Article ID: CJ-21-0312
Background: Because no data were available regarding the effect of preexisting left ventricular filling pressure (LVFP) on clinical outcomes in patients with acute myocardial infarction (AMI), we evaluated whether preexisting high LVFP can determine outcomes of subsequent AMI events.
Methods and Results: Among 399,613 subjects who underwent echocardiography for various reason from August 2004 to June 2019, 231 had experienced subsequent AMI and were stratified according to preexisting LVFP: low LVFP (E/e’ ≤14) and high LVFP (E/e’ >14). The primary outcome was cardiac death at 30 days and 1 year after AMI. Overall, 19.5% had high LVFP prior to AMI events. Preexisting high LVFP was associated with an increased risk of cardiac death at 30 days (3.8% vs. 11.6%; adjusted hazard ratio (HR) 4.56, 95% confidence interval (CI) 1.20–17.24, P=0.026) and 1 year after AMI (7.9% vs. 35.9%; adjusted HR 4.14, 95% CI 1.79–9.57, P<0.001). Preexisting E/e’ as a continuous value was significantly associated with 1-year risk of cardiac death (adjusted HR 1.08, 95% CI 1.02–1.15, P=0.007). Follow-up echocardiography showed that patients with high LVFP did not show improvement in systolic or diastolic function.
Conclusions: Preexisting high LVFP was associated with poor clinical course and 1-year cardiac death after subsequent AMI, as well as no improvement in systolic or diastolic function.
For several decades, the prognostic implication of systolic dysfunction after acute myocardial infarction (AMI) has been a major concern of researchers.1 However, unlike systolic function, which can be easily assessed by the left ventricular (LV) ejection fraction (LVEF), assessment of diastolic function has been a challenging issue in contemporary practice. Left ventricular filling pressure (LVFP) is a key parameter of LV diastolic function, which includes LV relaxation, stiffness, and early diastolic recoil.2
Editorial p ????
Invasive LV end-diastolic pressure measurement is the gold standard for reflecting LVFP and it can be used for prognostication during the index AMI event. A previous study reported that elevated LV end-diastolic pressure is associated with adverse outcomes among AMI patients who undergo percutaneous coronary intervention (PCI).3 In some cases, it is not easy to control pulmonary congestion or edema with diuretics therapy and renal replacement therapy after AMI, even in the absence of severe LV systolic dysfunction. Therefore, it is presumed that the preexisting LVFP, representing the degree of LV stiffness before MI, may affect the clinical course and outcome of AMI. However, there are no data on whether preexisting LVFP affects clinical outcomes in future MI.
LVFP can be noninvasively assessed by echocardiography, so we sought to identify whether preexisting LVFP using the ratio of early transmitral flow velocity to early diastolic mitral annular velocity (E/e’) by Doppler echocardiography, previously identified by screening echocardiography, can determine the outcome of subsequent AMI events.
The study population was selected from a general population who underwent transthoracic echocardiography (TTE) in Samsung Medical Center from August 2004 to June 2019. A total of 399,613 subjects were examined by TTE for various reasons, including routine examination, screening, and diagnosis of specific cardiac disease. Among them, only 676 patients (0.17%) re-presented for subsequent AMI (Supplementary Figure 1). Pre-AMI TTE was performed at a median of 645.0 days (interquartile range, 180.0–1,465.0 days) prior to the AMI. We excluded patients who underwent pre-AMI TTE more than 5 years prior (n=2), were not available to assess E/e’ (n=331), had previous history of AMI (n=111), and underwent thrombolysis (n=1). Finally, 231 patients who had undergone PCI were enrolled and divided into 2 groups according to E/e’ >14: preexisting high LVFP (n=45) and low LVFP (n=186). The study protocol was approved, the requirement for informed consent of the individual patients was waived by the Institutional Review Board of Samsung Medical Center, and the study was conducted according to the principles of the Declaration of Helsinki.
