% of apneas Tonsillectomy and adenoidectomy Over 45 Men> women Obesity, upper airway size Very common Impaired vigilance, cognitive impairment Decreased SWS and REM Cyclic obstruction At termination of most apneas CPAP SWS = slow wave sleep (stages 3 and 4) 208

Clinical Pearls I. Children with GSA often present with symptoms suggestive of ADHD rather than excessive daytime sleepiness. 2. The severity of sleep-disordered breathing does not correlate with the degree of tonsillar hypertrophy. 3. The peak incidence of pediatric GSA is in preschoolers (ages 4-6). 4. The predominant pattern of breathing abnormality is often obstructive hypoventilation without frequent arousals. REFERENCES I. Marcus CL: Sleep-disordered breathing in children. Am 1 Resp Crit Care Med 200 I; 164:16-30. 2. American Academy of Pediatrics: Clinical Practice Guideline: Diagnosis and Managment of Childhood Obstructive Sleep Apnea Syndrome. Pediatrics 2002; 109:704-712. 3. Chervin RD. Archbold KH. Dillon lE. et al: Inattention. hyperactivity. and symptoms of sleep-disordered breathing. Pediatrics 2002; 109:449-456. 209

PATIENT 64 A 6-year-old boy with large tonsils A 6-year-old boy was referred for evaluation of heavy snoring of 2-year duration. His parents had not noted apnea, but were concerned that he seemed to be "working hard to breathe during sleep" and was often sweaty during the night. While he was not sleepy during the day, he had trouble concentrating and was doing poorly in school. In the past the patient had been well-behaved, but he had become irritable and emotionally labile. Physical Examination: HEENT: bilateral, large tonsils (almost "kissing") with obstructed pharyngeal airway; boggy mucosa in nose. Otherwise unremarkable. Sleep Study: AHI 5/hr, long periods of increased end-tidal peoz to 55 mmHg, SaOz 92-93%. Figure: A sample tracing from a period of heavy snoring is shown below. Question: Should this patient have a tonsillectomy? PetC02 Chest Abdomen 1 sec 210

Answer: A tonsillectomy plus adenoidectomy (T&A) should be performed. Discussion: While excessive daytime sleepiness was the main presenting complaint in several initial studies of children with OSA, more recent studies have found this complaint in only a minority ofpatients. Snoring is common in children with OSA, although it is not universally present. As in adults, simple snoring is much more common than OSA. Parents may notice restless sleep with increased inspiratory effort and diaphoresis in a child with OSA. Daytime symptoms of mouth breathing, behavioral problems, or poor progress in school may be noted. Rarely, developmental delay occurs. With the exception of children with craniofacial abnormalities, most OSA in children is secondary to obstruction from adenotonsillar hypertrophy. Interestingly, the severity of the disorder does not correlate with tonsil size. Some children have large tonsils and snoring, yet little impairment in breathing. Polysomnographic findings in children with sleep apnea also can differ from those in adults. In adults, apnea in defined as a cessation of airflow for 10 seconds or more, and the normal AHI is considered to be < 5/hr. In children, any cessation in airflow greater than two normal respiratory cycles is considered an apnea, and an AHI l/hr is considered to be abnormal. Many children with OSA exhibit relatively few discrete apneas or hypopneas; instead they show long periods of hypoventilation and desaturation. For this reason, pediatric sleep laboratories often use end-tidal PC02 monitoring to assess the periods of hypoventilation. An increase in end-tidal PC02 to 55 mmHg of any duration or > 50 mmHg for longer than 10% of the total sleep time (or> 45 mmHg for longer than 60%) is considered abnormal. Today, most large sleep centers have special monitoring rooms for pediatric patients. These rooms feature children's decor and an extra bed in the room for a parent to stay. In most children with OSA, the treatment of choice is removal of enlarged adenoids and tonsils. Tonsillectomy generally is considered a routine surgery in children and is often performed as an outpatient procedure. However, in children with OSA, postoperative complications of T&A can occur in 16-27% of patients. Complications include upper airway obstruction, pulmonary edema, and respiratory depression from sedati ves and narcotics. One study suggests that the presence of any of the following is an indication for postoperative monitoring overnight in the hospital with oximetry: age < 2 years, craniofacial abnormalities, failure to thrive, morbid obesity, cor pulmonale, AHI > lO/hr, nadir Sa02 > 70%, and daytime hypoventilation. Overnight monitoring in a pediatric ICU should be considered for patients with elevated daytime PC02. Nasal CPAP has been used in the postoperative setting to help maintain upper airway patency. Postoperative polysomnograms 6-8 weeks after T&A are recommended for patients with additional risk factors (obesity, craniofacial abnormalities, cor pulmonale) or very high AHI values pre-op to ensure an adequate response. Nasal CPAP is problematic in young children, but has been used sucessfully in patients not responding to adenotonsillectomy. In severe cases of childhood OSA with hypoventilation and craniofacial abnormalities, tracheostomy may be indicated. In the present patient, despite an AHI of only 5/hr, there were long periods of snoring, shallow breathing, and increased end-tidal PC02, as well as desaturation. The patient was referred to an ENT surgeon, and tonsillectomy plus adenoidectomy was performed. After surgery, the patient was monitored in the hospital overnight with oximetry, but there were no signs of desaturation. Within days his disposition, behavior, and school work improved, and he slept better. 211

Clinical Pearls 1. In children, the most common cause of OSA is hypertrophy of the adenoids and tonsils. Tonsillectomy plus adenoidectomy (T&A) is the treatment of choice in most cases. 2. After T&A, high-risk children with OSA should be monitored in the hospital overnight with oximetry. 3. In children, an AHI 2:: l/hr is considered abnormal. 4. In a pediatric study, the sleep environment should be suitable for children. Arrange for a parent to sleep in the same room with younger children. 5. Monitoring end-tidal PCO? may be useful in detecting periods of obstructive hypoventilation in pediatric patients with sleep-disordered breathing. REFERENCES I. Rosen GM, Muckle RP, Mahowald MW, et al: Postoperative respiratory compromise in children with obstructive sleep apnea syndrome: Can it be anticipated') Pediatrics 1994; 93:784--788. 2. Carroll JL, Loughlin GM: Obstructive sleep apnea in infants and children: Diagnosis and management. In Ferber R, Kryger M (eds): Principles and Practice of Sleep Medicine in the Child. Philadelphia, WB Saunders, 1995; pp 193-230. 3. Marcus CLK, Ward SL, Mallory GB, et al: Use of nasal continuous positive airway pressure as treatment of childhood obstructive sleep apnea. J Pediatr 1995; 127:88-94. 4. American Thoracic Society: Standards and indications for cardiopulmonary sleep studies in children. Am J Respir Crit Care Med 1996; 153:866-878. 5. Schechter MS: Technical report: Diagnosis and management of childhood obstructive sleep apnea syndrome. Pediatrics 2002; 109:E69. 212

