Impact of untreated chronic obstructive coronary artery disease on outcomes after transcatheter aortic valve replacement

Abstract

Background and Aims

In transcatheter aortic valve replacement (TAVR) recipients, the optimal management of concomitant chronic obstructive coronary artery disease (CAD) remains unknown. Some advocate for pre-TAVR percutaneous coronary intervention, while others manage it expectantly. The aim of this study was to assess the impact of varying degrees and extent of untreated chronic obstructive CAD on TAVR and longer-term outcomes.

Methods

The authors conducted a retrospective cohort study of TAVR recipients from January 2015 to November 2021, separating patients into stable non-obstructive or varying degrees of obstructive CAD. The major outcomes of interest were procedural all-cause mortality and complications, major adverse cardiovascular events, and post-TAVR unplanned coronary revascularization.

Results

Of the 1911 patients meeting inclusion, 75%, 6%, 10%, and 9% had non-obstructive, intermediate-risk, high-risk, and extreme-risk CAD, respectively. Procedural complication rates overall were low (death 0.4%, shock 0.1%, extracorporeal membrane oxygenation 0.1%), with no difference across groups. At a median follow-up of 21 months, rates of acute coronary syndrome and unplanned coronary revascularization were 0.7% and 0.5%, respectively, in the non-obstructive population, rising in incidence with increasing severity of CAD (P < .001 for acute coronary syndrome/unplanned coronary revascularization). Multivariable analysis did not yield a significantly greater risk of all-cause mortality or major adverse cardiovascular events across groups. One-year acute coronary syndrome and unplanned coronary revascularization rates in time-to-event analyses were significantly greater in the non-obstructive (98%) vs. obstructive (94%) subsets (P_log-rank< .001).

Conclusions

Transcatheter aortic valve replacement can be performed safely in patients with untreated chronic obstructive CAD, without portending higher procedural complication rates and with relatively low rates of unplanned coronary revascularization and acute coronary syndrome at 1 year.

Structured Graphical Abstract

(A) Definitions for the four categories of coronary artery disease (CAD) studied alongside the breakdown of transcatheter aortic valve replacement (TAVR) valve type with 91% balloon-expandable valves (BEV) and 9% self-expanding valves (SEV). Procedural complication rates were minimal across the 1911 patients studied. (B) Cumulative incidence curve of major adverse cardiovascular events (MACE) [acute coronary syndrome (ACS), stroke, or heart failure hospitalizations] demonstrating no significant difference between obstructive and non-obstructive groups (P = .609) (C) Cumulative incidence curve of unplanned coronary revascularization showing a significant difference between obstructive and non-obstructive groups (P = .006). (D) Multivariable analyses for all-cause mortality and MACE. CI, confidence interval; ECMO, extracorporeal membrane oxygenation; HR, hazard ratio; LAD, left anterior descending artery; LCX, left circumflex artery; LM, left main; N/A, not available; RCA, right coronary artery.

Coronary artery disease, Aortic stenosis, Transcatheter aortic valve replacement, Percutaneous coronary intervention, Revascularization

See the editorial comment for this article ‘Transfemoral aortic valve implantation and concomitant CAD: the jury is out’, by H. Eltchaninoff and E. Durand, https://doi.org/10.1093/eurheartj/ehae141.

Introduction

Obstructive coronary artery disease (CAD) and severe aortic stenosis (AS) frequently co-exist.^1,2 Despite the increasing utility of transcatheter aortic valve replacement (TAVR), there is still much debate about the timing, or even necessity, for addressing coexisting stable CAD in these patients. Although many patients harbour significant concomitant valvular and obstructive CAD, their anginal symptoms may be mild or absent, or masked by concomitant severe AS. This begs the question of whether chronic obstructive CAD should be systematically treated pre-TAVR (akin to the surgical approach of bypassing all visually obstructive lesions at the time of aortic valve replacement),^3,4 vs. deferring until the post-TAVR phase when it may be easier to assess for presence/absence of residual symptoms posed from the CAD, and trialling guideline-directed medical therapy prior to embarking upon percutaneous coronary intervention (PCI) for symptom relief.^5–12 Delaying PCI could potentially render it safer when performed in the setting of a normally functioning aortic valve. That being said, re-accessing the coronary vessels across a bioprosthetic valve in situ may pose challenges in certain circumstances.¹³

Some studies evaluating the impact of CAD on patients undergoing TAVR have reported no difference in all-cause mortality, peri-procedural complications, or long-term morbidity and mortality, while others have shown contradictory results.^5–12 Similarly, studies examining revascularization before, concomitantly, afterward, or not at all in CAD patients undergoing TAVR have not yielded clear recommendations.^14,15 The above uncertainty has resulted in individual operators developing their own patient-specific preferences. Our institutional practice shifted towards a more conservative approach for managing concomitant stable obstructive CAD in the vast majority of TAVR recipients favouring a TAVR-first approach among those with symptoms most likely attributable to valve disease alone, even in high-risk patients with complex multivessel obstructive CAD. The present analysis assessed the impact of concomitant untreated chronic obstructive epicardial CAD upon peri-procedural TAVR-related outcomes, as well as longer-term clinical outcomes related to the severity and extent of baseline CAD.