Definitions and OutcomesPreexisting LVFP was assessed using E/e’ at the pre-AMI TTE. Cutoff values of E/e’ indicating increased LVFP vary from 8 to 15.4–8 We defined the cutoff of E/e’ for high LVFP as >14, according to a previous study validating its relationship with invasive catheterization.9 AMI was defined by increase and/or decrease in cardiac biomarkers, with at least 1 value above the 99th percentile upper reference limit, and with at least one of the following: (1) symptoms of acute myocardial ischemia; (2) new ischemic ECG changes; (3) development of pathologic Q waves; (4) imaging evidence of new loss of viable myocardium or new regional wall motion abnormality; and (5) identification of a coronary thrombus by angiography, including intracoronary imaging, or by autopsy.10 Diagnosis of ST-segment elevation MI (STEMI) was made in the presence of new ST-elevation at the J-point in 2 contiguous leads.10
The primary outcome of the current study was cardiac death during the first year. Secondary outcomes were all-cause death, MI, and re-hospitalization for heart failure at 1 year. All deaths were considered cardiac unless a definite non-cardiac cause could be established. If patients were unavailable to visit the hospital and the medical center, additional information was obtained by telephone interviews, if necessary. Mortality data for patients who were lost to follow-up were confirmed by national death records. All events were adjudicated by a cardiology expert who was blinded to the patients’ baseline characteristics.
Echocardiographic MeasurementsAll patients underwent comprehensive TTE before AMI, at the index event, and during follow-up after AMI using commercially available ultrasound systems (Vivid, GE Medical Systems, Milwaukee, WI, USA; Acuson, Siemens Medical Solution, Mountain View, CA, USA; or Sonos, Philips Medical System, Andover, MA, USA). Standard 2D, color, and tissue Doppler images were acquired with positional change and analyzed in the cardiovascular imaging core laboratory at Samsung Medical Center. Cardiac chamber quantification and assessment of systolic and diastolic function were performed according to ASE/EACVI recommendations.6,11 Routine TTE was recommended for all patients with AMI before discharge from hospital and followed up within 1 year. TTE at the AMI event was performed at a median of 1.0 day after AMI (interquartile range (IQR), 1.0–3.0 days). Follow-up TTE was performed at a median of 254.5 days (IQR, 158.5–395.2 days) after AMI.
Statistical AnalysisA post-hoc power calculation showed this study was adequately powered at >90% (0.975) with a two-sided alpha of 0.05.12 Discrete or categorical variables were compared using the Chi-square test or Fisher’s exact test, respectively. Continuous variables were analyzed using unpaired t-tests or Mann-Whitney rank-sum tests according to distribution, which was assessed by the Kolmogorov-Smirnov test and visual inspection of Q-Q plots. Survival analysis was performed with the Kaplan-Meier method and comparison with the log-rank test. Cox proportional hazards regression models were constructed for calculating hazard ratios (HRs) and 95% confidence intervals (CIs). Proportional hazards assumptions for the models were assessed by the log-minus-log plot and the Schoenfeld residuals. Survival analysis and Cox regression analysis were also performed for patients without previous coronary artery disease (CAD), defined as previous history of PCI.
To identify independent predictors of cardiac death, multivariable Cox proportional hazard models were constructed using variables with a P value <0.05 in the univariate analyses and variables that could be potentially relevant.13,14 Potentially relevant variables are listed in Supplementary Table 1. We performed a backward elimination based on an information criterion. The final model included age, sex, diagnosis of STEMI, and creatinine level. We adjusted the cohorts for the probability of treatment with the inverse probability treatment-weighting (IPTW). For IPTW analysis, the logistic regression model for the presence of high LVFP was applied to calculate propensity scores, based on differences in baseline characteristics including the variables of age, sex, body mass index, diagnosis of STEMI, hypertension, diabetes, current smoker, history of PCI, creatinine level, and preserved LVEF (≥50%) on pre-AMI echocardiography. With the IPTW method, no loss of sample occurs; therefore, this was considered beneficial for this small observational study to reduce selection bias and potential confounding.15,16 Residual differences in characteristics between adjusted cohorts were assessed by calculating the absolute standardized mean differences. Standardized mean differences were <0.1 across all matched covariates, indicating a good balance. All CIs for the inverse probability weighting analyses were assessed with bootstrapping methods with 1,000 iterations.
