When Adrienne Skinner was diagnosed with ampullary cancer, a rare gastrointestinal tumor, in early 2013, it didn’t come as a complete surprise. For nearly a decade, she had known her genes were not in her favor. What she didn’t know was that her genes would also point the way to a cure.
Cancer has long been defined by where it starts to grow—the lungs, the colon, the breasts. But the recent approval of a cancer immunotherapy for anyone harboring a specific genetic profile, regardless of the tissue in which it is found, signals a shift in how researchers think about cancer. Read on to learn more about how technology and drug development are converging to realize the promise of “personalized” medicine.
Skinner has Lynch syndrome, an inherited disorder caused by a defect in mismatch repair (MMR) genes, which encode for proteins that spot and fix mistakes occurring during DNA replication. People with Lynch syndrome have an up to 70% risk of developing colon cancer. Women with the disorder have similarly high chances of developing endometrial cancer at an early age.
The first time Skinner heard about the syndrome was in late 2004, after her sister was diagnosed with colon, ovarian, and endometrial cancers, the telltale trifecta associated with Lynch syndrome. It turned out that Skinner, her sister, and their mother were all carriers of deficient MMR genes.
After Skinner spent a year responding to—and then not responding to—two types of chemotherapy, her oncologist suggested she look into a novel trial under way at Johns Hopkins Kimmel Cancer Center. Clinicians there were testing a drug called Keytruda in cancer patients who have gene defects like hers.
Keytruda, developed by Merck & Co., is part of a wave of new treatments called checkpoint inhibitors that help the immune system recognize and attack cancer cells.
Although remarkably successful at treating skin and lung cancers, checkpoint inhibitors weren’t eliciting the same results with colon cancer. The team at Hopkins had a theory about why only a handful of colon cancer patients benefited: Like Skinner, they harbor defects in MMR genes. The researchers convinced Merck to give them the drug and found a nonprofit to support a study to test their hypothesis.
Once every two weeks, Skinner took a train from her home in Larchmont, N.Y., to Baltimore, where she was given an infusion of Keytruda.
Less than three months into the study, she went in for a biopsy to gauge whether the drug was kicking her immune system into gear. The surgeon who walked in after the procedure delivered incredible news. Skinner recalls he looked at her and said, “You know, if somebody hadn’t told me you had ampullary cancer, I wouldn’t have known, because there’s nothing in there.”
Skinner isn’t the only patient to experience that kind of dramatic response. Clinicians later reported that the immunotherapy works in people with all sorts of cancers that are characterized by MMR deficiency or a related condition known as microsatellite instability.
In a trial of 149 patients who had not responded to more conventional cancer drugs, tumors shrank in roughly 40% of those with colon cancer and 48% of those with other types of cancer.
The results were “unbelievable,” says Luis Diaz, who conceived the trial while at Hopkins and is now head of the solid tumor oncology division at Memorial Sloan Kettering Cancer Center. “Things never happen this way. I mean, 80% of the ideas one has fail.”
In May, swayed by the Keytruda data, the U.S. Food & Drug Administration granted its first-ever approval of a cancer therapy for patients harboring a specific molecular profile. The “tissue-agnostic” approval is perhaps the most public example of an ongoing redefinition of how cancer patients are treated.
Cancer care has always centered on the organ where a tumor is born—the lungs, the breasts, the colon. Even in this much-heralded era of “personalized” medicine, drugs that target genetic aberrations are still approved for use in specific organs. Moreover, companies still largely need to run separate trials to prove a drug’s efficacy in each organ.
Now, thanks to cheaper and faster genetic sequencing, researchers are thinking differently about cancer. With the technology, they can more easily match targeted treatments or cancer immunotherapies to the patients who have the molecular makeup to benefit from them.
More and more early- and midphase clinical trials, known as “basket studies,” are looking beyond the organ of origin and welcoming anyone with a specific genetic profile. Now that FDA appears open to a genetics-focused development approach, experts expect the oncology field to shift from its preoccupation with a tumor’s location.
The overarching goal of the tissue-agnostic approach is to make cancer drug development more efficient. Looking beyond the location of a tumor promises to speed up treatment options for people such as Skinner, who otherwise might not have known they would benefit from a drug.
Merck’s tissue-agnostic approval is “a watershed event,” says Razelle Kurzrock, head of the Center for Personalized Cancer Therapy at the University of California, San Diego’s Moores Cancer Center.
Kurzrock explains that oncology has for decades defined cancer cells by how they look under a microscope, leading to today’s organ-centered categories. Now, instead of looking at the surface of the cell, she says, oncologists are more often looking inside the cell to identify what is making it abnormal.
“Scientifically, that makes so much sense,” Kurzrock says. “You’re hitting the fundamental alteration that is driving cancer rather than the superficial appearance of the cancer.”
Still, the focus on the organ of origin persists, even as companies turn to designing drugs that block cancer-causing genetic mutations found in many tumor types. “The idea of precision medicine actually does work, but up until now, it has been largely tumor restricted,” says Roy Baynes, Merck’s senior vice president for global clinical development. “The real hope is this mechanism-based approach will translate more broadly.”
But challenges abound. Although oncologists laud the tissue-agnostic approval of Keytruda, they also caution against overselling genetics as a panacea.
“It’s definitely a step forward but not a definitive solution to all cancers,” says Igor Puzanov, director of early-phase clinical trials at Roswell Park Cancer Institute.
Among cancer researchers such as Puzanov, the BRAF gene is the fly in the ointment of tissue-agnostic drug development. BRAF mutations are ubiquitous in cancer but most commonly found in melanoma, colon cancer, and thyroid cancer. When the BRAF inhibitor vemurafenib was discovered, researchers hoped the molecule would destroy cancer cells in people with any of these three cancers.
It didn’t. Vemurafenib, which specifically blocks the BRAF V600E mutation, works remarkably well against skin cancer—Roche won approval for the drug to treat melanoma in 2011—but hardly any colon cancer patients respond to it. Researchers have spent several years trying to understand the discrepancy. Even as they come up with plausible theories, the failure of BRAF inhibitors to work in both kinds of tumor has dogged the field.
UCSD’s Kurzrock argues that dismissing tissue-agnostic approaches based on the BRAF story is shortsighted. No cancer drug works in every person. Even vemurafenib elicits a response in only half of melanoma patients with BRAF mutations. The other half have additional mutations that also need to be blocked.
Colon cancer is no different, Kurzrock says. The key is figuring out the other pathways involved so appropriate treatment combinations can be pursued. For example, adding an EGFR inhibitor to a BRAF inhibitor elicits a response in colon cancer patients.
Shifting how the medical establishment thinks about cancer means overcoming a mind-set that has persisted for decades.
“While we live in a research world where we sequence 600 genes for every tumor, you’re going to be worked up in a diagnostic setting that is different if you have breast cancer versus lung cancer,” says Wendy Winckler, head of next-generation diagnostics at the Novartis Institutes for BioMedical Research. Thus, a patient with, for example, breast cancer typically isn’t tested for EGFR mutation, which is commonly found in lung cancer.
Moreover, for drug companies it was simply easier to seek an approval for melanoma patients, 30–45% of whom have BRAF V600E mutations, than for non-small cell lung cancer patients, just 1–2% of whom have the mutation. To run a clinical trial in lung cancer, 100 patients would need to be screened to find just one to enroll. “That wasn’t happening before,” Winckler says.
Clinicians hope to make drug development more efficient by changing the way some trials are run.
But the availability of broad screening panels is changing that paradigm. “Historically, the world has been rate limited from doing this kind of drug development largely because of diagnostics,” says Joshua Bilenker, chief executive officer of Loxo Oncology. Just five years ago, he notes, the next-generation gene sequencers that can test for a large swath of molecular drivers of cancer didn’t even exist.
“Broad testing is how you find things you never even knew you were looking for,” Bilenker says. This was the case for the rare mutations targeted by Loxo’s most advanced drug candidate, larotrectinib.
Loxo is developing larotrectinib as a treatment for any cancer patient harboring TRK gene fusions, which occur when chromosomes break apart and then rejoin in the wrong place. Between 1,500 and 5,000 patients who are newly diagnosed with advanced cancer each year have a TRK fusion, Loxo estimates, meaning the genetic error appears in less than 1% of all cancers.
