Christian Guilleminault and Pierre Philip. Stanford University Sleep Disorders Center, Palo Alto, California, U.S.A.

Transcription

1 , Sleep, 19(9):SI17-S American Sleep Disorders Association and Sleep Research Society, Tiredness and Somnolence Despite Initial Treatment of Obstructive Sleep Apnea Syndrome (What to Do When an OSAS Patient Stays Hypersomnolent Despite Treatment) Christian Guilleminault and Pierre Philip Stanford University Sleep Disorders Center, Palo Alto, California, U.S.A. Summary: From a database of 4,129 patients with sleep-disordered breathing (SDB), we found 207 subjects (43 women) that still complained of daytime tiredness, fatigue, and/or sleepiness despite treatment. In 25 subjects the sleepiness developed I to 36 months following treatment and was related to noncompliance (8 subjects), significant weight increase and/or inappropriate treatment (10 subjects), or development of new medical problems (7 subjects). In the remaining 182 subjects, sleepiness was noted within 1 month after what was judged appropriate treatment for SDB. In this group, the reason for persistent complaint was divided into four categories: I) inappropriate treatment as a result of not using the measurement of esophageal pressure (P,,) in the initial diagnosis (41 subjects), 2) nonfunctional treatment (3 subjects), 3) associated narcolepsy-like syndrome (2 subjects), and 4) emergence of obesity and/or periodic leg movements as significant factors (135 subjects). The 135 subjects in this last category could be subdivided into three subgroups: I) younger subjects, severely overweight with lower mean nocturnal saturated arterial oxygen (Sa0 2 ) values; 2) older subjects, of normal weight, with high numbers of periodic leg movements (PLM); and 3) moderately overweight subjects, with a combination of PLM and lower mean Sa0 2 values than expected. Treatments were aimed at eliminating the identified problems; they included standard medications for PLM and nasal bilevel positive airway pressure (BiPAP) for low Sa0 2 measurements. These treatments were not effective in specific cases, and stimulant medications had to be prescribed, Key Words: Daytime somnolence-sleep-disordered breathing-nasal CPAP-Nasal BiPAP-Periodic leg movements-sleeping position Drug treatment-stimulants. 1 ~) Excessive daytime sleepiness (EDS) is a major complaint in patients with obstructive sleep apnea syndrome (OSAS). Clinicians and patients expect that appropriate treatment, whatever the type, will eliminate symptoms of the illness. However, a significant number of patients who supposedly have been appropriately treated for their syndrome still complain of daytime hypersomnolence. This is an important clinical problem because subjects may still be at risk of driving and job-related accidents even after "appropriate treatment". To investigate this complaint, we used our clinic database to generate a list of over 4,000 patients seen between January 1990 and August We surveyed this group to find patients who still presented a com- Accepted for publication July Address correspondence and reprint requests to Christian Guilleminault, M.D., Stanford University Sleep Disorders Center, 701 Welch Road, Suite no. 2226, Palo Alto, CA 94304, U.S.A. S117 plaint of daytime somnolence or daytime fatigue following initial treatment. For this study, we reviewed the initial treatment of these subjects, analyzed the results of a new polysomnographic investigation, and identified appropriate treatments following this new investigation. Sleep clinic database findings We identified 207 patients (43 women), about 5% of the total surveyed, with the complaint of hypersomnolence despite having had "appropriate treatment" of the upper airway resistance syndrome (UARS) or OSAS. Here, "appropriate treatment" indicates full treatment by a sleep clinic physician, including 1) in the case of surgery, a 6-month follow-up clinical and polysomnographic investigation after the planned last phase of the surgery; or 2) one recording night at the time of the initial nasal continuous positive airway

