Sleep suppression of ventricular arrhythmias: a predictor of beta-blocker efficacy

European Heart Journal (1996) 17, 917-925

Sleep suppression of ventricular arrhythmias: a predictor of beta-blocker efficacy M. V. Pitzalis, F. Mastropasqua*, F. Massari*, P. Totaro*, D. Scrutinio* and P. Rizzon Institute of Cardiology, University of Bari, Bari, Italy and *Division of Cardiology, "Salvatore Maugeri" Foundation, IRCCS, Cassano Murge, Bari, Italy

administration. In group 1, nadolol led to a mean reduction in the number of premature ventricular contractions of 90% (>70% in 21/23 patients). In group 2, the mean reduction was 76% (>70% in three out of six patients). In group 3, there was a mean increase in the number of premature ventricular contractions of 33%. The positive predictive accuracy of sleep suppression in relation to the antiarrhythmic efficacy of nadolol is very high (88%) when sleep suppression is present during two baseline Holter recordings. Sleep suppression is a sensitive parameter for identifying the premature ventricular contractions likely to benefit from beta-blocker administration. (Eur Heart J 1996; 17: 917-925)

Introduction

arrhythmias which arise under particular conditions, such as those which are exercise-induced'91. Although exercise cannot be considered a purely sympathetic stimulus, exercise-related arrhythmias may be at least partially due to modifications occurring in the balance of the autonomic nervous system. However, these are not the only arrhythmias affected by the autonomic nervous system, as any change in its activity may enhance or suppress various arrhythmogenic mechanisms. An example of this dependence on autonomic tone is the circadian variation in the incidence of premature ventricular contractions occurring in some patients. This variation follows a pattern which is somewhat similar to that of circadian variations in heart rate1'01, its highest incidence occurring during waking hours'""131; the increase in the number of arrhythmias during this period may depend either on the increase in sympathetic activity or on the concomitant increase in heart rate. Lown et a/.'141 observed that there was a significant decrease in the number of premature ventricular contractions during sleep in 22/35 (62%) patients, and called this phenomenon 'sleep suppression'. Its

Over the past few years, the role of the autonomic nervous system in the genesis of ventricular tachyarrhythmias has been elucidated by various experimental studies'1'21. It is generally accepted that autonomic nervous tone and, in particular, a high level of sympathetic activity also play an important role in the genesis of premature ventricular contractions'31. The identification of adrenergic-dependent premature ventricular contractions would be very useful in clinical practice as this type of arrhythmia is likely to be suppressed by beta-blockers. Beta-blockers are generally not considered to be highly efficacious in suppressing premature ventricular contractions'4""81; however, they are effective against Revision submitted 19 September 1995, and accepted 25 October 1995. Correspondence. Maria Vittoria Pitzalis, Institute of CardiologyUniversity of Bari, Policlinico, Piazza Giulio Cesare II, 70124 Ban, Italy. 0195-668X/96/060917 + 09 $18.00/0

Key Words: Premature ventricular contractions, heart rate, autonomic nervous system, beta-blockers, nadolol, sleep, sleep suppression, circadian variability.

1996 The European Society of Cardiology

Downloaded from eurheartj.oxfordjournals.org by guest on October 17, 2011

It has been reported that the frequency of premature ventricular contractions in some patients tend to decrease during the hours of sleep when modifications in autonomic tone and bradycardia occur. The aim of this study was to evaluate whether the phenomenon of sleep suppression may be a sensitive and specific parameter for predicting the antiarrhythmic effect of beta-blockers on premature ventricular contractions. The presence of sleep suppression was evaluated in 45 patients (mean age 50 ± 17 years) with frequent premature ventricular contractions at two baseline Holter recordings. Sleep suppression was defined as >50% reduction in the number of nighttime as opposed to daytime premature ventricular contractions. Three groups of patients were identified: those with sleep suppression at both Holter recordings (group 1); those with sleep suppression at only one Holter recording (group 2); and those without sleep suppression at either Holter recording (group 3). A third Holter was performed 5 days after nadolol

918 M. V. Pitzalis et al.

occurrence may be secondary to a reduction in heart rate and/or modifications in autonomic nervous system activity during sleep1'51. Beta-blockers may be considered drugs of first choice in the treatment of arrhythmias characterized by this phenomenon because of their ability to slow heart rate and reduce sympathetic activity. The purpose of this study was to investigate whether the presence of sleep suppression is a good index for predicting the efficacy of beta-blockers in suppressing premature ventricular contractions; some preliminary results have already been presented elsewhere'161.

