For over two decades, Genentech scientists have been researching and developing medicines that are targeted to treat specific patients. But in 2010, the company decided to make that practice a priority. For every potential new medicine in clinical development, researchers would look for a biomarker or a companion diagnostic at the outset. This would help determine which patients, even if those subsets of patients are small in number, could be most likely to benefit from the medicine. Today, more than 60 percent of the company’s medicines are being developed with a companion diagnostic.
Traditionally, new medicines have been developed with a kind of “one size fits all” approach. But as research techniques have grown more sophisticated and the complexity of patient biology has come into focus, scientists have realized that there are genetic differences among people with the same disease, and that those individual mutations can be targeted with more specific treatments.
In the 1980s, researchers first identified a gene that some breast cancer patients over-expressed. Then in the 90s, they created the first “personalized” medicine to target that particular gene. At the same time, a diagnostic tool was developed so that physicians and labs could test whether or not individual patients who had those genes could be targeted with the treatment.
“Years ago, a patient who presented with breast cancer was just labeled as a ‘breast cancer patient,’” says Priti Hegde, director of oncology biomarker development at Genentech. “Today she could have HER2-positive breast cancer, triple-negative breast cancer or hormone receptor-positive breast cancer.”
Now, genetic variations have been found in many other diseases. “Over time we’ve come to realize that most common diseases are very heterogeneous. Two lung tumors might look similar but have very different underlying causes,” says Garret Hampton, vice president of oncology biomarker development. “What we try to do is understand the heterogeneity of those diseases, so we can group patients by their genetic profile and then tailor our medicines to those patients.”
As Hegde puts it, the new realm of personalized medicine created a two-way street―science meeting the patient, but also the patient's biology meeting the science.
In order to target patients with specific genetic variations, scientists need to develop new tools, or companion diagnostic tests, to identify these patients. “It could be a blood test, an imaging test, a DNA test or something as simple as age,” says Andy Chan, senior vice president of research biology. “But in each of our programs, we want to think through how to select the right patients, and the right dose, in a scientifically driven way.”
“Now it’s not a hammer approach, it’s a scalpel,” says Hegde.
In swapping hammer for scalpel, Genentech began to match the pace of drug development with diagnostic capabilities. In 2010, Genentech’s Research Review Committee (RRC) formally instituted a requirement that for every molecule in early discovery, a companion diagnostic strategy would be developed simultaneously.
In some ways, this represented a pivotal moment for Genentech, when it became fully committed to personalized medicine. But in other ways, says Chan, it solidified a paradigm that had already begun to shift. “It was something we’d done [informally] for decades,” he says. “We’d been one of the leaders in this philosophy.”
In 2010, that philosophy become an official modus operandi and a bit of common sense was structured into a forward-thinking approach. At the time, Genentech had recently merged with Roche, a company with a robust diagnostics program. The entire organization united around this common commitment and goal of personalized medicine.
“It’s always been very much central to the way we developed medicines,” Hegde agrees. “By making it a common goal, it became an active approach rather than passive, with formal hypothesis testing early during the clinical development.”
In the past, a trial for a lung cancer drug that worked on only 5 percent of the patients might have been deemed a failure. But when a trial is aimed at that 5 percent, a subset of patients who have a particular biomarker, and it can be proven effective for those patients, then it is successful at helping that small, but important, group.
Researchers at Genentech exploring a potential new therapy now know that parallel teams are gathering diagnostics information at the same time. As a result, promising research doesn’t have to hit the brakes to await diagnostics data, so it helps speed up the development. What’s more, the company now assures that funding and resources for this approach will always be there.
“Ultimately, diagnostics is front and center to personalized health care,” Hegde says. "From a patient’s perspective, I’d want to know what my chance of responding to a particular therapy is. That's what diagnostics can do for patients.”
In the realm of drug development, creating an effective medicine is only part of the job. Matching that medicine to the right patient is every bit as crucial. As Genentech moves further into the future of personalized health care, combining those tasks from the very beginning has gone from a smart idea to an essential practice.