2 S118 C. GUILLEMINAULT AND P. PHILIP pressure (CPAP) treatment or the initial fitting of a dental appliance, with a second monitoring 1 month after this treatment initiation phase. In the majority of the subjects (182, or 88%) the complaint of hypersomnolence was noted at the time of this second, post-treatment, recording (i.e. within 1 month of the initial treatment recording). In the remaining 25 OS AS patients, the hypersomnolence complaint was reported between 2 and 36 months after the initial treatment monitoring. The mean age of the hypersomnolent follow-up patient group was 52 ± 14 years [not significantly (ns) different from that of the total group]. The initial respiratory disturbance index (RDI) was 43 ± 16 (ns), and the mean body mas~ index (BMI) was 29.8 ± 9 kg/m2 (the median was 32.1 kg/m2). This last measure shows a trend (p < 0.07, ns) toward a higher BMI. (The large standard deviation and the size of the total group vs. this subgroup are factors in the absence of statistical significance using a nonparametric Wilcoxon rank sign test.) All 4,129 OS AS subjects underwent the following in their original clinical evaluations: Clinical interview/evaluation. General clinical evaluation included oro-naso-facial clinical investigation, a standardized validated questionnaire on sleep disorders, and an evaluation of wakefulness (SQAW) (1). All-night recording. Polysomnographic recording systematically included electroencephalogram (C3/A2- C4/Al), electro-oculogram, chin and leg electromyogram, electrocardiogram (one lead), oro-nasal airflow (thermistors), non-calibrated inductive respiratory plethysmography (thoraco-abdominal bands,) pulseoximetry, video surveillance, and body position determination. Following a positive diagnosis of OSAS, patients were treated with nasal CPAP (after calibration), upper airway surgery, or a dental appliance. Follow-up evaluations were performed within 1 month for non-surgical cases or within 6 months for surgical cases after treatment. Persistent hypersomnolent group Subjects who had a persistent complaint of hypersomnolence, daytime tiredness, or fatigue were reevaluated through clinical interview and examination. Subjects who had received their initial treatment after 1992 were asked to fill out the Epworth sleepiness scale (ESS), so we also requested this at the time of reevaluation. The new polysomnographic recording included the same variables as those present at the time of the initial recording, with the addition of esophageal pressure (Pes) determination. (From 1993 onward, 70% of patients had systematic Pes monitoring at entry.) In addition, subjects underwent the Multiple sleep latency } <; 11 ~ C> c: ~ " '" [ '0 " Time (in minutes) 10 _ Subgroup 1 _ Subgroup2 -t#- Subgroup3...lI- Subgroup 4 FIG. 1. Results of the psychomotor vigilance task (PVT) tests. Shown are the number of lapses lasting 400 milliseconds or longer. Each group included 10 SUbjects. Subgroup 4 was a group of subjects treated with nasal CPAP. They were recruited from the database, and they are non-complainers, with same mean age and mean BMI as those of the 30 patients presented here. There were 10 patients per subgroup (I to 3). Each of the subgroups were formed by age matched subjects defined by cluster analysis (see text). test (MSLT) and, for those who entered the clinic after 1994, a psychomotor vigilance task (PVT; a tapping test administered for 10 minutes during which the time interval between two tapping movements is determined automatically by a computer) (Fig. 1). We measured nerve conductance in a subgroup of subjects. In addition, the following demographic and clinical variables were systematically analyzed: BMI, neck circumference, clinically determined oro-facial features, clinical symptoms associated with sleep disorders, SQAW questions related to hypersomnolence, family history of sleep disorders, past clinical history, medication intake, habitual drug use (alcohol, smoking, other drug intake), treatment compliance (questionnaire and "CPAP-counter" indicating the number of hours the CPAP machine was turned on for those patients treated from 1992 to date), ESS at entry and at reevaluation, polysomnographic variables, MSLT results, and, when available, the results of PVT and other tests (including barium swallow, esophageal ph determination, gastro-esophageal evaluation, 5 days of actigraphy, and, if not previously performed, spirometry in the supine position). Data analysis The "persistent sleepy and/or fatigue" group was subdivided, according to the results of the data collection, into subgroups based on clinical judgment. Cluster analyses were also performed with the use of the collected variables.

