Coronary microvascular resistance comparison of coronary arteries with and without considering vascular diameter: A retrospective, single‐center study

Abstract Background and Aims Measurement of coronary microvascular resistance (MR) is essential for diagnosing nonocclusive coronary artery ischemia, but whether coronary branches of different diameters can be similarly assessed using hyperemic microvascular resistance index (hMVRI) calculated from average peak velocity (APV) remains unclear. We investigated the relationship between coronary arteries of different diameters and hMVRI. Methods Thirty patients with suspected angina pectoris and nonobstructive coronary stenosis with fractional flow reserve >0.8 underwent evaluation of all coronary arteries using a Doppler velocity and pressure‐equipped guidewire. Quantitative coronary angiography (QCA) was used to analyze vessel diameter (DQCA). Coronary blood flow (CBFQCA) was calculated as πDQCA 2/4 (0.5 × APV) and hMVRIQCA as distal coronary pressure divided by CBFQCA during maximal hyperemia. Results The hMVRI was significantly higher for the right coronary artery than for the left anterior descending artery, but no significant differences between arteries were seen for CBFQCA and hMVRIQCA. Although the correlation between CBFQCA and APV was weak, CBFQCA divided into three groups according to DQCA showed very strong correlations with APV. Slopes of the straight line between APV and CBFQCA for small‐, middle‐, and large‐diameter groups were 0.48, 0.30, and 0.21, respectively, with slope decreasing as diameter increased. Conclusions Comparative evaluation of MR in coronary branches with varying vessel diameters requires vessel diameter to be accounted for.


| INTRODUCTION
Although angina pectoris affects approximately 112 million people globally, up to 70% of patients undergoing invasive angiography do not have obstructive coronary artery disease.Instead, the main cause is ischemia with no obstructive coronary arteries (INOCA), which can result from heterogeneous mechanisms, including coronary vasospasm and microvascular dysfunction, and is not a benign condition. 1Since no currently available modalities allow direct visualization of the human coronary microcirculation in vivo, microvascular evaluation relies on the measurement of parameters that reflect the functional status, such as coronary blood flow (CBF), coronary flow reserve, and microvascular resistance using a diagnostic guidewire.Both microvascular and epicardial vasospastic angina are then assessed with the acetylcholine test.
Lifestyle modification, risk factor management, and anti-anginal medications have been proposed for the management of INOCA.
However, studies of therapies to improve coronary microvascular dysfunction (CMD) have been small and heterogeneous in both design and methodology, and evidence-based treatments for CMD are not currently available.An updated standardization of criteria for microvascular angina (MVA) in patients presenting with angina pectoris or ischemia-like symptoms in the absence of flow-limiting coronary artery disease was proposed by the COVADIS group, with abnormal coronary microvascular resistance indices among the leading diagnostic parameters.Although these parameters can be investigated using the index of microcirculatory resistance (IMR) or hyperemic microvascular resistance index (hMVRI), evaluation in the catheterization laboratory has become easier in recent years and guidelines recommend IMR and hMVRI measurements to investigate INOCA.Many reports [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] have already examined the usefulness and pitfalls of guidewires in the examination of IMR or hMVRI, but which coronary artery should be targeted for evaluation in INOCA remains a vexing question.
8][9] On the flip side, when the LAD cannot be used for technical reasons, use of the left circumflex artery (LCX) is recommended, followed by the right coronary artery (RCA).The reason for this may be that previous reports have shown that hMVRI in the RCA is significantly higher than in other coronary branches. 14,15Average peak velocity (APV) is used instead of CBF for coronary microvascular resistance measurement, because IMR and hMVRI are used for simple measurements in the cardiac catheterization laboratory.However, whether coronary branches of widely varying diameters can be evaluated in the same manner using APV in humans remains highly questionable.
To determine whether differences in coronary microcirculatory resistance vary according to vessel size at the site of measurement of the coronary branch, we studied cases in which all coronary branches in the same patient were measured.

