Summary
Background
Despite improvement in revascularization strategies, microvascular obstruction (MO) lesions remain associated with poor outcome after ST-segment elevation myocardial infarction (STEMI).
Aims
To establish a bedside-available score for predicting MO lesions in STEMI, with cardiac magnetic resonance imaging (CMR) as the reference standard, and to test its prognostic value for clinical outcome.
Methods
Patients with STEMI of < 12 hours’ evolution treated by percutaneous coronary intervention (PCI) were included. CMR was performed 4–8 days later, to measure myocardial infarction (MI) extent, left ventricular ejection fraction (LVEF) and volumes, and to identify MO lesions. An MO score was built from multivariable logistic regression results and included clinical, angiographic and electrocardiographic criteria. Adverse cardiovascular events were recorded prospectively after STEMI.
Results
We analysed data from 112 patients. MO lesions were found in 63 (56%) patients and were associated with larger MI as assessed by higher peak creatine phosphokinase (3755 ± 351 vs 1467 ± 220 IU, p < 0.001), lower LVEF (46.7 ± 1.5 vs 53.4 ± 1.6%, p < 0.01) and larger MI extent (18.7 ± 1.2 vs 9.0 ± 1.3% LV, p < 0.001) on CMR. MO score > 4 accurately identified microcirculatory injuries (sensitivity 84%; specificity 82%) and independently predicted the presence of MO lesions on CMR. MO score > 4 predicted adverse cardiovascular events during the first year after STEMI (relative risk 2.60 [1.10–6.60], p = 0.03).
Conclusions
MO lesions are frequent in PCI-treated STEMI and are associated with larger MIs. MO score accurately predicted MO lesions and identified patients with poor outcome post-STEMI.
Résumé
Contexte
En dépit de l’amélioration des techniques de revascularisation, les lésions microvasculaires obstructives (LMO) grèvent le pronostic évolutif des syndromes coronariens aigus avec sus-decalage de ST (SCAST+).
Objectif
Établir un score prédictif de présence de LMO après revascularisation des SCAST+ en utilisant l’IRM cardiaque comme méthode de référence.
Méthodes
Les patients avec SCAST+ < 12 heures d’évolution et revascularisés par angioplastie coronaire étaient éligibles. Une IRM cardiaque était réalisée quatre à huit jours après revascularisation pour identifier les LMO, mesurer l’extension de l’infarctus et calculer les volumes et la fraction d’éjection du VG. Un score (MO score) a été construit sur la base des résultats d’une analyse multivariée incluant des facteurs cliniques, électriques et angiographiques. Les événements cardiaques défavorables ont été collectés dans le suivi.
Résultats
Cent douze patients ont été inclus dans l’étude. Les LMO ont été identifiées dans 56 % des cas et étaient associées à un infarctus plus étendu en IRM (18,7 ± 1,2 versus 9,0 ± 1,3 % VG, p < 0,001) et une FEVG plus basse (46,7 ± 1,5 versus 53,4 ± 1,6 %, p < 0,01). Une valeur de MO score supérieure à 4 prédisait efficacement la présence de LMO à l’IRM cardiaque (sensibilité 84 %; spécificité 82 %). De plus, un MO score supérieur à 4 était un prédicteur indépendant d’événements cliniques défavorables durant la première année après le SCAST+ (RR 2,6 [1,1–6,6], p = 0,03).
Conclusions
Les LMO sont fréquentes après angioplastie coronaire dans les SCAST+ et associées à un infarctus plus étendu. Un score simple base sur les paramètres cliniques, ECG et angiographiques prédit efficacement les LMO et identifie les patients à haut risque de complications.
Background
The restoration of coronary flow by transluminal coronary angioplasty has dramatically improved the prognosis of STEMI . However, the positive impact of advances in the quality of PCI has been limited by the persistence of coronary microcirculatory obstructive injuries that jeopardize optimal myocardial reperfusion.
MO injuries that develop during an acute myocardial infarction have been described as the “no-reflow” phenomenon, which remains a frequent, yet underestimated, complication during STEMI . MO lesions are associated with poor outcome in patients undergoing PCI, with increased rates of adverse cardiovascular events and an enhanced LV remodelling process . The development of microcirculatory injuries after PCI depends on multiple factors, including ischaemic and reperfusion injuries, but its precise pathophysiology remains partially unknown . Early detection of MO after PCI is still challenging, but it might help clinicians to identify patients at higher risk for complications .
