The Propofol cardioplegia for Myocardial Protection Trial (ProMPT) is designed as a single-center, placebo-controlled, randomized trial. Participants, health care staff, and members of the research team involved in data collection or providing health care, except for the perfusionist, are blinded to a participant’s allocation.

Patients undergoing nonemergency isolated CABG or AVR with CPB at the Bristol Heart Institute (BHI) are being invited to participate.

Participants may enter the trial if they are male or female, between 18 and 80 years, having elective or urgent CABG or AVR with CPB, and are able to give full informed consent for the trial. Patients may not enter the trial if any of the following apply: (1) previous cardiac surgery, (2) combined CABG and AVR, (3) emergency or salvage operation, (4) chronic renal failure requiring dialysis, (5) current congestive heart failure, (6) left ventricular (LV) ejection fraction <30% (ie, poor LV function), (7) allergy to peanuts, eggs, egg products, soybeans, or soy products (these are intralipid ingredients), or (8) participating in another interventional trial.

Potential trial participants are identified from operating theater lists and are sent or handed an invitation letter and information leaflet that has been approved by a NHS Research Ethics Committee (REC). They have at least 24 hours to consider whether to participate. If there is insufficient time to consider taking part, a potential participant is not approached to ask for written informed consent.

After admission to the BHI, patients are seen by a member of the research team who answers any questions, confirms eligibility, and requests and witnesses written informed consent. Details of all patients approached for the trial and reasons for nonparticipation (eg, reason for being ineligible or declining consent) are being documented.

Each participant is randomly assigned in a 1:1 ratio within CABG or AVR surgery stratum to standard cardioplegia with either propofol (intervention) or intralipid (placebo) supplementation using a secure, concealed, central Internet-based randomization system (Sealed Envelope Ltd). Standard cardioplegia is warm blood for CABG and cold blood for AVR. Randomization is minimized by diabetes mellitus status, defined as requirement for oral or insulin medication to control blood sugar at the time of admission. A member of the Clinical Trials and Evaluation Unit (CTEU) Bristol who is not involved in data collection or providing health care randomizes a participant shortly before surgery and communicates the allocation to the perfusionist. If a participant is unexpectedly rescheduled, the trial number and randomized allocation is retained.

The primary outcome is myocardial injury, assessed by measuring myocardial troponin T in serum from blood samples collected preoperatively and at 1, 6, 12, 24, and 48 hours after chest closure (Figure 3). Although absolute times of sampling after reperfusion could vary between participants, blinding of the intervention prevents any systematic difference between groups.

Data are collected to characterize the following secondary outcomes:

The primary outcome is myocardial injury, assessed by measuring myocardial troponin T in serum from blood samples collected preoperatively and at 1, 6, 12, 24, and 48 hours after chest closure. The concentration of plasma propofol is measured in blood samples collected from the cardioplegia/bypass circuit immediately before aortic cross-clamping, once during cardioplegia (after blood–cardioplegia mixing), and 10 minutes after cross-clamp release.

The trial is designed to test hypotheses about the superiority of the intervention.

A sample size of 96 has been chosen to enable the trial to detect a standardized difference of 0.5 between the propofol-supplemented and placebo groups, with 80% power at a significance level of 5% (2-tailed) within each surgical stratum. The target difference of 0.5 standard deviations is of moderate magnitude [35], is consistent with our experience of many of these outcomes in previous research to evaluate other cardiac surgery interventions, and is appropriate for an early phase trial. Estimation of the power of the trial assumed that the biomarkers, including the primary outcome, are measured at baseline and 5 times after the intervention and that the correlations between outcomes measured at baseline and after intervening and between postintervention outcomes measured on multiple occasions are 0.5.

