An update on progress with dual hormone (insulin and glucagon) closed loop pumps.
Initial trials in 2008-2009 used 2 pumps attached to patient with glucose sampling every 5 minutes for 27 hours.
Infusions were controlled via an algorithm on a laptop. 3 large unannounced standardized meals were given with no rescue snacks. There were significant rises after the meals but mean glucose was well within range and hypos were not a problem. The data suggests that there is a huge variability between subjects in insulin lispro absorption, but also that there is up to 50% variability in insulin absorption in individual subjects. It is noteworthy that there was no “sensor lag” in these studies as these were all based on real time venous glucose. In the same trial the accuracy and reliability of a variety of glucose sensors was also investigated and the Navigator was found to be better in both domains.
A 2-day feasibility trial was repeated using the Navigator as the glucose sensing mechanism with the added complexity of unannounced exercise (the other sensors were also attached and investigated although Navigator drove the system). What was shown was thatplasma glucagon was on average at the very low end of the normal physiological range implying that insulin-dosing calculation was good. Exercise induced fall in glucose was generally controlled well with the dual hormone closed loop. Next study will run in 2011-2012 and run a mobile dual hormone pump which will receive CGM every 5 minutes and study a 5 day period with the patient ambulant. The portable dual hormone pump is in development based on iPod Tough as the controller linked to insulin Omnipod and should be ready by late 2011. The study subjects will be chaperoned by a CGM nurse but be free to wander around the MGH campus to mimic real life. More information is available at: www.artificalpancreas.org
No comments:
Post a Comment