Memorandum

MannKind receives second complete response letter from FDA, requesting two additional clinical trials – January 19, 2011

Executive Highlights

  • MannKind announced its second CRL for the company’s inhalable insulin, Afrezza.

In another setback for MannKind, the FDA issued a second complete response letter late yesterday (1/18/2011) for the company’s inhalable insulin, Afrezza. The FDA has requested two additional randomized controlled phase 3 clinical trials (one in type 1 patients and one in type 2 patients) of the next-generation inhaler (Dreamboat). The letter specifically stated that the design of these studies should allow for 12 weeks of “relatively stable insulin dosing” after the titration period. The agency also recommended including a treatment arm using the MedTone inhaler (the old inhaler) to provide a direct head-to-head comparison of the two devices. Presumably, the in vitro data and clinical pharmacology data previously submitted was deemed insufficient – during Q&A, management noted that through discussions with the agency, they inferred that handling evaluations and bioequivalence data would suffice for a resubmission. These proposed studies of the Dreamboat device will be the first phase 3 trials of the next- generation inhaler. The complete response letter also requested further information on the performance characteristics, usage, handling, shipment, storage, and routine safety updates on Afrezza. In addition, the FDA has asked for additional details on the proposed user training and changes to the proposed labeling of the device, blister pack, foil wrap, and cartons. On some level, the fact that FDA requested such detailed information makes us think this is a “when” not “if” – on the other hand, FDA’s behavior has become increasingly hard to forecast.

As a reminder, the company has already initiated two treat-to-target studies of the next- generation device in patients with type 1 (Affinity-1) and type 2 (Affinity-2) diabetes. Interestingly, these studies closely resemble those outlined in the complete response letter, with a duration of 12 weeks (at stable dosing); however, MannKind would have to modify these trials to include a MedTone arm. The FDA did not specify the size or design of the trials in the complete response letter. The company is planning to request a meeting with the FDA as soon as possible to confirm the design of the two clinical trials. Beyond anticipating that the end-of- review meeting with the FDA would be held within roughly 60 days, management did not provide any timelines for resubmitting Afrezza. Given that Affinity-2 is expected to complete in December 2011 (and Affinity-1 has not yet started), we believe the earliest MannKind will be able to formally resubmit Afrezza is in early 2012, assuming the FDA accepts the ongoing Affinity trials and allows the company to incorporate a MedTone arm into at least one study (positioning Afrezza for a third PDUFA date in 2H12). We look forward to a more detailed update on the MannKind’s future plans during the 4Q10 financial update in early February. Overall, we found this to be a measured, informative call. We view the requests as fairly reasonable overall. We believe the FDA risks looking unreasonable if they receive the information requested and then request additional studies. It is very positive in our view that the additional studies requested relate to a device rather than Afrezza itself.

  • MannKind’s two-phase 3b treat-to-target studies, Affinity-1 and Affinity-2, are currently ongoing. The clinical sites have been initiated and Affinity-2 has enrolled a “small number” of patients. While no patients are currently enrolled in Affinity-1, the investigators have begun screening patients. Prior to initiation, MannKind submitted the protocol of both studies to the agency last fall. The treat-to-target goal for both studies is an A1c of 6.5% or less. Management believes that targeting roughly 143 completers in each treatment arm will adequately power the studies. During Q&A, management said that the majority of patients would be managed with either one or two cartridges at the highest dose (20 units) of Afrezza. While there was brief mention of the possibility of including other centers and countries into this phase 3b program to accelerate the trial or accommodate any additional patients, we strongly hope the company will not pursue anything that could add risk to the Afrezza development program or the regulatory process. We assume enrollment would not be a major issue.
  • Presumably, the head-to-head comparison between the Dreamboat and the MedTone C device will be a comparison of the pulmonary safety data. During Q&A, management referenced “very short-term” studies that demonstrated a reduction in the pulmonary function decline following administration with the Dreamboat device compared to the MedTone device. Management speculated that the FDA may be asking for clinical data to confirm this reduced effect on pulmonary function. Regarding manufacturing MedTone, although MannKind does not have a full assembly line, it has “some” assembled devices and “lots” of parts. CEO Al Mann stated that the company will initiate manufacturing efforts immediately and should have enough materials to manufacture a sufficient number of MedTone devices without a delay in the clinical trials.

