URINARY PRESEPSIN IMPROVES THE ACCURACY OF DIAGNOSING TUBULOINTERSTITIAL NEPHRITIS

Kidney biopsy is the gold standard for diagnosing kidney diseases. However, it carries a risk of bleeding complications and is sometimes infeasible. Additionally, the procedure requires time to schedule and perform. In urgent cases requiring prompt treatment, therapy may be initiated before or without performing a biopsy. Thus, there remains a clinical need for diagnostic tools that can serve as alternatives to, or supplements for, kidney biopsy.

Plasma presepsin (PSEP) is a well-established biomarker for early detection of sepsis. It is the N-terminal fragment of CD14 expressed on granulocytes and monocytes. During phagocytosis, PSEP is cleaved and released into the bloodstream. We previously reported that urinary PSEP (uPSEP) reflects inflammatory cell infiltration in the renal interstitium. Notably, elevated uPSEP levels are observed in tubulointerstitial nephritis (TIN) and T-cell-mediated rejection after kidney transplantation, both of which are characterized by interstitial inflammatory cell infiltration as the pathological feature. While uPSEP is expected to be clinically useful for detecting TIN, its utility in real-world clinical practice has not been fully validated. In the present study, we aimed to assess whether uPSEP provides clinically meaningful information to improve the prediction of TIN.

This cross-sectional study included patients registered in the Nagoya Kidney Disease Registry in 2020 and 2022. uPSEP levels were measured from urine samples collected and cryopreserved at the time of kidney biopsy. Patients were classified according to biopsy-proven diagnoses and the association with uPSEP was evaluated. In addition, before biopsy, attending physicians predicted diseases based on clinical history and laboratory test results (uPSEP not included). Patients were classified according to whether TIN was included in the predicted diagnoses, and this classification was compared with the biopsy-proven diagnoses.

A total of 543 patients (241 females and 302 males) with a median age of 59 years (IQR, 43–72 years) were included. Log-transformed urinary PSEP/creatinine (Cr) levels in patients with TIN were significantly higher than those in patients with focal segmental glomerulosclerosis, IgA nephropathy, minimal change nephrotic syndrome, and membranous nephropathy. ROC analysis for detecting TIN using urinary PSEP/Cr yielded an area under the curve (AUC) of 0.802 (95% CI, 0.741–0.862), with a cutoff value of 1526 ng/gCr. In a multivariable model adjusting for physician’s TIN prediction, the odds ratio for TIN diagnosis was 4.45 (95% CI, 1.84–10.70; P<0.001). In ROC analysis, the AUC of the extended model (based on physician’s TIN prediction and uPSEP) was significantly higher than that of the baseline model (based only on physician’s TIN prediction) by the DeLong test (0.928 vs 0.887; P=0.026). Reclassification analyses further demonstrated a total NRI of 0.61 (95% CI, 0.33–0.87) and an IDI of 0.043 (95% CI, 0.010–0.072); both confidence intervals did not cross zero.

Urinary presepsin is a promising biomarker that enhances diagnostic accuracy for TIN and may support clinical decision-making in routine practice.