Methods

Study design

We retrospectively screened patients over the age of 18 years who underwent TAVR at the Cleveland Clinic from January 2015 through December 2021. Those with a prior PCI, and no pre-TAVR coronary angiography, were excluded. Figure 1 summarizes patient selection of the final cohort included in the analyses (n = 1911). The Cleveland Clinic institutional review board approved this study. The need for written informed consent was waived as this was a retrospective analysis.

Figure 1

Flowchart of study population. Patient selection based on inclusion and exclusion criteria. CAD group categories: (i) non-obstructive (RCA, LAD, and LCX lesions all < 70% severity and LM < 50%), (ii) intermediate-risk (LCX or RCA or non-proximal LAD lesions ≥ 70%), (iii) high-risk (LCX and LAD lesions each ≥ 70% or RCA and either LAD or LCX lesions of ≥70% or LAD ≥ 70% or proximal LAD ≥ 70% or LM lesions ≥ 50% but <70%), (iv) extreme-risk (LCX, LAD, and RCA ≥ 70% or LM ≥ 70%). *Mean difference between PCI and TAVR is 79.6 months. TAVR, transcatheter aortic valve replacement; PCI, percutaneous coronary intervention; CABG, coronary artery bypass grafting; CAD, coronary artery disease

Study population

Patients were categorized into four groups according to baseline CAD severity pre-TAVR, extracted from pre-TAVR coronary angiography reports. ‘Extreme-risk CAD’ was defined as: (i) left circumflex artery (LCX), left anterior descending artery (LAD), and right coronary artery (RCA) all with lesions ≥ 70% (obstructive triple vessel disease) or (ii) left main (LM) lesions ≥ 70%. ‘High-risk CAD’ was defined as: (i) presence of concomitant LCX and LAD lesions each ≥ 70% severity or (ii) concomitant RCA and either LAD or LCX lesions of ≥70% severity, or (iii) proximal LAD lesion ≥ 70%, or (iv) LM lesion ≥ 50% but <70%. The ‘intermediate-risk CAD’ group was defined as: (i) LCX lesion ≥ 70% or (ii) RCA lesion ≥ 70% or (iii) non-proximal LAD lesion ≥ 70%. The ‘non-obstructive CAD’ group was defined as: (i) RCA, LAD, and LCX lesions all ≤ 70% in severity and (ii) LM lesions ≤ 50% severity. Those presenting with an ACS pre-TAVR were excluded.

Clinical and echocardiographic variables

Patient characteristics including demographics, comorbidities, medications, pre-TAVR laboratory results, Society of Thoracic Surgeons (STS) score, and New York Heart Association (NYHA) class were extracted from electronic medical records. Echocardiographic variables including left ventricular ejection fraction (LVEF), transvalvular gradients, aortic valve areas, stroke volume indices, left ventricular end-diastolic volume (LVEDV), and left ventricular end-systolic volume (LVESV) were extracted from pre-TAVR echocardiographic reports.

Study outcomes

The primary outcome was defined as either all-cause death, or major adverse cardiovascular events (MACE), or post-TAVR unplanned coronary revascularization (percutaneous or surgical) during follow-up, whichever came first. Procedural-related all-cause mortality and complications [defined as intra-aortic balloon pump (IABP) insertion post-procedure, post-procedural extracorporeal membrane oxygenation (ECMO), rhythm disturbances requiring defibrillation, and new-onset cardiogenic shock] comprised a special procedure-related outcome. Major adverse cardiovascular events were defined as ACS, stroke, or heart failure hospitalization.

Statistical analyses

Baseline characteristics were compared between study groups using a two-sided Student’s t-test, ANOVA (for more than two groups), or Mann–Whitney U test for continuous variables and χ² test for categorical variables. Continuous variables are represented as mean (standard deviation) or median (interquartile range), and categorical variables are reported as proportions.

Incidence rate was reported for each type of event, and incidence rate difference was calculated to assess the absolute risk and absolute risk difference between two comparison groups (obstructive vs. non-obstructive CAD). It is known that patients undergoing TAVR are elderly and experience competing risks of non-cardiac events. To appropriately account for competing risks of developing MACE, unplanned coronary revascularization, or death from any cause, we used competing risk analytical framework. Cumulative incidence functions were estimated, and curves were plotted using the Aalen-Johansen estimator, accounting for the respective other event types as competing risks.

We utilized cause-specific multivariable Cox proportional-hazard models, adjusted for age, gender, LVEF, peripheral artery disease, history of hypertension, creatinine, and low-density lipoprotein (LDL) levels.