3 HYPERSOMNOLENCE DESPITE OSAS TREATMENT S119 '(.' ~I ~.', Analysis of group A: developed hypersomnolence after standard follow-up evaluation The first analysis considered the subgroup of subjects (group A) who reported fatigue and sleepiness after initial improvement for at least 2 months and up to 36 months. This subgroup included 25 subjects with a mean age of 41 :t 5 years, and mean RDI of 38 :t 12. The mean time between initial treatment and new investigation was 13 :t 6 months; it ranged from 2 to 36 months. This was a heterogeneous group in terms of treatment: 4 subjects were treated with dental appliances, 3 had upper airway surgical treatment, and 18 subjects were recommended for nasal CPAP therapy. Eight of the 25 subjects were noncompliant with their treatment. Two with dental appliances were using their apparatus only intermittently, due to pain and poorly fitted appliances; six of those with nasal CPAP therapy used it only intermittently or had regular interruption of CP AP treatment early in the night. One subject was an elderly woman who woke up during the night with some degree of confusion and disorientation and was unable to adjust to the equipment. Three subjects were claustrophobic with the mask and Adam-circuit; two subjects developed mask-related chronic skin lesions and allergic reactions. None of these subjects, however, complained or reported the primary problem, but each came back for "persistent somnolence". Ten subjects had drastically increased their weight (the mean increase was 17 :t 5 kg) compared to their weight at the time of initial treatment. Two patients who had had initial positive results from surgery were in this category. The last seven subjects presented with additional medical complications. One surgical patient developed severe asthma, requiring continuous steroid treatment; he suffered both significant weight increase and poorly controlled, severe, lower airway disease. Two patients had had myocardial infarctions, one of them with secondary cardiac failure (despite initial appropriate nasal CPAP treatment). Two subjects greatly increased their daily alcohol consumption, with significant worsening of the symptoms of chronic alcoholism. Two patients presented a worsening of chronic obstructive lung disease (COLD), with development of clear dyspnea during wakefulness. New polysomnographic evaluations indicated the presence of upper airway partial or complete occlusion during sleep despite prior surgical success in 3 subjects or inappropriate low pressure in 10 subjects treated with nasal CPAP (8 with increased obesity and 2 with significant alcoholism). In four subjects the new monitoring indicated changes in the polysomnographic pattern. These changes included the presence of repetitive central apnea or hypopnea ("Cheyne-Stocking") and/or greater drops in oxygen saturation (myocardial infarctions and COLD), particularly during rapid eye movement (REM) sleep, with significant sleep fragmentation. In summary, lack of compliance to treatment or new medical problems explain the reappearance of daytime somnolence in this subgroup. Analysis group B: complained of hypersomnolence at standard follow-up evaluation The other analyses were performed on subjects (group B) who were still complaining of daytime sleepiness or daytime fatigue at the 1-month planned follow-up investigation. This subgroup was the larger of the two and included 182 subjects. Group B was significantly more overweight than group A, with mean BMI of 31 :t 4 kg/m2 (vs. 27 :t 3 kg/m2, p < 0.05, Wilcoxon rank sign test). This group included 7 subjects treated with dental appliances and 175 subjects treated with nasal CPAP. Nasal CPAP pressure ranged from 7 to 15 cm H20. To establish the documentation of EDS, MSLT was obtained on 180/182 subjects and ESS on 122/182. Despite the complaint of tiredness, fatigue, and sleepiness, the mean ESS score was 12 :t 2, and the sleep latency was 10.5 :t 2 minutes. The MSLT results were therefore borderline normal. The ESS indicated low abnormal scores. Thus there was a discrepancy between the subjective complaint of EDS and the scores obtained from the MSLT and from ESS. New polygraphic recordings and family member interviews revealed the following: In four subjects, the dental appliances were often dislodged during nocturnal sleep, particularly during REM sleep, due to normal loss of muscle tone. In three subjects, esophageal pressure measurements (PeJ revealed that, despite the disappearance