New Time Clock for ST-elevation MI Based on Biochemical Myocardial Infarction Onset Time (BIT-STEMI)

ST segment elevation myocardial infarction (STEMI) is one of the leading causes of death across the world (1) and immediate treatment with either thrombolytics or percutaneous coronary intervention (PCI) results in lower mortality (2,3,4). Prior studies have demonstrated that in patients with STEMI approximately 50% of jeopardized myocardium undergoes necrosis in the first hour, and after 6 hours, myocardial salvage is minimal or absent (5,6). Among such patients each minute of delay in reperfusion treatment raises 1-year mortality; in fact, the risk of 1-year mortality increases by 7.5% for each 30-minute delay (7).

As the outcome of reperfusion treatment in STEMI is time dependent, it is essential to accurately determine the time of onset of myocardial infarction. It is standard clinical practice to take the time of symptom onset as a surrogate for artery occlusion time, and this helps in decision making about mode of reperfusion in STEMI-patients. However symptom onset is a subjective parameter and affected by multiple factors such as recall issues in elderly patients and preceding unstable angina symptoms before artery occlusion. This mandates need for an objective method to know the artery occlusion time. Different time intervals are defined for STEMI patients that include, "artery occlusion to symptom onset time", "symptom onset to first medical contact time" and "door to balloon time" and combination of all these intervals is "Total Ischemic time" i.e. the time between artery occlusion and re-opening, which is strongly correlated with 1-year mortality (8).

In a recent study by Mahmoud et al. (9) an objective method, biochemical onset time is proposed for estimation of artery occlusion time using serial cardiac troponin T (cTnT) levels in patients with STEMI. However, this study was retrospective, had an average of two measurements of cTnT for each patient, peak troponin level was frequently missing and newer earlier detectable biomarkers such as high sensitive Troponin I (hsTnI) were not used.

The objective of this pilot study is to observe if there is any role of multiple samples of markers of myocardial injury in prediction of the biochemical onset time. We plan to use multiple samples (four) of hsTnI for each patient using the same method as above (9); and we will compare the biochemical ischemic time with the patient reported symptom onset time. Secondarily, we will try to determine whether a single sample of multiple cardiac biomarkers with different release kinetics drawn at time of patient presentation in emergency room (ER) can also predict precise time of onset of myocardial infarction, which would be a revolutionary finding in STEMI patients' early management. For this purpose we will use three markers including cardiac specific myoglobin (CS-Mb) and Creatinine kinase-Mb (CK-Mb) (10). We also plan to test three other novel markers which has previously shown to have some role in post MI prognosis and cardiac event risk prediction in general population {(Mean neutrophil volume (MNV), Gamma Prime fibrinogen (γ' Fibrinogen) and heart type fatty acid binding protein (h-FABP)} to assess whether it has a possible predictive value to assess time of onset of MI. Lastly, we plan to assess the prognostic value of biochemical ischemic time versus conventional ischemic time for the outcomes of major adverse cardiovascular events (MACE) at hospital discharge and 30 days post MI, infarct size using CMR and echo based parameters.

OBJECTIVES

1. To determine the biochemical onset time using multiple measurements of hsTnI from each patient (zero, 03, 08, 24hrs), and compare this biochemical time to the patient-reported symptoms onset time as an indicator of coronary artery occlusion.
2. To predict Biochemical occlusion at the time of presentation with the use of single sample of six different markers of myocardial injury.
3. To assess the association of conventional ischemic time and biochemical ischemic time with infarct size; using peak hsTnI, percent ejection fraction by Echocardiography and Cardiac Magnetic Resonance imaging (CMR) based infarct volume in grams.
4. To assess the association of conventional ischemic time and biochemical ischemic time with in-hospital and 30-days major adverse cardiac events, MACE; a composite of heart failure, shock, re MI or death.

METHODS This will be a single center nonintervention pilot study on 100 patients on consecutive patients coming to ER Tabba Heart Institute Karachi, Pakistan, (THI) with acute STEMI. Data will be collected prospectively.

Sampling technique:

Non probability consecutive sampling will be done to enroll patients from the ER of the hospital. Consecutive patients presenting at emergency of THI with diagnosis of STEMI within 24 hours of symptom onset will be assessed for eligibility.

Questionnaire and interview:

Research Officer (RO): three data collectors and one research officer will be employed full time for the duration of the study. Their job description will include to make sure consent forms and questionnaires are completed and to fill out the deficiencies from patients' medical record files, to keep log of the patient forms and to arrange team meetings every fortnight. Also they will follow the patients throughout the hospital stay to record relevant events. Also he or she will be responsible to call and coordinate the patients for follow up echo and CMR at 30 day post MI. This person will also be trained by the PI prior to the start of enrollment.

Ethical approval: Ethical approval has been taken from ethical review committee in THI.

Informed consent: Before enrollment, patient or the designated attendant will be asked to sign an informed consent form, provided both in Urdu and English as applicable to the patients and attendants. The patient will be explained about their rights during the study, regarding withdrawal and refusal anytime during the study. If the potential participant disagrees to consent then the staff will politely inquire about the possible reasons behind it and will offer reconsideration on next visit if possible. There will be no coercion, persuasion or resentment from the study staff on refusal to consent. One copy of the consent form will be handed to the patient along with important contact numbers.