Results

Baseline characteristics

From a total of 2952 patients who underwent TAVR between January 2015 and December 2021, 1911 patients with stable CAD were followed for a median time of 21 months (IQR: 10–35.2 months; 30% were lost to follow-up after one year) post-TAVR, among which 164 (9%) had extreme-risk obstructive CAD, 199 (10%) had high-risk obstructive CAD, 116 (6%) had intermediate-risk obstructive CAD, and 1432 (75%) had non-obstructive CAD (Figure 1). Table 1 summarizes characteristics of patients included in the study overall and according to the baseline extent and severity of CAD pre-TAVR. There were significant differences in age and sex across the CAD groups, with the highest percentage of men in the extreme-risk obstructive CAD group, along with this group harbouring a greater extent of comorbidities. Medical therapies also tended to be more prevalent in the higher risk CAD groups. LDL cholesterol (LDL-C) levels were lowest in the extreme-risk CAD group, which was in line with their more frequent statin use. The LVEF tended to be lowest in the extreme-risk CAD group, along with stroke volume index, while LV volumes were globally highest in this group.

Table 1

Pre-transcatheter aortic valve replacement patient characteristics

Variables	Total population (n = 1911)	Non-obstructive CAD (n = 1432)	Intermediate-risk CAD (n = 116)	High-risk CAD (n = 199)	Extreme-risk CAD (n = 164)	P-value
Clinical parameters
Age, years (mean (SD))	78.9 (9.6)	78.51 (9.86)	79.95 (9.07)	80.70 (8.49)	79.89 (8.09)	.006
Male sex, n (%)	1030 (53.9)	685 (47.8)	77 (66.4)	141 (70.9)	127 (77.4)	<.001
BMI, kg/m2 (mean (SD))	28.9 (7.56)	29.22 (7.96)	28.21 (7.21)	27.84 (5.44)	28.92 (6.04)	.065
STS risk score (mean (SD))	5.4 (4.0)	5.1 (3.7)	5.4 (3.2)	6.7 (5.5)	6.4 (3.9)	<.001
NYHA class, n (%)						.04
I	35 (1.8)	25 (1.7)	6 (5.5)	0 (0)	4 (2.2)
II	478 (25)	372 (26)	26 (22)	46 (23)	39 (24)
III	1337 (70)	1002 (70)	81 (70)	141 (71)	118 (72)
IV	61 (3.2)	33 (2.3)	3 (2.2)	12 (6)	3 (1.8)
Past medical history, n (%)
Prior MI	244 (13.0)	101 (7.2)	23 (20.0)	59 (30.1)	61 (37.7)	<.001
Atrial fibrillation	739 (39.3)	529 (37.6)	41 (35.7)	95 (48.5)	74 (45.7)	.007
Hypertension	1648 (87.7)	1210 (85.9)	104 (90.4)	184 (93.9)	150 (93.2)	.001
Diabetes	626 (33.3)	450 (32.0)	34 (29.6)	74 (37.8)	68 (42.2)	.024
Smokers	258 (13.7)	199 (14.2)	24 (20.9)	17 (8.7)	18 (11.2)	.016
Peripheral artery disease	1222 (65.0)	883 (62.7)	80 (69.6)	148 (75.5)	111 (68.9)	.002
Medications, n
ACE inhibitors	860 (45.1)	618 (43.2)	55 (47.8)	101 (51.3)	86 (52.8)	.026
Aldosterone antagonist	236 (12.4)	183 (12.8)	10 (8.7)	27 (13.7)	16 (9.8)	.340
Anticoagulants	1161 (60.9)	860 (60.1)	65 (56.5)	128 (65.0)	108 (66.3)	.20
Aspirin	1562 (82.0)	1126 (78.7)	103 (89.6)	183 (92.9)	150 (92.0)	<.001
Beta-blockers	1372 (72.0)	988 (69.1)	81 (70.4)	163 (82.7)	140 (85.9)	<.001
Calcium channel blockers	862 (45.2)	655 (45.8)	48 (41.7)	93 (47.2)	66 (40.5)	.47
P2Y12 inhibitors	406 (21.3)	217 (15.2)	41 (35.7)	94 (47.7)	54 (33.1)	<.001
Diuretics	1033 (54.2)	774 (54.1)	59 (51.3)	118 (59.9)	82 (50.3)	.27
Statins	1443 (75.7)	1007 (70.4)	106 (92.2)	179 (90.9)	151 (92.6)	<.001
Labs, mean (SD)						.01
Haemoglobin (g/dL)	11.6 (1.7)	11.6 (1.78)	11.7 (1.57)	11.9 (1.77)	12.00 (1.65)
Creatinine (mg/dL)	1.26 (1.4)	1.21 (1.38)	1.16 (1.11)	1.22 (1.07)	1.07 (0.70)	.63
HbA1c (%)	6.11 (1.0)	6.09 (1.02)	5.97 (0.83)	6.21 (0.98)	6.32 (1.00)	.14
HDL cholesterol (mg/dL)	49.7 (16.8)	54.1 (18.73)	51.3 (16.57)	48.9 (15.75)	44.4 (13.23)	<.001
LDL cholesterol (mg/dL)	80.4 (33.0)	84.9 (33.73)	79.8 (31.06)	79.1 (30.35)	76.1 (30.43)	.023
Triglycerides (mg/dL)	108.2 (68.4)	109.3 (67.81)	104.8 (50.79)	104.9 (52.06)	115.7 (53.14)	.58
Echocardiography
LVEF, mean (SD)	57.8 (11.4)	58.84 (10.77)	56.36 (11.56)	55.25 (12.64)	53.29 (12.99)	<.001

Data are represented as mean (SD) or n (%).