Questions and Answers

Q: It seems as if FDA disagreed with the approach more than the data. Could you comment on whether you think you misread FDA last June or whether you think something has changed from that time until now?

A: I think that is a very difficult question to answer. It’s something that we can really only speculate on. Certainly, we believe the science of the approach that we took was valid and the guidance we had previously received from the agency was that bioequivalence for device substitution could be appropriate. We know they have diligently reviewed this and have clearly made a decision in terms of what they expect to bridge the new device to that used in the pivotal clinical studies.

Q: I know there is some range of possibilities with regard to modifying existing clinical trials and seeing if FDA would be accepting of that, but could you give us some range for when you think you would be generating the kind of data FDA is looking for and when you would be in a position to refile?

A: We’re already collecting that data. The trials are still early, but I don’t think we can say anything specific. The trials that we are conducting, the language that we used in our protocols is almost the same thing that they are asking for it now, but they are asking for a cohort of the MedTone inhaler to compare directly to the device.

Q: What do you think would be the range of timeframes when you could be in the position to resubmit? Do you expect the six or 10-month review? And how do you expect to finance the company?

A: We can’t comment on that yet, we’re not ready. We’ll give you more information in our quarterly conference call early next month.

Q: The main issue I’m struggling to understand is the following: Has the FDA told you that they are satisfied with the safety and efficacy in the trials that were done with the MedTone device. If you end up doing these two new phase 3 trials and show that the Dreamboat is equivalent to MedTone, could the FDA then come back and say that Dreamboat is equivalent but there are issues with the levels of safety and efficacy you have shown? Have they commented on this at all?

A: No, but the complete response letter is a complete response and is supposed to list all of the questions. The wording of the request for the new trials suggest that they are comfortable with the safety, efficacy, and clinical utility of Afrezza using the MedTone inhaler, but that we have not done enough to show comparability of the next-gen inhaler with the old inhaler and that’s why they’re asking for new trials.

Q: I understand that, but if we go by what the last complete response letter said, they said they wanted information on the comparability of the commercial version to the earlier version and it seems like they have gone back on what they are asking for because you did meet with them in June...

A: No, no. That was not in the earlier CRL.

Q: But they said that they did not require any new clinical trials…

A: Yes, but they didn’t say anything about the next-generation inhaler.

I think that’s correct because that was the bioequivalence between the two inhalers, and we have submitted a bioequivalence study around that. We’ve also had extensive discussion with the agency in terms of the change to the assay methodology. Just to clarify the wording of the complete response letters, they’ve not asked us for anything further in terms of efficacy or clinical utility. I would see that as reassuring. What they have asked for are two relatively short (12 week) studies and they’ve asked us to ensure that we have relatively stable insulin dosing over that period as well as an adequate titration. That’s what we’re doing in the treat-to-target studies. Our anticipation is that these studies will continue to have a comparator of a rapid acting analog but we would choose to add MedTone C into one or both of the studies and look at the most optimal design. That’s why we can’t really give you hard and fast times because we have to discuss this with the agency.

Q: Did they specify the size of the trials in terms of the number of patients?

A: No, they didn’t. We’ll obviously address that with the agency. The studies that we’ve done are adequately powered knowing the variability of A1c we’ve seen in studies today.

Q: Have you asked for a meeting with the agency yet?

A: We will be doing so shortly. There’s obviously a process there and we need to make sure we ask the right questions. We would anticipate having an end-of-review meeting within 60 days or thereabouts.

Q: Regarding Affinity 1 and 2, I see the type 2 trial on clinicaltrials.gov but I don’t see the type 1 study. Has that one also started?

A: They have both started and there are sites initiated. We have a small number of patients in Affinity 2 and no patients in Affinity 1 as yet, although they are starting screening.

Q: When you think about the head-to-head comparison between MedTone and Dreamboat, what happens if they’re different? Even if Dreamboat is better than MedTone, does that potentially send you back to drawing board for longer term safety data?

A: The technosphere powder that goes in the blood is the same with both devices. The bioequivalence trial demonstrated that. There is nothing more that we really need to do to show it. If there is a difference, it would have to be due to some strange quirk with the patient selection.

Q: But if the FDA is asking you to run trials comparing them head-to-head in patients, does that imply that if there is any difference, they could come back to you to ask for longer term data?

A: I think that’s important in terms of the efficacy endpoint and anchoring it to a comparator of a rapid-acting analog. The head-to-head comparison with the MedTone C will specifically be a comparison of the pulmonary safety data. I will be interested in terms of that head-to-head comparison in terms of what we’ve seen in very short-term studies, where there is actually a reduction in the pulmonary reduction change acutely following administration from the new generation device compared to MedTone. So we anticipate this advantage but I think the FDA is seeing if we can confirm that in the longer-term study.

Q: Are you able to tell us what your year-end cash position?

A: We’ll be able to tell you more in less than two weeks. We’ll probably be providing a lot of additional information at that point because it will allow us to digest this a little further. We have said that we had cash through 3Q10 and that was with current projected spending, which by the way did include these trials.

Q: You indicated that FDA sent a complete response yesterday. Just interested why you didn’t disclose the news this morning. Obviously, it was a pretty volatile day with the stock and ended up having to be halted.

A: Well, first, we didn’t get it until late yesterday. We had to read it, understand it, and prepare a response, and that took several few hours.

Q: On the comparison to MedTone C, could you give us a little more color on what your experience has been with the MedTone device in terms of timeline of effects on pulmonary function and whether you think you will be able to demonstrate equivalence in terms of lung safety over a 12 week period?

A: First of all, this is specifically what has been asked in the CRL, so we can read into that if you wish but that is the specific timeframe for this from the agency. We have demonstrated in all of the studies that the small change that we do see with MedTone is completely apparent by three months, or 12 weeks, and does not progress thereafter. That is not the case, or as apparent, with Dreamboat.

In the short-term trial that we have done, we saw a much smaller, clinically negligible, amount of pulmonary function impact. And in the little study we have done with Dreamboat so far, it seems to be a small fraction of that amount. So I think what’s triggering the FDA’s request is that they would be pleased to see there is effectively no pulmonary function impact with Afrezza, or at least something that’s hardly noticeable.

Q: Could you give us an idea of whether there is any additional cost in terms of manufacturing to add MedTone in these trials?

A: No, we already have enough material to assemble enough MedTone devices to do this trial.

Q: I understand that the new inhaler requires less powder. As a result, I would think the FDA will be relatively happy to see it used appropriately, giving less of a pulmonary effect. I was wondering what exact measurements do you think you will take to follow efficacy and safety in the comparison study?

A: Just A1c and pulmonary function studies.

The agency has been fairly clear in terms of A1c. They have not specifically said in terms of pulmonary safety data. Usually, it would be standard pulmonary function testing. It may be a question we want to discuss with the agency, but we have already planned quite extensive pulmonary safety data assessment in the ongoing studies and we would anticipate that would cover any question they would have around pulmonary safety.

Q: So you don’t expect to bring the enrollment to a halt prior to discussing the CRL with the FDA?

A: I think we want to spend a little more time looking at our plan there. Having the opportunity to speak with the agency and get clarity with the agency is the primary course. But we can see what information we can get from the ongoing studies and see how we can use that, so I think that is a decision we have to make.

Q: So you will basically halt the studies temporarily at least?

A: No, we’re not going to.

Q: In the event that you have to manufacture additional MedTone devices, how long is that going to take, is that an easy switch for you to do? If that ends up taking longer than you expected, should we assume that you would have to wait at least until 2Q11 to start the trials?