PATIENT 65 A 20-year-old woman with daytime sleepiness since childhood A 20-year-old woman with moderate mental retardation was evaluated for persistent heavy snoring and leg edema. The patient had a long history of obstructive sleep apnea. Her condition deteriorated after she visited with her parents for 2 months and gained 30 pounds. Before this vacation, she was cared for at an institution for mentally retarded children and was doing well on nasal CPAP of 12 em H20. Physical Examination: Height 60 inches, weight 210 pounds. HEENT: upsloping palpebral fissures, crowded oropharynx/dependent palate; 17-inch neck circumference. Extremities: 2 + pitting edema, short calves. Neurological: oriented to person and place, but not date; evidence of mild retardation, but able to answer questions appropriately. Figure: This tracing occurred during stage 2 sleep in the initial, diagnostic portion of the study, off positive pressure. TcPC02 = transcutaneous PC02. Later in the study, frank obstructive apneas were noted. The awake TcPC02 was 50 mmHg. Question: Why is the Sa02 so low (74-79%) and the transcutaneous PC02 so high (70 mmHg) when no discrete breathing events are noted? nasal pressure C4-A2 02-Al ROC-A1 LOC-A2 chinl-chin3 -'*-...-........----"""""! EKG airflow chest abdomen Sa02 ------------~--~~-------~~ TcPC02 213

Answer: Hypoventilation associated with the Prader-Willi syndrome. Discussion: The Prader-Willi syndrome (PWS) is characterized by hypothalamic obesity, hyperphagia, hypogonadism, mental retardation, hypotonia (muscle weakness), and behavioral and sleep disorders. The prevalence of this disorder is 1 per 10,000 to 25,000 live births. The syndrome is associated with failure of expression of genes on the long arm of the paternally derived chromosome 15. There is a deletion of the long arm of this chromosome in approximately 50-70% of patients. PWS patients have a characteristic habitus including short stature, up-sloping palpebral fissures, and short arms and legs. Growth hormone secretion is abnormal. Some have compulsive behaviors. Daytime sleepiness may be an intrinsic manifestation of the disease, and in some patients it is worsened by sleep-disordered breathing. Some patients have sleep-onset REM. One important characteristic of patients with this syndrome is an insatiable appetite leading to massive obesity. In evaluating alterations in ventilatory control it has been difficult to separate intrinsic problems from those associated with severe obesity. However, the hypoxic ventilatory response was found to be absent or reduced in PWS patients with or without obesity. This is believed to be secondary to peripheral chemoreceptor dysfunction, although central processing of chemoreceptor information may be involved as well. The hypercapnic ventilatory responses appear to be decreased mainly in obese patients. The arousal response to hypoxia during sleep is virtually absent, and the arousal response to hypercapnia is impaired. Obese patients with Prader-Willi may have severe forms of OSA, with severe hypoxia and hypercapnia during sleep. They may also have daytime hypercapnia. Patients demon214 strating hypoventilation may require noninvasive venti lation. Treatment of OSA in these patients usually consists of weight loss and some form of positivepressure therapy or upper airway surgery. Weight reduction often results in considerable improvement. As PWS patients have mental retardation, weight loss is usually only possible in a very structured environment. Special group homes experienced in caring for persons with PWS have had success with weight loss. Positive-pressure treatment usually consists of nasal CPAP or bilevel pressure. Very severe cases may require nocturnal ventilatory support via tracheostomy - at least until weight loss has been effected. Recently, growth hormone replacement has been shown to improve muscle strength, decrease body fat, and improve ventilatory control in these patients. In the present case, the tracing shows what appears to be relatively normal airflow. However, the arterial oxygen saturation is very low and the transcutaneous PC02 high. In fact, the awake PC02 was also high. This patient has daytime hypoventilation that worsened during sleep. Later during the short diagnostic portion of the sleep study, traditional obstructive hypopneas and apneas were noted, with even more severe arterial oxygen desaturation. The patient required a combination of bilevel pressure 16/10 ern Hp and supplemental oxygen at 21pm to prevent nocturnal desaturation. After 2 months of treatment and an intensive weight-loss program (30- pound weight loss), the daytime transcutaneous PC02 returned to normal, and nocturnal oxygen was no longer needed (demonstrated by nocturnal oximetry). The patient continued treatment on bilevel pressure of 16/10 em H20, although lower levels might have sufficed.

Clinical Pearls 1. Prader-Willi Sydrome (PWS) is associated with severe obesity, obstructive sleep apnea, and abnormal ventilatory control. 2. The ventilatory response to hypercapnia and daytime CO2 retention (if present) may normalize with weight loss and treatment of OSA. 3. Most patients have an abnormality in chromosome 15. 4. Food restriction (weight control) is an essential component of treatment. 5. Obese PWS patients with OSA often have very severe desaturations and hypercapnia during sleep. REFERENCES I. Arens R, Gozal D. amlin KJ, et al: Hypoxic and hypercapnic ventilatory responses in Prader-Willi syndrome. J Appl Physiol 1994;77:231-2236. 2. Lindgren AC, Hellstron LG. Ritzen EM, Milerad J: Growth hormone treatment increased CO, response, ventilation. and central inspiratory drive in children with Prader-Willi syndrome. Eur J Pediatr 1999; 158:936-940. - 3. Manni R, Politini L, Nobili L, et al: Hypersomnia in the Prader-Willi syndrome: Clinic-electrophysiological features and underling factors. Clin NeurophysioI2001; 112:800-805. 215

PATIENT 66 A 55-year-old man with chronic obstructive pulmonary disease and nocturnal desaturation A 55-year-old man with severe chronic obstructive pulmonary disease (CapO) underwent nocturnal oximetry monitoring to determine if nocturnal desaturation might explain the presence of cor pulmonale. He admitted to snoring, but denied daytime sleepiness. Physical Examination: Height 5 feet 10 inches, weight 180 pounds. Blood pressure 130/90 mmHg, pulse 85, temperature 37°C, respirations 20/min. HEENT: edentulous, otherwise normal; 15Yz-inch neck circumference. Chest: decreased breath sounds. Cardiac: no murmurs or gallops. Extremities: I+ pedal edema. Laboratory Findings: Spirometry: FEV I 1.1 L (29% of predicted), FVC 2.5 L (52% of predicted), FEV /FVC 0.44. Arterial blood gas (room air): pH 7.43, PC02 38 mmHg, P02 65 mmHg, HC03 25 mmol/L. Figure: A tracing from the nocturnal oximetry monitoring is shown below. Question: What is causing the nocturnal arterial oxygen desaturation? - 100 0 ....- C 90 • - - 0 80 :.;:::; COL.. 70 :::J .- m 60 (J) C 50 Q) 0> >0- 40 X 0 30 .~ m 20 Q) t m 10 11 12 1 2 3 4 5 6 7 PM AM time FIGURE 1 216

