Numerical Validation of Diagnostic Uncertainties During In Vitro Assessment of Serial Coronary Stenoses

Fractional flow reserve (FFR: ratio of pressures distal and proximal to a stenosis), the commonly used diagnostic parameter for functional assessment of coronary stenosis, has a threshold range of 0.75–0.80 below which clinical intervention is recommended. This range is known as a “gray” zone resulting in uncertainty of diagnosis. An alternate diagnostic parameter, pressure drop coefficient (CDP: ratio of trans-stenotic pressure drop to the proximal dynamic pressure), developed based on fundamental fluid dynamics principles, has been suggested by our group. Pressure drop coefficient has a wide range of 0 ∼ 1000.In the presence of an additional downstream stenosis, FFR values of an upstream intermediate stenosis (80% area stenosis (AS)) may vary near the “gray” zone due to the reduction in hyperemic flow (flow at maximum arterial vasodilation), Q˜h, and pressure drop, Δp˜, leading to the possible misinterpretation of stenosis severity. Hence, using in vitro experiments and clinical data, we developed and validated a computational fluid dynamics (CFD) model to assess the variability of FFR and CDP of a fixed upstream intermediate stenosis when in series with a downstream stenosis of different area obstruction (mild: 64% AS, intermediate: 80% AS, and severe: 90% AS).Angiographic evidence [1] was used to construct rigid stenosis models which were paired in series to form three combinations of serial stenoses representing 80%–64% AS, 80%–80% AS and 80%–90% AS. In vitro experiments for the serial stenoses were conducted using a previously designed setup and procedure [2]. Δp˜-Q˜ characteristic curves of serial stenoses, obtained from experiments, were used in combination with the clinical coronary flow reserve (CFR: ratio of hyperemic flow to flow at rest)-distal perfusion pressure (p˜rh) plot in order to obtain the physiologic Q˜h for the serial stenoses combinations. FFR and CDP were then evaluated using the Q˜h and pressure values.A computational domain of serial stenoses, described above, consisted of upstream and downstream stenoses with a concentric 0.014” guidewire inserted through it (Fig. 1(a)). For each combination, a 3D model of the serial stenoses geometry was constructed and meshed with approximately 400,000 hexahedron elements, based on a mesh refinement study. Time-averaged hyperemic flow (Q˜h) obtained from experiments (136.4 ml/min for 80%–64% AS, 126.4 ml/min for 80%–80% AS, and 90.7 ml/min for 80%–90% AS), were used as steady-state laminar flow inputs for the CFD simulations (CFX, ANSYS Inc.). The boundary conditions comprised of a velocity inlet, stress-free outlet, and a no-slip artery and guidewire wall. Blood was modeled as a non-Newtonian Carreau fluid and the model constants were obtained from experimental data.Fractional flow reserve of the upstream intermediate stenosis increased when the downstream stenosis severity was increased due to the reduction in Q˜h. Fractional flow reserve for a single intermediate stenosis was reported to be 0.73 [2]. In the presence of a downstream mild, intermediate, and severe stenosis, the FFRe (superscript e: experimental) increased to 0.77, 0.79, and 0.88, respectively (Fig. 1(b)). Comparable FFRn (superscript n: numerical) values of 0.82, 0.85, and 0.92 were obtained for the corresponding cases and were within 7% of the FFRe values. Hence, FFR of the upstream intermediate stenosis, in the presence of a downstream stenosis, was either within or above the “gray” zone, leading to the possible misinterpretation of stenosis severity.Pressure drop coefficient of a single intermediate stenosis was reported to be 56 [2]. The CDPe values of the upstream intermediate stenosis, increased from 47 to 49 and further to 54 when the downstream severity was increased from mild to intermediate and further to a severe condition, respectively. The CDPn values of the upstream intermediate stenosis for the corresponding cases were 37, 36, and 39, respectively. A maximum difference of 28% was observed between CDPe and CDPn. However, considering the large range of CDP, the CDPe and CDPn values of the upstream intermediate stenosis showed minimal variation in the presence of a downstream stenosis of varying degree of severity.A similar trend of FFRn and CDPn values with FFRe and CDPe indicates that the complex flow through serial stenoses may be accurately modeled computationally and this may provide alternate means for cardiovascular diagnosis. More importantly, the artificial increase in FFR for a clinically relevant upstream intermediate stenosis in the presence of a downstream stenosis could lead to the incorrect deferral of coronary intervention. In contrast, CDP could better delineate the severity of the upstream stenosis, irrespective of the severity of the downstream stenosis.