|
|
Post by findingomega on Jan 6, 2016 10:14:01 GMT -5
I do believe in this drug/delivery system, but how will MNKD afford the studies that are essential in providing evidence of superiority, juvenile safety and efficacy etc. Maybe they will forgo the superiority study and let the market reveal Afrezza's positives but the lung function and juvenile studies seem to be a costly endeavor. Any ideas on how the company will afford these necessary tests?
|
|
|
|
Post by derek2 on Jan 6, 2016 10:18:47 GMT -5
I do believe in this drug/delivery system, but how will MNKD afford the studies that are essential in providing evidence of superiority, juvenile safety and efficacy etc. Maybe they will forgo the superiority study and let the market reveal Afrezza's positives but the lung function and juvenile studies seem to be a costly endeavor. Any ideas on how the company will afford these necessary tests? The necessary trials are the ones mandated by the FDA - the huge ongoing safety trial (yet to start), the pediatric trials, and the dosing variability trial. Perhaps one or more of these could be designed to measure a secondary outcome that would help the label, but do consider that after 6000 patients, it is likely that the true efficacy of Afrezza has already been measured. |
|
|
|
Post by jpg on Jan 6, 2016 10:31:57 GMT -5
I do believe in this drug/delivery system, but how will MNKD afford the studies that are essential in providing evidence of superiority, juvenile safety and efficacy etc. Maybe they will forgo the superiority study and let the market reveal Afrezza's positives but the lung function and juvenile studies seem to be a costly endeavor. Any ideas on how the company will afford these necessary tests? The necessary trials are the ones mandated by the FDA - the huge ongoing safety trial (yet to start), the pediatric trials, and the dosing variability trial. Perhaps one or more of these could be designed to measure a secondary outcome that would help the label, but do consider that after 6000 patients, it is likely that the true efficacy of Afrezza has already been measured. Agreed that there are a lot of very expensive mandated trials to do and that there is not a chance Mannkind or Al Mann has the $ to do these but I would strongly disagree that the trials done so far show what Afrezza can do. None of these trials were designed to show superiority and as you know design is the most important thing when looking at results... |
|
|
|
Post by bearcatmax on Jan 6, 2016 11:31:52 GMT -5
IF they can find another partner they need to work it into the deal if possible that the partner will conduct these studies, immediately. Take less of the profits for the partner to conduct these studies. If they can survive the next 2-3 years and get these studies completed in that time they will turn profitable. Maybe not make as much off Afrezza, but enough to keep the company running without worries and begin building the pipeline. Building the pipeline is the true value of the company. They just have to survive to get into that phase, IMO.
|
|
|
|
Post by derek2 on Jan 6, 2016 13:14:17 GMT -5
The necessary trials are the ones mandated by the FDA - the huge ongoing safety trial (yet to start), the pediatric trials, and the dosing variability trial. Perhaps one or more of these could be designed to measure a secondary outcome that would help the label, but do consider that after 6000 patients, it is likely that the true efficacy of Afrezza has already been measured. Agreed that there are a lot of very expensive mandated trials to do and that there is not a chance Mannkind or Al Mann has the $ to do these but I would strongly disagree that the trials done so far show what Afrezza can do. None of these trials were designed to show superiority and as you know design is the most important thing when looking at results... This is probably the wrong time for me to start (or more accurately, resume) acting like an abrasive know-it-all (I try to be polite on this more bullish forum and respect the tone of the board), however, the following document is worth a read: www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/EndocrinologicandMetabolicDrugsAdvisoryCommittee/UCM392917.pdfSlide 15 (page 16 in the pdf) talks about how non-inferiority trials are far more forgiving than superiority trials, and are designed to bias towards trial success, whereas superiority trials bias toward failure, with a far