Ongoing Methodological Approaches to Improve the In Vivo Assessment of Local Coronary Blood Flow and Endothelial Shear Stress

JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY

VOL. 66, NO. 2, 2015

ª 2015 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION

ISSN 0735-1097/$36.00

PUBLISHED BY ELSEVIER INC.

http://dx.doi.org/10.1016/j.jacc.2015.05.010

EDITORIAL COMMENT

Ongoing Methodological Approaches to Improve the In Vivo Assessment of Local Coronary Blood Flow and Endothelial Shear Stress The Devil Is in the Details* Peter H. Stone, MD,y Ahmet Umit Coskun, PHD,z Francesco Prati, MDx

A

ppreciation of the pathobiology of coronary

(3D) geometry of the arterial lumen and accurate

atherosclerosis and the factors responsible

measurement of the coronary blood flow. These

for its development and progression have

variables are then entered into a computational fluid

expanded dramatically in the past few years, along

dynamics (CFD) program to solve the fundamental

with a growing interest in the in vivo assessment of

equations of fluid flow (Navier-Stokes equations) to

factors that may predict future plaque outcomes and

determine the local blood flow patterns along the

coronary events. It has been known for years that

course of the artery. The detailed local shear stress

atherosclerosis develops and progresses preferen-

pattern along the lumen wall is calculated as the

tially in areas of low endothelial shear stress (ESS),

product of the velocity gradient at the lumen wall

located at the inner aspects of vascular curves, at bi-

and blood viscosity. Most methodologies using in-

furcations, and downstream from focal obstructions,

travascular ultrasound or optical coherence tomog-

but only recently have methodologies been devel-

raphy (OCT) now reconstruct the coronary artery by

oped that enable in vivo calculation of local shear

censoring the side branches and focus on detailed

stress patterns. In vivo hemodynamic assessments

blood flow patterns through the main artery. Because

on the basis of imaging studies hold promise for facil-

the side branches divert some portion of coronary

itating clinical decision making concerning the pres-

blood flow, ignoring the blood flow through the side

ence and natural history of atherosclerosis (1) and

branches will lead to inaccurate measurements of the

for flow simulations to predict pressure drop to model

blood flow through the main coronary artery: flow in

fractional flow reserve measurements in the evalua-

the main artery after the branching would be over-

tion of need for a revascularization procedure (2).

estimated, resulting in overestimation of local ESS as

Accurate measurement of local ESS is dependent on accurate reconstruction of the 3-dimensional

well as a net pressure drop after the branching. The technical difficulty of quantifying the blood flow diverted into the side branches, however, has been so anatomically and computationally complex that the side branch blood flow has generally been ignored or

*Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.

investigation of main coronary segments with large side branches is avoided. SEE PAGE 125

From the yCardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts; zMechanical and Industrial Engineering Department, Northeastern University, Boston,

In

this

issue

of

the

Journal,

Li

et

al.

(3)

Massachusetts; and the xDepartment of Interventional Cardiology, San

now introduce a different methodology with a goal

Giovanni Hospital, CLI Foundation, Rome, Italy. Dr. Prati has received

of more accurately measuring the local blood flow

consultant fees from St. Jude Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to

in the main coronary artery by including the

disclose.

more complex component of the blood flow that is

Stone et al.

JACC VOL. 66, NO. 2, 2015 JULY 14, 2015:136–8

The Devil Is in the Details

diverted through the side branches in the CFD

accurate manner, but it is not clear that the lumen

computations. The preliminary adoption of a co-

diameter of the side branch ostium is also measured

registration software, now commercially available,

in an accurate manner because the side branch

allowed for spatial combination of angiographic and

ostium diameter is in a different plane than the OCT

OCT images. The investigators then fused OCT im-

image acquisition plane. It is also not clear if the

ages of the main coronary artery, to identify the

lumen dimensions of the side branches beyond the

detailed lumen dimensions and the ostium size of

ostium are measured. If so, they could be on the basis

the side branches, with coronary angiography that

of the angiographic images, or alternatively on the

provided a 3D reconstruction of both the main artery

OCT images of the side branch take-off, obtained from

and the side branches. They used 2 angiographic

interrogating the main artery. But either solution may

projections with a difference in angulation >25  and

well be inaccurate, especially because the branches

included the whole course of all side branches of >1-

are so narrow. Similarly, the length of the side branch

mm diameter. The OCT was co-registered and fused

would be necessary to determine the side branch

with the 3D angiography using side branches as

blood flow and it is not clear this can be accurately

landmarks to correct for longitudinal and rotational

assessed by coronary angiography. It is also not stated

mismatch. The investigators then created a 3D

if the angle of the side branch from the main artery is

reconstruction of the OCT-derived lumen by match-

included in the computational program. Validation

ing the centerline of the reconstructed coronary

of the absolute accuracy of the tree model method

angiogram with the centerline of the OCT images.

