How Should the STICH Trial Results Affect Clinical Practice?

Michael J. Mack


Nat Rev Cardiol 

Abstract and Introduction


The STICH trial showed that CABG surgery does not necessarily improve cardiovascular outcomes in patients with coronary artery disease and left ventricular dysfunction who are receiving optimal medical therapy. However, surgical revascularization should still be considered if the coronary artery disease is severe and viable myocardium can be identified.


Many patients with coronary artery disease (CAD) present with impaired left ventricular function that might be reversible by revascularization. The presence of angina, the assessment of myocardial viability, or both are frequently used as decision-making tools to select patients who might benefit from CABG surgery. The rationale is that revascularization of viable, but hibernating, myocardium leads to improved left ventricular function and survival. The investigators in the Surgical Treatment for Ischemic Heart Failure (STICH) trial[1,2] tested this hypothesis by assessing the role of CABG surgery in patients with multivessel CAD and impaired left ventricular function, defined as an ejection fraction of 35% or less. The results of this trial, and those of a substudy on the role of myocardial-viability assessment for the identification of patients for whom CABG surgery would be beneficial, have now been published,[1,2] together with an accompanying editorial.[3]

The conclusion from the main STICH trial[1] was that no significant difference in survival existed between patients randomly assigned to receive medical therapy alone and those who received medical therapy plus CABG surgery. Furthermore, the substudy in which myocardial viability was assessed by either single-emission photon computed tomography or dobutamine echocardiography, demonstrated that although the presence of viable myocardium was associated with an increased likelihood of survival, the relationship was not significant after adjustment for other baseline variables.[2] James Fang, who wrote the editorial that accompanied these two papers in the New England Journal of Medicine, concluded that "medical therapy for coronary disease was underestimated … again".[3] Dr. Fang was referring to the results of the COURAGE trial,[4] which demonstrated that percutaneous coronary intervention did not reduce the risk of death, myocardial infarction, or other major cardiovascular events when added to optimal medical therapy.

So, how do we interpret the results of the STICH trial, and what are the implications for everyday decision-making in the care of these patients? Should clinical practice change? In order to answer these questions one must first ask whether the trial results are valid and, if so, whether they can be generalized to the population at large. In this respect, examining the original design of the trial, its limitations, and its exclusions is instructive. The researchers enrolled 1,212 patients into the STICH trial at 127 clinical sites in 26 countries over 5 years; an average of two patients per site per year.[1] The original plan was to recruit 2,000 patients and follow them up for 3 years, but enrollment was slow and this target was reduced. The follow-up period was lengthened to 5 years so that the trial would remain adequately powered. The STICH trial was actually designed to address two separate questions on the management of patients with heart failure, CAD amenable to surgical revascularization, and decreased left ventricular function.[5] First, whether medical therapy combined with CABG surgery was superior to medical therapy alone in prolonging survival in these patients. Second, whether the addition of surgical ventricular reconstruction to CABG surgery improved hospitalization-free survival among patients with anterior-wall dysfunction. The results of the study addressing the second question have previously been published, and showed that additional surgical ventricular reconstruction was not beneficial.[6]

The complexities of investigating these two separate hypotheses in the same trial,[1] as well as conducting the myocardial viability substudy,[2] led to an unwieldy screening process that required patients to be included in one of three strata before randomization—perhaps contributing to the lower than expected enrollment.[7] Although no screening-to-enrollment ratios are available, recruitment at our site, Medical City Dallas Hospital, was extremely low because of confusion among investigators and research coordinators regarding trial eligibility. Trial exclusion criteria might have further contributed to low trial enrollment and, therefore, limited the extent to which the results of the trial can be generalized beyond the studied population. First, patients with left main CAD and those with severe angina were excluded, conceivably leading to a bias against CABG surgery, which is the standard treatment for these conditions. Second, assessment of myocardial viability was not mandated; patients in whom viability was assessed were not subjected to randomization and assessment varied widely between sites, perhaps also leading to bias. Third, a substantial 17% of patients assigned to medical therapy crossed over and underwent CABG surgery. In the intent-to-treat analysis, these patients were included in the original assigned cohort, which might have led to underestimation of the benefits of CABG surgery. Fourth, because the primary end point of death from any cause was not met, the secondary end points have to be considered observational and hypothesis-generating only. Nonetheless, patients assigned to CABG surgery had a lower rate of death from cardiovascular causes and a lower composite of death from any cause or hospitalization for cardiovascular causes than those assigned to medical therapy only. The implications of the STICH trial[1,2] for clinical practice, are outlined in Box 1.

The authors and investigators of the STICH trial are to be congratulated for bringing to fruition an extremely difficult randomized clinical trial with robust data for us to consider in the everyday clinical management of patients with CAD and left ventricular dysfunction. However, it is my firm belief that we should be 'splitters' rather than 'lumpers' and take the overall findings of the trial, realize the constraints affecting the extent to which they can be generalized, and tailor therapy to the individual patient within the context of current guidelines.[8]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: