Iatrogenic (drug-induced) or potentially reversible bradyarrhythmia is a common but inadequately characterized clinical situation.1 The current European Society of Cardiology guidelines on cardiac pacing recommend discontinuation of iatrogenic medication or correction of the reversible cause before permanent pacemaker implantation.2,3

These patients often have underlying conduction system disease and there are few data on recurrence in this population or the need for pacemaker implantation in the future.4–6

According to published studies, around 25% of bradyarrhythmias are not caused by, but revealed by medication, and beta-blockers are the most common ‘offending’ drug involved.4

Metabolic abnormalities, particularly electrolyte imbalances, are generally benign and in most cases the resulting bradyarrhythmia can be resolved. They are therefore frequently excluded from studies of iatrogenic bradyarrhythmias.5

The present study aimed to analyze the need for permanent pacemaker implantation during hospital stay or after discharge in patients with iatrogenic bradycardia or bradycardia associated with other potentially reversible causes.

The baseline characteristics of the overall population and by type of bradyarrhythmia are presented in Tables 1 and 2.

We studied 153 patients, mean age 82±11 years, 47% male. Of these, 63% (n=97) presented CAVB, 16% (n=24) sinus bradycardia, 12% (n=18) second-degree AVB and 9% (n=14) AF with slow ventricular response. The most common form of presentation was syncope, seen in 42% (n=64) of patients.

Potentially reversible etiologies associated with electrocardiographic abnormalities were attributed to drug therapy alone in 85% (n=130), to drug therapy associated with electrolyte imbalances in 12% (n=19), and to electrolyte imbalances alone in 3% (n=4) (Figure 1).

Figure 1.

Diagram of the population with iatrogenic bradyarrhythmia according to etiology and response to correction of potentially reversible causes. PM: pacemaker implantation.

(0.16MB).

Of patients with purely drug-induced bradyarrhythmia, 38% (n=49) were taking beta-blockers, 15% (n=20) amiodarone, 12% (n=16) calcium channel blockers, 5% (n=6) digoxin, 2% (n=2) sotalol and 3% (n=4) propafenone, while 25% (n=33) were on two or three drugs (Table 1).

Permanent pacemaker implantation was needed in 55% (n=84) of the population, during the first hospitalization in 40% (n=61). Resolution of bradyarrhythmia was achieved in 15% (n=23) of the population following drug discontinuation or correction of electrolyte imbalance, but these patients suffered another episode of bradycardia without an apparently reversible cause and underwent pacemaker implantation.

Of the patients requiring a pacemaker, CAVB was the most frequent type of bradyarrhythmia, found in 77% (n=65) of cases, followed by second-degree AVB in 17% (n=14) (Table 2).

Of patients under drug therapy, 62% (n=81) required pacemaker implantation despite discontinuing therapy. Drug therapy in isolation was statistically significantly associated with less likelihood of resolving the bradyarrhythmia compared to other potentially reversible causes (Table 3).

Of the patients who received pacemakers, 46% were medicated with beta-blockers, 24% with amiodarone and/or digoxin and 12% with calcium channel blockers, all without statistical significance (Table 4).

Temporary pacing was required in 28% (n=43) of the population, mainly those with CAVB, of whom 53% (n=23) required permanent pacemaker implantation.

Patients in whom bradyarrhythmia was resolved without need for pacemaker implantation were significantly more likely to present renal failure (24 vs. 5; p=0.001). Most patients who did not need pacemaker implantation had hypertension (Table 1).

Four patients died during hospitalization, three before pacemaker implantation and one after. No deaths were recorded in patients after discharge without pacemaker implantation.

In this population, most patients required pacemaker implantation after iatrogenic drug discontinuation and/or correction of potentially reversible causes of bradyarrhythmias.

The proportion of patients with atrioventricular block who discontinue drug therapy and still need pacemaker implantation is unknown, as is the prognosis of these patients after hospital discharge without a pacemaker.2 There are few data on the situation concerning patients with bradyarrhythmias due to metabolic disorders, since their apparently benign course means they are excluded from most studies on iatrogenic bradyarrhythmias.

Although the number of patients with hyperkalemia in our study was low, electrolyte imbalances were associated with iatrogenic bradyarrhythmias in 75% of cases, in which normalization of heart rhythm was achieved after electrolyte correction, only one patient requiring pacemaker implantation. This is an indication of the benign nature of this condition, in contrast to the 96% (n=81) of patients under drug therapy requiring pacemaker implantation.

