A multicenter prospective validation cohort does not support the use of kidney/psoas [18F]FDG uptake in the diagnosis of kidney allograft subclinical rejection.

Abstract

Subclinical kidney allograft acute rejection (SCR) corresponds to "the unexpected histological evidence of acute rejection in a stable patient". The diagnosis of SCR relies on surveillance biopsy. Positron emission tomography (PET/CT) after injection of F18-fluorodeoxyglucose ([18F]FDG) has been proposed as a non-invasive screening approach. In the present multicenter prospective study, we assess the diagnostic yield [18F]FDGPET/CT to rule out SCR in stable KTR at 3 months post KTx. From 01/2021-03/2025, we prospectively combined surveillance biopsy and [18F]FDGPET/CT at ~3 months post transplantation in adult kidney transplant recipients from 4 independent imaging centers. The mean standardized uptake value (mSUV) was measured in kidney cortex and referenced as a ratio to psoas muscle mSUV (mSUVR). Our multicenter cohort of 185 patients was categorized according to the Banff-2022 classification as: normal (n = 158); borderline (n = 18); SCR (n = 9, including 6 T-cell-mediated rejection and 3 microvascular inflammation). No significant correlation was observed between the mSUVR and ti score (R = 0.032, p-value = 0.67). The mSUVR reached 2.33 [1.97-2.93], 2.71 [2.50-3.33] and 2.42 [2.27-3.14] in normal, borderline and SCR groups, respectively. In multivariate models stratified by center, the risk of non-normal histology (n = 27, including borderline and SCR) increased with donor age (OR=1.05 [1.01-1.1], p = 0.02) but not with the mSUVR (OR=4.11 [0.91-18.48], p = 0.07). The Z-score of mSUVR was significantly associated with the risk of non-normal histology (OR=1.542 [1.02-2.33, p = 0.04). The risk of biopsy-proven SCR (n = 9) was not significantly associated with mSUVR. The mSUVR of [18F]FDG PET/CT does not reliably rule out SCR on surveillance biopsy.