Patients and methods Patient characteristics

Eur Heart J, Vol. 17, June 1996

After discontinuing, for at least five half-lives, the administration of antiarrhythmic drugs and those drugs known to modify heart rate, all of the patients underwent two 24-h Holter recordings separated by an interval of 3 ± 1 days (range 2-7 days). The first baseline recording was performed 4 ± 2 days after hospitalization. The patients were instructed to go to bed at their usual time and to report whether they slept or not. Sleep suppression was defined as a reduction in the number of premature ventricular contractions during the night (0000 am to 0600 am) in comparison with those recorded during the day (1000 am to 0400 pm). On the basis of the entity of the nighttime reduction, three degrees of sleep suppression were denned: a reduction of at least 50%, 70%, and 90%. Three groups of patients were identified: those in whom sleep suppression was present at both baseline Holter recordings; those in whom it was present in only one of the two Holter recordings; and those in whom it was present during neither recording. After a mean period of 3 ± 1 days from the second Holter recording, all of the patients received nadolol 80 mg . day ~ ' divided into two administrations. On the fifth day of therapy, they underwent a third Holter recording in order to evaluate the effect of nadolol on the suppression of premature ventricular contractions. Nadolol was chosen because of its potent heart rate slowing effect and long half-life'17'181. At each Holter recording, the following parameters were evaluated: the length of the recording; the mean RR cycle length during the entire monitoring period as well as during the daytime and nighttime hours; the entity of the lengthening of the RR cycle during the nighttime period; and the daytime and nighttime incidence of premature ventricular contractions. The antiarrhythmic effect of nadolol was evaluated in terms of its average effect on the premature ventricular contractions of the three groups and on each patient in each group. The effect of nadolol on each group was calculated on the basis of the reduction in the mean number of premature ventricular contractions in comparison with the mean number of the two baseline recordings taken together. The reduction was evaluated over the entire Holter monitoring period, as well as over the daytime and nighttime periods. The efficacy of nadolol in each of the patients in the three groups was defined as a reduction in the number of premature ventricular contractions >90% or >70%. The effect of nadolol on couplets and runs was also evaluated. In order to reduce the effect of the spontaneous variability of these arrhythmias, this evaluation was made only if there were >70% couplets or > 10 runs in both baseline recordings. In order to evaluate the role of sleep-induced bradycardia in determining the occurrence of sleep suppression, we also analysed all of the recordings showing the phenomenon separately from those not showing it.

Downloaded from eurheartj.oxfordjournals.org by guest on October 17, 2011

The study population was selected from a series of 212 consecutive patients admitted to our Institution because of frequent, symptomatic premature ventricular contractions. The criteria for inclusion were the presence of sinus rhythm and frequent (>30 h~ '), stable premature ventricular contractions. When present, ventricular couplets and runs were also considered but only if they had the same morphology as the simple premature ventricular contractions. The length of the recordings had to be more than 19 h. The stability of the premature ventricular contractions was assessed by means of two Holter recordings performed at an interval of 3 ± 1 days, and defined as a variation, of less than 35%, in the number of premature ventricular contractions per hour obtained during each recording in comparison with the mean number of premature ventricular contractions per hour of the two recordings considered together. The exclusion criteria were: sick sinus syndrome; atrioventricular block of any degree; bundle branch block; current beta-blocker therapy or any contraindications to beta-blockers. Patients were also excluded if they were affected by sleep disturbances (such as sleep apnoea), needed to take sleeping pills, or did not sleep during the nights of the evaluation periods. None of the patients must have slept, or have been used to sleeping during the daytime period. Forty-five patients (35 male, 10 female) with a mean age of 50 ± 17 years (range 16-76 years) fulfilled all of the inclusion and exclusion criteria and, after having given their informed consent, formed our patient population. Thirty-six patients were affected by cardiac diseases of different aetiology (eight patients had had a myocardial infarction at least 6 months before the evaluation) and nine were considered as not having any cardiac disease by means of a full medical history, clinical evaluation, 12-lead electrocardiogram examination, chest X-ray, blood pressure measurements, mono and bidimensional echocardiography or exercise testing.

Methods

Beta-blockers and sleep suppression 919

In these two groups of recordings, the day/night increase in the RR cycle length was evaluated. In order to obtain data concerning the incidence of sleep suppression in the population of patients with frequent premature ventricular contractions, the presence of 50% sleep suppression was evaluated in all of the 176 consecutive patients who showed an incidence of premature ventricular contractions of > 3 0 h ~ ' during two 24-h electrocardiogram recordings separated by a period of 3-8 ± 2 days.

Instrumental and data analysis Holter monitoring was performed using a 2-channel Del Mar Avionics model 445A tape recorder and was always begun in the morning hours (between 8000 am and 0930 am). The data obtained from the recordings were analysed by means of a Del Mar Avionics Innovator Electrocardioscanner (model 750).

