In its earnings release last week, Seattle Genetics (SGEN) did not surprise anyone with the financial guidance and expected timelines for approval of its lead agent, SGN-35. However, on the business development front, the release did include an intriguing announcement that did not receive the attention it deserved. The company announced that Genentech recently advanced 3 new antibody drug conjugates [ADC] based on Seattle Genetics’ technology to phase I, this is in addition to the CD22 ADC already in clinical testing.
The announcement has several important implications for Seattle Genetics. First, the number of clinical programs in its partnered pipeline instantly jumped 50% from 6 to 9. By definition, this provides Seattle Genetics with more shots on goal and increases chances of substantial milestones and royalties down the road. More importantly, it establishes Seattle Genetics’ technology as Genentech’s (DNA) preferred ADC platform, an attractive position given Genentech’s dominance in oncology and ADCs in particular.
Genentech’s decision is a serious blow to Immunogen (IMGN), but it does not mean that Genentech will not use its technology at all. Genentech still has licenses for 4 targets from Immunogen based on deals from 2005 and 2008. The fact that until now no clinical program has resulted from these deals is a bad indication but it does not completely rule out this option.
Genentech’s growing ADC pipeline
When Genentech started developing antibody drug conjugates it worked with technologies from both Immunogen and Seattle Genetics. The first ADC it advanced to clinical testing was T-DM1, based on Immunogen’s technology. From the initial phase I, T-DM1 has demonstrated remarkable activity and a surprisingly benign safety profile. It is probably the success of this agent which ignited the excitement around ADCs and led to an industry-wide shift towards this field.
In parallel to T-DM1, Seattle Genetics was developing its own ADC, SGN-35 for Hodgkin’s Lymphoma, where the agent demonstrated strong efficacy with an excellent safety profile. Based on the clinical validation of both technologies and the fact Genentech had access to them, it seemed plausible that it would use both platforms in future projects. If anything, one could expect that T-DM1’s remarkable performance and the aggressive development plan Genentech is pursuing for this agent would make it lean towards Immunogen rather than Seattle Genetics.
But this was not the case. In June 2008, Genentech advanced a second ADC to phase I, which was based on Seattle Genetics’ technology, but the program was discontinued shortly afterwards, probably due to safety issues. This was followed by two follow-on licensing deals with Seattle Genetics for additional targets during 2010. The deals included $21.5M in upfront payments and probably over $1B in potential milestone payments.
Since then, Genentech advanced 4 additional ADCs to phase I, all of which employ Seattle Genetics’ technology. This brings the number of Seattle Genetics-based programs to 5 versus only one using Immunogen’s technology (T-DM1). Genentech disclosed the identity of only one of the 5 candidates, a CD22 ADC for the treatment of blood cancers (discussed here). The rest of the programs (the discontinued plus three new programs) are still undisclosed. The discontinued program was probably a MUC16 ADC whereas two of the active programs could be ADCs targeting CD79b (lymphoma) and TENB2 (prostate cancer).
Not all deals are created equal
In the past 2 years, Seattle Genetics managed to bring more deals, generating more cash and opportunities compared to Immunogen. It is unclear whether this reflects a real preference in the industry or simply Immunogen’s decision to be more selective in the deal it signs. During 2009-2010, Seattle reported rich deals with GSK (GSK), Astellas and Daiichi Sankyo in contrast to Immunogen, who signed a meaningful deal only recently with Novartis (NVS). The recent Novartis deal (discussed here) was an important validation after a 2 year drought and the price paid by Novartis certainly showed that Immunogen still has a lot to offer.
Not all deals are created equal, as some represent higher chances of actually generating clinical stage programs. In general, recent deals that involve a higher price tag indicate that a partner is more committed to a given program and is actively pursuing it. Immunogen’s partnerships with Novartis and Amgen (AMGN) are good examples.
Novartis agreed to very generous terms as part of last year’s deal, so one can assume that it would not do the deal unless it were highly excited about these programs. In contrast, Amgen secured Immunogen’s technology for two targets as part of an old collaboration it inherited from Abgenix. Because the agreement was signed when Immunogen’s technology was still unvalidated, it includes very modest financial terms ($1M upfront per program). Amgen’s 2009 licensing of two targets probably reflected the upcoming expiration of the Abgenix-Immunogen collaboration. This enabled Amgen to secure access to Immunogen’s technology for a fraction of the real market price, so it is still unclear what the driving force behind this deal was.
Fundamental questions regarding target selection
[Editor's note: The author has made changes in this section, January 2012]
Although it seems that when given a choice, Genentech prefers Seattle Genetics’ technology, there are cases in which Immunogen’s technology is superior to that of Seattle Genetics. In particular, targets that do not get internalized efficiently seem more suitable for Immunogen’s
cleavable disulfide linkers like those used in all of its programs except T-DM1. Genentech appears to avoid these targets, as ADCs based on disulfide linkers are thought to have a less favorable therapeutic window (efficacy/safety ratio), even though Immunogen and its partners were able to reach clinically active doses in most clinical trials.
