OUP user menu

Non-invasive coronary flow reserve after successful primary angioplasty for acute anterior myocardial infarction is an independent predictor of left ventricular adverse remodelling

Patrick Meimoun, Jacques Boulanger, Anne Luycx-Bore, Hamdane Zemir, Frederic Elmkies, Dorothée Malaquin, Luc Doutrelan, Christophe Tribouilloy
DOI: http://dx.doi.org/10.1093/ejechocard/jeq049 711-718 First published online: 8 April 2010


Aims To assess the usefulness of non-invasive coronary flow reserve (CFR) to predict left ventricular adverse remodelling (LVR) after ST-elevation myocardial infarction (STEMI).

Methods and results Sixty-five consecutive patients (mean age 58 ± 13 years, 24 women) with a first anterior STEMI, underwent prospectively CFR in the distal part of the left anterior descending artery (LAD), using intravenous adenosine infusion (0.14 mg/kg/min, within 2 min), and a standard echocardiography during the same exam, performed within 24 h after successful primary coronary angioplasty, and 6 months later, while the patients were in stable haemodynamic situation. CFR was defined as the peak hyperaemic LAD flow velocity divided by the baseline flow velocity. LV end-systolic volume (ESV) and end-diastolic volume (EDV), and LV ejection fraction (LVEF) were measured using the biplane Simpson's rule. LVR was defined as an absolute increase of ESV ≥15%. Compared with patients without LVR, patients with LVR (n = 18) had higher peak troponin T levels, wall motion score (WMS), a worse initial angiographic TIMI flow grade, and less improved electrocardiographic ST-segment resolution (all P < 0.05), and lower CFR (1.43 ± 0.2 vs. 1.97 ± 0.5, P < 0.01). At 6 months, patients with LVR had higher WMS, ESV, EDV, and lower LVEF compared with patients without LVR (all P < 0.01). Furthermore, acute CFR was significantly correlated to the 6-month LVEF and ESV, and to change of LVEF and ESV (all P < 0.01). In the multivariate analysis, acute CFR and initial angiographic TIMI flow grade were the independent predictors of LVR (all P ≤ 0.01). Receiver-operating characteristic curve analysis demonstrated that a cut-off value of 1.7 for CFR yields a sensitivity of 100% and a specificity of 62% to predict LVR at follow-up (P < 0.001, area under the curve 0.82).

Conclusion Non-invasive CFR is an independent predictor of LVR after successful primary angioplasty of anterior STEMI.

  • Coronary flow reserve
  • Myocardial infarction
  • Left ventricular remodelling


Left ventricular adverse remodelling (LVR) has been extensively described in last decades as a strong predictor of adverse events including cardiac death after acute ST-elevation myocardial infarction (STEMI).13 Despite improvement in STEMI management, LVR is still relatively frequent, complicating the course of ∼30% of anterior STEMI.4 The usefulness of various tools has already been tested in patients with acute myocardial infarction (AMI) to predict LVR.510 Despite successful recanalization of the infarct-related artery with an angiographic TIMI-3 flow after primary coronary angioplasty for STEMI, some patients have impaired microcirculatory reperfusion implying a lack of myocardial recovery.11 Therefore, it is not surprising that these patients are at increased risk for LVR and adverse events at follow-up.2,3,8 Non-invasive coronary flow reserve (CFR) is an attractive tool, easily available at bedside, to assess the coronary microcirculation in STEMI.12,13 Indeed, CFR assessed by transthoracic Doppler echocardiography (TDE) has already been used in various settings and compared favourably with the Doppler flow wire, nuclear cardiac imaging, and stress echocardiography.14 However, only one study which included few patients had assessed the usefulness of non-invasive CFR for the prediction of LVR.13 Besides, the independent prognostic value of CFR to predict LVR in STEMI has never been tested by comparison to clinical, biochemical, electrocardiogram (ECG), and angiographic parameters. Indeed, new acute reperfusion strategies have emerged last years that could have a positive impact on the coronary microcirculation, such as systematic stenting, the use of glycoprotein IIb–IIIa inhibitors, thrombus-aspirating device, and dual antiplatelet therapy.1517 Therefore, our objective was to assess the usefulness of non-invasive CFR to predict LVR in patients with successful primary angioplasty for anterior STEMI in the era of systematic use of antiremodelling medications such as angiotensin-converting enzyme (ACE)-inhibitors and beta-blockers.


