Inhaled Insulin's Long Journey to Commercialization

Published on: 
Evidence-Based Diabetes Management, May 2014, Volume 20, Issue SP8

What if patients with type 1 diabetes mellitus (T1DM) could be freed from needles, vials, and syringes? What if insulin didn’t have to be injected, but instead could be inhaled, like an asthma medication? What if a new technology considerably lowered their risk of hypoglycemia?

With the exception of a true artificial pancreas, this technology may top the wish list of persons with T1DM. Previous attempts at inhaled insulin won over the FDA but not the target audience, causing some pharmaceutical companies to reevaluate their commitment to this product. “The prospect of a viable inhaled insulin is exciting, but the patient community feels a bit jaded because of the experience with Exubera,” said Amy Tenderich, founder and editor of the website Diabetes Mine (

Indeed, a commercially viable inhaled insulin has turned out to be difficult to develop.1 Yet, one company may be getting close.

Pfizer’s 2006 Approval and 2007 Withdrawal

Earlier attempts to market an inhaled insulin and its delivery system revealed the difficulty in reaching the goal: not only must the product deliver insulin to the body, it must also be practical. An insulin product must be easy to use and small enough to carry around for use when needed.

Results have been mixed. Pfizer achieved FDA approval of its product Exubera in 2006 with considerable fanfare and lofty sales expectations of $2 billion a year, but then withdrew the product a year later, taking a $2.6-billion corporate charge.

Pfizer’s Exubera was successful in clinical trials. The real world was another story. Exubera did deliver insulin deep into the lungs. However, the delivery system was 1 foot long, ungainly, and much too large to fit into a purse. Dose adjustment was time consuming (selection of a specific dose was not possible) and required more patient education than Pfizer anticipated. A series

of blisters needed to be inserted into the device, followed by activation of an air pump. Needless to say, this was not a discreet activity.2

Patients with T1DM avoided using it, but providers also avoided prescribing it. Mark Feinglos, MD, professor of medicine, and division chief, endocrinology, metabolism, and nutrition at Duke University, commented, “What really caused Exubera’s failure was that in order to put people on it you had to explain to them how to use it very carefully, and pulmonary function tests had to be done. It was very labor intensive.”

The average primary care physician had too little time to invest in this new therapy. Pfizer soon realized the magnitude of the gap between sales expectations and reality, and Exubera was withdrawn at the end of 2007. To complicate matters, a connection (somewhat tenuous) was drawn between use of Exubera and subsequent cases of lung cancer. In 2008, 6 cases of lung cancer were reported in patients who used Exubera (and also smoked) versus 1 patient using a control inhaler. Although the numbers were not statistically significant, and causation has not been established, the possibility of a connection between lung cancer and inhaled insulin made later efforts to popularize inhaled insulin challenging.

Some researchers believe inhaled insulin is mediated through IGF-I receptor activation by high local concentrations of inhaled insulin that become lodged in bronchial tissue.3 Pfizer’s experience helped scuttle the developmental work of 2 other manufacturers (Lilly and Novo Nordisk) whose inhaled insulin technologies were in late-stage trials1

Afrezza on the Verge


MannKind, which was also working on its inhaled insulin product Afrezza, continued efforts and filed a new drug application in March 2009 for use with its first-generation MedTone inhaler.

Manny Hernandez, president of the Diabetes Hands Foundation as well as @AskManny, told Evidence-Based Diabetes Management, “I think it is fair to say that the Exubera failure made the demands on and the expectations for Afrezza higher. We want to make sure that the new drugs are safe and effective for patients.

“Judging from the form factor of the Exubera inhaler,” Hernandez continued, “it appears to me like [Pfizer] didn’t do enough market research, because Exubera was inconveniently large, drawing even more attention to the user than an insulin injection would. I think Afrezza has addressed this fairly well with its inhaler.”

Afrezza uses an ultra-rapid-acting formulation, and it showed promise in initial clinical trials. It comprises insulin formed into particles 2 to 3 μm in diameter, which are then lyophilized into a dry powder for inhalation. The insulin used is a type of regular human insulin. Unlike regular insulin, the dry form does not require refrigeration.

