Yesterday I had the chance to visit the Philadelphia headquarters of Echo Therapeutics, a small company working to develop a new non-invasive — that is, totally needle-free — continuous glucose monitoring system called the Symphony CGM.
I was welcomed by Christine Olimpio, Director of Investor Relations and David Walton, Vice President of Marketing and Commercial Development, who explained to me how it works. Basically, unlike the current CGMs on the market (Dexcom’s Platinum G4 and Medtronic’s Guardian), the Symphony does not actually puncture the skin. Instead it’s trans-dermal, meaning it reads glucose levels through the skin’s top layer. If I may paraphrase (with some recent updates) Karmel Allison’s feature on the Symphony from last year:
First, a special skin preparation device is used to permeate the skin before placing the sensor. The device abrades the skin, removing about 0.015 mm (about 1.5 microns) of the outer layer of skin, called the stratum corneum, using adaptive micro-abrasion technology. This process takes 10-20 seconds, and removal of the outer layer of skin allows the measuring of a number of physiological properties, including interstitial glucose levels. The abrasion is so shallow it is not even visible.
After the skin has been prepared, a biosensor is placed on the permeated site. Because the skin has been prepared, the sensors are only microns away from the blood vessels, and are able to measure interstitial glucose . . . as it flows from the blood into surrounding tissues.
The Symphony wirelessly provides the patient’s glucose level every minute to a remote monitor. The monitor then tracks the glucose levels and rate of glucose changes and provides visual and audible alarms if the patient’s glucose levels move outside the target range, which can be personalized for each patient.
As Christine demonstrated for me, what this looks like in practice is that you place the aforementioned skin prep device — which looks kind of like an electric shaver/nosehair clipper — against your skin, using a hole in the sensor’s adhesive to target the precise location. The device then removes the dead, top layer of your skin on the specific spot where the sensor will sit. It’s able to do this painlessly by sensing the hydration levels in your skin — the dead skin of the stratum corneum is only about 10 percent water, while the live layer of your epidermis is some 80 percent water — and it stops as soon as that fresh skin layer has been reached. (You don’t need to worry about accidentally grinding off live skin, in other words.) You pop the sensor on and after a one-hour priming, it begins to measure interstitial glucose levels, supposedly even more accurately than some of the other sensors on the market since it is so close to the skin’s capillary bed. Then it beams the glucose readings wirelessly to a receiver. Here’s a video demonstrating how this all works (note that neither the Symphony nor any sort of phone-based CGM display has been approved anywhere; this is just a demo, featuring a woman with an admirable stomach):
As nice as this all sounds, I must say that when I saw/learned about the Symphony in its current state, the idea of it competing with Dexcom and Medtronic did not make sense. First, the Symphony sensor only works while it’s touching fresh skin; once that skin dies off, which begins to happen within about three days, you need a new sensor. (At the moment, Echo is seeking approval for only 24 hours.) With the Dexcom already approved for seven (and with many users wearing it for much longer) that doesn’t seem viable. Second, as it stands now, the transmitter that sits on your skin is pretty large and unattractive (not that my G4 is winning any beauty contests). And third, considering how painless the Dexcom is once inserted, I don’t consider the non-invasive angle as a huge improvement. I mean, yes, if it could deliver the accuracy of the G4 and could be worn for a week, I might prefer skin abrasion over constant new puncture wounds. But that seems a long way away.
But while I think those are valid points (and ones of which the company is well aware), they overlook the Symphony’s primary focus at the moment: in-patient hospital use. Echo is currently doing a clinical trial of the Symphony system in preparation for submitting a CE Mark Technical File for approval in Europe (see press release here) and, if they get that approval, Echo believes that the Symphony system would be the first non-invasive continuous blood glucose monitoring system that would be approved for in-hospital use in the European Union.
Currently there are five invasive or minimally invasive hospital-based CGMs with CE Mark approval (invasive means it uses intravenous blood; minimally invasive means it uses interstitial fluid drawn from some sort of subcutaneous sensor, like the Dexcom or Minimed systems): Maquet, Menarini’s Glucoday, Edwards/DexCom’s Glucoclear, Optiscan’s OptiScanner, and Medtronic’s Sentrino. Glysure, Glumetrics, and Echo are in the process of developing/seeking CE Mark approval for hospital-based CGM systems, and there are currently no CGMs approved for in-hospital use in the United States.) The idea behind all these systems is that doctors and nurses would be able to monitor their patients’ blood glucose values in near real time on a specially designed bedside monitor, rather than have to rely on occasional spot checks; the particular benefit of the Symphony system is that it’s designed to do so without any sort of puncture wound at all.
If you’re like me, you now may be thinking, “Okay, that sounds great, but how many non-diabetic people really need to have their blood glucose levels monitored while in the hospital?”