Bold is meant to highlight significant values below a threshold of p = .05.

BMI, body mass index; MI, myocardial infarction; HDL, high-density lipoprotein; LDL, low-density lipoprotein; LVEF, left ventricular ejection fraction; HbA1c, glycated haemoglobin.

Table 2 describes procedural details of the TAVR recipients overall and according to the extent and severity of underlying CAD. The overall population received predominantly a balloon-expandable valve (91%) via the transfemoral approach (95%).

Table 2

Transcatheter aortic valve replacement procedural details

Variables	Total population (n = 1911)	Non-obstructive CAD (n = 1432)	Intermediate-risk CAD (n = 116)	High-risk CAD (n = 199)	Extreme-risk CAD (n = 164)	P-value
Valve type						.054
Edwards Sapien 3	1734 (91.0)	1294 (90.4)	100 (86.1)	181 (91.0)	159 (97.0)
Other	177 (9.0)	138 (9.6)	16 (13.9)	18 (9.0)	5 (3.0)
Valve size						<.001
20 mm	82 (4.0)	65 (4.5)	7 (6.0)	9 (4.5)	1 (0.6)
23 mm	634 (33.0)	518 (36.2)	26 (22.4)	52 (26.1)	38 (23.2)
26 mm	661 (35.0)	474 (33.1)	45 (38.8)	77 (38.7)	65 (39.6)
29 mm	439 (23.0)	305 (21.3)	30 (25.9)	49 (24.6)	55 (33.5)
Other	98 (5.0)	70 (4.9)	8 (6.9)	15 (7.5)	5 (3.1)
Access route						<.001
Transfemoral	1807 (95.0)	1372 (95.8)	108 (93.1)	180 (90.4)	147 (89.6)
Other	104 (5.0)	60 (4.2)	8 (6.9)	19 (9.6)	17 (9.4)

Data are represented as n (%).

Clinical outcomes

Table 3 describes the unadjusted rates of procedural and longer-term clinical outcomes of the differing populations according to underlying CAD extent and severity. There were no significant differences in peri-procedural death rates across the groups stratified by CAD extent/severity. Rates of peri- or post-procedural complications including shock, ventricular arrhythmias requiring defibrillation, IABP insertion, and ECMO use also did not differ significantly across groups (P = .60).

Table 3

Unadjusted post-transcatheter aortic valve replacement outcomes

Variables	Total population (n = 1911)	Non-obstructive CAD (n = 1432)	Intermediate-risk CAD (n = 116)	High-risk CAD (n = 199)	Extreme-risk CAD (n = 164)	P-value
Procedural complications, n (%)	13 (0.1)	9 (0.1)	0 (0.0)	1 (0.5)	3 (1.8)	.60
IABP	1 (0.1)	1 (0.1)	0 (0.0)	0 (0.0)	0 (0.0)
Ventricular arrhythmia requiring shock/ICD	2 (0.1)	2 (0.1)	0 (0.0)	0 (0.0)	0 (0.0)
ECMO	2 (0.1)	1 (0.1)	0 (0.0)	0 (0.0)	1 (0.6)
Shock	1 (0.1)	0 (0.0)	0 (0.0)	0 (0.0)	1 (0.6)
Peri-procedural death	7 (0.4)	5 (0.4)	0 (0.0)	1 (0.5)	1 (0.6)
ICU, hours (mean (SD))	24.56 (20.0)	23.86 (19.1)	25.23 (19.6)	26.74 (22.3)	27.55 (24.1)	.051
Major bleeding, n (%)	38 (2.0)	28 (2.0)	2 (1.7)	5 (2.5)	3 (1.8)	.95
Overall ACS, n (%)	22 (1.2)	10 (0.7)	2 (1.7)	8 (4.0)	2 (1.2)	<.001
UA	10 (0.5)	2 (0.1)	2 (1.7)	4 (2.0)	2 (1.2)
NSTEMI	8 (0.4)	6 (0.4)	0 (0.0)	2 (1.0)	0 (0.0)
STEMI	4 (0.2)	2 (0.1)	0 (0.0)	2 (1.0)	0 (0.0)
Overall unplanned coronary revascularization, n (%)	29 (1.5)	14 (1.0)	3 (2.6)	8 (4.0)	4 (2.4)	.006
CABG	9 (0.5)	7 (0.5)	0 (0.0)	1 (0.5)	1 (0.6)
PCI	20 (1.0)	7 (0.5)	3 (2.6)	7 (3.5)	3 (1.8)

Data are represented as mean (SD) or n (%).

