Abstract: TH-PO132
Improving Treatment of CKD-Mineral Bone Disorder (CKD-MBD) Through the Incorporation of Agatston Scores
Session Information
- Bone and Mineral Metabolism: CKD-MBD Updates
November 02, 2023 | Location: Exhibit Hall, Pennsylvania Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Bone and Mineral Metabolism
- 502 Bone and Mineral Metabolism: Clinical
Authors
- Gaweda, Adam E., University of Louisville School of Medicine, Louisville, Kentucky, United States
- Brier, Michael E., University of Louisville School of Medicine, Louisville, Kentucky, United States
- Lederer, Eleanor D., VA North Texas Health Care System, Dallas, Texas, United States
Background
KDIGO targets for CKD-MBD for Ca, P, and PTH are surrogates for the real damage that takes place in bone and the cardiovascular (CV) system. Vascular calcification (VC) correlates strongly with CV disease and death in individuals with CKD. We have modeled CKD-MBD through a series of mass balance equations describing movement of mineral between bone and soft tissue. We hypothesize that incorporation of measures and biomarkers of VC into a quantitative systems pharmacology model optimized using reinforcement learning will improve therapy by identifying alternative targets.
Methods
Data were abstracted from the Chronic Renal Insufficiency Cohort (CRIC) consisting of 5499 individual subjects with a mean CKD EPI eGFR of 47.7 ml/min/1.73m2 and a range between 3.0 and 126. Total and site specific Agatston scores were calculated and used for analysis. A total of 3230 Agatston scores were reported with a mean of 350 ± 767 and a range of 0 to 8247. A Cox Proportional hazard analysis was performed on CV death comparing Agatston score, site specific scores, and measured markers of inflammation (CaxP, Neutrophil/Lymphocyte Ratio (NLR), Fatty Acid Binding Protein (FABP), N-Acetyl-Beta-Glucosamide (NABG), Beta-2 Microglobulin (B2M), Beta-Trace Protein (BTP), Kidney Injury Molecule 1 (KIM1).
Results
The effects of inflammation markers on Total Agatston score and site-specific scores were not different, so only Total Agatston score will be reported. The results of the Cox model are shown in Table 1 following modeling with backwards elimination. All factors entered the model as an increased risk of CV death except for Beta-Trace Protein.
Conclusion
Substituting Agatston score for Ca net balance and modeling inflammatory intermediates NLR, B2M, BTP and FABP will optimize the CKD-MBD model. Leveraging artificial intelligence in the form of reinforcement learning will yield new pathways for treatment by targeting more proximal surrogates of disease.
| B | Sig | Exp(B) | |
| Agatston Score | 0.000 | <0.001 | 1.000 |
| CaxP Product | 0.023 | <0.001 | 1.023 |
| Neutrophil/Lymphocyte | 0.095 | <0.001 | 1.100 |
| Urine N-Acetyl-Beta Glucosaminidase | 0.052 | <0.001 | 1.054 |
| β-2 Microgloblin | 0.183 | <0.001 | 1.201 |
| Beta Trace Protein | -0.164 | <0.001 | 0.849 |
| Liver Fatty Acid-Binding Protein | 0.001 | <0.001 | 1.001 |
Funding
- Veterans Affairs Support