Study population

Seventy consecutive patients with a first acute anterior STEMI successfully treated with primary coronary angioplasty (PCI) were prospectively included in this single-centre study. Some of these patients were included in a previous report to assess the predictors of in-hospital complications and LV recovery.12 In the current study, all the patients underwent non-invasive CFR and a standard transthoracic echocardiography within 24 h after coronary angioplasty (23 ± 3 h) and 6 months later.

Patients were included if they had a first STEMI, left anterior descending artery (LAD) revascularization no more than 12 h after the onset of chest pain, successful PCI with stent implantation with a final angiographic TIMI-3 flow without residual stenosis, no contraindication to adenosine, no severe valvular disease, and stable haemodynamic situation at the time of testing (no cardiogenic shock). All patients gave inform consent to participate to the protocol.

The diagnosis of STEMI was based on chest pain lasting >30 min, ST-segment elevation >2 mm in at least two contiguous precordial ECG leads, and an increase in serum troponin T (normal < 0.05 µg/L in our hospital), and/or creatine kinase to more than three-fold the normal value. ECGs performed on admission to hospital and 60 min after primary angioplasty in the coronary care unit were used for ECG analysis. The sum of ST-segment elevation was measured 20 ms after the end of the QRS complex in leads I, aVL, and V1–V6, before (sum ST1) and after angioplasty (sum ST2). Besides the quantitative analysis, resolution of ST-segment elevation was classified into three categories according to previous studies:18 complete ST-segment elevation resolution (≥70% resolution), partial resolution (between 30 and 70% resolution), and no resolution (≤30% resolution). All ECGs were analysed by an investigator who was unaware of the patient data.

Coronary angiography and angioplasty

All patients received aspirin (250–500 mg), clopidogrel (300–600 mg), and intravenous bolus of heparin (5000 U) before angiography. Diagnostic coronary angiography was performed using the radial or femoral approach. Coronary angioplasty was performed by standard techniques, and after conventional wire crossing, stent implantation of the infarct-related vessel was performed, preceded by balloon pre-dilatation if necessary. During the procedure, glycoprotein IIb/IIIa inhibitors, thrombectomy (export Medtronic), and intracoronary vasodilators (adenosine and verapamil) were used at the discretion of the interventional cardiologist. The angiographic TIMI flow grade was assessed according to previous studies,19 before and after angioplasty. Collateral grading was done according to Rentrop grading system.20 After the procedure, all the patients received medical therapy according to the current guidelines for STEMI.21

Non-invasive coronary flow reserve

Non-invasive CFR was performed as previously described, in the mid-distal part of the LAD,22,23 using intravenous adenosine infusion (140 µg/kg/min over 2 min), with the commercially available machine (Acuson, Sequoia 256, Mountain View, CA, USA, and Vivid E9 system, General Electrics). The artery was visualized by colour Doppler flow mapping guidance, with a velocity range defined from 12 to 19 cm/s. CFR was calculated as the ratio of hyperaemic to basal peak diastolic flow velocity. A no-reflow pattern of the baseline flow velocity was defined as previously described with a deceleration time of diastolic flow velocity ≤600 ms and/or systolic flow reversal.24 Blood flow velocity measurements were performed offline by an experienced investigator blinded to patient data, by contouring the spectral Doppler signals, using the integrated software package of the ultrasound system. Final values of flow velocity represented an average of three cardiac cycles. The inter- and intra-observer variabilities for CFR measurements in our experience have been previously reported (around 5 and 4%, respectively).22,23

