Right ventricular (RV) dysfunction is a powerful risk marker after acute myocardial infarction (MI). Primary percutaneous coronary intervention (PCI) has markedly reduced myocardial damage of the left ventricle, but reliable data on RV damage using cardiac magnetic resonance imaging (MRI) are scarce. In a recent trial of patients with acute MI treated with primary PCI, in which the primary end point was left ventricular (LV) ejection fraction after 4 months measured with MRI, we conducted a prospectively defined substudy in which we examined RV function. RV ejection fraction (RVEF) and RV scar size were measured with MRI at 4 months. Tricuspid annular plane systolic excursion (TAPSE) and RV free wall longitudinal strain (FWLS) were assessed using echocardiography before discharge and at 4 months. We studied 258 patients without diabetes mellitus; their mean age was 58 ± 11 years, 79% men and mean LV ejection fraction was 54 ± 8%. Before discharge, 5.2% of patients had TAPSE <17 mm, 32% had FWLS > −20% and 11% had FWLS > −15%. During 4 months, TAPSE increased from 22.8 ± 3.6 to 25.1 ± 3.9 mm (p <0.001) and FWLS increased from −22.6 ± 5.8 to −25.9 ± 4.7% (p <0.001). After 4 months, mean RVEF on MRI was 64.1 ± 5.2% and RV scar was detected in 5 patients (2%). There was no correlation between LV scar size and RVEF (p = 0.9), TAPSE (p = 0.1), or RV FWLS (p = 0.9). In conclusion, RV dysfunction is reversible in most patients and permanent RV ischemic injury is very uncommon 4 months after acute MI treated with primary PCI.
Right ventricular (RV) dysfunction is a powerful predictor of outcome after acute myocardial infarction (MI). Most studies on RV injury after acute MI was performed in the thrombolysis era. Despite the current large-scale use of primary percutaneous coronary intervention (PCI), RV dysfunction after acute MI remains common and associated with worse prognosis. In most studies, results are based on conventional echocardiography early after MI, whereas the RV frequently recovers from ischemic injury in the post-MI period. In the current era of primary PCI, the degree of left ventricular (LV) damage is increasingly limited. In a recent study on the potential value of metformin in acute MI, it was demonstrated that mean left ventricular ejection fraction (LVEF) on magnetic resonance imaging (MRI) 4 months after PCI was 54%. Any effects on RV functional parameters of this early intervention, in such cohort of patients with large myocardial areas at risk, were not reported. Since the RV is relatively protected against ischemic injury, we hypothesize that RV systolic function remains preserved after early PCI. In the present study, we therefore aimed to evaluate RV function after acute MI treated with PCI using both MRI and echocardiography.
Methods
The design and results of the Glycometabolic Intervention as Adjunct to Primary Percutaneous Coronary Intervention in ST-Segment Elevation Myocardial Infarction (GIPS) III trial have been published previously. In brief, the GIPS-III was a prospective, double-blind, randomized placebo-controlled trial, in patients who underwent primary PCI for ST-segment elevation MI (STEMI). The trial was designed to study the effect of metformin. All patients had an electrocardiographic recording with ST-segment elevation of more than 0.1 mV in 2 or more leads and were treated with at least 1 stent sized ≥3.0 mm. Major exclusion criteria were: previous MI, diabetes, and need for coronary artery bypass grafting. In the GIPS-III clinical trial, 380 patients were randomized 1:1 to receive metformin or placebo. The primary efficacy measure was LVEF measured by MRI 4 months after PCI. The GIPS-III trial demonstrated no significant difference in LVEF. All patients underwent additional echocardiographic assessment before discharge and 4 months. The study complies with the Declaration of Helsinki and was approved by the local ethics committee (Groningen, the Netherlands). All patients included in the GIPS-III trial provided written informed consent.
MRI studies were performed on 3.0-T scanner (Achieva, Philips, Best, The Netherlands). Electrocardiogram-gated cine steady state, free precession images were acquired with breath holding in contiguous short-axis slices covering the entire RV and LV (TR 3.9 ms; TE 1.94 ms; flip angle 45°; matrix 180 × 220; voxel sixe 1.25 × 1.25 × 10 mm; slice thickness 10 mm; and slice gap 0 mm). Using identical slice locations, late gadolinium enhancement (LGE) images were acquired 10 minutes after intravenous administration of a gadolinium-based contrast agent (Dotarem, Gorinchem, the Netherlands; 0.2 mmol/kg) with an inversion recovery, gradient echo pulse sequence to identify the location and extent of MI. The inversion time was individually set to null the signal of viable myocardium (TR 5.2 ms; TE 2.5 ms; flip angle 5°; matrix 192 × 157; voxel sixe 1.25 × 1.25 × 10 mm; slice thickness 10 mm; and slice gap 0 mm).
