The Role of QRS Complex and ST-Segment in Major Adverse Cardiovascular Events Prediction in Patients with ST Elevated Myocardial Infarction: A 6-Year Follow-Up Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I am grateful to the editor for the opportunity to review the manuscript by Maletin et al, “The Role of QRS Complex and ST-segment in Major Adverse Cardiovascular Events Prediction: A 6-Year Follow-Up Study.” In this article, the authors carefully analyzed the dynamics of the QRS Complex and ST-segment during PCI against the background of ST-segment elevation myocardial infarction, and also studied the prognostic significance of these changes at 6-year follow-up. It should be noted that at present, ECG criteria for an unfavorable prognosis in myocardial infarction are not very popular, taking into account the emergence of other prognostic factors (biomarkers, echocardiogram data, coronary angiography, etc.). The authors were able to identify some associations of QRS Complex and ST-segment with MACE during a 6-year follow-up of patients after myocardial infarction; this may perhaps be of some interest.

However, during the review, I had comments and questions to which I would like to receive answers from the authors:

1. The main fundamental question that the authors would have to answer is the prognostic value of the QRS Complex and ST-segment in these patients, regardless of other known prognostic factors, or is it a reflection of them? In the binary regression analysis model used by the authors, it is impossible to answer this question, since only ECG indicators were included in it. The authors did not include other potentially significant prognostic factors (CPK and troponin levels, LVEF, Killip class, duration of symptoms - see ref. 1 below) in the model.

2. When clinically characterizing the groups, the authors did not provide important information about the patients (level of NT-proBNP, localization of myocardial infarction, number of affected coronary arteries during coronary angiography, number of stented vessels, etc.). These factors also deserved inclusion in the multivariate analysis model of MACE predictors.

3. The title is incomplete - there is no designation of the patients on whom the study was performed.

4. In the names of the tables there is also no mention of which category of patients the presented data were analyzed for.

5. Table 3 appears twice in the manuscript.

6. Abstract needs correction. The phrase “Statistical analysis was performed” (lines 18-19) is clearly redundant.

7. It is unclear how, when including patients, it was possible to form absolutely identical groups of 100 people. Was this a predetermined goal? I can't believe that this happened completely by accident. In any case, it is necessary to include a patient inclusion flowchart in the manuscript.

8. It is necessary to provide information about the equipment used (ECG, echocardiography, coronary angiography).

9. Table 2 does not show units of measurement for Dynamics of the QRS complex

10. In the discussion, I would like to see the opinion of the authors - is their proposed assessment of the QRS Complex and ST-segment really necessary to predict MACE, or will traditional risk markers be sufficient?

Refrences:

1. Deng W, Wang D, Wan Y, Lai S, Ding Y, Wang X. Prediction models for major adverse cardiovascular events after percutaneous coronary intervention: a systematic review. Front Cardiovasc Med. 2024 Jan 8;10:1287434. doi: 10.3389/fcvm.2023.1287434.

Comments on the Quality of English Language

No comments

Author Response

Dear Reviewers,

Thank you for your valuable feedback. We are confident that it will enhance the quality of our paper. We have made the necessary revisions following your instructions, and we have listed them below in the order in which they appear. We have also indicated the specific parts of the manuscript that were modified for each revision.

 

Reviewer, # 1

Comments and Suggestions for Authors

I am grateful to the editor for the opportunity to review the manuscript by Maletin et al, “The Role of QRS Complex and ST-segment in Major Adverse Cardiovascular Events Prediction: A 6-Year Follow-Up Study.” In this article, the authors carefully analyzed the dynamics of the QRS Complex and ST-segment during PCI against the background of ST-segment elevation myocardial infarction, and also studied the prognostic significance of these changes at 6-year follow-up. It should be noted that at present, ECG criteria for an unfavorable prognosis in myocardial infarction are not very popular, taking into account the emergence of other prognostic factors (biomarkers, echocardiogram data, coronary angiography, etc.). The authors were able to identify some associations of QRS Complex and ST-segment with MACE during a 6-year follow-up of patients after myocardial infarction; this may perhaps be of some interest.