Questionnaire: research officer will fill the questionnaire with the help of patient and the attendants. The information which is not supplied by the participant will be attempted to retrieve from the medical records and hospital data registry.

Assessment of outcomes Blood sampling Four trained phlebotomists will be utilized for blood sampling. The study staff will take the samples; label them with patients' study ID and medical record (MR) numbers, time and number of sample, and the biomarker being assessed. Separate tubes will be used for each biomarker. To ensure timely sampling for each patient a sheet with timed slots will be added to the patient's file to mark number of samples and their timings. The test tubes will be promptly delivered to the lab by the phlebotomist. Cost of blood testing of routine and additional cardiac biomarkers will be borne by the investigating team.

Assessment of Infarct size 24hrs hsTnI: Blood sample will be drawn at 24hours of admission for peak HsTnI as surrogate for the infarct size.

Transthoracic Echocardiogram: A post STEMI standard transthoracic echo will be performed on all patients during index hospitalization which included detailed assessment of ejection fraction and wall motion abnormality. At 30-days, re assessment of left ventricular (LV) function will be performed via 2D Echo.

Cardiac MRI: Myocardial infarction can be visualized and quantified using inversion recovery imaging 10 to 15 minutes after intravenous administration of gadolinium contrast (late enhancement imaging). Infarct size is expressed as absolute mass or percent of left ventricular mass (mass [grams] = volume [mL] × myocardial density [1.05 grams/mL]). Other additional parameters such as left ventricular volumes, ejection fraction, micro vascular obstruction (MO) and myocardial salvage index (MSI) will also be obtained. Infarct size and late MO will be assessed in late gadolinium enhancement (LGE) short-axis images covering the whole left ventricle (LV) approximately 15 min after injection of gadolinium chelate.

Assessment of Major adverse cardiovascular events (MACE):

Patients will continue to get their routine standard care in the inpatient areas. After the discharge in cardiology clinics, they will continue to receive standard medical care including adjustments in drug regimens according to primary physician's opinion. For in hospital MACE, data regarding heart failure (HF), cardiogenic shock, re-MI and in hospital death will be calculated. At 30 day, data on re MI and death will be collected.

Sample size Sample size is not determined using statistical power calculations. Sample size is determined by the number of available specimens of each cohort.

STATISTICAL ANALYSIS Means ± SD will be reported for continuous variables if normally distributed or median ± IQR if not normally distributed. Percentages will be used for categorical variables. Boxplot will be to show the data range for each biomarkers.

Scatter plots will be generated using the time between symptom onset time and sample collection time vs. biomarker values for each biomarker. PROC TRANSREG will be used to fit a curve to the data if applicable. Scatter plots will be generated using time between symptom onset time and sample collection time vs. change of biomarker values from previous time point for each biomarker. PROC TRANSREG will be used to fit a curve to the data if applicable.

Scatter plots will also be generated using time between symptom onset time and sample collection time vs. percent change of biomarker values from previous time point for each biomarker. PROC TRANSREG will be used to fit a curve to the data if applicable.

Regression Analysis Based on the above scatter plots, for each biomarker, a linear/nonlinear regression will be run using time between symptom onset time and collection time as dependent variable and using initial biomarker value, change of the biomarker value, age (in deciles), gender and any interaction terms of these variables as independent variables. Backward selection will be used to choose the best model for each biomarker.

Multiple Biomarkers Analysis Change for each biomarker will be calculated at the presentation using biomarker value at presentation against corresponding limit of detection (LOD).

1. Regressions will be run using time between symptom onset time and collection time as dependent variable and using initial biomarker value, change of the biomarker value, age (in deciles), gender and any interaction terms of these variables as independent variables. The best regression model selection based on the minimum of Akaike Information Criterion (AIC) will be chosen.
2. Principle component analysis (11) will be used to determine all the principal components with Eigen values greater than 1 using the following variables: change of each the six biomarkers, age, gender.

Regression using time between symptom onset time and collection time as dependent variable and principal components as independent variables will be run.

A P-value <0.05 will be taken as statistically significant. SAS version, 9.2 will be used for analysis.

Sensitivity analysis: In the sensitivity analyses, we will use mixed effects Tobit model with random effects that will account for repeated measures of biomarker within the same subject, and compare the results with the main fixed-effects analysis. We will also conduct sub-group analyses according to type of treatment i.e. thrombolytic therapy vs. PCI.

Secondary analysis: Estimates of first samples of the biomarkers (hsTnI, CS myoglobin, CK-Mb, MNV and gamma' fibrinogen) at the time of ER presentation will be used to calculate the biochemical onset time of MI using the same model. We will derive an equation using the regression coefficients of the biomarkers against their respective times of measurements. We will determine the discrimination and calibration (best fit) of the model with all these biomarkers combined or any of their combination compared to hsTnI only. If feasible, a nomogram will be developed using these regression coefficients to predict the biochemical MI onset time.

Additional analysis: To analyze the associations of patient-reported symptom onset time and biochemical onset time separately with infarct size using HsTnI, 2D echo derived EF and CMR derived infarct size, in-hospital and 30-day MACE, we will use Cox proportional hazards models.