Abstract
A pilot randomized controlled trial that evaluated the effect of remote ischemic preconditioning (RIPC) on clinical outcomes following major vascular surgery was performed. Eligible patients were those scheduled to undergo open abdominal aortic aneurysm repair, endovascular aortic aneurysm repair, carotid endarterectomy, and lower limb revascularization procedures. Patients were randomized to RIPC or to control groups. The primary outcome was a composite clinical end point comprising any of cardiovascular death, myocardial infarction, new-onset arrhythmia, cardiac arrest, congestive cardiac failure, cerebrovascular accident, renal failure requiring renal replacement therapy, mesenteric ischemia, and urgent cardiac revascularization. Secondary outcomes were components of the primary outcome and myocardial injury as assessed by serum troponin values. The primary outcome occurred in 19 (19.2%) of 99 controls and 14 (14.1%) of 99 RIPC group patients (P = .446). There were no significant differences in secondary outcomes. Our trial generated data that will guide future trials. Further trials are urgently needed.
Keywords
Introduction
Patients who require surgery for vascular disease constitute a high-risk group. Perioperative complications such as myocardial infarction (MI), cerebrovascular accident (CVA), renal failure, and death are common. 1 -3 These complications can be caused by multiple mechanisms such as plaque rupture and hypotension. 4 Therefore, it is desirable to have an intervention that can protect against injury via multiple mechanisms. Remote ischemic preconditioning (RIPC) may be suitable in this regard.
Ischemic preconditioning (IPC) is a phenomenon whereby brief periods of nonlethal ischemia in a tissue can render the tissue resistant to subsequent sustained ischemic episodes, 5 and proof-of-concept trials have confirmed its efficacy. 6 However, IPC is not clinically attractive, as it involves interfering directly with the blood supply of a vital organ such as the heart. RIPC refers to the initiation of an organ-protective phenotype by applying a brief ischemia–reperfusion stimulus to a distant tissue. Any tissue can provide the stimulus and any organ can be protected, although protection is relative and not absolute. 5 Clinically, the most attractive and easily achieved stimulus is skeletal muscle ischemia induced by blood pressure cuff inflation. Most of the trials to date have involved cardiac surgery patients although some vascular surgery trials have also been performed. Meta-analyses consistently found benefits in biochemical outcomes although firm data regarding clinical outcomes are lacking. 7,8
We performed a pilot multicenter randomized controlled trial to assess whether RIPC could improve clinical outcomes in patients who were undergoing major vascular surgery. We also examined the effect of RIPC on myocardial injury in these patients.
Methods
This was a prospective, multicenter, parallel group (1:1 allocation ratio) randomized controlled trial. It took place from January 1, 2012, to March 31, 2014, in 3 Irish tertiary vascular centers—University Hospital Limerick, Cork University Hospital, and University Hospital Waterford. Ethical approval was granted by the Institutional Review Boards of the 3 participating hospitals, and the trial was registered (NCT01691911). The study was compliant with the Declaration of Helsinki and Good Clinical Practice, and the participants gave written informed consent.
Eligible patients were those who were undergoing elective carotid endarterectomy, open abdominal aortic aneurysm (AAA) repair, endovascular aortic aneurysm repair (EVAR), or surgical lower limb revascularization (suprainguinal or infra infrainguinal). Patients were excluded for the following reasons: pregnancy, significant upper limb peripheral arterial disease, previous history of upper limb deep venous thrombosis (DVT), therapy with sulfonylurea or nicorandil medication, preoperative estimated glomerular filtration rate (eGFR) <30 mL/min/1.73m2 using the Modification of Diet in Renal Disease equation, previous history of myocarditis, pericarditis, amyloidosis, and the presence of severe hepatic disease defined as an international normalized ratio (INR) >2 in the absence of anticoagulation. Additionally, we excluded patients who were undergoing fenestrated or branched EVAR procedures. Patients were recruited consecutively in the participating hospitals before their procedure, either in outpatient clinics or in hospital wards.
RIPC comprised 4 cycles of 5 minutes of forearm ischemia with 5 minutes of reperfusion, requiring 35 minutes for an application. This was achieved by inflation and deflation of a blood pressure cuff placed around an upper limb. The cuff was inflated to 200 mm Hg or to at least 15 mm Hg higher than systolic pressure for those with systolic blood pressures of >185 mm Hg. The time of RIPC initiation in relation to onset of anesthesia and surgery was variable. We aimed to initiate RIPC prior to anesthesia induction and finish it before or after surgery began. In this way, preconditioned patients were within the initial 2-hour window of organ protection during their procedures. 9 In most previous trials, RIPC was applied uniformly at set times in relation to anesthesia and surgery—for logistical reasons we had to adopt a flexible approach to timing. Overall, the window between completion of RIPC and start of surgery was never more than 30 minutes for any patient, and for most patients it was 15 minutes or less—thus perioperative organ protection was potentially achieved for all patients with RIPC. Controls received no intervention.
