ADA 2016 Afrezza Abstracts

[peppy]

Jun 5, 2016 9:56:05 GMT -5 cathode said:

Jun 4, 2016 21:10:10 GMT -5 capnbob said:

I don't quite understand why MNKD chose to present these abstracts as they did. Three of them are done as meta-analyis on pooled data from other studies. These are not good studies to present at a meeting. They have a reputation for being little more than attempts to salvage something from past failures. Consequently, endos, insurers, and the FDA have practically no respect for them.

One of the studies is on healthy subjects and compares afrezza to regular insulin. What purpose did that serve? I'm pretty sure that everyone accepts that afrezza acts faster than regular insulin. Moreover, doing any study at this point in healthy individuals doesn't seem like a worthwhile investment.

The "Within-Subject Variability of Insulin Exposure" -- as I interpreted it -- was basically a failure. It showed within subject variability was, in fact, greater than what their model predicted. How is that supposed to help?

Finally, the one that had some potential -- "Technosphere Insulin Inhalation Powder (TI) Displays Earlier Onset" -- is poorly presented. They should have had graphs. Docs at meetings are flooded with information. The most effective presentation is visual, dramatic, and to the point. Second, they present no statistical analysis. By the same token, docs don't feel like wading through mounds of data; they usually try to go for the "meat" of the question -- was there a statistically significant effect. Third, with reference to a primary point -- speed of onset -- I don't see any data detailing exactly when each drug started to take effect. Another problem in this respect is their claim that afrezza started 25-35 minutes before lispro. Well, that contradicts pactically all other research that shows that lispro's effect starts within 10-20 minutes. That would mean that afrezza's effect actually started 20-30 minutes before the patient even inhaled it! The authors attempt no explanation of the remarkable difference that they measured. Fourth, they appear to try and imply that afrezza acts faster by comparing time from onst to 50% of peak and to peak activity to lispro. The problem there is that lispro peaks 90-120 minutes while afrezza peaks at around 60 minutes, so obviously afrezza reaches its peak faster and basically confirms what is already known. Finally, there is their conclusion:

"Cmax and AUC were dose proportional for TI but slightly sublinear for Lispro; saturable GIRmax was obtained over the dose range for both insulins. Onset of activity for TI occurred ca. 25-35 minutes earlier than for Lispro. TI duration of action is about 2 hours shorter than an equivalent dose of Lispro. Dose-response was almost linear up to 48U TI and 30 U Lispro."

That seems to me to be pretty weak. All of it is essentially already known except for the "25-35 minutes earlier" which, as I pointed out, is bound to be viewed as questionable by meeting attendees. 

MNKD should have forgotten all except the last one. For it, they should have collected as much data as possible to confirm the "25-35 minutes," done a statistical analysis to demonstrate a significant difference from lispro, and then provided a single large, dramatically colored graph to force the point home. That might at least given them some ammunition to claim "ultrafast acting." As it stands, I don't seen anything to be gained from these abstracts.

I want to make a few responses to the points raised here. First, this is a scientific session. The ethics of science require that there are no "good" or "bad" outcomes in terms of expectations. Rather, the results are used to either support, reject, or a make no impact on a hypothesis. With regard to the "Within-subject variability" abstract, I disagree with your interpretation that it was a failure. Rather, it seems that the linear mixing model they used does not accurately capture the variability. This is only a failure for the model. Is this finding inherently "bad"? No, it is informative. How does it help? It allows them to iterate on the results and find a model that better fits the experimental data.

On why "MNKD chose to present these abstracts as they did", look at the fact that MNKD employees are the first author on zero of the regular session abstracts. The abstracts were likely written by Sanofi employees or contractors. MNKD pulling abstracts would not be good publicity and downright stupid if the findings were positive, as they are. In addition, other posters above have stated that these are not the actual presentations. These are abstracts. The full set of statistical methods shouldn't be expected to be given in the abstract. Graphs will likely be shown on the posters.

