#Afrezza or How not to do Diabetes Pharma. Can @SanofiDiabetes shed a little more light? – Updated

#Afrezza or How not to do Diabetes Pharma. Can @SanofiDiabetes shed a little more light? – Updated
#Afrezza or How not to do Diabetes Pharma. Can @SanofiDiabetes shed a little more light? – Updated

***NOTE: This post has been updated following re-reading of the SEC documents relating to the Chang Yi LIang case due to mis-reading of the data***

I’ve written about Afrezza before. It’s an interesting product from Mannkind, and whilst I haven’t any experience, it seemed like a new paradigm in diabetes treatment. But we could just be looking at the ghost of what could have been, with the news yesterday that Sanofi and Mannkind have parted ways, the short sellers have won and that Mannkind may well run out of money sometime mid-2016.

The past: FDA issues and and short selling sharks

Mannkind had been working on Afrezza for a reasonably long time, and appeared to be about to get approval in 2010/2011 (after failing to properly carry out their testing process and being rejected by the FDA in 2009). it is widely reported that they were due approval on 29th December 2010, due to this being the PDUFA date they were given for the resubmission..

Then on December 25th 2010, the FDA received a letter from everyone’s favourite Hedge Fund Manager, Martin Shkreli (who happened to have a large short position in Mannkind and disclosed it) stating that the clinical process being undertaken by Mannkind wasn’t up to scratch. Now this wasn’t an issue with whether it worked, it was an issue with not sticking to the procedural rules, and the rules are that you have to trial with the device you are going to sell.

Mannkind hadn’t done this properly, and it had been picked up on. Whatever you might think of the intentions of Shkreli, this was also an error by the management team at Mannkind.

There are a number of unsubstantiated statements that in a face-to-face meeting, the FDA had agreed that the two devices in question were considered bioequivalent and that this wasn’t an issue, however I have been unable to find factual corroboration of this and raises the question as to whether Mannkind should have sought documentary evidence of such an agreement. It also seems to demonstrate that they were not fully aware of the short sellers sitting on their door step, and the extent that these guys would go to, and that could be considered a fault of their corporate brokers or internal investor relations team.

There are questions as to whether there had already been a planned approval that this stopped, and the actions of the FDA scientist, Chang Yi Liang would seem to confirm this.

On January 4th 2011, Chang Yi Liang sold Mannkind shares after the PDUFA date failed to materialise (page 25 of the linked document) – we assume originally he had seen and approval in the in the FDA database, bought shares, and when the approval failed to materialise, sold and avoided losses. He has since been found guilty of using insider information to make a lot of money on multiple stocks.

Some might say this was Shkreli’s attempt to drive the company into insolvency and cash in on those short positions. You can certainly believe that, given his hedge fund’s behaviour, and the letter, whilst saying that he had a conflict of interests in writing the letter, doesn’t really declare that he was hugely short and the associated drop in stock price with a delay would make him a lot of money.

I’ll leave you to make your own mind up on the altruism involved. As the CRL letter included detailed items such as more information about performance characteristics, usage, handling, shipment, and storage of the Dreamboat inhaler, an update of safety information related to Afrezza and information on proposed user training and changes to the proposed labelling of the device, blister pack, foil wrap, and cartons, and Chang had already bought Mannkind, one finds it hard to draw conclusions other than the CRL was caused by this letter, especially in light of similar behaviour with other products with the FDA, and mainly because the FDA felt it had to keep it’s reputation intact. 

The Class Action that swiftly followed, suggests that Mannkind’s submissions had not fully substantiated what was required for the FDA approval and that there were misleading statements from the Mannkind management team, but without those submissions, it’s a difficult one to corroborate. 

In the class action, there is a statement to the effect that Study 142 actually demonstrated that the two devices fell outside the bioequivalence requirement of the FDA and they had no choice but to issue the Complete Response Letter (CRL). That Mannkind agreed a settlement with the investors lends some credence to this point of view, (even with the Shkreli letter).  

It all suggests that there is a level of blame to be apportioned to each of the FDA, Mannkind and Shkreli, each for different items:

  • The FDA for suggesting the process could be side-stepped and not being squeaky clean all the way through; 
  • Mannkind for not getting this in documentary form and submitting studies that would refute this issue, basically leaving the door ajar;
  • Mannkind for not understanding the investor base – not having knowledge of the shorting behaviour is pretty naive and something that large company CEOs are very aware of; 
  • Shkreli’s team for trying to drive stock prices down for profit in MedTech (behaviour that they had regularly done).

Third time lucky with the FDA

Fast forwarding to 2014, after the New Drug Approval had been resubmitted for a third time, and reading through the FDA Committee Meeting documents from 2014, there seem to be a reasonable number of reviewer comments that highlight low levels of data, enough to require significant labelling overheads, plus adverse respiratory events that led to the need to undertake lung test before prescribing. 

Ahead of the meeting, some analysts were questioning whether it would be approved based on the reviewer comments that were publicly available.

