DICKKOPF-3 (UDKK3): A NEW BIOMARKER FOR LONG-TERM CKD PROGRESSION AND MORTALITY?

Abstract Background and Aims Kidney fibrosis has been reported to be a key progression hallmark of chronic kidney disease (CKD). It seems likely that a precise biomarker of renal fibrosis extension would contribute to predict accurately the risk of a decline in glomerular filtration rate (eGFR). Previous studies have shown that the assessment of urinary Dickkopf-3 (uDKK3), a stress induced tubular epithelial-derived profibrotic glycoprotein, might be a potential tubulointerstitial fibrosis biomarker and might identify patients at short-term risk of eGFR loss. We aim to evaluate uDKK3 as a potential biomarker for long-term CKD progression in a cohort with various etiologies of CKD, and subsequently in an overt diabetic nephropathy cohort. We also tested the role of treatment with RAAS blockers on the uDKK3 levels and if the treatment could modify them. Method We prospectively studied two independent cohorts consisted of 356 patients with stage 2-3 CKD. Progreser cohort comprised 255 patients with heterogeneous etiologies of CKD and Pronedi cohort 101 patients with overt diabetic nephropathy. The primary outcome was the time to the first event of the composite endpoint (>50% increase in serum creatinine concentration, end-stage kidney disease [ESKD], or death). We divided patients into tertiles according to their baseline uDKK3 levels: less than 1092 pg/mg (Tertile 1, T1), between 1092-5050 pg/mg (Tertile 2, T2) and higher than 5050 pg/mg (Tertile 3, T3). We used the cut point of T3 as an exploratory cut-off. Cox regression models were used to adjust for potential effects of confounders or modifiers: age, gender, mean arterial pressure, body mass index (BMI), urine albumin/creatinine ratio, urine protein/creatinine ratio and serum creatinine. Mixed-effects models were adjusted to study longitudinal data. Results Progreser cohort and Pronedi cohort did not differ in their clinical baseline characteristics except in proteinuria. At baseline, uDKK3 levels were not different between different etiologies or both cohorts, median uDKK3 was 2199 (IQR: 658-7618) pg/mg in the Progreser cohort and 3041 (IQR: 653-9777) pg/mg in the Pronedi cohort (p:0.56). Median time of follow-up was 36 months. Forty-nine patients (19 %) in Progreser cohort and 31 patients (31%) in Pronedi cohort reached the primary composite outcome. Baseline uDKK3 was significantly higher in patients who reached primary outcome. In Cox multivariate model, after adjustment for potential confounders, the highest levels of uDKK3 were found to be an independent factor for renal progression in Progreser cohort (HR 1.83, CI95% 1.11-3.31) and in Pronedi cohort (HR 2.74, CI95% 1.09-6.98). Both in the Progreser and the Pronedi cohorts, uDKK3 levels above 5050 pg/mg, were associated with a lower eGFR, higher proteinuria and were independently associated with a worse renal survival. uDKK3 gradually increased in the following months, especially in patients with higher proteinuria. Treatment with RAAS-blockers did not modify uDKK3 after 4 or 12 months of treatment. Conclusion In this study, uDKK3 ratio identified patients at high risk for long-term kidney disease progression regardless of the etiology of CKD. The hypothesis was generated in a cohort of patients with CKD of heterogeneous etiologies and was further confirmed in a cohort with overt diabetic nephropathy. The predictive role of uDKK3 persisted after adjusting by eGFR and proteinuria. uDKK3 is the first non-invasive biomarker of renal fibrosis and might serve as a useful biomarker for kidney disease progression. Therefore uDKK3 could be used by clinicians to optimize staging for renal progression and monitor therapeutic efficacy of different measures to halt CKD progression.