Background: Echocardiography has been used as a primary means to detect cellular rejection in infant heart transplant recipients. There is, however, limited information correlating echocardiography and biopsy-proven rejection in this age group.
Methods: Between September 1989 and July 1994, 32 consecutive heart transplantations were done in infants younger than 20 months old, who were followed up for 2 to 58 months (mean 28 months) with concurrent endomyocardial biopsy and M-mode echocardiography with digitization. M-mode data from all 16 episodes of rejection (International Society for Heart and Lung Transplantation grade 3A or greater) that occurred in 12 grafts were compared with data from the same grafts with histologic resolution of rejection 2 weeks after treatment and with data from biopsy-proven nonrejecting control grafts matched for sex, time after transplantation, donor weight, and donor age.
Results: Left ventricular mass index increased in rejection (86 +/- 9 gm/m2) versus resolution (64 +/- 6 gm/m2) and versus that in nonrejecting control grafts (59 +/- 8 gm/m2). Left ventricular shortening fraction increased in rejection (40% +/- 2%) versus resolution (38% +/- 10%). Septal thickening fraction decreased in rejection (33% +/- 9%) versus nonrejection (68% +/- 16%). These changes became significant only in grafts transplanted more than 1 month before study. Substantial overlap of measurements prevented identification of threshold values. Intraobserver and interobserver variabilities for standard M-mode data were 7% to 8% and 12% to 22%, respectively, whereas those for digitized parameters were markedly elevated at 37% to 71% and 51% to 81%, respectively.
Conclusions: We found (1) left ventricular mass index increases in cellular rejection but may be unreliable less than 1 month after transplantation and (2) significant interobserver and intraobserver variability may limit the applicability of digitized echo parameters to the detection of rejection in infant heart transplant recipients.