The choice of anesthetic was at the discretion of the consultant anesthetist who was responsible for the case. Both regional and general anesthesia and combinations of the 2 were utilized in the study. There were no restrictions regarding the use of volatile agents or opiates. The steps involved in surgical procedures were not specified and were left to the discretion of responsible consultant surgeons. Additional procedures such as peripheral angioplasty could be carried out at the surgeon’s discretion during EVAR or lower limb revascularization procedures.
The primary outcome was a composite clinical end point comprising any of cardiovascular death, MI, new-onset arrhythmia, cardiac arrest, congestive cardiac failure, CVA, renal failure requiring renal replacement therapy, mesenteric ischemia, and urgent cardiac revascularization within 30 days of operation. These are defined in Table 1. Prespecified secondary outcomes were duration of postoperative hospital and intensive care unit (ICU) stay, unplanned critical care admissions, and postprocedure renal injury. The individual components of the composite primary outcome were further secondary end points. Postoperative complications (wound infections, respiratory tract infections, deep venous thromboses, pulmonary emboli, limb ischemia, and limb amputations) were included as post hoc secondary outcomes. Perioperative myocardial injury, assessed by high-sensitivity cardiac troponin t (cTnT-hs) values, was a final secondary outcome. The cTnT-hs was measured via serum sampling preoperatively and on the first, second, and third postoperative days. It was measured using an electrochemiluminescence immunoassay on Elecsys and cobas e-immunoassay analyzers (Roche Diagnostics GmBH, Mannheim, Germany). The reference range was 0 to 5 ng/L, and a value of ≥5ng/L was considered abnormal.
Definitions of Components of the Composite Primary Outcome.
Abbreviations: cTnT-hs, high-sensitivity cardiac troponin T; CT, computed tomography; ECG, electrocardiography.
No sample size calculation was possible for this pilot trial because no reliable data existed on event rates for our composite primary outcome. Furthermore, no previous trial evaluated RIPC in this population.
A computer-generated random sequence was used. Randomization was stratified by procedure type and by center, and random block sizes between 4 and 8 were used. Allocation concealment was achieved by the use of sequential sealed opaque envelopes. A third party who was not involved with other aspects of the trial generated the random sequence and the sealed envelopes. Envelopes were opened sequentially prior to operations when patients were within operating theater complexes. Members of the surgical teams enrolled patients, assigned interventions, and applied RIPC where necessary. There was no blinding.
Categorical variables were compared using Fisher exact test. Continuous variables were compared using the 2-sample t test or the Mann Whitney U test as appropriate. Results were presented as means with standard deviations or as medians with interquartile ranges as indicated. Minitab version 16 (State College, Pennsylvania) was used for these analyses. Perioperative myocardial injury was compared between the groups by comparing area under the curve (AUC) for the first 3 postoperative days. The troponin measurements were log-transformed before analysis. To account for possible bias due to different missing data patterns in the preconditioned and nonpreconditioned individuals, the R-package norm 10 was used to create 100 different imputations of the missing log-troponin values, assuming a multivariate normal model. For each imputed data set, the AUC of log-troponin against time post operation (0, 1, 2 or 3 days) was computed for each patient. The difference in the average values of the log-troponin AUC for the RIPC group and control groups and the standard error of these differences were then computed for all 100 imputed data sets. These were then combined (over all 100 imputed data sets) using the technique described in Gelman et al 11 to calculate a single t test statistic. As sensitivity analyses, we repeated this procedure but instead based the t test statistic on comparing (1) the difference in log-troponin on day 3 and preoperation for each individual and (2) the least-squares regression slope of each individual’s log-troponin measurements over time, between the 2 arms. For all analyses, significance was set at 5%.
Results
Figure 1 summarizes patient flow through the trial. The trial ran between January 2012 and March 2014. It terminated when the 200th trial number was allocated.
Trial flow diagram.
Of the 231 patients assessed for eligibility 200 underwent randomization. Two patients were excluded following randomization: one underwent a branched EVAR procedure and thus was ineligible for inclusion and another patient had a baseline eGFR <30 mL/min/1.73m2 and thus was ineligible for inclusion. Ninety-nine participants were allocated to each treatment arm. Of the 99 patients randomized to the RIPC group, 94 received the intervention as described earlier. Five did not receive the allocated intervention: In one instance, the manual blood pressure cuff failed, 1 patient had previous axillary surgery, and on 3 occasions the anesthetic team required constant access to both upper limbs. There were no losses to follow-up at 30 days. Data from 99 patients in each treatment arm were finally analyzed using an intention-to-treat analysis.