Analyses will be based on the intention-to-treat principle, and will use data from all patients randomized (ie, the analyses will use the complete dataset). Continuous outcomes will be analyzed by regression modeling, transforming data logarithmically if required and adjusting for baseline values where available, and using mixed models for repeated measures. Mixed models allow all patients with data to be included in the analysis; that is, partial missing data (assumed missing at random) is permitted. Prerandomization and subsequent values will be modeled jointly, in preference to the prerandomization value being modeled as a covariate, to avoid the need either to exclude cases with missing preoperative measures or to impute missing preoperative values. Interactions between treatment and time will be examined and, if significant at the 5% level, results will be reported separately for each postoperative time point; otherwise, an overall treatment effect will be reported. Findings will be reported as effect sizes with 95% confidence intervals. Time in ICU/HDU will be analyzed as time-to-event data using Cox regression modeling.

The trial is not powered to detect differences in clinical outcomes and their frequencies will be tabulated descriptively in accordance with guidelines for reporting randomized controlled trials (RCTs) [36].

We do not have a prior expectation that the effect of propofol supplementation will differ by operation, but we will test for this possibility by fitting a treatment by surgery type interaction. If the interaction is statistically significant, surgery-specific effects will be reported along with the results of the interaction test. If, as anticipated, the interaction is not statistically significant at the 10% level, the overall treatment effect will be reported. Similarly, we will test the interaction of propofol supplementation with diabetic status and effects specific to diabetic status will be estimated if this interaction reaches 10% statistical significance.

The primary analysis will take place when follow-up is complete for all recruited patients. No interim analysis is planned.

The West Midlands REC approved the trial protocol on the November 16, 2009. The Medicines and Healthcare products Regulatory Agency (MHRA) granted clinical trial authorization on December 11, 2009.

Adverse events (AEs) will be recorded and reported according to University Hospitals Bristol NHS Foundation Trust and MHRA guidelines. In cardiac surgery, postoperative transient complications are expected and are not infrequent. The research team is required to notify the sponsor about deaths and unexpected nonfatal SAEs. Unexpected events are those not listed in the trial protocol or on the case report forms. The sponsor will inform the research team which SAEs should be reported to the REC and/or MHRA. Data on AEs are being collected from randomization for the duration of the participant’s postoperative hospital stay and for the 3-month follow-up period.

The trial includes several features designed to minimize the risk of bias. Concealed randomization will prevent selection bias. Blinding of the research team, clinical staff responsible for caring for patients (surgeons, anesthetists, and nurses), and participants will minimize performance and detection biases. Moreover, outcome measures have been defined as far as possible on the basis of objective criteria. An independent laboratory technician, without knowledge of treatment allocation, will measure biochemical markers.

The patient information leaflet and the process of obtaining informed consent describes the uncertainty about the effects of cardioplegia supplementation with propofol. Therefore, in the event of inadvertent unblinding of a participant, he or she should not have a strong expectation that either method should lead to a more favorable result.

The trial will be analyzed on an intention-to-treat basis (ie, outcomes will be analyzed according to the treatment allocation), irrespective of future management and events, and every effort will be made to include all randomized patients. The fact that trial recruitment is from a single center and coordination is by a UK Clinical Research Collaboration (UKCRC)-registered trials unit on site should promote the completeness of follow-up. Blood samples may be missed at some time points for some patients if their sampling times occur outside the working hours of the research team. In these instances, reliance is placed on the cooperation of ICU nursing staff for sample collection. When samples are missing, we will assume that they are missing at random (see Statistical Analysis) and will allow any participants with at least 1 sample to be included in the analysis.

The ProMPT trial is a novel trial to assess the hypothesis that propofol supplementation of cardioplegia will give better myocardial protection. If this hypothesis is confirmed, we will design a larger trial to test the effect of propofol supplementation on clinical outcomes (eg, postoperative complications).

