#Fiasp – Faster Acting, but what else is going on?

Having used Fiasp for nearly three weeks now, I’ve seen a couple of aspects to the insulin that have had me slightly mystified, and digging back into what I have been doing and what’s really going on.

For historical perspective, I used to use NovoRapid in the insulin pump reservoir for 6 days. I know that the recommendation is to change every 3 days, but I’m lazy. I had no issues with the insulin effectiveness failing, so I figured, Insulin Aspart? Should be okay? Right? I also change my cannula every three days (although historically, I’ve changed the tube every six or so). Well I’m not sure…

I’d been using the Fiasp for about a week, when I had the following.

Looking at this trace, we can see that breakfast had some sort of crazy issue (the rest of Saturday got rath broken thanks to Dexcom issues, that required me to change the sensor later on). I carb counted (I’ve been doing this for a while – breakfast is really quite straightforward) and I bolused. Here it is in close up:

2u just before, 2u just after, expecting OpenAPS to take over. Realising there was an issue I added some more. Then I watched it. And my glucose level continued to climb. Like the Fiasp was having no effect. Even if I’d carb counted at half what I’d eaten, 4u should have done something. Over the next hour I added another 6u . I didn’t feel like I was seeing the 20 minute reaction time that I’d seen in the week previously, and I also hadn’t eaten 100g of carbs with breakfast. Something wasn’t going according to plan.

We retired back to the hotel room, did another 2u and I replaced the set, the reservoir and the tube, and also applied a 4u shot via syringe. This amount of insulin and approach is not what I would suggest anyone else does. It could be extremely dangerous.  It seemed to resolve the issue, and the rest of that day is a cautionary tale in the use of technology. A combination of bad signal, poor communication between devices and just plain bad data really didn’t help anyone. Me the most, so the upshot was that my day was rather a mess by the end. I’d kept relatively flat throughout the afternoon, which was spent speaking at the Edinburgh Science Fair, where ironically, one of the topics we discussed was being on the cutting edge and the issues that sometimes came with it. Twofold in this case…

So I set to looking at why there was an issue? On this occasion, the insulin was into its fourth day and the set was at the end of its third. I’ve never experienced issues with cannula performance so I checked the set when I removed it. No issues showing. Four days had never been an issue with Novorapid, but, I wondered, was it in this case? The other option, Fiasp sometimes doesn’t work, had occurred to me. I needed to see more data.

Unfortunately, more data was to come along. And more rapidly than I was expecting. Here are two more occasions where I have seen almost exactly the same thing.

In both cases, a large bolus was given just before eating, and seems to have had no effect whatsoever. In both cases, I’d had insulin in the reservoir for over three days. In all cases, the cannula was relatively new.

I have a few hypotheses here that need some further investigation.

1. The first of these is that my carb counting has been horribly out in these scenarios and the end result is that my insulin amount simply didn’t cover what was going on. In this condition, the glucose conversion rates and glucagon release overwhelm Fiasp and it needs far more insulin than normal IC ratio to work. This feels a little off and I’m confident in at least one of these Carb Counts, if not two. That would leave the IC ratio as being incorrect, yet, in other cases that hasn’t been affected as my insulin sensitivity factor has remained the same (as previously tested).
2. The second is that the insulin, now older than three days, doesn’t work as effectively. Certainly, if we look at the OpenAPS performance prior to these spikes, I was being fed more insulin to handle what should have been normal conditions than in a normal situation.
3. Thirdly, there are a possible set of body conditions where Fiasp appears to have a delayed or limited effect. It’s possible that the Fiasp form of insulin aspart, due to the introduction of the B3, doesn’t get absorbed in the liver in the same (however small) amounts that the older form of insulin aspart does. As a result, glucagon generated spikes from liver dumps are not handled in the same way. This might tie in with comments from other new users that “getting up phenomenon” isn’t being noticeably affected by Fiasp. This is a difficult one to test and may require the use of glucagonogenic protein as a catalyst to see what happens.
4. Fourthly, due to the differences between the traditional fast acting analogues and Fiasp, the modified version of oref0 I’m using is not optimised to deal with the differences of the insulin’s effects and is therefore resulting in unexpected outcomes. I’m not sure this is the case, as I wouldn’t expect to see it work then not work in what seems like a cycle, but it’s something to keep an eye on.
5. Fifthly, it could be degraded insulin in the vial, but the times I’ve seen this have come from two different vials and it seems to have worked okay in the interim.
6. Finally, I could be starting to see degradation of my pump sites, in spite of rotating around them for the last two years. This again seems unlikely as the pump sites seem to work well, then simply not, which doesn’t fit with the normal pattern of site issues.

At this stage I am inclined towards number 2 or 3. Mitigating that more than three days in the pump reservoir really doesn’t work for Fiasp is something that is relatively easy. In the meantime, there is also some testing I can do in relation to 1 and 3, and I will follow those up over Easter.

For now though, I’m changing my insulin and set every three days religiously. If this proves to be the way to keep Fiasp working effectively, then it’s the right thing to do.