A: The trials have already been started. Secondly, we have enough material of the MedTone parts to include that very quickly, we just need the proper introduction into the trial protocol.

Q: If your trials show that the MedTone device is superior to Dreamboat, would you be in a position to move back to the MedTone device as your primary choice for commercialization?

A: First of all, the answer is that it won’t be superior. The Dreamboat device is clearly superior in many ways.

Q: How would it work to add a new arm to an already ongoing trial? Can you talk about how that may work? Is that something easily done?

A: I think that we started the discussion in terms of that. It depends on what stage you are at in the protocol and the number of patients that we have and also the methodology that we use. There are several methods we’re considering and when we go to the agency, we’ll have a more clear rationale and will look at the appropriate risks prior to meeting with the agency.

One thing that we could do, as we are enrolling patients, would be to randomize them, assigning the group for the MedTone, even if we delay the start of the actual usage and catch it up as we accelerate it later.

The thing to ensure is to avoid any bias in the study and to make sure that the comparisons in those patients in the MedTone arm are matched correctly. That’s an element of the study design that we’ll certainly be taking into account.

Q: After you meet with the FDA, I know last time they didn’t give you the minutes back from the meeting and you went ahead and filed your second filing. Would you wait on the minutes before you started clinical trials and what if they delayed the minutes again?

A: We are continuing with the clinical trials now. The trials we submitted to the agency last fall, Affinity 1 and 2, we are conducting. From the wording of their discussion, it appears to be that they almost want to see the results of the trials that we already submitted to them and the only difference is that they want us to add a MedTone cohort.

Q: So if they delayed giving you the minutes, that won’t affect your progress at all?

A: We don’t really think that will be an issue.

Q: On the titration phase, the press release says it will allow for adequate titration so that you’re at stable dosing for 12 weeks. Can you comment on how many weeks of titration that will likely require and is it not the case that in your prior studies, the dose was increasing throughout the entire duration of the study?

A: We anticipate the titration period being four weeks and that is what we have in the protocol. I think the point is to try and avoid the dose changing or increasing through the treatment period. By using forced titration design, which is the element that we have in terms of treat-to-target, that really overcomes that particular issue.

Q: So you’ll titrate for four weeks and then keep a stable dose in both the TI and the control arm?

A: The word that the agency has used is “relatively stable” because when you are having acute measures of glycemic control and changes, patients circumstances may change, and you have to allow an element of titration according to those measured, but I think the agency has used that word carefully.

Q: But that was not the case in your prior trials, correct?

A: The previous trials, as we’ve said, were not treat to target with forced titration algorithms, so they were different designs.

Q: And both Affinity 1 and 2 have a four-week titration period?

A: Yes they do. In fact, they have a titration period for the rapid-acting insulin and the long-acting insulin.

Q: What is the goal that patients are titrating towards in the Affinity trials?

A: A1c, and the design is non-inferiority compared with the comparator. But the treat-to-target goal is to get patients to 6.5% or less.

Q: What dose do you expect that will require with the Dreamboat inhaler?

A: I think that will depend upon the individual patient and should be no different from the dosing that we’ve seen in previous studies.

Q: And what has the dose been in previous studies?

A: We’re anticipating the majority of patients will be managed with either one or two cartridges at the highest dose.

Q: And what is the highest dose unit volume?

A: 20 units.

Q: Why don’t you get the MedTone inhaler approved?

A: There are a number of reasons. First of all, the agency was not satisfied with the insulin assay methodology that was used. In order to get approval of MedTone, we would have had to re-do that trial to show that MedTone C and D were equivalent. Secondly, they were not comfortable with the marking on the cartridges and what they wanted would have been virtually impossible to do. Thirdly, frankly, in my view, it would have been less safe to introduce a product with MedTone and then very quickly after, introduce a new device that requires only 2/3 as much powder. That would have been confusing to the patient. So we and all our partners didn’t want to launch with the MedTone, we wanted to launch with Dreamboat. In response to the first CRL, our plan was the to move forward immediately with Dreamboat.