| Patient population
We evaluated patients with clinical suspicion of stable angina pectoris based on the presence of chest pain or chest discomfort.
Evaluations were performed using elective coronary angiography and physiological assessments between September 2010 and May 2016.Of these, 30 patients (Group B) had all three coronary branches available for examination and an FFR > 0.8.Coronary microvascular resistance between coronary branches was then examined separately using both conventional indices and indices that considered vessel size using quantitative coronary angiography (QCA).All patients provided written informed consent for participation.The Ethics Committee at Ureshino Medical Center approved the protocol for the present study (approval no.10-10), which was performed in accordance with the principles laid out in the Declaration of Helsinki.

| Cardiac catheterization
After coronary angiography, aortic pressure was measured via a 5-or 6-Fr guiding catheter placed in the coronary ostium via a radial or femoral approach.Intracoronary pressure and coronary flow velocity were measured with a 0.014″ Doppler velocity and pressure-equipped guidewire (Volcano Corp), then the sensor was placed at a site more distal than the distal half of the coronary artery.Hyperemia was induced by papaverine hydrochloride delivered into the coronary artery (12 mg into the left coronary artery and 8 mg into the RCA within 15 s), then blood pressure was recorded at 20 s after the end of administration.
FFR was defined as the ratio of mean distal coronary pressure (Pd) to mean aortic pressure (Pa) in the target vessel beyond the lesion during maximal hyperemia.hMVRI was calculated as Pd divided by distal APV during maximal hyperemia.In the present study, those coronary arteries showing FFR > 0.8 were defined as non-obstructive coronary arteries unaffected by collateral effects, 5 and were investigated.The estimation of volumetric flow from Doppler flow velocity also incorporates vessel diameter.Volumetric CBF should always be calculated, using QCA to estimate epicardial diameter. 6,8Vessel diameter as analyzed by QCA (D QCA ) was measured at the site of the wire where blood flow was measured (Figure 1).QCA was performed using a validated densitometric analysis system (Pie Medical Imaging B.V., Maastricht, the Netherlands).CBF (CBF QCA ) was calculated as πD QCA 2 /4 (0.5 × APV) (D QCA , vessel diameter), then hMVRI QCA was calculated as Pd divided by CBF QCA during maximal hyperemia.
Coronary physiological values were calculated as follows:

| Patient characteristics
Patient characteristics in Groups A and B are presented in Table 1.
Median age was 69.0 years in Group A and 74 years in Group B. Men comprised 113 of the 160 patients (70.6%) in Group A and 23 of the 30 patients (76.7%) in Group B. No significant differences in patient characteristics were seen between Groups A and B, except for HbA1c.HbA1c was significantly lower in Group B than in Group A (p = 0.040), but no significant difference in the incidence of diabetes mellitus was identified (p = 0.301).
F I G U R E 1 Calculation of coronary physiological values.Hyperemic microvascular resistance index (hMVRI) was calculated as distal coronary pressure (Pd) divided by distal average peak velocity (APV) at maximal hyperemia.Vessel diameter was analyzed by quantitative coronary angiography (QCA) at the site of blood flow measurement by wire (white arrow).QCA was performed with a validated densitometric analysis system (Pie Medical Imaging B.V, Maastricht, the Netherlands).Coronary blood flow (CBF) was calculated as πD 2 /4 (0.5 × APV), where D is vessel diameter.CBF as measured by QCA is described as CBF QCA , then hMVRI QCA is calculated as Pd divided by CBF QCA during maximal hyperemia.As CBF QCA will be overestimated, hMVRI QCA will be higher than the original HMR.

| Relationship between coronary physiological assessments and coronary artery branches
Coronary physiological assessments of Groups A and B are presented in Table 2 and Figure 2, respectively.Median values of FFR, hMVRI, APV, and Pd in Group A were 0.92, 1.9, 37, and 72 mmHg/cm/s, respectively.Median values of FFR, hMVRI, APV, and Pd in Group B were 0.94, 1.9, 29, and 74 mmHg/cm/s, respectively.
In terms of the relationships between LAD, LCX, and RCA, similarly significant differences in FFR, hMVRI and Pd were seen in Groups A and B (Table 2 and Figure 2).In both Groups A and B, hMVRI was significantly higher for the RCA than for the LAD.(p = 0.152) (Table 3 and Figure 3A).Unlike the results examined in the APV, no significant differences between LAD, LCX, or RCA were seen for CBF QCA or hMVRI QCA (Table 3 and Figure 3B,C).
3.4 | APV and CBF QCA for small D QCA , middle D QCA , and large D QCA in Group B We found no collinearity between DQCA and APV in scattergrams (Figure 4A).We then examined the relationship between CBF QCA and APV in Group B, but the correlation was weak (p < 0.001, R 2 = 0.34) (Figure 4B).We further divided vessels into three equal groups according to D QCA : small diameter group,