higher burden of proof. Superiority trials must show MORE clinical effect, whereas non-inferiority trials must show _some_ effect, and in the case of Afrezza that _some_ effect on HbA1c was permitted to be 40% that of SC insulin. (-.4% Afrezza vs -1.0% for SC insulin) The idea being that as long as Afrezza gave some primary benefit on HbA1c, even if less than SC insulin, and some other benefit (convenience, lower hypo, etc), then there should be no reason to exclude it. Thing is, Afrezza barely met that non-inferiority criterion for effect on A1C, (40% of effect of SC insulin) and that's not just for the affinity trials - Afrezza has always shown lower efficacy compared to SC insulin. It's tough to imagine a superiority trial where MNKD would be able to prove beyond statistical significance that Afrezza has a greater than 100% of the effect of SC insulin. Afrezza was numerically inferior (-0.32%, IIRC) Put another way, in the Affinity trial, Afrezza was numerically inferior to SC insulin, even considering that the HbA1c effect goal was 40% of the efficacy of SC insulin. Afrezza was saved by the overlap of the margins of error, which gives some leeway. Not only would Afrezza not have this benefit of the doubt in a superiority trial, but it would have to prove beyond the margin of error that it was superior. [Within margin of error non-inferior] < [numericaly non-inferior] < [beyond margin of error non-inferior] < [within margin of error superior] < [numerically superior] < [beyond margin of error superior] As per page 84 in the PDF, Afrezza has never been shown to be superior to SC insulin even if you include all the way to the edge of the lower 90% margin of error, which gives 2 full standard deviations of forgiveness. In the case of a true superiority trial, you need to show > that 1% HbA1c benefit of SC insulin and then add on the 2 SD of effect to prove it isn't just chance. < .4% vs > 1%. That's a big bridge to cross, especially given that all of MNKD's trials against SC insulin with thousands of patients have shown otherwise. |
|
|
|
Post by peppy on Jan 6, 2016 13:28:06 GMT -5
I do believe in this drug/delivery system, but how will MNKD afford the studies that are essential in providing evidence of superiority, juvenile safety and efficacy etc. Maybe they will forgo the superiority study and let the market reveal Afrezza's positives but the lung function and juvenile studies seem to be a costly endeavor. Any ideas on how the company will afford these necessary tests? it is really just the mandated long term study we need to be concerned about.
clamp study; www.clinicaltrials.gov/ct2/show/NCT02470637 This study has been completed.
pediatric study, the 8 week portion completed. clinicaltrials.gov/ct2/show/NCT02527265?term=afrezza&recr=Open&no_unk=Y&rank=17
|
|
|
|
Post by james on Jan 6, 2016 14:16:15 GMT -5
I have no specific credibility to suggest this. However, it seems to me there is a very reasonable grounds for a petition to the FDA to delay the safety study until the product obtains more commercial success. The suggested scope of the study is nearly the size of the probable patient population at present.
|
|
|
|
Post by jpg on Jan 6, 2016 14:16:46 GMT -5
Agreed that there are a lot of very expensive mandated trials to do and that there is not a chance Mannkind or Al Mann has the $ to do these but I would strongly disagree that the trials done so far show what Afrezza can do. None of these trials were designed to show superiority and as you know design is the most important thing when looking at results... This is probably the wrong time for me to start (or more accurately, resume) acting like an abrasive know-it-all (I try to be polite on this more bullish forum and respect the tone of the board), however, the following document is worth a read: www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/EndocrinologicandMetabolicDrugsAdvisoryCommittee/UCM392917.pdfSlide 15 (page 16 in the pdf) talks about how non-inferiority trials are far more forgiving than superiority trials, and are designed to bias towards trial success, whereas superiority trials bias toward failure, with a far higher burden of proof. Superiority trials must show MORE clinical effect, whereas non-inferiority trials must show _some_ effect, and in the case of Afrezza that _some_ effect on HbA1c was permitted to be 40% that of SC insulin. (-.4% Afrezza vs -1.0% for SC insulin) The idea being that