Mucositis in cancer patients: Prototypic semi-mechanistic kinetic model

Abstract

19617 Background: Recent research advances have helped (i) define pathobiology of alimentary tract mucosal injury secondary to cancer therapy and (ii) link molecular mechanisms with clinically important outcomes. Recently-developed computational biology modeling may further enhance these advances. Semi-mechanistic (SM) modeling allows one to approach quantitative analysis of a biochemical system that is incompletely determined. In this study, data from sequential oral mucosal biopsies in 3 patients developing oral mucositis secondary to hematopoietic stem cell transplantation (HSCT) conditioning were utilized to establish a prototypic computational model for this toxicity. Methods: Plasma and oral mucosal biopsy specimens were obtained from 3 autologous HSCT patients before and after administration of conditioning chemotherapy: Day -10, +10, +28 and +100; Day 0 was day of transplant. Full-thickness tissue samples were measured by RT- PCR for COX-1, COX-2, IL-1β and TNF-a. Plasma samples were measured by ELISA for PGE2 and PGI2, markers of COX-2 activity. The SM model was implemented as a system of 6 ordinary differential equations with 15 parameters. Parameter estimation and simulations were conducted based on experimental results, using a combination of Mathematica, Berkeley Madonna and Virtual Cell software packages. Results: The SM model captured the behavior of COX-1, IL-1β and PGE2 dynamics, predicting an exponential decay for each of these species. Half-lives relative to average steady-state values were found to be 9.7 days, 8.7 days and 9.3 days for COX-1, IL-1β and PGE2 respectively. Correlation ratios for each of these species were calculated to be 0.62, 0.61 and 0.90 respectively. Conclusions: This prototypic model provides a basis for development of a detailed mathematical model for quantifying relevant components of the mucositis pathway. This combination of modeling and experiment could also identify gaps in the pathway that would be important targets for new hypotheses, including possible feedback mechanisms relative to inflammatory cytokines. No significant financial relationships to disclose.