Transthoracic Doppler echocardiography

Serial TDE measurements were performed during the same exam as the CFR evaluation,25 with the patient in the left lateral decubitus position. LV end-diastolic volume (EDV) and end-systolic volume (ESV), and left atrial volume were measured from the apical four- and two-chamber views, and LV ejection fraction (LVEF) was calculated from the modified biplane Simpson's rule. Wall motion score index (WMSI) was measured using the 16-segment model.26 Infarct zone wall motion score index (IZWMSI) was obtained using the nine segments assigned to the LAD territory.12 LV mass index was measured according to the ASE formula.27 Conventional Doppler parameters were also measured according to a standardized examination: early (E), and late (A), diastolic transmitral flow velocity, deceleration time of E (DTE), average of the septal and lateral annulus mitral early diastolic (Ea), late diastolic (Aa), and systolic (Sa) tissue velocity, and the ratio E/Ea. The pulmonary artery systolic pressure was calculated as previously described,28 adding right atrial pressure (from respiratory variation of inferior vena cava diameter). All echocardiograms were digitized online on optical discs for subsequent offline analysis by two other experienced observers blinded to patient data.

Adverse LV remodelling was defined as an absolute increase in the ESV ≥15% at follow-up.


Continuous variables were expressed as mean ± SD and categorical variables as percentages. When a variable was not normally distributed, a logarithmic transformation was performed [troponin peak, creatine kinase peak, and N-terminal pro-brain natriuretic peptide (NT-proBNP)]. Unpaired or paired Student's t-test and the χ2 test (or Fisher's exact test or direct comparisons of proportions as appropriate) were performed to assess differences according to the variables tested. To assess the relationship between acute CFR and ESV, LVEF, at follow-up, and their change, biochemical markers, ECG ST-segment scores, linear and non-linear correlations were tested and the best fit was retained. A stepwise multiple logistic regression analysis was performed to identify the independent predictor(s) of LVR. A variable with P< 0.1 in the univariate analysis was included in the multivariate model. For the independent predictors of LVR in the multivariate analysis, optimal threshold values were identified via the receiver-operating characteristic (ROC) curve analysis. Intra- and inter-observer variabilities of LVEF and ESV were tested in a random sample of 15 patients. For LVEF, the variabilities were 4 ± 2 and 5 ± 2%, respectively. For ESV, the respective values were 5 ± 2 and 6 ± 2%, respectively.

Statistical analysis was performed using MedCalc for Windows, version (MedCal Software, Mariakerke, Belgium). A P-value <0.05 was considered as significant.


From the 70 patients enrolled in the study, 65 had serial echocardiography evaluation for LVR assessment; 1 patient still alive refused the follow-up study, 2 were lost to follow-up, and 2 had adverse events at follow-up and could not undergo repeat echocardiography. Baseline characteristics of the study population are listed in Table 1.

