Brean Murray "We are initiating coverage of CELG with a Buy rating a target of $100---3 new product launches over the next 12 months could add $3-5B in peak sales within the next 3-5 years. Will EASILY meet 2015 goals even with conservative ests"
Brean Murray, Carret & Co
We are initiating coverage of Celgene Corp. with a Buy rating and 12-month target price of $100, based on a P/E multiple of 18x applied to our 2013 fully diluted EPS estimate of $5.53, in line with consensus. Our 18x price-to-earnings multiple is at the high end of the large-cap biotech group (13-20x), but appropriately reflects the 18% CAGR in EPS that we forecast between 2012 and 2015, driven by top-line growth from three new product launches over the next 12 months that we forecast could add $3-5B in peak sales within the next 3-5 years. CELG has offered long-term guidance of achieving $8-9B in peak revenue by 2015, and our conservative estimates suggest the company will easily meet the low end of that of that guidance
Sentiment: Strong Buy
Brean Cap detail comments on 3 drugs/launces they see adding $3 bill to $5 bill a yr including $2 billion for Apremilast.....
CELG, in April 2012, submitted an NDA to the FDA for Pomalidomide in relapsed refractory multiple myeloma. The application is based on the results of a Phase 2 study that randomized 221 patient with relapsed refractory multiple myeloma to receive pomalidomide (N=108) or pomalidomide plus dexamethasone (n=113). Results presented at the 2012 ASCO conference showed that patients who received the combination benefited from overall survival of 14.4 months. Those that received Pomalidomide alone benefited from an overall survival of 13.6 months. The combination of pomalidomide and dexamethasone was associated with an overall response rate of 34% vs. 13% for POM. Importantly, patients refractory to Revlimid as well as Revlimid and Velcade also responded to POM+LoDEX with ORRs of 29% and 30%, respectively.
POM was generally well tolerated, with no occurrence of severe peripheral neuropathy. Both arms compare very favorably with the historical control for overall survival in this patient population that peaks at 9 months (see Exhibit 8).
The FDA has set a February 10, 2013, PDUFA date for Pomalidomide, and we anticipate approval on or before that date. We forecast Pomalidomide sales in relapsed refractory myeloma of $730M in 2015, and we expect the drug to remain in the relapsed refractory setting peaking at $1.3B in total revenue before patent expiration in 2019. A Phase III is enrolling in the E.U. and a Velcade combination Phase III is planned in the U.S.
CELG plans to submit regulatory filings in the U.S. and Canada for Apremilast in psoriatic arthritis in early 2013. Filings for psoriasis in the U.S. and Canada will follow in 2H13. Soon after, CELG plans to submit regulatory filings in the E.U. for both indications also in 2H13. Apremilast is currently in Phase 3 trials for the treatment of psoriasis and psoriatic arthritis. This will put Apremilast on track to launch in the U.S. and Canada in 1H14 and in E.U. during 2H14.
View on Overall Efficacy and Tolerability
CELG's strategy is to position Apremilast as an option following methotrexate and prior to injectable TNF-alpha inhibitors. Disease-modifying activity could pave the way for favorable reimbursement from third-party payers looking to drive utilization towards a more convenient, lower-cost, orally available alternative to expensive injectable biologics. While the vast majority of patients with psoriatic arthritis will be treated with methotrexate at some point in their disease, the drug is not approved due to the dearth of controlled data. Kingsly et al published a study in 2012 that enrolled 200 patients in the U.K.Only 2/3 of the enrolled population made it through the trial and it took 5 years to enroll. Many of the patients wound up being too healthy for the study, as many clinicians held their more severe patients over for biologic treatment and there were a significant number of dropouts. The final analysis demonstrated that the trial failed to show a significant difference in treatment effect for methotrexate. The most significant drawback to the efficacy of the anti-TNF biologics is the loss of treatment effect over time, due to the accumulation of neutralizing antibodies and incidence of serious side effects. Many patients cycle through biologics at 1.5 year intervals, with progressively waning benefit.