of distinct apneas and hypopneas in the nocturnal recordings, the subjects still presented an abnormal degree of upper airway resistance. This could only be demonstrated with the measurement of Pes> which had not been used systematically during the initial patient evaluations. Similarly, it was found that 35 subjects titrated with nasal CPAP without the use of Pes had inappropriate CPAP settings. The settings were too low in 31 subjects, who showed an abnormal level of airway resistance, with increases in respiratory effort, as indicated by Pes measurements. The opposite, that is, abnormally high titrations, was noted in seven subjects, with the presence of short central hypopneas (defined as a significant decrease in respiratory efforts as indicated by Pes monitoring for at least one to two breaths, presence of arousals, and long awakenings.) Of the last 137 subjects, 2 presented with suspicion of a narcolepsy-like syndrome. The history of cata-

4 S120 C. GUILLEMINAULT AND P. PHILIP plexy was poor, but probably present, with rare events of muscle weakness involving the knees and legs, for 5-30 seconds, when there individuals were placed in a specific stressful situation (competition with siblings in one case and arguing with strangers in the other). Sleep paralysis and hypnagogic hallucinations were reported in both cases, as well as the presence of dreams at nap onset. Polysomnography with Pes measurements indicated adequate control of the breathing disorder, and the following day's MSLT indicated a mean sleep latency of 6 minutes and 6 minutes 30 seconds, respectively, with three sleep onset REM periods. The remaining 135 subjects could be subdivided into two major subgroups: 1) those with severe obesity and 2) those with normal weight or moderate obesity. 1) The subjects with major obesity (n = 61) had a mean BMI of 37 ± 4 kg/m2 (range ). 2) The second group (n = 74) had a mean BMI of 28 ± 3.5 kg/m2 (range 24-30). Cluster analysis performed on these 135 subjects indicated one group of low-bmi (mean 27 kg/m2) older subjects (mean age 54 years) with lower-pressure nasal CPAP (mean 8 cm H20) and appropriate mean values of saturated arterial oxygen (Sa02) under CPAP treatment (mean 95%), with a mean of 133 periodic leg movements during sleep. Another cluster defined a subgroup with a high BMI (mean 41 kg/m2) higherpressure nasal CPAP (mean 15 cm H20), younger age (mean 36 years), low mean Sa02 (mean 90%), median high CPAP (mean 12 cm H20) and a mean of 54 periodic leg movements during sleep. Two other clusters could be determined using the number of periodic leg movements as the main variable: one group with a high number of periodic leg movements (mean 191), older age (mean 51 years), low BMI (mean 25.8 kg/m2), and high mean Sa02 (96%), and a group with a low mean number of periodic leg movements (mean 24), younger age (mean 36 years), high BMI (mean 44 kg/m2), and low mean Sa0 2 (89%). Cluster analyses indicated that several major factors could be found in our 135 subjects, with a clear variable distribution. Using periodic leg movement, obesity, mean Sa02, and age, several subgroups could be defined: Subgroup 1. Very obese younger subjects with low mean Sa0 2, low, if any, periodic leg movements, and high nasal CPAP (n = 47). Subgroup 2. Normal weight older subjects with normal mean Sa02, lower nasal CPAp, and a high number of periodic leg movements during sleep (n = 27). Subgroup 3. Intermediate between subgroups 1 and 2, subjects with moderate obesity, intermediate mean Sa0 2, moderate number of periodic leg movements, and quite variable nasal CPAP (n = 61). We investigated daytime sleepiness as a function of TABLE 1. Subgroup I Subgroup 2 Subgroup 3 Daytime sleepiness in each of the three subgroups described in the text Mean MSLT (min- Nocturnal polysomnographic utes, Mean SQAW Mean sec- ESS (I to W ASO Mean Sa0 2 onds) (score) 5) (minutes) no. PLM (%) 8,40 11,30 11, % 96% 94% MSLT, multiple sleep latency test (the time indicates the mean sleep latency from the MSLT); ESS, Epworth sleepiness scale; SQAW, sleep questionnaire and assessment of wakefulness; WASO, wake after sleep onset; PLM, periodic leg movements. each subgroup. Table 1 presents the results obtained. As can be seen, despite their complaints, neither the ESS nor the MSLT results were severe. Subjectively, however, symptoms of fatigue were still important. In summary, Periodic leg movements and/or lower than expected Sa02 were the two major polysomnographic nocturnal findings in these three subgroups. Features of "crescendos" when measuring Pes were not seen. Tachypnea during sleep was noted in obese subjects. One can assume that persistent daytime somnolence was related to these findings. Psychomotor vigilance task tests were performed on 10 age matched subjects drawn from each of 1 to 3 subgroups and were compared to an age matched, appropriately treated with nasal CPAp, control group (see Fig. 1). Treatment approaches and response to treatments Our treatment approach was guided by the findings of our study: 1) Non-compliant patients were offered another mode of treatment (usually surgery). 