From Abby: Closed Loop Trial (Part I).
Okay, so the birthday girl (HAPPY YESTERDAY BIRTHDAY, ABBY!) was part of a clinical trial last week that involved the closed loop system, and she both Tweeted her experience and wrote it in two parts here on SUM. I think this trial is fascinating, and Abby's first-hand experience was inspiring. Here's part one of her feedback:
(During the summer of 2009 at CBC, I heard about a Closed Loop Trial in Boston. For one reason or another none of us on staff were able to participate at the time. I personally couldn't because I was unable to commit the time to coming to Boston, but I stayed in touch with the trial coordinator and this fall, she contacted me with some openings in the trial. Here is the extra-long story.)
There are a lot of requirements that factor in to being approved as a participant in this closed-loop-trial. For example, you can’t make any of your own insulin (you need a negative c-peptide blood test) and you also have to have an A1C under 9%. At the first screening visit, I had to fast from 10 pm the night before, then I had to deal with a ton of paperwork and a physical examination before they eventually gave me a Boost milkshake (that I wasn’t allowed to bolus for). The point of this was to draw blood 90 minutes later to screen for c-peptide. (Lucky me: I make zero of my own insulin.) So by the time they drew blood, I was starving and my blood sugar was pushing 300 mg/dL. It was a rough morning but worth it, because I ended up being accepted into the trial!
The next step took place about a month later when I had to select my food for the three days I’d be participating in the trial. They base your carb counts on your weight, and I had to eat between 85 and 120 grams of carbs at every meal, but the trick was that I would only be eating three times a day (will explain this more later). Choosing food was fun, but hard. That's a LOT of carbs for me, I generally don't eat more than 100g carbs all day, let alone in one sitting, three times a day. There were tons of good food choices, like pizza, and lasagna, and sandwiches. I was excited to stuff my face with delicious fatty food in the name of science.
Finally, it was trial time. The day before the trial started, I had to go in for a quick visit to have the continuous glucose sensors placed. I wore two Freestyle Navigators (one “active” for the computer program and one “dummied up” in case something happened to the active one. Navigator sensors take 10 hours to "warm up," so having two saved time in case there was an error) and one Dexcom for control purposes. All three sensors went on my belly, which was totally weird since I NEVER wear any diabetes gear there, but for research purposes they have to do everything in a uniform fashion. They also drew outlines of where the OmniPods would be placed the next day. Once the sensors were on, I went back home to get used to sleeping as a robot.
What won't this girl do for science?
The day of the trial, I couldn’t sleep. I woke up around 6:00 am, even though I didn't need to leave my apartment until 9:00 am. Whoops. I arrived (after taking the packed T with a giant backpack and a purse and a pillow in sweats while everyone else on the train was in fancy-shmancy work clothes - oh hey awkward) and was put into my hospital room for the weekend. Everyone who works on this research floor is SO NICE. I can't even explain how caring and wonderful these nurses and personnel are. Amazing.
After a quick physical and orientation to the floor, things were off and running. I had an IV placed that was hooked up to a machine called a GlucoScout which checks your blood sugar through this weird tubing and machine and then puts it back into your vein, so you don't lose a lot of blood and don't have to have your finger pricked all the time. It was great being able to just look at a machine and see what I was every 15 minutes and really trust the number. After this IV was placed, I ate lunch and manually bolused my insulin for the last time for the duration of the trial. At this point, the GlucoScout was checking me but I was reacting to everything, meaning they had no control over my insulin yet. I bolused for lunch, and watched my blood sugars from there using the Scout, not my meter - the nurses just recorded what I did. I went from 141mg/dL to somewhere around 180mg/dL and came back down, even though I ate 97g of carbs and took 12.6 units which is totally ridiculous amount of insulin (for me to be taking at one time). Things worked out fine.
