Cheyne-Stokes Respiration Presenting as Sleep Apnea Syndrome Clinical and Polysomnographic Features1,2

WILLIAM T. DOWDELL, SHAHROKH JAVAHERI, and WILLIAM MCGINNIS

Introduction

Cheyne-Stokes respiration (CSR) is a form of periodic breathing seen in a wide variety of medical conditions, including decompensated congestive heart failure, cerebrovascular disease, acclimatization to altitude, uremia, and in premature infants (1-7). In the majority of these patients, the ventilatory patterns may not be recognized, and the clinical features are generally dominated by ~he underlying disease process. In contrast, patients with sleep apnea syndrome (SAS) havea well-defined clinical presentation that most commonly includes loud snoring, excessivedaytime sleepiness, obesity, witnessed noctural apneas, and characteristic polysomnographic findings (8-10). During the last year, we have seen a number of patients who presented with clinical features of SAS, but whose polysomnography showeda breathing pattern noteworthy for the presence of CSR, either alone or in combination with obstructive or central sleep apneas. This report attempts to characterize the clinical and polysomnographic presentation of these patients and their response to therapy. A second major goal of the present study was to test the hypothesis that periodic breathing may lead to the development of upper airway obstruction in patients with sleep-disordered breathing. In this context, it has been noted that in patients with SAS, obstructive apneas occur at the nadir of periodic changes in breathing (11-14). Studies of Onal and associates (11) have been shown that in patients with SAS, the activities of the upper airway muscles and the inspiratory muscles of the chest wall decrease proportionately before obstructive apnea occurs. Decreased activity of upper airway muscles results in increased upper airway resistance and consequent in-

SUMMARY This study reports polysomnographic features of five patients with Cheyne-Stokesrespiration (CSR). They were referred for evaluation of presumptive sleep apnea syndrome on the basis of history and physical examination, but were found to have predominantly CSR on all-night sleep study. On the initial polysomnographic study, CSR comprised 47to 86~1l of all disordered-breathing events. Cheyne-Stokes respiration resulted in considerable oxyhemoglobin desaturation (mean baseline saturation was 95 ± 4 ± SO, and lowest saturation was 76 ± 8). More than one-half of all CSR events resulted in awakenings or arousals. Evidence of upper airway obstruction was noted in the majority of CSRevents in three of five patients. Four patients were treated with theophylline; one who refused drug therapy was treated with nasal continuous positive airway pressure (CPAP). Comparison of sleep studies before and after therapy showed a significant decrease in the CSR index (29 ± 11 versus 2 ± 2) and in the maximal oxyhemoglobin desaturation associated with CSR (13 ± 5 versus 3 ± 2), and an improvement in lowest 0, saturation associated with CSR (76 ± 8 versus 91 ± 4). Total disruptions in sleep architecture per hour of sleep improved significantly with therapy (46 ± 21 versus 20 ± 8). We conclude that the clinical presentation of CSR can be indistinguishable from that of the "traditional" sleep apnea hypopnea syndrome and can result in major oxyhemoglobin desaturatlon and sleep fragmentation. Theophylline results in considerable improvement in the disordered breathing of CSR during sleep. AM REV RESPIR DIS 1990; 141:871-879

creased negative intrathoracic pressure, which may result in upper airway collapse. If periodic breathing is a prerequisiteto the development of obstructive apneas, then patients with CSR who suffer from a considerable amount of respiratory periodicity should also be prone to the development of upper airway obstruction during sleep. Occlusive apneas should become particularly prominent in those patients with CSR who have definite central apneas. Because wemeasured esophageal pressure in the current study (see METHODS), we had the opportunity to examine closely the presence of obstructive events associated with CSR.

plaints with regard to sleep and the cardiopulmonary system, physical examination, thyroid function tests, chest roentgenograms, room air arterial blood gases and pH, pulmonary function studies, and cardiac radionuclidegated pool scans. We usually perform these tests routinely in our patients with SAS. The sleep studies of the patients were examined for evidence of sleep apnea and hypopnea. Additional variables examined in detail included indices for CSR with and without central apnea, duration of CSR events, oxyhemoglobin desaturation during CSR, presence of upper airwayobstruction associated with CSR events, and the presence of awakening or arousal upon termination of these events. Ap-

Methods The study population consisted of five patients who were evaluated in our sleep laboratory and were found to have CSR on an allnight sleep study. These patients weregathered from a pool of 52 patients studied in the sleep laboratory from June 1987to June 1988. The patients were examined by at least one investigator. Their medical records were reviewed, with special attention to the reason for referral to the sleep laboratory, presenting com-

(Received in original form April 27, 1989 and in revised form October 13, 1989) 1 From the Pulmonary Section, Department of Veterans Affairs Medical Center, and the Department of Medicine, University of Cincinnati College of Medicine, Cincinnati, Ohio. 2 Correspondence and requests for reprints should be addressed to Shahrokh Javaheri, M.D., Director of Sleep Laboratory, Pulmonary Section (111F), Department of Veterans Affairs Medical Center, 3200 Vine Street, Cincinnati, OR 45220.

