Quantitative (99m)Tc DTPA renal transplant scintigraphic parameters: assessment of interobserver agreement and correlation with graft pathologies

Abstract

Various (99m)Tc DTPA scintigraphic quantitative parameters for renal graft function assessment have been recommended, but none is universally accepted. In this study, 439 dynamic renal transplant scintigraphies (DRTS) were retrospectively analysed. In the first set of studies, four observers analysed the 47 random DRTS and interobserver agreement of eleven derived parameters was assessed. In the other set of studies, 181 instances of DRTS, performed on 127 recipients with renal biopsies within five days of each other were selected for correlation with pathology. Hilson's Perfusion index (HI), ΔP, P:Pl, P:U & T10 were selected for this analysis. The pathologies were categorized into renal vascular compromise (RVC; n = 20), acute tubular necrosis (ATN; n = 40), vascular rejection (VR; n = 34), interstitial rejection (IR; n = 33), normal (NOR; n = 36) and unclassified pathologies (n = 18). A majority of the parameters showed good Intraclass correlation (ICC). HI differentiated well between grafts with RVC and the remainder of the study cohort, (p < 0.0001; AUC = 0.84); at a cut-off > 278, it had 84% sensitivity and 78% specificity (Likelihood ratio = 3.8). At < 278, it had 98% 'negative' predictive value for RVC. HI also showed reasonable association with VR (p = 0.02; AUC = 0.62) and IR (p = 0.009; AUC = 0.65). However, significant overlap of HI values between various subgroups was noted. Other parameters had good ICC but were not effective in differentiating graft pathologies. Of the measured parameters, only HI proved to be useful for the pathological assessment, particularly in the identification of vascular compromise. This parameter, however, has lower specificity in differentiating the other pathologies.

Keywords: Acute tubular necrosis; Hilson’s index; interobserver agreement; quantitative renal transplant DTPA scintigraphy; rejection; renal artery stenosis; renal graft.

Figure 1

Flowchart demonstrating the method of the study. (DRTS, Dynamic renal transplant scintigraphy).

Figure 2

ROC curves Hilson’s Perfusion Index (HI) measurements in various pathologies (see also Table 4).

Figure 3

Box & whisker plot demonstrating spread of HI values in A. Normal biopsy versus abnormal biopsy (including RVC) and B. Various pathologies.

Figure 4

Bar chart showing number of studies at two HI cut-offs in each pathology subgroup. On X-Axis, pathology subgroups and on the Y-axis, total number of studies in percentage compared to all studies of that subgroup.

Figure 5

DRTS of a case of impaired graft function eight days post live donor transplantation showed high HI of 259. The biopsy showed findings consistent with vascular rejection. A. Perfusion phase (shown in 4 second per frame) show slightly delayed flow of tracer to the renal graft in the right iliac fossa. B. Clearance phase (1 minute per frame) show good extraction of tracer with some excretion of tracer activity over 20 minutes. C. Perfusion phase curves: pink- Iliac artery, white- renal and blue- background. D. Clearance phase curves: white- renal and blue- background.

Figure 6

DRTS of a case of poor graft function four months post deceased donor transplantation showed high HI of 563. (A) Perfusion phase (shown in 4 second per frame) show delayed flow of tracer to the renal graft in the left iliac fossa. (B) Clearance phase (1 minute per frame) show reasonable extraction of tracer with reduction of tracer activity over 20 minutes together with increased background activity. (C) Perfusion phase curves: pink- Iliac artery, white- renal and blue- background. (D) Clearance phase curves: white- renal and blue- background. (E) Angiogram revealed reduced perfusion to the graft due to a 95% stenosis just proximal to the U bend in the vessel, which is better appreciated in the supplementary 3D angiography film run (F, arrow).