Regarding the Lispro-Afrezza speed comparison -- You use the word "effect" exclusively in your analysis. That word is not found in the "Earlier Onset and Shorter Duration" abstract. Rather, they use the word "onset", which they are defining as the time to 50% of the maximum glucose infusion rate (GIR). I don't know whether the phrase "onset of activity" is legally defined by the FDA, but the 50% metric certainly demonstrates that Afrezza has a faster time of onset in that regard. This 50% metric seems to be used by Novo as well, from this press release/ article on "faster aspart". The current FDA label has this important caveat: "Despite the faster absorption of insulin (PK) from Afrezza, the onset of activity (PD) was comparable to insulin lispro." The clamp study seems to be able to directly address that statement and provide a more favorable view of Afrezza.

Quote; The current FDA label has this important caveat: "Despite the faster absorption of insulin (PK) from Afrezza, the onset of activity (PD) was comparable to insulin lispro." The clamp study seems to be able to directly address that statement and provide a more favorable view of Afrezza.

Reply: Thank you Cathode. Exacting you are.

Perhaps the basis for a label change.

perhaps ultra rapid.   www.mannkindcorp.com/mobile_site/product-pipeline-afrezza.html

Novo Nordisk to Begin Phase 3 Studies of Ultra-Rapid-Acting Version of Novolog in Late 2013

diatribe.org/issues/50/new-now-next/4

[Deleted]

Afrezza is absorbed nearly instantaneously upon lung contact vs a rapid acting insulin that requires the breakdown of the molecule. Why is that not sufficient to label Afrezza ultra rapid, assuming PD are the same for the two insulins?

[capnbob]

Jun 5, 2016 9:56:05 GMT -5 cathode said:

Jun 4, 2016 21:10:10 GMT -5 capnbob said:

I don't quite understand why MNKD chose to present these abstracts as they did. Three of them are done as meta-analyis on pooled data from other studies. These are not good studies to present at a meeting. They have a reputation for being little more than attempts to salvage something from past failures. Consequently, endos, insurers, and the FDA have practically no respect for them.

One of the studies is on healthy subjects and compares afrezza to regular insulin. What purpose did that serve? I'm pretty sure that everyone accepts that afrezza acts faster than regular insulin. Moreover, doing any study at this point in healthy individuals doesn't seem like a worthwhile investment.

The "Within-Subject Variability of Insulin Exposure" -- as I interpreted it -- was basically a failure. It showed within subject variability was, in fact, greater than what their model predicted. How is that supposed to help?

Finally, the one that had some potential -- "Technosphere Insulin Inhalation Powder (TI) Displays Earlier Onset" -- is poorly presented. They should have had graphs. Docs at meetings are flooded with information. The most effective presentation is visual, dramatic, and to the point. Second, they present no statistical analysis. By the same token, docs don't feel like wading through mounds of data; they usually try to go for the "meat" of the question -- was there a statistically significant effect. Third, with reference to a primary point -- speed of onset -- I don't see any data detailing exactly when each drug started to take effect. Another problem in this respect is their claim that afrezza started 25-35 minutes before lispro. Well, that contradicts pactically all other research that shows that lispro's effect starts within 10-20 minutes. That would mean that afrezza's effect actually started 20-30 minutes before the patient even inhaled it! The authors attempt no explanation of the remarkable difference that they measured. Fourth, they appear to try and imply that afrezza acts faster by comparing time from onst to 50% of peak and to peak activity to lispro. The problem there is that lispro peaks 90-120 minutes while afrezza peaks at around 60 minutes, so obviously afrezza reaches its peak faster and basically confirms what is already known. Finally, there is their conclusion:

"Cmax and AUC were dose proportional for TI but slightly sublinear for Lispro; saturable GIRmax was obtained over the dose range for both insulins. Onset of activity for TI occurred ca. 25-35 minutes earlier than for Lispro. TI duration of action is about 2 hours shorter than an equivalent dose of Lispro. Dose-response was almost linear up to 48U TI and 30 U Lispro."