871

872

nea and hypopnea indices for central, obstructive, and mixed events that were not associated with CSR were also recorded. All of the patients had full-night sleep studies performed before and after therapy, and the results of those studies were compared. Nocturnal polysomnography was performed using two-channel electro-oculogram (EOG), two-channel electroencephalogram (EEG), and submental electromyogram (EMG) to define the sleep stages, according to the criteria of Rechtschaffen and Kales(15). Respiratory activity was quantified by respiratory inductance plethysmography (Respitrace'"; Ambulatory Monitoring Inc., Ardsley, NY) using coils placed over the rib cage and abdomen. In all but one of the patients, esophageal pressure (PES) was quantified using a thin lO-cm latex balloon positioned in mid-esophagus and connected to a Validyne pressure transducer (Model MP45-14, Validyne Engineering Corporation, Northridge, CA). Airflow was qualitatively monitored usingan oral/nasal thermocouple (Model TCTlR; Grass Instrument Company, Quincy, MA) and an oral/nasal CO 2 gas analyzer (ModeI20089A2 Trimed 510;Biochem International, Waukesha, WI) simultaneously. Hemoglobin oxygensaturation was continuously recorded using an ear oximeter (Biox I1A; BT Inc., Boulder, CO). Data were recorded on a multichannel polygraph (Model 78D; Grass Instrument Company). Disordered-breathing (DB) events during sleep consisted of respiratory events defined as CSR (see below) and other events (the "traditional" apneas and hypopneas) not associated with CSR. Cheyne-Stokes respiration was defined as a periodic breathing pattern consisting of a smooth decrescendo pattern of diminishing tidal volume and PES deflection followed by a smooth crescendo pattern of increasing tidal ventilation and PES deflection (figures 1 and 2). A central apneic period of variable length (at least 10 s) was often present at the nadir of the ventilatory pattern (figure 2). An important component of the CSR event that helps distinguish it from the traditional apnea is the smooth crescendo arm of gradually increasing tidal volume and PES deflection occurring over several breaths (figures 1 and 2) - although the behavior of PES changes if upper airway occlusion occurs during a CSR event (figure 3). In contrast, in an apnea not associated with CSR, maximum respiratory excursion is usually seen with the first or second post-apnea breath. The same phenomenon is observed in hypopnea. At least three breaths were required to be present on either the Respitraces or PES tracing for each limb of the decrescendo-crescendo event of the CSR cycle, although four of five patients had considerably more than three breaths per limb (figures 1 and 2). The duration of CSR events was measured from the most negative deflection of the PES tracing on the decrescendo limb to the most negative PES deflection at the end of the crescendo limb (i.e., from peak to

DOWDELL, JAVAHERI, AND MCGINNIS

peak). In the one patient in whom the esophageal balloon could not be placed, measurements were made from the Respitrace'" sum tracing. Obstructive apnea (OA) was defined as the absence of airflow in the presence of rib cage and abdominal excursions for a period of at least 10 s, measured on the PES tracing from the beginning of inspiration of the first completely obstructed breath to the beginning of inspiration of the first breath associated with airflow (16).Central apnea (CA) was defined as the absence of rib cage and abdominal activity with absent PES deflection for a period of at least 10s, measured on the PES tracing from the end of inspiration of the breath preceding the apnea to the beginning of inspiration of the first breath ending in apnea (16). Mixed apnea was defined as the absence of airflow for at least 10s in conjunction with (1) an initially absent PES deflection for a period of one and one-half times the interval between the two previous inspiratory efforts (16), measured from peak to peak on the PES tracing, and (2) evidence of obstructive apnea after the central apnea. Hypopnea was defined as an episode of at least 10 s in which the amplitude of the sum of the Respitrace" was diminished by at least 50010 from the mean amplitude of the four previous breaths. An event was scored as obstructive hypopnea if at least one of the following three conditions was present: (1) paradoxical rib cage and abdominal wall movement, (2) increasingly negative PES deflection while airflow and Respitrace'" sum remained unchanged, and (3) diminished airflow while PES remained unchanged. An event was scored as central hypopnea if the Respitrace" and PES deflection were diminished and none of the above three conditions was present. All measurements of apnea/hypopnea duration were made from the PES tracing, or from the Respitrace'" sum tracing in the one patient in whom the PES tracing was not available. For all DB events, indices were calculated per hour of sleep. Awakening was defined as the emergence of alpha activity on the EEG tracing for at least 15s. Arousal was defined as alpha activity for a period of 2 to 15s (17). Data were analyzed by paired, two-tailed Student's t test and were expressed as mean ± SD. Differences were considered significant if the p value was 0.05 or less. Results