RV volumetric measurements were performed by a single experienced observer (TMG), blinded for medical history and treatment allocation. The endocardial and epicardial contours of the RV were traced manually on all end-systolic and end-diastolic slices, using QMass 7.6 (Medis, Leiden, the Netherlands). On the most basal slice, the right atrium and the pulmonary artery were excluded. Papillary muscle and trabeculae were included in the RV blood volume. RV end-systolic and end-diastolic volumes, stroke volume, ejection fraction (RVEF), and mass were calculated. The presence of RV LGE was visually determined by 2 independent observers, blinded for medical history and treatment allocation (TMG and TPW) and reviewed by a third observer (RN). RV LGE size was obtained by manually drawing regions of interest, using QMass 7.6 (Medis, Leiden, the Netherlands), and expressed as percentage of RV mass. An independent core laboratory (Image Analysis Center, VU University Medical Center, Amsterdam, the Netherlands) analyzed volumetric and LGE measurements of the LV. RVEF and RV end-diastolic volumes, indexed for body surface area were compared with normal reference values (i.e., males: RVEF ≥48% and indexed RV end-diastolic volume [RVEDVi] ≥114 ml/m 2 ; females: RVEF ≥50% and RVEDVi ≥103 ml/m 2 ). Body surface area was calculated using the simplified calculation of Mosteller.
All echocardiographic images were acquired using Vivid 7 and E9 systems (General Electric, Horton, Norway) with a 2.5- to 3.5-mHz probe and were digitally stored. Tricuspid annular plane systolic excursion (TAPSE) was obtained with M-mode parallel to the RV free wall and across the tricuspid annular plane and TAPSE <17 mm was defined abnormal. The longitudinal systolic movement of the lateral RV annulus was measured. RV free wall longitudinal strain (FWLS) was analyzed on the standard apical 4-chamber view using 2-dimensional (2-D) speckle tracking software by 2 investigators (TMG and YMH) using GE EchoPAC version BT12. Echocardiographic images with poor quality or poor tracking of 2 or more RV wall segments were excluded from the analysis. The endocardial border of the RV wall was manually traced according to current recommendations. On the most basal part, sampling of the right atrium, tricuspid annulus, and pericardium was avoided. RV FWLS was calculated as the mean of basal, mid, and apical segments of the RV free wall ( Supplementary Data ). RV FWLS > −20% is suggested moderately reduced and > −15% severely reduced. Intraobserver and interobserver reliability for RV FWLS measurements was assessed using 30 randomly selected studies.
Data were reported as mean ± SD for normally distributed data, median (interquartile range) for skewed distributed data or n (%) for categorical variables. Differences in continuous variables between subgroups were assessed using independent samples t tests, Mann–Whitney U tests or one-way analysis of variance, according to distribution and the level of subgroups. Differences in categorical variables between groups were performed using chi-square tests. Change in RV FWLS during the 4 months was compared using paired samples t tests. Correlations between continuous variables were performed using Pearson’s correlation. Intraobserver and interobserver variability of RV FWLS was assessed using 2-way mixed intraclass correlation coefficient. Statistical significance was considered achieved with a p value <0.05. Statistical analyses were performed using SPSS statistical software for Windows (version 20, 2011).
Results
In total, 380 patients were randomized in the GIPS-III trial, and a total of 258 patients were included in the present study ( Figure 1 ). In a large subset of patients, sufficient echocardiographic imaging was available both before discharge and at 4 months. Baseline characteristics of the study population are outlined in Table 1 .
| All patients (n=258) | Metformin (n=129) | Placebo (n=129) | |
|---|---|---|---|
| Age (years) | 57.6 ± 11.4 | 57.5 ± 11.4 | 57.7 ± 11.4 |
| Men | 203 (79%) | 102 (79%) | 101 (78%) |
| Body mass index (kg/m 2 ) | 27.0 ± 3.5 | 26.9 ± 3.7 | 27.0 ± 3.3 |
| Hypertension | 72 (28%) | 40 (31%) | 32 (25%) |
| Dyslipidemia | 165 (64%) | 77 (60%) | 88 (68%) |
| Current smoker | 130 (50%) | 69 (54%) | 61 (47%) |
| Stroke | 1 (0.4%) | 1 (0.8%) | 0 |
| Peripheral artery disease | 0 | ||