Answer: Chronic obstructive pulmonary disease. Discussion: Patients with capo can have arterial oxygen desaturation during sleep for several reasons. First, the normal sleep-associated 8- to 10- mmHg fall in P02 has much greater significance if the baseline presleep PO? value is on the steep part of the oxyhemoglobin saturation curve (P02 50-60 mmHg). At this point, a normal fall in P02 during NREM sleep results in significant arterial oxygen desaturation. Second, periods of sleep apnea of varying significance may occur. (See Patient 70 for a discussion of the overlap syndrome [CaPO + OSA]). Third, nonapneic arterial oxygen desaturation can be abrupt and severe during REM sleep in patients with capo. The REM-associated desaturations are believed to be secondary to hypo ventilation during periods of hypopnea, as well as to an increase in ventilationperfusion (V/Q) mismatch. Often, the hypopneic periods are not well-defined and consist of small, variable tidal volumes over periods of time as long as several minutes. During REM sleep, the diaphragm is the only active muscle of inspiration (REM-associated skeletal muscle hypotonia). In patients with capo, diaphragmatic function often is compromised secondary to hyperinflation. In addition, neural drive to the diaphragm may fall during bursts of eye movements in REM sleep, producing hypopnea (or central apnea). An increase in VIQmismatch during these episodes also may contribute to hypoxemia. The increase in V/Q mismatch is believed to be secondary to a decrease in functional residual capacity during these REMrelated hypopneic episodes. Patients with capo may have REM-associated desaturations without having sleep apnea even if their daytime P02 is 60 mmHg. As expected, these REM-associated desaturations occur every 90-120 minutes during the night. The most severe and longest periods of desaturation typically occur in the early morning hours, when REM periods are longer and the REM density (number of eye movements per minute) is greater. In contrast, the pattern of arterial oxygen desaturation on an all-night plot in patients with OSA shows a saw-tooth pattern consistent with repetitive, discrete episodes of desaturation. Several studies have determined equations to predict the severity of nocturnal desaturation in capo patients based on awake measurements. In general, patients with lower Sa02 and higher PC02 are more likely to have significant nocturnal desaturation. However, there is considerable individual variation. In the present patient, the nocturnal oximetry showed a fall in Sa02 to below 85% (Fig. I), with further episodes of steep desaturation (black bars) probably associated with REM sleep. A complete sleep study was performed because of the history of snoring (despite the fact that the oximetry was not suggestive of sleep apnea). The dramatic falls in Sa02 were associated with hypopneic breathing during REM sleep. The baseline Sa02 again fell to less than 85% for the majority of NREM sleep. The AHI was only 10/hr. A tracing from NREM sleep (Fig. 2) shows low Sa02 despite regular airflow. In contrast, a tracing from REM sleep shows irregular, small tidal volumes; reduced chest wall motion (chest wall muscle hypotonia); and a lower Sa02• The patient was treated with nocturnal oxygen, and improvement in pedal edema was noted. (See Patient 68 for a discussion of the treatment of nocturnal oxygen desaturation associated with capo.) 217

LOC - A 2 NREM REM ~~ ~~~~ ~~Vi~ ~~'1;J~ chin EMG ur,," I'~ ••, H'. 10 I .1.. Airflow Chest Abdomen 85% 75% FIGURE 2 Clinical Pearls II The usual pattern of nocturnal desaturation due to COPD is a fall in baseline 5a02 during NREM sleep, with more dramatic falls occurring during episodes of REM sleep. 2. A saw-tooth pattern in the 5a02 tracing suggests that substantial sleep apnea is present. 3. REM-associated desaturation in patients with COPD usually is secondary to periods of hypoventilation (hypopnea) rather than apnea. REFERENCES I. Fletcher EC, Gray BA, Levin DC: Nonapneic mechanisms of arterial oxygen desaturation during rapid-eye-movement sleep. J Appl Physiol 1983; 54:632-639. 2. Hudgel DW, Martin RJ, Capehart M. et al: Contribution of hypoventilation to sleep oxygen desaturation in chronic obstructive pulmonary disease. J Appl Physiol 1983; 55:669-677. 3. Douglas NJ. F1enley DC: Breathing during sleep in patients with obstructive lung disease. Am Rev Respir Dis 1990; 141: 1055-1069. 218

PATIENT 67 A 60-year-old man with leg swelling A 60-year-old man with a long history of heavy smoking and chronic obstructive pulmonary disease (COPD) was referred for evaluation. He had failed to qualify for home oxygen therapy on a recent examination. The patient had bouts of mild pedal edema during courses of steroid therapy for exacerbations of COPD. These bouts responded to diuretics. There was no history of snoring, and the patient denied daytime sleepiness. He used inhaled bronchodilators only on an as-needed basis. Physical Examination: Height 5 feet 8 inches, weight 160 pounds. Blood pressure 140/89 mmHg, pulse 78. HEENT: edentulous; 15 YJ-inch neck circumference. Chest: bilateral wheezing. Cardiac: distant sounds. Extremities: I+ pedal edema. Laboratory Findings: ABO (room air): pH 7.42, PCOz 38 mmHg, POz 62 mmHg, HC03 23 mmol/L. Spirometry: FEV I 1.6 L (46% of predicted), FVC 3.2 L (73% of predicted), FEV/FVC 0.50. Chest x-ray: possible mild enlargement of the pulmonary arteries. Question: Should this patient have polysomnography or nocturnal oximetry? 219

Answer: Nocturnal oximetry suffices in this case. Discussion: Some degree of nocturnal arterial oxygen desaturation is common in patients with moderate-to-severe capo. One study found that daytime measurements of lung function could predict nocturnal desaturation. Although there was considerable variability, the lower the SaOz and the higher the PCOz, the more likely nocturnal desaturation was to occur. The study also found that the survival of patients with greater-than-predicted nocturnal desaturation was no worse than that of patients with less nocturnal desaturation. The authors concluded that sleep studies were not useful in patients with capo unless sleep apnea was suspected. Another study of capo patients with daytime paz > 60 mmHg found that 27% showed some desaturation during sleep, although most of the desaturations were during REM sleep and often brief. Desaturation could not be predicted on the basis of daytime studies. In a subsequent investigation, modest improvements in daytime pulmonary artery pressures were documented in a group of patients with daytime paz > 60 mmHg and nocturnal oxygen desaturation who received oxygen treatment. However, it is not clear that these improvements are clinically significant. To date, there appears to be no clear benefit to diagnosing or treating isolated periods of REM-associated nocturnal desaturation. With the current state of knowledge, the clinician must individualize the decision for sleep studies and nocturnal oxygen treatment in patients with capo not qualifying for 24-hour oxygen treatment on the basis of daytime paz. A sleep study is indicated if obstructive sleep apnea (or another cause of excessive daytime sleepiness) issuspected. Patients with significant unexplained cor pulmonale might benefit from some type of sleep study. The usual criteria for continuous oxygen therapy are a daytime paz :S; 55 mmHg or 55-59 mmHg plus evidence of end-organ damage (cor pulmonale). Patients who qualify for oxygen on the basis of these criteria do not need a sleep study unless sleep apnea issuspected or they have failed to respond to oxygen treatment. For patients with a daytime paz 2: 60 mmHg and evidence of significant cor pulmonale, a sleep study could provide documentation of significant nocturnal desaturation. However, there are no clear criteria for what constitutes significant nocturnal desaturation, and the optimal type of sleep study is not known. It probably is reasonable to use nocturnal oxygen to treat capo patients with a daytime paz > 60 mm Hg who have significant desaturation « 85%) in both NREM and REM sleep-especially if cor pulmonale is present. However, no clear benefit of doing so has been documented. The type of sleep study used to evaluate for nocturnal desaturation often depends on local resources. Oximetry alone may suffice in many cases. However, if the tracings suggest the presence of sleep apnea, a full sleep study (and nasal CPAP titration) may be required. The clinician should remember that oxygen is not the only treatment for nocturnal desaturation. In the original Nocturnal Oxygen Treatment Trial, 21% of the patients screened no longer met criteria when they were placed on intensive bronchodilator therapy. Some patients with minimal acute improvement in the FEY I and FYC have steady improvement in oxygenation when treated with smoking cessation and bronchodilator therapy. In the present patient, sleep apnea was not suspected, and nocturnal oximetry revealed only brief periods of mild desaturation (to 85%), probably associated with stage REM sleep. The baseline SaOz remained above 92% for most of the night. The patient was treated with more aggressive bronchodilator therapy, including the long-acting inhaled beta agonist salmetrol. Clinical Pearls 1. The clinical suspicion of sleep apnea is the main indication for a sleep study in patients with capo. 2. In patients not qualifying for continuous oxygen treatment, sleep monitoring may be indicated if significant, unexplained cor pulmonale is present. 3. The criteria for what constitutes significant nocturnal desaturation and the benefits of treating such desaturation (in the absence of daytime hypoxemia) remain to be demonstrated. 220