We believe even conservative assumptions suggest that Apremilast could capture 100,000 of the psoriatic arthritis and psoriasis patients in the US, Canada and Europe. Our model forecasts Apremilast could be used by 60,000 of the 900,000 psoriatic arthritis patients and 45,000 of the 2.5M psoriasis patients. Taken together, our forecasts for Apremilast in both psoriatic arthritis and psoriasis peak at $2B (see Exhibit 10).
Psoriatic Arthritis Market
According to CELG market research, the vast majority of the 960,000 psoriatic arthritis patients (approximately 73%) patients in the U.S. and E.U. are treated either with an NSAID (30%) or DMARD (25%), or are DMARD failures (15%). These three groups of patients make up the primary target market for Apremilast. Based on the safety and tollerability demonstrated in the PALACE phase 3 program for Apremilast first presented in detail at the 2012 Annual ACR meeting, we estimate Apremilast could conservatively penetrate low-double-digit percent of the market or approximately 40,000 patients in the U.S. alone, which represents peak sales of approximately $750 million to $1 billion, assuming discount pricing to current biologic anti-TNF therapies.
Apremilasts market potential will become clearer in the Psoriasis segment, following release of results from the ESTEEM-1 (4Q12) and ESTEEM-2 (1Q13) studies. However, based on earlier data and the implications of the PALACE Phase 3 program, we believe a reasonable expectation for Apremilast's market capture is approximately 45,000 psoriasis patients in the U.S and E.U.
PALACE-1 Phase 3 Presentation at 2012 ACR Meeting
The PALACE program was the first, large Phase 3 program for an oral candidate in over 10 years. The entire PALACE program included over 2,000 patients enrolled in four studies. PALACE-1 included 504 patients alone with psoriatic arthritis diagnosis for at least 6 months and active disease characterized by three or more swollen/tender joints and skin involvement. Patients were allowed to be on concurrent DMARD agents. Over 70% of the patients enrolled in PALACE-1 had been exposed to DMARDs, while over 20% of patients had experience with prior biologics. Patients were randomized 1:1:1 to receive Apremilast 20 mg, 30 mg, or placebo for 24 weeks. Then, placebo patients are re-randomized to receive 20 mg or 30 mg for an additional 28 weeks, with an open-label extension for 5 years. The primary endpoint ACR 20 was measured at 16 weeks and achieved by 31% of patients treated with Apremilast 20 mg and 43% of those treated with Apremilast 30 mg versus 24% of patients treated with placebo. Patients treated with Apremilast 30 mg monotherapy experienced greater efficacy (ACR 20 = 59%) than any of the other treatment subgroups, including those treated with combination DMARD agents (See Exhibit 12).
Why the monotherapy arm did better than combination treatment is not yet fully understood, but could be related to pharmacodynamic interference, which is observed with other small molecules when combined with higher doses of methotrexate. While a bit puzzling, this dynamic is likely to facilitate switching patients from methotrexate to Apremilast.
Psoriatic Arthritis - Worst of Both Worlds
Psoriatic arthritis is distinct from rheumatoid arthritis. It is more involved with inate vs. adaptive immunity, which is why many therapies do not cross-over the treatment landscape to impact psoritic arthritis as effectively as RA. Patients with PsA experience the same swollen tender joints that RA patients experience and have added inflammation at the site of tendon insertion into bone as well as spinal inflammation. PsA patients also carry a higher emotional burden than RA and psoriasis patients. There are more co-morbities in PsA that include cardiovascular issues, obesity, respriatory problems, and depression. The cardiovascular co-morbidities in PsA can be as serious as those observed in RA, due to the inflammatory component. The metabolic syndromes leading to obesity are similar to those observed in psoriasis. This is why clinicians try to treat PsA aggressively and early on.
