Drug's Approval Marks a Milestone in Personalized Cancer Treatment
You might have missed it, but last month marked a sea change in how cancer is treated. When the U.S. Food and Drug Administration granted accelerated approval
in May to a cancer immunotherapy treatment based on disease genetics rather than type, or sites in the body, it marked the first official recognition of what research at Johns Hopkins and elsewhere has long suggested would be an effective strategy.
The treatment, called pembrolizumab, had already been approved for metastatic melanoma, non-small cell lung cancer, Hodgkin lymphoma, urothelial cancer and some head and neck cancers. Now it can be used in adults or children with any advanced cancer whose cells have defects in so-called mismatch repair genes. These mutations, first identified by Johns Hopkins researchers in 1993, occur in about 4 percent of cancers, disabling cells’ ability to correct errors in the DNA replication process and potentially triggering unchecked cellular growth, a hallmark of cancer.
The approval is “definitely a big deal, and there are several reasons to say that,” says oncology researcher Bert Vogelstein
, co-director of the Ludwig Center at the Johns Hopkins Kimmel Cancer Center
and a member of the Bloomberg~Kimmel Institute for Cancer Immunotherapy.
“First, it’s a treatment that’s applicable to 1 in 25 advanced cancer patients who otherwise have little hope.” In addition, says Vogelstein, it’s the first drug to be approved for cancers found anywhere in the body, solely on the basis of genetic abnormalities within the tumors. This means that all patients with advanced, treatment-resistant cancers could potentially be tested for mismatch repair deficiency and given pembrolizumab.
A third reason is that it combines two previously disparate therapeutic concepts: precision medicine, which targets specific genetic alterations within a specific tumor type, and immunotherapy, which enhances the body’s immune system to fight cancer. “So now you have what could be called targeted immunotherapy. You’re choosing patients to use immunotherapy on the basis of genomic targets,” Vogelstein says.
It All Started with Basic Sciences Research
The basic science research that led to pembrolizumab’s wider approval took place largely at Johns Hopkins, starting in the late 1980s when Vogelstein, working with longtime colleague Kenneth Kinzler
, collaborated with a group from the University of Helsinki to find that defects in mismatch repair genes lead to a predisposition to hereditary nonpolyposis colorectal cancer, or Lynch syndrome. These defects affect about 1 in 200 people. The work, published in 1993 in the journal Cell, led to a change in the standard of care for these patients and their family members. Those found to have an inherited defect in a mismatch repair gene would be regularly screened for cancers to which they were susceptible, while relief was provided to family members who did not inherit a defective gene.
Vogelstein and Kinzler’s lab continued their quest to define the genetics of cancer. One of the next major steps occurred in 2006
, when they reported the first genome-wide sequencing of breast and colon cancers in Science. “That study provided what is now called the landscape of genomic alterations in tumors,” says Vogelstein. In 2008
, they identified the complete genetic blueprint for lethal pancreatic and brain cancers, and with collaborators at Memorial Sloan Kettering Cancer Center, noted that many mutations found in cancers helped create new, foreign antigens. This laid the foundation for later research on what are now known as MANAs (mutation-associated neoantigens).
But the eureka moment for Vogelstein and Kinzler came in 2012, soon after Suzanne Topalian
, director of the Johns Hopkins Melanoma Program, published a landmark paper in the New England Journal of Medicine
. That paper showed that an experimental immunotherapy drug (now called nivolumab) worked well in 28 percent of patients with advanced melanomas and in 18 percent of those with advanced, non-small cell lung cancers—but in only a handful of patients with other cancer types. Just before the paper was published, Topalian and her husband and co-investigator, Drew Pardoll
, came to Vogelstein and Kinzler's lab for help making sense of the discrepancy. During the conversation, they mentioned that just one of 20 or so colon cancer patients treated responded to the drug.
“That piqued our interest,” says Vogelstein, who along with Kinzler and Luis Diaz, an oncologist in their lab, immediately suspected that the one who responded was mismatch repair-deficient. Most colon cancers (and breast, prostate, pancreatic and other cancers) have about 50 mutations, he says, only a small fraction of which are recognized by the immune system as foreign. But a tumor that’s mismatch repair-deficient can have 1,500 or more mutations, presenting many more chances of generating an immune response through immunotherapy treatment. Only melanoma and lung cancers start off with more mutations than most other cancer types, making them the likeliest candidates to be sensitive to drugs such as nivolumab.
“Right then and there we thought we had a good explanation for why melanomas and lung cancers might be more sensitive than these other tumor types,” Vogelstein says. The group also hypothesized that tumors with mismatch-repair deficiency would be extraordinarily sensitive to the drug. “We immediately wrote a paper about that hypothesis,” laughs Vogelstein, “and it was immediately rejected (by a journal) as speculative.”
From Lab Bench to Clinical Trials
Undaunted, the group forged ahead with a clinical trial of pembrolizumab in 20 patients who had mismatch-repair deficient colorectal and other cancers, led by oncologists Luis Diaz (now at Memorial Sloan Kettering Cancer Center) and Dung Le
. Merck, pembrolizumab’s manufacturer, provided the drug for study.
With only 4 percent of cancer patients eligible for the trial, tracking down potential study participants proved tricky, says Le. While tests for defects in mismatch repair are widely available, they are not often conducted for non-colorectal cancers. Initially, it took testing of many Johns Hopkins patients and lots of good old-fashioned networking with colleagues from other medical centers to recruit patients. As soon as they started seeing results—and sharing those results with referring physicians—recruiting became easier, says Le. As excitement grew, the trial was expanded to five additional sites across the country.
It was clear early on that the treatment was working. Trial patients received the drug as a 30-minute infusion every two weeks. While patients were not scheduled to receive a repeat scan to view the tumor until 12 weeks into therapy, there were noticeable differences in some patients after just one or two doses, Le says.
“We had patients who were weaning off of their pain medicine, who were able to do things they weren’t able to do before, and so we knew well before the scans that patients were responding,” Le recalls. “There’s always this potential that immunotherapy might take a while to work. But in these patients, it doesn’t.”
Some patient stories were extraordinary, says Vogelstein. One man in his 30s was just two days from entering hospice care when the lab discovered his cancer was mismatch repair-deficient, so he joined the trial. The man, a graduate student, has since finished his Ph.D. and gotten married. “That’s just one of many, many stories of people who had virtually no hope when they were treated, but whose lives have been significantly prolonged,” Vogelstein says.
The team published their results
in 2015. “These patients were resistant to treatment, so they’re expected to live four to six months,” Le says. First- or second-line treatments, such as chemotherapy, are effective in about 20 percent of these cases. With pembrolizumab, they are noting tumor shrinkages in half of patients, and stabilization of disease (meaning the tumors have not grown, shrunk or spread) in approximately 70 percent. "Most significantly, when patients respond, they tend to respond for a very long time. We’ve had patients who finished treatment three years ago who appear totally healthy and we’re still following.”
Some challenges remain for agents like pembrolizumab, like figuring out how long patients need to be treated. “That’s still the million-dollar question,” says Le. “We have no idea. So far, two years seems appropriate for these patients, because as we’re checking them, and patients are continuing to respond for that long. We hope that if their cancer returns in the future, they will respond again. But we need more time to figure that out.”
The Johns Hopkins investigators also hope to look at beginning pembrolizumab treatment earlier, perhaps even avoiding chemotherapy in some patients. And they want to use the experience with pembrolizumab to try to develop other therapeutics for patients who are not mismatch repair-deficient. “This applies to 1 in 25 cancers, and it can be life-saving, but that leaves 24 for which we have no therapies available once we exhaust the conventional ones,” Vogelstein says. “One step toward changing that will be to further study the cellular immune response to learn how this drug is really working. Drew Pardoll's world-class immunology team is hard at work on that."
For now, says Le, “We need to get the message out to find advanced cancer patients with these mutations so they can be given this therapy.”