Effect of Classic Secondary Prevention in Type 2 MI: A Target Trial Emulation Study

Overview

A classic heart attack is caused by a blockage to the coronary arteries that supplies the heart muscle with oxygenated blood. The medical term for this condition is type 1 myocardial infarction. There is strong scientific evidence that the usage of pharmacological drugs such as statins, beta blockers, Renin-Angiotensin-Aldosterone System blockers and platelet inhibitors after a type 1 myocardial infarction improves survival and reduces the risk for new myocardial infarctions. However, a myocardial infarction may also occur without blockage to the coronary arteries when other acute conditions causes either a decreased supply or an increased demand of oxygenated blood to the heart. The medical term for the latter is type 2 myocardial infarction. There are currently no scientific evidence that any pharmacological drug improves survival in patients with a type 2 myocardial infarction, of whom only one in three patients are alive after five years. The aim of this study is to investigate if those drugs that improves the prognosis after a type 1 myocardial infarction (Beta blockers, Renin-Angiotensin-Aldosterone System blockers, Statins and platelet inhibitors) also affects the prognosis after a type 2 myocardial infarction. The best way to answer this question would be to conduct clinical trials for each drug where type 2 myocardial infarction patients are randomized to either receiving the drug of interest or receiving placebo (sugar pills) and then compare the survival and outcomes in both groups over time. However, clinical trials are costly, time consuming and also difficult to conduct with type 2 myocardial infarction patients since these patients are treated at various hospital departments. Therefore, this study will instead include patients in a Swedish national register for myocardial infarction, in which myocardial infarction patients are reported from all Swedish hospitals, and compare type 2 myocardial infarction patients that did receive or did not receive each treatment. To minimize the risk of making inaccurate conclusions about the causal relationship between treatment and outcome, the study will define the optimal clinical trial for each treatment and then specifically emulate these trials in all possible aspects using the register data. This method is called "target trial emulation".

While acute and long term treatments in type 1 myocardial infarction (MI) are well documented, there are no evidence based treatments altering the prognosis in type 2 MI patients, of whom only one in three are alive after five years. Classic secondary preventive treatment in type 1 MI include the use of statins, beta blockers, renin-angiotensin-aldosterone system (RAAS) inhibitors, single antiplatelet therapy (SAPT) and dual antiplatelet therapy (DAPT). While some observational studies suggest classic secondary preventive MI treatment to be associated with better survival in type 2 MI, others do not. The aim of this study is to investigate if there is an association between treatment with each of statins, beta blockers, RAAS inhibitors, SAPT or DAPT; and all-cause mortality or long-term cardiovascular events in patients with type 2 MI. The study will be conducted as (several) registry-based observational studies emulating pre-specified target trials. A protocol of a pragmatic randomized controlled trial (a target trial) will be specified for each of statins, beta blockers, RAAS inhibitors, SAPT and DAPT; including eligibility criteria, treatment assignment, time zero and follow up, outcomes, causal contrast and statistical analyses. Each of these components will then be emulated using registry data. The target trial, as well as the registry-based study emulating the target trial, will differ slightly for each treatment. The study will include all, approximately 14 000, patients reported as type 2 MI (first reporting for each patient) in the Swedish Web-system for Enhancement and Development of Evidence-based care in Heart disease Evaluated According to Recommended Therapies (SWEDEHEART) from 2010 and onward. Data from SWEDEHEART will be merged with data from the National Patient Register, the Swedish Cause of Death Register and the Longitudinal Integrated Database for Health Insurance and Labour Market Studies. Patients will be followed from hospital discharge (time zero) until death, loss to follow-up or end of follow up (date of data collection). Treatment specific eligibility criteria will be applied at time zero. All important confounding factors at time zero will be identified by drawing causal diagrams, one for each treatment and outcome. Treatment specific models with registry data will then be created to balance the assigned groups using for example inverse probability weighting with multiple imputation of missing data among covariates. For all target trial emulations, the primary outcome of interest will be a composite endpoint of all-cause mortality, readmission for MI, stroke or heart failure. Secondary outcomes will be all-cause mortality, cardiovascular mortality, readmission for MI, readmission for stroke and readmission for heart failure individually. Safety outcomes will be specified separately for each treatment. Readmission for bacterial pneumonia will be used as a negative control outcome in all target trial emulations to check for residual confounding. Intention-to-treat analysis will be performed for each treatment and outcome. Competing risk methodology will be applied as appropriate.