| Popis: |
Seven patients underwent cardiac transplantation at Ste-Justine Hospital between June 1984 and October 1988 (age 3-16 years). Indices derived from M-mode echocardiography--left ventricular mass (LVM), isovolumetric relaxation time (IRT), shortening fraction (SF), velocity of circumferential fiber shortening (VCF), ejection fraction (EF), LV pre-ejection period/ejection time (PEP/ET), isometric contraction time (ICT)--and electrocardiography--summation of potentials (Sum Pot)--were used as non-invasive markers of rejection (367 studies), while rejection status was monitored by serial endomyocardial biopsy (55 biopsies). Spontaneous variation of such indices, as defined by the 90% distribution interval of their variation when biopsy was unchanged, was found to be important: -20% to +28% for IRT, -30% to +28% for LVM, -30% to 32% for Sum Pot, -32% to 42% for PEP/ET, etc. Using limits of spontaneous variation such as thresholds, none of the studied indices could consistently predict the 2 biopsy-documented episodes of rejection or the 7 episodes of improvement. Among all the studied indices, IRT was the most promising, correctly predicting 1 of the episodes of rejection and almost predicting the 2nd. Remeasurement of all M-mode tracings by an unique observer, following strict rules to maximize reproducibility, did not decrease the spontaneous variation of these indices, nor did it improve their performance as predictors of changes in the rejection status. These results warrant the continuing search for other non-invasive methods for accurate monitoring of the rejection status. |