However, during the review, I had comments and questions to which I would like to receive answers from the authors:

1. The main fundamental question that the authors would have to answer is the prognostic value of the QRS Complex and ST-segment in these patients, regardless of other known prognostic factors, or is it a reflection of them? In the binary regression analysis model used by the authors, it is impossible to answer this question, since only ECG indicators were included in it. The authors did not include other potentially significant prognostic factors (CPK and troponin levels, LVEF, Killip class, duration of symptoms - see ref. 1 below) in the model.

 

 

In the model where QRS width pre PCI controlled for only CPK and TNI QRS width remained significant.

 

MACE	Odds ratio	Std. err.	z	P>z	[95% conf.	interval]
QRS pre-PCI	1.027325	0.0129865	2.13	0.033	1.002185	1.053096
CPK MB pre-PCI	1.003071	0.0034383	0.89	0.371	.9963542	1.009832
TNI 0	1.012977	0.0970325	0.13	0.893	.8395824	1.222181
_cons	0.0243191	0.0316703	-2.85	0.004	.0018943	0.3122134

 

 

 

 

 

 

 

 

 

 

 

 

 

In multivariate model with all variables you suggested Killip code was the strongest predictor of MACE.

MACE	Odds ratio	Std. err.	z	P>z	[95% conf.	interval]
QRS pre-PCI	1.005941	0.0160751	0.37	0.711	0.9749229	1.037947
CPK MB pre-PCI	1.008128	0.0050329	1.62	0.105	0.998312	1.018041
TNI0	0.7274658	0.1039791	-2.23	0.026	0.5497275	0.9626706
EF hos	0.9473089	0.0292322	-1.75	0.079	0.8917131	1.006371
Kilip code	11.84407	5.042011	5.81	0.000	5.142158	27.28078
Sympthom duration	1.000925	0.0012829	0.72	0.471	0.9984135	1.003442
_cons	0.0652801	0.1736838	-1.03	0.305	0.0003549	12.00836

 

 

As our objective was to determine the impact of ECG QRS width and ST elevation magnitude on MACE. The univariate analysis revealed that these parameters are associated with MACE. It is already known from previous studies that Killip classification and EF are reliable predictors of MACE in STEMI patients. If you suggest I would recommend to include model with only CPKMB and TNI0 or to leave only univariate regression.

The similar situation is for ST elevation. Multivariable regression in tables

MACE	Odds ratio	Std. err.	z	P>z	[95% conf.	interval]
ST elevation pre-PCI mm	0.9282297	0.0818851	-0.84	0.399	0.7808461	1.103432
CPK MB pre-PCI	1.009098	0.0051637	1.77	0.077	0.9990277	1.019269
TNI 0	0.707397	0.1039681	-2.36	0.019	0.5303466	.9435537
EF hos	0.9351743	0.0301337	-2.08	0.038	0.8779397	.9961401
Kilip code	13.36212	6.036564	5.74	0.000	5.512274	32.39068
Symptom duration	1.000601	0.0013033	0.46	0.645	0.9980494	1.003158
_cons	0.2704874	0.5182964	-0.68	0.495	0.0063257	11.56599

 

MACE	Odds ratio	Std. err.	z	P>z	[95% conf.	interval]
ST elevation pre-PCI mm	1.165351	0.0758857	2.35	0.019	1.025718	1.323993
CPKMB pre-PCI	1.002997	0.0034725	0.86	0.387	0.996214	1.009826
TNI0	1.0225	0.0989468	0.23	0.818	0.8458492	1.236043
_cons	0.2018966	0.0689962	-4.68	0.000	0.1033331	0.3944742

 

 

2. When clinically characterizing the groups, the authors did not provide important information about the patients (level of NT-proBNP, localization of myocardial infarction, number of affected coronary arteries during coronary angiography, number of stented vessels, etc.). These factors also deserved inclusion in the multivariate analysis model of MACE predictors.

 

I am sorry currently levels of NT pro BNP’s are not available. Myocardial infarction localization is added.

In univariant analysis these factors were not significantly assossiated with MACE so they were not included in model.

 

3. The title is incomplete - there is no designation of the patients on whom the study was performed.

The study was performed on a specific group of patients, which is now clearly described in the title. (The Role of QRS Complex and ST-segment in Major Adverse Cardiovascular Events Prediction in patients with ST elevated myocardial infarction: A 6-year Follow-Up Study)

 

4. In the names of the tables there is also no mention of which category of patients the presented data were analyzed for.