REFERENCES 1. Nocturnal oxygen Therapy Trial Group: Continuous or nocturnal oxygen therapy in hypoxemic chronic obstructive lung disease. Ann Intern Med 1980; 93:391-398. 2. Cannaughton JJ, Catterall JR, Elton RA: Do sleep studies contribute to the management of patients with severe chronic obstructive pulmonary disease? Am Rev Respir Dis 1988; 138:341-344. 3. Fletcher EC, Luckett RA, Goodnight-White S, et al: A double-blind trial of nocturnal supplemental oxygen for sleep desaturation in patients with chronic obstructive pulmonary disease and a daytime P02 above 60 mmHg. Am Rev Respir Dis 1992; 145:1070-1076. 221

PATIENT 68 A 55-year-old man with chronic obstructive pulmonary disease and severe pedal edema A 55-year-old man with severe capo was referred for evaluation of cor pulmonale. He did not qualify for home oxygen on a recent examination. The patient denied daytime sleepiness, but complained of frequent awakenings. His wife noted that he frequently snored, but she had never observed episodes of apnea. Physical Examination: Pulse 90, blood pressure 130/90 mmHg. Height 5 feet 8 inches, weight 170 pounds. Chest: bilateral wheezing. Cardiac: regular rate and rhythm. Extremities: 3+ pedal edema. Laboratory Findings: Spirometry: FEV I 1.0 L (27% of predicted), FVC 3.0 L (64% of predicted), FEV ,/FVC 0.33 (80-120% of predicted). Chest radiograph: hyperinflation, large pulmonary arteries. Arterial blood gas (room air): pH 7.42, PC02 45 mmHg, P02 62 mmHg, HCO) 25 mmol/L. Sleep Study Time in bed 445 min (378-468) Sleep Stages %SPT Total sleep time 350 min (340-439) Stage Wake 18 (2-7) Sleep period time (SPT) 425 min (361-453) Sleep efficiency (%) 79 min (88-96) Stage I 13 (4-12) Sleep latency 10 min (1-22 min) Stage 2 49 (51-72) REM latency 2.5 min (65-104) Stages 3 and 4 5 (0-13) AHI 4/hr«5) Stage REM 15(17-25) Mean Sa02 NREM 84% REM 70% ( ) = normal values for age Question: What treatment do you recommend? 222

Answer: Optimize treatment of COPD and initiate low-flow nocturnal oxygen. Discussion: Low-flow oxygen by nasal cannula can prevent the typical, nonapneic arterial oxygen desaturation manifested by patients with COPD, without substantially increasing the nocturnal PC02• However, oxygen is expensive and therefore should be prescribed only when it is likely to be worth the cost. The benefits of chronic 24-hour oxygen therapy in patients with COPD have been well documented by the Nocturnal Oxygen Treatment Trial (NOTT) and other studies of patients meeting the standard criteria of a daytime P02 :5 55 mmHg breathing room air. The value of 55 mmHg was chosen because below this point pulmonary arterial pressure starts to increase significantly secondary to hypoxic vasoconstriction. In the NOTT study, patients also received oxygen if the P02 was 55-59 mmHg and evidence of end-organ damage was present (edema, hematocrit> 55%, or P pulmonale on EKG). Today most physicians would consider evidence of significant cor pulmonale or neurologic dysfunction an indication for oxygen treatment in this group with borderline oxygenation. In the group of patients meeting criteria for 24- hour oxygen, sleep studies are not indicated unless sleep apnea is suspected. However, the clinician often is faced with the difficult question of what to do about patients not meeting any of the above criteria (daytime P02 2:: 60 mmHg). Some patients may benefit from nocturnal oxygen therapy if significant and prolonged nocturnal desaturation is present. However, one study comparing patients with mild daytime hypoxemia could find no difference with respect to mortality or development of pulmonary hypertension between groups with and without nocturnal desaturation during short-term follow-up up to 6 years. Indeed, criteria for what constitutes significant nocturnal desaturation have not been standardized (see Patient 67). While low-flow supplemental oxygen induces little increase in CO2 in most stable COPD patients during sleep, this is not the case if significant sleep apnea also is present (the overlap syndrome). When such patients are treated with oxygen, varying amounts of desaturation persist, the apneas tend to lengthen, and the nocturnal PCO? may increase significantly. Patients with the overlap syndrome also may complain of a morning headache after oxygen is initiated. Oxygen alone is not the optimal treatment for this group of patients. (See Patient 70 for a detailed discussion of the overlap syndrome.) In the present patient, a sleep study was ordered because of the history of snoring (to rule out OSA). The study revealed a low AHI and relatively few discrete desaturations (changes in Sa02 2:: 4%). However, the baseline Sa02 during NREM sleep was below 85% and even lower in REM sleep (64 min). An echocardiogram was consistent with cor pulmonale (right ventricle dilation, normal left ventricle function). Therefore, it was believed that the patient would benefit from nocturnal oxygen therapy. A separate oximetry study documented that oxygen at a flow rate of 2 Llmin maintained a saturation above 90% for all but a few brief desaturations (probably in REM sleep). The patient was begun on nocturnal oxygen therapy and his pedal edema improved. He also reported improved sleep quality Clinical Pearls 1. The criteria for treatment of COPD with a daytime PO? 2:: 60 mmHg and nocturnal desaturation are not well defined. Patients with significant desaturation in both NREM and REM sleep may benefit from oxygen treatment, especially if evidence of significant cor pulmonale is present. 2. Low-flow oxygen treatment for COPD-associated nocturnal desaturation usually does not cause significant increases of nocturnal CO2 unless significant sleep apnea is present. REFERENCES I. Nocturnal Oxygen Therapy Trial Group: Continuous or nocturnal oxygen therapy in hypoxemic chronic obstructive lung disease. Ann Intern Med 1980; 93:391-398. 2. Goldstein RS, Ramcharan V. Bowes G. et al: Effect of supplemental nocturnal oxygen on gas exchange in patients with severe obstructive lung disease. N Engl J Med 1984; 310:425-429. 3. Douglas NJ. Flenley DC: Breathing during sleep in patients with obstructive lung disease. Am Rev Respir Dis 1990; 141:1055-1069. 4. Chaouat A. Weitzenblum E. Kessler R. et al: Outcome of COPD patients with mild daytime hypoxemia with or without sleeprelated oxygen desaturation. Eur Resp J 2001; 17:848-855. 223

PATIENT 69 A 60-year-old "pink puffer" with insomnia A 60-year-old man with severe chronic obstructive pulmonary disease (COPO) was referred for complaints of poor sleep. He had difficulty falling asleep and then awakened several times during the night. Sometimes the awakenings were associated with dyspnea. There was no history of snoring or daytime sleepiness. The patient's medications included theophylline 300 mg bid and albuterol by metered-dose inhaler. Physical Examination: Height 5 feet 10 inches, weight 150 pounds. Blood pressure 150/80 mmHg, pulse 95. General: thin, nervous; no acute distress. HEENT: unremarkable; IS-inch neck circumference. Chest: hyperresonant to percussion, diminished breath sounds. Cardiac: distant heart sounds. Extremities: no edema. Laboratory Findings: Spirometry: FEV I 1.5 L (40% of predicted), FVC 2.8 L (59% of predicted), FEV/FVC 0.54, OLCO 15 ml/min/mmHg (44% of predicted). Chest radiograph: hyperinflation. ABG (room air): pH 7.43, PC02 38 mmHg, P02 65 mmHg. Theophylline level: 11.5 mg/ml. Question: What treatment do you recommend? 224

