Trauma in silico: Individual-specific mathematical models and virtual clinical populations

See allHide authors and affiliations

Science Translational Medicine  29 Apr 2015:
Vol. 7, Issue 285, pp. 285ra61
DOI: 10.1126/scitranslmed.aaa3636

A virtual large sample size

Severe trauma or bleeding evokes an all-hands-on-deck immune response. When properly orchestrated, the myriad cytokines and peptides help to heal the patient. However, this process can easily go awry, and administering the right drug to compensate is a challenge. Brown and colleagues mathematically modeled the complicated immune responses from the data of 33 blunt trauma patients and then generated a larger cohort of 10,000 virtual trauma patients. This large virtual cohort predicted the reactions of smaller validation cohorts, but the surprise was that understanding the response details in a single patient did not predict how the population would act. The author’s virtual clinical trial indicated that inhibition of interleukin-6 (IL-6) produced a small survival benefit, whereas IL-1β inhibition did not help much and tumor necrosis factor–α made things worse.

View Full Text

Stay Connected to Science Translational Medicine