View this table:
Table 1

Baseline clinical, biochemical, and angiographic characteristics

N = 65LVR− (n = 47)LVR+ (n = 18)
Age (years)58 ± 1257 ± 1359 ± 10
Gender, women, n (%)24 (37)19 (40)5 (28)
BMI (kg/m2)27.3 ± 4.527.5 ± 4.626.7 ± 4.5
Hypertension, n (%)27 (41.5)19 (40)8 (44)
Diabetes, n (%)18 (28)11 (23)7 (39)
Smoking, n (%)28 (43)19 (40)9 (50)
Family history, n (%)12 (18.5)9 (19)3 (17)
Dyslipidaemia, n (%)35 (54)23 (49)12 (67)
Pre-infarction angina, n (%)26 (41)18 (38)8 (44)
Killip's class I/≥II (%)83/1787/1372/28
SBP/DBP (mmHg)136 ± 21/ 80 ± 15136 ± 20 / 78 ± 13138 ± 24/83 ± 19
Heart rate (bpm)76 ± 1975 ± 1778 ± 22
LDL cholesterol (g/L)1.27 ± 0.441.3 ± 0.41.1 ± 0.4
CK peak (log)3.15 ± 0.443.01 ± 0.43.47 ± 0.25*
Troponin T peak (log)0.60 ± 0.50.44 ± 0.50.97 ± 0.3*
Creatinine (µm/L)86 ± 2285 ± 2288 ± 23
NT-proBNP (log)3.23 ± 0.433.18 ± 0.453.34 ± 0.35
Hb (g/dL)14.6 ± 1.714.5 ± 1.615 ± 1.8
Leucocytes count (103/mm3)11.8 ± 311.5 ± 3.112.5 ± 2.7
Fasting glucose (mmol/L)6.96 ± 2.76.6 ± 2.97.6 ± 2.1
Sum ST1 (mm)12.1 ± 7.511.4±713.5 ± 6
Sum ST2 (mm)4.4 ± 3.33.46 ± 36.7 ± 3.2*
≥70% ECG-ST resolution, n (%)30 (46)24 (51)6 (33)
Time to reperfusion (h)4.9 ± 2.65 ± 2.74.6 ± 2.5
Initial TIMI flow grade
 0/1 (%)26 (40)/3 (5)12 (26)/3 (6)14 (78)/0*
 2/3 (%)21 (32)/15 (23)19 (40)/13 (28)2 (11)/2 (11)
Proximal LAD, n (%)36 (55.4)24 (51)12 (67)
Mid-LAD, n (%)29 (44.6)23 (49)6 (33)
Thromboaspiration, n (%)26 (40)14 (30)12 (67)*
Pre-dilatation, n (%)17 (26)10 (21)7 (39)
Anti-IIb/IIIa, n (%)42 (65)28 (60)14 (78)
Bare metal stent, n (%)45 (69)33 (70)12 (67)
Drug-eluting stent, n (%)12 (18.5)7 (15)5 (28)
Bioactif stent, n (%)8 (12)7 (15)1 (6)
Single-vessel disease, n (%)44 (68)32 (68)12 (67)
Two-vessel disease, n (%)15 (23)11 (23)4 (22)
Three-vessel disease, n (%)6 (9)4 (9)2 (11)
  • BMI, body mass index; family history of coronary artery disease; SBP/DBP, systolic blood pressure/diastolic blood pressure at presentation; CK, creatine-kinase; LAD, left anterior descending artery.

  • *P ≤ 0.01.

Left ventricular adverse remodelling

From the 65 patients who were followed up, 18 (28%) developed LVR. Characteristics of patients with and without LVR are summarized in Table 1. Troponin and CPK peak were significantly higher in patients with LVR vs. patients without LVR (P ≤ 0.02). After angioplasty, a less improvement was seen in patients with LVR when compared with patients without LVR, for sum ST2 (P < 0.01), but not for ≥70% ST resolution (P = 0.16). The only angiographic parameters which showed a significant difference between patients with and without LVR were the initial angiographic TIMI flow grade, more severely impaired in the former group, and thromboaspiration, less frequently used in the latter group (all P < 0.01).

Table 2 summarizes the cardiac drugs taken at follow-up. No significant difference was seen between patients with and without LVR, except aldosterone antagonists more frequently used in the former group. Regarding the acute echocardiography–Doppler parameters, no significant difference was seen between patients with and without LVR for ESV, EDV, LVEF, DTE, and E/Ea, at the acute phase (all P = NS) (Table 3). However, patients with LVR had, at the acute phase, worse regional wall motion abnormalities (WMSI and IZWMSI, all P < 0.01). Not surprisingly, patients with LVR at follow-up had a worse LVEF, EDV, ESV, WMSI, and IZWMSI, when compared with patients without LVR, with a mean per cent change of ESV of +39 vs. −29% (all P < 0.001). Furthermore, the former group had a more severely impaired diastolic function at follow-up (Table 4). Adenosine infusion was well tolerated and CFR was successfully performed in all patients with a mean value at the acute phase of 1.82 ± 0.52. The acute CFR was significantly more severely impaired in patients with LVR compared with patients without LVR (P < 0.01). The hyperaemic flow velocity was slightly but not significantly lower in patients with LVR when compared with patients without LVR. In contrast, the baseline flow velocity was significantly higher in the former group (Table 3). The proportion of a no-reflow pattern was of borderline significance between groups (P = 0.12).