Angiographic variables such as TIMI flow frame count or myocardial blush assessment provide partial information about quality of microcirculatory perfusion . ST-segment elevation resolution has been strongly correlated with outcome after MI but has not been specifically evaluated for microcirculatory injury detection . Previous groups reported the successful use of myocardial contrast echocardiography or intracoronary flow Doppler assessment/index of microcirculatory resistance to predict MO lesions. However, these latter methods might require specific devices and/or highly trained teams. Thus, physicians still need simple, accurate tools to promptly assess the success of reperfusion therapy in STEMI .
The aim of this study was to analyse the clinical, electrical and angiographic factors associated with microcirculatory injuries, and to establish a bedside score for prediction of MO at the acute stage of STEMI, using CMR imaging as the standard of reference. Moreover, we aimed to assess the prognostic value of this MO score for prediction of clinical outcome during the first year after STEMI.
Materials and methods
Patient selection
STEMI was defined as chest pain that persisted for ≥ 30 minutes associated with ST-segment elevation ≥ 0.1 mV in at least two contiguous electrocardiographic leads . Patients referred to our institution with STEMI, who underwent revascularization by primary or rescue PCI after thrombolysis failure, within the first 12 hours of evolution, were screened for inclusion. We excluded patients with cardiogenic shock, left bundle branch block on initial ECG, or severe chronic renal failure, and patients with contraindication to magnetic resonance imaging (claustrophobia, ferromagnetic foreign body, pacemaker, etc.).
Standard clinical and biological characteristics were assessed at baseline and creatine phosphokinase peak value was determined from consecutive blood sample measurements. Time to efficient revascularization (expressed in minutes) was calculated as the delay between onset of symptoms and completion of PCI. Overweight status was defined as a body mass index > 27 kg/m 2 . Creatinine clearance was calculated using the Cockroft-Gault method.
Our local ethics committee approved the study, which complied with the Declaration of Helsinki, and all patients gave their written consent before inclusion.
STEMI management
All patients included in the study were treated in accordance with international guidelines and the local protocols of our mobile care units . Thrombolytic therapy was given to patients when PCI could not be achieved within 2 hours after first medical contact, using tenecteplase (0.5 mg/kg) unless contraindicated. Rescue PCI was proposed in case of thrombolysis failure, defined as persistence of chest pain and ST-segment elevation regression of < 50%, 90 minutes after tenecteplase bolus .
Before catheterization, patients received aspirin 250 mg intravenously, a clopidogrel 300 mg bolus dose and an upstream abciximab loading dose (0.25 mg/kg of body weight), at the discretion of the physician. After coronary angiography, all patients (except those previously pretreated) received ad hoc abciximab as a 0.25 mg/kg bolus dose followed by a 12-hour infusion at 0.125 μg/kg/min, unless contraindicated. Heparin (initial bolus dose of 70 U/kg) or enoxaparin (initial bolus dose of 1 mg/kg) were administered during the procedure to achieve adequate anticoagulation goals. Immediately after the procedure, a clopidogrel loading dose of up to 600 mg was completed. Patients were treated routinely with aspirin (325 mg/day) and clopidogrel (75 mg/day). Angiotensin-converting enzyme inhibitors, beta-blockers and statins were administered according to current international guidelines . The culprit lesion was treated by stenting, that was performed directly, after balloon predilation or use of thrombectomy device, at the discretion of the operator. Pre- and post-PCI antegrade angiographic flows above the culprit lesion were reviewed retrospectively and classified according to the TIMI criteria by two operators who were unaware of the CMR results .
Electrocardiographic analysis
Standard 12-lead ECGs were registered before, then 90 minutes and 24 hours after revascularization, using a speed of 25 mm/s and amplitude of 10 mm/mV. ST-segment elevation was measured manually at 60 ms after the J point in three successive QRS-T complexes in all the derivations, and the sum of the ST-segment elevation (ST sum ) was calculated. The value of the ST-segment elevation in the single lead with the most prominent deviation (ST max ) was also measured. The percentage of resolution of both variables was assessed 90 minutes and 24 hours after PCI.
CMR analysis
Between day 4 and day 8 after PCI, all included patients underwent CMR by two investigators who were unaware of the clinical, electrocardiographic and angiographic findings. All CMR studies were carried out using a 1.5 T whole-body magnetic resonance scanner (Symphony TIM, Siemens, Erlangen, Germany), with a 12-element surface coil. Images were acquired in end-inspiration breath-hold and under ECG gating.
