Non-invasive remote ischemic preconditioning for patients with heart failure undergoing cardiac catheterization: a network meta-analysis of randomized controlled trials

Abstract

Objective

This study aimed to evaluate the efficacy of six non-invasive remote ischemic preconditioning (RIPC) interventions during the nursing care of patients with heart failure (HF) prior to cardiac catheterization.

Methods

A comprehensive search of nine Chinese and English online databases was conducted from the date of their inception to June 2023 to identify randomized controlled trials (RCTs) investigating RIPC in patients with HF prior to cardiac catheterization. Two independent investigators screened the articles, extracted data, and assessed their quality. The risk of bias was evaluated using the Cochrane risk-of-bias tool, and a network meta-analysis was conducted using R software.

Results

Four trials involving 511 patients with a low risk of bias were included in the analysis. Six non-invasive RIPC interventions were identified, all demonstrating effectiveness in reducing the incidence of contrast-induced acute kidney injury (CI-AKI). Among these, Intervention F (applying up to 50 mmHg above the resting systolic pressure for 5 min to the dominant leg or upper limb, repeated three times with an 18-minute interval) was deemed optimal, although the timing of the procedure was not specified. Intervention D (applying up to 200 mmHg pressure to the upper limb for 5 min, repeated four times with 5-minute intervals, within 45 min prior to cardiac catheterization, ) was considered suboptimal.

Conclusion

Although Intervention D was recommended as the preferred option, none of the four trials examined its impact on the cardiac function of patients with HF. Large-scale, multi-center RCTs are required, with outcome indicators including cardiac function and the occurrence of CI-AKI, to better understand the therapeutic effects of RIPC on HF and reduce the incidence of CI-AKI. This will provide a more robust foundation for clinical practice.

Introduction

Heart failure (HF) is a complex clinical syndrome characterized by symptoms and signs of ventricular contraction or filling disorders resulting from structural or functional cardiac abnormalities [1]. It represents an advanced manifestation of various cardiovascular diseases, and is associated with high morbidity, mortality, hospitalization rates, diminished quality of life, and substantial medical costs. The five-year mortality rate for HF ranges from 50 to 75% [2, 3].

According to the 2022 Guidelines for the Management of Heart Failure, etiological treatment has the potential to prevent or delay the occurrence of HF [4]. Coronary artery disease (CAD) is the primary contributor to HF, and prompt revascularization can mitigate the risk for HF [1]. Cardiac catheterization is a prerequisite and crucial step in the revascularization process. However, in addition to its negative inotropic effect and the potential to aggravate heart rate, the use of iodinated contrast during the procedure may induce contrast-induced acute kidney injury (CI-AKI) [5]. CI-AKI is the third leading cause of hospital-acquired AKI and is associated with up to 34% of fatalities [6].

Remote ischemic preconditioning (RIPC) involves the protective effect of transient, repeated, and reversible ischemic interventions in one organ on remote organs [7]. Several studies have demonstrated RIPC’s role in safeguarding kidney and heart tissues and reducing the incidence of various adverse outcomes during HF treatment [8]. RIPC encompasses both invasive and non-invasive procedures. The non-invasive approach, which involves the application of pressure to the limbs using a sphygmomanometer cuff, is simple to administer and carries minimal risk of complications. However, there is no consensus regarding the optimal operating cycle, interval, duration, pressure value, and other parameters [7, 9, 10].

This study utilized a network meta-analysis performed with R software to evaluate and compare the efficacy of various non-invasive RIPC interventions in the nursing care of patients with HF prior to cardiac catheterization. The aim was to identify the most effective procedure and offer a scientific foundation for clinical decision-making.

