3.5 Predictors of mortality after TAVI
We used Cox proportional hazard models to analyze the predictors of mortality after TAVI, as shown in Table 4. The univariate analysis revealed that preoperative serum albumin and hemoglobin were associated with mortality. In the multivariate analysis, only preoperative serum albumin showed a tendency towards predicting mortality, however the association did not reach statistical significance (hazard ratio: 0.54, 95% confidence interval CI: 0.65–1.03, p=0.08).
DISCUSSION
Life expectancy varies from country to country and in addition, the healthcare systems of each country are unique in themselves. In Japan, where health insurance is a requirement, the healthcare delivery environment is different as compared to other countries where health insurance is not a requirement. There are several reports on the outcomes of TAVI for nonagenarians in Western countries.6-8 However, to our knowledge, this is the first report to investigate outcomes in the Japanese population. Takeji et al. reported that the current hospital mortality rate of TAVI in Japan is 1.3%, suggesting that it has a high success rate and is being performed safely with very low mortality. 3 This study revealed that TAVI in nonagenarians have excellent early- and mid-term outcomes. The in-hospital mortality rate of our nonagenarian cohort was 0%, which was statistically equivalent to the younger patients’ cohort (0%). Although hospital stays were longer in the nonagenarian cohort, the other early outcomes were comparable, indicating that TAVI can be safely performed even in very elderly populations. Investigators from Western countries have shown that TAVI in nonagenarians can achieve acceptable in-hospital outcomes and that age alone should not exclude patients from treatment, 6-8 a viewpoint that is supported by our findings.
A decade ago, SAVR was the only effective therapy for aortic valve stenosis. The development of TAVI has dramatically changed the surgical treatment of aortic valve stenosis. Previous studies with nonagenarian cohorts showed that early mortality after SAVR ranged from 11% to 17%.9,10 In contrast, the in-hospital or 30-day mortality rates after TAVI in nonagenarians has been reported as 0%–8.7% in recent observational studies, showing a distinct advantage.6,7 In fact, there were no in-hospital or 30-day mortalities in our nonagenarian cohort. Although the long-term results of TAVI remain unclear, it is a useful therapeutic option and a reasonable alternative to SAVR in nonagenarians based on their relatively short life expectancy.
Our results showed that nonagenarian patients required longer hospital stays. Prolonged hospitalization can cause postoperative muscle weakness and deterioration of activities of daily living (ADL). Stehli et al. reported that deterioration of ADL after TAVI is more frequent in older patients and that 25% of nonagenarians transition to aged-care facilities within 1 year after TAVI. 7 We believe that this points to a particular disadvantage of TAVI in very elderly patients. In this cohort, only 15 of 23 nonagenarians had frailty assessment data available as these assessments were not routinely performed during the early period of our series. Although the sample size is limited, we could not find any deterioration in frailty status in the early phase after TAVI (Supplemental table1). However, the data is insufficient for assessing the possible future changes in ADL and quality of life. Unconscious scaling back of physical activity in patients with cardiac disease is a well-characterized phenomenon. Gradual improvement in activity may be possible after recovery of cardiac function (Supplemental Figure1). Further investigations into long-term ADL outcomes are required.
The freedom from cardiac events rates in each group were almost equivalent in this study. Interestingly no late cardiac deaths occurred in the nonagenarian cohort. The Japanese Ministry of Health, Labor and Welfare reported that the top 5 causes of mortality in nonagenarians are cardiac disease, pneumonia, senility, cerebrovascular disease, and malignancy. 11 In our nonagenarian cohort, the causes of late mortality were pneumonia in 3 patients, other infectious disease in 3, and malignancy in 1, but there were no cardiac deaths. Thus, TAVI may contribute to the avoidance of cardiac death in nonagenarians.
The wide-spread adoption of TAVI and the aging of the population have led to an increase in the number of procedures performed worldwide. The identification of prognostic factors is thus essential for patient selection and stratification in the TAVI era. Various predictors of prognosis after TAVI have been reported; for example, anemia, intra- or post-operative blood transfusion, psoas muscle area, and appetite immediately before discharge were associated with postprocedural mortality. 12-15 In this study, lower preoperative serum albumin seemed to be associated with poor long-term outcomes, but the results were not statistically significant (p=0.08). Serum albumin is a known frailty marker and may be a potential predictor of TAVI outcomes. In this study, age was not a predictor of prognosis despite generally being a significant factor in all procedures. Older patients have more co-morbidities and are at a higher operative risk. In fact, major operative risk scores, including the STS risk score, EuroSCORE, and JapanSCORE, identified older age as independent operative risk factor. 16-18 We believe that our results could be explained by the equivalent risk scores in the nonagenarian and the younger age groups (Table 1). This indicates that the nonagenarian patients may have been selected to some degree and be in relatively better condition as compared to the younger group, which would affect the assessment of age as a prognostic indicator.
This study has several limitations. Since this was a retrospective, observational study, selection bias may be present. The sample size was limited because our hospital is not a high-volume center. Furthermore, data regarding outcomes of conservative medical therapy in nonagenarian aortic stenosis patients were not available.
CONCLUSIONS
The early and mid-term clinical outcomes of TAVI in selected nonagenarians were comparable to those in younger patients and TAVI may contribute to the prevention of cardiac death in nonagenarians. These findings indicate that TAVI may be an effective treatment for aortic stenosis, even in nonagenarian patients.