The answer, it turns out, is lot: the stress and trauma caused by many surgeries and injuries often causes even non-diabetic patients’ blood glucose to spike – according to a 2010 study in Diabetes Care, hyperglycemia occurs in up to 90 percent of all critically ill patients. That’s because when you’re critically ill, injured, or just had a surgeon slice into you, your body starts pumping out all sorts of counterregulatory hormones, including glucagon (which makes your body release glucose into your blood), growth hormone (which stimulates the immune system but also causes insulin resistance and hyperglycemia), stress-related adrenal hormones called catecholamines (including epinephrine and norepinephrine) and a class of anti-inflammatory steroid hormones called glucocorticoids. These hormones are all important in helping your body to heal, but they also cause high blood sugars (among other things, they impair insulin’s ability both to encourage glucose uptake into muscle and to prevent the liver from dumping extra glucose into the blood). Throw in the fact that many patients in the ICU are receiving intravenous nutritional infusions (often glucose-based), and you’ve got a recipe for hyperglycemia.
In order to reduce blood sugar, hospitals — surprise, surprise – often give these patients intravenous insulin. Hospitals differ in their treatment goals (some might aim for a tight 80-110 mg/dl, others 140-180, etc), and the tighter their aim, the more often they run into the same problem that insulin-dependent diabetics do: they accidentally give too much insulin, and the patient goes low.
The implications of all these highs and lows are actually quite dramatic. An influential 2001 study of 1,548 patients published in the New England Journal of Medicine found that tight glycemic control in intensive care patients (defined as between 80 and 110 mg/dl) was associated with an estimated 32 percent reduction in mortality. Additionally, according to Christine, two or more blood glucose levels above 180 mg/dl have been associated with a two-fold increase in the cost of the hospital stay, and every 50 mg/dl increase in a glucose reading on the day of and after coronary artery bypass surgery is associated with a $2,800 increase in hospital charges.
On the flip side, the so-called NICE-SUGAR trial, the results of which were published in the New England Journal of Medicine in 2009, found that a similar tight glucose target increased the absolute risk of death at 90 days out by 2.6 percent compared to conventional treatment (with a goal of blood sugars less than 180 mg/dl), presumably because the tighter the target, the more likely the patient is to have a hypo-glycemic event.
Putting the results of these studies together, it would seem that keeping patients’ blood glucose between 80 mg/dl and 110 mg/dl is beneficial – as long as you can avoid severe hypos. This implies that any device that can help hospitals safely achieve tighter control without hypos could conceivably both improve patients’ health outcomes and reduce costs. In other words, bring on CGMs! (CGMs ability to reduce the incidence of lows in a hospital study was demonstrated in this 2010 study in Diabetes Care, though, interestingly, it did not find that they helped maintain normal glucose levels any better than a conventional testing schedule – a finding I personally wonder if could be explained by the novelty of the CGM systems to the nursing staff.)
But getting back to the Symphony: Christine and David showed me a demonstration of their version of a bedside monitor and it’s easy to see how it would work: a color display shows the patient’s past and current glucose values on a graph, similar in concept to the graph on your CGM receiver but larger and much easier to read. The doctor/nurse can set a customized glucose target, and alarms sound if the patient goes out of range. This would be a big improvement over the current non-CGM-based techniques to measure blood glucose in an in-patient setting, which usually occur only once every several hours and involve taking arterial blood samples (from IV lines that are already in place), which are either tested on an in-room Accu-chek glucometer, or sent to a lab. Different hospitals have different standards for how often blood glucose should be checked, which means a patient could spend hours in a hypo- or hyper-glycemic state without anyone realizing there was something wrong.
As noted, there are other invasive or minimally invasive continuous glucose monitoring systems currently approved for in-hospital use in the EU, though not in the US. (As a side note, you can’t just slap a Dexcom system onto each patient in the ICU, even if it were financially feasible, because the Dexcom interprets acetaminophen — that is, Tylenol, an extremely common hospital drug — as glucose and gives erroneously high results.) That means there are plenty of competitors trying to grab a share of this market. But the fact that the Symphony is completely non-invasive could make it stand out from the crowd: particularly in a germ-filled hospital setting amongst immune-compromised patients, the fewer puncture wounds you have, the better. Additionally, Echo is hoping to be able to price their sensors lower than those currently on the market.
Additionally, getting back to potential out-patient applications, while a 1-3 day sensor is never going to compete with Medtronic and Dexcom, it could conceivably make sense if it were to be incorporated into a patch pump, like the Omnipod, which has to be switched out every three days anyway. In that case you’d only have to deal with switching out one system (and you’d have only one thing stuck to your body), rather than the current two.
In short, the Symphony system is interesting stuff – and hopefully there will be good news on the in-hospital CE Mark soon. In the meantime, I applaud Echo for pursuing a novel approach to blood glucose monitoring, and hope to be able to report good news in the future.