Data are reported as mean values ± standard deviation. Multiple between-group comparisons of the mean value of each parameter were performed using one-way ANOVA. When the F value was statistically significant at ANOVA, multiple comparisons were made using the Student-Newman-Keuls test. These tests were used for each parameter both under baseline conditions and after nadolol administration. The effects of sleep and nadolol were evaluated using the Student t-test for paired-data. The statistical analyses were performed using an SPSS Base System statistical package for Windows version 5.0 (SPSS Inc. Chicago, IL, 1992). A value of P70% in 21 of the 23 patients in group 1 (91%) and of >90% in 17 (74%). In Group 2, there was a reduction of >70% in three out of six patients (50%) and of >90% in two out of six patients (33%). In Group 3, there was a reduction of >70% in two out of 16 patients (12-5%) and of >90% in one out of 16 patients (6%) (Table 4). Eur Heart J, Vol. 17, June 1996

Downloaded from eurheartj.oxfordjournals.org by guest on October 17, 2011

Statistical analysis

Table 1 Clinical characteristics of the three groups of patients

920 M. V. Pitzalis et al.

Table 2 Baseline characteristics of the three groups of patients Group 1

n

Group 2

23

HI Holter recording H duration 1323 ±96 835 ±99 RR 24 h 767 ± 94 RR day RR night 980 ± 116 213 ±68 D/N increase PVCs/h 322 ±313

Group 3

6

H2 1344±66 837 ±91 761 ±89 973± 125 212 ±52 312 ± 317

HI 1346 ±56 898 ± 88 852 ±71 1005 ±84 153 ±46 777 ± 567

16 HI 1356 ±47 804 ± 1 1 1 750± 115 966 ± 137 216±68 401 ±267

H2 1365 ±26 895 ± 47 827 ±61 1036 ±85 209 ±65 626 ± 493

H2 1356±40 798 ± 110 738 ± 112 940 ± 134 202 ± 69 402 ± 270

H duration = mean duration of the 24-h electrocardiogram monitoring period: RR 24h = mean value of the RR interval evaluated over the 24 h period, RR day = mean value of the RR interval evaluated over the 6 h daytime period; RR night = mean value of the RR interval evaluated over the 6 h nighttime period; DAN increase = mean increase in the value of the nighttime in comparison with the daytime RR interval; PVCs/h= mean value of the incidence of premature ventricular contractions per hour evaluated during the 24 h period.

Table 3

The effect of nadolol in the three groups of patients Nadolol

P

836 ± 92 763 9 ± 86-8 976-5 ± 118-3 212 5 ±52-3 317-3 ± 311 2 503-1 ±479-4 75-4 ± 119-7

1068 8 ± 114-5 1041-6 ± 113-5 1160-9 ± 136-8 119 3 ±76 8 52-6 ±110-8 38 4 ± 100-4 61-6 ± 133-5

70% and a >90% reduction in the number of premature ventricular contractions). The test was considered positive only if sleep suppression was present during both baseline Holter recordings, and negative when it was found in only one Holter recording or was not present at all. As shown in Fig. 1, the sensitivity and specificity of 50% sleep suppression in predicting a >70% reduction in the number of premature ventricular contractions are very high (80-8 and 89-5%, respectively), and positive predictive accuracy is 84-4%.

Pu 40 -

20 n

1

1

1

50

70 Sleep suppression (%)

90

Figure 2 Sensitivity ( - - O - - ) and specificity (—O—) of different degrees of sleep suppression (50, 70, 90%) in predicting the antiarrhythmic effect of nadolol evaluated as >90% reduction in the number of premature ventricular contractions. See text for further details.

When the 70% sleep suppression is considered, there is a considerable drop in sensitivity (61-5%); the specificity of the test remains the same (89-5%) and positive predictive accuracy is 73-3%. When 90% sleep suppression is considered, sensitivity becomes too low (308%) but the specificity of the test is very high (100%). Positive predictive accuracy is 60%. When the degree of nadolol antiarrhythmic efficacy was increased to 90%, no evident differences were found (Fig. 2). In particular, when 50% sleep suppression is considered, sensitivity is 85%, specificity is 76% and positive predictive accuracy is 80%. At a level of 70% sleep suppression, sensitivity is 60%, specificity is 76% and positive predictive accuracy is 68-9%. At 90% sleep suppression, sensitivity is 30%, specificity is 92% and positive predictive accuracy is 64-4%. It can therefore be said that, in patients with sleep suppression, the administration of nadolol always has an efficacious antiarrhythmic effect. Eur Heart J, Vol. 17, June 1996

Downloaded from eurheartj.oxfordjournals.org by guest on October 17, 2011

present, the reduction was >70% in four of 19 patients (21%), and only two of 19 patients (10-5%) showed a reduction of >90%.

922 M. V. Pitzalis et al.

Table 5 Effect of nadolol on couplets in the three groups ered together and compared with those not showing the phenomenon. The mean day/night increase in the RR of patients cycle length obtained from all of the recordings showing Baseline Nadolol sleep suppression (213 ± 60 ms) was not statistically different from that found in the other recordings (209 ± 68 ms). However, when this kind of analysis was 1-33 ± 1-7 Group 1 694-7 ±612-6