Clinical Trial of Kryptonite for Prevention of Sternal Complications in High Risk Patients Undergoing Cardiac Surgery

Purpose: To compare clinical outcomes following sternal closure with Kryptonite and conventional sternal closure, in cardiac surgery patients at high risk for sternal wound complications.

Hypothesis: Patients receiving sternal closure with Kryptonite will have a shorter post-operative length of stay, fewer sternal wound complications, improved respiratory function, better quality of life, and less pain than patients receiving a conventional sternal closure. The investigators hypothesize that patients receiving sternal closure with Kryptonite will be discharged from hospital, on average, 1 day earlier than patients receiving conventional sternal closure.

Justification: Kryptonite is a biologic polymer from the castor bean and is non-toxic. It is osteoconductive i.e. bone regrows through it. It is adhesive, and only requires 24 hours to completely cure, although sternal stability is achieved even sooner. In human cadaver studies, sternal edges adhered together with Kryptonite could not be broken after 2 hours. The proposed benefits of Kryptonite in cardiac surgery patients are the prevention of pathologic sternal displacement, the elimination of regional non-union, the enhancement of functional recovery, and the reduction of sternal wound complications.

The use of Kryptonite to close sternal wounds in cardiac surgery was first described in Calgary, Alberta. This group conducted a small, single-blind randomized clinical trial in 34 patients at low risk for sternal complications (personal communication, Fedak, P). Patients in the Kryptonite group experienced significantly less pain, required less analgesia, and demonstrated a faster return to baseline pulmonary function. The limitations of this study were the small sample size and blinding of patients only. Furthermore, hospital length of stay was not evaluated. A multi-center randomized clinical trial is planned, pending funding.

Objectives: The primary outcome of our trial is quality of life. Secondary endpoints are pain, sternal complications, and respiratory function.

Research Method: All patients undergoing non-emergent cardiac surgery through a primary sternotomy at Vancouver General Hospital will be screened for enrollment in the trial. Potentially eligible patients will be approached by a research nurse.

Forty-eight eligible patients will be randomized 2:1 intra-operatively to receive either conventional sternal closure or closure with Kryptonite. Sealed envelopes with treatment allocation will be utilized for the randomization process, and will be randomly selected in the operating room. Conventional sternal closure will be with single interrupted or figure-of-eight surgical steel wires. Conventional techniques such as vancomycin paste and bone wax can be utilized in the usual fashion to obtain hemostasis in these patients. Patients in the Kryptonite group will have surgical steel wires placed in the usual fashion, prior to the application of Kryptonite. However, bone wax and vancomycin paste will not be used in the Kryptonite group, as it impairs the adherent properties of Kryptonite.

Post-operatively, patients will not be made aware of which type of sternal closure they received. Furthermore, the post-operative care team in the Cardiac Surgery Intensive Care Unit (CSICU), as well as the surgical ward will also be blinded. The cardiac surgery ward is run by a single cardiologist who will be blinded to the patients' group allocations. Battery powered sternal saws will be available in the CSICU and on the cardiac surgery ward, in the event of emergent resternotomy.

All patients will be followed prospectively. The primary outcome will be quality of life. Secondary outcomes will be hospital length of stay, pain, sternal complications, and respiratory function. All patients will be required to answer a quality of life questionnaire pre-operatively (SF12), as well as at both 2 weeks and 5 weeks post-operatively. Pain scales will also be administered on post-operative day 2 and day 5, or on day of discharge if less than day 5. Questionnaires will be administered by the research nurse, either in hospital or by phone. Respiratory function will be measured by bedside spirometry, which will be measured pre-operatively, as well as on post-operative day 2 and 5. In-hospital sternal complications will be recorded, and following discharge the research nurse will ask about and sternal complications during follow-up telephone calls. Sternal complications of interest are sternal dehiscence requiring re-operation or sternal wound infection requiring antibiotics or re-intervention. Post-operative follow-up will otherwise be routine.

A data safety monitoring board will be assembled to perform an interim analysis when 24 of the 48 patients have been enrolled in the trial. If any safety risks are identified at this interim analysis, no further patients will be randomized in the trial.

Statistical Analysis:

Baseline characteristics will be compared between groups using Student's T-tests, for continuous variables and Chi-square tests for categorical variables.

The physical component score (PCS) and the mental component score (MCS) from the SF-12 will be compared between groups at baseline using the Student's T-test, as will the change in scores between baseline and 2 weeks and baseline and 5 weeks. The investigators will also explore the change over time using all three time points, baseline, 2 weeks, 5 weeks, using a general estimating equation (GEE).

Length of stay (LOS), change in pain score, rate of sternal complications, and respiratory function will also be evaluated. LOS will be reported as medians and interquartile ranges, given the high likelihood that this data will be significantly skewed. Differences between the groups will therefore be tested using the Mann-Whitney U test. Pain scores will be measured at 2 and 5 days. The mean and standard deviations for each of these time points will be reported and differences at 2 days will be assessed using the Student's T-test. The change in pain score between 2 and 5 days will also be compared using the Student's T-test.

The rate of sternal complications will be compared as simple proportions using the Chi-square test for differences. Respiratory function, specifically FEV1 and FVC, will be compared for the three time points using GEE, thus allowing us to compare the change in function over time.

With the exception of LOS, the investigators have assumed all continuous variables will be normally distributed. Outcomes will be tested for the normality assumption prior to carrying out the parametric tests. If a variable is not normally distributed, then either a log transformation will be applied or nonparametric tests (ie. Wilcoxon rank-sum test) will be considered to test for differences between groups.