Table 2 provides details on demographics, comorbidities, medications, baseline laboratory results, and operative details for each group. The groups were well matched at baseline. Mean age was 69 years in both groups, and 73 of 99 were males in the control group versus 78 of 99 in the RIPC group. Similar proportions of patients in each treatment arm underwent open AAA repair, EVAR, carotid endarterectomy, and lower limb revascularization procedures. Medication use and comorbidities were similar.
Baseline Demographics, Clinical Characteristics, and Operative Data.a
Abbreviations: cTnT-hs, high-sensitivity cardiac troponin T; IQR, interquartile range; SD, standard deviation; ACE, angiotensin-converting enzyme inhibitor.
aNumbers in parentheses represent percentages for categorical variables and standard deviations or interquartile ranges for continuous variables.
Table 3 provides details on the primary outcome and some secondary outcomes. Table 4 provides details on the additional secondary outcome of troponin leakage within the first 72 hours. The primary composite clinical end point occurred in 19 of 99 controls and in 14 of 99 patients with RIPC, representing a nonsignificant difference (P = .446). There were no significant differences in occurrences of individual components of the composite end point or in any of the other secondary outcomes. There was no significant difference between groups regarding mean AUC for troponin over the first 72 hours postoperatively (P = .4). Regarding our sensitivity analyses, when the t test statistic was based upon the difference in log troponin between the 72-hour point and preoperative levels, a P value of .44 was attained, and when it was based on the least squares regression slope of each individuals log-troponin measurements, the P value was .54.
Primary and Secondary Outcomes.a
Abbreviations: AAA, abdominal aortic aneurysm; EVAR, endovascular aneurysm repair; ICU, intensive care unit; N/A, not available; RIPC, Remote ischemic preconditioning.
aNumbers in parentheses represent percentages for categorical variables interquartile ranges for continuous variables.
Serum Troponin Values within the First 72 Hours Postoperatively.
Abbreviations: cTnT-hs, high-sensitivity cardiac troponin T; IQR, interquartile range; RIPC, remote ischemic preconditioning.
Discussion
In this pilot study of 200 patients undergoing major vascular surgery, we examined the effect of RIPC on clinical outcomes. We found no significant effect of RIPC on our predefined primary composite clinical outcome: It occurred in 19 (19.2%) of 99 control patients and 14 (14.1%) of 99 patients with RIPC (P = .446). Our secondary outcomes included the individual components of the primary outcome—unsurprisingly, we found no significant effect of RIPC on any of these. We found no difference in hospital or ICU length of stay. An additional secondary outcome was perioperative myocardial injury, and again we found no significant effect with RIPC when AUC for troponin leakage within 72 hours was compared between the groups (P = .4). Although this was a pilot study that has yielded a neutral result, we think that it makes a valuable contribution to our understanding of the clinical effects of RIPC.
Six other clinical trials have evaluated RIPC in major vascular surgery: 4 have examined RIPC in the setting of open AAA repair, 12 -15 and trials have also examined RIPC in EVAR 16 and carotid endarterectomy. 17 Ali et al randomized 82 open AAA patients to a lower limb RIPC stimulus achieved via iliac artery cross-clamping or to control groups. They found that RIPC reduced myocardial and renal injury at a biochemical level and also that myocardial infarction rates were reduced with RIPC. 12 Li et al randomized 62 open AAA patients and used an upper limb cuff-induced stimulus. 14 They found biochemical evidence for a protective effect of RIPC on pulmonary and intestinal injury, but notably they did not measure cardiac enzyme release, and they found no difference in rates of myocardial infarction. The studies by Ali and Li both found significant primary outcome results favoring RIPC—this highlights the organ protective potential of RIPC in AAA repair. However, the remaining vascular trials yielded neutral results as did ours. A third open AAA trial 13 involved 40 patients and a lower limb RIPC stimulus achieved by cross-clamping iliac arteries. The primary outcome was renal injury, and no difference was found. The final open AAA trial by Murphy et al 15 used an upper limb RIPC stimulus and had a sample size of 62 patients. Postoperative creatinine was the primary outcome, and no difference was found between the groups. The trials on EVAR 16 and carotid endarterectomy 17 involved 40 and 70 patients, respectively, and utilized a lower limb cuff-induced RIPC stimulus. They used surrogate outcome measures of neurological, cardiac, and renal injury and found no difference between the groups.