View this table:
Table 2

Cardiac medications at follow-up

Total (n = 65)LVR+ (n = 18)LVR− (n = 47)
Aspirin, n (%)65 (100)18 (100)47 (100)
Clopidogrel, n (%)65 (100)18 (100)46 (98)
Statin, n (%)62 (95)17 (94)45 (96)
Beta-blocker, n (%)63 (97)17 (94)46 (98)
ACE/ARA II inhibitors, n (%)61 (94)17 (94)44 (94)
Aldosterone antagonist, n (%)16 (25)9 (50)*7 (15)
  • ACE, angiotensin-converting enzyme inhibitor; ARA II, angiotensin type II receptor blocker.

  • *P < 0.05.

View this table:
Table 3

Echocardiography–Doppler parameters at the acute phase

LVR− (n = 47)LVR+ (n = 18)P-value
Acute CFR1.97 ± 0.541.43 ± 0.17<0.001
Baseline LAD flow velocity (cm/s)33 ± 1341 ± 150.04
Hyperaemic LAD flow velocity (cm/s)62 ± 2057 ± 19NS
No reflow pattern, n (%)10 (21)8 (44)0.12
EDV (mL)107 ± 20105 ± 20NS
EDV/m2 (mL/m2)56 ± 855 ± 9NS
ESV (mL)59 ± 1458.5 ± 12NS
ESV/m2 (mL/m2)31 ± 731 ± 6NS
LVEF (%)44 ± 644 ± 6NS
LV mass/m2 (g/m2)92 ± 1999 ± 35NS
E (cm/s)62.5 ± 1869.5 ± 22NS
E/A0.96 ± 0.31.01 ± 0.5NS
Ea (cm/s)7.6 ± 2.37.4 ± 2.2NS
E/Ea9.1 ± 3.410.2 ± 4.9NS
DTE (ms)180 ± 42170 ± 38NS
WMSI1.74 ± 0.231.91 ± 0.14<0.01
IZWMSI2.29 ± 0.42.5 ± 0.2<0.01
Number of LV segments6.2 ± 1.57.2 ± 1.1<0.01
LA vol/m223 ± 523 ± 5NS
MR (mild/moderate)13 (28%)/0%7 (39%)/1 (6%)NS
  • LAD, left anterior descending artery; EDV, end-diastolic volume; ESV, end-systolic volume; DTE, deceleration time of mitral E-wave; WMSI, wall motion score index; IZWMSI, infarct zone wall motion score index; LA vol/m2, left atrial volume index; MR, mitral regurgitation.

View this table:
Table 4

Echocardiography–Doppler parameters at follow-up

LVR− (n = 47)LVR+ (n = 18)P-value
CFR3 ± 0.72.4 ± 0.4<0.001
EDV (mL)106 ± 24140 ± 29<0.001
EDV/m2 (mL/m2)56 ± 1174 ± 15<0.001
ESV (mL)45 ± 1581 ± 22<0.001
ESV/m2 (mL/m2)24 ± 743 ± 11<0.001
LVEF (%)58 ± 842 ± 6<0.001
E (cm/s)71 ± 1978 ± 27NS
E/A1.3 ± 0.541.6 ± 0.99NS
Ea (cm/s)9.5 ± 2.57.3 ± 2.2<0.001
E/Ea7.7 ± 3.711.2 ± 4<0.01
DTE (ms)209 ± 52199 ± 51NS
WMSI1.3 ± 0.21.86 ± 0.2<0.001
IZWMSI1.5 ± 0.42.47 ± 0.3<0.001
MR (mild/moderate)9 (19%)/0%9 (50%)/3 (17%)<0.001
LA vol/m225 ± 532 ± 9<0.001
ΔLVEF14 ± 6−2 ± 7<0.001
ΔESV−16 ± 1023 ± 17<0.001
ΔEDV−2 ± 1335 ± 25<0.001
  • EDV, end-diastolic volume; ESV, end-systolic volume; DTE, deceleration time of mitral E-wave; WMSI, wall motion score index; IZWMSI, infarct zone wall motion score index; MR, mitral regurgitation; LA vol, left atrial volume; ΔESV and ΔEDV, change of ESV and EDV between the acute phase and follow-up.


