The Impact of Percutaneous Coronary Intervention on Mortality in Patients With Coronary Lesions Who Underwent Transcatheter Aortic Valve Replacement

Objectives. To analyze the effect of percutaneous coronary intervention (PCI) before transcatheter aortic valve replacement (TAVR) on all-cause and cardiovascular mortality after TAVR, differentiating between significant proximal lesions and the non-proximal (residual) lesions. Methods. An institutional TAVR database was complemented with data on the extent of coronary artery disease (CAD), lesion location, lesion severity, and the location of PCI. Survival analysis was performed to investigate the impact on 6-month and 3-year mortality after TAVR in all patients and in subgroups of patients with significant proximal lesions (>70% diameter stenosis [DS], >50% DS in left main), the non-proximal residual lesions, and in a propensity score matched cohort. Results. Among the 577 included patients, mean age was 83 years, 50% were female, and 31% had diabetes mellitus. Preprocedural PCI of unselected lesions was independently associated with increased 6-month mortality (hazard ratio, 2.2; 95% confidence interval, 1.0-4.6; P=.04), but selective PCI of significant proximal lesions did not have an association with higher mortality, nor did we find a significant effect of PCI on mortality in the propensity-matched cohort. Conclusion. Routine pre-TAVR PCI is not associated with mortality reduction in TAVR patients with coronary lesions in any segment or in patients with proximal coronary lesions. Despite the lack of a beneficial effect of routine pre-TAVR PCI, we cannot exclude a beneficial effect in a selection of patients with proximal lesions. Therefore, we strongly support the current clinical guidelines to only consider pre-TAVR PCI in proximal coronary lesions, while advocating a restrictive pre-TAVR PCI strategy.

J INVASIVE CARDIOL 2021;33(10):E823-E832.

Key words: coronary artery disease, percutaneous coronary intervention, transcatheter aortic valve replacement

Transcatheter aortic valve replacement (TAVR) is a minimally invasive treatment in patients with symptomatic severe aortic valve stenosis (AS).¹ Current guidelines recommend coronary angiography prior to TAVR to assess coronary artery disease (CAD), but only recommend consideration of percutaneous coronary intervention (PCI) in lesions with >70% diameter stenosis (DS) in a proximal segment of a coronary artery (class IIa recommendation, level of evidence C).¹ Surprisingly, these recommendations are not supported by clinical studies that assessed the effects of pre-TAVR PCI in patients with proximal coronary lesions. Recent meta-analyses of pre-TAVR PCI in any coronary lesion showed no advantageous effect of pre-TAVR PCI.^2,3

Therefore, we hypothesized that only pre-TAVR PCI of significant proximal lesions is associated with improved survival. In order to test the recommendation of the guidelines, we evaluated the impact of pre-TAVR PCI in all patients with coronary lesions >50% DS on all-cause and cardiovascular mortality, and differentiated between pre-TAVR PCI in significant proximal coronary lesions and pre-TAVR PCI in non-proximal lesions.

Study design and population. This observational cohort study was performed in accordance with the principles of the Declaration of Helsinki and data collection was in accordance with the policy of the Amsterdam UMC. The local institutional review board approved the study. For this cohort study, data were used from a registry with prospectively acquired data that comprised consecutive patients who underwent TAVR at the Amsterdam UMC in the Netherlands. The registry was supplemented with retrospective data from coronary angiography and PCI reports. In the Amsterdam UMC, eligibility for TAVR and the revascularization strategy were discussed by a multidisciplinary heart (valve) team. The default TAVR approach was transfemoral and device type and sizing were at the discretion of the operators. During the study period, the predominantly used valve type was the balloon-expandable Sapien 3 valve (Edwards Lifesciences). Mortality data were obtained from the Dutch national municipal registry in February 2019, ensuring complete follow-up. Causes of death were obtained via the general practitioner. Information on PCI procedures during follow-up post TAVR was obtained via the Netherlands heart registry.⁴ All patients undergoing TAVR between October 2007 and November 2018 were included in the analysis. Patients were excluded from the analysis if CAD data and coronary angiography images were missing, if the patients had a history of coronary artery bypass graft (CABG) surgery, or if vital status could not be obtained.

Study objective. The objective was to analyze the effect of pre-TAVR PCI on all-cause and cardiovascular mortality in all patients with coronary lesions with >50% diameter stenosis (DS). In this analysis, we subsequently differentiated between PCI in proximal significant coronary lesions, PCI in non-proximal lesions, and PCI in the left main (LM) coronary artery and/or proximal left anterior descending (LAD) coronary artery. We also analyzed the effect of PCI in a propensity-score matched cohort.

