Effect of ischaemic postconditioning on markers of myocardial injury in ST-elevation myocardial infarction: a meta-analysis

Abstract

Objectives

This study aimed to perform a meta-analysis of the short-term impact of ischaemic postconditioning (IPoC) on myocardial injury in ST elevation myocardial infarction (STEMI) using surrogate cardiac biomarkers.

Methods

Eligible studies were identified using several article databases. Randomised controlled trials published between 1 January 2000 and 1 December 2021 comparing IPoC to standard of therapy in STEMI patients were included in the search. Outcomes included surrogates of myocardial injury, specifically peak troponin, creatine-kinase (CK) and CK myoglobin binding (CK-MB) enzyme levels.

Results

11 articles involving 1273 patients reported on CK-MB and 8 studies involving 505 patients reported on CK. Few studies used troponin as an outcome, thus, a subanalysis of troponin dynamics was not performed. Meta-regression analysis demonstrated no significant effect of IPoC on peak CK-MB (effect size −0.41, 95% CI −1.15 to 0.34) or peak CK (effect size −0.42, 95% CI −1.20 to 0.36). Linear regression analysis demonstrated a significant correlation between a history of smoking and CK-MB in the IPoC group (p=0.038).

Conclusions

IPoC does not seem to protect against myocardial injury in STEMI, except possibly in smokers. These results resonate with some studies using imaging techniques to ascertain myocardial damage. More research using troponin and cardiac imaging should be pursued to better assess the effects of IPoC on cardiovascular outcomes in STEMI.

WHAT IS ALREADY KNOWN ON THIS TOPIC.

• Ischaemic postconditioning (IPoC) has demonstrated inconsistent benefits in limiting myocardial injury in percutaneous coronary intervention (PCI) among ST-elevation myocardial infarction (STEMI) patients. Cardiac biomarkers are an acceptable marker for extent of injury but their correlation with IPoC has not been systematically evaluated.

WHAT THIS STUDY ADDS

• Based on the study of certain cardiac biomarkers, there is no clear evidence that IPoC during PCI in STEMI reduces myocardial injury, except possibly in smokers.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• IPoC seems to have a limited role in STEMI patients. Further research is necessary using more updated and advanced markers of myocardial infarction to better clarify its ultimate effects.

Introduction

Myocardial reperfusion via primary percutaneous intervention (PCI) or thrombolytic therapy has dramatically improved the outcomes in patients with ST elevation myocardial infarction (MI) (STEMI). Despite this, mortality among these patients remains significant, with a 30-day and 1-year mortality rate of 4% and 7%, respectively.¹ A seemingly important contributor of continued mortality in these patients is reperfusion-associated myocardial injury, which involves local vascular and cellular damage following abrupt restoration of blood flow.² Indeed, infarct size is strongly associated with all-cause mortality and is an important prognostic measure in patients with STEMI.³ Among the methods used to quantify infarct size, measuring cardiac biomarkers is a simple, cost-effective and accurate method.⁴ One proposed therapy to address reperfusion injury includes ischaemic postconditioning (IPoC), which involves brief repetitive and controlled bouts of ischaemia to reperfused vascular territories in patients with MI.^{5 6} In early studies using this method, IPoC was associated with up to a 36% reduction in infarct size using cardiac biomarker release as a surrogate.⁶ Unfortunately, however, larger and more recent clinical studies have provided inconsistent results regarding the potential benefit of IPoC when using both cardiac biomarkers and imaging to reflect infarct size. Specifically, data on cardiac biomarkers in the context of STEMI treated with IPoC have not been systematically studied. For this reason, this study is aimed to provide an updated meta-analysis of randomised controlled trials (RCTs) comparing STEMI patients treated with conventional reperfusion therapy versus IPoC with stenting using cardiac biomarkers as surrogates of infarct size.

Methods

Search strategy

Three databases were screened, including PubMed, Embase and Web of Science for all RCTs between 1 January 2000 and 1 December 2021 assessing the effect of IPoC versus standard of practice in treatment of STEMI. The search was conducted using search terms appropriate for each database as well as manual keywords. Search terms included “coronary disease”, “myocardial infarction”, “acute coronary syndrome”, “percutaneous coronary intervention” and “ischemic postconditioning”.

Eligibility criteria

Inclusion criteria were: (1) original prospective RCTs in English; (2) full-text publications including unpublished data; (3) IPoC with PCI versus PCI alone in the setting of acute STEMI; (4) inclusion of human participants; (5) studies reporting cardiac biomarkers of infarct size (troponin, CK, CK-MB) associated with STEMI using both mean and SD and (6) TIMI (Thrombolysis in Myocardial Infarction) 0–1 prior to intervention.

Exclusion criteria were: (1) retrospective studies, posters or abstracts; (2) IPoC of non-coronary vessels; (3) utilisation of preconditioning; (4) indications for cardiac catheterisations other than STEMI; (5) cardiogenic shock or arrest; (6) articles in another language other than English; (7) animal studies; (8) articles using patients from the same cohort as an original earlier study; (9) TIMI 2–3 prior to intervention; (10) significant statistical differences in patient characteristics between experimental and control groups and (11) requiring mechanical circulatory support.

Data selection process and quality assessment

Studies were independently reviewed and assessed by two investigators (IB and YW) for relevance according to the inclusion and exclusion criteria. A third investigator (UN) later assessed the articles chosen and clarified any uncertainties regarding inclusion or exclusion of articles, if occurred.

Statistical analysis

The effect size (ES) of IPoC on cardiac biomarkers for each study was estimated based on mean values and the SD in each intervention and control group. Unless otherwise mentioned by the study authors, a normal distribution was assumed for all included articles. Some articles calculated data using mean and SD, while others used median and IQR. To make data comparable among all articles, a founded statistical method was used such that median and IQR data of all articles with more than 25 participants with normally distributed data were assumed equal to the mean and SD.⁷ Random effects model meta-regression analysis was performed to estimate the ES based on the studies included. Accordingly, ES equalling 1 reflected similar biomarker levels, while values above or below 1 reflected difference in such.

Correlation analysis was performed comparing biomarker levels in both IPoC and control groups in order to assess the standardised effect of the entire sample. Parameters with 5 or less total data points per study group were not included in the analysis. Weighted results were expressed as mean and SE of mean. Relationships between the dependent variable (eg, peak measured values of the studied biomarkers) and independent variables (eg, age, sex, body mass index, cardiovascular risk factors, vital signs on presentation, time from symptom onset to intervention, Rentrop grade, use of thrombectomy devices, IPoC protocol used, coronary artery disease extent on angiography, total balloon inflation time, Killip class, number of stents placed, myocardial blush grade, left anterior descending artery involvement or three vessel disease or pharmacotherapy use prior to PCI) were evaluated using linear regression analysis.

Correlation analyses were estimated according to proportion of the variance in the dependent variable that was predictable from the independent variable (ie, R²). The number of patients in the included studies was used as a weighted variable. A p<0.05 was considered significant; two-tailed p values were used for all comparisons. Analyses were performed using JMP V.15 (SAS Institute).

Several assumptions were made when tabling data. If there was no mention of Gp2a/3b use, we assumed participants did not receive such intervention. Unless explicitly stated otherwise in the article, all participants were assumed to have undergone stent placement.

Results

Identification of studies and data handling

The data selection process is detailed in figure 1. Overall, 4079 articles were identified through the preliminary database searching, with 2021 remaining after removal of duplicates. These records were then screened, resulting in 58 full-text articles.

Figure 1.

Article selection flow chart. CK-MB, creatine-kinase myoglobin binding; PCI, percutaneous coronary intervention; RCT, randomised controlled trial; STEMI, ST elevation myocardial infarction.

After in-depth assessment, 40 articles were excluded due to the following: duplicates or same patient cohort (16 articles), no data on cardiac biomarkers (10 articles), non-STEMI patients (4 articles), not a prospective RCT (3 articles), no postconditioning performed (2 articles), animal study (1 article), significant differences in baseline characteristics among the study groups (1 article), cardiogenic shock patients included (1 article), no PCI performed (2 articles). In total, 18 articles were included in this meta-analysis. After further review, 8 articles contained CK data alongside adequate SD analysis, while 11 articles contained CK-MB data (figure 1). Given the lack of data on cardiac troponin (that was reported on by just two studies who met the inclusion criteria), the primary investigated outcomes were peak-CK and peak-CK-MB.