Afrezza proved to be at least equivalent in combination with insulin glargine to multiple daily injections with an intermediate-acting insulin and basal insulin.4 Additionally, it did not interfere with pulmonary function.5 However, the FDA issued a complete response letter requiring additional information.

MannKind resubmitted the application in early 2010 (but this time with a smaller, second-generation inhaler called Dreamboat). FDA rebuffed the manufacturer again in January 2011, requesting 2 additional clinical trials.6 This time, the FDA also required comparisons of Afrezza’s Dreamboat inhaler with the MedTone Inhaler, to better understand whether clinical studies using the original inhaler could be applied to FDA decision making with the new inhaler.

The new “bridging” study would assess not only “performance characteristics, usage, handling, shipment, and storage,” but safety information and proposed user training.6 The 2 clinical trials posted positive results. The first study evaluated the inhaled insulin in 2 inhaler types (its older MedTone inhaler and newer-generation Dreamboat inhaler) compared with rapid-acting insulin aspart mealtime injections, in a 12-week observational, open-label study design for 518 patients with T1DM. All patients were first optimized on a basal insulin regimen.

The study results established similar levels of glycated hemoglobin (A1C) reduction for both insulin aspart and Afrezza and its next-generation inhaler. However, Afrezza was associated with a reduction in fasting blood glucose (FBG) levels, whereas aspart was associated with increased FBG concentrations. The frequency of all hypoglycemia events favored Afrezza and its next-generation inhaler, but there was no significant difference in severe hypoglycemic events. Patients experienced a mean weight loss of 0.4 kg in the Afrezza Dreamboat group, compared with a 0.9 kg weight gain in the insulin aspart group.7

Next, Afrezza was tested against oral therapy alone in a double-blind, placebo-controlled study of 353 patients with type 2 diabetes mellitus (T2DM) whose disease was uncontrolled on metformin with or without a second or third oral medication. Participants were given their oral therapy in addition to either Afrezza or an identical inhaler with placebo microspheres.

Over 24 weeks of treatment, Afrezza therapy was associated with significantly lower mean A1C concentrations (0.82- vs 0.42-point reduction in the inhaled insulin and oral therapy groups, respectively [P <.0001]) and 30% of the Afrezza group reaching A1C levels of 7.0% or below compared with 19% of the oral therapy group (P <.0005). Sixteen percent of those using inhaled insulin attained an A1C level of no greater than 6.5% compared with 4% of those assigned to receive oral therapy (P <.0021).

Patients receiving Afrezza had a higher rate of mild to moderate hypoglycemia (but no patient discontinued as a result of this side effect) and experienced a mean weight gain of 0.5 kg.8

In October 2013, MannKind filed its third new drug application for Afrezza. The FDA’s internal staff compiled an internal report indicating that some problems still existed and that claims of noninferiority to insulin aspart could not be supported. However, on April 1, 2014, the FDA’s Advisory Committee voted 13-1 to recommended approval for its use in T1DM and 14-0 for use in T2DM.9 The FDA then delayed the product’s prescription drug user fee act date (PDUFA) from April 15 to July 15, 2014, the deadline for the latest (or final) chapter in MannKind’s long, difficult

trek to approval.

However, as MannKind spent nearly $1.5 billion on the journey to approve Afrezza, industry consultants doubt that it has the capital to market the product.9,10 Undoubtedly, MannKind will need a marketing partner, and with some major players withdrawing from the inhaled insulin market, the company may find this challenging. Even if it does receive approval and find a commercialization partner, MannKind shouldn’t expect patients to run to therapy, said Tenderich, who has T1DM. “With any new formulation, we don’t really know what the long-term effects will be, like how it may affect the lungs,” she said. “The FDA is doing its best to mitigate those concerns, but it may take 5 to 15 years of real-world experience to find out how safe it actually is.”

She also pointed out that for patients such as herself with T1DM, “We need more precision dosing for the most part, matched to the amount of carbs we eat. It’s hard to believe that we can get this precise level of dosing, but we’re being told otherwise.”

