An engineering control system paradigm for quantitative understanding of hemostasis
Date
2016
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Hemostasis is an essential physiological process whose proper function relies
on the rapid formation of a blood clot in order to minimize blood loss after a vascular
injury. To that end, effective hemostasis is a result of the regulation of a delicate
balance between two opposite pathological conditions: hemophilia and thrombophilia.
When the hemostatic response is too weak, as in the case of hemophilia, blood clot
formation is delayed, causing excessive bleeding; while overactive hemostasis could
lead to thrombosis where a blood vessel is completely occluded by a blood clot, a
common manifestation of thrombophilia. Understanding the hemostatic process is of
significant importance because of the acute danger associated with hemostatic
disorders. In particular, they can lead to heart attack and stroke, two major causes of
death in the world. The complete process of hemostasis consists of numerous components
organized into two mutually interacting sub-processes known as primary and
secondary hemostasis. To date, quantitative modeling studies of hemostasis have been
restricted to a few individual components in isolation. Since in actuality these
components never function in isolation, such studies provide only limited
understanding of the individual components in question, not of the entire complex
process. Obtaining a useful, holistic and high fidelity representation of the complete
system requires a more comprehensive approach that adequately captures the full
characteristics of all components and their interactions. However, because of intrinsic system complexity, a traditional modeling approach will generate an overly complex,
computationally intractable, and hard-to-validate model.
In this dissertation, in recognition of the defining characteristic of hemostasis
as an automatic physiological control system, an engineering control system
framework was used to provide a modular structure for organizing the overwhelming
amount of information in hemostasis. In the case of primary hemostasis, mathematical
expressions, specifically algebraic equations and ordinary differential equations, were
developed to represent the biological events associated with each primary hemostasis
module: biological equivalents of a sensor, controller, and actuator. These modules
were then connected to form a mathematical model of primary hemostasis.
An existing ordinary differential equation model of secondary hemostasis was
updated and adapted to include additional recent biological findings on secondary
hemostasis. The primary and secondary hemostasis models were modified to
incorporate the interactions between the two subsystems before they were connected
to form a comprehensive mathematical representation of hemostasis. This
comprehensive model was used to simulate various physiological and pathological test
cases, generating results consistent with experimental observations and known
characteristics. We subsequently used the comprehensive model to study select disease
conditions representative of one of the four categories of hemostatic disorders:
hyperactive primary hemostasis, deficient primary hemostasis, hyperactive secondary
hemostasis, and deficient secondary hemostasis. This exercise enabled us to identify
new potential treatment targets that we recommend for future experimental
investigations, but which include targets that are already being used as the basis for current clinical treatment strategies for hemophilia A and hemophilia B, demonstrating
the effectiveness and predictive power of the comprehensive model.
Finally, we carried out an experimental and simulation case study involving an
actual patient with low platelet count due to immune thrombocytopenic purpura, a
pathology that represents a deficiency in the primary hemostasis controller and the
actuator. Using an existing model customized to describe the drug dose-response
relationship in the patient, a weekly treatment frequency was found to be capable of
achieving a stable platelet count, while biweekly treatment regimen always resulted in
oscillations due to the pharmacological properties of the drug.