Infrared Thermography as an Early Predictor of Mortality in a Rodent Model of Neonatal Endotoxic Sepsis

Abstract

Background Sepsis in the first four weeks of life kills over 1 million newborns annually. Early identification of neonatal sepsis has been shown to improve outcomes yet remains a major challenge in this population. Infrared thermography has been used in mass fever screening, tumor detection, anaphylaxis and in the evaluation of skin diseases but has not been studied extensively as a method for detecting the early thermoregulatory changes associated with sepsis. In this study, we utilize a neonatal rodent model of sepsis to evaluate the potential of infrared thermography to predict later mortality. Objective To use infrared thermography to determine whether early changes in skin surface temperature are associated with later mortality in a rodent model of neonatal sepsis. Design/Methods 10-day old rat pups (n=25) were isolated from their mothers and received an intraperitoneal injection of 0.1ug/kg of lipopolysaccharide (LPS). Previous studies from our lab have shown that this dose of LPS results in ~50% mortality 5-7 hours post-injection. Rats were placed on a heating pad (33.0°C) and in a thermoneutral neonatal isolette to maintain stable ambient air temperature (27.5°C). Thermal images were obtained using a hand-held thermal (FLIR) camera and temperatures of cranial (Tcr), scapula (Tsc), and rump (Tr) were evaluated using FLIR proprietary software. Images were taken at baseline (pre-injection) and at 10 to 30-minute intervals post-injection for 8 hours. Results LPS exposure resulted in 68% mortality and, therefore, rats were identified as survivors or non-survivors. Tcr pre-LPS injection was similar between groups. Tcr of surviving and non-surviving rats began to decrease at 60-70 minutes post-injection and reached a nadir at 150 minutes in both groups. In contrast to non-surviving rats, surviving animals had a subsequent recovery in Tcr (fig. 1a). Similar temperature profiles were observed in Tsc and Tr in respective surviving and non-surviving rats. Non-surviving rats began to expire at 5 hours and all had expired by 8 hours (fig. 1b). Conclusion(s) In this model of neonatal rodent endotoxic sepsis, survivors and non-survivors demonstrate different skin-surface temperature profiles over time. The rebound in Tcr post-LPS injection distinguishes survivors and non-survivors and may be a useful early identifier of lethal vulnerability to endotoxin.