An Experienced Dancer in the Wings

A relatively new (but not that new) player in the inhaled insulin development drama is Dance BioPharm. Started in 2010 by members of the team involved with Pfizer’s Exubera, Dance is working on a pocket-sized electronic insulin inhaler. This product (dubbed Adagio) has finished phase 2 trials (its clinical trial information has not been publicized) and is awaiting the beginning of its pivotal phase 3 testing. According to Dance’s website, insulin is provided as a liquid not a powder in Adagio, and drops are deposited into a reservoir on the top of the device for use at mealtimes.

Evidence-Based Diabetes Management contacted John Patton, PhD, CEO of Dance BioPharm, but he indicated that he could not comment because his company was preparing for an initial public offering.

If You Build It, Will They Come?

Should it be approved, MannKind believes that its product will retail for somewhere between $100 to $200 per inhaler. From the patient’s point of view, Tenderich believes that the needleless attribute of inhaled insulin may be a bit less attractive than proposed. If this product is to be used in patients requiring basal insulin, a daily injection is still needed, “and for most of us with type 1, the needles are so tiny, and they aren’t the biggest issue,” she said. “This may be more appealing to the type 2 patient, for the crowd who is worried about injecting,” Tenderich said. “In any case, patients are pretty savvy these days, and they will want to know about the side effects, and of course, whether their insurance will cover the therapy, and how much will it cost them.”

Feinglos agreed that Afrezza may be a better bet for patients with T2DM, but for a different reason. “The absorption is a very small percentage of what you give, so it’s really good for people with type 2 diabetes who need a boost in insulin,” Feinglos said. “But for someone with type 1 diabetes who happens to be very sensitive to insulin and needs to make very small changes in dose, it might not be the right thing for them” despite the advantages of the inhaler over injections.

He pointed out that providers have concerns with Afrezza that are similar to those with the original approval of Exubera. “The education and testing challenges for the physician’s office remain,” reminded Feinglos. “But how do you get around the problem of physicians actually prescribing it? Again, remember that most people with diabetes are cared for by primary care physicians, not by specialists. And you know very well how limited is the amount of time that the primary care physicians have. And not all of them have educators in the practice, not even group practices necessarily have educators.

So how do you get people onto a drug like this and how do you monitor them successfully? That’s where the problem may well come up again.” References

1. Kling J. Dreamboat sinks prospects for fast approval of inhaled insulin. Nat Biotechnol. 2011; 29:175-176.

2. Heineman L. The failure of Exubera: are we beating a dead horse? J Diabetes Sci Technol. 2008;2(3):518-529.

3. Mayer P, Reitzenstein U, Warnken M, Enzmann H, Racké K. Insulin action on H292 bronchial carcinoma cells as compared to normal bronchial epithelial cells. Pulm Pharmacol Ther. 2012;25(1):104-114.

4. Rosenstock J, Lorber DL, Gnudi L, et al. Prandial inhaled insulin plus basal insulin glargine versus twice daily biaspart insulin for type 2 diabetes: a multicentre randomised trial. Lancet. 2010;375:2244-2253.

5. Raskin P, Heller S, Honka M, et al. Pulmonary function over 2 years in diabetic patients treated with prandial inhaled Technosphere Insulin or usual antidiabetes treatment: a randomized trial. Diabetes Obes Metab. 2012;14:163-173.

6. MannKind Corporation receives complete response letter from the FDA for AFREZZA(R) [press release]. Valencia, CA: MannKind; January 11, 2011. ID=1517413&highlight=. Accessed April 22, 2014.

7. Mannkind reports positive data from a phase 3 clinical study of Afrezza in patients with type 1 diabetes [press release]. Valencia, CA: MannKind; August 14, 2013. Accessed April 23, 2014.

8. MannKind reports positive data from a phase 3 clinical study of AFREZZA in patients with type 2 diabetes [press release]. Valencia, CA: FierceBiotech; August 14, 2013. Accessed April 22, 2014.

9. Pollack A. Inhaled insulin clears hurdle toward F.D.A. approval. The New York Times, April 1, 2014. Accessed April 28, 2014.

10. Carroll J. Questions on MannKind’s inhaled insulin PhIII data cloud Afrezza’s future. Fierce Biotech; August 15, 2013. Accessed April 28, 2014.