Prevalence of Coronary Microvascular Disease and Coronary Vasospasm in Patients With Nonobstructive Coronary Artery Disease: Systematic Review and Meta-Analysis.

Background A relevant proportion of patients with suspected coronary artery disease undergo invasive coronary angiography showing normal or nonobstructive coronary arteries. However, the prevalence of coronary microvascular disease (CMD) and coronary spasm in patients with nonobstructive coronary artery disease remains to be determined. The objective of this study was to determine the prevalence of coronary CMD and coronary vasospastic angina in patients with no obstructive coronary artery disease. Methods and Results A systematic review and meta-analysis of studies assessing the prevalence of CMD and vasospastic angina in patients with no obstructive coronary artery disease was performed. Random-effects models were used to determine the prevalence of these 2 disease entities. Fifty-six studies comprising 14 427 patients were included. The pooled prevalence of CMD was 0.41 (95% CI, 0.36-0.47), epicardial vasospasm 0.40 (95% CI, 0.34-0.46) and microvascular spasm 24% (95% CI, 0.21-0.28). The prevalence of combined CMD and vasospastic angina was 0.23 (95% CI, 0.17-0.31). Female patients had a higher risk of presenting with CMD compared with male patients (risk ratio, 1.45 [95% CI, 1.11-1.90]). CMD prevalence was similar when assessed using noninvasive or invasive diagnostic methods. Conclusions In patients with no obstructive coronary artery disease, approximately half of the cases were reported to have CMD and/or coronary spasm. CMD was more prevalent among female patients. Greater awareness among physicians of ischemia with no obstructive coronary arteries is urgently needed for accurate diagnosis and patient-tailored management.

coronary flow reserve

coronary microvascular disease

Women’s Ischemia Syndrome Evaluation

Clinical Perspective

What Is New?

In patients with no obstructive coronary artery disease, approximately half of cases present with underlying disease, either coronary microvascular disease or coronary vasospasm.

Coronary microvascular disease is more prevalent in female patients; nonetheless, male patients are affected in a significant proportion.

Invasive and noninvasive diagnostic methods identified a similar proportion of patients with coronary microvascular disease.

What Are the Clinical Implications?

The large variability of methods, definitions, and thresholds for diagnosing coronary microvascular disease and coronary vasospasm is a call to a refinement and standardization of diagnostic tools.

Greater awareness among physicians of ischemia with no obstructive coronary arteries is urgently needed for proper diagnosis and patient‐tailored management.

Ischemic heart disease is the leading cause of mortality and morbidity globally. However, in clinical practice, a relevant proportion of patients with suspected coronary artery disease (CAD) undergo invasive coronary angiography showing normal or nonobstructive coronary arteries. Although many of these patients are considered as having normal coronary arteries, ischemia with no obstructive CAD has been associated with increased cardiovascular risk and higher rates of repeat coronary angiography. , , Recent guidelines reflect the wide spectrum of etiopathogenesis of ischemic heart disease and chronic coronary syndromes. Not only coronary atherosclerosis, but disorders of microcirculation and vasomotion may be part of the intricate process leading to myocardial ischemia. Coronary microvascular disease (CMD) is increasingly seen as an important contributor to the pathophysiology of ischemic heart disease. The diagnosis of CMD can be ascertained by means of invasive cardiac catheterization or noninvasive imaging techniques (Figure 1). Epicardial spasm, a separate clinical entity, can also lead to myocardial ischemia and myocardial infarction. , The diagnosis of coronary spasm ideally relies on the results of provocation tests performed in the catheterization laboratory. However, the prevalence of CMD and coronary spasm in patients with nonobstructive CAD remains to be determined.

Figure 1

Methods used for evaluation of microvascular disease.

A, Transthoracic echocardiography with Doppler of LAD. B, PET. C, MIBI SPECT. D, CMR. E, Doppler CFR. F, Absolute coronary blood flow measured by thermodilution. G, Thermodilution, CFR and IMR. H, Acetylcholine testing. CFR indicates coronary flow reserve; CMR, cardiac magnetic resonance; IMR, index of microcirculatory resistance. LAD, left anterior descending artery; MIBI SPECT, myocardial perfusion imaging on single photon emission computed tomography; and PET, positron emission tomography.