At the acute phase, baseline LAD flow velocity was significantly correlated to ECG parameters: sum ST2 and % ST-segment resolution (r = 0.34 and 0.4 respectively, all P < 0.01). Acute CFR was significantly correlated to troponin peak (log) (r = −0.63, P < 0.001), to acute WMSI (r = −0.41) and IZWMSI (r = −0.37) (all P < 0.01), and to sum ST2 (r = −0.47, P < 0.001). Furthermore, there was a significant inverse correlation between acute CFR and LVEF (r = 0.63), and ESV (r = −0.44), at follow-up (all P < 0.001), and between acute CFR and change of ESV (r = −0.56) (Figure 1), and LVEF (r = 0.66) (all P < 0.001).

Figure 1

Scatter plots of the significant curvilinear relationship between acute coronary flow reserve and per cent change of end-systolic volume.

In the multivariate analysis, the independent predictors of LVR were the acute CFR [odds ratio (OR) 0.06, 95% confidence interval (CI) 0.001–0.37, P = 0.01] and the initial angiographic TIMI flow grade (OR 0.25, 95% CI 0.1–0.59, P = 0.002). An ROC curve analysis demonstrated that a cut-off value of acute CFR of 1.7 was the best predictor of LVR, with a sensitivity of 100% (95% CI 82–100) and a specificity of 62% (95% CI 51–76) (area under the curve = 0.82, Z-statistic = 6.065, P < 0.001). By comparison, the initial angiographic TIMI flow grade = 0 had a sensitivity of 78% (95% CI 52–93) and a specificity of 74% (95% CI 60–86) to detect LVR (area under the curve = 0.74, Z-statistic = 3.94, P < 0.01) (Figure 2).

Figure 2

Superimposed receiver-operating characteristic curve analysis of acute coronary flow reserve and initial angiographic TIMI flow grade for predicting adverse left ventricular remodelling.

Patients with low coronary flow reserve

All the patients with LVR at follow-up had a low CFR at the acute phase (n = 18, sensitivity = 100%), but some patients with low CFR will not have adverse LVR at follow-up (n = 18). Baseline characteristics, sum ST2 ECG and ischaemic time, were not significantly different between these patients, and troponin peak (log) was of borderline significance (0.74 ± 0.4 vs. 0.97 ± 0.3, P = 0.056), However, the acute WMSI was more severely impaired in the subgroup of patients with LVR when compared with patients without LVR and low CFR (1.91 ± 0.14 vs. 1.77 ± 0.23, P = 0.03), and the initial angiographic TIMI 0 flow grade was more frequently observed in the former group (14/18 = 78% vs. 3/18 = 17%, P < 0.01). When combining a low CFR <1.7 with an initial angiographic TIMI flow grade = 0, the specificity for detecting LVR increased to 94% (Table 5). Only 7 (11%) patients were misclassified using the combined criteria when compared with 18 (28%) using CFR alone and 16 (25%) using the initial angiographic TIMI flow grade alone. A pair-wise comparison of ROC curve demonstrated that the combination of CFR and initial angiographic TIMI flow grade was superior to initial angiographic TIMI alone for predicting LVR (delta ROC = 0.11, 95% CI 0.01–0.22, P = 0.02). The combination of acute WMSI with either CFR or initial angiographic TIMI flow grade did not improve the accuracy to detect LVR.

