Abstract: SA-PO0509
Discriminative Performance of B-Type Natriuretic Peptide (BNP), Carbohydrate Antigen-125 (CA-125), Bioimpedance, and Venous Excess Ultrasound for Detecting Congestion in Nondialysis-Dependent CKD Stage 5: A Two-Center Pilot Study
Session Information
- CKD: Clinical - Therapies, CKM Health, Special Populations, and More
October 24, 2026 | Location: Exhibit Hall A, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: CKD (Non-Dialysis)
- 2202 CKD (Non-Dialysis): Clinical, Outcomes, and Trials
Author
- Ramirez Badillo, Erick Edgar, Hospital General de Ticomàn, Mexico City, Mexico
Background
Volume overload is common in advanced chronic kidney disease (CKD), but clinical assessment alone may underestimate venous and tissue congestion. BNP, CA-125, bioimpedance-derived fluid ratios, and Venous Excess Ultrasound Score (VExUS) may provide complementary information in non-dialysis CKD G5.
Methods
We conducted a two-center prospective observational pilot study including 47 adults with non-dialysis CKD G5 from two general hospitals in Mexico City. Patients were classified according to clinical congestion. BNP, CA-125, bioimpedance parameters, and VExUS were compared between groups. A combined logistic regression model was built using BNP, CA-125, ECW/TBW, and VExUS.
Results
Thirty-two patients had clinical congestion and 15 had no clinical congestion. Baseline age, sex, weight, diabetes, hypertension, loop diuretic use, albumin, and creatinine were generally similar between groups. Patients with clinical congestion had higher BNP levels [1169 (990-1638) vs. 889 (789-1050) pg/mL; p=0.007] and higher CA-125 levels [100 (89-110) vs. 79 (76-95) U/mL; p=0.002]. VExUS scores were higher in the congestion group (p=0.03), and ECW/TBW was also increased (0.40±0.04 vs. 0.33±0.18; p=0.01). The combined model showed high discriminatory performance for identifying clinical congestion: AUC 0.94, sensitivity 0.90, and specificity 1.00. In multivariable analysis, ECW/TBW remained significantly associated with clinical congestion (p=0.014), while VExUS showed a statistical trend (p=0.051).
Conclusion
In non-dialysis CKD G5, BNP, CA-125, VExUS, and bioimpedance identified a multimodal congestion phenotype. A combined model showed strong discriminatory performance for clinical congestion, supporting multimodal congestion assessment in advanced CKD and warranting larger validation studies.