Results

Third degree atrioventricular block was easily achieved in all patients with no complications, and was maintained at follow up. The mean number of burns required to achieve complete heart block was 2.6 (range 1 to 12) and it was not necessary to proceed to a combined left and right sided approach in any patient. Pacemaker function was satisfactory in all patients both immediately and at six week follow up. In particular, all patients had satisfactory atrial sensing. Pacemaker interrogation at this stage suggested that four patients (22%) had developed persistent atrial fibrillation; patients were left programmed to DDDR/MS in case sinus rhythm returned.

Table 1 summarises the quality of life scores, antiarrhythmic drug burden, left ventricular fractional shortening, and exercise capacity before and after intervention. It can be seen that we were able to reduce the number of antiarrhythmic drugs taken following the procedure (p < 0.01) and that there was an improvement in the psychological general wellbeing (p < 0.01; fig 1) and cardiac symptom scores (p < 0.01; fig 2). Before the intervention, all patients had normal left ventricular systolic function (FS > 30%). We were unable to show any significant change in left ventricular fractional shortening, exercise tolerance, or the MHI score after intervention.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Individual cardiac symptom scores before and six weeks after atrioventricular node ablation. Error bars show the mean score and standard deviation.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Individual psychological general wellbeing sores before and six weeks after atrioventricular node ablation. Error bars show the mean score and standard deviation.

Discussion

Previous studies of atrioventricular node ablation and pacemaker implantation have demonstrated its feasibility3 ,4 and an improvement in quality of life9 ,11 ,20 with its use in certain populations. However, these studies have not specifically addressed the use of dual chamber, mode switching technology in combination with atrioventricular node ablation for patients with paroxysmal atrial fibrillation. A criticism of some of the studies is that they have applied the technique to mixed populations of chronic and paroxysmal fibrillators10 or even to other supraventricular arrhythmias,14 thus making it difficult to apply the findings specifically to patients with either chronic or paroxysmal atrial fibrillation. Kay’s original series9did focus on paroxysmal fibrillators but this was before the availability of radiofrequency ablation and only VVIR pacing was used. It does not, therefore, directly support the use of DDDR/MS pacemakers in this population. The only other study which looked specifically at paroxysmal fibrillators did use dual chamber pacing11; however, this was a retrospective survey of 36 patients and the devices implanted did not incorporate mode switching technology.

Despite this lack of evidence, current pacing prescription guidelines21 ,22 advocate the use of DDDR/MS in this context, and prescription patterns would suggest that these guidelines are becoming widely accepted. Our study, however, shows that atrioventricular node ablation and implantation of a dual chamber mode switching pacemaker is an effective treatment for drug refractory paroxysmal atrial fibrillation. In the population studied it allowed a significant reduction in antiarrhythmic drug burden while producing an improvement in quality of life and cardiac symptoms. These improvements were detected by the psychological general wellbeing index and the cardiac symptom score. The MHI did not improve significantly; we suggest that this was because it assesses patients’ general health in terms of everyday functional ability, and our patients—although highly symptomatic and psychologically affected by their symptoms—were not very limited functionally. This may be exaggerated by the fact the MHI only asks patients to assess their functional ability on the day of assessment and hence the preoperative score may already be high if the patient is assessed in sinus rhythm, as all of ours were. Indeed, several of the patients attained a maximum score before intervention, thus allowing no room for improvement. This finding is in keeping with those of Fitzpatrick et al,11 whose retrospective assessment was able to show an improvement in quality of life and cardiac symptoms but not in the activities of daily living in paroxysmal fibrillators (though there was a significant improvement in all three in chronic fibrillators).

