Memorandum

Arena Announces BLOSSOM Phase 3 Data – September 20, 2009

Executive Highlights

  • Top-line data announced by Arena CEO Jack Lief for the phase 3 trial BLOSSOM, indicates that lorcaserin meets the FDA’s categorical endpoint, while falling short of meeting the agency’s average percent weight loss endpoint. Only one endpoint is required to be met for approval.
  • On an ITT-LOCF basis, 47.2% of patients on 10 mg BID lorcaserin achieved 5% or more body weight lost. When considering only patients who completed the study, 63.2% of patients on 10 mg BID lorcaserin achieved ≥5% weight loss.
  • Lorcaserin appears to be well tolerated, with the most frequent adverse events being respiratory infection (13-15% in the treatment arms v. 13% in placebo), nasopharyngitis (12- 13% in the treatment arms v. 12% in placebo), and headache (10-11% in the treatment arms v. 8% in placebo). The rate of serious adverse psychiatric events was not different between placebo and lorcaserin patients, which was a positive from our view given the various compounds where there are worries about psychiatric issues.
  • Extensive valuvulopathy assessments suggest lorcaserin does not increase cardiac valvular risk. Management noted that FDA-defined rates of valvulopathy were 2.0% in the lorcaserin 10 mg BID group, 1.4% in the lorcaserin 10 mg QD group, and 2.0% in the placebo group. We note that the ECHOs address valuvulopathy, not CV risk in terms of CV events (MIs, angina, revascularlizations, etc.)Arena plans to file the NDA for lorcaserin by the end of the year and to present more detailed data at the 2009 National Obesity Society Annual Meeting. While we do not view the results as striking as could have been, we do believe that the drug may be able to be combined with other compounds and could look considerably better.

Arena’s President and CEO, Jack Lief, led a conference call last Friday announcing top-line data from the Behavioral modification and LOrcaserin Second Study for Obesity Management (BLOSSOM) trial, the final phase 3 study for lorcaserin, an obesity drug targeting the serotonin 5HT2c receptor. The trial evaluated lorcaserin’s potential to reduce weight and improve a multitude of secondary endpoints using 10 mg of lorcaserin once- or twice-daily in 4,008 obese or overweight patients, some suffering from at least one co-morbid condition (clinicaltrials.gov says 30-45 with or without co-morbidity, 27-45 BMI with at least one co-morbidity. The results from the BLOSSOM trial are consistent with the BLOOM data – lorcaserin met the FDA’s categorical efficacy benchmark while falling short of the average percent weight loss endpoint. The weight loss data for lorcaserin showed 47% of patients achieving ≥5% weight loss on an Intent-to-Treat Last Observation Carried Forward (ITT-LOCF) basis and 63.2% of “completers” losing ≥5% body weight on lorcaserin 10 mg BID. Patients on the full dose of lorcaserin also experienced mean weight loss of 5.9% on an ITT-LOCF basis and 7.9% on a per protocol basis versus placebo of 2.8% and 3.9%, respectively. The weight loss data fell short of the Vivus Qnexa data also recently released (see our September 10 Closer Look). Qnexa showed a mean weight loss of 14.7% observed in the high-dose completers of Qnexa in the EQUIP trial as well as indications that drug can improve cardiovascular risk factors.

Lorcaserin appears to have an impressive safety profile, based on this data. The most frequent adverse events were upper respiratory infection (occurring 13-15% in the treatment arms versus 13% in placebo), nasopharyngitis (12-13% in treatment arms versus 12% in placebo), and headache (10-11% in treatment arms versus 8% in placebo). Dropout rates due to adverse events were very low at 6-7% in the treatment arms and 5% in the placebo group. However, these low rates are not entirely reflected in overall completion rates, which were 57-59% for the lorcaserin groups and 52% for the placebo group. Management highlighted findings from Decision Resources, an independent market research firm, which suggest that the majority (64%) of weight loss medication prescriptions are written by primary care physicians (PCPs), who are generally dissatisfied with the current treatment options. (We aren’t sure the extent to which this is true, but it sounds reasonable given the combination of dissatisfaction from efficacy and safety/tolerability). Lief also noted that most prescriptions are written by a small group of <10,000 PCPs. During the questions and answers session, management provided reassurance that the NDA would be filed by year’s end; however, Lief refrained from providing more detailed data, which will be presented at The Obesity Society meeting in Washington, DC in late October. Overall we believe these data are positive, especially in regards to safety but that there will be considerable competition, assuming Qnexa and Contrave make it through FDA. The key question from our view is whether it will be easy to identify super-responders to this therapy and whether there will be major sub-groups who do not respond to other drugs in development; if so, of course, given the size of the obese population in the US alone, there will be more than enough room for a number of drugs to do well.