Q: Coming back to the A1c goal, how do they know that they’re in goal in that short period of time? I thought A1c took a lot longer to reach steady state?

A: You titrate according to the seven-point glucose, which is the standard methodology, as well as the fasting glucose. So there are various short-term goals. The A1c, as you say, represents the average glycemic change over the preceding 12 weeks, which is why the 12-week period of relatively stable dosing is important.

Q: For the measurements that have shown a fasting glucose benefit, have you measured insulin and c-peptide levels that correlate to the time those blood glucose values were drawn and if so, what was found?

A: No they haven’t been measured at the same time.

Q: So how do you know there isn’t an insulin depot effect in the lung that is causing a prolonged release of insulin over time? Did the FDA have any questions on that?

A: They haven’t had any. We’d be speculating in terms of that, but the best data going against a depot effect in the lung are studies that show quite rapid clearance from the lung. We also have data looking at circulating insulin levels and the time following that. We see them returning to baseline and zero in type 1 patients not on basal insulin, within a few hours, so there is no evidence of a depot effect.

Q: Is there anything in the second CRL that was also in first CRL or are the FDA’s concerns completely distinct and separate? Is there any overlap at all?

A: None at all.

There are some relatively standard areas where the agency has maintained consistency regarding the nature of the REMS we will have and there are really quite a lot of things that are essentially the same in terms of the safety update. These things are quite standard in a complete response letter, so if those are what you are referring to, there has been no change in those areas.

Q: I was just referring to the concerns that you mentioned on this release and on this call. Were any of the concerns in the first CRL repeated in the second CRL?

A: No, specifically no mention of clinical utility.

Q: You have cash until 3Q11. Does that assume your stock price stays above $6.50?

A: No, that doesn’t assume any additional inflows of cash.

Q: Is the FDA requiring that the next-generation device look superior to MedTone on lung function and what would the labeling or REMS outcome if it looked the same as the MedTone device?

A: That would be considerable speculation. They’re certainly not requiring or mentioning anything about superiority, their wording was head-to-head comparison. I would anticipate any differences would be represented on the label.

We had submitted data to the agency that showed a lower pulmonary function drop with Dreamboat and they may want to see that as a positive signal and not as a concern.

Q: At this stage, why would you try to retrofit an ongoing study to meet the new CRL rather than going down with a blank page and asking what can be done, design something that would meet it, and may be take a hit on timing, but be at a lower risk position?

A: That’s something we’re carefully assessing. The point is that the study designs we have mentioned are actually very similar to that which is being asked of us by the agency. So, we would be proposing to the agency many events and study designs based on our own assessments. That’s how it’s done, they don’t tell you how to do it. And these are our very best efforts at doing treat-to- target forced titration designs. And the timeframe the agency has chosen is exactly consistent with what we think in our previous studies.

Q: You mentioned Affinity 1 and 2 would have appropriate power in terms of size of the population enrolled. Can you give us a sense of the enrollment that would be needed for Affinity 1 and 2? You said you’re just beginning enrollment; what is the typical timeframe for enrolling this size of a population?

A: Until we know any details of protocol, I would be cautious about providing any timeframes. For the study, which has been pointed out, you can look on clinicaltrials.gov, but we’re looking for about 143 completers per arm. At the moment, they are two arm studies. The recruitment period tends to be in the range of months. Clearly, there are changes we can make to that, such as bringing on other centers and countries, and those are things we need to carefully consider to make sure we get the right result as opposed to the speed question.

Q: You mentioned you have enough of the current MedTone in-house to use them immediately for a clinical study, is that correct?

A: We don’t have a full assembly. We have some assembled devices but we have a lot of parts, so we have to assemble enough and we’ll initiate that effort right away.

Q: Is there any assembly line issue or potential for a delay there?

A: I don’t imagine so. The message we got from our team was that there should be no problem adding a MedTone cohort.

--by Sanjay Trehan and Kelly Close