Our patients’ pacemaker prescription was dictated by physician preference and hence the mode switching algorithms employed varied between patients. The CPI Vigor DR and the Medtronic Thera DR both have counter based algorithms which require a certain number of cardiac cycles to detect atrial fibrillation and hence may take 5–10 seconds to mode switch; in contrast the Vitatron Diamond device has a “beat to beat” algorithm which will switch on the first beat of atrial fibrillation (or on a single atrial extrasystole). The advantage of the latter algorithm is that a more constant heart rate is maintained at the onset of atrial fibrillation rather than the pacemaker tracking to the upper rate limit and only falling back to the sensor rate once the arrhythmia is diagnosed. Despite the fact that the majority of our patients had slow mode switching devices, we were still able to demonstrate the improvements in quality of life described above. It is possible that the improvements may have been greater if all patients had received a device with a “faster” algorithm.23

In any study of the impact of a procedure on quality of life and wellbeing, a placebo effect must be considered. None of the studies of atrioventricular node ablation and pacing for paroxysmal atrial fibrillation (including ours) assesses this problem. However, Brignoleet al have addressed this in a short term study of a mixed group of patients with chronic atrial fibrillation and flutter.20 They randomised patients to undergo either pacemaker implantation alone or atrioventricular node ablation, and VVIR pacing at the same session. Patients were assessed before and 15 days after their procedure, using a symptom score, New York Heart Association classification, and an activity scale. While there were improvements in the group treated with pacing alone, the improvement in the ablated patients was nearly twice as great, suggesting that although there may be some placebo effect associated with undergoing a procedure, this does not account for all the benefits observed. It might also be argued that even if there is a placebo effect associated with the procedure, it is still beneficial if patients’ perception of wellbeing is improved.

In contrast to the effects on quality of life, we were unable to show any change in exercise capacity after the procedure. Only Kay’s study9 has looked at this in paroxysmal fibrillators, when an improvement was seen. Brignole et al found a similar effect in chronic fibrillators, as might be expected.20 We suggest that, as for the MHI scores, we did not detect a change in exercise capacity because the functional status of our patients was generally good before intervention. This difference from Kay’s population9 may reflect the fact that atrioventricular node ablation is now relatively commonplace and that at the time of Kay’s study it would have been considered much more a treatment of “last resort.” The lack of improvement in exercise capacity in our population may also be related to the fact that all our patients were exercised in sinus rhythm at baseline, but three were in atrial fibrillation at follow up; hence there would have been a loss of atrial transport in these patients at the time of their postoperative test. It is important to note, however, that despite these factors we did not find a deterioration in exercise capacity after intervention.

With regard to the effects of atrioventricular node ablation and pacing for atrial fibrillation on left ventricular systolic function, published reports to date contain no studies using the DDDR/MS pacing mode and, as with previous quality of life studies, the populations studied have been heterogeneous with regard to the chronic or paroxysmal nature of the arrhythmia.12 ,13 ,24 However, in patients with impaired left ventricular function and chronic atrial fibrillation, atrioventricular node ablation and pacing seems to improve left ventricular systolic function.20 ,25 It is suggested that in paroxysmal atrial fibrillation the degree of improvement is related to the frequency and duration of atrial fibrillation and is again restricted to those patients with marked left ventricular impairment.13 In our study there was no beneficial effect on left ventricular systolic function. The lack of improvement is perhaps not surprising as most of the patients were assessed in sinus rhythm and all had normal left ventricular function before ablation. It is important to note, however, that as with studies of patients with chronic atrial fibrillation and normal left ventricular function, we did not show any detrimental effect of the procedure on left ventricular fractional shortening, despite the potentially negative inotropic effect of right ventricular apical pacing.26

The results of this study included four patients (22%) in whom atrial fibrillation was paroxysmal at the time of ablation and pacing, but who rapidly progressed to persistent atrial fibrillation on withdrawal of antiarrhythmic drugs. All of these patients had an improvement in their quality of life scores after intervention. It seems likely that they would have fared equally well with VVIR pacing; however, as their atrial fibrillation was clearly paroxysmal at the time of ablation and their pacemakers were left programmed to DDDR/MS on an intention to treat basis, we feel their inclusion in the results was justified. It is evident, however, that further work is required to identify risk factors for progression to chronic atrial fibrillation in a population of patients undergoing ablation and pacing for paroxysmal atrial fibrillation.

In conclusion, we found that atrioventricular node ablation and implantation of a DDDR/MS pacemaker was an effective treatment for paroxysmal atrial fibrillation. We were unable to show any detrimental effect on left ventricular systolic function or exercise capacity. However, it remains unclear whether DDDR/MS pacing is superior to VVIR, or indeed whether optimising medical treatment may achieve the same results.