Two CMR sequences were used, as follows. Cine sequences (steady-state free precession) were acquired in short axis for global function analysis with the following parameters: repetition time/echo time, 40/1.8 ms; slice thickness, 6 mm; no gap between slice; flip angle, 65°; field of view, 350 × 350 mm; matrix, 148 × 256; temporal resolution, 40 ms. Delayed enhancement (inversion-recovery gradient-echo) sequences were acquired 10 minutes after intravenous administration of Gd-DOTA 0.2 mmol/kg, encompassing the LV in short-axis, four-chamber and two-chamber orientations. The sequences parameters were: repetition time/echo time, 780/1.56 ms; slice thickness, 6 mm; spacing, 0; flip angle, 10°; field of view, 350 × 400 cm; matrix, 152 × 134. The inversion time was 270–325 ms (set to null normal myocardial). End-diastolic phase was chosen to set the acquisition window.
All examinations were transferred to a computer workstation (Leonardo, Siemens, Erlanguen, Germany). Quantitative and qualitative analyses were performed by two independent observers blinded to other data. LV volumes (indexed to body surface area), LVEF and mass were measured according to Simpson’s rule.
The presence and extent of myocardial infarction and MO were analysed qualitatively and quantitatively, using a 17-segment LV model. Myocardial infarction was defined on delayed contrast images as a hyperintense area in the LV myocardial wall supplied by the infarct-related artery. Myocardial necrosis extent was defined qualitatively as subendocardial (extent < 50% of whole wall thickness in more than half of the segments showing myocardial enhancement) or transmural (extent > 50% in more than half of the segments showing myocardial enhancement). MO was defined on delayed contrast enhanced images as a hypointense zone within the hyperenhanced infarcted zone . The quantitative size of myocardial infarct and MO were computed by Simpson’s method . The extent of myocardial infarct and MO were expressed as a percentage of the total LV mass as follows: myocardial infarct extent (%) = (hyperenhanced area/LV area) × 100. MO extent (%) was computed as follows: hypoenhanced extent = (hypoenhanced area/LV area) × 100.
Follow-up
Clinical follow-up was carried out at clinical visits and/or through phone contact. We collected data prospectively regarding adverse cardiovascular events during the first year after STEMI. Cardiovascular adverse events were defined as cardiac death, acute congestive heart failure requiring hospital admission and intravenous infusion of diuretics, stroke, acute coronary syndrome and urgent PCI.
Statistical analysis
Statistical analysis was performed with SPSS 15.0 software for Windows (SPSS, Chicago, IL, USA). Data are expressed as mean and standard error of the mean and the normality of their distribution was assessed by the Kolmogorov-Smirnov test. Categorical and continuous variables were compared using the Chi 2 or Fisher’s exact tests and Student’s t test, respectively. Continuous variables were correlated using Pearson’s coefficients.
For MO score estimation, we first performed univariate analysis to identify variables potentially associated with the presence of MO lesions. Next, we estimated the independent predictors of MO using a multivariable binary logistic regression analysis, including the different variables identified by the univariate analysis ( α = 0.25). Interactions between variables were introduced into the model to take the potential colinearities into account. Logistic regression was used to perform multivariable analysis and gain an estimation of the variables for the MO score ( α = 0.05). Finally, we designed an MO score based on this logistic model in which clinical relevant variables were balanced by a user-friendly coefficient.
For quality assessment, the MO score was first validated by the CART method, with cross-validation based on the jack-knife algorithm . The CART methodology provides a hierarchical evaluation of prognostic variables. Firstly, it determines, for each variable, the cut-off values that optimize, in this instance, the identification of MO. Secondly, CART determines the optimal prognostic variable. This algorithm involves recursive subdivisions of a group of subjects on the basis of the choice of optimal cut-off points of the optimal prognostic variables, maximizing classical split criteria . This method has been successfully applied in various clinical cardiovascular research studies . We provided a cross-validation approach using the classical jack-knife algorithm . This sampling technique involves discarding one patient at a time and fitting a new model for the remaining patients ( N -1), examining the score for its effectiveness in prediction.
Patients’ classification after identification of MO lesions was used as a reference for receiver operating characteristic curve analysis. AUCs were compared using Hanley’s method and z statistics.
