Cardiovascular Magnetic Resonance-Guided Rejection Surveillance in Cardiac Transplantation: Five-year Safety Outcomes

Introduction: The NIDACT-2 study demonstrated that cardiovascular magnetic resonance (CMR) surveillance of cardiac allograft rejection (CAR) provided a non-invasive, non-inferior alternative to traditional endomyocardial biopsy (EMB) during the first year after orthotopic heart transplantation (OHT). We assessed 5-year safety outcomes in this cohort.
Method: Forty patients randomised to either CMR- or EMB-based surveillance to guide immunosuppression during the first-year post OHT were followed for five years. Outcomes collected included mortality, CAR, all-cause and infection-related hospitalisation, coronary allograft vasculopathy (CAV), and graft function. Clinically-indicated CAR surveillance by CMR or EMB beyond the first year were performed at the discretion of the treating physician. Invasive angiogram, CT coronary angiography, exercise stress echocardiography, and sestamibi myocardial perfusion scanning were performed at the discretion of the treating physician for CAV assessment. Clinical outcomes and analyses included Chi-square, Kruskal-Wallis test, Mann-Whitney-U testing, negative binomial regression, and Kaplan-Meier survival analysis.
Results: Eight deaths occurred over 5 years (CMR n=4, EMB n=4), with no difference in mortality risk (HR 0.985; 95%CI 0.246 – 3.940, p=0.983). The number of patients who experienced high-grade CAR (≥2R) did not differ significantly between groups (CMR n=9 (45%), EMB n=8 (40%); Chi2 = 0.102, p=0.749) nor did time-to-first high-grade rejection event (HR 1.126, 95% CI 0.434 – 2.921, p=0.807). Hospitalisation event rates (CMR n=39, EMB n=96; OR 0.81, 95% CI 0.55-1.17, p= 0.274) and total length of stay per patient (19±28 vs 37±50 days, p= 0.140) were similar, although time-to-first hospitalisation was longer in the CMR group (HR 0.411; 95% CI 0.194 – 0.870, p=0.020).  Similar infection-related hospitalisation event rates (CMR n=20, EMB n=38; OR 0.38 95% CI 0.13-1.10, p=0.06) and time-to-first infection-related hospitalisation (HR 0.464; 95% CI 0.192 – 1.124, p=0.089) were observed. There was no significant difference in CAV incidence (HR 1.262, 95% CI 0.449 – 3.548, p=0.659) nor any significant change in left ventricular ejection fraction between groups (CMR 61.8±3.4 vs EMB 55.9±13.9, p=0.26).
Conclusion: CMR-guided rejection surveillance after cardiac transplantation demonstrates comparable 5-year safety outcomes to EMB-guided surveillance.