as long as Afrezza gave some primary benefit on HbA1c, even if less than SC insulin, and some other benefit (convenience, lower hypo, etc), then there should be no reason to exclude it. Thing is, Afrezza barely met that non-inferiority criterion for effect on A1C, (40% of effect of SC insulin) and that's not just for the affinity trials - Afrezza has always shown lower efficacy compared to SC insulin. It's tough to imagine a superiority trial where MNKD would be able to prove beyond statistical significance that Afrezza has a greater than 100% of the effect of SC insulin. Afrezza was numerically inferior (-0.32%, IIRC) Put another way, in the Affinity trial, Afrezza was numerically inferior to SC insulin, even considering that the HbA1c effect goal was 40% of the efficacy of SC insulin. Afrezza was saved by the overlap of the margins of error, which gives some leeway. Not only would Afrezza not have this benefit of the doubt in a superiority trial, but it would have to prove beyond the margin of error that it was superior. [Within margin of error non-inferior] < [numericaly non-inferior] < [beyond margin of error non-inferior] < [within margin of error superior] < [numerically superior] < [beyond margin of error superior] As per page 84 in the PDF, Afrezza has never been shown to be superior to SC insulin even if you include all the way to the edge of the lower 90% margin of error, which gives 2 full standard deviations of forgiveness. In the case of a true superiority trial, you need to show > that 1% HbA1c benefit of SC insulin and then add on the 2 SD of effect to prove it isn't just chance. < .4% vs > 1%. That's a big bridge to cross, especially given that all of MNKD's trials against SC insulin with thousands of patients have shown otherwise. As so often I don't agree it seems... You seem to be mixing a few concepts of trial design maybe? Analaysing past trials results do not reflect what 'real world' superiority trials would or could show. Simply based on the fact all the trials run so far were done based on how a trial looking at 'slow in slow out insulin' needs to be designed and never with the flexibility needed to show the advantages of a 'fast in fast out' insulin. Titration to effect is key. It's like comparing nitroprussiate to labetolol for blood pressure control. A protocol designed to work for labetolol will look bad and even dangerous if applied to nitroprusiate. Many scientists (including sadly many at the FDA and some MNKD pundits...) don't have the clinical experience to understand real world biological applications of highly variable drug therapy. And I certainly see your point about you thinking you know it all... Turns out a lot of us have been very wrong on this one. Who has made more (or lost least for some)? It's a shame Mannkind ends this way but they were taking on a very big market and endos love simplicity (like the trial explanation you provide above) and Sanofi never intended to push for a good CMG 'beat to beat' superiority trial. Sadly Mann is right but will be run over by partially literate scientists... I have simply and very consistantly disagreed with you and 'your friend' on most 'points of science'... You have been 100 times the gentlemen though. We will probably never read or get science the same way but thank you for being such a polite poster. |
|
|
|
Post by derek2 on Jan 6, 2016 15:16:31 GMT -5
This is probably the wrong time for me to start (or more accurately, resume) acting like an abrasive know-it-all (I try to be polite on this more bullish forum and respect the tone of the board), however, the following document is worth a read: www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/EndocrinologicandMetabolicDrugsAdvisoryCommittee/UCM392917.pdfSlide 15 (page 16 in the pdf) talks about how non-inferiority trials are far more forgiving than superiority trials, and are designed to bias towards trial success, whereas superiority trials bias toward failure, with a far higher burden of proof. Superiority trials must show MORE clinical effect, whereas non-inferiority trials must show _some_ effect, and in the case of Afrezza that _some_ effect on HbA1c was permitted to be 40% that of SC insulin. (-.4% Afrezza vs -1.0% for SC insulin) The idea being that as long as Afrezza gave some primary benefit on HbA1c, even if less than SC insulin, and some other benefit (convenience, lower hypo, etc), then there should be no reason to