T1 mapping for diagnosis of early cardiac cellular allograft rejection

Abstract

Background Non-invasive surveillance of acute cellular rejection (ACR) has become a primary focus in heart transplantation (HTx), aiming to reduce the frequency of invasive endomyocardial biopsies (EMB). While cardiac magnetic resonance (CMR) with T1 mapping has shown promise as a diagnostic tool, its application during the early post-transplant phase remains a significant knowledge gap, as perioperative tissue injury can confound imaging interpretation. Objective To evaluate the accuracy of cardiac T1 mapping for ACR in heart transplant recipients within the first 60 days post-transplantation, using EMB as the reference standard. Methods This prospective single-center study included adult HTx recipients within 60 days post-transplant. All patients underwent CMR with native T1 mapping and extracellular volume (ECV) quantification. EMB served as the gold standard, classifying patients into non-significant (<2R) and significant (≥2R) rejection groups. Results Forty-five recipients were included; 14 (31.1%) presented with cellular rejection ≥2R. Baseline characteristics such as recipient mean age (50.1 vs 49.6 years, p = 0.79) and the interval from HTx to CMR (32 vs 31.4 days, p = 0.86) were comparable between groups. Native T1 (1048 ± 46.2 vs 1006.6 ± 60.6 ms, p = 0.028) and ECV (36.8 ± 5.3% vs 33.2 ± 5.5%, p = 0.043) were significantly elevated in the ≥2R group. Native T1 demonstrated an AUC of 0.700 for diagnosing rejection ≥2R (cutoff: 1036.5 ms; sensitivity 64.3%, specificity 74.2%, PPV 52.9%, NPV 82.1%), whereas ECV yielded an AUC of 0.726 (cutoff: 37.9%; sensitivity 57.1%, specificity 90.3%, PPV 72.7%, NPV 82.4%). Multivariate analysis identified indexed RV end-diastolic volume and the presence of late gadolinium enhancement (LGE) as independent predictors of rejection. However, a multiparametric approach combining native T1 <1037 ms, ECV <38%, and absence of LGE yielded a NPV of 100% within this small subgroup. Conclusion The diagnostic accuracy of standalone T1 mapping for identifying ACR was modest in this early post-HTx period. However, native T1 and ECV were significantly higher in the rejection group, and their integration into a multiparametric approach, including LGE, demonstrated a high NPV in this cohort. These findings highlight CMR’s potential for monitoring myocardial changes even in the early post-transplant stage.