Blood-induced interference of glucose sensor function in vitro: implications for in vivo sensor function.

BACKGROUND Although tissue hemorrhages, with resulting blood clots, are associated with glucose sensor implantation, virtually nothing known is about the impact of red blood cells and red blood cell clots on sensor function in vitro or in vivo. In these studies, we tested the hypothesis that blood can directly interfere with glucose sensor function in vitro. METHODS To test this hypothesis, heparinized human whole blood (HWB) and nonheparinized human whole blood (WB) were obtained from normal individuals. Aliquots of HWB and WB samples were also fractionated into plasma, serum, and total leukocyte (TL) components. Resulting HWB, WB, and WB components were incubated in vitro with an amperometric glucose sensor for 24 hours at 37 degrees C. During incubation, blood glucose levels were determined periodically using a glucose monitor, and glucose sensor function (GSF) was monitored continuously as nanoampere output. RESULTS Heparinized human whole blood had no significant effect on GSF in vitro, nor did TL, serum, or plasmaderived clots from WB. Sensors incubated with WB displayed a rapid signal loss associated with clot formation at 37 degrees C. The half-life was 0.8 +/- 0.2 hours (n = 16) for sensors incubated with WB compared to 3.2 +/- 0.5 (n = 12) for sensors incubated with HWB with a blood glucose level of approximately 100 mg/dl. CONCLUSIONS These studies demonstrated that human whole blood interfered with GSF in vitro. These studies further demonstrated that this interference was related to blood clot formation, as HWB, serum, plasma-derived clots, or TL did not interfere with GSF in vitro in the same way that WB did. These in vitro studies supported the concept that the formation of blood clots at sites of glucose sensor implantation could have a negative impact on GSF in vivo.