Simultaneous Estimation of Gender Male and Atrial Fibrillation as Risk Factors for Adverse Outcomes Following Transcatheter Aortic Valve Implantation

Post-TAVI outcomes by gender and AF

During the five-year period following TAVI, 141 patients (13.0%) died from any cause, 59 (5.4%) died from cardiovascular causes, and 183 (16.8%) experienced the composite endpoint of all-cause mortality or hospitalization for heart failure. Rates of most in-hospital complications were similar across groups, except for acute kidney injury, which showed some variability (Supplementary Table S1). Long-term event rates, expressed per 1,000 person-years, varied notably between the four groups. Patients who were male with AF consistently had the highest occurrence of all endpoints, whereas females without AF experienced the fewest events. Male patients without AF and females with AF had intermediate event frequencies (Figure 1).

Unadjusted Kaplan-Meier curves reinforced these trends, demonstrating that females without AF had the lowest cumulative incidence for all outcomes compared with the other three groups (Figure 2a; Supplementary Figure S1a). Interestingly, the presence of AF was associated with higher cumulative all-cause and cardiovascular mortality primarily among female patients, while the combined endpoint of death and heart failure hospitalization was elevated in AF patients of both sexes (Figure 2b).

To further explore the combined influence of male sex and AF, multivariate Cox regression analyses were conducted using females without AF as the reference group. Analyses incorporating a comprehensive set of covariates revealed that both male gender and AF independently increased the risk of all-cause mortality, and their coexistence was linked to the highest risk. When a more limited set of covariates was applied to evaluate multiple endpoints, including all-cause and cardiovascular mortality, the composite of death and heart failure hospitalization, and heart failure events alone, results were consistent. Male patients with AF had more than a threefold higher risk of all-cause (HR 3.4) and cardiovascular mortality (HR 4.4) compared with females without AF. For heart failure hospitalization specifically, AF emerged as the predominant risk factor, while gender exerted a smaller influence. These findings indicate that male sex and AF each contribute independently to poorer long-term outcomes after TAVI, with their combination particularly amplifying the risk of mortality.

This study examined the combined influence of gender and pre-existing atrial fibrillation (AF) on long-term outcomes after TAVI, evaluating all-cause and cardiovascular mortality as well as the composite of death and heart failure hospitalization across four patient groups defined by sex and AF status. By categorizing patients in this way, potential statistical interactions between AF and gender were avoided, allowing a clearer assessment of their independent effects.

The principal findings of this analysis are as follows. In-hospital mortality and major perioperative complications were comparable across all groups. However, when outcomes were tracked over a five-year period, significant differences emerged. Patients who were female without AF experienced the lowest incidence of adverse events, whereas male patients with AF consistently exhibited the highest rates. Kaplan-Meier analyses confirmed these trends, with females without AF demonstrating the lowest cumulative incidence across all endpoints. Multivariate Cox regression further demonstrated that both male sex and AF independently increased the risk of all adverse outcomes. Notably, the coexistence of male sex and AF did not result in a synergistic elevation of risk, highlighting their independent contributions to long-term prognosis after TAVI.

Aortic valve replacement improves cardiac hemodynamics by reducing afterload in patients with severe AS, and technical refinements in TAVI have progressively reduced perioperative risk [15]. Despite these advances, long-term mortality after TAVI remains substantial [16], emphasizing the need for accurate risk stratification not only for procedural safety but also for long-term outcomes, particularly as TAVI expands to lower-risk populations [5]. Traditional risk scores, such as EuroSCORE, EuroSCORE II, and STS-PROM, remain widely used for short-term risk assessment, but these tools were originally developed for surgical populations and may not reliably predict long-term outcomes in TAVI recipients [17,18]. In prior analyses of this registry, the STS-ACC TAVR risk score, specifically designed for TAVI patients, demonstrated superior predictive performance compared with conventional models [19]. The addition of AF, along with serum albumin and body mass index, further improved the accuracy of risk prediction, underscoring the importance of AF as a prognostic factor in TAVI.

