Abstract: TH-PO0855
Metabolite Removal by High-Flux Hemodialysis (HFX-HD) and High-Volume Hemodiafiltration (hvHDF): A Cross-Over Study
Session Information
- Dialysis: Prescription, Physiology, and Treatment Optimization
October 22, 2026 | Location: Exhibit Hall A, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Dialysis
- 801 Dialysis: Hemodialysis and Frequent Dialysis
Authors
- Wang, Xiaoling, Fresenius Medical Care, New York, New York, United States
- Grobe, Nadja, Fresenius Medical Care, New York, New York, United States
- Lopez-Gil, Salvador, Instituto Nacional de Cardiologia Ignacio Chavez, Mexico City, Mexico
- Madero, Magdalena, Instituto Nacional de Cardiologia Ignacio Chavez, Mexico City, Mexico
- Kotanko, Peter, Renal Research Institute, New York, New York, United States
- Stauss-Grabo, Manuela, Fresenius Medical Care Deutschland GmbH, Bad Homburg, HE, Germany
Background
Comparisons of solute removal during HFX-HD and hvHDF are mostly limited to a small set of solutes and thus fail to capture the diversity of metabolites affected by dialysis. Untargeted metabolomics enables unbiased characterization of dialysis-associated changes in blood metabolites. We performed untargeted liquid chromatography–mass spectrometry (LC-MS) analyses of plasma samples in patients undergoing HFX-HD and hvHDF.
Methods
In a cross-over study, 12 HD patients underwent a week of hvHDF (3x) followed by a week of HFX-HD (3x) using the FX CorDiax 120 dialyzer (Armenta-Alvarez A, et al. Kidney360 4: 1095, 2023). Pre- and post-HD plasma samples were collected from two sessions per modality, one day apart.
LC-MS using a Phenomenex Luna C18 column was conducted on an Agilent Q-TOF 6545 in positive ionization mode. Metabolites were identified using an in-house library. Reduction ratios (RR) were calculated from their MS intensities (normalized to D5-tryptophan) and corrected for hemoconcentration. RR coefficient of variation (CV) per metabolite were calculated across 24 RRs per modality.
Results
Among 91 identified metabolites, 35 showed CV >100% (e.g., tryptophan, CMPF) and were excluded from further analysis. Of the remaining 56 metabolites, median RR was significantly higher with hvHDF (74%; range 36–90%) than HFX-HD (67%; 28–88%; p<0.0001, paired t-test).
At the individual metabolite level, hvHDF showed higher RRs for 48 of 56 metabolites, with 15 reaching significance, including uremic solutes, amino acids and derivatives (p<0.05, paired t-test) (Figure 1A). The RR of ADMA/SDMA (Figure 1B) and creatinine (Figure 1C) was most pronounced (p=0.001). Eight metabolites had non-significantly lower RR with hvHDF.
Conclusion
HvHDF delivered higher RRs than HFX-HD for most metabolites. These findings extend beyond traditional small-molecule markers (e.g., urea), highlight the utility of metabolomics for characterizing modality-specific clearance profiles, and support investigation in larger cohorts.
Funding
- Commercial Support – Fresenius Medical Care