Mechanisms of hemoglobin cycling in anemia patients treated with erythropoiesis-stimulating agents

Abstract

Patients with renal anemia are frequently treated with erythropoiesis-stimulating agents (ESAs), which are dynamically dosed in order to stabilize blood hemoglobin levels within a specified target range. During typical ESA treatments, a fraction of patients experience hemoglobin ‘cycling’ periods during which hemoglobin levels periodically over- and undershoot the target range. Here we report a specific mechanism of hemoglobin cycling, whereby cycles emerge from the patient’s delayed physiological response to ESAs and concurrent ESA dose adjustments. We introduce a minimal theoretical model that can explain dynamic hallmarks of observed hemoglobin cycling events in clinical time series and elucidates how physiological factors (such as red blood cell lifespan and ESA responsiveness) and treatment-related factors (such as dosing schemes) affect cycling. These results show that in general, hemoglobin cycling cannot be attributed to patient physiology or ESA treatment alone but emerges through an interplay of both, with consequences for the design of ESA treatment strategies. Author summary Patients with kidney failure often suffer from anemia, a disease characterized by a shortage of red blood cells. In clinical practice, this shortage is assessed through measurements of blood hemoglobin concentration. Anemia is typically treated using drugs that mimic erythropoietin, the body’s own hormone regulating red blood cell generation in healthy subjects. During anemia treatment, a widespread phenomenon termed “hemoglobin cycling” may emerge, whereby blood hemoglobin levels oscillate with high amplitude around a target concentration which the therapy aims to achieve. This is an undesired condition, which also necessitates frequent adjustment of the drug dose. The causes of hemoglobin cycling are not understood on a fundamental level. We developed a simple mathematical model that describes physiological and treatment-related effects. In this model, cycling emerges because of long delays between drug administration and their effect on hemoglobin concentration and vice versa, leading to a negative-feedback loop that reinforces oscillations. We use our model to develop statistical indicators of this type of hemoglobin cycling in clinical data and analyze how physiological and treatment-related factors may contribute to the appearance of cycling. These results can help to understand how to modify existing treatment schemes to avoid hemoglobin cycling in the first place.