That seems to me to be pretty weak. All of it is essentially already known except for the "25-35 minutes earlier" which, as I pointed out, is bound to be viewed as questionable by meeting attendees. 

MNKD should have forgotten all except the last one. For it, they should have collected as much data as possible to confirm the "25-35 minutes," done a statistical analysis to demonstrate a significant difference from lispro, and then provided a single large, dramatically colored graph to force the point home. That might at least given them some ammunition to claim "ultrafast acting." As it stands, I don't seen anything to be gained from these abstracts.

I want to make a few responses to the points raised here. First, this is a scientific session. The ethics of science require that there are no "good" or "bad" outcomes in terms of expectations. Rather, the results are used to either support, reject, or a make no impact on a hypothesis. With regard to the "Within-subject variability" abstract, I disagree with your interpretation that it was a failure. Rather, it seems that the linear mixing model they used does not accurately capture the variability. This is only a failure for the model. Is this finding inherently "bad"? No, it is informative. How does it help? It allows them to iterate on the results and find a model that better fits the experimental data.

On why "MNKD chose to present these abstracts as they did", look at the fact that MNKD employees are the first author on zero of the regular session abstracts. The abstracts were likely written by Sanofi employees or contractors. MNKD pulling abstracts would not be good publicity and downright stupid if the findings were positive, as they are. In addition, other posters above have stated that these are not the actual presentations. These are abstracts. The full set of statistical methods shouldn't be expected to be given in the abstract. Graphs will likely be shown on the posters.

Regarding the Lispro-Afrezza speed comparison -- You use the word "effect" exclusively in your analysis. That word is not found in the "Earlier Onset and Shorter Duration" abstract. Rather, they use the word "onset", which they are defining as the time to 50% of the maximum glucose infusion rate (GIR). I don't know whether the phrase "onset of activity" is legally defined by the FDA, but the 50% metric certainly demonstrates that Afrezza has a faster time of onset in that regard. This 50% metric seems to be used by Novo as well, from this press release/ article on "faster aspart". The current FDA label has this important caveat: "Despite the faster absorption of insulin (PK) from Afrezza, the onset of activity (PD) was comparable to insulin lispro." The clamp study seems to be able to directly address that statement and provide a more favorable view of Afrezza.

"I disagree with your interpretation that it was a failure." But it was a failure in the sense that their model -- whatever it was -- failed to predict what afrezza would do. Granted, that is not a failure of afrezza itself. However, this is a meeting and docs are moving past rapidly. They see "Within-Subject Variability of Insulin Exposure and Metabolic Activity following Replicate Doses of Technosphere® Insulin Inhalation Powder (TI) in Patients with T1DM" and they glance down and see that there was more variability than was expected per the model. They don't stop and ask whether the model was correct or not. They walk away with the impression that afrezza was more variable than expected.

I didn't mean to imply they should have "pulled" the abstracts. Five of them I just didn't think should have been submitted -- I mentioned those above. Their data wasn't that impressive -- to me at least -- and they dilute the effect of whatever favorable impression might be gained from the one that said something suggestive of signifcance. As far as the "posters" go, I, for one, looked at past afrezza presentations at the ADA and could find nothing but the abstracts. Has Mannkind ever published the posters anywhere. Regardless, if a statistical analysis was done, then the abstract should include at least a statement to that effect and whether the results were significant or not. Again, this is a meeting with tons of data being thrown at viewers, the big stuff like statistical significance needs to stand out and really be pushed in their faces.