The aim of the present systematic review and meta‐analysis was to determine the prevalence of CMD and coronary spasm assessed by invasive and noninvasive methods in patients with no obstructive CAD.

Methods

The data that support the findings of this study are available from the first author upon reasonable request.

Search Strategy and Selection Criteria

Studies describing prevalence of coronary microvascular disease and coronary spasm among patients with no obstructive CAD were reviewed. Two reviewers (N.M. and G.M.) systematically searched PubMed and Scopus. The search was conducted in August 2021, starting from inception, and was performed separately for coronary microvascular dysfunction and coronary vasospasm (Table S1). No restrictions were applied for language. Additionally, reference lists of the eligible studies and recent systematic reviews were screened to identify relevant studies. In case of multiple publications with the same population, the latest report was used. The inclusion criteria were: (1) studies comprising patients with suspected CAD, (2) presenting with no obstructive coronary disease, and (3) undergoing a diagnostic test for CMD, spasm, or both with a report of the number of patients testing positive and the total number of patients evaluated. Studies were divided into 2 groups according to the pathophysiology assessed: CMD and coronary vasospasm, respectively. The definition of no obstructive coronary disease and the threshold of diagnostics tests used to define the presence of CMD were based on each individual study. The present systematic review and meta‐analysis is presented in agreement with Preferred Reporting Items for Systematic Reviews and Meta‐Analyses reporting guidelines (Table S2). Quality of included studies was assessed by the Quality Assessment of Diagnostic Accuracy Studies tool. Risk of bias was evaluated across 4 domains: patient selection, index test, reference standard, and flow and timing. This systematic review and meta‐analysis was registered in PROSPERO (International Prospective Register of Systematic Reviews) (CRD42020220077).

Outcomes of Interest

The primary outcome of interest was the prevalence of CMD and/or coronary vasospasm among patients with no obstructive CAD. Patients’ demographic and clinical characteristics, diagnostic methods performed, and number of positive patients were collected. In the present meta‐analysis, definitions of CMD and vasospasm were used according to the ones defined in each study.

Statistical Analysis

Categorical variables are reported as percentages, and continuous variables are reported as mean±SD. To account for heterogeneity between studies, a random‐effects model based on the Der Simonian‐Laird method was used. Weighted events are reported with 95% CIs. Heterogeneity was assessed using the I 2 value. I 2 values of 25%, 50%, and 75% represented mild, moderate, and severe inconsistency, respectively. Random‐effects meta‐regression analyses were used to explore the influence of sex, clinical characteristics, type of diagnostic method, different inclusion, and exclusion criteria on the outcome of interest. Linearity was assessed visually. Pairwise meta‐analysis was performed to compare the risk of CMD between sexes. All analyses were performed using R version 4.0.2 meta and metafor packages (R Foundation for Statistical Computing, Vienna, Austria).

Results

One hundred fifty‐five articles received a complete review, and 56 studies met inclusion criteria and were included in the meta‐analysis (Figure 2). , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , Overall, 14 427 patients were included. The mean age was 59±5 years, 65% were women, and 21% had diabetes. Most of the patients (75%) underwent invasive evaluation. Studies included in the systematic review, methods used, inclusion criteria, and definitions are described in Table S3.

Figure 2

Preferred Reporting Items for Systematic Reviews and Meta‐Analyses flowchart.

CAD indicates coronary artery disease.

The risk of bias was low on the index test, reference standard, flow, and timing. Nevertheless, in 11% (6/56) of the studies, the risk of bias in the patient selection was considered high because of inclusion of women only (Figure S1). The assessment of the quality of the studies included in the meta‐analysis is presented in (Table S4).

Coronary Microvascular Disease

Thirty‐seven studies reporting rates of CMD in patients with no obstructive CAD were included. They comprised 7212 participants; the mean age was 59±5 years, 61% were women, 66% had hypertension, 22% had diabetes, and 19% were smokers. Twenty‐four studies used invasive methods for diagnosing CMD, whereas 14 used noninvasive methods. Assessment of invasive coronary flow reserve (CFR), either by Doppler or thermodilution techniques, was the most used method (45%), followed by positron emission tomography in 32% of patients (Figure 3). Table 1 shows baseline clinical characteristics of patients undergoing CMD investigations.

Figure 3

Bar plot chart with studies evaluating the prevalence of CMD assessed by invasive (different shades of red) and noninvasive (different shades of blue) methods.

