CMD Prevalence After TAVI Depends on How It Is Measured
Coronary microvascular dysfunction CMD is common in severe aortic stenosis, but the apparent burden shifts depending on the diagnostic tool used. In a systematic review and meta-analysis of 11 primary studies, pooled CMD prevalence before transcatheter aortic valve implantation TAVI was 46.8% when assessed with wire-based thermodilution indices and 38.4% with angiographic indices. Those are not interchangeable numbers. They reflect different physiologic constructs, different cut-points, and different sensitivities to the hemodynamic state of the ventricle and microcirculation at the time of testing.
The distinction matters at the bedside because severe aortic stenosis does not injure the coronary circulation through one pathway alone. Chronic pressure overload drives left ventricular hypertrophy, raises extravascular compressive forces, shortens diastolic perfusion time, and alters coronary hemodynamics. The result is a microvascular bed under constant stress, where reduced capillary density, perivascular fibrosis, and impaired vasodilatory reserve can coexist. When a diagnostic modality asks a different physiologic question, it may label different patients as having CMD even when the underlying disease burden is similar.
Coronary Flow Reserve Improves After Valve Replacement, but the Mechanism Looks Demand-Driven
A separate signal emerged across the paired before-and-after studies: coronary flow reserve CFR increased after TAVI by a pooled mean difference of +0.49. Five studies contributed 103 paired observations to that analysis, and the heterogeneity was substantial. Even so, the direction of change was consistent enough to suggest that relieving the stenotic valve alters coronary physiology in a measurable way.
The biology behind that improvement is unlikely to be simple structural repair. The post-TAVI rise in CFR probably reflects reduced resting myocardial demand after afterload relief rather than a rapid reversal of microvascular remodeling. When the left ventricle no longer has to generate the same pressure against a stenotic valve, baseline oxygen consumption falls and the coronary circulation is less constrained at rest. A higher CFR can therefore appear even if the anatomic microvascular substrate has not meaningfully regenerated.
A single point captures the clinical consequence: improved CFR does not necessarily mean the microcirculation has healed.
Microvascular Resistance Did Not Uniformly Normalize
Changes in microvascular resistance were more selective than the CFR signal. Reductions were confined to pressure-to-velocity and reserve ratio indices, while absolute minimal microvascular resistance remained unchanged in the available studies. Four studies, with 85 observations, informed that narrative synthesis.
This pattern fits a physiologic ceiling on what TAVI can accomplish in the short term. If the most load-sensitive indices improve while absolute minimal resistance does not, the valve replacement may be relieving an imposed hemodynamic burden without erasing the structural or fixed component of microvascular disease. In practical terms, that means a patient can show better coronary reserve after TAVI yet still harbor a microcirculation that is not fully normal. The coronary bed may have more room to breathe, but not necessarily a repaired vascular architecture.
Prognostic Signal for MACE Remains Suggestive, Not Settled
Baseline CMD was associated with major adverse cardiovascular events MACE in the pooled prognostic analysis, but the estimate was imprecise and difficult to generalize. The pooled hazard ratio was 3.22, with a 95% confidence interval from 0.98 to 10.60, based on only two studies and 585 participants. Heterogeneity was substantial.
Here, the interpretation has to stay cautious. The point estimate points toward excess risk, yet the interval is wide enough to include little or no association. The inconsistency reflects a deeper problem in the literature: CMD definitions were not equivalent across studies, and the hazard ratios were not calculated in a uniform way. Some studies used different invasive indices, others used different thresholds, and the timing of assessment relative to TAVI also varied. Those differences can produce a misleading sense of agreement when the underlying physiologic definitions are not aligned.
Standardization Is the Main Barrier to Clinical Use
The review’s most important contribution may be its exposure of methodological fragmentation. Eleven studies are enough to suggest that CMD is frequent in severe aortic stenosis and often persists in some form around the time of valve replacement. They are not enough to define a single prevalence estimate that can be applied broadly to clinical decision-making. The same is true for prognostic use. Without standardized CMD definitions, it is difficult to know whether a positive test identifies diffuse microvascular disease, afterload-related hemodynamic distortion, or both.
The authors therefore stratified prevalence by diagnostic modality rather than pooling unlike indices into a single summary figure. That choice is clinically sensible. Wire-based thermodilution and angiography-derived approaches do not measure the same thing with the same sensitivity, and pooling them without separation would blur the physiology rather than clarify it. The same logic explains why microvascular resistance changes were narrated rather than pooled across dissimilar constructs such as hyperemic microvascular resistance, microvascular resistance reserve, and minimal microvascular resistance.
Prospective multicenter studies will need to solve three practical problems before CMD can move from academic signal to routine risk stratifier in TAVI care. First, investigators need common thresholds. Second, they need consistent timing of post-TAVI reassessment. Third, they need outcome studies large enough to determine whether CMD tracks with events independently of the hemodynamic burden of aortic stenosis itself.
For now, the evidence supports a restrained conclusion: CMD is common before TAVI, CFR often improves afterward, and the microvascular story is not yet mature enough to guide prognosis with confidence. The pooled associations are informative, but they remain limited by small study counts, heterogeneous physiologic definitions, and incompatible methods for summarizing risk.