Introduction: Diagnosing acute rejection (AR) after kidney transplantation relies on an invasive kidney biopsy. Novel, preferably non-invasive, markers are needed to indicate AR. Kidney-derived urinary extracellular vesicles (kd-uEVs) hold promise as candidate markers, but a comprehensive understanding of their release after transplantation is lacking. This study aims to elucidate kd-uEV dynamics post-transplant and their potential as biomarkers to distinguish AR from another common allograft complication, acute tubular necrosis (ATN).
Method: In the discovery cohort, urine was collected from donor-recipient pairs (n = 73) pre-transplant and from the recipients on day 3 (D3), day 7 (D7), day 180 (D180), and before a for-cause biopsy post-transplant. Urine from recipients (n=30) with biopsies within two weeks post-transplant served as validation cohort. Urine was centrifuged to remove cells and then stained with the uEV-marker CD63 combined with a kidney-specific marker, aquaporin 2 (AQP2; collecting duct), or podocalyxin (PODXL; podocyte). Kd-uEVs were quantified using imaging flow cytometry. Urine dilution was corrected by calculating kd-uEV percentages in the total CD63+ uEVs.
Results: In the discovery cohort, of the CD63+ uEVs from donor urine, 4.2 [0.9 – 11.9]% were AQP2+, and 6.4 [1.5 – 10.7]% were PODXL+. Compared with donors, pre-transplant recipients displayed minimal levels, with 1.1 [0.2 – 1.7]% (p < 0.001; AQP2+) and 1.5 [1.0 – 3.1]% (p < 0.001; PODXL+). These recipient levels remained minimal on D3 but increased on D7 to 7.3 [3.1 – 18.2]% (p < 0.001; AQP2+; Figure 1) and 10.0 [3.2 – 16.3]% (p < 0.001; PODXL+) which persisted until D180. In the discovery cohort, 78% of AR cases occurred in the first two weeks post-transplant. During this period, AR showed a significantly higher level of AQP2+ uEV (16.1 [7.9 – 31.9]%) than the ATN group (1.6 [0.5 – 6.4]%; p = 0.006; Figure 2). Similarly, PODXL+ uEV were also higher in the AR, 13.4 [6.6 – 27.4]%, than the ATN, 3.1 [0.6 – 7.0]% (p = 0.027). The receiver operating characteristic (ROC) analysis demonstrated a 0.934 area under the curve (AUC) for AQP2+ uEV (p = 0.002), 0.868 for PODXL+ uEV (p = 0.008), and 0.925 for the combination (p = 0.002) to distinguish AR from ATN in the first 2 weeks after transplantation. In the validation cohort, still, AQP2+ uEV was observed with an increase in the same period of AR (14.3 [7.1 – 20.4]%) than the ATN (3.3 [1.5 – 5.5]%; p < 0.001). In parallel, PODXL+ uEV was elevated in the AR (9.4 [6.6 – 19.3]%) compared to the ATN (2.9 [4.7 – 10.8]%; p = 0.02).
Conclusion: Post-transplant, AQP2+ uEV release initiates more from D7 than from D3 onwards. For patients with a for-cause biopsy within the initial two weeks after transplantation, increased kd-uEV release indicates AR. These findings underscore the promising potential of kd-uEVs as non-invasive biomarkers for identifying AR early after kidney transplantation.