will be very difficult, but necessary, because the side

They performed CFD analyses using 2 models which

branch size and blood flow are very small and direct

were compared: the single-conduit model, which

measurement of actual flow and pressure would

ignored the side branches; and the tree model,

be very problematic. It will be necessary as well to

which included the lumina of the side branches

validate this approach with fractional flow reserve

derived from the 3D angiography.

during maximal hyperemia. It is also important to be

The investigators observed that the calculation of mean blood flow rate through the coronary artery was

aware of the incremental time required to perform the complex tree model methodology.

significantly lower using the tree model than the

The proposed method may have clinical implica-

single-conduit model, as expected, and the estimated

tions. More accurate assessment would be invaluable

pressure at the outflow was significantly higher in the

to more reliably determine the local ESS patterns

tree model compared to the single-conduit model.

responsible for the development and progression of

The ESS in the tree model was an average of 4.64 Pa

coronary plaques and their risk of disruption leading

lower than in the single-conduit model.

to a new clinical event. More reliable assessment of

The investigators’ goal is important, and the

the functional severity of an obstruction (distal cor-

magnitude of ESS difference between the 2 methods

onary pressure to aortic pressure ratio, Pd/Pa) would

is not trivial. Improving the computational model to

also be very helpful to improve the physician’s

include the blood flow through the side branches so

decision-making process, especially in the presence

the blood flow remaining in the main artery can be

of uncertain/intermediate lesions. Furthermore, OCT

accurately measured would improve the accuracy of

findings are extremely helpful to optimize percuta-

the calculations of local blood flow in the main artery,

neous coronary intervention results (4). If the method

and consequently the calculations of local ESS and

is accurate then it would also be ideal to have the

the local pressures. It is not surprising that there

method readily available to other investigators to

is not ideal agreement between the computations

facilitate more in vivo investigation of the pathobio-

derived from the 2 computational models; the critical

logic and clinical implications of local blood flow and

issue is whether the hemodynamic values from

local pressure. The time requirements for the ambi-

the tree model method are indeed more accurate

tious post-processing 3D reconstruction may limit

than the values from the single-conduit model, or

its online clinical application in the catheterization

simply different approximations with comparable

laboratory.

overall accuracy considering all the uncertainties in

The field of intravascular assessment of local ESS

measurements, natural dynamic variations in flow,

and blood pressure has expanded dramatically in

geometry, and pressure levels. The fundamental

recent years, and more accurate, and more readily

issue to this point is whether the anatomy of the

available, methodologies will be invaluable to mea-

side branches is accurately identified using the

sure these variables and inform clinical decision-

investigators’ methodology. OCT images certainly

making in a reliable, practical, and timely manner.

identify the cross-sectional lumen margins in a very

The clinical applications are very important, both for

137

138

Stone et al.

JACC VOL. 66, NO. 2, 2015 JULY 14, 2015:136–8

The Devil Is in the Details

local hemodynamic assessment for investigation of

need to be sure the new calculations are indeed more

the pathobiology of atherosclerosis and for local

accurate, not just different.

pressure assessment. The methods proposed by Li et al. (3) may be very important to enhance the ac-

REPRINT REQUESTS AND CORRESPONDENCE: Dr.

curacy of our coronary invasive measurements and

Peter H. Stone, Cardiovascular Division, Brigham &

improve the accuracy of our insights compared to the

Women’s Hospital, 75 Francis Street, Boston, Massa-

more routine method of single-conduit model, but we

chusetts 02115. E-mail: [email protected].

REFERENCES 1. Stone PH, Saito S, Takahashi S, et al. Prediction of progression of coronary artery disease and clinical outcomes using vascular profiling of endothelial shear stress and arterial plaque characteristics: the PREDICTION Study. Circulation 2012;126:172–81. 2. Taylor CA, Fonte TA, Min JK. Computational fluid dynamics applied to cardiac computed tomography for noninvasive quantification of

fractional flow reserve: scientific basis. J Am Coll Cardiol 2013;61:2233–41. 3. Li Y, Gutiérrez-Chico JL, Holm NR, et al. Impact of side branch modeling on computation of endothelial shear stress in coronary artery disease: coronary tree reconstruction by fusion of 3D angiography and OCT. J Am Coll Cardiol 2015;66: 125–35.

4. Prati F, DiVito L, Biondi-Zoccai G, et al. Angiography alone versus angiography plus optical coherence tomography to guide decision-making during percutaneous coronary intervention, et al. EuroIntervention 2013;8:823–9.

KEY WORDS computational fluid dynamics, coronary blood flow, endothelial shear stress