In a quarter of patients with drug-related bradycardia, the arrhythmia is not caused, but revealed by medication, of which beta-blockers are the most common ‘offending’ drugs.4

In a study by Zelter et al., 56% of patients whose AVB resolved after drug discontinuation presented recurrence of bradyarrhythmia,6 while in Knudsen et al. the corresponding figure was 25%.2

In this population, 25% of patients discharged without a pacemaker following apparent resolution of bradyarrhythmia presented recurrence of electrocardiographic abnormalities with need for pacemaker implantation during follow-up.

Studies in the literature indicate that drugs at therapeutic doses do not usually cause significant bradycardia in patients without structural heart disease, but when there is latent sinus node or AV node disease, drug therapy can trigger bradyarrhythmias.4–6

Although not reaching statistical significance, bradyarrhythmias were less likely to be resolved by discontinuation of beta-blockers than other antiarrhythmics, as shown in other studies.1,2,5

Advanced age and a history of ischemic heart disease are predictors of arrhythmic risk, but in our population the rate of pacemaker implantation was no higher in these patients, even though the mean age of the study population was over 65 years and 26% had coronary artery disease.1,5

Bradycardia in the context of AF is relatively common. According to some studies, pacemaker implantation is required in 1.4% of patients taking amiodarone and 2.5% of those taking sotalol. In a study by Israel et al., among 78 patients with a DDDR system implanted for symptomatic bradycardia and paroxysmal or persistent atrial fibrillation, the bradycardia was drug-induced in 33%.1,5 In our study, only two patients with AF with slow ventricular response required pacemaker implantation after drug discontinuation or correction of a potentially reversible cause, which suggests that this condition is more amenable to resolution, and hence more benign, than AV node disease. None of the patients with AF with slow ventricular response who were discharged without a pacemaker presented a new episode of arrhythmia requiring pacemaker implantation during follow-up.

Bradyarrhythmia was less likely to be resolved in cases of AV node disease (CAVB and second-degree AVB), with 55% (n=79) of patients requiring pacemaker implantation. Of those with AV node disease who were discharged without a pacemaker, 37% (n=21) presented recurrence during follow-up. This poses the important question as to whether pacemaker implantation should be considered immediately after the first episode of bradycardia. Various authors have commented on this issue, to which there is as yet no definitive answer. Several studies have shown that AV node disease does not occur in structurally normal hearts and that even when it is induced by drug therapy, there is underlying conduction system disease.2,4

In the present study, most patients with renal failure achieved resolution of bradyarrhythmia and hence did not need pacemaker implantation. The literature shows that permanent pacing is required more often in patients with chronic renal failure than in the general population. This could not only be the consequence of hyperkalemia but might also reflect increased conduction system disease due to abnormalities in calcium metabolism leading to fibrosis and myocardial calcification.7 The level of severe renal failure in our population was low, and as a result resolution of bradyarrhythmia could be achieved by correcting the iatrogenic cause (drug therapy and/or electrolyte imbalance).

Hypertension is a common cardiovascular comorbidity in the general population as well as in our study. Depending on its severity and duration, hypertension can induce cardiac remodeling, which may affect electrical activity, leading to rhythm disturbances, particularly supraventricular and ventricular arrhythmias. There have been few multicenter studies on AV conduction disorders associated with hypertension.8

In our study, hemodynamic instability and consequent need for temporary pacing were not significantly associated with greater need for permanent pacemaker implantation, although a larger sample might enable more definite conclusions to be drawn. Hemodynamic instability was not associated with recurrence of bradyarrhythmia.

The three patients who died during hospitalization required temporary pacing due to hemodynamic instability, but did not respond. The deaths were not considered to be a complication of the procedure. These patients were old (aged over 80 years) and consequently had multiple comorbidities.

Although there were recurrences of bradycardia, there were no deaths during follow-up, which is an indication of the safety of current clinical practice, as most patients were discharged without a pacemaker following resolution of the iatrogenic cause of their bradyarrhythmia.

The study's retrospective nature and its small sample size are its main limitations. In addition, the relatively short follow-up period may have led to the frequency of bradycardia being underestimated.

We were unable to analyze the dosage of each drug associated with bradyarrhythmias for every patient, since this was not recorded at admission in all cases. Information on escape rhythm and QRS width was also incomplete, the latter because digitally stored ECGs were not available throughout the study period.

Patients did not undergo 24-hour Holter outpatient ECG monitoring, which might have identified asymptomatic bradycardia or AVB in some cases. Only symptomatic brady-arrhythmias with indication for permanent pacemaker implantation were assessed in this study.

In a high percentage of patients admitted with bradyarrhythmia associated with a potentially reversible cause, mainly drug-induced, the arrhythmia recurs or does not resolve, requiring placement of a permanent pacemaker.

Patients with AV node disease constitute a subgroup with a higher risk of recurrence who require greater vigilance during follow-up and should be considered for pacemaker implantation after the first episode.

The authors have no conflicts of interest to declare.