Larotrectinib, which inhibits the fusions, stole the spotlight at last month’s annual meeting of the American Society of Clinical Oncology in Chicago. Researchers presented a study that tested the drug in 50 children and adults who had TRK fusions across 17 cancer types, ranging from rare tumors to common cancers, such as colon and lung. An astounding 76% saw their tumors shrink, and the drug continued to work for those responders a year into the trial.
“The truly tumor-agnostic activity we’ve seen is a bit surprising, even to us,” Bilenker says. Loxo plans to apply by early 2018 for FDA approval of the drug for anyone with TRK fusions.
Last month, the biotech firm also began a trial of its next-generation TRK inhibitor, LOXO-195. Although many patients have seen sustained responses to larotrectinib, cancer cells inevitably develop resistance to targeted agents. LOXO-195 was designed to lock the conformation of TRK into place, overcoming resistance.
Another company, Ignyta, will also seek approval next year of a drug for people with TRK fusions.
Ignyta is taking a slightly different approach for its lead compound, entrectinib, which blocks fusions in TRKs, ROS1, and ALK. It plans in 2018 to seek a tissue-agnostic approval in people with TRK fusions and an approval in lung cancer for people with ROS1 mutations.
The tissue-agnostic development pathways that Loxo and Ignyta chose are outliers. Although several other large basket trials are under way, most are geared toward finding signals of efficacy before companies move on to trials in specific organs or tissues. Still, researchers hope that some of those studies will reveal drugs with broad efficacy.
The National Cancer Institute (NCI) recently announced that it has sequenced the tumors of 6,000 people as part of its Molecular Analysis for Therapy Choice, or NCI-MATCH, trial. The study started enrolling patients in August 2015 with the goal of pairing anyone whose tumor has a particular molecular makeup with one of 21 drugs or drug combinations.
So far, about 19% of the people recruited have been matched with a drug or drug combination, and more than half of them have rare cancers, says Barbara Conley, the associate director of NCI’s Cancer Diagnosis Program. Conley is responsible for overseeing the NCI-MATCH study.
The trial is designed to find signals that the targeted treatments are effective. Still to be seen is whether the signals point to broad use in patients who share mutations or suggest efficacy in specific organs. “There are going to be some drivers that are so strong that they will drive a response and benefit across tumors,” Conley says, but she expects responses limited to individual organs to be more common.
Novartis, meanwhile, has for several years been running what it calls Signature trials, which similarly match patients to one of its targeted therapies.
Since the program launched in 2013, Novartis has studied more than 600 patients who have 15 types of cancer and were given a range of experimental compounds, says Richard Woodman, Novartis’s head of North American oncology clinical development.
Looking ahead, researchers see several opportunities for tissue-agnostic drug approvals. Everyone points out that high tumor mutational burden, a measure of the number of gene mutations in cancer cells, anecdotally correlates with positive response to checkpoint inhibitors such as Keytruda.
Researchers are also interested in exploring whether PARP inhibitors—compounds that block an enzyme that helps patch up tumor DNA and are already approved to treat BRCA-mutated ovarian cancer—could be broadly effective against all BRCA-mutated cancers. And ongoing studies are testing whether HER2-targeted breast cancer treatments could be effective in other HER2-mutated tumors.
“The futuristic world is broad sequencing assays are used in nearly all routine cancer workups to find whatever it is that leads to the right therapeutic option for the patient,” Loxo’s Bilenker says.
FDA seems invested in clearing the path for developing drugs based on genetics. Last month, in testimony to a Senate subcommittee, FDA Commissioner Scott Gottlieb said the agency will this year release a new policy that “will address the issue of targeted drugs and how we simplify the development of drugs targeted to rare disorders that are driven by genetic variations, and where diseases all have a similar genetic fingerprint, even if they have a slightly different clinical expression.”
Regardless of what comes next, patients such as Skinner who have rare mutations are thankful for the recent advances. Skinner stopped treatment with Keytruda in April 2016 and more than a year later remains tumor-free. Given her genetic makeup, she knows her cancer could return. But she is relieved to have access to an approved drug that could address whatever comes next.
That relief isn’t just about her own future. Skinner has four daughters, three of whom have tested positive for Lynch syndrome. “I can face near-certain death—and I did—and I’ll tell you that I was ready for it,” she says. “But the fact that my kids are at risk was the worst. That now makes this drug trial and fabulous result that much more meaningful for me.”