exclude it. Thing is, Afrezza barely met that non-inferiority criterion for effect on A1C, (40% of effect of SC insulin) and that's not just for the affinity trials - Afrezza has always shown lower efficacy compared to SC insulin. It's tough to imagine a superiority trial where MNKD would be able to prove beyond statistical significance that Afrezza has a greater than 100% of the effect of SC insulin. Afrezza was numerically inferior (-0.32%, IIRC) Put another way, in the Affinity trial, Afrezza was numerically inferior to SC insulin, even considering that the HbA1c effect goal was 40% of the efficacy of SC insulin. Afrezza was saved by the overlap of the margins of error, which gives some leeway. Not only would Afrezza not have this benefit of the doubt in a superiority trial, but it would have to prove beyond the margin of error that it was superior. [Within margin of error non-inferior] < [numericaly non-inferior] < [beyond margin of error non-inferior] < [within margin of error superior] < [numerically superior] < [beyond margin of error superior] As per page 84 in the PDF, Afrezza has never been shown to be superior to SC insulin even if you include all the way to the edge of the lower 90% margin of error, which gives 2 full standard deviations of forgiveness. In the case of a true superiority trial, you need to show > that 1% HbA1c benefit of SC insulin and then add on the 2 SD of effect to prove it isn't just chance. < .4% vs > 1%. That's a big bridge to cross, especially given that all of MNKD's trials against SC insulin with thousands of patients have shown otherwise. As so often I don't agree it seems... You seem to be mixing a few concepts of trial design maybe? Analaysing past trials results do not reflect what 'real world' superiority trials would or could show. Simply based on the fact all the trials run so far were done based on how a trial looking at 'slow in slow out insulin' needs to be designed and never with the flexibility needed to show the advantages of a 'fast in fast out' insulin. Titration to effect is key. It's like comparing nitroprussiate to labetolol for blood pressure control. A protocol designed to work for labetolol will look bad and even dangerous if applied to nitroprusiate. Many scientists (including sadly many at the FDA and some MNKD pundits...) don't have the clinical experience to understand real world biological applications of highly variable drug therapy. And I certainly see your point about you thinking you know it all... Turns out a lot of us have been very wrong on this one. Who has made more (or lost least for some)? It's a shame Mannkind ends this way but they were taking on a very big market and endos love simplicity (like the trial explanation you provide above) and Sanofi never intended to push for a good CMG 'beat to beat' superiority trial. Sadly Mann is right but will be run over by partially literate scientists... I have simply and very consistantly disagreed with you and 'your friend' on most 'points of science'... You have been 100 times the gentlemen though. We will probably never read or get science the same way but thank you for being such a polite poster. Thanks! BTW - I totally agree with you that a way forward could be to show the FDA in, a clinically controlled trial, how to get the best out of Afrezza. I'm open to the possibility that the trials so far have not been conducive to the best possible outcomes. Where we differ on that, I think, is that I would want to see proof before changing my assessment, whereas I think you have some faith already, as given evidence to you by the experience of Sam, Eric et al. Just a fair difference of opinion. |
|
|
|
Post by esstan2001 on Jan 6, 2016 15:30:39 GMT -5
All this stuck in the mud A1C standard for evaluating- how to change the paradigm?
Why not have PWD like Sam, Eric, Clevland Clinic etc. lobby / petition / start a movement to organize an expert advisory panel of thought leader endocrinologists to develop new standards / guidelines regarding diabetes care, using the newer technologies available (CGM)? Then (for trials) define the obvious, that maintaining real time blood glucose levels within such and such a tighter range leads to less cell stress, health complications etc.-
Run these trials against injectable RAA and compare the (CGM data) blood glucose excursions, show (prove) the obvious superiority, along with some A1C data...
Get the FDA into this century already.
