That's the conclusion of an article in the Financial Times earlier this week that highlights the growing importance of companion diagnostics for the pharmaceutical industry.
Traditionally, medicines have been given to large numbers of patients with an apparently common disease, all the while accepting that they will be a failure for many and cause significant side effects. Genetic testing identifies the smaller numbers of sufferers in whom the drugs work, reducing costly and ineffective treatment in others.
The article notes that twenty years ago, cancer drugs might only be effective in ten percent of the patients treated. Today, new diagnostics linked to drugs like Erbitux, used to treat colon cancer, can identify the 60 percent of patients without a mutation in the KRAS gene, which makes them more likely to respond to treatment. The diagnostic spares patients who don't benefit the risk of serious side effects, and ensures that drug spending goes to the patients who are most likely to benefit. It also allows drugmakers to charge higher prices to offset the tremendous costs of developing sophisticated new medicines.
The development of personalized medicines still faces significant hurdles. Cancers typically develop resistance to even targeted drugs, effectively evolving in real time to bring new cancer promoting growth mechanisms online. Companies and researchers need to do more to understand the complex signaling networks driving cancer growth and mutation, and develop cocktails that can check or slow the development of drug resistant cancers.
Over time, diagnostics willl likely shift away from single gene mutations and towards more complex proteomic (and other "-omics") tests to measure these network interactions and target cocktail therapies appropriately from the initiation of cancer treatment. Some of these cancer roadmaps are already in development for leukemia.
And cancer is far from the only example: it has also become increasingly apparent that complex chronic diseases are rarely the result of a single gene mutation, but are driven by networks of complex gene and protein interactions. Scaling up our understanding of these networks and translating them into the clinic will be an enormous undertaking - and will be unworkable if taking a single new drug to market still takes over a decade and costs over a billion dollars.
We're starting to see the first glimmers of resarch networks take shape that can take advantage of new genomic technologies and sophisticated IT architecture - through, for instance, the NIH's cancer Biomedical Informatics Grid, defined by "information liquidity" and breaking down the "invisible wall" that has traditinally separated the research and treatment communities.
There's a long way to go before the full promise of personalized medicine is realized.
But there's no going back.