The EASD Virtual Meeting place has been a revelation. It’s allowed me to view the sessions that I wanted the DOC that’s present at the annual meetings to go and see, and as a result, given me the chance to see the sessions first hand.
Having been able to see the slides and watch the presentations, I wanted to show the two sets of data side by side before sharing my thoughts. The images here are all screenshots from the EASD Virtual Meetings hosted at http://www.easdvirtualmeetings.org/#Home, and were from presentations by Dr Tim Heise on BCLispro and Dr David Russell-Jones on Faster Aspart.
Faster Insulins – the Study details
This update is from two of the sessions that were shown in the Insulin Insights grouping. Namely, some clinical trials of Bio Chaperoned Lispro and Faster Aspart. These are basically faster acting versions of Humalog from Lilly and NovoRapid from Novo Nordisk, but we don’t know what they’ll be called in the future.
I’ve added the slides that are of interest from the presentations and some of my own commentary based on this. It’s certainly an interesting set of data.
Faster Aspart from Novo Nordisk
First up, Faster Aspart.
What is it? It’s NovoRapid with a form of Vitamin B3 which makes it absorb fasterand L-Arginine which makes it stable.
The absorption profile shows quicker absorption:
The trial split the participants into those confident and those not confident with carb counting and dose adjustment and gave them different mechanisms to use accordingly. In this trial, titration was undertaken, unlike, the BCLispro trial. The trial used a similar liquid meal to that of the BCLispro trial.
Post-prandial glucose levels were show to be 1.18 mmol/l lower after 1 hour and 0.67 mmol/l lower at 2 hours.
And the Hba1C reductions showed up okay, with a reduction in those taking it pre-prandially from average 7.6% to average 7.3%.
In addition, hypoglycaemia was not made worse by the faster acting version:
So all in all, it looks good. Given the size of this trial, it suggests that CE and FDA approval are not too far off.
That’s the Novo offering.
Bio Chaperoned Lispro from Lilly
What about the Lilly offering? Bio-Chaperoned Lispro (or Faster Humalog).
Lilly have taken a slightly different approach in their offering, using a small oligosaccharide-based molecule which accelerates the insulin absorption.
What does this mean? Well, just like the Faster Aspart, the BCLispro is absorbed noticeably faster, reaches a higher max and drops of more quickly than its non-accelerated predecessor.
If anything, if you look at the picture below, it looks as though it has faster absorption than that of faster aspart, but that would need to be compared in side by side trials.
It’s also faster out of the body, as shown below.
Checking out the trial, this insulin is at an earlier stage than the Faster Aspart, so we see that the trial is smaller. In this case there were 38 participants and the trial was undertaken by with a standardised liquid meal and used 0.2u/kg of insulin for all participants.
If we move on to the post prandial glucose results though, the data looks astounding. The difference between Humalog and BCLispro is much bigger than between Aspart and Faster Aspart. At one hour the mean difference is 2.3mmol/l and at 2 hours it’s 1.5mmol/l. Look at how flat the curve has become!
Looking at the area under the curve we see that there is a 61% reduction with the BCLispro. This equates to a much greater period remaining within a lower post prandial glucose range. Essentially what the graph says is that compared to the pre-meal baseline, the spike with BCLispro is half that of Humalog, whereas with Faster Aspart, it was 83% (although these numbers should be taken with a healthy level of cynicism – in this study, remember, it was a set meal with a set insulin amount per kg of body weight – an untitrated sample).
Similarly to the faster Aspart, post prandial hypoglycaemia levels did not vary between the older, slower analogue and the new formulation.
From the Aspart trial we see:
And from the BCLispro study:
Looking at the two, there are broadly similar results, although it would appear as though the BCLispro is more effective than the Faster Aspart. The only way to really determine that would be through a full, independent side by side study though.
Otherwise, the number of hypos is not abnormally higher with the faster insulins, and the reduction in post-prandial glucose really looks to be something that is worth pursuing.
Two things stand out for me from the data. The first is that the BCLispro data seems way better than that of Faster Aspart. To such an extent, a larger trial is clearly necessary to see whether it holds true over a decent sample size.
Secondly, the area under the curve for BCLispro reminds me a lot of that of Afrezza. It’s not the same, but as injectable it’s doing well to even start to approach that!
It’s worth noting that all these results were obtained using a meal that was very rapidly absorbed – liquid meals have this unfortunate tendency, and as such, skews the data. It remains to be seen what happens in a clinical setting when using normal, whole food as the meal form and the effects that has on post prandial hypoglycaemia with faster insulin.
Which brings me on to the next step with these faster insulins. Faster in/faster out is the mantra. If you look at the clamp curves for both of these, they’ve still got quite long tails, but with the much earlier take up, how will this change the way they play in pumps? From what I can guess, it will mean that your basal rate, which is currently set for roughly two hours ahead, will need to be reconsidered, and that looks like needing to be reconsidered only an hour ahead instead of two. I can see some confusion at the pump clinics coming…
The other issue that arises here is that of insulin profiles in pumps. Let’s think about that for a moment. When you bolus, the pump calculates IOB to a specific model for rapid acting. Most of the rapid analogues have similar profiles. These extra-rapid ones appear, initially at least, to be a bit different. They are certainly different from the current profiles, so it means that the IOB calculations in existing pumps are unlikely to be correct, and given the ease with with which updates are done to pump firmware, this could get interesting.
In a similar vein, the Artificial Pancreases coming to the market now are all firmwared up. The code runs in the pump/”brain” module. These profiles are critical in determining the IOB of both bolus and basal, and making decisions as to what to do next. Does the advent of these new insulins raise questions over the FDA requirement of a separate, secure “brain”? It certainly presents a huge advantage to OpenAPS, Loop and HAPP as software products that simply get a new set of curves, downloaded through a normal, easy mechanism!
Overall then, these are some very interesting results and show that a lot is going on. The improved performance of both insulins is going to be really interesting in use and can only add to the arsenal of what we have available to improve our glucose level management. They also raise many interesting questions.
I am looking forward to getting my hands on:
- More data!
- BCLispro, when it’s released, if it’s as good as this shows!