2) CPAP patients with inappropriate air pressure levels were retitrated using Pes to identify persistence or reappearance of abnormal upper airway resistance. 3) Associated medical problems were treated with the help of specialists. Stimulants were prescribed in the narcolepsy-like cases. Periodic leg movements were treated if they were numerous (present in more than one-third of recordings) and/or if present with arousals (n = 61). Our first approach to treatment was L-DOPA with carbidopa (Sinemet ). Our usual initial dosage was a mg tablet given 20 minutes before bedtime. None of our patients is currently at this low dosage. The slow release 200/50 mg (Sinemet ) is currently used by most of our patients. Dosage requirements may vary, and mg may be added if patients wake up in the middle of the night or early in the morning. Failure of treatment. About 16% (n = 10) of the I..

5 HYPERSOMNOLENCE DESPITE OSAS TREATMENT S121 Of '.,, ) t) patients treated with Sinemet developed problems again or did not respond to treatment. These patients were switched to Pergolide, with a good response for most (n = 6/10.) The four patients who did not respond were treated with codeine at bedtime. Two subjects, despite appropriate treatment of periodic leg movements and the absence of obesity, still complained of daytime fatigue and had a borderline MSLT at 11 minutes 30 seconds, with an ESS score of 9. Despite the risk of exacerbating the periodic leg movements, 20 mg Fluoxetine was administered in the morning, with a resulting resolution of the subjective complaint. 4) In the group of obese patients with low Sa0 2, all patients were switched to bilevel positive airway pressure (BiPAP; spontaneous trigger machine). If the mean Sa0 2 was still below 92%, 0.5 to 0.75 I 100% oxygen was bled into the port of the nasal mask. All subjects were retitrated, with measurement of transcutaneous partial pressure of carbon dioxide (PC0 2 ) during sleep. After 1 month of treatment, patients underwent a complete reevaluation with all-night polysomnographic recording, including Pes measurement, and next-day MSLT. All subjects presented a mean Sa0 2 above or equal to 92%. Failure of treatment. Despite the improvement of mean Sa0 2 and the absence of a "crescendo" abnormal breathing pattern while monitoring Pes' or the presence of visually seen repetitive electroencephalographic arousals, including transient arousals, 17 subjects complained again of daytime fatigue. Follow-up MSLTs showed a borderline result in these individuals, with a mean ESS score of 11 and mean MSLT of 9 minutes 15 seconds. The weight range of these subjects was to 227 kg. All subjects were placed on stimulant medication (most commonly mg dextro-amphetamine). Follow-up MSLT with stimulant medication showed a mean MSLT of 12 minutes and 20 seconds. The subjects included in subgroup 3 benefited from periodic limb movements (PLM) treatment and a bilevel positive airway pressure approach. There was no failure. Comments Many workers, including ourselves (2), have already commented upon the potential effect of periodic leg movements on the persistent complaint of sleepiness. We had indicated that, following treatment with nasal CPAp, some patients may improve, whereas others may experience an increase in periodic leg movements. Fry et al. (3), Criollo et al. (4), MacFarlane et al. (5), Shaffer et al. (6), and George et al. (7) have considered the problems and have emphasized that the implementation of nasal CPAP may lead to the emergence of periodic leg movements. George et al. (7) reported that no more than 8% of the noted increase is associated with the use of nasal CPAP. Our belief is that it is not nasal CPAP per se but rather the time spent supine, or near supine, while using nasal CPAP that is the most important factor. We performed nerve conduction