Thank you for this suggestion we corrected the names of the tables accordingly:

Table 1. Basic ST-segment elevation myocardial infarction (STEMI) patients characteristics

Table 2. ECG characteristics at baseline and during the follow-up

Table 3. MACE in STEMI patients

Table 4. Association of QRS complex width and MACE in STEMI patients

also in this table we corrected the typo instead of mm width was changed in msec

Table 5. Association of ST segment elevation and MACE in STEMI patients

 

 

5. Table 3 appears twice in the manuscript.

Corrected

6. Abstract needs correction. The phrase “Statistical analysis was performed” (lines 18-19) is clearly redundant.

Corrected

7. It is unclear how, when including patients, it was possible to form absolutely identical groups of 100 people. Was this a predetermined goal? I can't believe that this happened completely by accident. In any case, it is necessary to include a patient inclusion flowchart in the manuscript.

Thank you for your valuable comment we do realize that it is unclear how the sample and groups were formed we corrected it in the methodology. Consecutive patients who fulfilled inclusion and exclusion criteria were included up to target of 100 patients per group. One group consisted of patients who had symptom onset  to balloon time less then 6 hours and the other group consisted of patients who had symptom onset to balloon time in time frame 6-12 hours. We changed the text accordingly:

Consecutive patents who fulfilled inclusion and exclusion criteria were included in the study. Based on the duration of the symptoms to balloon time, patients were divided into two groups: (I) Group A, in which symptoms lasted up to 6 hours from the onset to the balloon, and (II) Group B, in which symptoms lasted from 6 to 12 hours from the onset to the balloon. The target number of the patients was same for the both groups and it was one hundred.

 

8. It is necessary to provide information about the equipment used (ECG, echocardiography, coronary angiography).

We have added text in metodology:

2.2. Electrocardiographic data

All patients underwent a resting 12-lead surface ECG record at a paper speed of 50mm/sec and gain of 10mm/mV using Schiller ECG machine CARDIOVIT AT-102, Baar, Switzerland. One cardiologist interpreted each ECG.

2.3 Coronary angiography

All patients underwent urgent coronary angiography which was performed using Artis Z catheterization laboratory (Simens Medical Solutions USA, Malvern, PA, United States of America).

 

2.4 Echocardiography

All participants underwent two-dimensional transthoracic echocardiography (ECHO) using a GE Healthcare Vivid E9 machine equipped with an M5S-D, 1.5–4.6 MHz transducer, while continuous ECG monitoring was carried out during the examination.

9. Table 2 does not show units of measurement for Dynamics of the QRS complex

Units added

10. In the discussion, I would like to see the opinion of the authors - is their proposed assessment of the QRS Complex and ST-segment really necessary to predict MACE, or will traditional risk markers be sufficient?

Refrences:

1. Deng W, Wang D, Wan Y, Lai S, Ding Y, Wang X. Prediction models for major adverse cardiovascular events after percutaneous coronary intervention: a systematic review. Front Cardiovasc Med. 2024 Jan 8;10:1287434. doi: 10.3389/fcvm.2023.1287434.

Deng W et al. in recent publication concluded that there is rising number of modeling studies with questionable predictability and that there is a need for external validation. Our paper analyzed two ECG parameters which applicability in practice is easy and simple. We do agree that sensitivity and specificity is moderate however, we found that there is a good correlation of this parameters with MACE. One of the biggest flaws of this parameters is that they can be affected with other diseases such as conduction system disorders etc. So, application of this parameters can be used in patients with first manifestation of STEM without prior structural and rhythmical disorders. However, future research and external validation might possibly include these easily measured ECG parameters in predictive models, where their real predictive power would be obtained.

Reviewer 2 Report

Comments and Suggestions for Authors

The reviewed manuscript “The Role of QRS Complex and ST-segment in Major Adverse Cardiovascular Events Prediction: A 6-year Follow-Up Study” by Srđan Maletin et al concerns a possible predictive role of some electrocardiographic indices in the development of cardiac complications following the STEMI PCI treatment.Indeed, in a number of previously published studies it has been shown that some electrocardiographic changes can serve as a predictor of the development of, for example, heart failure. The authors of this study focused on the prognostic value of the width of the QRS complex and the dynamics of the ST segment.