It suggests a level of naivety on the part of the programme managers at Mannkind with respect to what they needed to produce to get the product signed off. Likewise, the comparisons with injectable insulin could have been done a lot more effectively and would have provided better marketing material. The full transcript of the meeting is in the link. I’ll admit to not having read the entire piece as there are 500 pages of it, but there are some difficult questions raised in the session, even after 154 pages!


Even after all these delays, it was eventually approved in 2014 and a license agreement with Sanofi was agreed, in order to broaden the range and usage of the product. In theory, at least, it was supposed to be a global product, with Sanofi (a diabetes expert, remember) marketing it (and presumably seeking non-US regulatory approvals) after success in the US. But at the same time, Sanofi paid only $150mn upfront to acquire the license to do this. In Pharma deals of this type, these end stage agreements usually have much larger numbers attached to them. This suggests a lack of interest from the big pharma players in what some saw as a revolutionary new treatment. Why could that be?

The required pulmonary tests weren’t cheap and weren’t easily accessible in the prescribing location. A simple strike against Afrezza, without even trying, upping both the complexity and cost of getting a prescription.. The fall out from Exubera (Pfizer writing down $2.8bn) will have weighed heavily on the big players and driven many of the issues they saw. The requirement for a lung function test will also have weighed on the decision of potential patients to look at it. Injecting insulin isn’t going to cause lung problems and the generalised risks of inhaling anything abnormal are pretty well known amongst the wider population. Anyone who had heard about Exubera was likely to have concerns.

At last, a product on the market

Once Sanofi had bitten the bullet, and jumped on the Afrezza bandwagon, reasonably big things were expected. Sales of around $600mn annually were estimated. But the launch was slow in early 2015 and by August there had been little advertising, 35% of US doctors hadn’t heard of the product (a year after it had been given approval) and sales had reached only $3.3mn.

In addition, the cost of Afrezza was significantly more expensive than existing injectables (even after the short term deals expired), something that would always make the insurance companies in the US sit up and say “No way”. The partnership with Sanofi was supposed to improve the production costs of the product which would have gone some way to minimising this.

And yet the buzz on social media and amongst the patient community was very loud. People wanted to know about the product. The utopia of not injecting possibly? I don’t recall seeing anyone from Mannkind or Sanofi join in and spread the word. I don’t recall seeing anything relating to sign off in Europe, Asia or Australia. What were Sanofi doing? To be at $3.3mn versus what should have been $300mn expectation is astonishing.

In the meantime, Sanofi had been marketing Toujeo widely and obviously sold a competing product in Apidra (Insulin Lispro), a different fast acting injectable. Could this have been the reason?

Something was clearly amiss, and I’d originally thought to point the finger at Sanofi being not as interested about it as we’d like to think. Of course the result is that yesterday Mannkind and Sanofi parted ways. 

What went wrong?

Well World of DTC Marketing has undertaken an interesting autopsy of the result of this failure and for me there are a few key points:

  • Failure of basic research to uncover basic value communication: What were the fundamental selling points of Afrezza, and to whom… Was this really never done?
  • Failure to do a head to head clinical study against insulin products, especially when it comes to compliance and A1C control: This isn’t just a no-brainer, it was identified in the reviewers comments in the FDA docs as an issue!
  • Failure to convince physicians that “lung capacity” would not be an issue with a majority of patients and to use science to back up those claims: No-one managed to get the doctors onboard, especially after the issues with Exubera. Memories are long, especially bad ones…
  • Failure to take advantage of the Social Media buzz: and there was plenty, but I don’t recall seeing anything official.
  • Failure to let Sanofi take over the marketing of the product: Now this is an interesting one. I was all for blasting Sanofi’s marketing of the product. They know how to do Diabetes. Lantus is the biggest selling Diabetic treatment on the planet with $8bn annual sales. But maybe the marketing was nothing to do with Sanofi, and that’s really where the problem lay.

Looking back through it all, it seems as though the product regulatory approval, launch and subsequent marketing may well have been worthy of a text book on “How not to do diabetes pharma”, which is a pity, as the product shows great promise. There appear to be bad guys and good guys within the story, it is enacted over a long time and the plucky hero may yet come back and win the heart of the damsel. But it isn’t going to be cheap. Someone is going to need to invest a lot of money to keep the ball rolling, and with the share price of Mannkind where it’s at, it’s unlikely to be outside investors.

As a footnote, I hope this isn’t the end of Afrezza. I’d like to see alternatives like this funded, but I think this may be required in privately owned firms, where the share price can’t be manipulated for gain and there is no point in making reports to the FDA just to burn money, time and the share price (whether it had any impact or not).

Who knows, but without some of the shenanigans, would Afrezza now have four years under its belt and be a much more widespread treatment? Or would the ghosts of inhaled insulins past have stymied progress equally badly five years ago, as they were, if anything more recent in memory?


Be the first to comment

Leave a Reply

Your email address will not be published.


*