| Previous PCI | 4 (1.6%) | 1 (0.8%) | 3 (2.3%) |
| Clinical presentation | |||
| Systolic blood pressure (mm Hg) | 133 ± 22 | 133 ± 21 | 132 ± 23 |
| Diastolic blood pressure (mm Hg) | 84 ± 15 | 85 ± 15 | 83 ± 15 |
| Heart rate (beats/min) | 76 ± 16 | 76 ± 15 | 76 ± 17 |
| Total ischemic time (min) | 156 (105 – 243) | 163 (109 – 265) | 146 (102 – 232) |
| Single-vessel coronary disease | 187 (73%) | 89 (69%) | 98 (76%) |
| Infarct-related coronary artery | |||
| Left main coronary artery | 0 | ||
| Left anterior descending | 106 (41%) | 52 (40%) | 54 (42%) |
| Proximal | 70 (27%) | 32 (25%) | 38 (30%) |
| Mid | 32 (12%) | 17 (13%) | 15 (12%) |
| Left circumflex artery | 43 (17%) | 21 (16%) | 22 (17%) |
| Right coronary artery | 109 (42%) | 56 (43%) | 53 (41%) |
| Proximal | 37 (14%) | 15 (12%) | 22 (17%) |
| Mid | 48 (19%) | 28 (22%) | 20 (16%) |
| Infarct-related artery TIMI flow | |||
| Pre intervention grade | |||
| 0 | 148 (57%) | 71 (55%) | 77 (60%) |
| 1 | 18 (7%) | 9 (7%) | 9 (7%) |
| 2 | 44 (17%) | 27 (21%) | 17 (13%) |
| 3 | 48 (19%) | 22 (17%) | 26 (20%) |
| Post intervention grade | |||
| 2 | 16 (6%) | 10 (8%) | 6 (6%) |
| 3 | 242 (94%) | 119 (92%) | 123 (95%) |
| Myocardial blush grade | |||
| 0 | 4 (2%) | 1 (1%) | 3 (2%) |
| 1 | 18 (7%) | 11 (9%) | 7 (5%) |
| 2 | 56 (22%) | 27 (22%) | 29 (23%) |
| 3 | 178 (69%) | 88 (68%) | 90 (70%) |
| Laboratory values at admission | |||
| Creatine kinase (U/L) | 133 (90 – 215) | 143 (97 – 285) | 123 (87 – 181) |
| Myocardial band of creatine kinase (U/L) | 17 (13 – 25) | 18 (14 – 34) | 16 (13 – 23) |
| Creatinine (μmol/L) | 72 (64 – 82) | 71 (62 – 85) | 73 (64 – 80) |
| NT-proBNP (ng/L) | 79 (38 – 180) | 80 (41 – 218) | 78 (37 – 118) |
| Glucose (mmol/L) | 8.2 (6.9 – 9.4) | 8.3 (6.9 – 9.5) | 8.1 (6.9 – 9.3) |
| HbA1c (%) | 5.8 (5.6 – 6.0) | 5.7 (5.6 – 6.0) | 5.8 (5.6 – 6.0) |
Table 2 summarizes RV and LV function on MRI at 4 months. No patient had reduced RVEF. RVEF was not different according to culprit lesion (i.e., left anterior descending coronary artery [LAD], right coronary artery [RCA], and left circumflex coronary artery [LCX]; p = 0.11). There was no difference in RVEF between patients with LV LGE septal involvement (n = 171) and without septal involvement (n = 72); p = 0.43. RVEF was higher in women compared with men (67% vs 63%, p <0.001). Age was weakly correlated with RVEF ( r = 0.15, p = 0.02). RVEF was not different according to myocardial blush grade (p = 0.24). LVEF was fairly correlated with RVEF ( r = 0.25, p <0.001); LV end-diastolic volume index did not correlate with RVEF ( r = −0.04, p = 0.55) and also LV scare size was not associated with RVEF ( r = −0.01, p = 0.91). There was a slight difference in RVEF between the metformin and placebo group (63% vs 65%, p = 0.03). After adjustment for age, gender and LVEF, this difference did not remain significant (p = 0.11). Indexed RVEDVis above normal reference values were seen in 2 patients (0.8%). RVEDVi was higher in men compared with women (80 vs 70 ml/m 2 , p <0.001) and was fairly correlated with age ( r = −0.27, p <0.001). RV LGE was detected in 5 patients (2%; Table 3 ). All RV infarctions were not isolated but extended throughout the inferior wall. Figure 2 is an example of such patient.
| MRI after 4 months | |
| RV end-diastolic volume (ml) | 159 ± 36 |
| RV end-systolic volume (ml) | 58 ± 18 |
| RV ejection fraction (%) | 64 ± 5 |
| RV infarct size (%, n=5) | 7.4 ± 4.9 |
| LV end-diastolic volume (ml) | 193 ± 44 |
| LV end-systolic volume (ml) | 90 ± 33 |
| LV ejection fraction (%) | 54 ± 8 |
| LV infarct size (%, n=243) | 8.9 ± 7.6 |
| Echocardiography before discharge | |
| Tricuspid annular plane systolic excursion (mm) | 22.8 ± 3.6 |
| RV free wall longitudinal strain (%) | -22.6 ± 5.8 |
| Echocardiography after 4 months | |
| Tricuspid annular plane systolic excursion (mm) | 25.1 ± 3.9 ∗ |
| RV free wall longitudinal strain (%) | -25.9 ±4.7 ∗ |
| Subject | Sex/age (yrs) | Right coronary artery | RV scar size | LV scar size | Location |
|---|---|---|---|---|---|
| 1 | M/45 | Mid | 1.3% | 7.0% | Inferoseptal |
| 2 | M/53 | Mid | 13.2% | 6.5% | Inferolateral |
| 3 | M/81 | Proximal | 7.0% | 8.9% | Inferoseptal |
| 4 | M/63 | Mid | 11.4% | 20.6% | Inferoseptal |
| 5 | F/68 | Proximal | 4.1% | 13.2% | Inferoseptal |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