ACS, acute coronary syndrome; UA, unstable angina; NSTEMI, non-ST-elevation myocardial infarction; STEMI, ST-elevation myocardial infarction; CABG, coronary artery bypass graft; PCI, percutaneous coronary intervention; IABP, intra-aortic balloon pump; ICU, intensive care unit; ICD, implantable cardioverter defibrillator; ECMO, extracorporeal membrane oxygenation; Shock, defined as new-onset cardiogenic shock as a result of TAVR.

During a median of 1.32 years at risk, the composite outcome was observed in a total of 737 patients (207 in obstructive CAD and 530 in non-obstructive CAD). The incidence rate of the primary composite outcome was similar in both comparison groups: in the non-obstructive CAD group 227 per 1000 person-years of follow-up (95% CI 208–247), and in the obstructive CAD group 240 per 1000 person-years of follow-up (95% CI 210–276). Incidence rate difference was 0.01 (95% CI −0.02 to 0.05); P = .47. There was no significant difference in the incidence rate of all-cause death between groups: 68.1 per 1000 person-years of follow-up in non-obstructive CAD (95% CI 58.3–79.6) and 82.6 per 1000 person-years of follow-up (95% CI 65.5–104.3) in the obstructive CAD group. Incidence rate difference for all-cause death was 0.01 (95% CI −0.007 to 0.04); P = .18. Notably, the incidence rate of competing MACE outcome was similar between groups: 153 per 1000 person-years of follow-up (95% CI 138–170) in the non-obstructive CAD group, as compared to 142 per 1000 person-years of follow-up (95% CI 119–170) in obstructive CAD group. Incidence rate difference was—0.01 (95% CI −0.04 to 0.02); P = .48. The incidence rate of competing unplanned coronary revascularization was higher in obstructive CAD group [16.3 per 1000 person-years of follow-up (95% CI 9.6–27.5)] than in non-obstructive CAD group [(6.0 per 1000 person-years of follow-up (95% CI 3.4–10.1)]. The incidence rate difference for unplanned coronary revascularization was 0.01 (95% CI 0.001–0.02); P = .010.

The overall rates of ACS were higher in those patients with obstructive CAD (compared with the non-obstructive CAD group, P < .001), however, ACS rates were not stepwise incrementally higher across the intermediate, high, and extreme-risk CAD populations (1.7%, 4%, and 1.2%, respectively). The overall coronary revascularization rate was 1.5%, with 1% of the non-obstructive CAD group undergoing revascularization, compared with rates of 2.6%, 4%, and 2.4% of the intermediate, high, and extreme-risk CAD patients, respectively (P = .006). Figures 2 and 3 show cumulative incidence curves of all-cause death and MACE, respectively, outlining no significant differences between the obstructive and non-obstructive disease groups (P= .62 and .61, respectively).

Figure 2

Cumulative incidence of all-cause death. Cumulative incidence of all-cause death over total follow-up showing no significant difference between obstructive and non-obstructive groups (P = .615). CAD, coronary artery disease

Figure 3

Cumulative incidence of MACE. Cumulative incidence of MACE (ACS, stroke, or heart failure hospitalizations) showing no significant difference (P = .609). CAD, coronary artery disease; MACE, major adverse cardiovascular events

Supplementary data online, Table S1 assesses unadjusted procedural and longer-term outcomes further stratified by baseline LVEF. These data did not unravel any relationships between both the severity/extent of obstructive CAD and the degree of baseline LV function with procedural events.

Supplementary data online, Table S2 describes the specific management breakdown of ACS patients. There were no patients in whom coronary angiography was aborted or PCI not completed due to the presence of a transcatheter aortic valve prosthesis posing technical difficulties for guide catheter engagement.

Table 4 describes univariable and multivariable analyses evaluating all-cause mortality, MACE, and unplanned coronary revascularization. Compared with the non-obstructive CAD group, increasing severity/extent of baseline untreated obstructive CAD did not associate with all-cause mortality or MACE. Univariable analyses point to the high-risk obstructive CAD group demonstrating a significantly greater risk of unplanned coronary revascularization (HR 4.07, 95% CI 1.71–9.73; P = .002) (Figure 4). Due to the limited number of unplanned coronary revascularization procedures, multivariable adjustment was not plausible.