Answer: Try a long-acting inhaled bronchodilator. Discussion: The sleep of patients with COPO is poor, with low sleep efficiencies and, often, reduced amounts of slow wave and REM sleep. Patients may complain of frequent awakenings. Many different approaches have been tried to improve sleep in these patients, but all of the approaches share one limitation: no one actually knows what is waking patients up. For example, isolated hypoxemia is a poor arousal stimulus. One study suggested that supplemental oxygen improves sleep, while another did not find an improvement. Cough or wheezing also could awaken patients. However, cough usually does not occur during sleep. Nocturnal dyspnea is another possible cause of disturbed sleep. Patients with COPO have an exaggeration of the normal diurnal variation in lung function, with FEV I worsening around 6 AM. Many bronchodilators taken at bedtime may have worn off by the time they are most needed. Sustained-action theophylline might have an advantage; however, this drug's stimulant properties could disturb sleep. Studies have suggested that the advantages may balance the side effects in some patients. Moreover, the stimulatory side effects vary in severity among individuals. Salmeterol, a long-acting beta-agonist, has the potential of being an effective bronchodilator with less central nervous system stimulation. The effects of theophylline and salmeterol on sleep quality have not been directly compared in COPO patients, but one study of patients with asthma showed only a minimal advantage with salmeterol (slightly fewer arousals). Thus, both of these medications may be useful in patients with COPO and nocturnal/early morning dyspnea. Another alternative is to use an increased dose of ipratropium bromide at bedtime (4 puffs qhs). This medication has few systemic side effects, and the higher dose may give a longer duration of effective bronchodilation. One study of ipratropium (0.02% solution) nebulized qid found an improvement in nocturnal oxygenation and an increase in REM sleep compared to placebo in a group of patients with moderate to severe COPO. A new, long-acting anticholinergic, tiotropium, is now available in the U.S. and may be useful in treating nocturnal symptoms in patients with COPO. Some patients with COPO may still complain of disturbed sleep despite optimal medical management. They often request sleeping pills, and many take over-the-counter medications. The question arises: Are hypnotics safe in these patients? Numerous studies of hypnotic medications have found minimal worsening of nocturnal saturation. One important caveat is that most of the patients in these studies were stable (no acute exacerbations) and nonhypercapnic. Hypnotics can worsen obstructive sleep apnea; therefore, recipients should not have the overlap syndrome. This said, one would probably want to use shorter-acting benzodiazepines (triazolam, temazepam) or the nonbenzodiazepines zolpidem or zaleplon. Other alternatives would be to use sedating tricyclic antidepressants (Sinequan and others) in low doses. Certainly each case must be individualized. Question patients carefully about morning confusion or memory loss. In the present case, the patient reported that theophylline made him "jumpy." He was switched from theophylline to salmeterol (2 puffs every 12 hours). He also was given a limited supply of zolpidem 5 mg to use on occasional nights when he was unable to fall asleep. On this regimen the patient noted improved sleep most nights. Clinical Pearls I. Patients with COPO have poor sleep quality with a low sleep efficiency and reductions in REM and slow wave sleep. 2. There is conflicting evidence about whether oxygen therapy improves sleep quality in hypoxemic patients. 3. Inmany patients the benefits of long-acting bronchodilator medications may balance potential side effects secondary to central nervous system stimulation. 4. Some patients with complaints of insomnia on theophylline may improve on longacting inhaled beta-agonists. Ipratropium bromide may also be useful. 5. Short- or intermediate-duration benzodiazepine hypnotics and the nonbenzodiazepine hypnotics usually result in only minimal worsening in breathing during sleep in stable nonhypercapnic patients with COPO. Caution still is indicated. 225

REFERENCES I. Calverly PMA, Brezinova V, Douglas NJ, et al: The effect of oxygenation on sleep quality in chronic bronchitis and emphysema. Am Rev Respir Dis 1982; 126:206-210. 2. Berry RB, Desa MM, Branum JP, et al: Effect of theophylline on sleep and sleep-disordered breathing in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 1991; 143:245-250. 3. Girault C, Muir JF, Mihaltan F, et al: Effects of repeated administration of zolpidem on sleep, diurnal and nocturnal respiratory function, vigilance, and physical performance in patients with COPD. Chest 1996; 110:1203-1211. 4. Selby C, Engleman HM, Fitzpatrick MF, et al: Inhaled salmeterol or oral theophylline in nocturnal asthma. Am J Respir Crit Care Med 1997; 155:104-108. 5. Martin RJ, Bartelson BL, Smith P, et al: Effect of ipratropium bromide treatment on oxygen saturation and sleep quality in COPD. Chest 1999;115: 1338-1345. 226

PATIENT 70 A 52-year-old man with chronic obstructive pulmonary disease and pedal edema A 52-year-old man was being treated for severe COPD with bronchodilators and continuous oxygen therapy at 1 Llmin. Despite this treatment he had severe, persistent pedal edema and CO 2 retention. Large doses of diuretics had not improved the pedal edema. His wife reported that he snored and fell asleep in front of the television during the day. The patient attributed this to poor sleep at night. Physical Examination: Height 5 feet 9 inches, weight 200 pounds. Blood pressure 150/90 mmHg, pulse 88. HEENT: edematous uvula, dependent palate; 17-inch neck circumference. Chest: bilateral wheezes. Cardiac: distant heart sounds. Extremities: 3+ pedal edema. Laboratory Findings: Spirometry: FEY I 1.7 L (46% of predicted), FYC 3.0 L (64% of predicted), FEY /FYC 0.57. ABG: pH 7.36, PC02 55 mmHg, P02 58 mmHg on 1 Llmin of oxygen by nasal cannula. Chest radiograph: large pulmonary arteries, no pulmonary edema. Figure: Below is a trace of the initial part of a nocturnal recording of Sa02 (on oxygen at 1 Llmin by nasal cannula). Question: Would complete polysomnography be useful? A 15 minutes time • Lights out ..--.. 100 0 C 90 0 += 80 0 '- ::J 70 -+- 0 en 60 C 50 <D 0> 40 0 30 0 ">:: 20 <D t: « 10 227

Answer: Yes. A sleep study (diagnostic/nasal positive airway pressure titration) is indicated in this patient with the overlap syndrome. Discussion: This patient fits the classic description of the "blue bloater" variant of COPD (hypercapnia plus cor pulmonale). Recently, it has been appreciated that many such coPO patients also have obstructive sleep apnea (OSA). Such patients are said to have the overlap syndrome (COPD+ OSA). This combination has important consequences in terms of morbidity and treatment. A low baseline POz and/or ventilation-perfusion mismatch secondary to COPD results in more significant oxygen desaturation during apnea. Thus, OSA patients with COPD tend to have more severe arterial oxygen desaturation and cor pulmonale. Most OSA patients do not have daytime CO? retention. Those that do usually have a component of COPD or the obesity hypoventilation syndrome. However, the presence of OSA may impair ventilatory drive, so that while patients with hypercapnic COPD usually have an FEV I < I L, overlap patients may retain CO2 with more moderate degrees of airflow obstruction. The amount of CO? retention in overlap patients does not necessarily correlate with the AHI. One study comparing hypercapnic and nonhypercapnic patients with the overlap syndrome found no difference in the FEVI and AHI between the two groups. The hypercapnic group was heavier and had a history of heavy ethanol use. The authors hypothesized that the hypercapnic patients had depressed respiratory drives possibly secondary to the effects of alcohol. As effective treatment of OSA in patients with the overlap syndrome can result in a reduction of daytime PCOz, nocturnal COzretention secondary to apnea probably is contributory to the development of hypercapnia in overlap patients. What are the implications of this combination of diseases (COPD + OSA)? The most obvious is that both diseases need to be treated for optimal results. Thus, the daytime gas exchange of the patient with OSA and mild COPO may improve when bronchodilator therapy is added to treatment of OSA (nasal CPAP or bilevel therapy). In the patient with 228 COPD and significant sleep apnea, treatment with oxygen alone may not adequately prevent nocturnal oxygen desaturation. In this group, low-flow oxygen also can result in significant nocturnal increases in CO? Studies suggest that oxygen alone does not improve right-sided heart failure if apnea is not adequately treated. Conversely, OSA patients with low POzvalues when awake may need oxygen plus nasal positive pressure to prevent nocturnal desaturation. When the upper airway obstruction of hypercapnic patients with the overlap syndrome is adequately treated, the daytime PCOz frequently improves. As previously discussed in Patient 45, some patients with COPD find bilevel pressure more tolerable than CPAP. When should a sleep study be ordered in a patient with known COPD? The major indication is a suspicion of sleep apnea (snoring, daytime sleepiness). Another is evaluation of a patient already using nocturnal oxygen who is not showing improvement in cor pulmonale. If resources are limited, a simple overnight pulse oximetry test may show the sawtooth pattern suggestive of sleep apnea. However, this approach can be more costly in the long run if a nasal positive airway pressure titration is then required. In the present patient, oximetry showed the sawtooth pattern of desaturation consistent with OSA (refer to figure). The worsening of SaOz at A is probably secondary to an episode of REM sleep. Note the low baseline SaOz and the persistence of decreased SaOz after each event. The patient underwent polysomnography while receiving his usual oxygen. An AHI of 50/hr was documented. Nasal CPAP was titrated, and apnea and hypopnea were abolished at 10 em HzO. Addition of oxygen at 2 Umin was needed to maintain SaOz above 90%. Treatment with a combination of nasal CPAP and oxygen was begun, and the patient's daytime sleepiness and pedal edema improved. After I month of nasal CPAP, the daytime PCOz had decreased to 45 mmHg.

Clinical Pearls I. A saw-tooth pattern on the Sa02 trace in a patient with capo suggests that OSA also is present. 2. If a capo patient with only moderately severe airflow obstruction has daytime COo retention, suspect the presence of the overlap syndrome. - 3. In the overlap syndrome, adequate treatment of both capo and OSA is required. 4. Supplemental oxygen therapy of nocturnal desaturation alone is not optimal treatment for most patients with capo and significant OSA. 5. Nasal positive airway pressure (CPAP or bilevel pressure) plus oxygen if needed is probably the most effective therapy for patients with capo and significant OSA who have a low P02 when awake. REFERENCES I. Bradley TD, Rutherford R, Lue F. et al: Role of diffuse airway obstruction in the hypercapnia of obstructive sleep apnea. Am Rev Respir Dis 1986; 134:920-924. 2. Fletcher EC. Schaaf JW. Miller J, Fletcher JG: Long-term cardiopulmonary sequelae in patients with sleep apnea and chronic lung disease. Am Rev Respir Dis 1987; 135:525-533. 3. Chan CS. Grunstein RR. Bye PTP. et al: Obstructive sleep apnea with chronic airflow limitation: Comparison of hypercapnic and eucapnic patients. Am Rev Respir Dis 1989; 140:1274-1278. 4. Sampol G. Sagales MT, Roca A. et al: Nasal continuous positive airway pressure with supplemental oxygen in coexistent sleep apnea-hypopnea syndrome and severe chronic obstructive pulmonary disease. Eur Respir J 1996; 9:111-116. 229

PATIENT 71 A 35-year-old woman with asthma and poor sleep at night A 35-year-old woman had been treated for moderate-to-severe asthma since age 15. Her medications included sustained-action theophylline 450 mg qam and 300 mg at 6 PM, inhaled albuterol 2 puffs qid, and inhaled f1uticasone 220 meg/spray 2 puffs bid. Her symptoms were improved with higher doses of f1uticasone, but she was worried about potential side effects. During several severe exacerbations the patient was treated with oral prednisone. She had gained about 15 pounds in the last 2 years. Over the past year, the patient had reported frequent awakenings with shortness of breath. In the morning she felt sleepy and somewhat wheezy. The patient's roommate reported loud snoring and gasping for air during the night. Physical Examination: Height 5 feet 2 inches, weight 130 pounds. General: slightly "cushinoid" appearance. HEENT: edematous uvula and palate; 15-inch neck circumference. Chest: bilateral expiratory wheezes. Cardiac: normal. Extremities: no edema. Laboratory Findings: Spirometry (3 PM): FEY I 1.8 L (63% of predicted), FYC 3.0 L (90% of predicted), FEY/FYC 0.60. Theophylline level (12 noon): 12.0 mg/ml. Peak Flow Diary (Umin) 6 AM 10 PM AWAKENINGS* 6 AM 10 PM AWAKENINGS* Day I Day 2 Day 3 200 225 225 300 325 300 3 2 3 Day 4 Day 5 Day 6 225 180 200 325 300 300 4 3 3 • Awakenings are from the preceding night. Questions: What is your diagnosis? What other evaluation do you suggest? 230

Answers: Nocturnal asthma (documented diurnal variation in airflow). Consider a sleep study to rule out obstructive sleep apnea. Discussion: Nocturnal worsening of symptoms and sleep disturbance are significant problems for patients with asthma. In one study, up to 40% experienced symptoms every night. There is a normal circadian variation in airway function, with the highest airflow in the late afternoon (4 PM) and the lowest in the early morning (4 AM). This normal variation is exaggerated in patients with obstructive airway diseases: FEY1or peak flow can fall as much as 20-40% in the morning hours ("morning dippers"). The etiology of this variation is multifactorial and includes circadian changes in the amounts of circulating steroids and, possibly, inflammatory mediators in the lungs, as well as changes in cholinergic tone. Sleep also appears to have an adverse effect on asthma, independent of other factors. The easiest way to diagnose severe nocturnal worsening of asthma is to have the patient record peak flow measurements at bedtime and upon awakening. Treatment of patients with nocturnal asthma should begin with inhaled corticosteroids. This medication has been shown to reduce the circadian fluctuation in airway tone. Patients with continued nocturnal symptoms despite an adequate dose of inhaled corticosteroids can then be treated with a long-acting bronchodilator. Theophylline has been proven effective despite the stimulatory effects of the medication. In dosing theophylline, the goal should be to obtain the highest levels during the time of greatest airflow obstruction (at night and early morning). Long-acting oral beta agonists have also proven helpful. However, long-acting inhaled beta agonists (salmeterol and formoterol) are probably the most popular long-acting bronchodilators. These drugs are effective treatment for nocturnal asthma and may cause less systemic stimulation and sleep disruption than theophylline. Interestingly, in terms ofsleep quality (arousals), Selby and coworkers found only a slight advantage for salmeterol compared to theophylline: the falls in morning flow rates were similar, but awakenings were less frequent on salmeterol. Weigand et al found salmeterol to be more effective than theophylline at preventing the morning drop in flow rates. The drugs did not differ in polysomnographic findings, but patients perceived better sleep with salmeterol than theophylline. In any case, the longacting beta agonists require less attention to dosing than theophylline. While asthmatics generally have a greater response to beta-agonists than anticholinergic medications, vagal tone is increased at night. Studies have shown an improvement in nocturnal peak flow after ipratropium bromide. A higher bedtime dose (4 puffs) may be needed for a more prolonged duration of action. Finally, asthmatic patients may have other sleep disorders. If obstructive sleep apnea (OSA) is present, adequate treatment may improve the control of asthma as well as symptoms of daytime sleepiness. The reasons that OSA may worsen asthma are at present unknown. In the current case, the peak flow diary confirmed severe morning dipping. The theophylline dosing was changed to 450 mg at 10 PM and 300 mg qam to provide the highest levels when needed. The patient reported fewer nocturnal symptoms, but felt her sleep was not as sound (more awakenings). Theophylline was changed to salmeterol, 2 puffs every 12 hours, with inhaled albuterol as needed between salmeterol doses. Morning peak flows improved to 250-325 Llmin, and the patient reported a decrease in the number of prolonged awakenings to one or less per night. Despite these changes, complaints of falling asleep during the day persisted. A sleep study revealed an AHI of 30/hr. During the second half of the sleep study, a nasal CPAP titration documented elimination of snoring and apnea at 7 em Hp. Treatment with nasal CPAP resulted in improvement in symptoms of daytime sleepiness and further improvement in nocturnal symptoms of asthma. 231

