Changes in polysomnographic findings following transcatheter aortic valve implantation in a patient with aortic stenosis

Abstract

Patients with aortic stenosis (AS) are likely to have sleep-disordered breathing (SDB) and improvements in AS by transcatheter or surgical aortic valve replacement alter the type and severity of SDB. However, limited data are available whether polysomnographic findings changed following transcatheter aortic valve implantation (TAVI). In this report, we describe the case of a patient with severe AS and SDB whose polysomnographic findings showed that after TAVI, sleep disturbances occurred in association with worsened periodic leg movements despite improvement in the SDB.

<Learning objectives: Patients with aortic stenosis often have sleep-disordered breathing (SDB) which can be improved following transcatheter aortic valve implantation (TAVI). However, no studies assessed SDB by polysomnography before and after TAVI. Following TAVI, our case revealed sleep disturbances associated with worsened periodic leg movements (PLM) despite improvement of SDB. Therefore, polysomnography may be considered following TAVI to detect residual SDB and sleep disturbance in association with PLM.>

Introduction

Aortic stenosis (AS) is a valvular heart disease (VHD) that commonly occurs in adults, and the prevalence of AS and severe AS among elderly is 12.4% and 3.4%, respectively [1]. Patients with VHDs are likely to present with sleep-disordered breathing (SDB), possibly caused by an underlying heart failure (HF) [2]. In general, SDB can manifest as two types: central sleep apnea (CSA) and obstructive sleep apnea (OSA). A recent study reported that approximately 70% of patients with severe AS exhibited SDB with a similar proportion of predominant OSA and CSA [3]. In patients with AS, low cardiac output and increased left ventricular (LV) filling pressure may play key roles for the coexistence of SDB [4], especially CSA [5]. Patients with AS who are candidates for valve replacement can present with low cardiac output and high LV filling pressure [6]. Following aortic valve replacement, cardiac output increases whereas LV filling pressure decreases, leading to alterations in the type and severity of SDB.

Recently, transcatheter aortic valve implantation (TAVI) has become an important alternative for aortic valve replacement in patients with AS especially in Japan that is a recognized aged society. In this report, we describe a case of a patient with severe AS and SDB, whose polysomnographic findings changed following TAVI.

Case report

A 75-year-old man with severe AS (body mass index, 23.4 kg/m²) presented with shortness of breath even during low-intensity activity, indicating New York Heart Association (NYHA) functional class III. The patient had been taking immunosuppressive drugs for the past 25 years for treating pure red blood cell dysplasia. Electrocardiography revealed sinus rhythm without any bundle branch block. Transthoracic echocardiography showed a normal LV ejection fraction (70%) with severe AS (aortic valve area, 0.6 cm²; mean pressure gradient, 42 mmHg). Computed tomography (CT) revealed severe calcification of the right and non-coronary cusps. The diameter of both the ilio-femoral arteries was >7 mm; coronary angiogram showed no significant stenosis. The patient had a clinical frailty scale of 4, an STS score of 4.5%, and a logistic Euroscore of 4.3%. Our heart team recommended TAVI.

Since the patient presented with severe snoring and daytime drowsiness, polysomnography was performed overnight prior to undergoing TAVI, which revealed an apnea–hypopnea index (AHI) of 28.2 events per hour of sleep with a mean and minimum oxyhemoglobin saturation (SO₂) of 95% and 87%, respectively. Since the central event accounted for 54.8%, the patient presented predominant CSA with prolonged lung-to-finger circulation time (LFCT) (Table 1 and Fig. 1A).

Table 1.

Changes in clinical parameters before and 7 months after TAVI.