Demographic data and some clinical and laboratory findings at the time of the first sleep study of the five patients are listed in table 1. The mean age was 58 yr. All patients were obese, with a mean weight of 100.6 kg and a mean body mass index (18) of 32.2 kg/m', All snored loudly, three reported excessive daytime sleepiness, all had a history of witnessed nocturnal apneas by spouse or bed partner, and two reported chronic morning head-

aches. Two patients (both with a history of congestive heart failure, New York Heart Association Functional Class III) reported orthopnea and one of these patients also complained of paroxysmal nocturnal dyspnea. On physical examination, end-inspiratory crackles, third or fourth heart sound gallops, and elevated jugular venous pulsations were uniformly absent. Chest roentgenograms showed cardiomegaly in two patients with a history of congestive heart failure; however, there was no radiographic evidence of pulmonary edema. Cardiac radionuclidegated pool scans showed left ventricular hypokinesis or segmental akinesis in four patients; however, severe left ventricular dysfunction was noted only in two patients (table 1). Pulmonary function studies were normal in three patients. The remaining two patients had mild restrictive ventilatory defects on spirometry with normal diffusion capacities, consistent with obesity. The mean FVC was 3.8 ± 1.1 L (82010 predicted); FEV h 3.1 ± 0.9 L (80010 predicted); FEVt/FVC ratio, 81 ± 5; FRC, 3.0 ± 0.6 L (94010 predicted); and DLco, 26.5 ± 9.6 mllmin/mm Hg (97010 predicted). Pa02 varied from 72 to 88 mm Hg, with a mean of 79 ± 7 mm Hg. Pac02 varied from 37 to 40 mm Hg, with a mean of 38 ± 1 mm Hg. The mean pH was 7.44 ± 0.04. The flow volume loop in Patient 2 showed a sawtooth pattern on both the inspiratory and expiratory limbs. However, this patient had no obstructive apneas or hypopneas on the initial study, and only 7 of 37 (50/0) of his CSR events were associated with some degree of upper airway obstruction. On the initial study, CSR comprised 47 to 86010 (mean 63 ± 17010) of all disordered-breathing events. The remaining events were the usual traditional obstructive, central, or mixed apneas and hypopneas that were not associated with CSR events (table 2). Most CSR events occurred in nonREM sleep; the CheyneStokes Respiration Index (CSRI) during nonREM sleep was 26 ± 10 per hour which was signficantly (p = 0.003) less than CSRI of 2 ± 2 during REM sleep. This finding is similar to that in high altitude (5). Interestingly, however, the duration of the CSR cycles was similar for events occurring in REM and nonREM sleep (50 ± 13 versus 46 ± 12 s, p = 0.20). After the initial sleep study, four patients were treated with theophylline alone. Patient 5 was unwilling to take any medication, and thereafter was treated

EM~

t. . . . .

fIllJeIt'I ....~" '1..149

~ ..~ ••,W.'.ltIi,~.IIII!W~t.~.'Iit.. ~I-. "'W ,...-.. . ,I ....,'........,.. b.i.-....... "'I'iIlI•• h!'."'~"I~~.~b"I.~,,~

"

'1

I

sOntV

I.'.

n

THERMOCOUPLE

I

n

r

1$

..

.......

PES tCInHzOl

5.0 Me

\

I

;

"

I

I

I

I'

I

II

f

"

I

I II!

I

I

,

r

,

, I

,

,

"

,

t

I

r

I

i'l

i"1 ') , 1

h,J~;=J~~~

,

h',

Fig. 1. Representative sleep study recording showing a CSR event without central apnea during Stage II nonREM sleep. EOG = electro-oculogram, EEG = electroencephalogram, EKG = electrocardiogram, CO. = naso-oral CO. tracing representing airflow, RC = rib cage and ABO = abdominal wall movement measured by Respitrace@ belts, RC + ABO = Respitrace@ sum tracing, PES = esophageal pressure, Sao, = arterial oxyhemoglobin saturation. Note the gradual decrease in PES, ABO, RC, and RC + ABO followed by a gradual symmetric increase in these tracings.

I.

7.

1$

-

10

20

.------

r

I' I I 1 ; : I

r' ~ ~ L - J '-----J

I

. ' 'E$ I : ';

. ;~\ ~,~;~,''''''''1,~~,,';~\t.,'II~~t~#)tJ,'~m~{I~\~, ' , ' ~~~~II\"\~\I~I~~t~I\4\rVA1\~~~~Y~'~~1Mi."~~fI~r 1 1·1 r i"'," l' =1 1: ' W\!\~_, !\ '/ i~ II r1'1 r .! 1 I I 1I .: (,~/~,.rI/''(rtl~r. /rrr(rYTlrV'"~~V~~~~~rV'~rrrV'~~ ~~ ~ r ~ ~~r r Vrrr~~~ r~~r~~V-VY~lrVvu-t-v--y-

1""",,,:,. , r

0

EEG C.'.

J I.JjL.!..Il.oIllJ.AtA.... """'""l~"".~I~}~M~wJJNI..MlJ..lAhu.t'LI II I I~~I ,.~l II I 'I ·!~i