In addition to categorical analysis, the associations between preexisting E/e’ as continuous variables and the risk of 1-year cardiac death were graphically presented with restricted cubic spline with 3 degrees of freedom.17 For comparison of serial changes of echocardiographic parameters according to preexisting high LVFP, we performed two-way repeated-measures analysis of variance for comparing changes of parameters with a group and time interaction.
All analyses were two-tailed, and statistical significance was defined as a P<0.05. All statistical analyses were performed using IBM SPSS Statistics for Windows, Version 22.0 (Armonk, NY, USA) and R version 3.6.0 software (R Foundation for Statistical Computing, Vienna, Austria).
Among the study population, 45 patients (19.5%) had preexisting high LVFP prior to index AMI events (Table 1). Patients with high LVFP were older (72.6±9.7 vs. 66.1±10.9 years, P<0.001) and there was a smaller proportion of men (60.0% vs. 78.0%, P=0.022) compared with those with low LVFP. The incidence of current smoker was lower (35.6% vs. 56.5%, P=0.019) and the incidence of previous history of PCI was higher (33.3% vs. 15.6%, P=0.012) in the high LVFP group than in the low LVFP group. At presentation, the incidence of STEMI was lower in the preexisting high LVFP group than in the low LVFP group (44.4% vs. 68.3%, P=0.005). Creatinine level was significantly higher in the preexisting high LVFP group than in the low LVFP group (2.2±2.3 vs. 1.1±0.8 mg/dL, P=0.003). There were no significant differences in other laboratory findings, procedural findings, or in-hospital management, including mechanical circulatory support at the time of index AMI event.
| Preexisting low LVFP (n=186) |
Preexisting high LVFP (n=45) |
P value | |
|---|---|---|---|
| Age, years | 66.1±10.9 | 72.6±9.7 | <0.001 |
| Men | 145 (78.0) | 27 (60.0) | 0.022 |
| Body mass index, mg/m2 | 24.4±3.4 | 24.2±3.5 | 0.707 |
| STEMI | 127 (68.3) | 20 (44.4) | 0.005 |
| Killip class III or IV | 32 (17.2) | 16 (35.6) | 0.012 |
| Pulmonary congestion or edema | 29 (15.6) | 15 (33.3) | 0.012 |
| Cardiogenic shock | 4 (2.2) | 2 (4.4) | 0.729 |
| Cardiovascular risk factors | |||
| Hypertension | 107 (57.5) | 29 (64.4) | 0.498 |
| Diabetes | 136 (73.1) | 32 (71.1) | 0.932 |
| Hyperlipidemia | 86 (46.2) | 20 (44.4) | 0.960 |
| Current smoker | 105 (56.5) | 16 (35.6) | 0.019 |
| History of PCI | 29 (15.6) | 15 (33.3) | 0.012 |
| Diastolic dysfunctiona | <0.001 | ||
| Normal | 53 (33.3) | 0 (0) | |
| Grade I | 96 (60.4) | 21 (70.0) | |
| Grade II–III | 10 (6.3) | 9 (30.0) | |
| Laboratory findings | |||
| Creatinine, mg/dL | 1.1±0.8 | 2.2±2.3 | 0.003 |
| hs-CRP, mg/dL | 4.2±23.5 | 3.3±7.9 | 0.672 |
| CK-MB, ng/mL | 32.7±65.6 | 34.4±55.2 | 0.873 |
| Troponin-I, ng/mL | 15.8±51.0 | 9.5±20.7 | 0.218 |
| Medications | |||
| ACE inhibitors | 43 (23.1) | 5 (11.1) | 0.115 |
| Angiotensin-receptor blockers | 76 (40.9) | 14 (31.1) | 0.302 |
| β-blockers | 118 (63.4) | 28 (62.2) | 1.000 |
| Loop diuretics | 56 (30.1) | 24 (53.3) | 0.006 |
| Spironolactone | 12 (6.5) | 11 (24.4) | 0.001 |
| Procedural and in-hospital management | |||