As regards "use the word "onset," I interpret that to mean "onset" of the glucose lowering effect. And yes, the first number is to 50%-GIRmax and the second number is time to 100%-GIRmax. The problem is that lispro does not reach its GIRmax until 90-120 minutes AFTER afrezza:







You can see that going from onset at "0" to 50% of afrezza's peak is about 30 minutes while its peak is 55 minutes. For lispro, it doesn't hit the 50% point until around 60 minutes while it's peak is around 90-120. You can't say afrezza's onset is faster simply because time expended from onset to 50% or 100% is less than lispro's beeause, obviously, afrezza's number necessarily must be less since it peaks sooner. Consequently, you can't conclude anything about actual time of onset from that data alone. The above graph suggests time of onset of effect was near identical. What is really needed is a graph of the abstract data to clarify the "25-35" minute earlier onset claim. That is more or less the "crux" of the matter.  

[capnbob]

Jun 5, 2016 11:50:51 GMT -5 @kastanes said:

Afrezza is absorbed nearly instantaneously upon lung contact vs a rapid acting insulin that requires the breakdown of the molecule. Why is that not sufficient to label Afrezza ultra rapid, assuming PD are the same for the two insulins?

Lispro does not require "breakdown." It is the active monomer that has been genetically modified to prevent conversion to the inactive forms. That is the reason for the long tail. Faster absorption does not necessarily translate into a "ultra rapid" PD graph. Other variables besides absorption are at play.

[Deleted]

Jun 5, 2016 14:32:35 GMT -5 capnbob said:

"I disagree with your interpretation that it was a failure." But it was a failure in the sense that their model -- whatever it was -- failed to predict what afrezza would do. Granted, that is not a failure of afrezza itself. However, this is a meeting and docs are moving past rapidly. They see "Within-Subject Variability of Insulin Exposure and Metabolic Activity following Replicate Doses of Technosphere® Insulin Inhalation Powder (TI) in Patients with T1DM" and they glance down and see that there was more variability than was expected per the model. They don't stop and ask whether the model was correct or not. They walk away with the impression that afrezza was more variable than expected.

I didn't mean to imply they should have "pulled" the abstracts. Five of them I just didn't think should have been submitted -- I mentioned those above. Their data wasn't that impressive -- to me at least -- and they dilute the effect of whatever favorable impression might be gained from the one that said something suggestive of signifcance. As far as the "posters" go, I, for one, looked at past afrezza presentations at the ADA and could find nothing but the abstracts. Has Mannkind ever published the posters anywhere. Regargless, if a statistical analysis was done, then the abstract should include at least a statement to that effect and whether the results were significant or not. Again, this is a meeting with tons of data being thrown at viewers, the big stuff like statistical significance needs to stand out and really be pushed in their faces.

As regards "use the word "onset," I interpret that to mean "onset" of the glucose lowering effect. And yes, the first number is to 50%-GIRmax and the second number is time to 100%-GIRmax. The problem is that lispro does not reach its GIRmax until 90-120 minutes AFTER afrezza:







You can see that going from onset at "0" to 50% of afrezza's peak is about 30 minutes while its peak is 55 minutes. For lispro, it doesn't hit the 50% point until around 60 minutes while it's peak is around 90-120. You can't say afrezza's onset is faster simply because time expended from onset to 50% or 100% is less than lispro's beeause, obviously, afrezza's number necessarily must be less since it peaks sooner. Consequently, you can't conclude anything about actual time of onset from that data alone. The above graph suggests time of onset of effect was near identical. What is really needed is a graph of the abstract data to clarify the "25-35" minute earlier onset claim. That is more or less the "crux" of the matter.  

so can you break down it down for us please in short concise form? of all the pros and cons
or a better to way put it so the docs get it...for good or bad

[LosingMyBullishness]

Jun 5, 2016 14:43:25 GMT -5 capnbob said:

Jun 5, 2016 11:50:51 GMT -5 @kastanes said:

Afrezza is absorbed nearly instantaneously upon lung contact vs a rapid acting insulin that requires the breakdown of the molecule. Why is that not sufficient to label Afrezza ultra rapid, assuming PD are the same for the two insulins?

