Coronary Hemodynamics in Patients With Severe Aortic Stenosis and Coronary Artery Disease Undergoing Transcatheter Aortic Valve Replacement

Objectives In this study, a systematic analysis was conducted of phasic intracoronary pressure and flow velocity in patients with severe aortic stenosis (AS) and coronary artery disease, undergoing transcatheter aortic valve replacement (TAVR), to determine how AS affects: 1) phasic coronary flow; 2) hyperemic coronary flow; and 3) the most common clinically used indices of coronary stenosis severity, instantaneous wave-free ratio and fractional flow reserve. Background A significant proportion of patients with severe aortic stenosis (AS) have concomitant coronary artery disease. The effect of the valve on coronary pressure, flow, and the established invasive clinical indices of stenosis severity have not been studied. Methods Twenty-eight patients (30 lesions, 50.0% men, mean age 82.1 ± 6.5 years) with severe AS and coronary artery disease were included. Intracoronary pressure and flow assessments were performed at rest and during hyperemia immediately before and after TAVR. Results Flow during the wave-free period of diastole did not change post-TAVR (29.78 ± 14.9 cm/s vs. 30.81 ± 19.6 cm/s; p = 0.64). Whole-cycle hyperemic flow increased significantly post-TAVR (33.44 ± 13.4 cm/s pre-TAVR vs. 40.33 ± 17.4 cm/s post-TAVR; p = 0.006); this was secondary to significant increases in systolic hyperemic flow post-TAVR (27.67 ± 12.1 cm/s pre-TAVR vs. 34.15 ± 17.5 cm/s post-TAVR; p = 0.02). Instantaneous wave-free ratio values did not change post-TAVR (0.88 ± 0.09 pre-TAVR vs. 0.88 ± 0.09 post-TAVR; p = 0.73), whereas fractional flow reserve decreased significantly post-TAVR (0.87 ± 0.08 pre-TAVR vs. 0.85 ± 0.09 post-TAVR; p = 0.001). Conclusions Systolic and hyperemic coronary flow increased significantly post-TAVR; consequently, hyperemic indices that include systole underestimated coronary stenosis severity in patients with severe AS. Flow during the wave-free period of diastole did not change post-TAVR, suggesting that indices calculated during this period are not vulnerable to the confounding effect of the stenotic aortic valve.

A significant proportion of patients with severe aortic stenosis (AS) have concomitant coronary artery disease (CAD) (1,2). Determining the significance of CAD is challenging because traditional noninvasive and invasive indices of ischemia have not been validated in this setting (3). At present the decision to revascularize a coronary lesion in a patient with severe AS is based on angiography (3). This anatomic approach is unlikely to correctly identify those lesions that are truly flow limiting and may therefore lead to inappropriate treatment decisions (4).
Invasive indices of coronary stenosis severity provide more accurate localization of ischemia than noninvasive indices (5). There are several invasive indices of coronary artery stenosis severity. These are measured either during resting or hyperemic conditions and can be further divided into those that use the complete cardiac cycle (Pd/ Pa, fractional flow reserve [FFR] [5]) or only a period within diastole (instantaneous wave-free ratio [iFR] [6]).
To validate whether an invasive index is accurate in determining lesion significance, in patients with severe AS, an understanding of how AS affects coronary flow is required. Transcatheter aortic valve replacement (TAVR) permits unique insights into the acute effects of AS on coronary physiology (7) because intracoronary physiology assessment can be made immediately before and after valve insertion, thereby minimizing any potential confounding factors. In this study, we aimed to use the TAVR model to determine how AS affects 1) phasic coronary flow; 2) hyperemic coronary flow; and 3) the most common clinically used indices of coronary stenosis severity, iFR and FFR.