| Multivessel disease | 95 (51.1) | 27 (60.0) | 0.363 |
| Non-culprit PCIb | 30 (16.1) | 5 (11.1) | 0.541 |
| Target vessel location | |||
| Left main disease | 10 (5.4) | 3 (6.7) | 1.000 |
| Left anterior descending artery | 128 (68.8) | 37 (82.2) | 0.109 |
| Left circumflex artery | 75 (40.3) | 22 (48.9) | 0.381 |
| Right coronary artery | 118 (63.4) | 27 (60.0) | 0.797 |
| Stent diameter, mm | 3.1±0.5 | 3.0±0.4 | 0.228 |
| Stent length, mm | 34.7±19.8 | 35.1±17.6 | 0.895 |
| No. of stents | 1.1±0.7 | 1.0±0.7 | 0.171 |
| Intra-aortic balloon pump | 6 (3.2) | 4 (8.9) | 0.205 |
| Extracorporeal membrane oxygenation | 9 (4.8) | 4 (8.9) | 0.486 |
| Post-PCI TIMI 3 flow | 182 (97.8) | 42 (93.3) | 0.271 |
Values are expressed as mean±SD or number (%). aTotal of 189 patients were available for assessing diastolic dysfunction. bImmediate procedure at index event. ACE, angiotensin-converting enzyme; CK-MB, creatinine kinase myocardial band; hs-CRP, high-sensitivity C-reactive protein; LVFP, left ventricular filling pressure; PCI, percutaneous coronary intervention; STEMI, ST-segment elevation myocardial infarction; TIMI, Thrombolysis in Myocardial Infarction.
When patients with preexisting high LVFP presented with AMI, they had longer hospital stay (14.2±24.7 vs. 5.6±5.7 days, P=0.024) and intensive care unit stay (3.3±4.7 vs. 1.7±3.8 days, P=0.013) compared with those with low LVFP (Figure 1). In addition, the incidence of pulmonary congestion/edema (Killip class III) or cardiogenic shock (Killip class IV) was significantly higher in patients with preexisting high LVFP (33.3% vs. 15.6%, P=0.012), in whom loop diuretics were more frequently used (53.3% vs. 30.1%, P=0.006).
In-hospital outcomes at index event of AMI. There are significant differences in (A) length of hospital stay, (B) length of ICU stay, (C) pulmonary congestion or edema, and (D) use of loop diuretics according to preexisting LVFP. The bars indicate mean of continuous variables and frequencies of categorical variables. AMI, acute myocardial infarction; ICU, intensive care unit; LVFP, left ventricular filling pressure.
Preexisting high LVFP was associated with increased risks of both 30-day cardiac death (3.8% vs. 11.6%; adjusted HR 4.56, 95% CI 1.20–17.24, P=0.026) and 1-year cardiac death (7.9% vs. 35.9%; adjusted HR 4.14, 95% CI 1.79–9.57, P<0.001) following an AMI event (Table 2, Figure 2). When we compared patients without a previous history of PCI, patients with high LVFP were consistently associated with an increased risk of cardiac death (Supplementary Figure 2). Subgroup analysis consistently showed that preexisting high LVFP was associated with an increased risk of 1-year cardiac death. However, there were no statistically significant differences between subgroups (i.e., no significant interaction; Supplementary Figure 3). For sensitivity analysis, IPTW was performed (Table 2). Findings consistently showed that preexisting LVFP was associated with increased risks of 30-day cardiac death (IPTW-adjusted HR 6.63, 95% CI 1.73–25.37, P=0.006) and 1-year cardiac death (IPTW-adjusted HR 4.82, 95% CI 1.85–12.58, P=0.001).