Solid gray line illustrates the 42% pooled prevalence of CMD, and the dashed lines illustrate 95% CIs. CFR indicates invasive measurement of coronary flow reserve, Doppler, and thermodilution method; CMR, cardiac magnetic resonance; IMR, index of microcirculatory resistance; PET, positron emission tomography; and SPECT, myocardial perfusion imaging on single photon emission computed tomography.

Table 1

Number of Positive Patients and Baseline Clinical Characteristics of the Patients Included in the Studies Investigating the Prevalence of Coronary Microvascular Disease

Study	Patients included	No. positive, n (%)	Age, y	Women, n (%)	Hypertension, n (%)	Diabetes, n (%)	Dyslipidemia, n (%)	Current smoker, n (%)
Cassar, 2009 13	376	170 (45%)	49±11	254 (68%)	157 (42%)	36 (10%)	208 (55%)	NA
Godo, 2020 32	148	91 (62%)	44±9	111 (75%)	79 (53%)	11 (7%)	91 (62%)	60 (41%)
Ford, 2018 33	151	78 (52%)	61±10	111 (74%)	125 (81%)	29 (19.2%)	120 (79.5%)	24 (15.9%)
Graf, 2006 35	58	42 (72%)	58±10	39 (67%)	NA	8 (18%)	NA	17 (29%)
Hasdai, 1998 36	203	118 (58%)	51 (17–78)	158 (78%)	59 (29%)	8 (4%)	88 (43.3%)	28 (27%)
Kobayashi, 2015 39	157	39 (25%)	64±12	117 (29%)	77 (49%)	38 (24%)	91 (58%)	47 (30%)
Kotecha, 2019 40	23	16 (70%)	63±8	NA	6 (26%)	NA	NA	NA
Lee, 2015 42	137	38 (28%)	54±11	107 (77%)	74 (53%)	32 (23%)	87 (63%)	11 (8%)
Michelsen, 2018 43	919	241 (26%)	62±9	919 (100%)	467 (51%)	117 (13%)	580 (63%)	149 (16%)
Murthy, 2014 44	1218	641 (53%)	62 (53–69)	813 (67%)	894 (73%)	363 (30%)	663 (54%)	121 (10%)
Pargaonkar, 2019 47	155	34 (22%)	54±13	119 (77%)	68 (44%)	26 (17%)	90 (58%)	23 (15%)
Pargaonkar, 2020 48	88	32 (36%)	NA	53 (60%%)	NA	NA	NA	NA
Pepine, 2010 49	152	74 (49%)	55±10	189 (100%)	57 (30%)	21 (11%)	50 (26%)	19 (10%)
Quesada, 2020 50	150	67 (45%)	54±12	36 (24%)	75 (50%)	25 (17%)	90 (60%)	22 (15%)
Sade, 2009 53	65	27 (40%)	55±8	68 (100%)	37 (54%)	NA	35 (52%)	16 (24%)
Safdar, 2020 54	124	81 (65%)	51±11	91 (73%)	81 (65%)	42 (34%)	53 (43%)	20 (16%)
Sakamoto, 2012 55	73	12 (16%)	65±8	36 (49%)	33 (45%)	6 (8%)	17 (23%)	11 (15%)
Sara, 2016 56	926	281 (30%)	52±13	567 (61%)	371 (40%)	59 (6%)	485 (52%)	111 (12%)
Schindler, 2005 58	72	50 (69%)	58 _ 8	28 (39%)	50 (69%)	3 (4%)	30 (42%)	18 (25%)
Sicari, 2009 61	394	87 (22%)	61±10	223 (57%)	238 (60%)	69 (18%)	NA	120 (31%)
Suda, 2019 63	187	75 (40%)	63±12	74 (40%)	100 (54%)	52 (28%)	66 (35%)	52 (28%)
Taqueti, 2018 64	201	108 (54%)	66 (57–79)	130 (65%)	152 (76%)	129 (64%)	66 (33%)	16 (8%)
Uemura, 2016 65	61	16 (26%)	59±15	18 (30%)	37 (61%)	15 (25%)	NA	37 (61%)
Verna, 2018 66	101	45 (45%)	60±11	48 (48%)	58 (57%)	9 (9%)	53 (53%)	21 (21%)
Solberg, 2019 62	66	11 (17%)	54±9	66 (100%)	15 (23%)	2 (3%)	8 (12%)	44 (67%)
Schroder, 2019 59	174	49 (28%)	64±10	NA	NA	NA	NA	NA
Sara, 2019 57	129	49 (38%)	50±12	61 (47%)	NA	NA	NA	NA
Kumar, 2020 41	163	107 (66%)	57±12	79 (48%)	118 (72%)	37 (23%)	122 (75%)	30 (18%)
De Vita, 2019 34	30	18 (60%)	67±10	19 (63%)	19 (63%)	4 (13%)	16 (53%)	15 (50%1)
Mygind, 2016 45	54	20 (37%)	62±8	54 (100%)	29 (54%)	NA	34 (63%)	34 (63%)
Panza, 1997 46	66	13 (20%)	49±10	44 (67%)	NA	Na	NA	NA
Schroder, 2018 60	97	37 (38%)	62 (31–79)	97 (100%)	NA	NA	NA	NA
Reis, 1999 52	48	29 (60%)	54±10	48 (100%)	23 (48%)	6 (13%)	24 (49%)	NA
Kim, 2013 38	40	11 (28%)	53±11	NA	NA	NA	NA	NA
Ishimori, 2011 37	18	8 (44%)	41±11	18	NA	NA	NA	NA
Rahman, 2019 51	85	45 (53%)	57±10	66 (78%)	25 (29%)	11 (13%)	23 (27%)	12 (14%)
Konst, 2020 67	103	38 (37%)	62±9	NA	NA	NA	NA	NA