	Before TAVI	7 months after TAVI
Body weight, kg	61.5	61.3
Blood pressure, mmHg	110/56	120/64
Heart rhythm, and rate, /min	Sinus, 70	Sinus, 70
Plasma NT pro-BNP level, pg/mL	1801	397.6
Echocardiographic parameters
LVEF, %	70	72
LVEDV, mL	118	112
LVESV, mL	40	31
LAVI, mL/m2	53.1	38.2
E/e'	30.7	11.5
AVA, cm2	0.6	1.5
Peak velocity, m/s	4.3	2.8
Mean pressure gradient, mmHg	42	16
AR grade	Mild	Trivial
MR grade	Mild	Trivial
RVSP, mmHg	41	30
Sleep study parameters
TST, min	334.5	363
N1 sleep	23.5	37.1
N2 sleep	60.8	53.7
Slow-wave sleep, % of TST	0.4	0.6
REM sleep, % of TST	15.2	8.7
Supine sleep, % of TST	19.5	17.2
Total AHI, events/h	28.2	11.7
Obstructive AHI, events/h	8.6	10.7
Central AHI, events/h	14.1	1
Minimum SO2, %	87	88
Arousal index, events/h	29.8	43.5
PLM index, events/h	18.1	86
PLM arousal index, events/h	2.9	19.3
LFCT, s	25.75	16.95

AHI, apnea–hypopnea index; AR, aortic regurgitation; AVA, aortic valve area; NT pro-BNP, N-terminal pro-B-type natriuretic peptide; LAVI, left atrial volume index; LVEF, left ventricular ejection fraction; LVEDV, left ventricular end-systolic volume; LVESV, left ventricular end-diastolic volume; LFCT, lung-to-finger circulation time; MR, mitral regurgitation; PLM, periodic leg movements; REM, rapid eye movement; RVSP, right ventricular systolic pressure; SO₂, oxyhemoglobin saturation; TAVI, transcatheter aortic valve implantation; TST, total sleep time.

Fig. 1.

(A) Diagnostic polysomnography before TAVI. Eight occurrences of central apnea were detected, with subsequent waxing and waning hyperventilation. During central apneas, the movements of the thorax and abdomen were absent. Duration from the onset of the first breath terminating the apnea to the nadir of the subsequent dip in the SO₂ measured at the finger indicates the LFCT, which is considerably long (an average of 10 consecutive apnea–hyperpnea cycles during the first episode of stage 2 sleep, 26 s). (B) Follow-up polysomnography performed 7 months after TAVI. Five occurrences of obstructive apnea were detected. During obstructive apneas, the out-of-phase movements of the thorax and abdomen were observed. LFCT was shorter than that before TAVI (an average of 10 consecutive apnea–hyperpnea cycles during the first episode of stage 2 sleep, 17 s).

TAVI, transcatheter aortic valve implantation; EEG, electroencephalogram; EOG, electro-oculogram; EMG, submental electromyogram; ECG, electrocardiography; SO₂, oxyhemoglobin saturation; LFCT, lung-to-finger circulation time.

TAVI was performed by puncturing the left femoral artery under general anesthesia. CT revealed an aortic annulus (area, 437 mm²); transesophageal echocardiography also showed an aortic annulus with a dimension of 22 mm. Therefore, he was implanted a 23-mm Edwards SAPIEN 3 valve (Edwards Lifesciences, Irvine, CA, USA). Pre-dilatation was performed using a 20-mm balloon, after which a 23-mm Edwards SAPIEN 3 valve with 1-ml overfilling was successfully implanted. The subsequent course was good, and the patient was discharged on the 8th postoperative day.

Seven months after performing TAVI, the patient’s shortness of breath improved up to NYHA class I, with improvements in the N-terminal pro B-type natriuretic peptide (NT-proBNP) level and echo parameters for LV filling pressure. However, as the patient continued to snore, a follow-up polysomnography was performed, which revealed a marked reduction in AHI (11.7 events per hour of sleep) with an improvement in oxygenation (mean and minimum SO₂ of 97% and 88%, respectively) (Table 1). Despite the alleviation of the central respiratory events, obstructive respiratory events with shortened LFCT became obvious and were accompanied by a worsening of periodic leg movements (PLM) and related arousal (Table 1 and Fig. 1B).