Due to the limited data available, correlations were not conducted for Killip class, number of stents placed, myocardial blush grade, left anterior descending artery involvement, the presence or absence of three vessel disease or pharmacotherapy use prior to PCI. Median data of 4 included articles (all including more than 25 participants) were converted to means, according to acceptable standards.⁷

Study Characteristics

Eleven studies (1273 patients) reporting on CK-MB and 8 studies (505 patients) reporting on CK were included in this study. In the CK-MB analysis, 638 patients underwent PCI with IPoC and 635 underwent standard PCI. In the CK analysis, 259 underwent PCI with IPoC while 246 patients underwent standard PCI. The main characteristics of the included articles involving CK and CK-MB are found in tables 1 and 2, respectively. Average levels of peak CK and CK-MB are demonstrated in tables 3 and 4.

Table 1.

CK studies—baseline characteristics

First author (refs.)	Year	Total no of patients (N)	IPoC/control (n)	Maximum time from onset of symptoms (hours)	Age (years)	Males (%)	DM (%)	HTN (%)	Smoking (%)	Average time symptoms to PCI (min)	LAD (%)	Pre-PCI TIMI 0 (%)	No of inflations (n)	Duration of inflation (sec)	Reflow between inflations (sec)	Total time of inflation (sec)
Traverse29	2019	122	65/57	6	59	85	18	53	37	n/a	45	100	4	30	30	120
Araszkiewicz30	2014	72	35/37	6	57	75	15	53	53	212	31	100	4	60	60	240
Freixa24	2012	79	39/40	12	60	78	20	50	57	328	48	96	4	60	60	240
Garcia31	2011	43	22/21	12	58	81	12	56	33	267	30	100	4	30	30	120
Laskey32	2008	24	12/12	6	59	58	40	79	68	225	100	71	2	90	180	180
Yang33	2007	41	23/18	n/a	n/a	n/a	n/a	n/a	n/a	n/a	n/a	44	3	30	30	90
Ma21	2006	94	47/47	12	64	68	41	59	0	411	51	67	3	30	30	90
Staat6	2005	30	16/14	6	57	83	17	37	57	324	40	100	4	60	60	240

CK, creatine-kinase; DM, diabetes mellitus; HTN, hypertension; IPoC, ischaemic postconditioning; LAD, left anterior descending ; n/a, not available; PCI, percutaneous coronary intervention; TIMI, Thrombolysis in Myocardial Infarction.

Table 2.

CK-MB studies—baseline characteristics

Authors (refs.)	Year	Total no of patients (N)	IPoC/control (n)	Maximum time from onset of symptoms (hours)	Age (years)	Males (%)	DM (%)	HTN (%)	Smoking (%)	Average time symptoms to PCI (min)	LAD (%)	Pre-PCI TIMI 0 (%)	No of inflations (n)	Duration of inflation (sec)	Reflow between inflations (sec)	Total time of inflation (sec)
Mukherjee and Jain34	2019	40	20/20	12	57	63	18	45	35	n/a	53	100	4	60	60	240
Liu et al35	2016	56	28/28	12	59	70	20	61	46	285	75	n/a	3	30	30	90
Araszkiewicz et al30	2014	72	35/37	6	57	75	15	53	53	212	31	100	4	60	60	240
Dong et al25	2014	62	32/30	12	69	68	36	68	45	297	50	100	3	30	30	90
Elżbieciak et al36	2013	39	18/21	12	59	77	23	85	59	275	100	n/a	4	60	60	240
Hahn et al37	2013	700	350/350	12	60	77	25	46	53	n/a	22	47	4	60	60	240
Liu et al38	2013	45	24/21	n/a	60	60	24	65	55	283	65	100	3	30	30	90
Freixa et al24	2012	79	39/40	12	60	78	20	50	57	328	48	96	4	60	60	240
Garcia et al31	2011	43	22/21	12	58	81	12	56	33	267	30	100	4	30	30	120
Xue et al39	2010	43	23/20	12	58	95	25	53	67	282	50	100	4	60	60	240
Ma et al21	2006	94	47/47	12	64	68	41	59	0	411	51	67	3	30	30	90

CK-MB, creatine-kinase myoglobin binding; DM, diabetes mellitus; HTN, hypertension; IPoC, ischaemic postconditioning; LAD, left anterior descending ; n/a, not available; PCI, percutaneous coronary intervention; TIMI, Thrombolysis in Myocardial Infarction.

Table 3.

