Experimental high-frequency ultrasound can detect graft rejection after small bowel transplantation

Abstract

Early diagnosis of graft rejection after small bowel transplantation (SBT) can allow prompt institution of vigorous immunosuppressive therapy, with resultant reversal of the rejection process. The current method for graft monitoring is random mucosal biopsy from a stomal site or through an endoscope. However, because early rejection often has a patchy distribution, it could be missed by random biopsy. We hypothesized that the pathological process of rejection would alter acoustic impedance of the tissue and thus change the ultrasonic patterns of the graft intestinal wall. If this hypothesis is correct, then high-frequency endoscopic ultrasound (US) could be used to monitor the entire transplanted bowel and guide the biopsy, with improved yields. This hypothesis was tested in a rat orthotopic SBT model. Sixty-two intestinal specimens (9 isografts, 12 allografts treated with cyclosporine A [CsA], 22 untreated allografts, and 19 intestines from normal rats) were collected for in vitro transluminal US imaging (30 MHz) and histopathologic study. The echo pattern of normal rat intestinal wall consisted of five echo layers that correlated spatially with the histological layers: the innermost hyperechoic layer 1, plus hypoechoic layer 2, corresponded to the mucosa; hyperechoic layer 3, the submucosa; anechoic layer 4, the muscularis propria; and hyperechoic layer 5, the serosa. The isografts and CsA-treated allografts were identical histologically and ultrasonically to normal intestine. However, the echo patterns of the untreated allografts had progressive loss of architectural stratification, with worsening rejection. The change began with patchy indistinctness and disruption of hyperechoic layers 1, 3 and 5, and progressed to total obliteration of the layers, with the intestinal wall becoming a nonstratified hypoechoic structure. In addition, the normal thickness ratio (2–4:1) of layers 1 + 2 to the remaining deeper layers (3, 4, and 5) was decreased or inverted in the rejected intestine. US detected histologically confirmed graft rejection, with a sensitivity and specificity of 100% and 80%, respectively. Furthermore, there was a strong correlation between the US prediction of rejection severity and actual severity determined by histological study (correlation coefficient = 0.91; P < .001). These data confirm our hypothesis and indicate that high-frequency endoscopic US may be a useful method to monitor graft rejection after SBT.