Discussion

The present case revealed a significant alleviation of CSA but not of OSA following TAVI. This result is consistent with that reported by previous studies, in which TAVI was associated with a significant alleviation of CSA but not of OSA [7]. The alleviation of CSA can be associated with a reduction of the LV filling pressure and an increase in the cardiac output following successful TAVI. In fact, in the present case, TAVI was associated with a decrease in the right ventricular systolic pressure, left atrial volume index, and NT-proBNP level, suggesting decreased LV filling pressure and LFCT, which in turn suggested a shortened cardiac output. LFCT was measured from the onset of the first breath after apnea to the lowest point of the subsequent SO₂ dip. A delay in LFCT may reflect a prolonged lung-to-carotid body circulation time. A previous study reported CSA improvement and shortened LFCT in patients with rheumatic VHDs after cardiac valve replacement [8]. In the present case, no changes were observed in the heart rate, LV ejection fraction, LV end-diastolic volume, and LV end-systolic volume following TAVI. However, since the grade of mitral regurgitation decreased, which decreased the regurgitation flow, the forward output increased, thus shortening LFCT.

The present and previous studies assessing SDB before and after TAVI differ with respect to the type of sleep study. In the present case, SDB was assessed using polysomnography, whereas in previous studies, it was assessed using polygraphy, which is not recommended for diagnosing CSA and cannot determine the total sleep time, sleep stages, and arousal and PLM indices. One important finding of the present case is the alleviation of CSA following TAVI, which was first confirmed even on polysomnography that is the standard for diagnosing CSA.

Other important findings include the absence of an increase in slow-wave sleep and in the arousal index by 1.5 times, despite the alleviation of CSA. Such deterioration in sleep quality following TAVI is associated with marked increases in PLM and PLM arousal indices. These are novel findings because no data have been reported regarding changes in PLM before and after TAVI and regarding the increase in PLM after improvement of HF with an alleviation of CSA. In addition, no data have suggested that after TAVI, patients with SDB still have severe sleep fragmentation and insufficient amount of slow wave sleep in association with PLM even after the alleviation of CSA with a reduction in total AHI.

In patients with HF, PLM is common and is associated with a poor, long-term outcome [9]. Increased sympathetic nervous system activity in patients with HF may facilitate PLM development. In addition, reduced peripheral blood flow due to diminished cardiac output is associated with PLM. However, in the present case, PLM markedly worsened following the improvement in cardiac output and HF condition by TAVI. In patients with severe OSA, PLM may increase during the alleviation of OSA using continuous positive airway pressure (CPAP) due to the unmasking of underlying PLM [10]. Although in the present case, CSA was alleviated after TAVI, OSA was not severe before TAVI and remained unchanged following TAVI. CSA may have similar masking effects on PLM. However, no specific data regarding an increase in PLM following an alleviation in CSA are available; therefore, the mechanisms for the worsening of PLM after TAVI remain unclear.

A previous study suggested the reduction of PLM index on CPAP in patients with mild OSA [10]. In the present case, as AHI did not fulfill the criteria for CPAP initiation under the Japanese health insurance and the patient did not provide consent to use an oral appliance, the effects of residual OSA treatment on PLM index were not assessed. In addition, the patient refused to conduct any interventions to assess PLM.

Nevertheless, in our case, if the patient underwent polygraphy rather than polysomnography following TAVI, neither objective sleep disturbance (severe sleep fragmentation and insufficient amount of slow wave sleep) nor PLM was detected. Therefore, clinical implications of this case report include that polysomnography, not polygraphy, may be considered at least once after TAVI to determine whether patients have SDB and objective sleep disturbance in association with PLM. Although further studies are needed to determine whether sleep disturbance in association with worsened PLM following TAVI is common, and whether interventions for such PLM-related sleep disturbances should be considered, this case may contribute to generating a hypothesis and act as a rationale to conduct such studies.

Conflict of interest

Nothing to disclose.