Lispro does not require "breakdown." It is the active monomer that has been genetically modified to prevent conversion to the inactive forms. That is the reason for the long tail. Faster absorption does not necessarily translate into a "ultra rapid" PD graph. Other variables besides absorption are at play.

This is not correct according to two different articles I just looked up for verificiation.
Lispro (and I assume Aspart and Glulisin) are engineered in such a way that they dissociate faster as the stability of the oligomers are reduced.
Insulin lispro  is engineered such that lysine and proline resting on the C-terminal end of the B chain are reversed,
Aspart:  Prolin-28 is substituted by Asparaginacid.
Glulisin: Lysin-29 with Glutaminacid and at 3B Asparaginacidamid with Lysin.

Al and others stated repetitively that the Insulin in Afrezza is a monomer. No engineering to weaken the structure of the oligomers for quicker dissociation can beat a monomer.
Nothing can beat a monomer. That is what I would like to see as definition for ultra-rapid.

[Deleted]

If Afrezza is the only monomer, why the confusion? It is easy for a doubter to claim more variables are in play when it come to absorption, but they are never specific as to what the other variables are.

Assuming Afrezza is the only monomer approved and if PD are equal for Afrezza and rapid acting insulins, then Afrezza is the only true ultra rapid acting insulin available.

[cm5]

Lispro is an hexamer--

The dissociation of formulated insulin lispro hexamer in subcutaneous space--

The surrounding subcutaneous tissue absorbs the hydrophobic phenolic (ie, formalin) preservative, (thus) destabilizing the R6-hexamer, allowing zinc to diffuse away, and then the hexamer dissociates into monomers that can be readily absorbed. 

Successful Drug Discovery, Volume 1
edited by Janos Fischer, David P. Rotella
p.47, Figure 3.7
ISBN: 978-3-527-33685-2
256 pages
May 2015

[agedhippie]

Jun 5, 2016 14:47:26 GMT -5 @iam2sekc4u2002 said:

Jun 5, 2016 14:32:35 GMT -5 capnbob said:

"I disagree with your interpretation that it was a failure." But it was a failure in the sense that their model -- whatever it was -- failed to predict what afrezza would do. Granted, that is not a failure of afrezza itself. However, this is a meeting and docs are moving past rapidly. They see "Within-Subject Variability of Insulin Exposure and Metabolic Activity following Replicate Doses of Technosphere® Insulin Inhalation Powder (TI) in Patients with T1DM" and they glance down and see that there was more variability than was expected per the model. They don't stop and ask whether the model was correct or not. They walk away with the impression that afrezza was more variable than expected.

I didn't mean to imply they should have "pulled" the abstracts. Five of them I just didn't think should have been submitted -- I mentioned those above. Their data wasn't that impressive -- to me at least -- and they dilute the effect of whatever favorable impression might be gained from the one that said something suggestive of signifcance. As far as the "posters" go, I, for one, looked at past afrezza presentations at the ADA and could find nothing but the abstracts. Has Mannkind ever published the posters anywhere. Regargless, if a statistical analysis was done, then the abstract should include at least a statement to that effect and whether the results were significant or not. Again, this is a meeting with tons of data being thrown at viewers, the big stuff like statistical significance needs to stand out and really be pushed in their faces.

As regards "use the word "onset," I interpret that to mean "onset" of the glucose lowering effect. And yes, the first number is to 50%-GIRmax and the second number is time to 100%-GIRmax. The problem is that lispro does not reach its GIRmax until 90-120 minutes AFTER afrezza:







You can see that going from onset at "0" to 50% of afrezza's peak is about 30 minutes while its peak is 55 minutes. For lispro, it doesn't hit the 50% point until around 60 minutes while it's peak is around 90-120. You can't say afrezza's onset is faster simply because time expended from onset to 50% or 100% is less than lispro's beeause, obviously, afrezza's number necessarily must be less since it peaks sooner. Consequently, you can't conclude anything about actual time of onset from that data alone. The above graph suggests time of onset of effect was near identical. What is really needed is a graph of the abstract data to clarify the "25-35" minute earlier onset claim. That is more or less the "crux" of the matter.  

so can you break down it down for us please in short concise form? of all the pros and cons
or a better to way put it so the docs get it...for good or bad

It's difficult! The good thing about Afrezza is the fast start. The neutral thing is that it clears fast. Fast clearance can be good (harder to stack insulin and get a hypo accidentally) and bad (you need a second shot for slow digestion).

