Biomedical Engineering Reference
In-Depth Information
tactics such as the adoption of changing "disguises" and the secretion of mole-
cules that interfere with immune system communication. The trillion-cell im-
mune "army" has no general; signals from the brain modulate the action of the
immune system to some extent but do not control its detailed performance.
In this essay I shall survey some of my attempts, with colleagues, to apply
theoretical investigations in order to better understand what controls the immune
system, with emphasis on ideas that might be of interest to physicians. A unify-
ing concept is the role of information, defined simply as "knowing what's going
on," in guiding immune system performance and medical interventions.
Before beginning the main body of this chapter, let me note that a natural
analogy of the principal task facing the immune system is the task facing the
physician in curing disease, or at least blunting its impact. I have argued that a
better analogy is with a public health system (1). Such a system may well have
to contend with several diseases at once, selected from a wide spectrum of pos-
sibilities. Triage is a key problem: how to allocate the available resources ap-
propriately to meet a changing panoply of challenges. Feedbacks are essential to
the proper operation of a public health system. As information flows in concern-
ing the performance of the system, shifts in resource allocation can be made to
better contend with changing circumstances.
2.
THERAPY AS AN INFORMATION
-
YIELDING PERTURBATION
Shochat and colleagues (2) constructed a detailed model (about 25 differen-
tial equations, almost 40 parameters) for multiple hematopoietic stages, starting
from pluripotent stem cells and ending in the generation of monocytes and po-
lymorphonucleic cells (PMN). The role of granulocyte colony-stimulating factor
(G-CSF) was taken into account. Values for all but five parameters could be
found in the literature. The remaining parameters were estimated by requiring
model predictions to match data on blood counts of monocytes and PMN. Of
crucial importance were measurements of shifting monocyte and PMN levels by
frequent blood counts routinely taken in patients during high-dose chemotherapy
followed by autologous peripheral blood stem cell transplantation and injection
of G-CSF. (The lymphoid lineage was neglected in the model, since appropriate
data were not available.) By choosing parameters to minimize the least-squared
difference between the actual blood counts and the theoretical predictions, Sho-
chat et al. produced estimates of parameters that cannot as yet be measured di-
rectly in humans, notably stem cell density and division rate. The calibrated
model was used to explore alternative strategies to reduce the post-transplant
nadir period of deficient immunity. One somewhat surprising suggestion is that
G-CSF injections before day 5 post-transplant may have little influence in di-
minishing the nadir period.