Data and methods

Inclusion and exclusion criteria for articles

The inclusion criteria for the study were as follows. (1) Study type: randomized controlled trials (RCTs). (2) Study group: participants diagnosed with HF who underwent angiography or angioplasty, including elective percutaneous coronary intervention (PCI) or cardiac catheterization, and who provided informed consent. (3) Control group: Participants who received routine preoperative care and hydration. Interventions: Participants in the RIPC group received non-invasive RIPC in addition to routine preoperative nursing. This included variations in timing, intervals, durations, and pressure values of RIPC application. (4) Outcome indicators: occurrence of CI-AKI; cardiac function parameters such as the last left ventricular ejection fraction (LVEF) before admission and New York Heart Association (NYHA) classification of cardiac function; renal function parameters including creatinine, urea nitrogen, serum creatinine clearance, and uric acid levels; other relevant indicators such as the need for reoperation or hemodialysis, unplanned readmission, and the number of readmissions due to cardiovascular and cerebrovascular adverse events within 1 year post-discharge. The exclusion criteria were: (1) articles not in Chinese or English; (2) articles lacking data or pertinent content; (3) meeting abstracts; (4) studies featuring invasive RIPC interventions.

Search strategies

The computerized search strategy included the databases Pubmed, Cochrane, Embase, EBCSO, Scopus, Ovid, CNKI, Wanfang, and Sinomed from their inception to June 2023, utilizing a combination of subject terms and free terms. English terms included variations of HF, cardiac failure, heart decompensation, CHF, myocardial failure, cardiac catheterization, PCI, PTCA, TRI, coronary intervention, transradial access, coronary atherectomy, radial artery, rotational atherectomy, coronary stent, coronary angioplasty, intracoronary stent, balloon angioplasty, coronary artery angiography, TRA, transradial artery, remote ischemic preconditioning, and RIPC. The Ovid database was utilized for this study. Additionally, published systematic reviews and meta-analyses were scrutinized, and their references were searched to ensure comprehensive coverage of relevant literature.

Article screening and data extraction

The searched articles were organized and managed using EndNote X9 software. Two investigators independently screened the articles and extracted data according to predefined inclusion and exclusion criteria. The results were then cross-checked for consistency. In the event of disputes, a third investigator was consulted to resolve the issue. The extracted data encompassed basic information, research methodologies, potential bias risks, subject characteristics, sample sizes, interventions, outcome indicators, adverse reactions, and research findings.

Article quality evaluation

Two investigators independently assessed the quality of the included trials using the Cochrane risk-of-bias evaluation tool. This assessment involved examining various aspects such as the random allocation method, allocation concealment, blinding of study participants, implementers of the therapeutic scheme, and outcome assessors, as well as the integrity of outcome data, selective reporting, and other potential sources of bias. In the event of disagreements or disputes, a third investigator was consulted to resolve the issue. The results of the risk-of-bias assessment were entered into the RevMan statistical software for visualization and analysis. Bias risk in each domain was categorized as low, high, or unclear.

Statistical methods

The quality of articles was assessed using RevMan 5.3 statistical software. The primary outcome indicator, occurrence of CI-AKI, was analyzed using a forest plot. Due to the inclusion of only 4 articles in the study (less than 10), funnel plot analysis was not performed. In the forest plot analysis, if the confidence interval (CI) of each trial intersects the null line, it suggests no statistical difference. Additionally, the p-value and I² statistic were utilized to assess heterogeneity, where p > 0.1 indicates no heterogeneity, and p ≤ 0.1 suggests heterogeneity. Furthermore, a network meta-analysis was conducted using R4.3 software, with binary variables as the primary outcome indicator. Within the Bayesian framework, consistent tests, heterogeneity tests, and convergence evaluations were performed through both direct and indirect comparisons. The overall quality of the trials was assessed using trajectory diagrams, density plots, and Brooks-Gelman-Rubin diagnostic plots. The optimal non-invasive RIPC intervention was determined through order ranking and comprehensive ranking methods.

Results

Article screening process

The preliminary search yielded a total of 431 articles, which was reduced to 245 after deduplication using Endnote X9 software. Subsequently, 224 articles were excluded based on their titles and abstracts. The remaining 21 articles, all in English, were deemed eligible for inclusion after full-text review. The article screening process is illustrated in Fig. 1.

Fig. 1

Article screening process

Basic information and quality evaluation results of articles

Basic characteristics of included articles

The included trials comprised a total of 511 patients from 4 countries (Turkey, Iran, Netherlands, and Australia), and investigated 6 different RIPC interventions, as detailed in Table 1. The primary outcome indicator across all trials was the occurrence of CI-AKI. Table 2 presents the basic characteristics of the included articles.