The problem with analogs is that they can have a variable speed uptake depending on a number of factors (how close the injection is to a blood vessel or muscle, location on the body, and body temperature). There is not a huge amount in it but it does vary. Afrezza bypasses this problem by using the lungs.

The lungs as a route may not be that great judging from the repeatability study. However the variance seems to be largely masked by the speed of clearance so it looks like it isn't that important.

I was always told that analogs started acting within 15 minutes however the emphasis was on started to act. It is not going to seriously help with a meal but it may take you low if you are borderline (say 75) hence the warning.

[Deleted]

It seems people are trying to obfuscate PK values of Afrezza and other rapid acting insulins. I believe it is time for MannKind to emphasize Afrezza is the only monomer approved, therefore making it the only ultra rapid acting insulin available.

[hillsave]

Summary of all abstracts


Sent from my iPhone

Begin forwarded message:

From: Hillard Saveth <[e-mail]>
Date: June 3, 2016 at 12:52:35 PM EDT
To: [e-mail]
Subject: Summary of all 6 Abstracts on Afrezza to be presented at ADA next week

Treatment Satisfaction in Adults with Type 1 Diabetes Using Mealtime Inhaled Technosphere[reg] Insulin (Afrezza[reg]) in Combination with Insulin Glargine (Lantus[reg]) Versus Insulin Lispro (Humalog[reg]) in Combination with Glargine
Treatment satisfaction increased significantly in both groups ([italic]P[/italic][lt]0.05), with no significant difference between groups in change from baseline. Perceptions of insulin therapy improved significantly in the TI+G group only ([italic]P[/italic][lt]0.001), with greater improvement in the TI+G group than in the L+G group ([italic]P[/italic][lt]0.002). Insulin perception subscales for convenience, comfort, and ease of regimen adherence showed significant ([italic]P[/italic][lt]0.01) improvement in the TI+G group, with greater improvement than in the L+G group ([italic]P[/italic][lt]0.05). In the TI+G group, improvement in treatment satisfaction was significantly ([italic]P[/italic][lt]0.01) associated with improvement in convenience, comfort, and ease of regimen adherence.
The findings indicate that, compared with patients using injected rapid-acting insulin for mealtime boluses, those using TI for mealtime boluses came to see insulin therapy more positively (more convenient, more comfortable or less painful, easier to adhere to regimen) during the course of the study.

No Cardiac Effects Found with Therapeutic and Supratherapeutic Doses of Technosphere[reg] Inhalation Powder: Results from a Thorough QTc Clinical Study
The predominant treatment-related AE was cough (54%), which was highest in the TIP groups. Other safety assessments were unremarkable.
This study showed that inhaled TIP did not produce clinically significant effects on heart rate, PR and QRS interval duration, or cardiac morphology. The data also showed that TIP did not affect cardiac repolarization.

