Abstract
Background
Ischemic cardiomyopathy, commonly attributable to coronary artery disease, leads to ventricular remodeling. Surgical ventricular reconstruction (SVR) aims to reverse this by restoring cardiac morphology. This review evaluates SVR’s influence on ventricular volume reduction, comparing <30% to >30% reduction effects on left ventricular ejection fraction (LVEF) and patient outcomes.
Methods
A systematic search following Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines was performed in PubMed, Embase, Web of Science, and the Cochrane Collaboration. Data extracted included left ventricular end systolic volume, LVEF, New York Heart Association classification, mitral regurgitation grade, and mortality rates.
Results
Of the 37 articles included, 29 reported a mean volume reduction of >30%, while 8 reported a reduction <30%. A total of 4975 participants were included across all studies. Mean volume reduction was 43% in group 1 (>30%) and 28% in group 2 (<30%). The mean relative increase in LVEF for group 1 was 38%, which was larger compared to group 2 with 30%. Additionally, mean reduction in New York Heart Association class was 1.5 in group 1 and 1.2 in group 2. There was no difference in mitral regurgitation grade reduction between the 2 groups. Postoperative mortality within 30 days was 5.8% in group 1 vs. 5.2% in group 2.
Conclusions
Both volume reduction groups showed improved LVEF and outcomes post-SVR, indicating the efficacy of SVR. The majority of studies achieved a volume reduction exceeding 30%, surpassing the 19% reported in the Surgical Treatment for Ischemic Heart Failure trial. Further meta-analysis can determine optimal volume reduction for enhanced cardiac function and outcomes.
Highlights
- •
Thirty-seven studies: <30% vs. >30% volume reductions in surgical ventricular reconstruction.
- •
Compared cardiac function and mortality in <30% vs. >30% volume groups.
- •
Improved left ventricular ejection fraction in both volume groups.
- •
Surgical ventricular reconstruction improves cardiac function and patient outcomes.
Introduction
Approximately 64 million people are diagnosed with heart failure (HF) worldwide. As life expectancy rises, the prevalence of HF is expected to increase. Ischemic cardiomyopathy (ICM), accounting for roughly 60% of HF cases, often involves anterior myocardial infarctions, which represent about 33% of all myocardial infarctions. , Negative left ventricular (LV) remodeling occurs in around 30% of patients after acute myocardial infarction despite timely percutaneous coronary intervention and optimal guideline-directed medical therapy and occurs as a response to increased wall stress. Failure to normalize increased wall stress finally results in progressive LV dilation and contractile function deterioration.
In patients with severe LV systolic dysfunction and coronary artery disease suitable for intervention, coronary artery bypass grafting (CABG) is recommended as the primary revascularization strategy in patients with multivessel disease and acceptable surgical risk. Current guidelines of the European Society of Cardiology furthermore state that in experienced centers, surgical ventricular reconstruction (SVR) of the LV may be performed concomitant to CABG if HF symptoms are more predominant than angina and if myocardial scar and LV remodeling are present. By exclusion of akinetic and/or dyskinetic LV scar, SVR aims to restore the physiological volume and normal elliptical shape of the LV. If indicated, CABG and SVR can be performed in combination with mitral valve (MV) repair to address secondary mitral regurgitation (MR).