  • The BLOSSOM trial (n=4,008) evaluated the safety and efficacy of 10 mg of lorcaserin (once or twice daily). Obese patients (BMI between 30 and 45 kg/m2) were included with or without co-morbid conditions and overweight patients (BMI between 27 and 30 kg/m2) were recruited only if the patient had a co-morbid condition. At baseline, average BMI was 35.9 kg/m2, with an average weight of 220 pounds (100 kg). Roughly 80% of patients enrolled in the BLOSSOM trial were women. Patients were randomized in a 2:1:2 ratio to 10 mg lorcaserin twice-daily (BID), 10 mg lorcaserin once- daily (QD), or placebo.
  • The BLOSSOM data shows 10 mg lorcaserin twice-daily clearly meets the FDA benchmark for categorical efficacy. Arena reported the percentage of patients achieving ≥5% weight loss and ≥10% weight loss in each arm of the study using an ITT- LOCF analysis as well as a per protocol analysis. There was a significant increase in the percentage of patients achieving ≥5% weight loss in the 10 mg BID and 10 mg QD compared to placebo, however, it is not clear whether the percentage of patients achieving≥5% weight loss in the 10 mg QD arm (40.2%) is “approximately double” that of the placebo arm (25.0%)—this must be achieved to satisfy the categorical efficacy benchmark set by the FDA. The following are the one-year efficacy data for the BLOSSOM trial:

 

10 mg BID

10 mg QD

Placebo

≥5% weight loss (Per protocol)

63.2%

53.1%

34.9%

≥5% weight loss (ITT-LOCF)

47.2%

40.2%

25.0%

≥10% weight loss (Per protocol)

35.1%

26.3%

16.1%

≥10% weight loss (ITT-LOCF)

22.6%

17.4%

9.7%

Mean weight loss (Per protocol)

7.9%

6.5%

3.9%

Mean weight loss (ITT-LOCF)

5.9%

4.8%

2.8%

  • Although only top-line data was reported, lorcaserin seems to have a favorable side effect profile. Management noted that no adverse event rate in the treatment groups ever exceeded placebo by more than 4% (we note we have not seen AE information presented in this way before). The most common adverse events were upper respiratory infection (12.7%, 14.5%, 12.6%), nasopharyngitis (12.5%, 11.7%, 11.8%), and headache (10.0%, 10.5%, 7.6%) for lorcaserin twice-daily, lorcaserin once-daily, and placebo, respectively. Depression, anxiety, and suicidal ideation occurred at similar rates within each group and lorcaserin patients did not experience any more neuropsychiatric serious adverse events (SAEs) than patients in the placebo group. There were no SAEs of seizure. Withdrawal rates due to adverse events were 7.2% in the lorcaserin BID group, 6.2% in the lorcaserin QD group, and 4.6% in the placebo group – these are very low numbers, especially for an obesity candidate. Overall, one-year completion rates were 57.2%, 59.0%, and 52%, respectively, slightly better than average for obesity trials. We are curious what the main reasons were that patients discontinued if not due to adverse events; we assume it was relative lack of efficacy – from discussions we have had with doctors and in industry, we gather many patients and providers are most satisfied with 10% weight loss or higher.
  • Pooled echocardiography data from BLOOM and BLOSSOM show no signs of risk for valvulopathy following lorcaserin treatment. Management noted that FDA-defined rates of valvulopathy were 2.0% in the lorcaserin 10 mg BID group, 1.4% in the lorcaserin 10 mg QD group, and 2.0% in the placebo group.
  • Management did not delve into lorcaserin’s efficacy with respect to secondary endpoints. However, management noted there were “significant improvements” in blood pressure, HDL, and triglycerides using a per protocol analysis (only HDL and triglycerides were significantly improved using ITT-LOCF analysis). Only the p-values were provided for each analysis and we would like to understand better the degree of the significance. We look forward to more detailed data of both the BLOOM and the BLOSSOM trials at The Obesity Society annual meeting in October.