The authors found (statistically!) some correlations between the width of the QRS complex and ST-segment dynamics with the outcomes of percutaneous intervention after STEMI. However, it seems to me that many of the conclusions in the paper are the result of excessive reliance on statistical methods, while they require serious analysis and comprehension.

First, it is obvious that the difference in heart rate in the patients of the first and second groups should have led to a difference in the duration of the QRS complex. What is more prognostically significant here: HR or QRS width? This issue needs to be discussed.

Second. As can be seen from Table 1, the enzymes indicating cardiomyocyte damage (MB CK and Troponin) were significantly lower in the patients of the first group at all times (after 6 hours, after 72 hours), except when measured after 24 hours. This is obviously a problem of statistics and measurement accuracy, which should be paid attention to. If the authors believe that this fact has any practical meaning, then such an assumption should be explicitly formulated and discussed. Statistics played the same joke on the authors when finding out the presence of signs of heart failure in patients upon admission. Killip 1 in patients of the first group is significantly higher than in patients of the second group, but when divided into Killip 2, 3 and 4, the reliability of the differences disappears. Obviously, the total number of patients with signs (more or less serious) of heart failure in the second group should be significantly higher than in the first, and this should be pointed out.

Lines 162-166. The duration of the QRS complex is indicated in millimeters. It is clearly erroneous; there should be milliseconds. Also, in Table 2, you need to specify the dimension of the width of the QRS complex (there are also obviously milliseconds).

Table 2 also shows the ST segment elevation after 1 month and 6 months of 0.18 and 0.14 mm, which, in my opinion, is meaningless. Even less clear is the indication of a high significant difference between these values in patients of the first and second groups, despite the fact that the mean in both cases is 0. In general, it seems to me that it would be more informative if the authors did not give the spread of results for each parameter, but the standard deviation (Mean + s.d.).

Finally, the authors conclude (line 244) that "an increase of pre-PCI QRS complex width by 1 msec increases the risk of 6-year MACE by 3%". An increase compared to what? If we are talking about the elongation of the QRS complex by 1 msec, I must say that the variability of the duration of this complex in the analysis from pulse to pulse (beat by beat) not less than 1 msec and therefore, in my opinion, such a change cannot have any predictive value.

Minor point.

It seems unfortunate to me to label a highly reliable difference as 0.000 and a non-significant difference as 1.000. Of course, a program on a computer can produce a large number of digits after a decimal point, but it looks very strange. Maybe it's better to go back to the standard p<0.001 and p>0.1?

 

Finally, some phrases look very strange and difficult to understand. The article clearly needs editing by a native English speaker.

 

Comments on the Quality of English Language

The article needs editing by native English speaker.

Author Response

Dear Reviewers,

Thank you for your valuable feedback. We are confident that it will enhance the quality of our paper. We have made the necessary revisions following your instructions, and we have listed them below in the order in which they appear. We have also indicated the specific parts of the manuscript that were modified for each revision.

 

1. First,it is obvious that the difference in heart rate in the patients of the first and second groups should have led to a difference in the duration of the QRS  What is more prognostically significant here: HR or QRS width?

Thank you for this interesting comment we found that there is no signified difference in QRS complex width at the baseline despite the fact that there was significant difference in heart rate. Therefore, we believe that differences in QRS width during the follow up were consequence on grater damage condacting system in patients who had longer ishemia time

2. the enzymes indicating cardiomyocyte damage (MB CK and Troponin) were significantly lower in the patients of the first group at all times (after 6 hours, after 72 hours), except when measured after 24 hours. This is obviously a problem of statistics and measurement accuracy, which should be paid attention to. If the authors believe that this fact has any practical meaning, then such an assumption should be explicitly formulated and discussed. Statistics played the same joke on the authors when finding out the presence of signs of heart failure in patients upon admission. 

Thank you for your comment. Patients with shorter ischemia time had lower values of cardio-specific enzymes due to the fact that myocardial damage is lower. However after revascularization and reperfusion there is wash-out phenomenon which probably led to elevation of enzymes I both groups. However normalization of the enzymes was faster in the group with shorter ischemia time

3. Killip 1 in patients of the first group is significantly higher than in patients of the second group, but when divided into Killip 2, 3 and 4, the reliability of the differences disappears. Obviously, the total number of patients with signs (more or less serious) of heart failure in the second group should be significantly higher than in the first, and this should be pointed out.