NA indicates information is not available.

The pooled prevalence of CMD was 0.41 (95% CI, 0.36–0.47; I 2=94%; Figure 4). In 18 studies, CMD prevalence were reported separately for men and women. In the meta‐regression analysis, there was no association between the proportion of women included in each study and prevalence of CMD. However, the risk of testing positive for CMD was 1.45 times greater than for men (Figure 5). The prevalence of CMD derived from invasive and noninvasive diagnostic methods was similar (0.43 [95% CI, 0.33–0.53] for invasive methods versus 0.42 [95% CI, 0.36–0.49] for noninvasive methods (P=0.993; Figure S2). Among noninvasive methods, a higher rate of CMD was found in patients who underwent positron emission tomography examination (0.46 [95% CI, 0.46–0.65]) compared with other noninvasive techniques (0.40 [95% CI, 0.30–0.55]; P=0.019).

Figure 4

Prevalence of coronary microvascular dysfunction.

The vertical black line indicates the pooled averaged prevalence rate estimate, and the red diamond represents the overall estimated prevalence with 95% CI in a random‐effects model. Gray squares indicate weighted‐point estimates of incidence for each single study, with gray horizontal lines indicating 95% CI. I 2 indicates Higgins index of heterogeneity. Pos indicates positive; and Tot, total.

Figure 5

Sensitivity analysis of prevalence of microvascular disease according to sex.

A, Forest plot illustrating the risk ratio (RR) and 95% CI of prevalence of coronary microvascular disease according to sex. B, Metaregression plot showing association between the prevalence of coronary microvascular resistance (y‐axis) and the proportion of women included in each study. The size of the bubble represents the number of patients included in each study. Neg indicates negative; and Pos, positive.

Sensitivity analyses addressing definitions of CMD based on different CFR thresholds (eg, abnormal CFR considered ≤2.5 or ≤2.0) found no significant difference in rate of CMD (0.43 [95% CI, 0.35–0.51] for CFR ≤2.5 versus 0.46 [95% CI, 0.33–0.60] for CFR ≤2.0 (P=0.986; Figure S3). A separate analysis including only studies with at least 200 patients, performed to prevent overestimation bias seen in small studies, found similar prevalence of CMD (0.42 [95% CI, 0.36–0.49]).