and electromyographic studies in about 50% of the patients reported on in this study. As shown before, about half of that subgroup had reduced nerve conduction velocity. The subjects with reduced nerve conduction velocity were more overweight and usually older. Mention of backaches, an unfortunately common complaint, was found in 82% of these cases. When we analyzed the diagnostic night versus the CPAP titration and the CPAP followup nights, we found out that subjects spent two-thirds more time on their backs on CPAP nights when compared to diagnostic nights. In these overweight subjects, with mild lower back problems, and probably mild nerve root trauma, as indicated by nerve conduction velocity studies, one could perfectly explain, using a simple physiological scheme, the increase in periodic leg movements with nasal CPAP therapy. Increased stimulation of the lower motor neurons, due to the increased sensory input associated with repetitive increase in microstimuli, and secondary to increased weight applied to the lower back itself was related to longer supine sleep in these overweight older subjects with poor muscle girdles. The increased lower motor neuron stimulation would partially depolarize the lower motor neuron and partially open a gate for other signals. These signals had been too weak, previously, to lead to a motor response but were additive now and led to periodic leg movements. Independent of periodic leg movements, it is also obvious that appropriate treatment of the upper airway does not completely eliminate daytime sleepiness in patients who present chest wall restriction due to their obesity. As we have already reported (8), the complaint of sleepiness is not necessarily related to the presence of visually scored sleep fragmentation, including determination of transient electroencephalographic arousals (7), and Pes crescendo is often missing, but it does not necessarily eliminate the increased work of breathing during sleep in overweight subjects. As shown here, the complaint of daytime fatigue persists despite the use of BiPAP therapy, and even despite the addition of low flow oxygen given through the nasal mask, with the elimination of any clear hypoxemia without induction of hypercapnea. We believe that these patients expend an abnormal amount of effort breathing. This increase in respiratory effort is currently not detected by the diagnostic techniques used, but we believe that computerized analysis of sleep electroencephalography will demonstrate the Sleep, Vo!' 19, No.9, 1996

6 '.. S122 C. GUILLEMINAULT AND P. PHILIP presence of abnormal sleep features. These studies are currently being performed and final data are not available. Independent of the results, these patients responded well to stimulant treatment. Last but not least, once again we have found a discrepancy between subjective complaints and MSLT results. As often mentioned, there are gray areas when using MSLT scores for the delineation of normal and pathological sleep tendency. Our patients had clear lapses at PVT testing despite MSLT scores between 8 and 11. The need to refine our tests for determination of somnolence is indicated once again. REFERENCES I. Douglass AB, Bornstein R, Nino-Murcia G, et al. The sleep disorders questionnaire 1: creation and multivariate structure of the sleep disorders questionnaire. Sleep 1994;17: Guilleminault C, Crowe C, Quera-Salva MA, Partinen M, Miles L, Nino-Murcia G. Periodic leg movements (PLM) and sleep apnea. Sleep Res 1987;16: Fry JM, DiPhillipo MA, Pressman MR. Periodic leg movements in sleep following treatment of obstructive sleep apnea with nasal continuous positive airway pressure. Chest 1989;36: Criollo M, MacFarlane J, Chapman K, Moldofsky H. Periodic leg movements in sleep and continuous positive aireway pressure in obstructive sleep apnea. Sleep Res 1990;19:148 (abstract). 5. MacFarlane J, Moldofsky H, Shahal B, Chapman K. Periodic leg movements and apneas during CPAP titration. Sleep Res 1990; 19: 248 (abstract). 6. Shaffer JI, Boecker MR, Neeb MJ, Klempert AR, Plenzer SC, Mahajan V. The emergence of periodic limb movements in obstructive sleep apnea patients treated with nasal CPAP. Sleep Res 1993;22:269 (abstract). 7. George CFP. Fergusson KA, Flaherty BA. Periodic limb movements in obstructive sleep apnea. Sleep Res 1995;24:237 (abstract). 8. Guilleminault C, Stoohs R. Kim YD, Chervin R, Black J, Clerk A. Upper airway sleep disordered breathing in women. Ann Intern Med 1995;122: ; 4