Although the Surgical Treatment for Ischemic Heart Failure (STICH) trial was designed and powered to evaluate the added value of SVR in ischemic HF patients, it showed unfavorable outcomes for SVR. First, the STICH trial investigators demonstrated that the addition of SVR to CABG did not result in a benefit in overall survival or survival free from cardiac hospitalization compared with CABG alone in patients with ICM. However, the results caused widespread discussion and were finally re-examined as the volume reduction achieved with SVR within the STICH trial was insufficient. Importantly, the probability of all-cause death was found to be significantly higher in patients that had a post-SVR LV end systolic volume index (LVESVI) of 60 mL/m 2 or more. A similar trend in survival was correlated to the magnitude of decrease in LV volumes. Although not significant in the STICH population, a survival benefit was observed for patients in whom at least 30% reduction of LVESVI was achieved. , In line with this, Isomura et al. showed that SVR in patients with a >33% volume reduction resulted in a higher survival rate. Moreover, positive effects of a 30% volume reduction on ventricular function and New York Heart Association (NYHA) classification were observed by the RESTORE group. We ther2efore performed a systematic review to evaluate the effect of the level of volume reduction (<30% vs. >30%) on LV function after SVR in patients with ICM. Secondary outcomes were the effects of volume reduction on ventricular function, NYHA classification, MR, and mortality.
Methods
Search Strategy
A systematic literature search was performed in PubMed, Embase, Web of Science, and The Cochrane Collaboration according to the Guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis on April 16, 2021. A clinical librarian was consulted to assist with the search process. The complete search strategy can be found in Appendix A .
Study Selection
Studies were included when the following characteristics were reported: SVR procedure performed in patients with ICM, left ventricular end systolic volume (LVESV), LVESVI, left ventricular ejection fraction (LVEF), NYHA class, MR grade, and mortality rates.
Articles were screened by 2 independent reviewers (H.O. and P.C.H.) on title and abstract. Divergences were resolved by consensus. We assessed whether studies described SVR in patients with ICM. Articles describing patients in whom SVR was performed for reasons other than ICM were excluded. Only articles written in English were included. A full text review was done by 2 reviewers (H.O. and P.C.H.) on selected studies. For the qualitative synthesis, studies were included if they met the inclusion criteria.
Data Analysis
Data were extracted using Microsoft Excel (version 2019). The following data were extracted: author, year, title, study design, number of participants, mean age, inclusion criteria for SVR, type of intervention, follow-up period, method for measuring ventricular function, LVESVI or LVESV, LVEF, MR grade, NYHA class, and mortality. Results were divided into 2 groups: mean <30% volume reduction and mean >30% volume reduction. The outcomes were reported in tables, with the difference in LVEF between the 2 groups as the primary outcome and the differences in NYHA class, mortality, and MR grade as secondary outcomes. For each outcome, a weighted average was calculated using the formula: ∑( n in each study/Total n × Outcome value), where n represents the number of participants in each study for whom the specific outcome was measured, and Total n represents the total number of participants across all studies for that particular outcome. This method was applied to all outcomes, including LVEF, NYHA class, mortality, and MR grade, ensuring that the results were adjusted for the sample sizes reported in each study. Sensitivity analyses were conducted per outcome, excluding studies with fewer than 30 participants for each specific outcome.
Quality Assessment
In the quality assessment, 2 specific tools tailored to the study designs of our included studies were employed: the Newcastle-Ottawa Scale for observational studies and the Risk of Bias 2 (RoB 2) tool for randomized studies. , For observational studies, there was a focus on 3 main domains: selection, comparability, and outcome. In this process, a numeric score was assigned to each item within the Selection and Outcome categories, with the Comparability category allowing for a maximum of 2 scores per item. We then aggregated these scores to compute a total score for each study.
Results
Search Results
The literature search yielded a total of 1168 articles. From these, n = 973 were retrieved from PubMed, n = 59 from Embase, n = 70 from Web of Science, and n = 66 from Cochrane. The search results are summarized in Figure 1 . For the qualitative synthesis, 37 studies were included. Two randomized controlled trials and 35 observational studies were identified. The number of articles that reported the outcomes of interest were as follows: LVESVI and LVEF were reported in 37 articles, MR grade was reported in 8 articles, NYHA class was reported in 16 articles, and mortality was reported in 21 articles.
Quality Assessment
Quality assessment was conducted for all studies in our systematic review. Both randomized controlled trials in our review, by Jones et al. and Aguiar Ribeiro et al., were evaluated using the RoB 2 tool. The study by Jones et al. was found to have a high risk of bias, whereas the study by Aguiar Ribeiro et al. was identified with some concerns regarding bias. The observational studies, assessed by the Newcastle-Ottawa Scale, displayed varying scores, reflecting diverse methodological quality. Detailed results of these assessments are presented in Tables B1 and B2 in Appendix B .
Volume Reduction
The amount of volume reduction was calculated by using the mean preoperative and postoperative LVESVI data from 37 included articles. All articles reported mean LVESVI of the study participants. There was no data collection and analysis performed on patient-level related to the amount of volume reduction. In 2 publications, LVESV was reported and used to calculate the volume reduction. , The amount of volume reduction achieved in the different studies are presented in Tables 1 and 2 . The weighted average volume reduction was 43% in group 1 ( n = 1562) and 28% in group 2 ( n = 1488) ( Tables 1 and 2 ). The lowest mean volume reduction was performed by Nomura et al. with a mean volume reduction of 16.7%. Di Donato et al. reported the highest mean volume reduction of 58.9%. The range of mean volume reduction in articles reporting >30% volume reduction (group 1) ranged between 34.2% and 58.9% ( Table 1 ). Studies with a mean <30% volume reduction (group 2) ranged between 16.7% and 29.6% ( Table 2 ). The sensitivity analyses in both groups demonstrated no notable changes in the weighted average volume reduction ( Tables 1 and 2 ).
| Author and year | N | n | Mean LVESVI (mL/m 2 ) | Reduction in % | p -value | |
|---|---|---|---|---|---|---|
| Preoperative | Postoperative | |||||
| Athanasuleas, 1998 | 8 | 8 | 132 ± 41.6 | 80.5 ± 30.3 | 39.0 | <0.0002 |
| Di Donato, 2001 | 207 | 207 | 112 ± 64 | 46 ± 26 | 58.9 | <0.0001 |
| Suma, 2001 | 50 | 50 | 113 ± 45.7 | 66.4 ± 28.4 | 41.5 | <0.05 |
| Athanasuleas, 2001 | 439 | 185 | 109 ± 71 | 69 ± 42 | 36.7 | <0.0001 |
| Di Donato, 2001 | 44 | 44 | 137 ± 70 | 77 ± 33 | 43.8 | <0.0001 |
| Menicanti, 2002 | 46 | 46 | 98 ± 36 | 63 ± 22 | 35.7 | 0.001 |
| Fuji, 2004 | 14 | 14 | 78 ± 37 | 51 ± 34 | 34.6 | <0.05 |
| Kokaji, 2004 | 16 | 15 | 115 ± 34 | 54 ± 14 | 53.0 | <0.05 |
| Di Donato, 2004 | 30 | 30 | 144 ± 69 | 69 ± 40 | 52.1 | 0.001 |
| Cirillo, 2004 | 69 | 65 | 65.8 ± 29.8 | 37.7 ± 12 | 42.7 | <0.0001 |
| Menicanti, 2004 | 108 | 108 | 170 ± 63 a | 107 ± 41 a | 37.1 | 0.000 |
| Yamaguchi, 2005 | 20 | 20 | 137 ± 24 | 65 ± 19 | 52.6 | <0.05 |
| Tulner, 2006 | 21 | 21 | 186 ± 77 | 101 ± 50 | 45.7 | <0.001 |