Figure 4

Cumulative incidence of unplanned coronary revascularization. Cumulative incidence of unplanned coronary revascularization showing a significant difference between obstructive and non-obstructive groups (P = .006). CAD, coronary artery disease

Table 4

Univariate and multivariate-adjusted Cox proportional-hazard regression of all-cause mortality, major adverse cardiovascular events, and unplanned coronary revascularization

Group	Unadjusted		Multivariable adjusteda
	HR (95% CI)	P-value	HR (95% CI)	P-value
All-cause mortality
Non-obstructive CAD	Ref	N/A	Ref	N/A
Intermediate-risk CAD	1.32 (0.89–1.94)	.17	1.17 (0.70–1.95)	.55
High-risk CAD	1.17 (0.83–1.65)	.36	0.93 (0.57–1.51)	.77
Extreme-risk CAD	0.80 (0.53–1.20)	.28	0.74 (0.45–1.23)	.25
Major adverse cardiovascular events
Non-obstructive CAD	Ref	N/A	Ref	N/A
Intermediate-risk CAD	1.03 (0.71–1.49)	.87	1.05 (0.68–1.63)	.83
High-risk CAD	1.20 (0.91–1.56)	.19	1.07 (0.77–1.48)	.69
Extreme-risk CAD	0.79 (0.57–1.10)	.16	0.76 (0.52–1.11)	.16
Unplanned coronary revascularization
Non-obstructive CAD	Ref	N/A	N/A	N/A
Intermediate-risk CAD	2.42 (0.69–8.44)	.17	N/A	N/A
High-risk CAD	4.07 (1.71–9.73)	.002	N/A	N/A
Extreme-risk CAD	1.63 (0.47–5.67)	.45	N/A	N/A

CAD, coronary artery disease; CI, confidence interval; HR, hazard ratio; N/A, not available.

^aAdjusted for age, sex, baseline left ventricular ejection fraction, peripheral artery disease, hypertension, creatinine, and baseline low-density lipoprotein cholesterol levels.

Supplementary data online, Figures S1 and S2 outline Kaplan–Meier survival curves at 1 year for freedom from MACE (ACS, heart failure hospitalization, or stroke) and unplanned coronary revascularization or ACS, respectively. These show no significant 1-year MACE differences between the CAD groups (P_log-rank= .91), whereas significant separation of curves occurs around the 8–12 week period post-TAVR in terms of unplanned coronary revascularization/ACS in those with obstructive disease (irrespective of baseline severity/extent) vs. those without underlying obstructive lesions (absolute differences between the non-obstructive CAD group and the obstructive CAD was 4%–5% at 1 year) (P_log-rank< .001). Supplementary data online, Figure S3 shows this effect to remain stable out to 3-year follow-up (P_log-rank< .001).

Sensitivity analyses were performed to better understand any potential confounding by the exclusion of 735 patients with pre-TAVR PCI. Supplementary data online, Tables S3–S5 summarize these data. Time from PCI to TAVR varied between risk groups (with only 13% having their PCI within 3 months pre-TAVR). The overall mean time between pre-TAVR PCI to the TAVR procedure was 79.6 months, with 72% of these patients having their PCI ≥ 1-year pre-TAVR. These 735 excluded patients with PCI pre-TAVR had significantly higher rates of severe disease than the included patient population (14% vs. 10% and 11% vs. 9% in high-risk and extreme-risk CAD cohorts, respectively, P < .001). However, there was no difference in ACS events between risk groups during follow-up post-TAVR. Of the 13% of patients who underwent PCI within 3 months pre-TAVR, 97% had CCS class 2–4 angina pre-TAVR. Of these 96 patients, 16 harboured ostial LM or ostial RCA lesions.

Discussion

The present analysis is unique in demonstrating the procedural safety of TAVR in a large number of recipients with varying degrees of untreated stable obstructive CAD. We found a very low rate of procedural TAVR-related complications that overall did not significantly differ across those with non-obstructive or obstructive CAD, as well as relatively low rates of ACS or unplanned coronary revascularization at 1 year. There was no significant difference in cumulative incidence of all-cause death and MACE over the total follow-up period between the obstructive and non-obstructive CAD patient groups. Furthermore, adjusted analyses did not reveal a correlation between the extent/severity of untreated obstructive CAD and all-cause mortality or MACE in the medium term. Notably, the presence of a transcatheter aortic bioprosthesis did not impede downstream coronary angiography or PCI when indicated with no noted difficulties or complications regarding coronary cannulation during left heart catheterization post-TAVR (Structured Graphical Abstract). These findings support the premise that stable, obstructive CAD, regardless of its severity/extent and pre-existing left ventricular function, can be initially managed medically in TAVR candidates. Future randomized trials are eagerly anticipated to further illuminate the optimal clinical strategy for this common clinical conundrum.

Peri-procedural safety

Although TAVR harbours the typical risks associated with other transcatheter cardiac procedures, it also has unique steps that could pose a risk to those with underlying untreated CAD. During valve deployment, patients often undergo a period of rapid ventricular pacing to temporarily lower cardiac output in order to secure proper device positioning. While there is a potential risk of decreasing myocardial oxygen delivery due to the drop in coronary perfusion during rapid pacing,¹⁶ we showed no significant difference in procedural outcomes such as shock requiring mechanical circulatory support (IABP or ECMO), rhythm disturbances requiring defibrillation, or intensive care unit length of stay. Additionally, complications were rare across the study population, with only one patient developing a brief period of new-onset cardiogenic shock following TAVR. Moreover, ∼10% of cases in both the high and extreme-risk cohorts involved alternate access TAVR using general anaesthesia. Despite the increase in cardiovascular risk associated with general anaesthesia, these cases were still performed safely in spite of significant obstructive CAD. Furthermore, a sensitivity analysis of subgroups with varying LVEF cut-offs did not demonstrate differing rates of procedural complications according to LVEF; this is in line with previous studies.¹⁷