| DISCUSSION
The present study examined whether differences exist in microcirculatory resistance in human coronary arteries of different sizes.
The main findings were as follows.First, as reported previously, hMVRI was significantly higher for the RCA than for the LAD in Groups A and B. Second, unlike the results for the APV, no significant differences in hMVRI QCA as calculated using D QCA were seen between the LAD, LCX, and RCA.Moreover, the correlation between CBF QCA and APV was weak.However, CBF QCA in patients divided into three groups according to D QCA showed a very strong correlation with APV.
T A B L E 1 Patient characteristics in Groups A and B.  3 whereas hMVRI incorporates Doppler flow velocity. 16Both indices have separately been shown to predict infarct size, 17,18 microvascular obstruction, 18 regional wall motion, 17 and adverse left ventricular remodeling.Williams et al. 19 reported that the correlation between independent invasive and noninvasive measurements of microvascular function was better with hMVRI than with IMR.However, neither of these invasive measures of microvascular resistance consider vessel size.

| Relevance of vessel diameter in evaluating CBF and hMVRI
Unlike the results for APV, no significant difference in hMVRI QCA was found for LAD, LCX, or RCA, respectively, when examined using CBF QCA .Furthermore, the correlation between APV and CBF QCA was weak.Quantitative measurement of volumetric CBF (in milliliters per minute) during catheterization in humans is not currently possible.In evaluating CBF, a method that accounts for vessel size by QCA has been reported in addition to the semi-quantitative method using APV.Perera et al. 8 evaluated angina and non-obstructive coronary artery disease by calculating epicardial diameter using QCA and calculating CBF for the purpose of evaluating endothelial-dependent microvascular function by acetylcholine.
Although no collinearity was evident between D QCA and APV in scattergrams, vessel size as divided into three groups allowed identification of a strong correlation between CBF QCA and APV.
According to Doucette et al., 6 APV showed a very strong linear correlation with CBF in small-sized straight tubes in vitro and in normal LCX in dogs, but slope decreased as diameter increased.
Similarly, in the present study, the slope of the straight line between APV and CBF QCA for small, medium, and large diameters in humans was 0.48, 0.30, and 0.21, respectively, decreasing as diameter increased.Moreover, we assessed associations between APV and CBF QCA for coronary artery branches.The LAD is the preferred target vessel in many studies, reflecting its myocardial mass and coronary arterial predominance, and much of the evidence for physiological assessments in INOCA has been obtained using the T A B L E 2 Coronary physiological assessments by Doppler and pressure sensor-equipped guide-wire in Groups A and B.

