Screening and managing obstructive sleep apnoea in nocturnal heart block patients: an observational study

Obstructive sleep apnoea (OSA) is an underestimated clinical problem and is widely accepted to be associated with significantly raised cardiovascular morbidity and mortality [1]. Nocturnal heart block often occurs in patients with OSA and is reportedly a consequence of the autonomic effects of recurrent apnoea with subsequent oxygen desaturation, cyclic fluctuations in sympathovagal balance and cardiac hemodynamic changes [2]. As a result, OSA has been identified as a risk factor for almost all types of arrhythmias, ranging from asymptomatic sinus bradycardia to sudden death [3, 4]. An initial study [5] reported that almost 18 % of OSA patients developed bradycardic arrhythmias, while Miller et al. [6] observed that the prevalence of heart block in patients with sleep apnoea was 9–13 %. Screening significant rhythm disturbances assumes particular clinical significance in view of the adverse outcomes [7].

Nearly 80 % of men and 93 % of women with moderate-to-severe sleep apnoea remain undiagnosed within the community [8]. It is more likely to be unrecognized in heart block patients who are ignorant of the symptoms of sleep disorder. The incidence of OSA in patients with nocturnal atrioventricular block or sinus arrest is less defined. Overnight polysomnography (PSG) is the gold standard for OSA diagnosis, but is not generally available and is expensive to perform, requiring highly trained personnel, sophisticated equipment, and an entire night to record data. Berlin Questionnaire (BQ), a reliable method of determining the risk factors of OSA based on symptoms of snoring and daytime sleepiness along with the presence of hypertension or obesity, has good sensitivity ranging from 73 % to 86 %, with a variable specificity of 44 % in sleep laboratory [9] and of 77 % in primary care sites [10]. Considering that not all patients who are observed to have heart block need to undergo PSG, this screening strategy to identify heart block patients at risk of developing OSA should be revalued for validity and feasibility.

It has long been speculated that the vast majority of OSA-related arrhythmias can be reversed during effective continuous positive airway pressure (CPAP) therapy [11, 12]. However, clinical studies on nocturnal heart block patients are limited and conflicting. One randomised controlled trial by Sonya Craig et al. [13] confirmed that there was no significant reduction in frequencies of dysrhythmias comparing therapeutic CPAP groups with the subtherapeutic control, in contrast to the results of another randomized controlled study [14] that confirmed the alleviation of ventricular ectopy by CPAP in patients with heart failure. Moreover, one case report [15] demonstrated that despite atrial pacing, occasional atrioventricular (AV) block still persisted during patient’s follow-up. According to the study of Fietze [16], the prevalence of sleep disorder breathing (SDB) in cardiac pacemaker patients was similar to those without pacemakers. Consequently, screening for sleep apnoea would be necessary before and after a pacemaker. Therefore, we conducted this study to find a simpler method of identifying sleep apnoea in these patients and to examine the effect of CPAP in patients with OSA and heart block.

The present study observed that respiratory sleep disorders are common among nocturnal heart block patients, especially in patients with high risk factors associated with OSA, such as obesity, habitual snoring or hypertension. To the best of our knowledge, this is the first report of Berlin Questionnaire being administered to nocturnal heart block patients. The principal findings include that BQ is an effective instrument for the recognition of sleep apnoea in heart block patients and this kind of heart block, which occurs frequently during sleep rather than wakefulness in OSA patients, can be largely abolished following the institution of CPAP therapy. OSA appears to be a trigger for heart block during sleep, which may be a potential mechanism for nocturnal bradyarrhythmia.

Previous reports have examined the association between OSA and cardiac arrhythmias in selected group with different severities of OSA or in different populations, such as frequent ventricular premature complexes, conduction delay arrhythmias [3], supraventricular arrhythmias and atrial fibrillation [19]. They also suggested that the incidence of cardiac conduction disorders was more frequent in OSA patients than in the general population [5, 6]. Moreover, cardiac arrhythmias have been documented in up to 48 % of patients suffering from OSA [20]. However, this literature is sparse and conflicting on heart block. A population based study [21] did not observe differences in the occurrence of nocturnal cardiac break among the OSA and matched control groups. Additionally, Harbison et al. [22] only observed 7 cases of nocturnal cardiac break in 45 OSA patients. Compared with other arrhythmias, the Sleep Heart Health Study reported sinus pause was not common in patients with OSA [20]. Little information is available regarding arrhythmia in Asian patients with OSA up to now. In our study all the patients included were mainland Chinese. We performed PSG on them and calculated the occurrence of sleep apnoea as high as 87.5 %, complementing the early epidemiological studies in which a high rate of cardiac arrhythmia patients identified from the general population developed OSA in Asian population. Data from the epidemiological studies in China have estimated that the prevalence of OSA in general population is approximately 3.62–4.81 % [23, 24]. In particular, these characteristics of obesity, disruptive snoring and hypersomnolence were more related to the occurrence of heart block than syncope and palpitations suggested the target population that we emphasized. As a result, the defined population was largely suspected of sleep disorder, which explaining the higher occurrence of OSA. It is notable that those patients conscious of their snoring and sleepiness were more willing to be enrolled in our study. The evaluation and selection from both physicians and patients probably bias the occurrence of OSA in all the arrhythmias but reflect the results in this small kind of arrhythmias. Meanwhile, the high occurrence of OSA, confirmed by PSG, indicates why our study population was most responsive to CPAP.