Revascularization timing in transcatheter aortic valve replacement

The current data landscape demonstrates heterogeneity with regard to outcomes following coronary revascularization pre-TAVR, with most studies reporting no difference in short-term prognosis but variable results on longer-term follow-up. A 2017 meta-analysis of 3858 participants included patients with CAD who did or did not undergo revascularization pre-TAVR. Investigators found no difference in either 30-day cardiovascular mortality or 1-year overall mortality. Moreover, patients who underwent PCI pre-TAVR demonstrated significantly greater major vascular complication rates and 30-day mortality, which may ultimately be attributed to having undergone an additional percutaneous procedure carrying its own inherent risks and/or the need for dual antiplatelet therapy post-PCI.¹⁰ Similarly, Sankaramangalam et al. undertook a meta-analysis of 8013 patients, demonstrating no difference between CAD and non-CAD patients undergoing TAVR in terms of 30-day all-cause mortality. Despite these promising short-term results, there was a significant increase in 1-year all-cause mortality (OR 1.21; 95% CI 1.07–1.36; P = .002).¹⁴

The ACTIVATION (PercutAneous Coronary inTervention prIor to transcatheter aortic VAlve implantaTION) trial randomized patients with Canadian Cardiovascular Society (CCS) angina class 2 or less to either undergo PCI (circa 20% bare-metal stent use) or not pre-TAVR. Pre-TAVR PCI was found to offer no benefit in terms of mortality or rehospitalization rates at 1 year but did lead to higher rates of bleeding complications.¹⁵ Similarly, the recently published REVASC-TAVI registry included 1603 patients with significant, stable CAD and divided these patients into those who had PCI before (n = 1052), after (n = 157), or concomitantly (n = 394) with TAVR. All-cause death and the composite endpoint [all-cause death, stroke, myocardial infarction (MI), or heart failure rehospitalization] at 2 years were significantly lower in the cohort who underwent PCI post-TAVR.¹⁸ While these emerging data further support a ‘TAVR-first approach,’ the present study expands upon these findings by including patients with much greater extent of obstructive CAD across broad LVEF categories, while evaluating additional outcomes such as peri-procedural stability and post-TAVR ACS/unplanned revascularization rates over a longer follow-up period.

Effect of coronary artery disease severity

Khawaja et al.⁵ demonstrated that worsening CAD severity, as measured by higher SYNTAX scores, did portend worse outcomes in TAVR recipients. On the other hand, Puymirat et al.¹⁹ studied the impact of CAD on TAVR recipients from the FRANCE 2 registry noting that only significant LAD lesions were associated with increased 3-year mortality (HR 1.42, 95% CI 1.10–1.87), but neither the presence nor the extent of CAD was associated with mortality at 3 years. The present study further supports these findings with an expanded, more granular definition of disease severity and data that signal short-term peri-procedural safety.

In light of the variety of practice differences, the European Association of Percutaneous Cardiovascular Interventions released a 2023 statement on managing patients with CAD undergoing TAVR. One of the major consensus points was that PCI should be performed pre-TAVR in those who have severe CAD (defined as stenosis > 50% in LM or >70% in all other coronary arteries) only within proximal segments and particularly in those presenting with ACS or symptomatic angina.²⁰ Our results challenge these recent recommendations and suggest that in stable, obstructive CAD, even in very high-risk disease substrates, PCI pre-TAVR is not necessary for mitigating TAVR procedural risk, with unplanned revascularization and ACS rates remaining relatively low at 1 year, without significant effect upon MACE. That being said, other considerations such as valve choice may come into play, which could influence an operator’s preference for the need and/or timing of revascularization. The present study does not however mandate that all patients with obstructive CAD should undergo TAVR first, but rather it can be performed with no increased risk. In those who have significant burden of disease and more severe anginal symptoms that seem more attributable to coronary rather than valvular disease, a PCI-first strategy is a viable option as evidenced by the 735 patients excluded from the original analysis (13% of whom had PCI within 3 months pre-TAVR).

We defined CAD according to angiographic burden and severity. The occurrence of ACS was rare occurring in only 1.2% of patients [10 unstable angina (UA), 8 non-ST-elevation myocardial infarction (NSTEMI), and 4 ST-elevation myocardial infarction (STEMI) cases] across the entire stable CAD study cohort. The distribution of ACS events across the obstructive CAD severity subgroups was somewhat heterogeneous. A greater proportion of NSTEMI cases occurred in the non-obstructive CAD and high-risk CAD groups, while the majority of UA and STEMI cases were found in the high-risk and extreme-risk groups. In terms of managing these patients, there was a relatively small portion of the overall study population that ultimately required PCI or coronary artery bypass grafting (CABG) (1% and 0.5%, respectively). Within the NSTEMI cohort, only 1 of the 8 cases underwent left heart catheterization but did not receive PCI due to the diffuse extent of disease discovered, while the remaining 7 cases were managed medically (see Supplementary data online, Table S2 and Supplementary data online, Figure S4).

