Effects of Mediterranean Diet Intervention in Diabetic Heart Disease

Study Design: The study is a prospective, un-blinded, randomized interventional clinical trial comparing ModMeD to standard cardiac diet in 48 (24 per arm) inpatients with type II DM scheduled for CABG at MMC and who are scheduled to stay in the hospital for four (4) or more days prior to surgery. While it is not feasible to blind the subjects or clinicians, laboratory investigators processing specimens and analyzing biological responses to the intervention will be blinded. Specifically, tissue samples will be de-identified and coded before transferring to the laboratory. Primary Objective: To determine if a Mediterranean diet causes rapid changes in lipid content (as determined by mass spectrometry) and in cardiac cell activity (as determined by ERBB2 expression level) among patients with TIIDM undergoing CABG surgery at MMC. Study Arms: 1. Standard Cardiac DM Diet (SCaDMD) during the inpatient stay (before and after CABG surgery) 2. Mediterranean Diet (MedD) during the inpatient stay (before and after CABG surgery) Subject recruitment: 48 inpatient adults will be enrolled (age ≥ 18 years diagnosed with type II DM with hemoglobin A1c level greater than 7.5% and coronary artery disease with planned surgical revascularization (CABG)) at MMC. Other inclusion criteria include planned pre-surgical hospital stay of at least 4 days, and willingness to have the Mediterranean Diet restricted menu for the duration of their inpatient stay. Informed Consent Process: Via an IRB-approved partial HIPAA waiver, the clinical team will screen for eligible patients who are scheduled for CABG surgery. Eligible subjects will be approached by their clinical team to gauge interest in study participation. Patients who indicate interest will be approached by a study team member to explain the study fully and obtain informed consent. Patients who sign the informed consent form for this research will have agreed to: 1) have baseline body measurements 2) be randomized to consume a Mediterranean diet or the standard cardiac DM diet during their hospital stay, 3) undergo a biopsy procedure at the time of their CABG and to donate ventricular myocardium tissue approximately the size of a grain of rice, 4) donate tissue usually discarded during surgery (cardiac right atrial tissue, skeletal muscle and fat), 5) donate a total of 30 ml of blood collected at 3 time points: prior to, during, and 4 days post CABG surgery, 6) to fill out questionnaires about their food consumption and 7) allow the review of their medical record for their cardiovascular health information. The informed consent will include itemized patient responsibilities such as completing dietary questionnaire accurately, and adhering to diet restrictions, including the requirement that patient families do not bring outside food items to patients. Interested patients and their families will be given ample opportunity to ask questions. Study Procedures Dietary Questionnaire: Upon obtaining informed consent all subjects will complete a dietary questionnaire (DHQ II) provided to them by a registered dietitian staffing Cardiology Inpatient floors at MMC. The DHQ II is an online food frequency questionnaire advanced by the NIH (https://epi.grants.cancer.gov/dhq2/about). The original DHQI was developed by the Risk Factor Assessment Branch and validated to show that it provides reasonable nutrient estimates 26-29. The DHQII is an updated version with 10 more food items (134 instead of 124) and 8 dietary supplements. The DHQII has 4 versions that differ by time frame and portion size questions. We will utilize the one-month option (Appendix A) that includes portion sizes. This is a self-administered questionnaire to be completed by the patient that takes about an hour. Subjects will be offered the online version via portable tablet or a paper version. DHQII data will be analyzed using the Diet\*Calc software program freely available from the NCI (https://epi.grants.cancer.gov/dhq2/dietcalc/). This program generates nutrient and food group intake estimates for the standard versions of the DHQII used in this study. Data from the dietary questionnaire will allow us to conduct a number of exploratory analyses that have not been attempted before. For example, we will look for associations between the abundance of certain lipid species with patient unsaturated and saturated fat intake. Baseline anthropometrics: Height (m) and weight (kg) will be measured using standard clinical equipment and will be verified by a study team member. This information will be used to calculate body mass index (BMI), at the time of enrollment. A study team member will also measure waist and hip circumference. Randomization: After subjects have been screened for eligibility and have provided informed consent, they will be randomized in a 1:1 ratio to one of two dietary interventions. Randomization will be implemented by the study statistician and will be involve stratification by gender in alternating blocks of 4 and 6. nQuery randomizer will be used to create the randomization scheme; the randomization list will be transferred to sealed, opaque, sequentially numbered envelopes, one set for each stratum. At the time of randomization, the next numbered envelope will be opened to determine study group assignment. Study dietitians will ensure diet assignments are communicated and followed. Standard Cardiac DM Diet (SCaDMD): As per current nutritional guidelines, all subjects will receive standard instruction and the dietary menu (Appendix B) from an MMC dietitians. based upon their diagnoses