Table 1 Six RIPC interventions

Table 2 Basic characteristics of included articles (n = 4)

Quality evaluation of results of included articles

The quality evaluation results of the included articles are summarized as follows: The random allocation method was mentioned in 3 articles; [11,12,13] the allocation concealment method was mentioned in 1 article; [12] blinding was mentioned in 4 articles; [12,13,14,15] the integrity of outcome data was discussed in 3 articles; [11, 12, 14] none of the articles reported selective outcome reporting; [11,12,13,14] and other potential sources of bias were noted in 2 articles [12, 13]. The overall quality evaluation of the articles is depicted in Fig. 2.

Fig. 2

Summary of risk-of-bias

Results of R language-based network meta-analysis

General condition of included trials: analysis of convergence, consistency, and heterogeneity

• (1) Analysis of convergence: The trajectory diagram (Fig. 3) shows that after exceeding 5000 iterations, the Markov Chain Monte Carlo (MCMC) chain (blue segment) exhibited steady fluctuations with good overlap. In the density plot, with 20,000 iterations, the bandwidth approached 0 and stabilized, indicating effective model convergence. The Brooks-Gelman-Rubin diagnostic diagram (Fig. 4) showed that the potential scale reduction factor (PSRF) for each intervention approached 1 and remained stable after iterative calculations. This suggests that the model converged effectively and can accurately predict outcomes accurately.

• (2) Analysis of consistency test: For direct comparisons of Interventions A, E and F, p > 0.05, indicating no statistical difference or good consistency. Further details are provided in Supplementary Fig. 1.

• (3) Analysis of heterogeneity test: Heterogeneity was observed in the direct comparison of Interventions B and E, with an I² value of 81.9%. However, network comparison among Interventions C, D, and E met the assumption of homogeneity, with an I² value of 0%. Additional information is provided in Supplementary Fig. 2.

Fig. 3

(A) Trajectory diagram and density map-1. (B) Trajectory diagram and density map-2

Fig. 4

Brooks-Gelman-Rubin diagnostic diagram

Comparison of interventions

• (1) Analysis of effect size of each trial: As depicted in Fig. 5, the heterogeneity test yielded p = 0.17 > 0.1, indicating no significant heterogeneity among trials. However, the Z test yielded p = 0.03 < 0.05, indicating statistically significant differences between trials.

• (2) Analysis of outcome indicator probability for each intervention and optimal intervention: Table 3 presents the probability of CI-AKI occurrence as the first and sixth indicators for each intervention. For Intervention A, the probabilities were 0.1970 and 0.0000, respectively. For Intervention B, the probabilities were 0.2780 and 0.0010. The probability of CI-AKI occurrence was highest for Intervention A (0.7490325) and lowest for Intervention F (0.0068350). For Interventions B, C, D and E, the probabilities were 0.6941175, 0.7146625, 0.3036875 and 0.5316650, respectively. Therefore, Intervention F was identified as the most effective in reducing the occurrence of CI-AKI, followed by Intervention D. Intervention A was the least favorable option.

Fig. 5

Forest plot of the occurrence of CI-AKI

Table 3 Order ranking table

The comprehensive analysis using RevMan and R software showed that the consistency model employed in this study provided a good fit and achieved convergence, meeting the assumptions of consistency. The predictive results from this model were deemed highly reliable. The ranking table indicated that Intervention A was associated with the highest risk of CI-AKI, while Intervention F was associated with the lowest risk. Consequently, Intervention F was identified as the optimal intervention for minimizing the risk of CI-AKI, followed by Intervention D.