The influence of gender on TAVI outcomes has been inconsistently reported in the literature. Some studies indicate improved mid- to long-term survival in women [20, 21], while others have found comparable or worse outcomes compared with men [22, 23]. Large registry analyses and meta-analyses have suggested higher rates of vascular complications and short-term mortality among women, potentially related to higher utilization of non-transfemoral TAVI approaches [24, 25]. In contrast, the present study observed similar short-term outcomes between men and women, but women experienced significantly better long-term survival. The low proportion of non-transfemoral TAVI (8.7%) in this cohort, together with similar access site distributions across the four groups, likely contributed to the absence of sex-based differences in early outcomes. These observations suggest that procedural factors had a limited effect on long-term prognosis in this study population, highlighting the independent roles of gender and AF in shaping post-TAVI outcomes.

Aortic stenosis (AS) imposes pressure overload on the left atrium, promoting structural remodeling that predisposes patients to atrial fibrillation (AF), which in turn increases the risk of adverse events such as stroke and heart failure [26, 27]. Increasing evidence supports the negative prognostic impact of both pre-existing and new-onset AF in patients undergoing TAVI [28, 29]. In the present study, the overall prevalence of pre-existing AF was 24.3%, aligning with prior reports that ranged between 16% and 51.1% [13]. Contrary to earlier studies reporting higher AF prevalence among men, AF occurred at a similar rate in both sexes in this cohort [25]. Interestingly, the adverse impact of AF on all-cause and cardiovascular mortality was more pronounced in women than in men. While the smaller number of male participants may have limited the statistical power to detect a significant effect, these findings suggest the possibility of a gender-specific influence of AF on post-TAVI prognosis. Given established sex differences in AF epidemiology, clinical presentation, and outcomes [30], it is plausible that male sex and AF could have additive effects on adverse events, although synergistic interactions remain uncertain.

Previous research has typically examined the effects of gender or AF separately on short- and long-term TAVI outcomes, with few studies considering their combined impact. The present analysis demonstrates that both male sex and pre-existing AF independently increased the risk of all-cause and cardiovascular mortality as well as the composite of death and heart failure hospitalization. While patients with both risk factors exhibited the highest hazard ratios for all endpoints, the combined presence of male sex and AF did not result in a statistically significant synergistic increase in risk. These observations suggest a primarily independent, with a minor additive, effect of these factors on post-TAVI outcomes.

Limitations

Several limitations should be noted when interpreting these findings. First, the registry included a relatively small number of institutions and patients, and the retrospective design may allow unmeasured confounding factors to influence outcomes. Second, temporal changes in the prognostic impact of gender and AF should be considered. Previous work has suggested that the survival advantage of female TAVI patients has diminished in more contemporary cohorts [31]. Although the majority of participants in this study (96.3%) were classified as contemporary, and the multivariate model accounted for the year of TAVI, residual temporal effects cannot be excluded. Third, while this study found no significant synergistic effect of male sex and AF in the overall cohort, such an effect cannot be completely ruled out in specific subpopulations. Fourth, similar to other large Japanese TAVI registries [32], females constituted a higher proportion of patients than in many international studies [7, 31, 33], which may have influenced the observed results. Future investigations incorporating data from diverse global registries are needed to validate these findings.

Conclusion

In summary, this retrospective cohort study from a multicenter Japanese registry demonstrates that male sex and pre-existing AF are independently associated with an increased risk of all-cause and cardiovascular mortality, as well as the composite of death and heart failure hospitalization following TAVI. Although no evidence of a synergistic effect between these two factors was observed, their individual prognostic significance underscores the importance of considering both gender and AF in clinical decision-making and post-procedural management. These findings highlight the need for tailored treatment strategies and vigilant follow-up in patients undergoing TAVI to optimize long-term outcomes.

Acknowledgments: The authors wish to thank Morimasa Takayama and Tetsuya Tobaru for their significant contributions to this registry database. The authors also appreciate the statistical assistance of Ryo Naito during the process of preparing the revised version of the manuscript.

Conflict of interest: None.

Financial support: This study was supported by the Sakakibara Clinical Research Grant for Promotion of Sciences, 2020 year. grant number H-4-2020.

Ethics statement: None.