of DM and coronary disease. Patients will be prescribed a cardiac diet, which is a lower fat diet, and if desired by the practitioner, additional sodium or carbohydrate restrictions may be included on a case by case basis. Study dietitians will supervise will the recording of record food consumption daily from enrollment until day 4 post CABG surgery by 1) direct questioning, 2) review of food orders placed by the patient, and 3) examination of food trays upon completion of meals. Mediterranean Diet (MedD): The MMC registered dietitians have created a sub-menu (Appendix C) from the existing MMC inpatient menu that is compliant with the Mediterranean diet as described by Casas et al. (2016)20 and was used in the PREDIMED trial. The only difference between the Mediterranean diet used in this study that used in the PREDIMED trial is the exclusion of wine and sofrito, a sauce used as a base in Spanish, Italian, Portuguese and Latin American cooking. This was due to the limits of MMC cafeteria. Upon randomization, study dietitians will meet with the subject to review the ModMeD and guide their food selection. Study dietitians will also meet with floor nurses to explain the difference between the standard diet and the Mediterranean diet and to explain the study procedures to record food consumption. Study dietitians will supervise the recording of food consumption daily from enrollment until day 4 post CABG surgery by 1) direct questioning, 2) review of food orders placed by the patient, and 3) examination of food trays upon completion of meals. Biologic Sample collection: Ventricular cardiac tissue, atrial cardiac tissue, and skeletal intercostal muscle will be collected from study participants during their CABG surgery. Blood samples will be collected at three time points: 1) at enrollment (on the day of the questionnaire and anthropomorphic measurements), 2) on the day of surgery, and 3) before discharge from the hospital. All samples will be de-identified, coded, packaged per Category B regulations and transported by courier to MMCRI. Samples will be received, logged and processed by trained lab personnel. Any excess tissue not utilized for this study will be banked and distributed by the BioBank for non-specified research if patient consent has been obtained for this purpose. Specimen handling: Blood specimens will be left at room temperature for 15-20 minutes and then centrifuged for 30 minutes. The supernatant (serum) will be collected and stored on ice for transport to MMCRI. We have performed tests to evaluate the stability of lipids in blood serum and have found that lipids are stable for up to 3 hours at room temperature. All tissue will be frozen as soon as possible and transported frozen on dry ice to MMCRI. All serum and tissue samples will be stored at -80ºC. Experience has shown that samples remain unchanged for more than 1 year under these conditions. Sample analysis will be performed when all samples have been collected. Ventricular tissue: Anterior left ventricular (LV), free wall epicardial biopsies (1.5mm x 1.5-7mm deep/length) will be obtained from patients undergoing bypass surgery by their cardiac surgeon at the time of the surgery25,30. Specifically, the biopsy will be performed after cardiopulmonary bypass perfusion is initiated and once the aorta has been clamped and cardioplegia administered. An area approximately the size of a grain of rice on the surface of the anterior wall of the left ventricle free of epicardial fat will be chosen for the biopsy. The biopsy will be performed in parallel with other procedures to prepare the patient for grafts and will not add to the overall surgery time. The tissue will be placed in physiological buffer solution for cell isolations or lipid extraction for lipidomic profiling. Cell isolations will be performed as soon as possible after tissue is received at MMCRI. Tissue that will be used for lipid extractions can be frozen at -80ºC. Atrial tissue: The atrial tissue will come from the discarded right atrial appendage (RAA) during surgery. Atrial tissue is considerably larger than the ventricular biopsy and allows us to subdivide each sample. One piece will be placed into physiologic buffer for cell isolation. A second piece will be placed in formalin and used for immunofluorescence. The third will be stored in RNAlater® Solution to stabilize RNA until extraction. The final piece will be frozen for isolation of caveolae fractions, lipidomic and protein profiling, and western blotting. Skeletal intercostal muscle tissue: Intercostal muscle is removed during surgery in order to access the left internal mammary artery, which is used to provide a bypass graft. The discarded intercostal skeletal muscle will be divided into two pieces, one frozen for isolation of caveolae fractions, lipidomic and protein profiling, and western blotting, and the second either placed in formalin and used for immunofluorescence, placed in physiologic buffer for cell isolation, or placed in RNAlater® Solution to stabilize RNA until extraction. Blood: A baseline blood sample will be drawn via venipuncture once before surgery (on the day of diet questionnaire administration when patient is having daily clinically-indicated blood draws) to evaluate monocyte expression of CAV as well as plasma lipidomic profiles. Follow-up blood samples will be taken at two other times points: the day of the surgery via a central line and 4 days after the surgery and before discharge via venipucture. Blood samples will be collected into 2 5-ml tubes, an ACD (Acid Citrate Dextrose) anticoagulant tube and a red-capped SST tube and refrigerated