| Preexisting low LVFP (n=186) |
Preexisting high LVFP (n=45) |
Unadjusted HR (95% CI) |
P value | Adjusted HRa (95% CI) |
P value | IPTW-adjusted HR (95% CI) |
P value | |
|---|---|---|---|---|---|---|---|---|
| 30-day outcomes | ||||||||
| Cardiac death | 7 (3.8%) | 5 (11.6%) | 3.12 (0.99–9.84) | 0.052 | 4.56 (1.20–17.24) | 0.026 | 6.63 (1.73–25.37) | 0.006 |
| All-cause death | 13 (7.0%) | 7 (15.6%) | 2.35 (0.94–5.89) | 0.068 | 3.17 (1.13–8.93) | 0.029 | 4.12 (1.26–13.45) | 0.019 |
| Myocardial infarction | 0 (0%) | 1 (2.4%) | NA | NA | NA | NA | NA | NA |
| Stroke | 1 (0.6%) | 1 (2.2%) | 4.34 (0.27–69.41) | 0.300 | 4.84 (0.21–113.89) | 0.328 | 6.47 (0.44–94.21) | 0.172 |
| MACCE | 15 (8.1%) | 8 (17.8%) | 2.37 (1.01–5.58) | 0.049 | 2.76 (1.04–7.30) | 0.041 | 3.55 (1.12–11.24) | 0.031 |
| 1-year outcomes | ||||||||
| Cardiac death | 14 (7.9%) | 15 (35.9%) | 5.02 (2.42–10.40) | <0.001 | 4.14 (1.79–9.57) | <0.001 | 4.82 (1.85–12.58) | 0.001 |
| All-cause death | 22 (12.0%) | 19 (42.4%) | 4.00 (2.17–7.40) | <0.001 | 3.63 (1.80–7.32) | <0.001 | 3.85 (1.64–9.05) | 0.002 |
| Myocardial infarction | 3 (2.0%) | 3 (8.3%) | 5.15 (1.04–25.57) | 0.045 | 1.63 (0.21–12.39) | 0.637 | 0.70 (0.11–4.49) | 0.706 |
| Stroke | 1 (0.6%) | 1 (2.2%) | 4.34 (0.27–69.41) | 0.300 | 4.84 (0.21–113.89) | 0.328 | 6.47 (0.44–94.21) | 0.172 |
| MACCE | 27 (14.9%) | 21 (47.1%) | 3.74 (2.11–6.62) | <0.001 | 3.54 (1.84–6.81) | <0.001 | 3.01 (1.30–6.96) | 0.010 |
Data are expressed as number of events and the cumulative incidence of clinical outcomes. Cumulative incidences of clinical outcomes represent Kaplan-Meier estimates at 30 days and 1 year, respectively. aAdjusted for variables of age, sex, diagnosis of ST-elevation myocardial infarction, and creatinine level. CI, confidence interval; HR, hazard ratio; IPTW, inverse probability treatment-weighting; MACCE, major adverse cardiovascular and cerebrovascular events.
Comparison of outcomes according to preexisting LVFP (E/e’ >14). The cumulative incidence cardiac death compared according to preexisting high LFVP at (A) 30 days and (B) 1 year. Preexisting high LVFP was related to increased risk of both 30-day and 1-year cardiac death. CI, confidence interval; HR, hazard ratio; LVFP, left ventricular filling pressure.
Preexisting E/e’ as a continuous value was also significantly associated with 1-year risk of cardiac death after adjustment of LVEF (adjusted HR 1.08, 95% CI 1.02–1.15, P=0.007; Figure 3). However, preceding LVEF fraction was not significantly associated with 1-year risk of cardiac death (adjusted HR 0.98, 95% CI 0.94–1.02, P=0.306). Among patients without preexisting high LVFP, there was no significant difference in outcomes regardless of the subsequent LVFP at index AMI event (4.5% vs. 6.1%; adjusted HR 1.12, 95% CI 0.20–6.31, P=0.901; Supplementary Figure 4).
Independent association of LVEF and E/e’ with estimated risk of cardiac death at 1 year. The restricted cubic spline model presents the linear relationship between left ventricular ejection fraction (LVEF) or E/e’ and estimated risk of 1-year cardiac death. (A) Preexisting LVEF-adjusted E/e’ was not significantly associated with risk of cardiac death; (B) preexisting E/e’ was significantly associated with risk of 1-year cardiac death, even after adjustment for LVEF. CI, confidence interval; HR, hazard ratio; LVFP, left ventricular filling pressure.