Regarding the high prevalence of OSA and its potential prognostic implications in nocturnal heart block patients, a combination of clinical symptoms and categorized questionnaires helped to predict its occurrence. The BQ, used as a screening tool, requires a high sensitivity with an acceptable specificity. After systematically performing and examining its validity, the BQ achieved the highest sensitivity at a cut-off point ≥ 30/h for predicting AHI, which is similar to the latest finding of Ulalsi et al. [9]. The performance of BQ in cardiovascular patients has been highly efficient and has a sensitivity of 0.86 and specificity of 0.89 compared with sleep studies [25]. Such results with BQ were also achieved in patients with atrial fibrillation [26] and resistant hypertension [27]. However, the screening instrument in our study was limited by a lack of specificity, as an elevated value may result from snoring. This issue is currently of interest because prior studies showed the occurrence of an atrioventricular block was more likely induced by observed apnoea and nocturnal hypoxemia, independent of snoring. Therefore, without sleep apnoea or hypoxemia, sole snoring hardly affects arrhythmia [28, 29]. This could be a possible reason why the BQ is not a reliable discriminating tool to rule out the non-OSA group. Although BQ is a poor predictor of OSA with a specificity of 0.44, which is valid also for other references [30], that does not mean the Berlin Questionnaire is not recommended as an appropriate instrument to identify patients with sleep apnoea. Given the high cost of sleep monitoring and the lack of dedicated sleep centres, this screening should be considered in the evaluation of arrhythmias with possible risk factors of OSA for its accuracy.

The number of episodes of bradycardia and sinus pauses increased in parallel to the severity of OSA as reported by Roche et al. [31]. Nevertheless, we did not find severity of sleep apnoea assessed by AHI significantly related to bradyarrhythmia, contrary to those who suggested a graded risk for arrhythmias depending on OSA severity [6]. In light of these contradictory results, we must take the large inter-individual variability in the episodes of heart block into account. As an observational study, any confounding risks by unmeasured variables cannot be excluded, even though we had excluded patients with most pre-existing cardiac disease to avoid bias. Our observational study based on nocturnal heart block patients, and was constructed to be neither a fully representative population-based epidemiological study, nor a general arrhythmia-based epidemiological study. Therefore, the sympathovagal predominance, atrial dilatation, impaired baroreflex [32] and hemodynamic fluctuation may predispose subjects to develop arrhythmia. Of note, the influence of O₂ desaturation on the occurrence of heart block was obvious that supported the correlation between them. Furthermore, through the sub-group analysis at the cut-point of AHI of 30/h, RR interval time in patients with AHI ≥ 30/h was significantly higher than in those with AHI < 30/h as demonstrated in our results, which is consistent with the main conclusions of the previous study [33].

After analysing the features of the subgroup (AHI < 10), we observed that most of the arrhythmia patients with a lack of symptoms were stratified as low risk OSA group, with a minimal or mild residual AHI. It has been hypothesized that a REM-sleep-related AV block occurs independently of episodes of apnoea or hypopnoea. Similar features have been documented in several cases report [34, 35]. Most of them remained irrelevant arrhythmia changes despite CPAP therapy, consistently, the improvement of O₂ saturation was also not distinct. We concluded that CPAP was not completely preventative in this subgroup, as opposed to those with more severe OSA.