or processed within 3h. Data Collection: The medical records of the participants will be reviewed to abstract demographic and clinical data as specified below. Patient data collection will occur at enrollment and approximately 1 month post surgery (for a maximum of three six months). The data will be entered into a study specific, HIPAA-compliant, REDCap database, along with intervention-specific data (questionnaire data, anthropomorphic measurements and diet information). These data, which includes patient names and other identifiers, will only be accessible to clinical investigators. Laboratory data will be maintained separately. Clinical data will be coded and exported for collation with laboratory data and analysis without other personal identifiers. The link to the code will be accessible only to clinical investigators. Personally identifiable information 1. Name 2. Date of birth 3. Medical record number 4. Day of hospital admittance 5. Date of surgery 6. Date of discharge 7.7.2 Demographics 1\. Age (years) 2. Gender (F/M) 3. Race/Ethnicity (OMB categories) 4. Smoking history (never/former/current; if former/current, pack years) 7.7.2 Medical History 1. Date of diagnosis 2. HbA1c levels 3. Standard lipid profile data 7.7.2 Medication Information at time of enrollment 1\. Anti-diabetic medication (drugs, dose, frequency) 2. Lipid-lowering medication (drugs, dose, frequency) 3. Anti-hypertensive medication (drugs, dose, frequency) 7.7.3 Trial-specific data 1. Dietary questionnaire scores for nutrient and food group intake 2. Anthropomorphic measurements (height, weight, BMI, hip \& waist circumference) 3. Daily diet adherence (yes or no; if no what food item was not in adherence) 4. Daily caloric intake 5. Weight during intervention 6. Time of blood sample procurements 7. Research laboratory data (ERBB2 expression data, lipidomics data) 7.7.4 Clinical outcomes 1\. Medications at discharge 2. Post-operative morbidity and survival at 30 days post surgery 8.0 Laboratory Procedures 8.1 Expression levels of ERBB and other proteins: Cell surface ERBB2 expression in atrial endothelial cells will be assessed by flow cytometry. Also, levels of ERBB2, CAV1/3, and IRS-1, will be determined by Western blot using specific antibodies (CAV1: BD# 610058, mouse or Cell Signaling; CAV3: BD# 610420, mouse IRS-1: BD#611395, mouse. pY989-IRS-1: Santa Cruz, SC-17200-R, rabbit, total ERBB2: Life Technologies. pERBB2/Tyr1248: Millipore, 06-229, in combination with goat anti-rabbit IgG conjugated FITC, Abcam, ab6717). pAkt levels will be assessed in circulating monocytes collected from whole blood as well as in isolated atrial and ventricular endothelial cells. Additional lipid and protein analysis: Lipid composition of the baseline plasma samples will be compared with that obtained from the corresponding atrial and ventricular tissue samples to assess the degree that fasting circulating lipids represent tissue content. Non-myocyte single cell suspension will be prepared from atrial tissue using collagenase II/Dispase II digestion as described36. ERBB receptor and CAV1 and 2 levels in conjunction with cell-specific markers will be measured by flow cytometry. Statistical Considerations Sample size/Power: Pilot data was used to determine the likely levels of several measures in our study population and have chosen one to use for a power analysis, cell surface ERBB2 expression in atrial endothelial cells. Pilot data using fluorescence based flow cytometry revealed a mean of 9.56 and SD of 3.14 for this measure in diabetic patients (Fig. 3). Thus, target sample size of 24 patients in each study group will allow for detection of a difference of 2.64 in this measure (hypothesized to be similar in the control group to our preliminary data and higher in the intervention group) with 80% power and a significance level of 0.05. Secondary analyses of plasma lipid content and monocyte signaling will be repeated measures analysis, which will add additional power given that we will have multiple observations per person, and take away some power given the necessity for adjusting p values for multiple comparisons by time point. Analyses: This will be an intention-to-treat analysis. The success of randomization will be evaluated by comparing patient characteristics at baseline using proportions or means and SDs using by chi-square tests or Student's t-tests, as appropriate. Within group changes (i.e. delta score) and between group differences in measures taken at serial time points (blood serum) will be analyzed by mixed model repeated measures ANOVA; the mixed model will be required because patients are likely to have different duration of dietary exposure and, therefore, different numbers of repeated measures. Variables of interest are time, treatment group, and the time by treatment interaction. Model results will be used to estimate mean values and change over time for within-group and between-group comparisons. Graphical techniques and time point-specific p values will be used to identify the time course of any observed changes, e.g., how quickly the dietary intervention might lead to differences in the measures examined. Significance will be identified at the P\<0.05 level, with appropriate adjustment for multiple comparisons where applicable. In the lipidomic analyses a false discovery rate of 0.05 will be set and FDR-corrected P values will be reported. The linear relationship between the change in specific signaling measures (ERBB levels) and the change in lipid species (SM and other classes) will be examined using Pearson's correlation coefficient. Analysis will be conducted in SAS.