EVALUATION OF ENDOMYOCARDIAL BIOPSY IN THE DIAGNOSIS

SUMMARY The reliability of endomyocardial bioptome samples in detecting cardiac rejection was assessed in 26 formalin-fixed previously transplanted hearts. Thirteen human donor hearts (mean postoperative survival 128 days) and 13 baboon donor hearts (mean survival 16.5 days) were studied. Twenty samples were taken under direct vision from the endomyocardium of each heart with an Olympus bioptome catheter. A total of 397 tissue samples was examined “blindly” histologically (177 human and 220 baboon). The bioptome samples were assigned a histological rejection score and then compared with the score accorded multiple tissue sections from the same heart. Sample scores agreed with tissue section scores as follows: humans 86% (samples showed more severe alterations in 5% and less severe in 9%) and baboons 57% (samples more severe in 40% and less severe in 3%). Only 2 false-negative samples were encountered among 285 tissue samples from hearts showing rejection. Changes of rejection were equally distributed between the left and right ventricles. Endomyocardial sampling proved an accurate means of detecting the presence of rejection. In the baboon hearts the endomyocardium tended to show more severe changes than the rest of the myocardium. Endomyocardial biopsy histology is superior to electrocardiography in the diagnosis of early cardiac rejection ( 1–3). The Stanford group ( 1, 2) has shown that serial transvenous biopsy of the transplanted human heart can lead to improved management of acute rejection episodes. Despite the crucial role of graft biopsy histology in the management of patients with cardiac transplantation, we are unaware of any study assessing the accuracy of endomyocardial biopsy in reflecting the state of the general myocardium. The patient who shows cardiac rejection on endomyocardial biopsy is treated with increased immunosuppression. Even if death occurs within a few days, one is unable to gauge how accurately the biopsy represented the state of the myocardium at the time of taking the biopsy. We thus decided to take multiple endomyocardial samples from formalin-fixed donor hearts to allow correlation with the whole heart at the same point in time. We had gained the impression from our human and animal cardiac transplant material that the light microscopic alterations in mild and moderate cardiac rejection did not involve all portions of the heart uniformly (Fig. 1). Our concern in this regard was heightened by reports that rejection injury in the kidney is not distributed uniformly throughout the transplanted kidney (4). It has been stated that the right ventricle is more severely involved than the left ventricle in cardiac rejection (4). If such differential ventricular involvement were true, it would be of importance in our assessment of biopsies taken from heterotopic cardiac transplants with retention of the recipient's heart. In such cases, for technical reasons, it is easier and quicker to biopsy the left rather than the right ventricle, using the Seldin-ger technique. In view of the importance attached to endomyocardial sampling in diagnosing early cardiac rejection and the lack of data regarding its accuracy, we decided to test its reliability in our formalin-fixed human and baboon cardiac transplant material.