Clinical Pearls I. The degree of diurnal variation in airflow can most easily be documented by peak flow measurements at bedtime and on awakening. 2. When OSA is present in asthmatic patients, adequate treatment of OS A may improve the asthma. 3. The addition of a long-acting bronchodilator is indicated if patients with nocturnal asthma do not respond to inhaled steroids, or if the required dose of inhaled steroids is higher than desired. 4. Long-acting inhaled beta agonists taken at bedtime have been shown to improve morning flow rates as well as, if not better than, theophylline. Patients may also perceive better sleep quality. 5. If theophylline is used, dosing should be such that the highest levels are during the night or the early morning hours. REFERENCES I. Chan CS. Woolcock AI. Sullivan CE: Nocturnal asthma: Role of snoring and obstructive sleep apnea. Am Rev Respir Dis 1988; 137:1502-1504. 2. Beam WR, Weiner DE. Martin RI: Timing of prednisone and alterations of airways inflammation in nocturnal asthma. Am Rev Respir Dis 1992; 146:1524-1530. 3. Martin RI: Nocturnal asthma: Circadian rhythms and therapeutic interventions. Am Rev Respir Dis 1993; 147:525-528. 4. Selby C. Engleman HM, Fitzpatrick MF, et al: Inhaled salmeterol or oral theophylline in nocturnal asthma? Am I Respir Crit Care Med 1997; 155:104-108. 5. Weersink ElM. Douma RR, Postma DS. et al: Fluticasone propionate, salmeterol xinafoate, and their combination in the treatment of nocturnal asthma. Am I Resp Crit Care Med 1997; 155:1241-1246. 6. Weigand L. Mende CN, Zaidel G, et al: Salmeterol vs theophylline. Sleep and efficacy outcomes in patients with nocturnal asthma. Chest 1999;115:1525-1532. 232

FUNDAMENTALS OF SLEEP MEDICINE 16 Central Sleep Apnea Central apnea is defined as a cessation in airflow of 10 seconds or longer that is associated with an absence of respiratory effort. Patients with CSA comprise less than 15% of patients with sleep apnea evaluated at most sleep centers. Some central apneas are common in patients with obstructive sleep apnea (OSA). However, the diagnosis of central sleep apnea (CSA) syndrome requires that the majority of apneic events be central in nature. The exact proportion of central events required is not clear; some physicians diagnose CSA when 50-80% of the events are central. Of note, many patients who exhibit central apneas also exhibit central hypopneas (reduction in airflow associated with a reduction in inspiratory effort) or simply hypoventilate during sleep with relatively few discrete events. Unless you measure esophageal pressure, it may be difficult to differentiate central from obstructive hypopnea. However, central hypopneas usually exhibit a rounded nasal pressure signal profile (Fig. 1). nasal pressure chest abdomen -'\,-/\..-JVVV~--'\---'\--'\..-/\,..JVVVVVll\l\r.l\,-J\,.f\-..J\..- -'\,-/\..-J\,..-'V'1r~-'\---".-"\..-~.I\,-J\,.f\-..J\..- FIGURE 1 The CSA syndrome affects a heterogeneous mix of patients that can be subdivided into a group with daytime hypoventilation (hypercapnic) and a group without hypoventilation (non-hypercapnic; see table). The group with hypoventilation includes patients with a defect in ventilatory control (primary central alveolar hypoventilation) and patients with neuromuscular disorders. These patients usually have a history of bouts of respiratory failure and cor pulmonale. The nonhypercapnic group is composed of patients with Cheyne-Stokes breathing (usually secondary to congestive heart failure [CHF]) and patients in whom no obvious cause for the CSA exists (idiopathic CSA). Patients With Central Sleep Apnea Syndrome HYPERCAPNIC NON-HYPERCAPNIC Idiopathic central hypoventilation Neuromuscular diseases Brainstem disease Idiopathic central sleep apnea Cheyne-stokes breathing • Associated with congestive heart failure • Associated with neurological disease Sleep at high altitude During nasal CPAP titration 233

Cheyne-Stokes breathing (CSB) is a specific type of periodic breathing that is associated with a crescendo-decrescendo pattern of ventilation, with central apneas (Fig. 2) or central hypopneas at the nadir. Arousals tend to occur near the maximum point of ventilatory effort (sometime after apnea cessation) in Cheyne-Stokes breathing. In idiopathic central apnea, arousals tend to occur at apnea termination, and ventilatory drive returns abruptly (see Fig. 2). Most cases of Cheyne-Stokes breathing secondary to CHF also have a longer cycle time (longer ventilatory phase between apneas) and a long delay in the nadir in Sa02 (long circulation time). The length of the ventilatory phase tends to correlate with the length of circulation time (decreased cardiac output). The apnea length depends on the amount of hypocapnia at sleep onset. Cheyne-Stokes Breathing A airflow I B I ------.-/~ Idiopathic CSA A airflow A = position of arousal B = delay in saturation nadir FIGURE 2 Patients may develop central apnea during positive-pressure titration. In some patients with combined OSA and Cheyne-Stokes breathing secondary to heart failure, CSB-CSA will be manifest once the obstructive component is eliminated with positive pressure. In other OSA patients undergoing positive-pressure titration, central apneas may occur following arousals from respiratory effort-related arousals (inadequate pressure) or from arousals secondary to excessive pressure or mouth leak. In the first case, higher pressure may eliminate the events. Of note, early studies of OSA patients after tracheostomy documented persistent central apneas that tended to resolve over time. Sleep Apnea and Congestive Heart Failure. The interaction between sleep apnea and the heart has attracted considerable attention. It has been realized that complaints of shortness of breath and frequent awakening at night are not necessarily secondary to paroxysmal nocturnal dyspnea due to CHF. Types of Sleep Apnea in Patients with CHF Obstructive sleep apnea Mixed obstructive and central sleep apnea Central sleep apnea and Cheyne-Stokes breathing (CSA-CSB) One recent study of 450 men and women with CHF identified risk factors for the presence of obstructive and central sleep apnea. Risk factors for CSA included male sex, atrial fibrillation, age> 60, and hypopcapnia (PC02 < 38 mmHg) during wakefulness. Risk factors for OSA included a high body mass index (BMI) in men and increased age in women. Studies have suggested that the presence of CSA in patients with CHF means a worse prognosis. Treatment of patients with OSA and cardiomyopathy has been shown to improve cardiac function. Treatment of patients with CHF and CSA with nasal CPAP was associated with a significant improvement in left ventricular ejection fraction at 3 months and a relative risk reduction of 81% in mortality-cardiac transplantation rates." 234

