Along with the acquisition of Immunex in 2001 - mostly to gain rights to Enbrel, which DID turn into a ginormous blockbuster - Amgen acquired a hodgepodge of early and mid-stage leads in the Immunex pipeline.
In most acquisitions like this, it seems like there is little value placed on the early leads of the target company - most of the attention is on the flagship, late stage product (Enbrel, in this case), and even then, the incoming programs & leads have to compete for resources with the incumbent R&D programs. Since these early stage leads have little data accumulated, and may involve new biology, little ever becomes of the "bonus" leads. (Also, after a merger, the acquiring company becomes even more of a "product" company, and also needs to cut cash costs post-merger to meet the financial expectations justifying the merger.)
Hats off to Amgen, though, for sticking with a lead sourced from Immunex, as today it was disclosed that an antibody targeting HGF (for cancer) has enough anti-MET activity that it generated strong response in certain gastric cancers.
(The Amgen antibody is referred to as either AMG 102 or rilotumumab. An excellent press release with lots of technical details is here, including the survival data. (11.1 months overall survival WITH AMG 102 versus 5.7 months without for patients with high MET expression.)
This story is really exciting for a reason beyond the notion that we might have another weapon in the war on cancer, and the notion that Amgen is a great research organization.
Searching Clinicaltrials.gov leads me to believe that this trial is the one that generated today's good news. If so, note that the trial was launched nearly four years ago, and there is NO mention of MET protein or biomarkers in the trial design.
It seems to me, then, that some astute Amgen researchers - after the trial started and likely after Amgen had already concluded that AMG 102 wasn't effective enough to continue development - decided to sift through the results on a patient by patient basis to see if there was a genomic commonality with the responders. The researcher (or team) then used either a shotgun profiling approach to identify a biomarker to stratify the patients or decided to specifically assay for MET expression. I'd guess that the first approach was used, but either way, Amgen boldly applied genomic technology to do something oft-rumored but seldom seen - a biomarker-driven compound rescue.
(And even if I'm wrong, and Amgen was onto MET stratification all along, it is still an impressive finding.)
To temper the excitement somewhat, the clinical trial found only 27 of 82 (32%) gastric cancer patients in the trial had over-expressed MET protein, so it is likely then that the FDA label for AMG 102 - like Xalkori - will be very targeted and linked to a companion diagnostic. However, as discussed here before, MET is a hot target with likely broad applicability, so AMG 102 might someday reach blockbuster status.
Also, due to its' origins back in Immunex's labs, 11 or more years of the patent life of AMG102 may have already lapsed. I would guess that someone within Amgen has just been tasked with conjugating AMG 102 with something interesting in order to boost efficacy and reset the patent clock. (Immunomedics: make sure someone is manning your phones!)
The other possible winner here: Francis Collins & his NCATS initiative. Amgen's work on '102 reinforces that there is value in rummaging through pharma programs to find new applications.
AMG 102 is another manifestation of the Molecular Future hypothesis, and more proof that targeted therapy is effective. It sure seems like targeted therapies have a higher-than-average clinical success rate once safety has been demonstrated in a Phase I trial. (And maybe those targeted therapies that haven't succeeded just need the right biomarker.)
ADDENDUM: on reflection, I do not believe that Amgen's results are predictive of success for NCATS. My understanding now is that NCATS is less about further qualifying a lead through advanced technology (genomics, etc.) than about starting de novo development in new applications for drugs that have previously made it into clinical trials. Other than the fact that the compounds under consideration are likely safe for humans (i.e. the leads had successful Phase I trials), the NCATS approach provides no running start or any reason to anticipate better than average discovery success. (In fact, I'd argue that a bureaucratic organization like the NIH is LESS likely to have discovery success.)
On a positive note, let's hope that Amgen's success can be similarly applied to the rest of the bio-pharma industry.