#2956 NT-proBNP and Serum CA-125 in peritoneal dialysis: developing a decision algorithm for fluid overload assessment

Abstract Background and Aims Fluid overload (FO) in peritoneal dialysis (PD) increases cardiovascular morbidity and mortality, but clinical signs often appear late, making early detection challenging. NT-proBNP is an established marker of hypervolemia, while serum CA-125 may reflect extracellular volume. This study proposes a biomarker-based algorithm to enhance FO diagnosis and management in PD. Method We conducted a retrospective analysis of peritoneal dialysis (PD) patients assessed between July 2023 and December 2024 at a Portuguese PD center. On the same day, patients underwent adequacy testing, bioimpedance analysis, and biomarker quantification (NT-proBNP and serum CA-125). Fluid overload (FO) was defined as overhydration (OH) ≥5% of total body water (TBW). Statistical analyses were performed using SPSS v29, applying parametric tests when appropriate. To evaluate diagnostic performance, we calculated sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and the area under the curve (AUC) from receiver operating characteristic (ROC) curves. Results A total of 88 measurements were collected from 46 chronic PD patients: 38.80% (n = 16) were female, with a mean age at PD initiation of 61.5 ± 10.6 years (PD duration 15.00 ± 10.60 months) and automated PD was used by 52.17% (n = 24). Serum CA-125 was 11.85 ± 14.28 U/mL, NT-proBNP 1091.0 ± 3055.0 U/mL, OH 1.76 ± 1.90 L, and %OH 9.55 ± 13.30. FO (%OH ≥5%) was present in 71.6% (n = 63). Edema was detected in 23.9% of the general population, and 33.9% of FO cases, showing low sensivity (33.0%) and NPV (37.9%). No significant correlation was found between serum CA-125 and OH (r = 0.154, P = 0.151) while its diagnostic performance for FO detection showed an AUC of 0.597. However, a significant association with FO was observed (χ² = 5.730, P = 0.017; Fisher's exact test P = 0.021). NT-proBNP (with cutoff 801 pg/mL) showed 84.1% sensitivity, 24.0% specificity and an AUC 0.857. In patients with NT-proBNP ≥801 pg/mL, serum CA-125 (cutoff 15.5 U/mL) had 43.3% sensitivity, but 100% specificity and PPV. Conclusion This study underscores the relevance of a biomarker-based algorithm in refining the diagnostic approach to fluid overload in peritoneal dialysis. The combination of NT-proBNP ≥801 pg/mL and serum CA-125 >15.5 U/mL demonstrates high specificity, strongly supporting the diagnosis of fluid overload in this population. While clinical assessment remains fundamental in volume evaluation, this algorithm provides an objective and reproducible tool to support decision-making, particularly in scenarios of diagnostic uncertainty or when direct patient assessment is not possible.