Early and late AMR in heart transplantation—Distinct entities?

Abstract

Acute cellular rejection (ACR) has been appreciated as an important cause of morbidity and mortality since the early days of heart transplantation. Substantial improvements in survival were only made with the advent of therapies to prevent and treat this specific form of rejection. In contrast, only in the last decade or so has antibodymediated rejection (AMR) been accepted and recognized as an important and distinct entity associated with adverse outcomes. This late recognition has been in no small measure due to the significant challenges in developing and establishing an acceptable consensus on diagnostic criteria for AMR. 1 Early reports suggested that AMR was seen early after transplant and more likely to be associated with hemodynamic compromise and development of cardiac allograft vasculopathy (CAV). 2 More recently, there has been increased recognition that AMR may be more sporadic and occur at any time after transplant. Coutance et al observed a very high mortality, higher incidence of graft dysfunction, and fulminant CAV in patients with late AMR. 3 In this issue of the journal, Clerkin et al report that, although AMR may be more common within the first year of transplant, its development several years post-transplant may be of greater concern. In a single-center, retrospective analysis almost 10% of the transplant cohort (689 patients) was identified with AMR. 4 More than a third of these were diagnosed late after transplant (mean 1,084 days, compared with 23 days for those with early AMR). Early AMR was associated with a lower incidence of graft dysfunction (25.6%), but with subsequent survival comparable to that of transplant recipients without AMR. In contrast, more than half the patients (56%) had graft dysfunction when AMR occurred late. Late AMR correlated with a very high incidence of accelerated CAV (50% at 1 year) and a higher mortality. It is notable that both AMR groups were compared with all patients without AMR, which presumably included patients with ACR. These observations raise the question of whether early and late AMR are truly distinct entities. Outcomes after rejection depend on a number of major factors, including severity, timing of diagnosis and efficacy of treatment. Early AMR has the distinct advantage that it is likely to be detected on protocol biopsies performed more frequently in the first year and therefore promptly treated. In this study, 98% of early AMR was detected on protocol biopsies compared with only 32% of late AMR, and 490% of all AMR episodes were treated. The majority of late AMR episodes were detected because, presumably, symptoms prompted the biopsy, which infers a selection bias for more severe disease. This may explain the worse outcomes observed in the setting of graft dysfunction in the study. Patients with graft dysfunction received significantly more treatments and there was a trend for a need for more treatments in the late AMR group overall compared with the early AMR group. The presence of graft dysfunction also more likely prompted evaluation for CAV increasing the likelihood of its early detection. Is it possible that patients with late AMR may have had the disease for some time, and would earlier diagnosis improve outcomes? Recently, Loupy et al reported that unrecognized sub-clinical AMR episodes may be observed