Diagnostic Accuracy of CMR for Cardiac Transplant Rejection

Quick Takes

  • In this meta-analysis, 10 studies were included that compared diagnostic accuracy of CMR imaging to endomyocardial biopsy (EMB) for cardiac transplant rejection. It is limited by significant heterogeneity and bias.
  • T2 imaging had the highest diagnostic accuracy with a pooled sensitivity of 86.5% and specificity of 86% in detecting cardiac allograft rejection compared to EMB.
  • Late gadolinium enhancement had the lowest diagnostic accuracy in detecting rejection in cardiac allograft and does not have a role in detecting rejection.

Study Questions:

What is the diagnostic accuracy of cardiac magnetic resonance (CMR) imaging for the diagnosis of cardiac allograft rejection?


The authors performed a meta-analysis including studies in adults undergoing a cardiac transplant and CMR using a 1.5T to 3T magnet characterizing myocardial tissue with measurement of T1, T2, extracellular volume (ECV), or late gadolinium enhancement (LGE). All included studies used endomyocardial biopsy (EMB) for reference.


After an initial review of 478 potential relevant manuscripts, 10 studies were included. Mean age was 34-71 years and 67% were men. All CMRs were performed using 1.5T machines. Diagnostic accuracy was best for T2 imaging that was evaluated in five studies with 276 CMR-EMB pairs with 86.5% pooled sensitivity and 86% pooled specificity for rejection. This was followed by native T1 imaging, studied in four studies with 267 CMR-EMB pairs, and had a pooled sensitivity of 85% and pooled specificity of 70% for rejection. ECV was next, evaluated in five studies with 132 CMR-EMB pairs, and had a pooled sensitivity of 91% with a pooled specificity of 68%. Diagnostic accuracy was least for LGE, evaluated in five studies with 364 CMR-EMB pairs with a pooled sensitivity of 50% and specificity of 60%. There was significant heterogeneity in sensitivity and specificity for nearly all four CMR parameters assessed. Half of the included studies had an unclear risk for bias.


In this meta-analysis, T2 imaging on CMR had the highest diagnostic accuracy of all parameters assessed on CMR in detection of cardiac transplant rejection, while LGE had the least. There was substantial heterogeneity across studies, and risk of bias could not be determined in 50% of studies.


Cardiac allograft rejection is the leading cause of death in heart transplant recipients in the first year after transplantation. EMB is considered the gold standard for detection of rejection but is associated with 3% risk of complications and can be limited by sampling error. Accordingly, there has been an increased focus on finding alternate noninvasive means of detecting rejection, and CMR imaging is one modality that has been studied for the same. This meta-analysis, however, reveals significant heterogeneity and risk of bias among included studies. Furthermore, only 10 studies were included, reducing generalizability. T2 imaging on CMR had the highest pooled sensitivity and specificity, and LGE the least for detection of allograft rejection. Nonetheless, pooled sensitivity and specificity for T2 imaging were 86%, raising concern for missing rejection in a fair number of transplant patients. Additional prospective studies are needed to examine the diagnostic accuracy of CMR in transplant rejection, and ideally, the diagnostic accuracy of a combination of these different parameters should be studied.

Keywords: Allografts, Biopsy, Cardiac Surgical Procedures, Contrast Media, Diagnostic Imaging, Gadolinium, Graft Rejection, Heart Failure, Heart Transplantation, Magnetic Resonance Imaging, Magnetic Resonance Spectroscopy, Myocardium

< Back to Listings