Treatment Strategies for Chronic Coronary Heart Disease with Left Ventricular Systolic Dysfunction or Preserved Ejection Fraction—A Systematic Review and Meta-Analysis

Round 1

Reviewer 1 Report

This paper describes results of a systematic review and meta-analysis of randomised trials comparing invasive with medical treatments in the management of coronary heart disease. It appears to have been performed with due attention to the specifics of this study type, and it has been written with good technical use of language. The study and its findings are important, though I find that the paper's construction and the manner in which the results have been presented, together with some ambiguity in use of terminology and perhaps understated limitations, make the study's findings difficult to follow and assess. There are also a few issues that might not have received sufficiently focused attention. I present these below.

There is a recurrent ambivalence in this paper in which terminology changes without clear reason. The title speaks of chronic ischaemic heart disease whereas the text often speaks of coronary heart disease - consistency would aid clarity. Similarly, comparison is both between the primary disease with and without revascularisation therapy and between patients with and without reduced left ventricular function, essentially creating a four-way classification. On top of this there is inconsistency in terminology and use of acronyms that are not always clear. In presenting the results of the investigation these complexities could be better managed so as to make the findings more readily available to the reader. I would strongly advise the authors to specifically consider reviewing the paper to improve its consistency, while if terms are viewed as synonymous, that might be declared rather than assume appropriate interpretation, then use the preferred term consistently.

Be consistent in use of terms, e.g., MR & OMT, invasive versus medical, invasive versus conservative, optimal versus appropriate. Please ensure you define all acronyms at first use and in all tables and figures. Since this is an open access journal, the familiarity of potential readers with these terms may be incomplete. Readers should not need to search for interpretation.

Please define "unplanned coronary revascularisation" at the outset.

Figures 2 & 3. Please review the legend, the text with the asterisk is confusing. The meaning only becomes clear at line 460 - too late for readers reviewing the paper superficially. Additionally, the procedure adopted of subtracting results is questionable, especially without the ability to consider data on the characteristics of the "created" group, even if partially addressed by weighting. The potential impacts of this strategy deserve more space in addressing limitations.

All figures - please ensure the legends make figures fully interpretable without need to refer to the text and are consistent in style throughout the paper. Study year does not need to be entered twice. Please consider placing outcome of interest (e.g., all-cause mortality, cardiovascular mortality) somewhere in the headers of the tables and restructure to clearly differentiate between events (meaning number of events of outcome of interest) and total study group size. It will also help to indicate in the legend that the totals referred to in the header are the total number of events for which information on all-cause mortality was included in each study. Please indicate or reference how forest plots are structured - there are different options that can be employed in building a forest plot, please indicate the convention you have followed.

Table 1. Perhaps change the header for the second column to "Study Years/Follow-Up".

Tables 2 & 3. These tables are extremely difficult to follow and appear to be incomplete. The numbers in parentheses are presumably percentages - please provide correction if I am wrong - and if so, percentages of what? Layout can be vastly improved by using right justification rather than central justification within the columns. It is not clear what the reader is supposed to interpret from these tables.

Line 266. Consider changing "regarding the" to "in relation to"

Line 269. "… reduced recommendations" - presumably this means reduced recommendations for revascularisation?

Line 285. This decision is entirely legitimate in the context of what the authors are attempting to achieve, but might indicate the need for greater emphasis on this constraint in the abstract and introduction. This is a question of exclusion criteria.

Line 313. Consider changing "may be" to "may also be"

Line 360 - is this for each scenario?

Line 404. Should this not say "and in the production of mitochondrial energy"?

Line 407. This paragraph should be joined to the following paragraph, plus its relevance is not entirely clear. Surely, if non-compliance is so high in revascularisation patients then this factor needs to be taken into account in assessing efficacy in randomly controlled trials? Perhaps efficacy would be far higher if patients were compliant with medical therapy, introducing a "wildcard" factor that potentially invalidates the results from the trials quoted. Perhaps this issue needs to be given much greater emphasis and the difference between the two cohorts studied in the current meta-analysis controlled. Certainly, this needs to be given greater emphasis (if you are going to mention it at all), and could have a similar impact to the paucity of patient data in general.

Line 463. Change "to that our analysis" to "to note that our analysis".

Line 483. "appropriate" - is this the same as optimal, seems odd to change the word at this point.

Line 451. "Standard practice in analysed studies" could be replaced with "frequent feature of trials studying treatment outcomes in ischaemic heart disease". The term used raises the question of whether the studies you actually used in your meta analysis might not in fact indicate a systematic bias in the conclusions drawn here.

Line 461. This is another example of ambiguity in use of terminology. Presumably the term heart failure is used as being synonymous with LV dysfunction. If so, a declaration early in the paper to that effect as to how terms are being used might avoid confusion rather than rely on assumptions concerning how different clinicians use terminology. If these two are seen as being different, then that needs to be explained.

Line 463. The lack of patient-specific information should be declared in Materials And Methods and the issue given more space in Limitations. Failure to address this issue could indeed be a major source of bias in many supposedly randomised trials and thus meta-analyses.

Author Response

The team of authors would like to thank the Reviewer for detailed analysis of our article and constructive criticism. We believe that taking into account the Reviewer's comments allowed us to significantly improve the quality of the manuscript.

According to the recommendations of the respected Reviewer, the authors have made changes to the text of the article to avoid discrepancies and ambiguities in terminology, in particular: “chronic coronary heart disease”, “optimal medical therapy”.