Note:
Values are presented as n (%) or median ± standard deviation.
Abbreviations: APV, average peak velocity; FFR, fractional flow reserve; hMVRI, hyperemic microvascular resistance index; LAD, left anterior descending coronary artery; LCX, left circumflex; Pd, mean distal coronary pressure; RCA, right coronary artery.8][9] On the flip side, when the LAD cannot be used for technical reasons, use of the LCX is recommended, followed by the RCA.Nonetheless, RCAs have been documented to exhibit considerably greater resistance indices than LADs and LCXs when measured through conventional means.Consequently, utilizing the same measurement methodology for RCAs as that for LADs poses a predicament.However, adopting resistance indices that account for vessel diameter is anticipated to yield decreased inaccuracies.Furthermore, measuring indices of microcirculatory resistance in multiple coronary artery regions may be particularly useful in patients with typical angina pectoris without obstructive coronary artery disease.However, the abnormality in cases of coronary microvascular dysfunction (CMVD) 20 may not involve all coronary microvessels of a major coronary branch uniformly, but may instead be distributed in the myocardium in a scattered manner, with a sparse distribution of myocardial ischemia.In such situations, significant differences may exist in microvascular coronary resistance even among the three coronary branches.Therefore, measuring the IMR in multiple coronary territories may be useful, particularly in patients with typical angina in the absence of obstructive coronary artery disease.
In addition, abnormalities in cases of CMVD are not consistently present in all coronary microvessels in the main coronary branches.
Instead, these abnormalities are distributed unevenly in the local myocardium, causing myocardial ischemia to be sparsely scattered.
Measuring microvascular resistance in multiple coronary artery territories is thus crucial in comprehending the extent of damage distribution.
In the present study, similar to investigations using continuous coronary thermodilution 21 or positron emission tomography (PET), [22][23][24] interindividual variability in CBF QCA and hMVRI QCA values was observed.The main factor explaining the large ranges in flow and resistance values appears to be the dependence on myocardial mass.PET-derived flow and resistance express these measurements per unit of tissue mass (i.e., milliliters per minute per gram of tissue).Indeed, when adjusting CBF and microvascular resistance for mass, similar values were found for the three different myocardial territories within the same patients.Even so, even when adjusted for mass, inter-individual values for CBF and microvascular resistance show a considerable range. 23,24Further confirming the hypothesis of natural variation in hyperemic

| STUDY LIMITATIONS
Several study limitations to this investigation should be acknowledged.
First, the present study was performed at a single facility and involved a relatively small population.
(A) Coronary angiography.(B) Insertion of the Doppler flow guidewire into the left anterior descending artery.(C) Final position of the Doppler flow guidewire as confirmed by contrast injection.(D) Physiological measurements at rest.(E) Physiological measurements under hyperemic conditions.CFR, coronary flow reserve; HMR, hyperemic microvascular resistance; HSR, hyperemic stenosis resistance.

F
I G U R E 2 Evaluation of FFR, hMVRI, APV, and Pd in Groups A and B. In comparisons between LAD, LCX, and RCA, significant differences in FFR, hMVRI, and Pd are seen in Group A (A-D) and Group B (E-H).(A) FFR is significantly lower in the LAD than in the LCX or RCA (p < 0.001) in Group A. (B) hMVRI is significantly lower in the LAD than in the LCX or RCA (p = 0.003) in Group A. (C) No significant differences in APV are seen between the three coronary branches in Group A. (D) Pd is significantly lower in the LAD than in the RCA (p < 0.001) in Group A. (E) FFR is significantly lower in the LAD than in the LCX or RCA (p < 0.001) in Group B. (F) hMVRI is significantly lower in the LAD than in the RCA (p = 0.019) in Group B. (G) No significant differences in APV are seen between the three coronary branches in Group B. (H) Pd is significantly lower in the LAD and LCX than in the RCA (p = 0.004) in Group B. † p < 0.05; † † p < 0.001.FFR, fractional flow reserve; hMVRI, hyperemic microvascular resistance index; LAD, left anterior descending coronary artery; LCX, left circumflex; Pd, mean distal coronary pressure; RCA, right coronary artery.T A B L E 3 Re-evaluation of Group B using hMVRI QCA and CBF QCA .
myocardial perfusion, rather than measurement inaccuracies, Everaars et al.24 found hyperemic flow values in the range of approximately 50-450 mL/min in arteries.The large ranges of hyperemic flow values also found in the present study thus seem likely to correspond to a naturally occurring large variability of normal hyperemic values.
3.3 | Relationships between D QCA , CBF QCA , and hMVRI QCA of coronary artery branches in Group B D QCA was analyzed by QCA at the site of blood flow measurement by diagnostic wire in Group B, with hMVRI re-evaluated as hMVRI QCA and CBF as CBF QCA .D QCA of the LAD, LCX, and RCA was 2.27, 2.81, and 2.91 mm, respectively; D QCA was significantly larger for the LCX and RCA than for the LAD (LCX, p = 0.029; RCA, p < 0.001), but no significant difference was seen between D QCA of the LCX and RCA