Mean peak CK-MB

Authors (refs.)	Year	IPoC Mean peak CK-MB±SD (U/L)	Control Mean peak CK-MB±SD (U/L)
Mukherjee and Jain34	2019	290±16	414±51
Liu et al35	2016	117±31	144±41
Araszkiewicz et al30	2014	334±228	383±184
Dong et al25	2014	2160±486	2398±470
Elżbieciak et al36	2013	242±167	188±240
Hahn et al37	2013	232±172	229±204
Liu et al38	2013	93±35	93±35
Freixa et al24	2012	251±29	195±18
Garcia et al31	2011	195±33	242±40
Xue et al39	2010	248±118	352±154
Ma et al21	2006	117±76	172±93

CK-MB, creatine-kinase myoglobin binding; IPoC, ischaemic postconditioning.

Table 4.

Mean peak CK

Authors (refs)	Year	IPoC Mean peak CK±SD (U/L)	Control Mean peak CK±SD (U/L)
Traverse et al29	2019	1664±1339	1770±1601
Araszkiewicz et al30	2014	2233±1412	3043±1361
Freixa et al24	2012	3909±485	3122±379
Garcia et al31	2011	2182±1717	2444±1928
Laskey et al32	2008	1524±435	1862±461
Yang 33	2007	2229±255	2699±634
Ma et al21	2006	1237±813	1697±965
Staat et al6	2005	2831±404	4234±722

CK, creatine-kinase; IPoC, ischaemic postconditioning.

In the CK-MB analysis, 1 article enrolled participants who presented within <6 hours from symptom onset while 10 articles enrolled those who presented within <12 hours. Thrombectomy was performed in five articles. In terms of IPoC protocol, the most prevalent number of ischaemia/reperfusion cycles and duration of each were 4 and 60 s, respectively. The prevalence of each protocol included 6 articles reporting of 4 cycles×60 s, 4 of 3 cycles×30 s and 1 article of 4 cycles×30 s.

In the CK analysis, four studies enrolled participants who presented within <6 hours from symptom onset while three studies enrolled those who presented within <12 hours. Thrombectomy was performed in three studies. In terms of IPoC protocol, the most prevalent number of ischaemia/reperfusion cycles and duration of each was 4 and 60 s, respectively. The prevalence of each protocol included three articles reporting of 4 cycles×60 s, 2 of 3 cycles×30 s, 2 articles of 4 cycles×30 s and 1 article of 2 cycles×90 s.

ES and correlations

Estimated individual study ES of IPoC on peak CK-MB ranged between −3.22 and 2.32. Meta-regression revealed a statistically insignificant effect of IPoC on peak CK-MB (ES −0.41, 95% CI −1.15 to 0.34) (figure 2). Also, estimated individual study ES of IPoC on peak CK ranged between −2.38 and 1.79. Meta-regression revealed a statistically insignificant effect of IPoC on peak CK (ES −0.42, 95% CI −1.20 to 0.36) (figure 3).

Figure 2.

CK-MB forest plot. CK-MB, creatine-kinase myoglobin binding; IPoC, ischaemic postconditioning.

Figure 3.

CK forest plot. CK-MB, creatine-kinase myoglobin binding.

Parameters with 5 or less total data points per study group were not included in analysis. Such parameters included postprocedure maximum ST deviation (five articles), percentage ST resolution (five articles), area under the curve (AUC) CK and AUC CK-MB (four articles). Furthermore, correlations were not conducted for Killip class, number of stents placed, myocardial blush grade, left anterior descending artery involvement or three vessel disease or pharmacotherapy use prior to PCI, due to the limited data available.

Linear regression analysis demonstrated a statistically significant negative correlation between a history of smoking and CK-MB (p=0.038) levels in the IPoC group, although the coefficient of determination was low (R²=0.397). Besides this, no other correlation was found when comparing IPoC and control protocols (table 5).

Table 5.