Vasospastic Angina

Twenty‐four studies investigating the presence of coronary vasospasm were included. They comprised 6553 patients; the mean age was 60.5±8.0 years, 39% were women, 21% had diabetes and 32% were smokers. Table 2 shows baseline and clinical characteristics of the patients undergoing coronary spasm investigations. Among studies investigating the presence of coronary vasospasm, 21 addressed epicardial spasm only, and 13 also reported the proportion of patients with microvascular spasm. The overall prevalence of coronary epicardial and microvascular spasm was 0.49 (95% CI, 0.43–0.56; I 2=96%; Figure 6). The prevalence of epicardial spasm was 0.40 (95% CI, 0.33–0.47; I 2=96%), whereas the prevalence of microvascular spasm was 0.24 (95% CI, 0.21–0.28; I 2=87%; Figure 7, Figure S4). For most of the patients, acetylcholine was used for the provocation test (98%), , , , , , , , , , , , , and 2 studies used ergonovine. , No significant difference was found considering the type of provocation test and prevalence of spasm 0.49 (95% CI, 0.38–0.55) for acetylcholine versus 0.48 (95% CI, 0.39–0.57) for ergonovine (P=0.935). In 12 studies, coronary spasm prevalence was reported separately for men and woman. The prevalence of coronary spasm was similar between sexes 0.28 (95% CI, 0.22–0.53) in women versus 0.25 (95% CI, 0.18–0.35) in men (Figure 8). From subgroup analyses considering different definitions of epicardial spasm (ie, based on ≥90% or ≥70% coronary vasoconstriction), no significant difference in rate of spasm was detected: 0.47 (95% CI, 0.35–0.50) for ≥90 constriction versus 0.49 (95% CI, 0.42–0.55) for ≥70% constriction (P=0.133).

Table 2

Number of Positive Patients and Baseline Clinical Characteristics of the Patients Included in the Studies Investigating the Prevalence of Vasospasm

Study	Patients included	No. positive, n (%)	Age, y	Women, n (%)	Hypertension, n (%)	Diabetes, n (%)	Dyslipidemia, n (%)	Current smoker, n (%)
Aziz, 2017 14	1379	813 (59%)	62±11.9	799 (58%)	970 (70%)	237 (17%)	841 (61%)	502 (36%)
Ford, 2018 33	151	56 (37%)	61 (53–68)	111 (74%)	NA	29 (19%)	120 (80%)	24 (16%)
Hoshino, 2016 15	292	90 (30%)	64±11	156 (51.7%)	114 (39%)	33 (11%)	98 (34%)	130 (45%)
Kim, 2018 16	328	128 (39%)	58±10.4	233 (71%)	128 (39%)	31 (9.4%)	72 (22%)	39 (12%)
Mohri, 1998 17	117	81 (74%)	63 (54–68)	59 (50%)	56 (48%)	26 (22%)	49 (42%)	50 (43%)
Montone, 2018 18	80	37 (46%)	63±11	40 (50%)	32 (40%)	8 (10%)	19 (24%)	17 (21%)
Montone, 2020 19	210	118 (56%)	62±11	82 (39%)	79 (38%)	13 (6%)	54 (26%)	27 (13%)
Oh, 2019 20	464	156 (34%)	57±11	164 (35%)	60 (13%)	23 (5%)	94 (20%)	48 (10%)
Ohba, 2012 21	370	264 (71%)	63±11	211 (57%)	197 (53%)	73 (20%)	193 (52%)	107 (29%)
Ong, 2014 23	847	488 (58%)	62±12	485 (57%)	609 (72%)	142 (17%)	460 (54%)	307 (36%)
Ong, 2012 22	124	77 (53%)	64±10	100 (%)	102 (71%)	31 (22%)	83 (58%)	22 (15%)
Ong, 2014 24	137	69 (50%)	63±11	93 (68%)	105 (77%)	27 (20%)	73 (53%)	38 (28%)
Pirozzolo, 2020 25	96	56 (58%)	65±12	49 (51%)	84 (88%)	15 (16%)	84 (88%)	25 (26%)
Quyyumi, 1992 26	51	5 (10%)	51±11	31 (61%)	20 (39%)	NA	NA	NA
Suda, 2019 63	187	126 (67%)	63±12	74 (40%)	100 (54%)	52 (28%)	66 (35%)	52 (28%)
Sun, 2002 29	55	14 (26%)	60±10	23 (42%)	26 (47%)	9 (16%)	26 (47%)	30 (55%)
Sun, 2005 28	131	101 (79%)	59±11	69 (53%)	59 (45%)	30 (13%)	50 (38%)	36 (27%)
Tsuchida, 2005 30	102	74 (77%)	57±11	15 (15%)	43 (42%)	31 (30%)	NA	82 (80%)
Uemura, 2016 65	61	15 (28%)	59±15	18 (30%)	37 (61%)	15 (25%)	NA	37 (61%)
Verna, 2018 66	101	57 (57%)	60±11	48 (48%)	58 (57%)	9 (9%)	53 (52%)	21 (20%)
Seitz, 2020 27	847	283 (33%)	64±11	529 (63%)	533 (63%)	129 (15%)	411 (49%)	260 (31%)
Yamanaga, 2015 31	50	29 (58%)	62±13	24 (48%)	28 (56%)	10 (20%)	29 (58%)	10 (20%)
Quesada, 2020 50	150	83 (55%)	54±12	36 (24%)	75 (50%)	25 (17%)	90 (60%)	22 (15%)
Hasdai, 1998 36	203	59 (29%)	51 [17–78]	158 (78%)	59 (29%)	8 (4%)	88 (43%)	28 (14%)