Data collection and definitions. The collected data consisted of patient demographics, patient comorbidities, preoperative assessment parameters, and procedural characteristics. Demographic characteristics included information on gender, age, and body mass index (BMI). BMI was classified accordingly as normal weight (23-30 kg/m²), underweight (<23 kg/m²), or overweight (>30 kg/m²). Patient comorbidities included cardiovascular as well as chronic pulmonary and renal dysfunction. Preoperative assessment included risk scores (Society of Thoracic Surgeons [STS] risk score and EuroScore II) and echocardiographic-measured left ventricular ejection fraction (LVEF), aortic valve area (AVA), and maximum pressure gradient. LVEF was scored as normal function (>55%), moderately impaired function (30%-55%), or poor function (<30%). The TAVR access route, implanted valve type, and year of TAVR were documented. Causes of death and information on PCI during follow-up were collected. Procedural mortality was defined as all-cause mortality within 30 days or during index procedure hospitalization, as defined by the Valve Academic Research Consortium 2 criteria.⁵

Data on the extent of CAD and pre-TAVR PCI were collected from the coronary angiography and PCI reports. For the purpose of this study, coronary assessments performed up to 12 months prior to TAVR were included. The coronary tree was divided into 16 segments, as suggested by the American Heart Association. Lesion location was noted. Lesion severity (expressed as percentage of DS) was noted per segment and divided between <50% DS, 50%-70% DS, 70%-90% DS, 90%-100% DS, and 100% DS/chronic total occlusion (CTO). Significant proximal coronary lesions were defined as >70% DS in a proximal coronary segment(s) and/or as >50% DS in the LM. Pre-TAVR PCI was defined as PCI during or after the coronary angiography that was performed during the work-up for TAVR.

Statistical analysis. All statistical analyses were performed with R software, version 3.5.1 (R Foundation for Statistical Computing). To analyze the effect of PCI, patients with coronary lesions were subdivided into those with PCI vs those without PCI treatment in the work-up for TAVR. Continuous baseline variables are presented as mean ± standard deviations or median (interquartile range [IQR]). The distribution of continuous variables was tested with the Shapiro-Wilk test and parametric one-way analysis of variance (ANOVA) test. The non-parametric Kruskal-Wallis test was used for comparison of continuous variables. Categorical baseline variables are presented as frequencies and percentages and compared using the Chi-square test or Fisher’s exact test. Survival analysis after TAVR (baseline) was performed using Kaplan-Meier analysis and subgroups were compared with the use of the log-rank test. Before Cox regression analysis, variables with missing data were complemented with imputed data. No variables exceeded 10% missing values; they were imputed using the multivariate imputation by chained equations (mice) package in R, version 3.5.1.⁶ Multivariable Cox proportional hazards regression models were used to calculate hazard ratios (HRs) with corresponding 95% confidence intervals (CIs) for the follow-up periods of 6 months and 3 years. The multivariable Cox analysis model consisted of all available baseline characteristics (Table 1), supplemented with the year of the TAVR procedure and the lesion location and severity. Multivariable Cox proportional hazard models were created to investigate the impact of the CAD extent (total cohort) and to investigate the impact of the location of coronary lesions and the effect of PCI (patients with lesions >50% DS and without CABG).

An additional analysis was performed for the effect of PCI using a propensity-score matched cohort. Propensity-score matching was performed with the MatchIt package in R statistics.⁷ We used predictors for PCI treatment that were identified with multivariable logistic regression, together with predictors that were identified as being important for the decision to perform PCI before TAVR by 5 interventional cardiologists. The cohort was matched using nearest-neighbor matching. Predictors included age, gender, history of PCI, extent of CAD and location of the coronary lesions, aortic valve area, diabetes, estimated glomerular filtration rate, and type of TAVR (balloon-expandable valve vs self-expanding valve).

Descriptive statistics. A total of 1546 patients underwent TAVR in the Amsterdam UMC between 2007 and November 2018. From the included patients, 27 were excluded with missing reports of coronary angiography or PCI, 5 were missing vital status, and 191 had a history of CABG (Figure 1). Of these 1323 patients, 746 did not have coronary lesions of >50% DS. Of the remaining 577 patients with coronary lesions, 150 were treated with PCI and 427 were not. The median age of the study population was 83 years (IQR, 78-86 years) and 287 patients (50%) were women (Table 1). Patients with pre-TAVR PCI had a significantly more frequent medical history of PCI and myocardial infarction, and were more frequently implanted with balloon-expandable valves. Pre-TAVR PCI was performed significantly more often in patients with triple-vessel disease and in patients with significant LAD lesions. The 3-year mortality rate in patients with coronary lesions with >50% DS was 31.0% (n = 179). As a reference, for the patients without coronary lesions and without CABG (not included in further analysis), the 3-year mortality rate was 24.9% (n = 185; HR, 1.1; 95% CI, 0.89-1.5; P=.30). The causes of death of the patients with coronary lesions are listed in Table 2.