In order to identify the pattern or differences in the effect of revascularization depending on the presence of LV systolic dysfunction, we distinguished 2 scenarios in the analysis: 1 - comparison of studies with preserved LV EF and 2 - studies with reduced LV EF (<45%). Thus, meta-analysis of the results of studies with reduced LVEF allows us to emphasize the higher efficiency of revascularization in this category of patients.

The terms "revascularization", "invasive and conservative therapy" are widespread and generally accepted. The authors believe that their use in the article allows a varied presentation of the terminology. All abbreviations are defined when first used in the main text, figures and tables;

The term "unplanned coronary revascularization" is defined at first mention.

Comment «Figures 2 and 3. Please review the legend, the text with the asterisk is confusing. The meaning only becomes clear at line 460 - too late for readers reviewing the paper superficially» is taken into account.

Comment «All figures - please ensure the legends make figures fully interpretable without need to refer to the text and are consistent in style throughout the paper» - is taken into account. The figure titles include endpoints, the legend indicates that total means the sum of all events in each study.

According to a comment «Study year does not need to be entered twice», group of authors considers it necessary to leave the years of RCTs, as intermediate results of these studies are found in the literature to avoid contradictions and there was no need to apply to the text to clarify the year of publication.

Comment «Please indicate or reference how forest plots are structured - there are different options that can be employed in building a forest plot», we would like to clarify that the meta-analysis used a random-effects model or a fixed-effect model. When the heterogeneity of study results exceeded 40%, the random effects model was used. There are indications in the figures: Random or Fixed.

Comments on «Table 1 and Table 2-3; Line 266; Line 269; Line 313; Line 360; Line 404; Line 407; Line 451; Line 463; Line 483» is taken into account.

As per the comment in line 285, we would like to clarify that we did not include studies from the past in which medical therapy differed from current views. We also did not include studies that did not directly compare the outcomes of optimal medical therapy with revascularization. There are numerous publications in the literature analyzing past studies that detail the limitations, and the author team would like to refrain from repeating already published material.

According to Line 461 comment «heart failure or LV dysfunction», It should be clarified that the name of the subgroup was borrowed from the original source Lopes RD et al. «Initial Invasive Versus Conservative Management of Stable Ischemic Heart Disease in Patients With a History of Heart Failure or Left Ventricular Dysfunction Insights From the ISCHEMIA Trial»,  according to the design of the study presented.

The authors agree with Reviewer Line 463's comment that lack of baseline patient data may be a major source of systematic bias in many studies and hence in meta-analyses. To mitigate this risk, we used only published data from RCTs. The possible presence of bias was analyzed by visual assessment of the Funnel diagram, and no pronounced asymmetry was found (Supplementary Appendix, Figure 1.1). RoB2 was also used to assess the risk of bias and level of evidence.

Reviewer 2 Report

pathophysiology-2760014, Treatment strategies for chronic ischemic heart disease with left ventricular systolic dysfunction or preserved ejection fraction - a systematic review and meta-analysis by Elena Zelikovna Golukhova et al.

Comments:

Abstract: 

- The reviewer suggests including the name of the statistical methods or tools used in your analysis.

- The contrast between outcomes in patients with preserved LV EF and those with LV systolic dysfunction is well-presented. However, you might want to emphasize this contrast a bit more to highlight the significance of your findings.

Introduction: 

- The reviewer suggests highlighting  how the findings of this meta-analysis might impact current clinical guidelines or practice.

- A little more depth could be added to the discussion on myocardial revascularization. For instance, briefly discussing different revascularization techniques and how they might differ in efficacy for the two patient groups.

- If possible, reference a few key studies that have attempted to address this issue but have left gaps that your research aims to fill.

The search strategies and study selection:

- A brief explanation of why studies conducted after 2007 were specifically included (e.g., advancements in revascularization techniques) would be beneficial. This adds context to your selection criteria.

- Expand on why certain types of studies (e.g., clinical case reports, expert opinions) were excluded. This helps in understanding how these exclusions might influence the study's outcomes.

Discussion:

- Given the significant findings in patients with LV systolic dysfunction, a more detailed discussion on the implications for this patient group would be beneficial.

- Are there findings that contradict existing literature? If possible, discuss potential reasons for these discrepancies.

- Provide specific recommendations for future studies, such as exploring long-term outcomes or examining different patient subgroups.

Minor editing of English language required

Author Response

The team of authors would like to thank the Reviewer for careful reading of the article and comments. We tried to take into account all recommendations and added the name of statistical methods and tools used in our meta-analysis to the abstract. We presented the main studies on this topic and discussed different methods of revascularization.

The Reviewer recommends explaining why certain types of studies were excluded. We respond that only RCTs were included as the most accurate way of identifying causal relations between treatment and the underlying disease. Non-randomized controlled trials cannot exclude the possibility of some random factor influencing therapy or outcome. Clinical cases or expert opinion do not provide statistical power and do not minimize reliance on individual outcomes or expert incorrect judgement.

In accordance with the Reviewer's comments in the Discussion, we clarify that the results of the meta-analysis do not contradict the existing clinical guidelines for the management of patients with chronic CHD. Based on the meta-analysis, the following conclusions were drawn: new data are needed to objectively validate the findings to explore the optimal treatment strategy for patients with chronic CHD and LV systolic dysfunction. In future studies, it is necessary to randomize patients according to the degree of LV EF reduction and to evaluate separately the event rate depending on the method of myocardial revascularization in patients with chronic CHD.

Round 2

Reviewer 2 Report

I would like to thank the authors for addressing the comments raised in the initial round of review.