Next, two Omnipods were placed on my abdomen (using up every last available spot of skin, I looked ridiculous and felt even more ridiculous). One Omnipod had glucagon in it and the other contained Humalog. (Fun fact: OmniPods don't hurt nearly as much as you'd think they would considering that very loud click they make when the needle goes in.)
At 2:50 pm, I took my personal insulin pump off, and at 3:00 pm the whole closed loop system was turned on. At first, it wasn't totally weird, because I don't pay attention to my pump between meal times, but it was super cool to see this computer reacting to the Navigator number and bolusing me every five minutes.
The closed-loop system is totally complicated, and even after the engineer came in and completely explained it to me for over an hour, I still only have the general gist of how it works (which is fine since I'm not an engineer nor do I want to be). Basically, the computer relies on the Navigator CGM, and gives you insulin or glucagon units based on the number. It keeps track of how much insulin you have in your body, gives you a 45% pre-meal bolus when it knows you're eating (which for the trial was at 8 am, 12 pm, and 6 pm, for uniformity’s sake), then reacts to your blood sugar from there.
So for example, I got 3.2u of Humalog right before I ate dinner, and then as my blood sugar rose after my meal, I got between 0.1u an 0.3u (depending on the result), every five minutes. The system administered a basal and correction dose at the same time. This pre-meal bolus was why I had to eat so many carbs, to keep the environment controlled. The researchers realize this isn't a “real life” scenario, but they aren't at the point of testing real life yet. Also, if you're below 120mg/dL you get a small amount of glucagon (I think the most I got was 2u at once, when my blood sugar was somewhere around 74mg/dL) but it also gives you insulin because you need basal.
The whole thing is mind-boggling and I'm glad it makes sense to the doctors and engineers.
I had pizza for dinner, with a salad, an apple, and some milk. It was really tough to eat all of that food, but it was so darned delicious (which helped). My blood sugars rose steadily, I peaked at 217 mg/dL around 9 pm, but then came back down. Apparently it came down too much, because while I was sleeping, I started receiving regular doses of glucagon from the system. I wasn't ever low, never had any hypo symptoms, but I was hovering in the 80s and 90s. This is when the trouble started.
In addition to the fifteen minute GlucoScout checks, the nurses drew a little bit of blood every 30 minutes while I was sleeping to check the glucose in the "gold standard" machine, as well as keeping some fluid for testing after (insulin, glucagon, and c-peptide levels, I think), so I wasn't sleeping well to begin with. Then around 3 am, I woke up feeling really nauseous and like I was going to pass out - that strange feeling where you stop hearing and your body gets really flushed. I told the nurse and he helped by turning the heat down and giving me a cold cloth and some diet ginger ale. I was ok for a while, but from 4 am to 6 am I was (to spare you gruesome details) vomiting and not sleeping. This immediately disqualified me from the study, which was a HUGE bummer. (The first time I vomited they took all of the pumps and sensors off me and I reconnected my own pump, but we kept the GlucoScout on so they could monitor me). By 6:45 am I felt much better and ended up falling asleep until around 8 am.
The whole time I was there they were checking my ketones, and I never spilled any. When I woke up in the morning I was 89 mg/dL and I ate a piece of toast. After that, the nurse checked again and I had large ketones with a blood sugar of 119 mg/dL. The whole thing was just bizarre. I'm sure I had ketones from dehydration and vomiting but it's still disconcerting to have ketones with a perfectly fine blood glucose and not even feeling sick. I drank my weight in ginger ale (diet and regular, so I could take more insulin to clear the ketones) and by 10:30 am my ketones were gone and my blood sugar was 178 mg/dL.
So the moral of my clinical trial story is that even though the researchers only got 12 hours of data from me, they saw that someone could be very sensitive to glucagon, even in small doses (I received 15 units over three hours total) which tells them that maybe a glucagon pump isn't for everyone, or maybe a more conservative dosing would be required.
Part two of Abby's feedback will be posted in a little bit. :) But for now, I just want to thank Abby for doing this, and for sharing her experiences with us. What you did helped move us forward, Abby. You rock.