NA indicates information is not available.

Figure 6

Prevalence of coronary vasospasm.

The vertical black line indicates the pooled averaged prevalence rate estimate, and the red diamond represents the overall estimated prevalence with 95% CI in a random‐effects model. Gray squares indicate weighted‐point estimates of incidence for each single study, with gray horizontal lines indicating 95% CI. I 2 indicates Higgins index of heterogeneity.

Figure 7

Prevalence of coronary microvascular spasm.

 

Figure 8

Sensitivity analysis of prevalence of coronary vasospasm according to sex.

A, Forest plot illustrating the risk ratio (RR) and 95% CI of the prevalence of coronary vasospasm according to sex. B, Metaregression plot showing association between the prevalence of coronary vasospasm (y‐axis) and the proportion of women included in each study. The size of the bubble represents the number of patients included in each study. Neg indicates negative; and Pos, positive.

Combined Prevalence of CMD and Coronary Vasospasm

In 3 of the studies, , , patients underwent evaluation for CMD and spasm. Overall, 541 patients, with a mean age of 58±10.2 years and comprising 63% women, were included. The prevalence of CMD alone was 0.23 (95% CI, 0.10–0.45), coronary spasm alone (either epicardial or microvascular) 0.19 (95% CI, 0.10–0.33), and coexistent CMD and coronary vasospasm in 0.23 (95% CI, 0.17–0.31).

Discussion

The main findings of the present systematic review and meta‐analysis can be summarized as follows: (1) The proportion of patients with no obstructive coronary arteries presenting with CMD was 41%, whereas coronary spasm (epicardial and/or microvascular) was present in 49% of the cases. (2) Women are more likely than men to be affected by CMD. (3) Invasive and noninvasive diagnostic methods identified similar proportions of patients with CMD. (4) There was high heterogeneity between studies in the observed prevalence of CMD and vasospastic angina.

There is an increasing awareness among clinicians of the importance of microvascular function testing in patients with nonobstructive coronary arteries. , Murthy et al reported that even in the absence of obstructive coronary atherosclerosis, 53% of patients who present with chest pain have evidence of inducible myocardial ischemia. Moreover, it was shown that the presence of CMD identifies patients at increased risk of death and myocardial infarction. , The present meta‐analysis found that almost half of patients with no obstructive coronary arteries undergoing evaluation of the coronary microcirculation have CMD. Coronary function testing enables stratifying management of patients from different endotypes of ischemia with no obstructive CAD. Individualized treatment strategies are required, given the different pathophysiological mechanisms underlying these distinct disease endotypes. Objective evidence of the cause of chest pain and stratified therapy positively influence the quality of life of these patients. , Furthermore, identification of CMD or coronary spasm as the cause of symptoms prevents patients from undergoing repeated invasive diagnostic evaluations, which may reduce health care costs and allows for medical therapy optimization according to a specific diagnosis.

Coronary microvascular dysfunction has been deceivingly recognized as a women’s disease. The WISE (Women’s Ischemia Syndrome Evaluation) study demonstrated that 39% of women who present with chest pain and no obstructive CAD have evidence of induced myocardial ischemia and coronary vasomotor dysfunction. However, Murthy et al showed, using positron emission tomography, that CMD was highly prevalent in both sexes (51% in men versus 54% in women). The present meta‐analysis found that CMD is highly prevalent in both sexes; however, women are more likely to have CMD. , , An important fact to consider is that a substantial number of the studies did not evaluate men in a similar proportion to women.

