The concept of 'new technologies' for type 1 diabetes has expanded in recent years at a rate that some might consider comparable to 'Moore’s Law', and the sheer number of new technologies entering into the type 1 diabetes marketplace is also growing at a remarkable rate. From the patient’s perspective, this is not only exciting but can lead to a sense of optimism ('things will get better'). Technologies that today are growing commonplace (e.g. insulin pumps, continuous glucose monitoring devices, rapid HbA1c monitoring, etc.) were new technologies not that long ago. Indeed, it could be argued that the major advances in type 1 diabetes care made within the last quarter of a century have come from technology rather than biology. At the same time, not all new technologies succeed (e.g. the Glucowatch), regardless of their purported promise. Both type 1 diabetes patients and their healthcare providers will soon see a series of further new technologies whose basis is tied to the notion of improving the lives of those living with the disease.
Despite dramatic improvements in therapies for type 1 diabetes including improved insulins, more accurate monitors, insulin pumps – and more – most patients spend a majority of their 24 hour day outside of the ideal glycaemic zone, including significant time being hypoglycaemic.
This information has directed major energies into the development of the 'artificial pancreas'. This device (or, more appropriately, series of devices) would sense glucose and secrete an appropriate quantity of insulin. This notion is not a new one, but has been 'just around the corner' for many years.
Considerable engineering, technological, and regulatory hurdles have (thus far) stood in the way of its introduction. This said, it does appear that introduction of such a system is now feasible. While certainly not a 'cure' for type 1 diabetes as some have claimed, it does have the theoretical potential to change the landscape of type 1 diabetes treatment. The current state of affairs for the three main components (insulin pump, continuous glucose monitoring (CGM) and control algorithm) will be discussed further.
The insulin pump
Since their introduction, insulin pumps have proved to be effective at improving the management of type 1 diabetes, leading not only to a reduced risk of hypoglycemia but (often, but not invariably) to a drop in HbA1c. At present, there are a number of insulin pumps on the market (e.g. Medtronic Paradigm Revel, Tandem t:slim, Animas Vibe and others) as well as disposable patch pumps (e.g. the Omnipod from Insulet).
Insulin pump with infusion set
Standard pumps include a pump and insulin reservoir attached via tubing to the infusion site, whereas patch pumps are tubeless; designed with a remote controller (wireless) and an insulin filled unit that can be worn. Timing represents a major issue for insulin pumps since the action of insulin delivered under the skin is altered relative to that akin to pancreatic production. Nonetheless, their technological state now approaches that required for deployment as part of an closed loop system.
Continuous glucose monitors
For more than 30 years, the management of those with type 1 diabetes has been subject to marked improvements through the introduction of glucose monitoring technologies. What started out with rudimentary (and bulky) devices that required over a minute to provide a reliable blood glucose level are now near weightless and provide accurate values within 5 seconds.
These devices have traditionally used enzymatic methods to ascertain blood glucose concentrations at a given point of time. This is, however, undergoing a dramatic change. CGM devices have been approved and are designed to provide not only more detailed data on glucose excursions (high and low) but in addition, do so in near continuous form. To achieve this, a sensor is placed just under the skin where it detects glucose from the interstitial fluid that fills the capillary spaces rather than directly from the blood). A transmitter sends the signals to a receiver that displays the values.
Currently, three CGM systems have seen widespread applications as stand alone sytems: Medtronic's Enlite sensor, Dexcom's Gen4 and Gen5 sensors, and Abbot's FreeStyle Navigator. Beyond this, three insulin pump-CGM integrated systems are being utilized: Medtronic Paradigm's REAL-Time the Animas Vibe and Tandem’s t:slim G4.
These devices are now poised for future application as part of a closed loop insulin delivery and assessment system (the so called artificial pancreas). Here, a key for continuous glucose monitoring will be a durable and accurate sensor. Indeed, the development of such a sensor has, until now, proved to be quite problematic; given the need for it to be rapid, sensitive and selective for glucose, durable, and safe (when implanted) for continuous use in patients. While it is likely that glucose sensors will become available as a new technology in the not-too-distant future, predicting which form will attract the most use is currently unknown as a variety of modalities (e.g. transdermal, optical) are being pursued.
New insulin delivery and glucose monitoring systems will predict blood glucose values
The third major component of the so-called artificial pancreas is that of a control algorithm. In many ways, the algorithm could be considered 'the brain' within this system as its purpose is to take glucose measurements from the CGM, interpret them, and then instruct the insulin pump to dose appropriately. At it surface, this might seem like a simple process: rise in glucose, infuse insulin; glucose lowers, stop insulin. However, the development of a competent control algorithm has been extremely challenging.
To begin, unlike the setting in which beta cells respond immediately to insulin, with a first and second phase release of insulin, within an artificial pancreas, a delay will be inherent due to the time required for the system to 'understand' that a rise in glucose has occurred, together with the time for the infused insulin to work, followed by the response of the algorithm to assure that the desired effect occurred. This in turn leads to a second major hurdle; what to do in settings where too much insulin is provided? Beyond this, programmes must be able to respond to life’s variables (e.g. sickness, exercise, complex meals, etc.). Such events could result in marked excursions (high and low) that would prove problematic to an algorithm optimised for a particular set of ranges. To meet this need, algorithms must be developed that rapidly identify and adapt themselves to any influence that modifies blood glucose and do so rapidly. And then there is the issue of having a single algorithm that would effectively deal with daytime life versus night.
To meet these challenges, three formats of algorithms are undergoing active investigation. These include those designed for Proportional, Integrative, Derivative Control; MD-Logic Control; and Model Predictive Control. It goes without saying that each format has a series of strengths and weaknesses but as a potential future technology, there is optimism that one or more will address the aforementioned challenges.
One more constituent deserves mention in any discussion of a closed loop system. The three current rapid-acting insulin analogues all begin working about 10–20 minutes after injection and last for several hours. While this may prove sufficient for basal–bolus treatment regimens, these insulins are simply not sufficiently fast-acting to match peak glucose with peak insulin levels, and thus to control postprandial hyperglycaemia. Faster-acting insulins are therefore required for any form of closed loop system. To this end, a number of formulations are undergoing development.