View this table:
Table 5

Value of coronary flow reserve, initial angiographic TIMI flow grade, and their combination to predict adverse left ventricular remodelling

N = 65LVR+No LVR
CFR < 1.7a1818
CFR ≥ 1.7a029
TIMI = 0b1412
TIMI > 0b435
CFR < 1.7 and TIMI = 0c143
CFR ≥ 1.7 or TIMI > 0c444
  • Se, sensitivity; Sp, specificity; PPV, positive predictive value; NPV, negative predictive value.

  • aSe = 100%, Sp = 62%, PPV = 50%, NPV = 100%, accuracy = 72%.

  • bSe = 78%, Sp = 74%, PPV = 54%, NPV = 90%, accuracy = 75%.

  • cSe = 78%, Sp = 94%, PPV = 82%, NPV = 92%, accuracy = 89%.


In patients with successful epicardial reperfusion after first acute anterior myocardial infarction, non-invasive CFR is an independent predictor of LVR at follow-up. There is a strong correlation between acute CFR and, ESV and LV systolic function at follow-up. There is also a strong correlation between acute CFR and, ECG parameters of myocardial reperfusion, biochemical markers of amount of necrosis. A cut-off of 1.7 of acute CFR could predict LV remodelling with a sensitivity of 100% and a specificity of 62%.

Predictors of left ventricular remodelling and comparison with previous studies

Despite its widespread application in various settings, non-invasive CFR was used in only one study including 31 patients to predict LVR.13 This study included several patients (16%) with a final TIMI-2 or lower flow, which from a pathophysiological point of view means incomplete epicardial revascularization or microvascular obstruction,29 and implies a poorer prognosis.19 All our patients had successful primary angioplasty with a final angiographic TIMI-3 flow and no residual stenosis. Therefore, CFR is a good surrogate of microvascular function, not redundant with angiographic variables, giving an additional prognostic value in this setting. Apart from the final perfusional status of the infarct-related artery, various factors could contribute to LV remodelling, at different stages, after reperfused AMI. Non-invasive CFR remained an independent predictor of LVR after adjusting for parameters such as the proximal or mid-LAD location of the infarct, the pattern of LV filling, as assessed by the mitral inflow by Doppler echocardiography,6 or the ratio E/Ea, the infarct size, as reflected by the IZWMS and the extent of enzymatic and troponin release, and heart failure at admission, as reflected by the Killip class.9 Finally, the transmural extent of necrosis, which is best assessed by magnetic resonance imaging, is an important factor of LV remodelling.30 We did not systematically perform magnetic resonance imaging in our patients. However, a recent study31 showed that ischaemic time is an independent predictor of transmural extent of necrosis. In the present study, time from symptom onset to reperfusion was not significantly different in patients with and without LVR, and even after adjusting for this parameter, CFR remained an independent predictor of LVR. The independent predictive value of CFR was not assessed in the study of Rigo et al.13 However, this study demonstrated that early assessment of CFR, within 24 h after primary angioplasty, was better correlated to LVR at follow-up than later assessment in the course of the disease.

Baseline coronary flow velocity

Baseline acute LAD flow velocity was slightly higher in the group of patients with LVR when compared with patients without LVR. Baseline haemodynamic parameters, LV mass index, ageing, and cardiac medications, which could influence baseline LAD flow velocity,14 did not differ in each subgroup. Furthermore, at the acute phase, EDV, the ratio E/Ea, and DTE, which reflect LV preload and could also influence baseline LAD flow velocity, were not significantly different in each subgroup. Maybe this difference in baseline LAD flow velocity was due to a persistent autoregulation disturbance of the coronary microcirculation after the episode of ischaemia–reperfusion due to the infarct. Spontaneous or angioplasty-induced distal embolization could have also play a role for explaining this difference in baseline LAD flow velocity. However, the pre-dilatation, a supplier of distal embolization, rarely used, was not significantly different in each subgroup, as the intracoronary vasodilator employ. Interestingly, we found a significant positive correlation between baseline LAD flow velocity and ECG parameters after successful primary angioplasty (a surrogate of microcirculatory disturbance).