Figure 4 and Supplementary Table 2 summarize the details of the echocardiographic parameters pre-AMI, at AMI, and at follow-up. On pre-AMI TTE, the preexisting high LVFP group had lower LVEF and e’ velocity than the low LVFP group, while E/e’, right ventricular systolic pressure, LV mass index, and left atrial volume index were higher in the preexisting high LVFP group. After subsequent AMI, follow-up TTE showed that patients with preexisting low LVFP showed improvements in systolic function (LVEF at AMI 50.7±11.9% vs. follow-up 54.0±11.6%; P=0.003) compared with the TTE at the AMI event; however, there was no significant improvement among the patients with preexisting high LVFP (LVEF at AMI 49.9±13.9% vs. follow-up 48.5±15.7%; P=0.599). Diastolic function, which can be assessed by LVFP using the parameter of E/e’, also had showed no significant improvement after AMI among the patients with preexisting high LVFP (AMI 20.6±7.5 vs. follow-up 18.3±8.0; P=0.226). The temporal trend of e›, another representative parameter of diastolic function, was also similar to E/e’ among the patients with preexisting high LVFP and there was no significant improvement.
Serial echocardiographic parameters according to time point. Changes of (A) LVEF, (B) e’, (C) E/e’, and (D) RVSP at pre-AMI, AMI, and follow-up. Means and standard errors are shown. The P values were calculated by two-way repeated-measures ANOVA for the time-by-group interaction of continuous variables. AMI, acute myocardial infarction; CI, confidence interval; HR, hazard ratio; LVEF, left ventricular ejection fraction; LVFP, left ventricular filling pressure; RVSP, right ventricular systolic pressure.
The current study investigated the clinical background and outcomes according to preexisting high LVFP among patients who experienced a subsequent AMI event. We screened about 0.4 million subjects and 676 patients (0.17%) revisited with a subsequent AMI event. The major findings were as follows. First, approximately 20% of patients already had preexisting high LVFP. They tended to have a higher incidence of pulmonary edema or congestion, more frequent use of loop diuretics and spironolactone, and longer hospital and intensive care unit stays when they experienced AMI. Second, when we assessed the clinical effect of preexisting high LVFP, high LVFP was significantly associated with increased risk of 30-day and 1-year cardiac death after AMI. Third, although all patients underwent PCI to restore coronary flow, patients with preexisting high LVFP showed no improvement in either systolic or diastolic function at follow-up echocardiography compared with TTE at the index AMI event.
Clinical Effect of Preexisting High LVFP on Subsequent AMI EventsLV diastolic function is characterized by LV relaxation, early diastolic recoil, and chamber stiffness, all of which are determined by LVFP.2 Previous studies have focused on the association between diastolic dysfunction and LVFP at the index AMI event and clinical outcomes to investigate the clinical effect of LVFP in AMI patients. Those studies showed that elevated LVFP in the acute stage of AMI was associated with poor clinical outcomes.1–3,18 However, their findings have been limited by not considering the effect of underlying LV stiffness. Because elevated LVFP at the index AMI event may be determined by synergy between preexisting stiff LV myocardium and newly developed ischemic and infarcted myocardium, previous research has been unable to clearly investigate the pure effect of preexisting LVFP after subsequent AMI events. Therefore, we firstly screened a general population that underwent TTE and further analyzed those who experienced AMI after the TTE, excluding the elevation of infarct-related filling pressure.
Patients with preexisting high LVFP were older and more often female than those with low LVFP. These different clinical profiles are attributed to major factors, such as arterial stiffness and myocardial fibrosis, causing vulnerability to increased LVFP. Further, the rate of PCI was higher in patients with preexisting high LVFP, suggesting that underlying CAD also affected LVFP, in agreement with a previous study.19 When such patients present with AMI, they may have different clinical courses because of their preexisting high LVFP. In fact, in the present study, we found significantly higher incidence of pulmonary congestion or edema among patients with preexisting high LVFP after AMI, and more than half of these patients should be treated with diuretics.