Questions and Answers:

Q: Can you provide us with more details on the market research? Most of the data we have looked at says primary care physicians will use some combination of drugs and not just a monotherapy.

A: Phentermine is a 50 year old amphetamine class of drug with its own set of side effects. Lorcaserin is not meaningfully different with respect to side effects from placebo, however patients lose a significant amount of weight. Independent market research has been conducted that shows physicians would be willing to switch from phentermine to lorcaserin.

Q: For patients not having problems on phentermine, what would be the logic to go on lorcaserin?

A: A general practitioner would prefer starting patients on single agents and would consider adding or switching therapies if they were not satisfied. While it’s possible some physicians might want to add a combination treatment, we believe lorcaserin alone will give physicians what they want.

Q: Can you give us the rates of SAEs for all arms?

A: We don’t have exact numbers for the rates of SAEs, however, it is approximately the same in each of the treatment groups. (We note that typically companies get this information early as we understand it; we are not sure why the company is not speaking to this.)

Q: What were the baseline levels for valvulopathy and did you see any difference with patients who already had valvulopathy?

A: Around 4-5% of patients at baseline had FDA-defined valvulopathy. There were slightly more in the lorcaserin BID group but they are all between 4-5%. We haven’t received all the analysis but it appears as if pre-existing FDA-defined valvulopathy does not appear to have progressed.

Q: It seems as if the placebo response rate is on the high side – 2.8%. Was there any difference in behavior modification or anything to explain that?

A: There were no differences. It is just due to the variations in the trial. You can see that lorcaserin response was extremely consistent between the trials.

Q: Did any specific adverse event (AE) lead to a majority of the safety or tolerability- related dropouts?

A: Overall, dropouts for AEs were quite low, below 8% in all of the groups, which speaks again to the tolerability. In terms of individual AEs, there is not one that stands out as the main cause.

Q: Can you provide extra context for the secondary endpoints - how important do you think they are from a regulatory and reimbursement standpoint?

A: FDA wants to see improvements in secondary outcomes measures. We do see better results when we look at the per protocol analysis. We will present more detailed analysis at The Obesity Society.

Q: Can you comment on the extent of due diligence by large pharmaceutical companies?

A: There is a lot of diligence currently underway. Meetings are continually being scheduled for prospective partners to come visit. We don’t have anything else to provide at this time but we will update you when we have more information.

Q: Can you give us details behind the pooled valvulopathy data from BLOOM and BLOSSOM?

A: Pooled data will include all patients from both studies. The overall rate of valvulopathy in placebo was 2.2% vs. 2.3% in the lorcaserin-treated groups. This excludes the risk of valvulopathy with the FDA and provides evidence for lorcaserin being highly specific.

Q: Regarding FDA guidelines, can you give some color on how we should interpret “approximately two times placebo rate” rule for the efficacy benchmark – for BLOSSOM you were approximately there but not quite.

A: The agency guidance is that the proportion of patients who lose at least 5% should be at least 35% and approximately double the proportion in the placebo group and difference between groups is statistically significant. As we said, what’s important, beyond benchmarks, is the risk- benefit profile of lorcaserin. Lorcaserin met all primary safety endpoints and a significant proportion of patients lost 35 pounds (as we understand it, this is essentially the “super-responders” or the 25% that lost the most weight – we are curious to know more about standard deviations, etc.). This gives us a lot of confidence that lorcaserin can change the way PCPs address weight management to the broad cross-section of their patients. BLOSSOM results are amazing because they confirm the results of BLOOM.

Q: Can you give what you currently view as the net cash position following the recent financing transaction and what you need to be spending for the next 6-12 months to get the NDA filed. Also, what should we be looking at in terms of cash burn?

A: In our last call, June 30th cash was about $175 million. Year-end cash will be between $120 and $130 million. Our cash usage goes down very significantly now that the 4000 patient BLOSSOM study is complete. BLOOM-DM is only 600 patients so that is not terribly costly so I think we’re in good shape.