Stratified Approach

The prevalence of CMD in patients with angina and no obstructive CAD undergoing invasive angiography depends on the methods and cutoffs applied. Assessment of invasive CFR was found to be the most‐used method for detecting CMD. However, it was derived mainly using a Doppler* or thermodilution technique. , , , In addition, some studies used a cutoff value of ≤2.5, , , , , , , , , whereas others used ≤2.0. , , , , The different methods and cutoffs may partially explain the high between‐study heterogeneity. However, we found that the prevalence of CMD was similar between methods and cutoffs. The recently published consensus document on diagnosis of CMD defined specific thresholds for identification of distinct endotypes of ischemia with no obstructive CAD. Here, CMD is defined as the presence of symptoms of myocardial ischemia, unobstructed coronary arteries (ie, diameter stenosis <50% or fractional flow reserve >0.80), and any of the following: index of microcirculatory resistance >25, CFR ≤2.0, and hyperemic microvascular resistance >1.9. Vasospastic angina, assessed with an acetylcholine provocation test, is considered positive for epicardial spasm when ≥90% diameter stenosis (compared with the angiography performed after nitrate administration) occurs with angina and ischemic ECG changes, whereas microvascular spasm is defined as the presence of angina and ischemic ECG changes without severe epicardial narrowing.

Despite the increasing awareness of CMD as a cause of chest pain, diagnostic methods to assess its presence remain underused. There are 2 main barriers to the widespread adoption of these methods in clinical practice. One refers to the limited availability of methods to diagnose CMD, such as positron emission tomography and invasive measurements. The second arises from the lack of effective medical therapies to treat CMD. Therefore, future research should focus on the evaluation of therapies to improve quality of life in patients with CMD. A breakthrough in this field would potentially facilitate the widespread adoption of CMD and vaso‐function testing in clinical practice.

Limitations

The main limitation of the present meta‐analysis is the lack of data on individual patients, which would have allowed for a standardization of CMD and coronary spasm definitions. Moreover, we observed a high level of heterogeneity between studies. The possibility of publication bias cannot be excluded (Figure S5). We were unable to identify specific variables leading to heterogeneity; however, this is most likely related to the inclusion criteria of each individual study and the difference between definitions of CMD and spasm that were used across the studies (Table S2). Another fact that should be accounted for is the possibility of false‐positive cases, especially in the studies with noninvasive imaging. , During the past years, more attention has been drawn to the fact that CFR is unable to define the pathophysiologic substrate for all cases of angina with no obstructive coronary arteries. It has been suggested that assessing the full range coronary pathophysiology requires concepts beyond CFR, such as regional size–severity quantification versus global perfusion and subendocardial perfusion on relative tomographic images.

Conclusions

In patients with no obstructive CAD, approximately half of the cases present with underlying disease, either CMD or coronary vasospasm. CMD is more prevalent in women; nonetheless, men are affected in a significant proportion. The large variability of methods, definitions, and thresholds for diagnosing these conditions is a call to a refinement and standardization of diagnostic tools. Greater awareness among physicians of ischemia with no obstructive coronary arteries is urgently needed for accurate diagnosis and patient‐tailored management.

Sources of Funding

None.

Disclosures

Dr Sonck reports research grants provided by the Cardiopath PhD program. Dr Berry is employed by the University of Glasgow, which holds research and/or consultancy agreements with AstraZeneca, Abbott Vascular, Boehringer Ingelheim, GSK, HeartFlow, Opsens, and Novartis. Dr Berry received research funding from the British Heart Foundation (RE/18/6134217). Dr De Bruyne has a consulting relationship with Boston Scientific, Abbott Vascular, CathWorks, Siemens, and Coroventis Research; receives research grants from Abbott Vascular, Coroventis Research, CathWorks, and Boston Scientific; and holds minor equities in Philips, Siemens, GE Healthcare, Edwards Life Sciences, HeartFlow, Opsens, and Celyad. Dr Collet reports receiving research grants from Biosensor, Coroventis Research, Medis Medical Imaging, Pie Medical Imaging, CathWorks, Boston Scientific, Siemens, HeartFlow, and Abbott Vascular; and consultancy fees from HeartFlow, Opsens, Abbott Vascular, and Philips. The remaining authors have no disclosures to report.

Figures S1–S5

Click here for additional data file.