| Ferrazzi, 2006 | 85 | 65 | 89.6 ± 27.6 | 56.6 ± 34.5 | 36.9 | <0.0001 |
| Cirillo, 2006 | 12 | 12 | 70.8 ± 12.6 | 42.4 ± 8.6 | 40.1 | <0.05 |
| Aguiar Ribeiro, 2006 | 39 | 38 | 107 ± 19 | 65 ± 14 | 39.3 | <0.001 |
| Takeda, 2008 | 72 | 72 | 111 ± 40 | 59 ± 25 | 46.8 | <0.0001 |
| Suma, 2009 | 76 | 70 | 123.3 ± 38.9 | 74.0 ± 27.5 | 40.0 | <0.001 |
| Bové, 2009 | 23 | 21 | 77 ± 17 | 40 ± 4 | 48.1 | <0.05 |
| Di Donato, 2009 | 67 | 67 | 96 ± 36 | 57 ± 25 | 40.6 | 0.0001 |
| Baravelli, 2010 | 65 | 65 | 88 ± 33 | 57 ± 28 | 35.2 | <0.0001 |
| ten Brinke, 2010 | 9 | 9 | 173 ± 77 a | 103 ± 40 a | 40.5 | <0.001 |
| Zhong, 2011 | 40 | 40 | 117 ± 39 | 77 ± 31 | 34.2 | <0.001 |
| Dor, 2011 | 117 | 101 | 96 ± 45 | 50 ± 21 | 47.9 | 0.01 |
| Shudo, 2011 | 21 | 21 | 123 ± 43 | 71 ± 31 | 42.3 | <0.05 |
| Skelley, 2011 | 87 | 9 | 100.1 ± 50.0 | 57.7 ± 31.2 | 42.4 | <0.05 |
| Cho, 2012 | 40 | 40 | 110.3 ± 35.9 | 57.8 ± 20.7 | 47.6 | <0.0001 |
| Cho, 2014 | 102 | 99 | 104.1 ± 37.4 | 61.4 ± 21.9 | 41.0 | <0.0001 |
| Castelvecchio, 2021 | 20 | 20 | 90.8 | 51.6 | 43.2 | 0.0002 |
| Total n = 1562 | Weighted average = 43.0% | |||||
| Sensitivity analysis | Total n = 1392 | Weighted average = 40.0% | ||||
| Author and year | N | n | Mean LVESVI (mL/m 2 ) | Reduction in % | p -value | |
|---|---|---|---|---|---|---|
| Preoperative | Postoperative | |||||
| Athanasuleas, 2004 | 1198 | 671 | 80.4 ± 51.4 | 56.6 ± 34 | 29.6 | <0.001 |
| Nomura, 2006 | 26 | 26 | 112 ± 43 | 94 ± 47 | 16.7 | 0.0955 |
| Menicanti, 2007 | 1161 | 300 | 145 ± 64 a | 104 ± 50 a | 28.3 | 0.001 |
| Ogawa, 2007 | 4 | 4 | 135 ± 36 | 98 ± 28 | 27.4 | NS |
| Jones, 2009 | 501 | 161 | 83 | 67 | 19.3 | NS |
| Calafiore, 2010 | 44 | 44 | 67 ± 28 | 52 ± 19 | 22.4 | NS |
| Isomura, 2011 | 90 | 90 | 87.2 ± 31.8 | 63.1 ± 25.6 | 27.6 | <0.0001 |
| Zhong, 2011 | 4 | 4 | 96 ± 22 | 63.1 ± 25.6 | 28.1 | <0.05 |
| Total n = 1488 | Weighted average = 28.0% | |||||
| Sensitivity analysis | Total n = 1266 | Weighted average = 28.0% | ||||
LVEF and NYHA Class
Mean relative LVEF increase between preoperative and postoperative LVEF ranged from 10% to 72% in group 1 and ranged between 5.1% and 33.4% in group 2 ( Tables 3 and 4 ). In both groups, volume reduction was effective in improving NYHA class reduction. In group 1, the mean NYHA reduction ranged from 0.9 to 2.1 ( Table 3 ). Mean NYHA reduction in group 2 ranged between 1.1 and 1.7 ( Table 4 ). The weighted average LVEF increase was 38% in group 1 ( n = 1741) and 30.0% in group 2 ( n = 1586) ( Tables 3 and 4 ). The weighted average NYHA class reduction was 1.5 in group 1 and 1.2 in group 2 ( n = 1691) ( Tables 3 and 4 ). The weighted averages for LVEF increase and NYHA class reduction remained effectively unchanged in the sensitivity analysis for both groups ( Tables 3 and 4 ).