REFERENCES I. Bradley TO. McNicholas WT. Rutherford R. et al: Clinical and physiologic heterogeneity of the central sleep apnea syndrome. Am Rev Respir Dis 1986; 134:217-221. 2. Malone S. Liu PP. Holloway, et al: Obstructive sleep apnea in patients with dilated cardiomyopathy: Effects of continuous positive airway pressure. Lancet 1991; 33: 1480-1484. 3. Javaheri S. Parker Tl. Wexler L. et al: Occult sleep-disordered breathing in stable congestion heart failure. Ann Intern Med 1995;122: 487-492. 4. Hanly PI, Zuberi-Khokhar NS: Increased mortality associated with Cheyne-Stokes respiration in patients with congestive heart failure. Am I Resp Crit Care Med 1996; 153:272-276. 5. Sin DO. Logan AG. Fitzgerald FS et al: Effects of continuous positive airway pressure on cardiovascular outcomes in heart failure patients with and without Cheyne-Stokes respiration. Circulation 2000; 102:61--66. 6. Van AT. Bradley TO. Liu PP: The role of continuous positive airway pressure in the treatment of congestive heart failure. Chest 2001; 120:1675-1685. 235

.t.'11 Lli l_ I.! j I J I I I J I •• I I I 1 I ~; .,1 Il 'I' J t ~dj'. PATIENT 72 A 58-year-old man with daytime sleepiness A 58-year-old man complained of daytime sleepiness of 2-year duration. His wife reported that he occasionally snored and was a "restless sleeper." There was no history of muscle weakness, orthopnea, pedal edema, or respiratory failure. Physical Examination: Blood pressure 150/85 mmHg, pulse 80, temperature 37°C, respiratory rate 15. General: thin. HEENT: unremarkable; 15-inch neck circumference. Chest: clear to auscultation and percussion. Cardiac: normal. Extremities: no edema. Neurologic: normal. Sleep Study: AHI 35/hr. No periodic limb movements. Figure: Over 70% of the respiratory events (apneas and hypopneas) were similar to the one illustrated below. Question: What is the cause of the patient's daytime sleepiness? C4 - A1 "'Y~'(~'''''\JI'''~'vr!'~~.0MW.0~1II,~~~\IiI~YII~",l1{o.ti,i'''vJl,f'r';'·+J O 2- A1 .~~'lt-WW...,J;~~;A'I...l'l~w.~-.A ROC-A1 ~t~·,,~ LOC-A1 "'~~jv"'~vt-'I'''fl~~'''''Y~+>tM.~I'N EMG Airflow Chest Abdomen --------------- FIGURE 1 236

Diagnosis: Idiopathic central sleep apnea. Discussion: Central sleep apnea (CSA) occurs because the PCO? is below the apneic thresholdthe level of PCO~ during sleep below which ventilation is absent. -The apneic threshold is usually only 1-2 mmHg below the awake PCO? values. Most patients in stable sleep have PC02 levels around 5 mmHg higher than the wake values. Hyperventilation during wakefulness does not cause apnea because of the presence of the wakefulness stimulus-a poorly defined but important component of ventilatory drive that is lost during sleep. During NREM sleep, ventilation depends on metabolic control. If the PC02 falls below the apneic threshold for any reason (even in normals), central apnea is the result. The presentation of idiopathic CSA is somewhat variable, including complaints of insomnia, daytime sleepiness, or choking during the night. In a recent series, the symptom of excessive daytime sleepiness was the major presenting complaint. Snoring may occur in idiopathic CSA, but is less prominent than in OSA. Patients with idiopathic CSA also tend to be thinner than those with OSA. Ofthe patients with non-hypercapnic CSA, Cheyne-Stokes breathing is more common than idiopathic CSA. In one study, only 5% of over 300 patients with sleep apnea had idiopathic CSA. Polysomnography in idiopathic CSA typically reveals frequent, isolated central apneas or runs of central apneas (one form of periodic breathing). A run of central apneas may follow arousal from a non-respiratory stimulus. Central apneas during NREM sleep occur most commonly in stage I or 2 sleep. Central apnea is believed to occur because the PC02 level is below the apneic threshold (the lowest PC02 triggering ventilation during sleep). This is consistent with the findings that most patients with idiopathic CSA have relatively low PC02 values when awake and asleep, and that central apneas usually follow periods of increased ventilation. In Figure 2, a central apnea (*) follows a big breath (arrow). Patients with idiopathic CSA also have increased ventilatory responses to CO2 compared to normal individuals. In a recent study, an increase in the baseline C4-Al 02-Al~ ROC-Al :;;::~ LOC-A2 chin EMG flow chest abdomen 5002 94% FIGURE 2 awake PC02-by either CO2 administration or the addition of dead space-decreased the amount of central apnea. The periods of increased ventilation triggering central apneas often are associated with arousal. Arousal may trigger a transient increase in ventilation and a fall in PC02. This transient fall in PC02 is then associated with a central apnea as the patient returns to sleep (below the apneic threshold). Thus, arousal may initiate or predispose to continuation of central apnea. In some patients with idiopathic CSA, central apnea occurs mainly in the supine position. Some investigators have hypothesized that reflexes triggered by upper airway collapse may inhibit respiration in these patients. These may be the reasons that some patients with CSA respond to nasal CPAP treatment. In the present patient, over 70% of the respiratory events were central apnea (Fig. 1). Note the absence of movement in the chest and abdominal tracings. The predominance of central apneas that were not of the Cheyne-Stokes type and the absence of symptoms or signs of congestive heart failure resulted in a diagnosis of idiopathic CSA. (See Patient 73 for a discussion of treatment for idiopathic CSA). 237

Clinical Pearls 1. Idiopathic eSA occurs in nonhypercapnic patients without an obvious associated disease (neurologic disorder or congestive heart failure). 2. While many patients with idiopathic eSA complain of excessive daytime sleepiness and have a history of snoring, some complain primarily of insomnia (frequent awakenings). 3. Idiopathic eSA is uncommon and these patients comprise less than 5% of patients with sleep apnea. 4. Patients with idiopathic eSA have low daytime peo2 levels. Arousal from sleep may trigger several large breaths with a subsequent central apnea. REFERENCES I. Dempsey JA. Skatrud JB: A sleep-induced apneic threshold and its consequences. Am Rev Respir Dis 1986; 133: 1163-1 170. 2. Xie A, Wong B, Phillipson EA, et al: Interaction of hyperventilation and arousal in pathogenesis of idiopathic central sleep apnea. Am J Respir Crit Care Med 1994; 150:489-495. 3. Xie A, Rutherford R, Rankin F, et al: Hypocapnia and increased ventilatory responsiveness in patients with idiopathic central sleep apnea. Am J Resp Crit Care Med 1995; 152: 1950-1955. 4. Xie A, Rankin F. Rutherford R, et al: Effects of inhaled CO, and added dead space on idiopathic central sleep apnea. Am Rev Respir Dis 1997; 82:918-926. - 238