Survival analysis. In the 577 patients with coronary lesions >50% DS, mortality rates and Kaplan-Meier estimates for all-cause and cardiovascular mortality at 6-month follow-up were significantly higher for patients who underwent PCI vs patients who did not undergo PCI in the work-up for TAVR (P=.01) (Table 3 and Figure 2A). After adjustment for covariates, pre-TAVR PCI was associated with higher all-cause mortality at 6-month follow-up, but not for cardiovascular mortality (all-cause mortality: HR, 2.2; 95% CI, 1.0-4.6; P=.04; cardiovascular mortality: HR, 1.9; 95% CI, 0.87-4.0; P=.11). Patients who underwent pre-TAVR PCI had a significantly higher procedural mortality (Table 3) and more frequently underwent PCI within 3 years after TAVR than patients who did not undergo pre-TAVR PCI (5.3% vs 2.3%, respectively).

Subgroup analysis. In the 235 patients with significant proximal coronary lesions (>50% DS in LM and/or >70% DS in proximal segment), all-cause Kaplan-Meier mortality curves and estimates were comparable between the patients who did and did not undergo pre-TAVR PCI (Table 3 and Figure 2B). HRs were not significantly different for 6-month and 3-year follow-up (all-cause mortality: HR, 1.8; 95% CI, 0.59-5.4; P=.31 and HR, 1.1; 95% CI, 0.53-2.2; P=.86, respectively; cardiovascular mortality: HR, 1.3; 95% CI, 0.43-4.1; P=.63 and HR, 0.94; 95% CI, 0.43-2.1; P=.88, respectively). Procedural mortality rates (5.2% in the PCI group vs 7.0% in the no-PCI group) and the percentage of post-TAVR PCI treatment (5.2% in the PCI group vs 3.2% in the no-PCI group) were comparable (Table 3).

For the 342 patients with non-proximal (residual) coronary lesions, Kaplan-Meier mortality curves and estimates were significantly higher in the pre-TAVR PCI group (Table 3 and Figure 2C). After adjustment for covariates, there was no significant association between PCI treatment and all-cause and cardiovascular mortality at 6-month and 3-year follow-up (all-cause mortality: HR, 2.6; 95% CI, 0.85-7.7; P=.09 and HR, 1.7; 95% CI, 0.86-3.5; P=.13, respectively; cardiovascular mortality: HR, 1.7; 95% CI, 0.86-3.5; P=.12 and HR, 1.8; 95% CI, 0.78-4.0; P=.17, respectively). In comparison with patients who had significant proximal coronary lesions, procedural mortality was significantly higher in the patients with PCI in the non-proximal coronary lesions (13.7% in the PCI group vs 3.0% in the no-PCI group) (Table 3). The percentage of post-TAVR PCI was higher in the PCI group, but not significantly different (5.5% in the PCI group vs 1.9% in the no-PCI group).

PCI treatment in the LM and proximal LAD (Figure 2D) resulted in comparable mortality rates and there was no significant associations with all-cause and cardiovascular mortality at 3-year and 6-month follow-up.

Propensity-score matched cohort. The matched cohort consisted of 90 PCI-treated patients and 90 controls. Baseline characteristics did not differ significantly for patient characteristics, procedural characteristics, the extent of CAD, and/or the severity and location of coronary stenosis (Table 4). Although not significantly different, standardized mean differences were above 0.10 for 7 variables. Patients in the matched cohort who underwent PCI had more hypertension, less chronic obstructive pulmonary disease, less atrial fibrillation, less smoking, lower EuroScore II, more triple-vessel disease, and more non-significant lesions. Kaplan-Meier curves were comparable (Figure 3) and multivariable Cox analysis revealed no association between PCI and a higher mortality at 6-month and 3-year follow-up (all-cause mortality: HR, 0.9; 95% CI, 0.32-2.6; P=.85 and HR, 1.2; 95% CI, 0.64-2.1; P=.63, respectively; cardiovascular mortality: HR, 0.9; 95% CI, 0.32-2.6; P=.85 and HR, 1.1; 95% CI, 0.59-2.1; P=.73, respectively).

In the present analysis, pre-TAVR PCI was not associated with reduced mortality. Surprisingly, pre-TAVR PCI in unselected coronary lesions was independently associated with a higher 6-month mortality rate. Pre-TAVR PCI of significant proximal coronary lesions, as recommended in the current guidelines, showed no association with mortality. Accordingly, we did not find an association between pre-TAVR PCI and mortality in the additional propensity-score matched cohort. Despite the absence of a beneficial effect of pre-TAVR PCI in patients with significant proximal lesions in our analysis, some beneficial effect of pre-TAVR PCI on an individual patient basis cannot be excluded. Therefore, our study supports the current clinical guidelines to only consider pre-TAVR PCI in proximal coronary lesions.