Differential Effect of Preexisting LVFP on Clinical Outcomes After AMIAfter AMI, early death was significantly higher among patients with preexisting high LVFP, compared with those with low LVFP. The major cause of death after AMI is cardiac death related to pump failure.20 Because LVFP is affected by both systolic and diastolic dysfunction, patients with preexisting high LVFP may require more preload reduction to achieve adequate LVFP even if they have similar infarct-related injury to patients with low LVFP and overt preload reduction easily leads to tissue hypoperfusion. Furthermore, patients with preexisting high LVFP may be more vulnerable to pulmonary congestion, which subsequently leads to the aggravation of hypoxemia. Hypoxemia decreases oxygen delivery to the myocardium and finally results in pump failure. The present study has provided evidence for this plausible explanation: the patients with preexisting high LVFP were more frequently admitted with Killip III/IV at subsequent AMI events, and they were also treated with diuretics more frequently during the admission. Furthermore, when we indirectly assessed the clinical severity of AMI patients based on length of stay, the preexisting high LVFP group had longer durations than those without preexisting high LVFP.
At 1 year after the index AMI event, patients with preexisting high LVFP showed increased risk of death, which was mostly driven by cardiac death. This trend was consistent after exclusion of patients with previous history of CAD before AMI. Additionally, preexisting E/e’ was associated with increased risk of cardiac death as a continuous variable, even after adjustment for LVEF. However, preexisting LVEF fraction was not a determinant of 1-year cardiac death. When we compared patients according to newly developed high LVFP at the index AMI among the patients with preexisting low LVFP, the LVFP at index AMI was not related to increased risk of cardiac death. This finding suggested that assessment of LVFP at the acute phase of AMI might be limited to its prognostic discriminating value. Although previous studies have established the prognostic impact of diastolic dysfunction after AMI,1–3,18 it is possible that any influence of preexisting LV stiffness prior to AMI might be masked. Therefore, we should consider the underlying mechanism causing high LVFP when interpreting diastolic function at the index procedure for AMI.
Serial Changes of Echocardiographic Parameters After AMIAfter subsequent AMI events, changes in systolic and diastolic function according to preexisting LVFP were assessed by serial TTE. When we compared the echocardiographic parameters at the AMI event and follow-up, neither systolic nor diastolic function improved among patients with preexisting high LVFP. This phenomenon might be the result of differences in amount of ischemic or infarcted myocardium; however, there was no significant difference in the amount of damaged myocardium, as assessed by cardiac enzymes, or the incidence of CAD between the groups. In the same line, patients with preexisting low LVFP also showed transient elevation of E/e’ in the acute phase of AMI, and some of those had newly developed high LVFP at the index AMI event. A plausible explanation of the fundamental difference is LV stiffness, which is confirmed by preexisting high LFVP. LV stiffness is associated with cardiomyocyte injury from coronary microvascular disease.21,22 Therefore, patients with preexisting high LVFP may be more vulnerable to ischemic damage, and their stiff myocardium will be somewhat less viable even though revascularization is deemed successful.
Study LimitationsSome limitations of the current study are presented. First, this was a single-center retrospective study. Although we screened all patients who underwent TTE over 15 years, the proportion of AMI was relatively low (0.17%). According to the Korea Acute Myocardial Infarction Registry, the annual incidence of AMI is about 0.01% (5,000 patients/50 million) in Korea,23 which suggests we recruited a representative sample of real-world practice. Second, there will be selection bias or potential confounders because of the study’s retrospective design. Although we tried to adjust for differences in baseline characteristics using IPTW, other sensitivity analyses based on alternation methods such as propensity-score matching could not be performed due to the small sample size. Third, although a previous study has reported that E/e’ shows a strong relationship with invasively measured LVFP,9 it cannot completely reflect the real LVFP. Fourth, we analyzed the clinical outcomes during 1 year after AMI, which may not be sufficient to conclude the long-term effect of preexisting LVFP. However, it should be noted that this study focused on the differential effect of underlying LV stiffness during the acute stage of AMI, with regard to the clinical picture and cardiac death.
Preexisting high LVFP was associated with poor clinical course and increased risk of early death within 30 days of the index AMI. Additionally, those with preexisting high LVFP showed significantly increased risk of cardiac death at 1 year. Notably, despite successful revascularization, patients with preexisting LVFP showed no improvements in systolic and diastolic function after subsequent AMI events. Therefore, consideration of the underlying status of LV stiffness is crucial for prognostication and management of AMI patients.
None.
None of the authors have any conflicts of interest to declare.
Name of the ethics committee: Samsung Medical Center. Reference number: 2020-08-029-001.
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-21-0312