Koehler et al. [36] reported that the magnitude of bradycardia appeared to be related to the severity of hypoxaemia. Several previous studies convincingly demonstrated that the administration of oxygen attenuated the bradycardia [37]. However, a case reported by Ken-ichi Maeno [38] indicated that OSA can provoke heart block without oxygen desaturation. Our findings show the frequent occurrence of nocturnal heart block in prominent oxygen desaturation and provide further evidence that hypoxia with cessation of airflow is a predisposing condition for these arrhythmias. The reversibility of nocturnal arrhythmias by CPAP has been demonstrated in most patients with OSA [39]. In general, the present study shows a favourable outcome with CPAP by abolishing 80 % of nocturnal heart block in 51 consecutive patients, similar to the Becker’s reversal outcomes [40], and an elimination in 72 % of patients of Wolfram’s study [12]. It is well acknowledged that atrioventricular block is observed during both REM and non-REM sleep, but it is hard to differentiate from a REM-sleep–related AV block, because it is often caused by an increased dominated surge of the parasympathetic nervous system. One case report observed that CPAP treatment did not abolish REM-sleep–related AV blocks in patients with mild or no OSA [41], which further supports our subgroup analysis results. Why CPAP could not result in a complete disappearance of nocturnal arrhythmias in several severe patients, irrespective of an obvious cardiac pathology or lack of compliance, remains unknown. After plotting the association of treatment pressure with outcomes, we observed a dose-relative effect on the therapy results. Our study confirms the previous findings [42] in which CPAP application revealed an absence of a sinus and AV block in mild sleep apnoea, which was better explained by pressure effects. Different from an optimal pressure of 15 cm H₂O to make the observed arrhythmia disappeared as recorded by Ki-Hwan Ji [15], our curve show a pressure of 12 H₂O cm generates the greatest benefit except for the improvement in O₂ saturation.

Those symptomatic heart block patients, with lethal risk of long duration of RR pauses exceeding 3 seconds, indeed present a classic indication for the pacemaker implantation according to the current national and international guidelines. In our study, patients whose blocks occurred during nocturnal sleep or daytime nap with risks of sleep apnoea seldom developed long lasting duration of second-degree AV block or sinus block exceeding 5 seconds. A detailed analysis of hourly summary of electrocardiogram result (shown as Additional file 1) and PSG together with Holter recording in a patient (shown as Additional file 2) showed 4957 episodes of heart block with RR interval exceeding 2 seconds during sleep. It has to be noted that these blocks of about 2 seconds duration occurred toward the end of the apnoea episodes, which were abolished largely by CPAP later. This directly demonstrates a strong association between OSA and heart block. A meta-analysis of ten randomized controlled trials has shown that atrial overdrive pacing (AOP) is inferior to CPAP for treating OSA-related heart block and there is a possibility that CPAP therapy may reduce the need for cardiac pacemakers, if patients are screened and treated early. However, the role of AOP in OSA by another meta-analysis was not confirmed [43]. It was reported that SDB was identified in 59 % of those bradyarrhythmia patients implanted with pacemakers [44]. Furthermore, Simantirakis EN et al. [45] indicated AOP was inferior to CPAP on reducing OSA severity. Based on this evidence, those patients with heart block occurring exclusively during obstructive sleep apnoea without obvious pathological abnormality are the best target population for CPAP therapy. Regarding the controversial indication of pacemaker implantation, we emphasize the short waiting time for diagnosis and treatment in our study, since all patients had baseline and follow-up studies just completed within 3 nights. Most often, in clinical practice, after comprehensive evaluation of the medical history, electrophysiological abnormalities and other cardiac symptoms, pacemaker implantation is considered and performed according to the patient’s wish, or based on the recommendation of his or her attending physician.

There should be a low threshold method such as BQ for the identification of OSA that is mandatory to prevent being misdiagnosed by cardiologist or neglected by respiratory physician. Most importantly, this non-pharmacological approach spares patients with concomitant pacemaker therapy. Since patients with long-term pacemaker still exhibit a high prevalence of OSA [44], suggestion of screening OSA is an urgent option that should draw the highest attention of cardiologists.

Several limitations of the present study should be noted. Firstly, lack of electrophysiologic evaluation for underlying disease may obscure the study population. Secondly, it is difficult to match important determinants for arrhythmia patients in a case-control design, owing to inadequate numbers of patient.

Considering the high prevalence of OSA in heart block patients, a systematic PSG should be ideally offered to patients suffering from nocturnal heart block. Our study complements the previous results that patients with nocturnal cardiac arrhythmia have an increased vulnerability to OSA, and provides a method for systematic OSA detection via BQ screening. For better management in the sleep laboratory, it is feasible to apply CPAP that abolishes the majority of apnoea and apnoea-associated heart block.