Correlation of cardiac biomarkers between IPoC and control groups

Parameter	CK-MB group		CK group
	P value	R²	P value	R²
Average age	0.724	0.015	0.847	0.007
Male (%)	0.509	0.050	0.633	0.040
Avg BMI	0.266	0.135	0.319	0.165
DM (%)	0.386	0.081	0.605	0.047
HTN (%)	0.843	0.005	0.984	0.000
Hyperlipidaemia (%)	0.810	0.023	0.496	0.254
History of smoking (%)	0.038	0.397	0.318	0.165
Systolic blood pressure (mm Hg)	0.746	0.151	0.580	0.113
Diastolic blood pressure (mm Hg)	0.590	0.361	0.158	0.710
Mean heart rate (bpm)	0.325	0.315	0.636	0.061
Time symptom onset to procedure (min)	0.600	0.041	0.962	0.001
Thrombus aspiration (%)	0.328	0.757	0.330	0.755
LAD involved (%)	0.782	0.009	0.883	0.005
Multivessel disease (%)	0.621	0.0667	n/a	n/a
Pre-PCI TIMI 0 (%)	0.458	0.081	0.418	0.119
Inflations (n)	0.139	0.226	n/a	n/a
Length of inflation (s)	0.093	0.282	n/a	n/a
Total inflation time (s)	0.095	0.279	n/a	n/a
Time of reflow between Inflations	0.093	0.282	n/a	n/a
Max time of symptom onset for inclusion (h)	n/a	n/a	n/a	n/a

Values in bold indicate statistically significant findings (p<0.05)

BMI, body mass index; CK-MB, creatine-kinase myoglobin binding; DM, diabetes mellitus; HTN, hypertension; IPoC, ischaemic postconditioning; LAD, left anterior descending; n/a, not available; PCI, percutaneous coronary intervention; TIMI, Thrombolysis in Myocardial Infarction.

Discussion

We conducted an updated meta-analysis of RCTs comparing the effect of IPoC versus standard of practice on cardiac biomarkers, specifically CK and CK-MB. When comparing IPoC and control groups, our results demonstrated a correlation between smoking history and relative decrease in peak serum CK-MB levels, suggesting that IPoC may be more beneficial in smokers. On the other hand, meta-regression analysis demonstrated no significant effect of IPoC on peak CK and CK-MB. Lastly, our analysis indicates that until now, minimal studies have examined cardiac troponin as a surrogate of infarct size in STEMI patients.

Despite the fact that rapid restoration of coronary perfusion in STEMI patients reduces infarct size and the development of heart failure,⁸ it is also associated with reperfusion injury due to various mechanisms.^{2 9} Optimising PCI methods to reduce reperfusion injury is an important step towards reducing morbidity and mortality associated with STEMI. Zhao et al demonstrated preclinical data on dogs suggesting that postconditioning improves infarct size and endothelial function.⁵ Several theories have been proposed outlining the possible molecular pathways associated with IPoC, including the activation of reperfusion-injury salvage kinase pathway,^{10 11} and the reduction of the number of necrotic, apoptotic and autophagic cells.¹² Since then, numerous clinical studies have been published on the topic, overall providing mixed results regarding the effectiveness of IPoC.^{6 13 14} Endpoints have usually included infarct size either via surrogate markers or imaging, or clinical outcomes.

Several meta-analyses have been conducted in the past decade assessing the effect of IPoC on infarct size in STEMI patients using cardiac biomarkers as surrogates. The latest study was published by Gao et al approximately 7 years ago, which included 2289 participants and demonstrated significant decreases in peak CK and CK-MB in a subgroup analysis in those undergoing direct stenting alongside IPoC.¹⁵ Two other older studies demonstrated significant decreases in peak CK or CK-MB in 673 and 244 participants, respectively.^{16 17} One meta-analysis demonstrated reduced AUC of CK, but with high percentages of heterogeneity between the studies. When they pooled data of troponin AUC, there was no significant reduction in the IPoC group.¹⁸ Finally, several other meta-analyses have been conducted assessing cardiac biomarkers in IPoC, but they calculated outcomes as aggregates of cardiac biomarkers (CK, CK-MB, troponin) rather than individual markers.¹⁹ In one of these studies, subanalysis of more specific biomarkers including CK-MB and troponin was performed, but no effect of IPoC was found.²⁰

To the best of our knowledge, our meta-analyses is the most up to date systemic assessment for the effect of IPoC on infarct size based on cardiac biomarkers in STEMI patients undergoing PCI. We included numerous studies from the past several years that had yet to be included in any meta-analysis on the topic thus far. Furthermore, our study is unique in that separate analyses were conducted for each biomarker (CK, CK-MB) using relatively large sample sizes. As mentioned, only two meta-analyses have thus far demonstrated significations reductions in peak CK and CK-MB, but these studies either used similar patient populations in their analysis or were limited to relatively small sample sizes.16 17 21 22 Our analysis, on the other hand, included a significantly larger sample size (1273 patients in the CK-MB group) with a larger ratio of men to women and lower ischaemic time, as compared with previous studies, among other differences.

