Abstract: TH-PO234
Simultaneous Dosing of Iron and ESA Using Model Predictive Control (MPC)
Session Information
- Anemia and Iron Metabolism: Clinical
October 25, 2018 | Location: Exhibit Hall, San Diego Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Anemia and Iron Metabolism
- 202 Anemia and Iron Metabolism: Clinical
Authors
- Brier, Michael E., University of Louisville, Louisville, Kentucky, United States
- Gaweda, Adam E., University of Louisville, Louisville, Kentucky, United States
Background
Advances have been made to assist the dosing of erythropoietic stimulating agents (ESA) beyond the use of a "paper protocol". These methods have been dependent on either assuming that iron stores are replete or that a separate iron dosing protocol is followed. Since ESA response is influenced by the iron status and changes as patients become iron replete, it would be beneficial to perform these dosing procedures in parallel. We tested the hypothesis that we could develop a MPC algorithm for the simultaneous dosing of both ESA and iron.
Methods
ESA dose prediction was performed using our previously published work (J Am Soc Nephrol 25:159-66 2014). Dose determination is based on an estimate of the red blood cell life span, sensitivity to the ESA and dose of ESA. Further, a close-loop design was used to "feed back" the error between observed and desired Hb. A second MPC model was developed for Iron dosing that determines the dose of iron based on baseline ferritin, iron and Tsat sensitivity to iron administration based on our previously published model ( Clin J Am Soc Nephrol 5:576-81, 2010). The current model was developed in Matlab/Simulink. We simulated several iron deficient states with ferritin between 80 and 120 with ESA sensitivity values from 0.1 to 2.0. Simulations were performed to determine the impact of iron supplementation on ESA dose.
Results
The results of the simulations are shown in the following table where the effect of simultaneous iron dosing on total ESA dose is shown. Total iron administered was independent of ESA dosing in this simulation and resulted in 10,000, 8,000 and 4,000 mg over 2 years for ferritins of 80, 100, and 120, respectively.
Conclusion
We simulated the effect of concurrent ESA and iron dosing in ESRD using advanced computational approach. The algorithm optimizes both, ESA and iron dose. Further work needs to be performed on the dynamics of the response during all phases of iron repletion and in the face of elevated ferritin levels due to inflammation.
Total ESA dosed over 2 years when model preforms iron dosing (Iron Dosed) or does not (No Iron Dosed)
| Iron Dosed | Ferr=80 | Ferr=100 | Ferr=120 | No Iron Dosed | Ferr=80 | Ferr=100 | Ferr=120 |
| K=0.1 | 6455 | 6975 | 6925 | K=0.1 | 8435 | 7830 | 7240 |
| K=0.5 | 1805 | 2000 | 1905 | K=0.5 | 5715 | 2240 | 2000 |
| K=1.0 | 935 | 1030 | 980 | K=1.0 | 3060 | 1170 | 1045 |
| K=2.0 | 475 | 550 | 495 | K=2.0 | 1565 | 595 | 535 |
Funding
- NIDDK Support