Thank you for this interesting comment we believe that exclusion criteria had influence on Killip class since the patient with severe heart failure were not include in study

4. Lines162-166. The duration of the QRS complex is indicated in  It is clearly erroneous; there should be milliseconds.Also, in Table 2, you need to specify the dimension of the width of the QRS complex (there are also obviously milliseconds).

Corrected

5. Finally,the authors conclude (line 244) that "an increase of pre-PCI QRS complex width by 1 msec increases the risk of 6-year MACE by 3%". An increase compared to what? If we are talking about the elongation of the QRS complex by 1 msec, I must say that the variability of the duration of this complex in the analysis from pulse to pulse (beat by beat) not less than 1 msec and therefore, in my opinion, such a change cannot have any predictive value.

 

Thank you for this important comment. We missed adding that the increase is related to the normal cut-off value for QRS width which is 99 msec. We change the text accordingly. 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors did some work to correct the manuscript. However, I still have questions about the statistics.

It is not clear to me why the authors, out of many known prognostic factors, left in their model only those that indicate the prognostic value of ECG data. At a minimum, it is necessary to add to the text of the manuscript those tables that the authors provided in response to my comments. Then discuss them in the Discussion section.

Comments on the Quality of English Language

No comments

Author Response

Reviewer 1

It is not clear to me why the authors, out of many known prognostic factors, left in their model only those that indicate the prognostic value of ECG data. At a minimum, it is necessary to add to the text of the manuscript those tables that the authors provided in response to my comments. Then discuss them in the Discussion section.

The answer is that study aims to examine the prognostic significance of ECG factors as predictors of outcome.

Tables were added and discussed thoroughly.

Reviewer 2 Report

Comments and Suggestions for Authors

After editing, the article began to look noticeably better, but I still have a number of questions.

11.    Statistical factors play a very important role in the work. The good thing is that a number of variables are now represented as an average plus-minus some value. But what is this value? Is it standard deviation or standard error? This should be clearly defined in the “Statistical analysis” section.

22.    It needs to be clarified why some variables are represented as an average and a spread while others as an average plus-minus an error. Refer to the phrase "Continuous variables were presented as medians (interquartile range) and analyzed by...". (line 223), one can think that BMI is a continuous variable whereas body surface not. And so on.

33.    I am still surprised by the ST segment elevation data highlighted in yellow in Table 2. The authors in the Discussion section again simply point out that "ST-segment elevation after one month by 1mm increases the risk of 6-year MACE by 3.26 times" (Line 458), although according to Table 2 such an elevation of ST segment as well as significant difference in ST segment levels between the groups after 1 and 6 months seems impossible.  Still, this issue needs not a simple statement, but a serious discussion.

Author Response

Reviewer 2

After editing, the article began to look noticeably better, but I still have a number of questions.

11. Statistical factors play a very important role in the work. The good thing is that a number of variables are now represented as an average plus-minus some value. But what is this value? Is it standard deviation or standard error? This should be clearly defined in the “Statistical analysis” section.

 

This value is standard deviation (SD). Statistical analysis and table corrections have been made accordingly.

 

22. It needs to be clarified why some variables are represented as an average and a spread while others as an average plus-minus an error. Refer to the phrase "Continuous variables were presented as medians (interquartile range) and analyzed by...". (line 223), one can think that BMI is a continuous variable whereas body surface not. And so on.

 

This has been corrected

 

33. I am still surprised by the ST segment elevation data highlighted in yellow in Table 2. The authors in the Discussion section again simply point out that "ST-segment elevation after one month by 1mm increases the risk of 6-year MACE by 3.26 times" (Line 458), although according to Table 2 such an elevation of ST segment as well as significant difference in ST segment levels between the groups after 1 and 6 months seems impossible.  Still, this issue needs not a simple statement, but a serious discussion.

 

ST segment data highlighted in table 2 are median [IQR] values of different ST segment elevation time points compared between groups.

And in table 5 we reported a logistic regression analysis of the Association of ST-segment elevation and MACE in STEMI patients. Based ot these OR we stated that for 1mm increase in ST segment elevation at one month risk for MACE increases for 3.26 times.

This has been discussed in the discussion section.

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

The authors responded to my comments and made corrections to the text of the manuscript. I have no other comments.

Comments on the Quality of English Language

No comments

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have made the necessary corrections and I think that the article may be published as it stands.