3.1 Sinus and paced P-wave duration analysis
The distribution of sinus P-wave duration was as follows: < 120 ms (64%), 120-139 ms (26%), and ≥ 140 ms (10%); and for paced P-wave, < 120 ms (11%), 120-159 ms (49%), and ≥ 160 ms (40%). Table 1 reflects sinus and paced P-wave duration (ms) according to the presence of AHRE. Mean paced P-wave was significantly longer than sinus P-wave (154±27 vs 115±18 ms; p < 0.001), both in patients with AHRE (161±29 ms vs 119±20 ms; p < 0.001) and without AHRE (148±23 ms vs 112±16 ms; p < 0.001). IAB was present in 36% of the patients in sinus rhythm (31% partial, 5% advanced) and 89% during atrial stimulation (p = 0.011). IAB in sinus rhythm was present in 43% of the patients with AHRE and 30% of the patients without AHRE (p = 0.039) and i4% and 86%, respectively, during atrial stimulation (p = 0.034).
ROC curves showed the following values for area under the curve (AUC): sinus P-wave: 0.594 ± 0.038 (p = 0.016) and paced P-wave: 0.635 ± 0.038 (p < 0.001) (Figure 2). No significant difference between sinuthe s and paced P-wave duration for AHRE development was found, although paced P-wave showed a better performance.
In a subgroup analysis dividing the population according to paced P-wave duration (Table 3), a paced P-wave ≥ 160 ms showed the strongest association with AHRE development, especially for those lasting > 24 h (OR 4.2; 95% CI [1.6-11.4] p < 0.004)
Table 4 shows the results of a univariate analysis in which several cut-off points in sinus/paced P-waves and clinical factors were included. Among the statistically significant predictors for AHRE, a paced P-wave ≥160 ms, a CHA2DS2-VASc score ≥4, and a LA SI dimension ≥ 55 mm were included in a multivariable model which yielded the following 3 independent predictors augmenting AHRE: paced P-wave ≥160 ms (OR: 1.84; 95% CI [1.03–3.28]; p = 0.038), CHA2DS2-VASc score ≥4 (OR: 2.01; 95% CI [1.13–3.55]; p = 0.016), LA dimension ≥55 (OR 2.69; 95% CI [1.48–4.90]; p = 0.001).
DISCUSSION
We have found that patients with CIED using right atrial stimulation in conventional atrium lead location presented significant prolongation of the P-wave duration and a higher prevalence of IAB. Furthermore, our results show that a longer P-wave duration and a higher presence of IAB are significantly related to the development of AHRE during long-term follow-up. A paced P-wave duration ≥160 ms had the highest predictive value for AHRE of any duration, especially for those lasting >24 h. Only patients without prior known AF were included, since a history of this entity is considered an independent risk factor for the development of AHRE.
The development of AHRE is relatively common among patients with CIED13, 14. In the recently published IMPACT study, AHRE episodes of >6 min to <6 h were the most prevalent among all duration categories, and age, history of AF, and hypertension were the most important risk factors 14. Although they have been related to the development of clinical AF, their clinical implications are not yet sufficiently clarified15. Even so, finding prognostic markers for the development of AHRE and therefore of possible clinical AF is currently an important goal due to the possible therapeutic implications, especially as concerns anticoagulation 16.
Our group previously analyzed 380 patients who had undergone CIED implantation, showing that IAB is a significant predictor of AHRE, both in subjects with and without a previous history of AF8. Tekkesin et al included 367 patients with CIED with no previous history of AF, and 32% of them had IAB (29% partial and 3% complete) 10. This proportion was also higher in patients who developed AHRE during follow-up (12). In the present study, this proportion was similar: 36% of the patients had IAB, and its presence was significantly higher in patients with AHRE than without it (43% vs 30%; p = 0.039). Ariyarajah et al found a higher IAB prevalence in patients with previous embolic stroke than in patients with a non-embolic stroke (61% vs. 41%; p < 0.05)5. Some other studies also support the same relation between the presence of advanced IAB and the development of AF and ischemic stroke 17. Tse et al. carried out a meta-analysis of 16 studies with 18204 patients in which they found that IAB, especially advanced IAB, was a significant predictor of de novo and recurrent AF 18. In the study by Lehtonen et al with 5667 patients, AF, male sex, hypertension, age, and obesity were identified as clinically significant factors that prolonged the duration of the P-wave 19. The REGICOR registry showed a higher IAB prevalence in patients with AF vs. without AF (63% partial and 11% advanced vs 46% partial and 9% advanced; p = 0.0006) 3.
However, there are almost no studies analyzing the paced P-wave, the incidence of IAB, and the development of AHRE. Only Kristensen et al.20 conducted a study with 109 patients who had undergone CIED implantation due to sinus node disease or brady-tachy syndrome, in which the duration and dispersion of the sinus and paced P-waves were measured, analyzing their relationship with AHRE and AF during follow-up. Although the paced P-wave was wider than the sinus P-wave during high atrial pacing (155 ± 27 ms vs 117 ± 26 ms), there was no significant difference in the paced P-value duration between patients with and without AHRE, and it was not a significant predictor of AHRE/AF. Only a previous diagnosis of brady-tachy syndrome was a significant predictor. These findings differ from ours since we demonstrated that the paced P-wave duration or the presence of IAB during atrial pacing are significant predictors of AHRE during follow-up. In addition, as in the study by Kristensen et al, we found that with atrial stimulation the duration of the P-wave increases by 39 ± 29 ms on average and, consequently, the proportion of patients with IAB (89% with stimulation vs. 36% without stimulation).
All traditional works define the presence of IAB by taking a P-wave duration ≥ 120 ms, though this value may not be appropriate if atrial stimulation alone increases the duration of the P-wave and generates a greater degree of IAB or develops a previously nonexistent IAB. Perhaps values ​​greater than 120 ms should be considered to define IAB in the presence of atrial stimulation, making them better predictors of AHRE. That is why we have carried out a predictive analysis of AHRE as a function of the width of the paced P-wave, finding that a greater width increases the risk of presenting AHRE of longer duration. A paced P-wave ≥160 ms has the best prognostic value for any duration of AHRE episodes, but especially for those lasting >24 h, which are perhaps those with the most important clinical implications.
We found that in addition to paced P-wave duration, the presence of left atrial enlargement and a CHA2DS2-VASc scale ≥4 are significant independent predictors of AHRE development in the general population with CIED. Atrial enlargement by itself is associated with IAB 2, 21, 22. All these findings may have an important clinical impact since some studies have shown that a higher percentage of atrial pacing is associated with an increased risk of AHRE 23. This could also explain why the vast majority of studies on the prevention of AHRE/AF with different lead atrial placements have failed 24. Whether atrial stimulation itself could be also pro-arrhythmogenic, especially in patients with previous IAB, is subject to debate, and better control of the paced P-wave duration during implantation and follow-up of patients with CIED would be necessary to avoid future episodes of AHRE.