Treat-to-Target Technosphere[reg] Insulin in Patients with Type 1 Diabetes
This single-arm, open-label, treat-to-target pilot study was designed to examine the effects of TI on A1C, postprandial glucose control, and daily insulin requirements in 15 patients with T1DM over 45 days. The study consisted of 4 visits, at 0, 15, 30, and 45 days. All patients were on intensive insulin treatment and wore a continuous glucose monitor (CGM) with real-time display. Patients were instructed to take a second dose of TI if blood glucose (BG) was [ge]180 mg/dl 2 hrs after meals.
Baseline patient demographics were mean age 38.3 yrs, diabetes duration 20.1 yrs, BMI 26.4 kg/m², total daily insulin dose 52.1 U (basal insulin 31.1 IU, prandial insulin 21.0 U [5.5, 6.3, and 6.7 U at breakfast, lunch, and dinner]); 67% of patients were on insulin pumps and 33% were on multi-dose insulin.
Final mean mealtime TI doses were 8.6, 9.8, and 9.6 U (breakfast, lunch, and dinner). During the study, the mean basal insulin dose remained the same. A second supplemental dose of TI was taken 37.7% of the times 90-120 mins after meals (39% between days 0-15; 36% between days 16-30; 34.6% between days 31-45). Mean supplemental doses at day 45 were 5.9, 6.6, and 7.9 U (Figure). Prandial TI doses increased during the first 30 days before stabilizing. A1C values decreased from 7.86% to 7.47% during the study. There were no changes in CGM values with regard to percent of time with CGM values below, above, and within target range of 60-150 mg/dl.
Supplemental TI improved glucose control in T1DM patients without any increase in time spent with BG [lt]60 mg/dl.

Analysis of Cardiovascular Adverse Events in Patients with Type 1 or Type 2 Diabetes Enrolled in Selected Therapeutic Trials in the Phase 2/3 Technosphere[reg] Insulin Development Program
The prevalence of prior cardiovascular disease included in the MedDRA terms myocardial infarction, coronary artery disease, or stroke was [lt]10% in patients with type 1 diabetes but approached 40% in patients with type 2 diabetes. Risk factors were more prevalent in patients with type 2 diabetes (i.e., higher mean age, higher diastolic and systolic blood pressure, use of antihypertensive drugs, abnormal lipoprotein patterns, history of tobacco use, and a higher incidence of a positive medical history for myocardial infarction, heart disease, or stroke).
In controlled trials, the incidence of cardiovascular events did not show increased risk with TI use in type 1, type 2, or the combined type 1 plus type 2 diabetes populations

Doppler Echocardiography in Patients with Type 1 Diabetes Treated with Technosphere[reg] Insulin
Baseline cardiac function was comparable between the 2 treatment groups (table). After 16 weeks of treatment, no clinically meaningful changes from baseline in any of the ECHO parameters were noted in either treatment group. In addition, RV size and function remained unchanged after 16 weeks of therapy.
Technosphere[reg] Insulin vs Insulin Lispro in Patients with Type 1 Diabetes Using Multiple Daily Injections
Prespecified efficacy endpoints included change in A1C, fasting plasma glucose (FPG), and 1- and 2-hr PPG following a standard meal challenge.
At 16 wks, change in A1C was similar in the 2 groups and there was no difference in G doses. The least-squares mean A1C treatment difference from baseline was -0.10% (SE 0.09; 95% CI -0.28, 0.07) for TI+G and -0.03% (SE 0.08; 95% CI -0.19, 0.13) for lispro+G. Treatment group differences for change from baseline in FPG (-32.4 mg/dl [SE 12.4; 95% CI -57.1, -7.7; [italic]P[/italic]=0.0107]) and 1- and 2-hr PPG (-66.29 mg/dl, [italic]P[/italic][lt]0.0001; -34.40 mg/dl, [italic]P[/italic]=0.0175) were significantly lower with TI+G vs lispro+G. Incidence of hypoglycemia was similar; however, the overall total and mild/moderate hypoglycemic event rates (events/pt-month) were significantly reduced with TI+G vs lispro+G. No differences in pulmonary function tests were observed between treatment groups.
TI use was comparable in A1C reductions to insulin lispro with significantly lower 1- and 2-hour PPG and FPG and fewer hypoglycemia events/pt-month in pts with T1DM on MDI.