Discussion

The results of this study indicated that all non-invasive RIPC methods effectively reduced the incidence of CI-AKI. According to the comprehensive ranking table, the non-invasive RIPC intervention described by Sami et al. (2022) was deemed the most effective [13]. This study considered both the baseline renal function of patients and the type of contrast agent used. In contrast, the intervention described by Seyed et al. (2021) was considered suboptimal, despite having complete data and being clinically feasible [12]. The kidneys are particularly sensitive to ischemia-reperfusion injury and may benefit from ischemic preconditioning [15]. This may be attributed to the release of injury-related molecules that interact with toll-like receptors (TLR) on proximal tubular cells. This interaction mediates the activation of the immune response, thereby protecting the kidneys against subsequent inflammatory or ischemic stress responses [8]. Recent research has also indicated that RIPC can effectively prevent acute mountain sickness, with its pathophysiological process potentially linked to hypoxia [16]. Coincidentally, the occurrence of CI-AKI is also associated with renal ischemia and hypoxia. Through one or more transient episodes of ischemia-reperfusion applied to remote tissues or organs, RIPC can render the myocardium less sensitive to subsequent prolonged ischemic events. Consequently, both RIPC methods are promising for effectively safeguarding renal function in patients following cardiac catheterization. However, it is noteworthy that Intervention D is considered the top recommendation, primarily due to its clinical viability.

Data obtained from 21 full-text articles, including 4 RCTs, revealed that indicators related to HF, such as cardiac function classification, brain natriuretic peptide (BNP) levels, LVEF, and the 6-minute walk test (6MWT), were not assessed in any of the trials. Transient muscle ischemia may provide protection to heart tissue [8]. Although the exact mechanism remains unclear, research indicates that transient limb ischemia releases a low molecular weight (< 15 kDa) factor, which protects the myocardium against ischemia/reperfusion injury. This factor requires activation by opioid receptors and induces changes in mitochondrial function via the ATP sensitive potassium channel [17, 18]. While RIPC may reduce the volume load in patients with HF, its impact on their cardiac function remains uncertain. As living standards improve and populations age, an increasing number of patients with HF are at risk of cardiovascular adverse events, such as symptom aggravation, prolonged hospital stays, higher hospitalization costs, and even sudden death. This situation underscores the need for interventions like cardiac catheterization. Managing the perioperative period of cardiac catheterization in patients with HF has become a critical issue.

HF is characterized by the heart’s inability to supply the surrounding tissues with sufficient blood and oxygen to meet their metabolic needs, often leading to tissue ischemia and anoxia [19]. Although the precise mechanism of RIPC remains elusive, numerous studies have suggested that occluding blood flow to remote organs through pressurization can activate protective mechanisms in target organs (e.g., heart, kidneys), potentially reducing their susceptibility to ischemia and anoxia. Therefore, it is recommended that future RIPC studies consider cardiac function as an evaluation indicator to further elucidate the impact of RIPC on patients with HF.

There are certain limitations in this study that should be acknowledged. First, the included trials were limited in number, with only four RCTs being included. This small sample size may affect the generalizability of the findings. All these RCTs were conducted at single centers, which had a high risk of potential biases. All included articles were in English. Second, there are deficiencies in the outcome indicators assessed in the included trials. None of the trials evaluated the effect of remote RIPC on cardiac function, which is a crucial aspect to consider in patients with HF. Therefore, the therapeutic effect of RIPC on patients with HF could not be conclusively determined. To address these limitations, it is recommended that future research endeavors focus on conducting large-sample multi-center RCTs.

Conclusion

This study included four RCTs, examining the efficacy of six different non-invasive RIPC methods in reducing the occurrence of CI-AKI in patients with HF undergoing cardiac catheterization. The network meta-analysis revealed that Intervention F, which involved pressurizing the dominant leg or upper limb to 50 mmHg above the resting systolic pressure of the patient for 5 min, repeated 3 times, with an 18-minute interval, was optimal in reducing CI-AKI occurrence. However, the timing of this intervention was not specified.

On the other hand, Intervention D, which involved applying 200 mmHg pressure to the upper limb for 5 min within 45 min prior to cardiac catheterization, repeated 4 times with 5-minute intervals, was considered suboptimal. Despite this, Intervention D is noted for its clinical viability. However, none of the trials investigated its impact on cardiac function in patients with HF undergoing cardiac catheterization. Additional research is needed to clarify the potential benefits of RIPC in improving cardiac function outcomes in this patient population, and more rigorous study designs are necessary to confirm the clinical role of RIPC.