In a Global Landmark survey of psoriatic arthritis patients conducted in 2012, the majority of respondents reported that their therapy was burdensome. Among patients treated with oral DMARDs, 48% considered the side effects of therapy a significant burden (see Exhibit 13). Among the most difficult things to tolerate are the persistent GI side effects, frequent blood tests, and liver monitoring. Apremilast’s emerging profile suggests it could have an advantage, as the associative GI side effects wane over the initial treatment period, peaking in the first week of dosing before beginning to resolve. There also appears to be no need for blood or liver function monitoring. The alcohol restriction for patients treated with methotrexate due to the risk of liver damage is also considered a burden. The survey also uncovered a significant prevalence of needle
phobia among patients treated with biologics who describe fear and anxiety in preparing for injection. Patients who travel frequently are inconvenienced by the need to refrigerate and store their medication.
When the efficacy of Apremilast in psoriasis patients is placebo-corrected in comparison with injectable TNF inhibitors, Apremilast appears to be just as effective as the TNF inhibitors. Phase 2 data In psoriasis patients illustrated a PASI 75 response rate of 41% for Apremilast at 16 weeks, which is just below the low end of the range in PASI 75 response rates of 47-78% for Enbrel and Humira. In psoriatic arthritis patients, Apremilast’s 12-week ACR 50 response rate of 13% is a more inferior comparison to Humira’s 12-week ACR 50 response rate of 36%. While acknowledging that Apremilast may not have the same efficacy demonstrated by Enbrel or Humira in the short term, a possibility was raised over Apremilast potentially being a diseasemodifying drug. Currently, there is no data available regarding Apremilast’s ability to delay, stop, or reverse the progressive damage to joints experienced particularly by psoriatic arthritis patients over time. However, Apremilast’s inhibition of PDE-4 reduces a number of pro-inflammatory mediators including TNF-α, interleukin-2, interferon-γ, leukotrienes, and nitric oxide synthase. Perhaps, this broad spectrum of activity over time will demonstrate reduction in joint damage.
Enbrel’s (anti-TNF-α) ability to halt joint damage is evident on radiographs of patients after 12 months of treatment, while Humira (anti-TNF-α) shows a benefit at six months. In both cases, the disease-modifying benefit amounts really to just a halt in the progression of disease, making this a tough endpoint to achieve.
CELG presented positive data from a pivotal Phase 3 trial for Abraxane in melanoma cancer earlier this week at the Society for Melanoma Research annual meeting. In January 2013, CELG will present the results of its Phase 3 registration study for Abraxane in pancreatic cancer at the ASCO GI Symposium. We expect there will be much debate in the coming weeks over Abraxane’s potential ahead of ASCO GI. Historically, older generation taxanes have not shown a significant clinical benefit over gemcitabine alone in front-line pancreatic cancer, thereby creating low expectations for Abraxane. Single-agent paclitaxel has been associated with a response rate of 3- 6% and a median overall survival of 5-6.5 months (See Exhibit 14; Whitehead et al, Saif et al). However, there have been more promising results with the inclusion of taxanes in drug regimens involving two or more agents in combination. The Phase 2 study for Abraxane in combination with gemcitabine doubled the median OS typically observed with gemcitabine (6.0 months) to 12.2 months, with a much higher response rate of 46% (See Exhibit 14; VonHoff et al.) versus gemcitabine alone (10-20%). Abraxane is not the only newly formulated paclitaxel to improve upon the efficacy of first-generation paclitaxel. Medigene’s EndoTAG liposomal-encapsulated paclitaxel was evaluated in combination with gemcitabine in a Phase 2 trial of 212 patients who received doses ranging from 11mg to 44mg twice weekly combined with standard-dose gemcitabine for seven weeks. The combination elicited a response rates that ranged 14-16%, with overall survival of 8.1-9.3 months (see Exhibit 14; Lohr et al). Medigene is initiating a Phase 3 trial for EndoTAG in breast cancer, which is the lead indication for the formulation with its partner and is planning to file that indication no earlier than 2018. We believe these results, coupled with the early data for Abraxane, suggest that, at a minimum, Abraxane’s approval in pancreatic cancer will at least open the door for further development with other chemotherapies and newer-targeted agents.
