Mechanistic simulations of inflammation: current state and future prospects

Abstract

Inflammation is a normal, robust physiological process. It can also be viewed as a complex system that senses and attempts to resolve homeostatic perturbations initiated from within the body (for example, in autoimmune disease) or from the outside (for example, in infections). Virtually all acute and chronic diseases are either driven or modulated by inflammation. The complex interplay between beneficial and harmful arms of the inflammatory response may underlie the lack of fully effective therapies for many diseases. Mathematical modeling is emerging as a frontline tool for understanding the complexity of the inflammatory response. A series of articles in this issue highlights various modeling approaches to inflammation in the larger context of health and disease, from intracellular signaling to whole-animal physiology. Here we discuss the state of this emerging field. We note several common features of inflammation models, as well as challenges and prospects for future studies.

Figure 1. The role of alarm/danger signals in inflammation, distilled for mathematical modeling purposes

Solid arrow: induction; dashed line: suppression. An initiating stimulus (e.g., pathogen (Panel A) or trauma (Panel B)) stimulates both pro- and anti-inflammatory pathways. In the setting of infection, pro-inflammatory agents (e.g., TNF) cause tissue damage/dysfunction, which in turn stimulates further inflammation (e.g., through the release of “danger signals”). In the case of trauma, tissue damage occurs immediately and further simulates inflammation. Anti-inflammatory agents (e.g., TGF-β1) both suppress inflammation and stimulate healing.

Figure 2. Generic branched pathway used to illustrate the personalization of an average pathway model

Equations are formulated in the manner of Biochemical Systems Theory. The given initial values correspond to the stable steady state, which is obtained with parameter settings as indicated. To personalize the model, one or more of the parameter values are altered and the responses are checked.

Figure 3. Results of select simulations of the pathway model in Figure 2

Y₁, …,Y₄ denote X₁, …, X₄ divided by their “normal” steady-state initial values, respectively. Y₅ denotes the ratio X₃/X₄. A: Increase of p₁ by 50%. B: Doubling of p₂. C: Increase in strength of inhibition (p₃) from −0.5 to −1. D: Doubling of p₄.

Figure 4. The “fragmented continuum” from pre-clinical studies to ultimate clinical utility

The current paradigm of healthcare delivery can be thought of as starting with in vitro and in vivo pre-clinical studies. Candidate therapies from these efforts are tested in clinical trials, yet the ultimate efficacy and safety of these therapies is revealed only via the process of in-hospital care over an often prolonged timeframe that may involve chronic and/or rehabilitative care. These domains are generally separate, and there is currently no framework under which these domains can be linked. Both the United States Food and Drug Administration, in its “Critical Path” document [81] and the United States National Institutes of Health, in its “Roadmap” statement [82] have called for the use of computational simulations to bridge this “bench to bedside” gap. Translational systems biology focused on the inflammatory response may serve this role through the creation of computational simulations that span pre-clinical studies, in silico clinical trials, patient diagnostics and other aspects of in-hospital care, and ultimately long-term care and rehabilitation.