Agent-based simulation of acute phonotrauma in cases with various cytokine profiles
Section snippets
Background
A common source of vocal fold injury is phonotrauma, which is a complex process accompanied by both acute and chronic inflammation. To overcome the prohibitively large number of experiments needed to characterize all possible combinations of various subject conditions and specific treatment doses in human clinical studies we are conducting, we propose using biosimulation as an adjunct. We hypothesized that agent-based modeling (ABM) of inflammation and healing will predict differentiated acute
Methods
We used our previously reported ABM of acute vocal fold inflammation, which is based on data from the wound healing literature and from inflammatory mediators in human laryngeal secretions. The ABM consists of cells (neutrophil, macrophage, and fibroblast), inflammatory mediators (IL-1β, TNF-α, transforming growth factor–β1, and IL-10), extracellular matrix (collagen), and a tissue damage function that is induced by and, in turn, induces further inflammation. We performed 5 runs of the ABM up
Results
Table 1 depicts the levels of IL-1β and TNF-α in our human phonotrauma study and shows that subjects with high baseline values of IL-1β and TNF-α did not increase the levels of these mediators immediately after induced phonotrauma but rebounded at 4 or 24 hours post trauma. These results, on which the ABM was not calibrated, were surprisingly predicted by our ABM qualitatively (Fig. 1). Analysis of our simulations suggested that this proinflammatory cytokine rebound was temporally associated
Conclusions
Our ABM reproduced basic cellular and molecular patterns of the inflammatory and wound healing responses in acute phonotrauma and predicted that long-term wound healing outcomes could be dramatically differentiated by defined variations in the initial magnitude of biomechanical stress as well as pre-existing inflammatory mediator profile. Systems level modeling may shed light on the basic biology of wound healing in the vocal folds and may ultimately allow for in silico testing of treatment