Patient survival of events was analysed and Kaplan-Meier curves were constructed. Differences between survival curves were evaluated using the log-rank test. Cox proportional hazards regression was used to identify independent correlates of 1-year occurrence of adverse events ( α = 0.05).
Results
Baseline characteristics of the patients
Between January 2006 and August 2008, 165 patients with STEMI were admitted to our institution, 112 of who fulfilled the inclusion criteria ( Fig. 1 ). Baseline characteristics of the patients are displayed in Table 1 .
| All patients | MO group | No MO group | p | |
|---|---|---|---|---|
| ( n = 112) | ( n = 63) | ( n = 49) | ||
| Men | 93 (83) | 54 (86) | 39 (79) | 0.39 |
| Age (years) | 57.8 ± 1.1 | 59.7 ± 1.4 | 56.1 ± 1.8 | 0.18 |
| TIMI risk score (STEMI) | 2.6 ± 0.2 | 3.1 ± 0.3 | 2.1 ± 0.3 | 0.01 |
| Cardiovascular risk factors | ||||
| Hypertension | 42 (38) | 22 (36) | 20 (41) | 0.56 |
| Diabetes | 19 (17) | 12 (19) | 7 (14) | 0.48 |
| Active smoking | 66 (59) | 31 (50) | 35 (71) | 0.02 |
| Body mass index (kg/m 2 ) | 26.6 ± 0.4 | 26.9 ± 0.4 | 26.1 ± 0.7 | 0.34 |
| Dyslipidaemia | 44 (42) | 19 (30) | 20 (40) | 0.27 |
| Previous history of CAD | 17 (15) | 6 (10) | 11 (22) | 0.06 |
| Time to efficient revascularization (min) | 205 ± 11 | 244 ± 17 | 154 ± 10 | < 0.001 |
| Thrombolysis failure | 10 (9) | 9 (15) | 1 (2) | 0.02 |
| Culprit lesion | ||||
| Left anterior descending artery | 56 (50) | 34 (54) | 22 (45) | 0.35 |
| Circumflex artery | 11 (10) | 5 (8) | 6 (12) | 0.44 |
| Right coronary artery | 45 (40) | 24 (38) | 21 (43) | 0.61 |
| Peak CPK (IU/L) | 2735 ± 243 | 3755 ± 351 | 1467 ± 220 | < 0.001 |
| Initial fibrinogen (g/L) | 3.78 ± 0.16 | 3.89 ± 0.25 | 3.72 ± 0.19 | 0.59 |
| Initial WBC count (/mm 3 ) | 11.390 ± 322 | 11.380 ± 430 | 11.400 ± 514 | 0.97 |
| Pharmacological management | ||||
| Upstream abciximab treatment | 31 (28) | 19 (31) | 12 (25) | 0.44 |
| Post-PCI abciximab infusion | 96 (88) | 54 (88) | 42 (88) | 0.87 |
| Coronary angiography findings | ||||
| Initial TIMI flow | 0.88 ± 0.1 | 0.76 ± 0.13 | 1.04 ± 0.17 | 0.19 |
| Post-PCI TIMI flow | 2.69 ± 0.05 | 2.52 ± 0.07 | 2.9 ± 0.05 | < 0.001 |
| Direct stenting | 57 (51) | 32 (51) | 25 (52) | 0.89 |
| Use of thrombectomy device | 29 (26) | 22 (35) | 7 (14) | 0.02 |
| One-vessel disease | 46 (41) | 29 (46) | 17 (35) | 0.26 |
| Two-vessel disease | 40 (36) | 22 (35) | 18 (37) | 0.82 |
| Three-vessel disease | 26 (23) | 12 (19) | 14 (29) | 0.21 |
| Presence of collateral vessels | 36 (32) | 15 (24) | 21 (42) | 0.10 |
| Electrocardiographic findings | ||||
| 90-minute ST sum regression (%) | 70.9 ± 2.6 | 61.5 ± 3.5 | 84.8 ± 2.8 | < 0.001 |
| 24-hour ST sum regression (%) | 71.6 ± 2.8 | 66.5 ± 3.9 | 80.5 ± 3.7 | 0.01 |
| 90-minute ST max regression (%) | 66.7 ± 2.6 | 57.6 ± 3.4 | 80.1 ± 3.5 | < 0.001 |
| 24-hour ST max regression (%) | 70.3 ± 2.5 | 63.4 ± 3.3 | 81.3 ± 3.1 | < 0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