In 2010, CELG acquired Abraxis Biosciences for $2.9B in cash and stock plus performance milestones of up to $650M payable to holders of contingent value rights that also possess a 2.5- 10%-tiered royalty on sales of the company (see Exhibit 15). The performance milestones, which included a PFS claim for Abraxane in NSCLC and approval in pancreatic cancer by April 1, 2013, will not be met, thus leaving the royalty as the only ongoing obligation CELG has to holders of the CVR. CELG is obligated to pay a royalty of 2.5% for any sales between $1-2B, a 5% royalty on sales between $2-3B, and a 10% royalty on any sales above $3B. Our peak sales estimate for Abraxane are $1-2B, so our model contemplates only the 2.5% royalty in the out-years (see Exhibit 15). We estimate FY12 sales of $460M for Abraxane in breast and lung cancer, with breast cancer indication contributing the majority of that revenue ( $400M). We extimate the revenue potential from melanoma is in the $200-250M range, while pancreatic cancer could add another $500 million to $1 billion (see Exhibit 15), depending on the magnatude of the clinical benefit that will be further elucidated at the January 2013 ASCO GI symposium Abraxane is a cremophor-free albumin-encapsulated paclitaxel particles that are nanometer-sized.
The paclitaxel is immediately bioavailable and in a non-crystalline form allowing for rapid release. Abraxane exploits the SPARC (secreted protein, acidic, rich in cysteine) pathway that has been identified as an albumin-binding protein expressed on the surface of several tumor types. SPARC is also associated with poorer outcomes and greater metastasis, thus making it a relevant drug target. Abraxane’s benefit is that it can deliver more toxic chemotherapy payload to the site of the tumor, but is also less toxic than generic paclitaxel (Taxol) due to its formulation. Abraxane was initially approved for second-line treatment of metastatic breast cancer in 2005 in the U.S. and in the E.U. in 2008 after data showed an improvement in response rate of 33% vs. 19% for solvent based paclitaxel in patients who received 2 or more prior therapies. Patients treated with Abraxane further benefited from an improvement in survival of 56.4 weeks versus 46.7 weeks (see Exhibit 16).
On November 8, 2012, CELG announced that its Phase 3 trial for Abraxane in pancreatic cancer achieved its primary endpoint, with a statistically significant improvement in overall survival. The magnatude of the OS benefit will not be known to the public until the data is presented as a late breaker at the ASCO GI Symposium in January 2013. The Phase 3, open-label randomized, multicenter trial compares Abraxane 125 mg/m2 every week for three weeks in combination with gemcitabine administered weekly to gemcitabine monotherapy, with a primary endpoint of overall survival. Secondary endpoints include objective tumor response rate and PFS. In a Phase 2 study that was published in the Journal of Clinical Oncology in October 2011, a total of 67 patients were treated with Abraxane in doses of 100 mg/m2, 125 mg/m2, and 150 mg/m2
weekly for three weeks in combination with standard gemcitabine. The maximum tolerated dose was found to be 125 mg/m2 (only 4 patients received 150 mg/m2). Side effects were considered to be manageble with fatigue, neuropathy and hematological abnormalities among the most prevalent Grade 3 or higher side effects. The only Grade 4 side effects were leukopenia, neutropenia, and thrombocytopenia, all of which were manageable. Patients who received Abraxane 125 mg/m2 benefited from a median survival of 12.2 months, a 50% improvement vs. the historical benefit with gemcitabine alone (See Exhibit 17).
In a Phase 2 study that was published in the Journal of Clinical Oncology in October 2011, a total of 67 patients were treated with Abraxane in doses of 100 mg/m2, 125 mg/m2, and 150 mg/m2 weekly for three weeks in combination with standard gemcitabine. The maximum tolerated dose was found to be 125 mg/m2 (only 4 patients received 150 mg/m2). Side effects were considered to be manageble with fatigue, neuropathy and hematological abnormalities among the most prevalent Grade 3 or higher side effects. The only Grade 4 side effects were leukopenia, neutropenia, and thrombocytopenia, all of which were manageable. Patients who received Abraxane 125 mg/m2 benefited from a median survival of 12.2 months, a 50% improvement vs. the historical benefit with gemcitabine alone (See Exhibit 17).