RESUMEN
This study aims to explore the utility of serial duplex scanning and to compare its results with those of single time-point scans of renal allografts in the diagnosis of acute rejection (AR). A retrospective analysis of 6017 serial duplex scans (mean: 9.8 scans per patient, 5.7 of which were done during the first 10 days) was performed in 614 patients with 462 episodes of AR from 1992-2000. Even in the absence of AR (n=278), there were day-to-day fluctuations in pulsatility index (PI) and resistive index (RI). An increase of >10% in intra-renal indices was noted 0.95 days (mean) before the commencement of treatment for AR (SD 1.3, range 1-6 days). In patients with acute tubular necrosis (ATN), who have high base line indices, sensitivity of single value of PI and RI was 58% (cut-off level 1.8) and 68% (cut-off level 0.8), with specificity of 66% and 56%, respectively. By contrast, a >10% increase over the previous 'best' in PI and RI had a sensitivity of 78% and 60% respectively, and a specificity of 78% and 90%, respectively. Reversal of flow during diastole (n=50) was found to be associated with 22% graft loss within 3 months of transplantation. We can conclude that a considerable overlap between the indices of patients with AR and those with ATN greatly limits the diagnostic yield of duplex scanning. Nonetheless, serial scanning of renal allografts is more likely to herald the need for biopsy in the diagnosis of AR than one-time scanning.