It is difficult to ascertain the attributable risk of the TAVR procedure per se in those with concomitant, stable, high-risk obstructive CAD. In the 5-year follow-up of the MASS study, Lopes et al. studied patients with either single, two-, or three-vessel disease (TVD) stratified to either PCI, CABG, or medical treatment. In the medical treatment TVD cohort, they reported 88% and 65% event-point cumulative rates of death, MI, or refractory angina needing revascularization at 12- and 60-month follow-up, respectively.²¹ We demonstrate similar ACS or unplanned coronary revascularization rates in our high- and extreme-risk cohorts that harboured either obstructive TVD or severe LM disease. Circumstantially, it therefore appears that the TAVR procedure itself did not significantly contribute to higher rates of ACS, death, or revascularization in our study population. Rather, the coronary event rates were typical of the natural history and progression of the background burden of CAD. Closer inspection of the time-to-event curves highlights a clustering of ACS/unplanned coronary revascularization procedures ∼8–12 weeks post-TAVR, without any major ongoing curve separation thereafter. This highlights the importance of early close surveillance and intense secondary preventive measures in those harbouring residual obstructive CAD post-TAVR, with perhaps a low threshold for repeat cardiac catheterization/coronary revascularization should there be residual symptoms despite successful TAVR.

Limitations

Several limitations of the present analysis warrant further consideration. The present study was retrospective from a single high-volume centre with predominantly BEV use. There is potential for differences in root flow characteristics between short stent frame balloon vs. tall stent-frame SEV that could, in theory, have confounded our findings. We excluded a total of 735 and 279 patients who underwent PCI and CABG, respectively, prior to undergoing the TAVR procedure. Of the 735 patients excluded, 13% underwent PCI within 3 months pre-TAVR. In this 13% subgroup, 97% had CCS angina scores of 2–4. This is representative of the cohort we sought to exclude given these patients were not stable from an angina perspective. Most of the pre-TAVR PCI cases were performed because of unstable angina/angina (60%), with 16 patients harbouring ostial LM or RCA lesions. However, despite the excluded subset of PCI patients harbouring significantly higher rates of severe disease than the included patient population, there was no significant difference in post-TAVR ACS events. Symptoms attributable to CAD vs. AS are often hard to truly delineate, and randomized trials have not shown the presence of ischaemia to bear major prognostic effects in stable CAD.²² Our method of assessing risk categories was based on the vessel affected and degree of occlusion—representative of clinical workflow. We recognize that there are more comprehensive measures of lesion assessment such as the SYNTAX score, which was not routinely recorded in our patient population. Despite this, our extreme-risk category took into account patients with severe proximal disease. Given the large catchment area of our quaternary referral centre, there could have been instances of inconsistent follow-up as patients typically return to their local institutions for more routine care post-TAVR. This could have resulted in an underestimation of ACS or unplanned coronary revascularization events captured post-TAVR.

To confirm the present findings, the Staged Complete Revascularization for Coronary Artery Disease vs Medical Management Alone in Patients With AS Undergoing Transcatheter Aortic Valve Replacement (COMPLETE TAVR; NCT04634240) randomized trial will shed further light on whether systematic PCI in the post-TAVR setting of angiographically severe lesions in vessels ≥ 2.5 mm will influence cardiovascular death, new MI, ischaemia-driven revascularization, hospitalization for UA, or heart failure at long-term follow-up compared with optimal medical management.

Conclusion

While there are patients with more dominant coronary symptoms or unstable coronary syndromes that could be considered for PCI pre-TAVR, the present study suggests that untreated stable obstructive CAD, even high-risk anatomic substrates, poses no meaningful added risk to TAVR procedural outcomes. Rates of ACS and coronary revascularization rates were relatively low at 1 year, ultimately mirroring the known natural history of native CAD burden, suggesting that underlying obstructive stable CAD can be managed according to current CAD guidelines. While these data support a TAVR-first approach in patients with stable obstructive CAD, valve choice and anatomical factors (i.e. ostial LM or RCA lesions) may still influence an operator’s decision on the need and/or timing for PCI pre-TAVR. Further trials are required to assess the preferred treatment strategy for patients with both significant coronary and aortic valve disease.

Supplementary data

Supplementary data are available at European Heart Journal online.

Declarations

Disclosure of Interest

All authors declare no disclosure of interest for this contribution.

Data Availability

The data underlying this article are available in the article and in its online supplementary material.

Funding

All authors declare no funding for this contribution.

Ethical Approval

The Cleveland Clinic Institutional Review Board approved this study. Given the retrospective design, the need for written informed consent was waived.

Pre-registered Clinical Trial Number

None supplied.