Q: What about the debt?

A: To Deerfield, net debt is $90 million and we’ve paid $10 million with the last equity raise. The next payment will be $20 million in mid-2011. The debt repayment doesn’t enter the picture prior to the expected PDUFA date.

Q: On the partnership front, can you launch this without a partner and at what point do you need to start making contingency plans. When do we start to worry?

A: We’ve been drawing up contingency plans the entire time. While a small sales force can address a small portion of the market, we really need a large PCP sales force on the order of 1000 sales reps to really address the full potential of this market. More will be visible as we progress to filing an NDA late this year and once we have it accepted early next year. Of course, we will update you as soon as possible.

Q: If you take 47% who lost 5% weight and exclude those who were carried forward are you still above 35%?

A: Absolutely.

Q: On valvulopathy, what is the upper limit on the confidence interval?

A: We needed to show the upper bound of the 90% confidence interval was below the noninferiority margin. We’re saving some of the thunder for the obesity society meeting. The upper bound is well below the noninferiority margin so we did accomplish what we negotiated with the FDA.

Q: Have you completed an abuse liability study? Presumably you have to include that in the NDA submission?

A: We have completed that study. Our abuse liability study has been submitted for presentation at a couple scientific meetings. Results were quite favorable and show very little to no abuse potential for lorcaserin.

Q: Can you refresh our memories as to the amount of capital that has been expended on both BLOOM and BLOSSOM?

A: For our phase 3 program, external costs have been at least $200 million. However, the internal costs are quite significant. I would estimate somewhere between $500 million and a $1 billion will have been spent by the time lorcaserin is approved, sometime hopefully next year. The good news is that we’ve spent the vast proportion of that and very little is left to do. Because we’ve tested our drug for two years (editor’s note - as is required for NCEs) I think most physicians will be comfortable with it.

Q: If the total cost approaches $1 billion, that is more than double your current market capitalization. What kind of price tag would you would be looking for in terms of partnerships? Have any negotiations started with large pharmaceutical companies at this time?

A: We have a good manufacturing facility that currently manufactures GMP products that is used by numerous companies throughout the world. We expect to provide finished product to our partners that’s manufactured in our Swiss facility. In terms of up front and milestone payments, obviously we’re still in discussions. I think most partners are waiting to see this data. This is a very important event and we expect to capture the benefit of lorcaserin in our spending via transfer pricing, to a large extent.

Q: Are you expecting to recover the $500 million to $1 billion in the form of upfront payments or gross margins on the manufacturing of the drug?

A: Yes definitely. Obviously we need to get the drug approved before we can see gross margins. There is no question in my mind that we will recover the cost to compensate us for the risk we’ve taken in developing lorcaserin. I suspect lorcaserin will be quite successful as a first-line therapy. I think this really is a game changer in the weight management area. If you look at drugs to treat hypertension, physicians have numerous choices of mechanisms to use. In weight management, there are only two and the side effect profiles limit the use of these drugs.

Q: What were the patient numbers with valvulopathy in each arm?

A: I only have the pooled data. The number of total patients in the lorcaserin group is slightly higher than in the placebo group at week 52. The actual number is slightly higher in the lorcaserin group to give you the matched percentages. We did rule out the risk of valvulopathy, however, we are trying to save this data for The Obesity Society meeting in a few weeks.

Q: Have there been any seizures in the clinical trial program so far in any of the arms?

A: We did not see any in the BLOOM or BLOSSOM studies. We had one patient with a history of fainting and having seizure-like activity during blood-drawing and he had one of these episodes during this trial. He happened to be in the lorcaserin 10 mg BID arm.

Q: How do you anticipate the label will read in regard to echocardiograms, will patients need one at baseline and then at follow up?

A: Short answer is that we don’t know that yet because the FDA will decide but we believe our data have ruled out any risk of valvulopathy. No mater how you look at these data, they just don’t show a signal.

Q: Is there any one rate-limiting step to submitting the NDA?

A: We’ve put together pretty much all the data we now need for this NDA. We have favorable results on everything we’ve compiled. We have the people, processes, and passion in place to get the NDA together. The key challenge right now is sitting down and writing it. I think the rate- limiting factor is sleep.

-- by Sanjay Trehan and Kelly Close