| Author and year | Volume reduction in % | Assessed for LVEF ( n ) | % of LVEF increase | p -value | Assessed for NYHA class ( n ) | NYHA class reduction | p -value |
|---|---|---|---|---|---|---|---|
| Athanasuleas, 1998 | 39.0 | 8 | 72.0 | <0.008 | |||
| Di Donato, 2001 | 58.9 | 207 | 37.1 | <0.0001 | |||
| Suma, 2001 | 41.5 | 50 | 58.7 | <0.05 | |||
| Athanasuleas, 2001 | 36.7 | 354 | 34.5 | <0.0001 | |||
| Di Donato, 2001 | 43.8 | 44 | 35.3 | <0.0001 | |||
| Menicanti, 2002 | 35.7 | 46 | 10.0 | 0.11 | 38 | 1.6 | <0.01 |
| Fuji, 2004 | 34.6 | 14 | 16.7 | <0.05 | |||
| Kokaji, 2004 | 53.0 | 15 | 62.1 | <0.05 | |||
| Di Donato, 2004 | 52.1 | 30 | 50.0 | 0.001 | |||
| Menicanti, 2004 | 37.1 | 108 | 17.2 | 0.001 | 108 | 1.5 | 0.0001 |
| Cirillo, 2004 | 42.7 | 65 | 39.1 | <0.0001 | 61 | 1.1 | NM |
| Yamaguchi, 2005 | 52.6 | 20 | 75.0 | <0.0001 | |||
| Tulner, 2006 | 45.7 | 21 | 33.3 | 0.007 | 21 | 2.0 | <0.05 |
| Ferrazzi, 2006 | 36.9 | 65 | 55.8 | <0.0001 | 65 | 1.6 | <0.0001 |
| Cirillo, 2006 | 40.1 | 12 | 53.7 | <0.05 | |||
| Aguiar Ribeiro, 2006 | 39.3 | 38 | 52.5 | <0.001 | 38 | 1.7 | NS |
| Takeda 2008 | 46.8 | 72 | 56.0 | <0.0001 | 49 | 1.3 | <0.0001 |
| Suma,2009 | 40.0 | 70 | 33.7 | <0.001 | 42 | 2.1 | NM |
| Bové, 2009 | 48.7 | 21 | 44.4 | <0.05 | 16 | 1.8 | <0.001 |
| Di Donato, 2009 | 40.6 | 67 | 25.9 | 0.0001 | 67 | 1.2 | 0.0001 |
| Baravelli, 2010 | 35.2 | 65 | 31.0 | <0.0001 | 65 | 0.9 | <0.0001 |
| ten Brinke, 2010 | 40.5 | 9 | 27.8 | <0.001 | 9 | 1.9 | <0.01 |
| Zhong, 2011 | 34.2 | 40 | 19.2 | <0.001 | |||
| Dor, 2011 | 51.0 | 101 | 69.2 | <0.001 | |||
| Shudo, 2011 | 42.3 | 21 | 40.0 | <0.05 | |||
| Skelley, 2011 | 42.4 | 19 | 41.4 | <0.05 | |||
| Cho, 2012 | 47.6 | 40 | 20.7 | <0.0001 | |||
| Cho, 2014 | 41.0 | 99 | 25.9 | <0.0001 | 99 | 1.5 | 0.0001 |
| Castelvecchio, 2021 | 43.2 | 20 | 56.1 | 0.0009 | |||
| Total n = 1741 | Weighted average = 38.0% | Total n = 678 | Weighted average = 1.5 | ||||
| Sensitivity analysis | Total n = 1561 | Weighted average = 37.0% | Total n = 632 | Weighted average = 1.4 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