|
|
|
|
Post by jpg on Jan 6, 2016 15:47:39 GMT -5
Thanks! BTW - I totally agree with you that a way forward could be to show the FDA in, a clinically controlled trial, how to get the best out of Afrezza. I'm open to the possibility that the trials so far have not been conducive to the best possible outcomes. Where we differ on that, I think, is that I would want to see proof before changing my assessment, whereas I think you have some faith already, as given evidence to you by the experience of Sam, Eric et al. Just a fair difference of opinion. Not to ramble on to much, but I will... My original impression of what could be done was based on simply understanding the PK/ PD of various insulins and comparing it with Afrezza. It seemed onvious to me that this would be a much much better insulin if you had a way to get beat to beat target points (which CGM provides). The diabetics who report their results simply reinforced my view of the superior PK/PD and added a few twists I hadn't thought about (and that made retrospectively a lot of sense). So no it wasn't the end users but the confirmation of what should happen. Are we getting only a partial picture? Probably and most likely but their experience does fit well with most of what we know of how native insulin is thought to work. |
|
|
|
Post by jpg on Jan 6, 2016 15:50:46 GMT -5
All this stuck in the mud A1C standard for evaluating- how to change the paradigm? Why not have PWD like Sam, Eric, Clevland Clinic etc. lobby / petition / start a movement to organize an expert advisory panel of thought leader endocrinologists to develop new standards / guidelines regarding diabetes care, using the newer technologies available (CGM)? Then (for trials) define the obvious, that maintaining real time blood glucose levels within such and such a tighter range leads to less cell stress, health complications etc.- Run these trials against injectable RAA and compare the (CGM data) blood glucose excursions, show (prove) the obvious superiority, along with some A1C data... Get the FDA into this century already. It would seem like a good idea but has as much chance of happening as peace in the Middle East. There are currently numerous bodies setup to help guide standard of care. Part of these standards are based on trials (and CMG is still early in the game), part is dogma and part is simply resistance to change. Also the ADA isn't exactly known to be quick thinking... They are more like the types who laughed at Pasteur for his germ stuff! |
|
|
|
Post by mnkdnut on Jan 6, 2016 20:36:12 GMT -5
Seems to me, a LOT has been learned this past year about how to get the most benefit out of "A". Sam and his cohort, through their own inquisitiveness and CGM use, have learned and shared a lot about how to dose, when to dose, and when to dose again. The speed of "A" makes the decision of when you dose critically important to the results you can expect. Much more so than for RAAs which have to be taken much earlier, and last much longer. It appears that the FDA trial protocol took none/little of this in consideration and left patients to dose "at the start of the meal", which to no surprise resulted in a so-so comparison to RAAs. Taken too early, and perhaps taking too little, would naturally leave "A" with much less effect on reducing A1C than its potential. A properly designed study that allowed both arms (Afrezza and RAA) to be used in their own optimal way would really showcase the difference. This might also explain why we didn't notice an expected rush of former Trial patients to get "A" when it became available - a lot of them probably didn't notice a big difference based on how they were under-utilizing it!
Interesting reading the NY Times article about SNY dropping "A" last night. The author mentions the comparison with Exubera and how the size is much more discreet (whistle vs. tennis ball can), but that "Mannkind also said its insulin had some desirable medical characteristics, though there was debate on that." That's what the FDA Study and it's non-inferior results leave us with! I hope MNKD follows through on SNY's late effort to get Dr. Steve Edelman to design (and presumably execute) a Study to demonstrate "A" at its best. It's getting very expensive not to have this.
|
|
|
|
Post by mbseeking on Jan 7, 2016 2:33:15 GMT -5
There is a thread or theme emerging here which has not fully formed but I suspect might be important. Let me see if I can capture it.
To improve insurance coverage (presumably) we need a better label. The best way to get a better label would be a superiority study. But what would a superiority study show? Better A1C? Within the current medical paradigm that would be the most convincing.
But this seems unlikely. Why? Because within the confines of a clinical study people properly administering Afrezza are unlikely to get A1Cs statistically much better than people properly administering SC insulin. Why? Well , they are both insulin , when used properly. Insulin is insulin.
But isn't that the rub,, and what Sam has been trying to tell us. Afrezza is far simpler to use and more likely to give better results in the real world. Even for a T1 where diabetes is an acute disease rather than T2 where it is chronic.
Jump to conclusion.. isn't what we really need a study of a large population using this product at home with the habits / approaches they adopt everyday.. and track things like A1Cs , hypos over time , CVD events, neuropathy events, .. one group with Afrezza , one or more not. Would such a study still be a clinical or superiority study ( I suspect not). .. or is it more an observational study adding to a body of evidence supporting the efficacy of Afrezza.
If you agree with where I'm heading here.. then they're are cheaper technologies becoming available to mount these kind of studies. The new COO of Apple , Jeff Williams , launched a set of tools and methologoies a year ago called ResearchKit aimed at 'democratising' medical research. I might append: make it cheaper. Maybe the newly released MNKD could poke a few of their precious (sincerely) dollars at a university to run such a study using such tool. Five or $10 million over a few years, but started soon: surely there would be a faculty somewhere that would take that up? Interim results, peer reviewed, in 6 or 9 months published adding 'weight' to the consensus on Afrezza vs SC. It would additive to the compulsory safety studies that have to be done.
|
|