Value of coronary flow reserve to predict left ventricular adverse remodelling by comparison to initial angiographic TIMI flow grade

Although the sensitivity of CFR to detect LVR was very high (100%), its specificity was relatively low (62%), implying that a substantial proportion of patients (18/47, 38%) will not have LVR, despite a very impaired acute CFR. As microvascular impairment in STEMI is a dynamic process, serial measurements of CFR could have improved the specificity of the test. However, the best time to perform CFR after reperfused STEMI is not established, and the combination of CFR with the initial angiographic TIMI flow grade improved the specificity to detect the patients at high risk for LVR at follow-up. Indeed, only 3 of 47 (6%) patients without LVR had low CFR and initial angiographic TIMI flow = 0 (specificity = 94% to detect LVR).


We did not perform a systematic coronary angiography at follow-up to assess the patency of the LAD, and therefore, some patients could have suffered a silent restenosis or reocclusion influencing the rate of LVR. However, all the patients underwent non-invasive CFR in the distal part of the LAD at follow-up, and it was >2 in all but four cases (two in each subgroup, with and without LVR), strongly suggesting not only the patency of the vessel but also the lack of significant stenosis upstream of the vessel.14 Besides, coronary stenting of the infarct-related artery has been performed in all patients and results in extremely high long-term patency rates.32

Our results do not necessarily apply to all STEMI populations. The prognostic value of CFR in our study was assessed in patients with first anterior STEMI, without haemodynamic compromise, and they were treated with primary percutaneous angioplasty and successful LAD recanalization with a final angiographic TIMI-3 flow grade and a majority of single-vessel disease. Whether LVR in our study translates into a worse clinical outcome has not been determined, given the short follow-up of the study and the relatively few number of patients. However, LVR is a strong predictor of prognosis according to previous reports with longer follow-up and more patients.2,3

We have chosen an arbitrary value of ≥15% increase in ESV to define LVR (i) in order to be clearly above the variability of ESV measurement and (ii) to have a potential clinical meaning. Indeed, this cut-off was chosen in several studies demonstrating the benefit of resynchronization in heart failure.

LVEF, ESV, and EDV measured by the Simpson's method are affected by several factors including load dependency and observer variability and are based on an assumption of symmetric LV geometry which is not necessarily true in STEMI where regional LV dysfunction alters LV geometry. The assessment of WMSI (and IZWMSI) is less affected by these limitations and seems more sensitive to detect subtle differences between patients with and without LVR, as seen in our study at the acute phase. Finally, the assessment of LV volumes later at the acute phase, in larger sample groups, might have influenced the results.

Clinical implications

Despite improvement in STEMI management including all patients with a final angiographic TIMI-3 flow and the use of ACE-inhibitors and beta-blockers in nearly 100% of cases, we found as previous studies around 30% of LVR in our study. As LVR is a strong prognostic parameter after STEMI, non-invasive CFR could give important prognostic information in this setting, easily, without sophisticated rules, or radiation exposure. A more aggressive approach should be done in a patient with a very low CFR at the acute phase of STEMI, particularly if the initial angiographic TIMI flow grade = 0. For instance, the systematic use of aldosterone antagonist—an inhibitor of cardiac adverse remodelling—should be tested, before the occurrence of heart failure.

In conclusion, non-invasive CFR performed within 24 h after primary angioplasty of acute anterior STEMI is an independent predictor of LVR. Its combination with an initial angiographic TIMI 0 flow grade could dramatically improve the diagnostic accuracy to detect the patients at high risk for LVR.

Conflict of interest: none declared.


View Abstract