Indeed, to date, the only two ADCs which demonstrated sufficient potency as single agents are T-DM1 and SGN-35, both utilize
non-cleavable more stable linkers (either uncleavable or peptide based) for targets that are good internalizers. The rest of ADCs in Immunogen’s pipeline have not performed as nicely as T-DM1, although it is hard to attribute a given clinical profile to a single element. IMGN901, for example, had activity in multiple indications including lung cancer and multiple myeloma, but it was not potent enough as mono-therapy. This forced Immunogen to evaluate IMGN901 in combination with standard of care in three indications.
Developing a drug in combination is by definition more challenging. It requires comparative trials that are longer and more expensive, and it also has safety implications that may prevent optimal dosing. Most importantly, proof of concept is reached only after a large randomized phase II trial, a formidable challenge on its own (discussed here). This brings up the debate on whether targets that require
cleavable disulfide linkers should be pursued at all.
There may still be cases where ADCs based on
cleavable disulfide linkers show sufficient potency as single agents, especially in blood cancers that are considered less challenging than solid tumors. This year at ASCO, Sanofi is expected to present data for SAR3419, a CD19 ADC, given weekly as mono-therapy. Sanofi already announced this program will move into phase II as single agent, so one could expect data to be positive (although there are a lot of other promising agents in clinical development for NHL).
But what about cases where an ADC needs to be given in combination? This is still an open question as there is not enough clinical data out there regarding ADCs as part of combination regimens. One encouraging indication comes from Endocyte (ECYT), who is developing small-molecule drug conjugates (SMDC). Although SMDCs have a totally different clinical profile than ADCs, both rely on a similar approach of a targeting moiety used to deliver a toxic payload into cancer cells. Endocyte’s lead agent, EC145, was not effective enough to be given alone, so the company evaluated it in a randomized phase II in combination with chemotherapy. Results demonstrated a remarkable benefit in a subset of patients who expressed EC145’s target in all of their lesions, so this can be viewed as a conceptual proof of concept. Next year, Immunogen expects to start phase I with IMGN853 which is a direct competitor of Endocyte’s EC145.
Interestingly, Sanofi’s SAR3419 utilizes Immunogen’s disulfide linker, even though CD19 is highly internalized. In one of Genentech’s seminal publications on ADCs for blood cancers, researchers also found that CD19 ADCs using Immunogen’s non-cleavable linkers were ineffective. Seattle Genetics published a paper in 2008, showing good efficacy with its non-cleavable linker.
Immunogen’s partnered pipeline in 2011
Immunogen expects three new programs in its partnered pipeline to start phase I this year, and there is still a hypothetical chance that one of these programs will be with Genentech. Another partner that could start phase I with an ADC based on Immunogen’s technology is Bayer who is developing an anti-mesothelin ADC. Bayer had a pre-existing collaboration with Seattle Genetics prior to signing the deal with Immunogen, but licensed Immunogen’s technology for this ADC. One potential explanation is the lower activity observed with non-cleavable linkers for mesothelin targeted ADCs. Another active partner is obviously Novartis, who has been very aggressive in bringing antibodies to the clinic through its broad collaboration with Morphosys (MOR.DE).
Sanofi-Aventis already has two ADCs based on Immunogen’s technology in clinical testing. Last month, Oxford Biotherapeutics out-licensed an antibody program to Sanofi-Aventis, who intends to pursue it as an ADC. As the target for this antibody is proprietary of Oxford Bio, Sanofi will probably have to negotiate a new licensing deal with either Seattle Genetics or Immunogen. Based on the broad collaboration between Sanofi and Immunogen, Immunogen stands a fair chance of getting this project. Based on recent ADC deals, this deal will be much more lucrative than the current deals Sanofi has with Immunogen, with an upfront fee of $5-8M and $200M in milestones.
The two big catalysts for Imuunogen’s partnered pipeline will be data presentation for SAR3419 and T-DM1 in the second and third quarters this year, respectively.
Sanofi will publish the long anticipated phase I data using a once weekly schedule. There are additional CD19 programs in development, including Micromet’s (MITI) MT-103 that continues to generate solid data in NHL. Competition is thinning out following the discontinuation of Medarex’s Fc engineered CD19 antibody, probably due to safety issues. This does not bode well for Morphosys and Medimmune who also have Fc engineered CD19 antibodies in phase I.
Roche will present PFS data from a randomized phase II evaluating T-DM1 vs. Herceptin+ chemo in 1st line HER2+ breast cancer patients. This trial generated a positive signal in terms of response rate in favor of T-DM1 but the PFS data will be crucial as a more reliable endpoint. Since Roche is already enrolling patients in a pivotal study using a similar design in a similar patient population, one can assume that T-DM1 was at least comparable to standard of care.
We are selling our positions in Allos (ALTH) and NeurogesX (NGSX). In addition, we are initiating positions in Pharmacyclics (PCYC) and Endocyte, as both companies have wholly owned agents with solid clinical data. With pharma industry constantly looking for advanced stage products with high likelihood of success, both companies could strike lucrative licensing deals during 2011.
Portfolio holdings as of Feb 13th, 2011