METHODS
PATIENT POPULATION. Twenty-eight consecutive patients (30 lesions) with severe AS planned for TAVR and moderate to severe CAD were included. Recruiting centers were the Hammersmith Hospital, Imperial College NHS Trust (London, United Kingdom) and Skane University Hospital (Lund, Sweden). TAVR was indicated by international guidelines (3), and the treatment decision was made at a heart team meeting. Exclusion criteria were known nonviable myocardium in the area of the corresponding coronary artery being studied, contraindication to the administration of adenosine, inability to consent, and weight more than 200 kg. All participants gave written informed consent, and the study was given full ethical approval (14/SC/1103).   Coronary flow velocity (centimeters per second) was measured at baseline and during hyperemia.
Definitions of hemodynamic variables were as follows: (5) iFR ¼ Pd wfp /Pa wfp (6) Flow during the wave-free period of diastole (9) Systolic resistance ¼ Pd systole /v systole Basal stenosis resistance ¼ DP b /v b (10) Hyperemic stenosis resistance ¼ DP h /v h (11) where Pa is mean aortic pressure; Pd is mean intracoronary pressure distal to a stenosis; wfp is the wave-free period of diastole; v h is mean flow velocity distal to a stenosis during hyperemia; v b is mean flow velocity distal to a stenosis at baseline; DP h is Pa À Pd during hyperemia; and DP b is Pa À Pd at baseline.
Phasic analysis was performed to identify pressure and flow characteristics during different periods of the cardiac cycle. The wave-free period was identified using wave-intensity analysis as previously described (12). A custom-written MATLAB algorithm was used to separate systole, diastole, and the wave-free period to facilitate phasic analysis of hemodynamic data. A schematic outlining how this was performed is shown in Figure 2.
STATISTICAL ANALYSIS. Continuous variables are presented as mean AE SD unless otherwise stated.
Comparisons before and after TAVR were performed using a Wilcoxon signed rank test. The threshold for statistical significance was set at 0.05.   Table 3).

RESULTS
Following TAVR, 15 patients (53.6%) had no paravalvular leak, 13 patients (46.4%) had trivial to mild paravalvular leak, and no patients had mild to moderate, moderate, or severe paravalvular leak ( Table 3).  Table 4. An example of    Values are mean AE SD or n (%). Values are n or mean AE SD.

DISCUSSION
In this study, we have shown that 1) iFR-flow does  Although these are both pressure-derived indices of    Ahmad et al.  There is a paucity of available data regarding coronary stenosis assessment in patients with severe AS. Existing studies have not measured coronary flow and assumed that it is not affected by  Our results cannot therefore be generalized to patients with more mild degrees of AS.
Adenosine was administered as an intracoronary bolus and not via intravenous infusion. We cannot therefore exclude the possibility that intravenous adenosine infusion would yield different results.
However, intracoronary adenosine is recognized as a valid approach to FFR assessment (31,32), and such assessments have been included in all the large randomized trials of physiology to date (33,34).
Intravenous infusion was avoided because of the recognized potential for a 15% reduction in aortic pressure (35) that could potentially destabilize a patient with severe AS.
Post-TAVR physiological measurements were made immediately after the valve had been replaced and within the same catheter laboratory procedure.
We cannot therefore comment on any more long-term changes in coronary hemodynamics.
The prevalence of severe aortic regurgitation has been significantly reduced with the development of the current generation of TAVR valves (36). This is reflected in the presence of only trivial to mild aortic regurgitation in our dataset compared with other groups that used earlier generation valves (37,38). It is therefore unlikely that the degree of AR, which was mild at most in a minority of our patients, would explain the large differences seen in this study between systolic and diastolic parameters and hyperemia and resting parameters.
The sample size of our study may be considered small, with 30 coronary lesions across 28 patients.
However, this is the largest study to date of invasive coronary flow in patients with severe AS and the first to study patients with stenosed coronary arteries. It is also the first study to include phasic analysis, permitting an increase in our understanding of the coronary physiology in this complex hemodynamic condition. This was a mechanistic study, aiming to provide a comprehensive insight to coronary hemodynamic status in patients with severe AS undergoing TAVR. A decision-making strategy for revascularization in patients with severe AS, on the basis of current FFR or iFR data, cannot be made.
This study was designed to compare hyperemic and resting coronary flow and to perform a phasic analysis to delineate differences between systole and diastole. It was not, however, powered to detect differences between resting indices of coronary stenosis severity. Our phasic analysis suggests that there is a significant change in systolic