In this study, linear regression analysis demonstrated a statistically significant correlation between smoking history and CK-MB in the IPoC group. Despite the low coefficient of determination found, it appears then that smokers may benefit more from the possible cardioprotective effects of IPoC in reducing infarct size. The physiological mechanism of this correlation is unclear, and more research is needed to elucidate such. Moreover, no correlation was found among other common cardiovascular risk factors including diabetes or hyperlipidaemia. A possible explanation for this specific effect may be related to endothelial injury associated with smoking, which increases vascular permeability and oedema; the intermittent nature of reperfusion inherent to IPoC may limit injury in this case.²³

Interestingly, only two articles meeting our inclusion criteria examined cardiac troponin as a surrogate marker of infarct size in STEMI patients. In contrast to our meta-analysis, both studies demonstrated significantly lower levels of cardiac biomarkers (in this case troponin-I) after IPoC as compared with controls.^{24 25} This may be due to the greater specificity of troponin to cardiac ischaemia and differences in study design and the investigated population. For this reason, more studies are necessary in order to specifically analyse troponin as a surrogate of infarct size after IPoC.

Using meta-regression analysis, our study did not demonstrate a significant effect of IPoC on cardiac biomarkers; several explanations may be proposed. There was a large amount of variability in the results from each study included in our meta-analysis, as demonstrated by the large CIs in each of the forest plots. Furthermore, variability in protocol among the studies may have contributed to variability in our results. Differences in protocol included, among other things, ischaemic time, the number and duration of postconditioning balloon inflations and performance of thrombus aspiration. We know that shorter ischaemic times and performance of thrombus aspiration have been associated with less ‘no-reflow’,²⁶ a phenomenon by which there is suboptimal myocardial perfusion through an area of the coronary circulation without angiographic evidence of vessel obstruction.²⁷ Indeed, it is possible that the markers chosen are not necessarily the most reflective of infarct size. Other biomarkers that have been proposed include AUC of CK or CK-MB curves, per cent resolution of ST-elevation on ECG and troponin I or T, among others.¹⁸ In clinical practice, troponin is the preferred indicator of acute myocardial injury given its high degree of sensitivity and specificity,²⁸ and may have been a more accurate measure of infarct size than the biomarkers used in this analysis. Unfortunately, too few RCTs measuring the effect of IPoC on troponin have been conducted to be included in this meta-analysis.

There were several limitations to our study. As mentioned earlier, in the previous studies, mostly CK and CK-MB were used as measures of infarct size, despite their minimal use in contemporary clinical practice. As in any meta-analysis, variability in protocol among included studies may skew results. Indeed, the studies included in this study differed especially in terms of catheterisation and IPoC protocols, which were yet to be standardised across research groups. Recent studies assessing the impact of IPoC have proposed that robust measures of infarct size, including echocardiography and cardiac MRI, are more reflective of true infarct size and long-term outcomes.^{13 29} Despite the history of using cardiac biomarkers as correlations of infarct size, given their imperfect specificity and temporal relationship with onset of ischaemia, it indeed is possible that cardiac imaging may reveal differing results regarding the impact of IPoC.

In this updated meta-analysis assessing the effect of IPoC on cardiac biomarkers among RCTs, no significant effect was demonstrated between IPoC and overall peak CK and CK-MB levels using meta-regression analysis. Among patients undergoing IPoC, the presence of a greater percentage of smokers was associated with a more positive effect of IPoC than the control group. Our results add to the growing evidence that IPoC has minimal if any effect on improving infarct size in the general study population, at least based on cardiac biomarkers. As such, it cannot be generally recommended to all patients. More research should be performed using cardiac troponin as well as imaging such as cardiac MRI in order to obtain a more accurate quantification of infarct size. Furthermore, longitudinal studies comparing specific IPoC protocol may shed further light on the technical aspects of this intervention.

Data availability statement

All data relevant to the study are included in the article or uploaded as online supplemental information.

Ethics statements

Patient consent for publication

Not applicable.