We have used broad eligibility criteria for the trial and expect the findings to apply to almost all patients undergoing standard CABG or AVR using CPB. The findings will be disseminated through usual academic channels (ie, presentation at international meetings, peer-reviewed publications, and through patient organizations and newsletters to patients, where available). Because propofol is widely available in the acute care setting and is relatively inexpensive, there should be few obstacles to adoption. We do not anticipate that the findings will be commercially exploitable.

Despite numerous reports supporting a cardioprotective effect of propofol during surgery [17-19,37], there is some evidence to the contrary. For example, studies comparing the volatile anesthetic sevoflurane with propofol for anesthesia during CABG surgery reported protection to be conferred by sevoflurane only [15,38]. This inconsistency may be explained in part by the chosen dose regimen because the cardioprotective effect of propofol has been shown to be dose dependent [19]. Clinical benefits appear to be more evident at higher doses [39] with a maintenance dose of approximately 4.2 μg/mL attenuating postoperative cellular damage and improving clinical outcome in patients undergoing CABG with CPB [40]. This dose has also been shown to confer significant cardioprotection against global ischemia in rats when used as an adjunct to warm or cold cardioplegia [30,32]. The underlying mechanism of protection is independent of any protection conferred by cardioplegia and hypothermia alone [41]. In addition, the intralipid emulsion used as a vehicle for propofol administration exhibits no protection against cardiac injury [32].

One of the major challenges of this trial was to develop a method to deliver a specific dose of propofol/intralipid to the heart during surgery. Although delivering as an adjunct to cardioplegia was the most practicable approach, differences in the temperature and method of delivery of cardioplegia for CABG and AVR procedures had to be addressed. In addition, because propofol had not been used as an adjunct to cardioplegia in the human clinical setting before this trial, it was necessary to determine the occurrence and extent of any physical or physiochemical changes to propofol or intralipid when mixed with cardioplegia solution and 0.9% sodium chloride. The Fresenius Kabi Stability Assessment Unit at Cardiff University performed pH, visual, microscopic, and particle size distribution analysis of relevant combinations of samples at time points ranging from 0 to 24 hours. Evidence of instability was found in the propofol-containing samples after 8 hours. Therefore, a period of 6 hours was deemed a reasonable, safe period within which to use the trial solutions after they had been made.

Propofol maintenance during CPB is common practice at the BHI. An infusion rate of 33 to 100 μg/kg/min (equating to 1.98-6 mg/kg/h) is used resulting in a circulating blood propofol concentration of between 1.3 and 3.6 μg/mL.

The trial has adopted a pragmatic approach, giving anesthetists the option to continue with this practice, resulting in baseline concentrations of up to 3.6 μg/mL of propofol for placebo arm participants. To ensure at least a doubling of the propofol concentration between placebo and intervention participants, anesthetists are asked to cap their usage of propofol during maintenance to a 3 μg/mL where possible. A 3 μg/mL blood propofol concentration is well under the dose required for clinical benefit [39]. Our supplementation concentration was set at 6 μg/mL to ensure that a potentially cardioprotective dose of between 6 and 9 μg/mL was achieved in patients allocated to the intervention arm.

The proposed maximum dose of 9 μg/mL does not exceed the level routinely observed in the circulation during induction of anesthesia for cardiac surgery (ie, a safe concentration to which the myocardium is routinely exposed).

Propofol clearance follows a 3-compartmental open model, with a first exponential phase half-life of approximately 1.6 to 4.0 minutes [42]. A short half-life and rapid plasma distribution of propofol results in fast onset and short duration of propofol action. Very little of the supplementary propofol remains in the circulation when the next dose of cardioplegia is administered. Hence, it is exceptionally unlikely that any AEs could be attributed to propofol supplementation postoperatively.

Designing a practicable method for delivering a potentially protective dose of propofol to the heart during cardiac surgery has been challenging, but recruitment has been steady and straightforward. If our approach confirms the potential of propofol to reduce damage to the heart during cardiac surgery, we will apply for funding to carry out a larger multicenter trial powered to detect a difference in a primary clinical outcome.