The percentage of TAVR patients with concurrent CAD and the percentage of patients treated with pre-TAVR PCI in our study was comparable with previous studies.^2,3 At our institution, patients are critically evaluated for the necessity of performing pre-TAVR PCI. Furthermore, this high-volume center was an early adopter of TAVR, starting its program in 2007, and was a forerunner in performing TAVR procedures in younger patients and patients with lower surgical risk. This has resulted in relatively low STS risk scores and EuroScore II scores in the included population compared with the aforementioned studies.

Kotronias et al and Lateef et al both performed a meta-analysis investigating the impact of pre-TAVR PCI on 30-day mortality after TAVR.^2,3 Both studies reported higher 30-day mortality in patients who underwent PCI, as well as a higher rate of major vascular complications. Lateef et al also reported higher 1-year mortality in patients who underwent PCI. The chance of treatment bias in these patient populations is high, and some included studies may have compared patients without CAD to patients who underwent pre-TAVR PCI for significant coronary lesions. Therefore, assumptions on the negative effect of PCI should not be made too easily.

Nevertheless, our study confirms that routine pre-TAVR PCI, tested in a population with any >50% DS coronary lesions, is associated with higher 6-month mortality. This association was mainly driven by the difference in mortality in the non-proximal lesions. This finding could also be the result of treatment bias, especially because the 6-month mortality in patients without PCI in non-proximal lesions is relatively low, indicating that patients without PCI treatment were relatively healthy compared with the patients who underwent PCI. Another explanation may be that performing a pre-TAVR PCI induces extra risks in this frail patient population, leading to the significantly increased procedural mortality in this subgroup. This negative association between pre-TAVR PCI and mortality was not present in the subgroups of patients with significant proximal coronary lesions and significant LM and proximal LAD lesions. The results of the propensity-score matched cohort were qualitatively similar to the non-matched analyses of these latter subgroups. Although we could not find a beneficial effect from pre-TAVR PCI, PCI may be considered on an individual patient basis. Future studies are needed to identify specific patients who may benefit from a pre-TAVR PCI strategy, such as a staged PCI treatment strategy in patients in whom anginal complaints persist despite increased coronary blood flow and vasodilator reserve directly after TAVR.^8,9

Impact on daily practice. The results of our study support the European guideline recommendations to consider PCI in patients with significant proximal coronary lesions. Although we cannot exclude a beneficial effect of pre-TAVR PCI in selected individual patients, we believe that most patients will not benefit from routine pre-TAVR PCI and that these invasive procedures inevitably carry more risks, especially in TAVR patients, who are generally older and more frail. Therefore, we recommend a restrictive but tailored case-by-case multidisciplinary evaluation to perform preprocedural PCI by the local heart team based on clinical symptoms, such as pharmacological refractory angina together with significant lesions in the proximal coronary arteries. Another consideration might be to perform pre-TAVR PCI in patients in whom coronary access is thought to be compromised after the procedure. A staged post-TAVR PCI strategy can always be considered in patients with persistent anginal complaints despite increased coronary blood flow and vasodilator reserve after TAVR.

Finally, if only the proximal coronary segments need evaluation before TAVR, a computed tomography coronary angiography may provide sufficient information in a large proportion of patients, thereby reducing the invasiveness of the TAVR work-up.

Study limitations. First, we retrospectively analyzed a single-center, non-randomized cohort, which is a design that is more prone to errors and bias in comparison with a multicenter, randomized, clinical trial. Despite the control for confounders by regression analysis and cohort matching to minimize the effect of treatment bias, residual confounding is possible due to the observational design of this study. Second, interpretation of the location and severity of coronary lesions by coronary angiography was performed by the attending cardiologists. The decision of whether to perform PCI and timing of PCI were at the discretion of the heart team, and we included the year of enrollment in this cohort in order to correct for changing guidelines over time. Because coronary angiography was performed by the attending cardiologist, the angiographic analysis was not always performed using quantitative analysis, but was left at the discretion of the attending cardiologist.

Routine pre-TAVR PCI is not associated with mortality reduction in TAVR patients with coronary lesions in any segment or in patients with proximal coronary lesions. Despite the lack of a beneficial effect of routine pre-TAVR PCI, we cannot exclude a beneficial effect in a selection of patients with proximal lesions. Therefore, we strongly support the current clinical guidelines to only consider pre-TAVR PCI in proximal coronary lesions, while advocating a restrictive pre-TAVR PCI strategy.