The Engineering of Anemia Management Protocols: Modeling, Control Design, and Clinical Results

End stage renal disease (ESRD) patients often suffer from anemia due to endogenously produced erythropoietin (EPO) that is inappropriately low for the level of anemia, reduced red blood cell lifespan, EPO resistance, chronic inflammation, and acute loss of blood. The discovery of recombinant human EPO (rHuEPO) over 30 years ago has shifted the treatment of anemia for patients on dialysis from blood transfusions to rHuEPO therapy. Anemia management protocols (AMPs) comprise a set of experience-based rules for weekly-to-monthly titration of rHuEPO doses based on hemoglobin (Hb) measurements. Unfortunately, AMPs have been designed using open-loop ideas.
Clinical studies suggested that current AMPs fail to stabilize Hb levels to target, and may lead to undesired behavior, such as Hb cycling. The FDA recently issued a warning (June 24, 2011) "… Procrit, Epogen, and Aranesp are used in patients with chronic kidney disease (CKD) may put them at risk of heart attack, stroke, and death…" and in information for healthcare professionals who treat patients with CKD suggested that "No clinical trial to date has identified a hemoglobin target level, ESA dose, or dosing strategy that does not increase these risks."
This seminar introduces a "novel' approach for personalized medicine: designing patient-specific AMPs based on feedback control principles. To this end, we present a physiologically-relevant erythropoiesis model validated based on a 15-month long retrospective study of 44 ESRD patients from one dialysis facility. We motivate the approximation of the nonlinear dynamics in order to facilitate AMP design and discuss aspects of the robust control design. We conclude with preliminary results from a pilot clinical study.

Yossi Chait is a Professor in the department of Mechanical & Industrial Engineering at the University of Massachusetts. He graduated from OSU with BS (1982), MSU with MS (1984), and PhD (1988) degrees in Mechanical Engineering. Since 1988, his research activities have been continuously funded by NSF covering diverse topics including the Quantitative Feedback Theory (QFT), robust multivariable control, reset control, Internet congestion control, modeling of the mammalian master clock, and dynamics of the human thyroid. His recent research interests lie at the interface of engineering and medicine. He has consulted internationally in the area of robust control, and is a co-developer of the QFT Control Design MATLAB Toolbox. He held visiting professor positions at the Swiss Federal Institute of Technology, Switzerland; the Technion, Israel, and Delft University of Technology, The Netherlands. He is a fellow of ASME.