Some ideas sound good in theory but just never manage to work out in real-world experience, and cancer vaccines are on their way to that destination. Following the failures of Vical and Merck KGaA with their respective cancer vaccines, GlaxoSmithKline (GSK -0.83%) has announced the second pivotal trial failure for its MAGE-A3 cancer vaccine. Glaxo will continue to try to identify patient sub-populations in both melanoma and lung cancer that could/do respond to a clinically significant degree, but the odds are good that this high-risk/high-reward program is on its last legs.

Another miss for MAGE-A3
Glaxo announced on Thursday that a Phase III study (MAGRIT) of its MAGE-A3 cancer vaccine failed to meet its co-primary endpoints in a study in non-small cell lung cancer. This follows the failure of MAGE-A3 to show sufficient efficacy in patients with melanoma last year and is broadly consistent with expectations that there was only a slim chance that this compound was going to prove effective (most sell-side analysts seemed to be factoring in only a 10% to 20% chance of success).

Glaxo is not entirely finished with MAGE-A3 just yet. The company's scientists are going over the data from both melanoma and lung cancer in the hope of finding identifiable patient subgroups that show a stronger, clinically relevant response to the vaccine. Glaxo is not generally a company that refuses to accept failure and looking for subgroups is a valid notion, as tailoring cancer treatments to particular genotypes (personalized medicine) is a validated approach.

This news was a minor negative for Glaxo, in no small part because the odds of success were seen as low and the resulting risk-adjusted expected contribution to Glaxo's future sales was quite modest. For Agenus, the small biotech that developed the adjuvant for MAGE-A3, the consequences were more serious and the shares fell nearly 12% on the news.

Vaccines on their heels...
Cancer vaccines sound great in theory, but they have yet to really make a mark in real-world oncology. Dendreon's (NASDAQ: DNDN) Provenge is approved and actively marketed, but the drug has failed to catch on as a prostate cancer treatment. Provenge adds about four months to median survival time for metastatic prostate cancer, but the procedures involved (blood is taken from the patient and sent to Dendreon, who manufactures a customized vaccine) and the high cost of treatment led to disappointing adoption and less than $300 million in annual revenue.

On a relative basis, though, Dendreon's Provenge has been a rousing success. Merck KgaA's Stimuvax / tecemotide (partnered from Oncothyreon) also failed in a lung cancer study, but like Glaxo Merck is hoping to find an identifiable patient subgroup that responds better. Vical announced the failure of its Allovectin melanoma vaccine in 2013 and abandoned the program.

... but not done yet
The mechanism of action of oncology vaccines make too much sense to abandon the approach entirely. Cancer vaccines can be broadly divided between those that stimulate/prompt the patient's own immune system to destroy cancer cells and those that directly target and attack cancer cells, and there are still products in pipelines worth watching.

Glaxo has its early stage PRAME ASCI vaccine for lung cancer and the WT1 vaccine for breast cancer. Aduro has had some encouraging early stage success with its GVAX/CRS-207 approach, and Celldex (CLDX -2.78%) is approaching an interim analysis in the pivotal study of its rindopepimut vaccine for EGFRvIII-positive glioblastoma. Rindopepimut showed a nine-month improvement in overall survival in a Phase II study, which is quite encouraging for this difficult-to-treat cancer. Celldex also has the early stage vaccine CDX-1401 in trials in various solid tumors.

Last and not least is Amgen's (AMGN -0.50%) T-Vec (talimogene laherparepvec). T-Vec works as both an oncolytic vaccine (directly destroying cancer cells) and a stimulant to the immune system and clinical results have been encouraging. Phase III overall survival data in melanoma should be available soon, and although PD-1/PDL-1 therapies may ultimately prove more successful in melanoma, success in this indication would provide some much-needed validation to the approach.

The bottom line
The good news for Glaxo is that although it seems unlikely that the company will recoup the money spent on developing MAGE-A3, the failure of this vaccine only reduces the long-term fair value by about 1% or 2%. Looking at the wider field, it remains to be seen whether cancer vaccines have a permanent seat at the table as a viable oncology approach, with Celldex and Amgen likely the best bets for near-term success.