Keeping you informed

Hillard Saveth
LPL Financial
9 South Long Beach Road
Rockville Centre, NY 11570
516-763-9700- OFFICE
516-967-8334- CELL

Member FINRA/SIPC

[agedhippie]

Jun 5, 2016 20:01:23 GMT -5 @kastanes said:

It seems people are trying to obfuscate PK values of Afrezza and other rapid acting insulins. I believe it is time for MannKind to emphasize Afrezza is the only monomer approved, therefore making it the only ultra rapid acting insulin available.

No!  Analogs are largely monomer that is why they are analogs and not insulin. The difference is the delivery route - Afrezza goes almost immediately into the bloodstream. The only thing that is faster is IV delivered Regular insulin, despite not being monomeric , because it is delivered directly into the bloodstream.

The key to rapid action is the time to the bloodstream, not the form. Rapid Acting has to percolate from subcutaneous fat, to capillaries, to the proper cardiovascular system and that all takes time. 

[cathode]

Jun 5, 2016 14:43:25 GMT -5 capnbob said:

Jun 5, 2016 11:50:51 GMT -5 @kastanes said:

Afrezza is absorbed nearly instantaneously upon lung contact vs a rapid acting insulin that requires the breakdown of the molecule. Why is that not sufficient to label Afrezza ultra rapid, assuming PD are the same for the two insulins?

Lispro does not require "breakdown." It is the active monomer that has been genetically modified to prevent conversion to the inactive forms. That is the reason for the long tail. Faster absorption does not necessarily translate into a "ultra rapid" PD graph. Other variables besides absorption are at play.

I think you have the right idea, but aren't stating it correctly. Insulin Lispro, by definition, is a single protein molecule. It does not breakdown and is active in the exact same way as human insulin. Lispro can form dimers and hexamers just like regular human insulin, especially in the presence of zinc (the same element that enables stable hexameric insulin to be stored in the beta cells of a healthy pancreas). High concentration of Lispro in solution will also cause Lispro to crystallize. The pharmaceutical benefit of Lispro compared to human insulin is that it less-readily forms dimers (paired insulin molecules) that combine to make hexamers. Lispro is also quicker to dissociate from the hexameric state than regular human insulin.

A vial of Humalog (Insulin Lispro) that a patient receives from the pharmacy contains twice as many Lispro molecules as zinc ions. It also contains about 50 times as many meta-Cresol molecules as Lispro molecules. Meta-Cresol performs a similar function as the zinc in solution, which is to say that it acts as a nucleus for Lispro association and hexamer formation. Eli Lilly calls Meta-Cresol a preservative. The vial of Humalog from the pharmacy contains hexameric Insulin Lispro.

[cathode]

Jun 5, 2016 20:39:34 GMT -5 agedhippie said:

Jun 5, 2016 20:01:23 GMT -5 @kastanes said:

It seems people are trying to obfuscate PK values of Afrezza and other rapid acting insulins. I believe it is time for MannKind to emphasize Afrezza is the only monomer approved, therefore making it the only ultra rapid acting insulin available.

No!  Analogs are largely monomer that is why they are analogs and not insulin. The difference is the delivery route - Afrezza goes almost immediately into the bloodstream. The only thing that is faster is IV delivered Regular insulin, despite not being monomeric , because it is delivered directly into the bloodstream.

The key to rapid action is the time to the bloodstream, not the form. Rapid Acting has to percolate from subcutaneous fat, to capillaries, to the proper cardiovascular system and that all takes time. 

Actually no. Insulin analogs are called analogs because the molecular structure is somehow altered from that of regular human insulin. FDA approved analogs have a similar form and function as regular human insulin but the slight alterations in the amino chains cause distinct differences in their stability and kinetics. The form does matter for subcutaneous insulin, because it is the form that dictates the rate of dissociation into monomers.

[cm5]

(Wikipedia) Meta-Cresol:

meta-Cresol, also 3-methylphenol, is an organic compound with the formula CH3C6H4(OH). It is a colourless, viscous liquid that is an intermediate in the production of other chemicals. It is a derivative of phenol, an isomer of p-cresol and o-cresol