Regarding cardiovascular interventions in general, most of the trials on RIPC have focused on surrogate outcomes, and only a few studies 18,19 have had clinical primary end points. Systematic reviews involving cardiac surgery, 20 CABG surgery, 21 percutaneous coronary intervention (PCI), 22 and cardiovascular surgery combined with PCI 8 have found results favoring RIPC. A consistent conclusion is that RIPC can reduce myocardial injury when determined by cardiac enzyme release, 8,9,20,23 and notably reviews involving PCI only 22 and cardiac and vascular surgery combined with PCI 8 found reduced myocardial infarction rates. A recent large-scale meta-analysis that examined major clinical outcomes following cardiovascular surgery found no evidence for a significant benefit with RIPC 7 but notably the analysis was underpowered to evaluate clinical outcomes and was limited by clinical heterogeneity. Nonetheless, the conclusions provided grounds for optimism regarding the potential for RIPC to reduce perioperative MI—pooled data from 17 trials (1777 patients) on RIPC in cardiovascular surgery 12 -14,16,17,19,24 -34 generated a pooled risk ratio of 0.69 (95% CI 0.34-1.40) favoring RIPC. When the MIs in the current trial and the 2014 AAA trial by Murphy et al 15 are additionally included, the pooled risk ratio becomes 0.68 (95% CI 0.41-1.14; Figure 2). This analysis used RevMan version 5.3 (Copenhagen, Denmark).
Forest plot for perioperative myocardial infarction including all trials on remote ischemic preconditioning in cardiovascular surgery.
Despite the lack of adequately powered trials with hard clinical outcomes as primary end points, convincing “proof-of-concept” evidence underpins the biological plausibility of RIPC-induced cardioprotection. In contrast to cardiac surgery, there is a paucity of data relating to vascular surgery, and conclusions from the vascular trials are less consistent. Five of the 7 vascular trials to date, including the current trial, had neutral primary outcome results. Reasons for this are uncertain although aspects that are probably implicated are small sample sizes and heterogeneity in terms of populations, outcome measures, and procedures. Most of the cardiac surgery trials evaluated cardiac enzyme levels. In contrast, only 4 vascular trials evaluated such outcomes (Ali et al, 12 Murphy et al, 15 Walsh et al, 17 and the current trial). The current report provides important data that may guide the design of future trials involving RIPC and any of the included vascular procedures. Although the major challenge regarding RIPC in the wider cardiovascular context is to generate evidence on clinical effects, we think the most feasible next step in vascular surgery is to harden the evidence regarding proof of concept. As such, an adequately powered vascular surgery trial is needed.
We now propose a larger feasibility trial of RIPC in vascular surgery (NCT02097186). 35 We have set 3 main objectives: to further evaluate the ability of arm-induced RIPC to confer protection in major vascular surgery using cardiac enzyme release as a surrogate marker of efficacy, to assess the effect size of RIPC using this surrogate marker, and to confirm that RIPC can offer cardioprotection to vascular surgery patients. The primary analysis will be based on troponin-positive events, and we aim to recruit 400 patients. Details on other aspects of methodology including sample size estimation are available in the protocol. 35 Recruitment has commenced, and we hope that it will yield convincing evidence to confirm the role of RIPC in vascular surgery.
The principle strength of the current trial is its pragmatic multicenter design, which involves an emphasis on clinical as well as surrogate outcomes. Regarding limitations, the chief concern is the omission of a sham intervention. Although blinding surgeons, patients, and outcome assessors would have increased the validity of the trial, its omission has allowed for a greater sample size than would otherwise have been possible. We highlight that it was possible to achieve blinding in the current trial in relation to perioperative MI—a blinded cardiologist assessed this outcome, making this analysis quite robust. A further drawback relates to the length of follow up—we limited this to 30 days, and therefore conclusions beyond this point cannot be made. Our study is underpowered—given the observed event rates, a sample size of 900 in each arm would have been required in order to demonstrate significance with 80% power. Finally, we wish to highlight that we elected to include patients undergoing a diverse range of procedures rather than focusing on one type of procedure. This allowed us to maximize recruitment, although we accept that this variability reduced the chance of achieving a significant result.
Conclusion
In our pilot trial, we found no evidence to support the hypothesis that RIPC offers perioperative protective to patients undergoing major vascular surgery. Most of the trials to date on RIPC in vascular surgery have yielded neutral results. Despite this, RIPC represents a theoretically attractive risk reduction strategy as convincing mechanistic data confirm its potential. Although the long-term goal is to evaluate patient important outcomes, we think that future trials on RIPC in major vascular surgery should aim to clarify “proof of concept.”
Footnotes
Authors’ Note
All members of the Preconditioning SAVES Group contributed in each of the following areas: substantial contribution to conception and design or acquisition of data or analysis and interpretation of data; drafting the article or revising it critically for important intellectual content; and final approval of the version to be published.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.