Exceptional Responders
From Paths of Progress Fall/Winter 2014
Finding Genetic Causes of Dramatic Cancer Drug Responses Could Have Broad Benefits
by Richard Saltus
When he decided to try a drug called everolimus for the first time in a rare, highly lethal type of thyroid cancer, oncologist Jochen Lorch, MD, had little hope it would help 57-year-old Grace Silva. "No therapies have ever worked for anaplastic thyroid cancer," he said. In 2011, Silva had exhausted standard treatment and tumors had spread to her lungs.
What happened next some might call a miracle. "Much to our surprise, after about six months, her tumor had almost completely disappeared," says Lorch. This response lasted for 18 months, until the cancer became resistant to everolimus, according to Lorch.
"I feel blessed," says Silva, who continues to battle the disease in 2014. "I thought I would never have grandchildren — and now I have two!"
Researchers call such dramatic, unexpected outcomes "exceptional responses." They typically involve experimental drugs that have failed in most similar cases but are extraordinarily effective in one or a few individuals. Until now, they have mystified physicians, cancer researchers, and drug makers.
But very recently, in a handful of recent reports, researchers have begun solving these puzzles. The answers lie hidden in the genes and chromosomes of a patient's cancer, which were found to contain rare, unsuspected DNA mutations that made the cancer cells highly sensitive to a particular drug.
These "response mutations" — some of them previously unknown to scientists — weren't present in the tumors of most patients, explaining why they did not benefit from the same drug. Silva's case and that of another patient who had an unusually long-lasting remission after drug treatment were reported in 2014 by scientists at Dana-Farber, Brigham and Women's Hospital, and the Broad Institute of MIT and Harvard, along with collaborators at the Whitehead Institute for Biomedical Research.
These discoveries are opening new windows on what determines a tumor's response — or lack of one — to cancer drugs. Studying exceptional responders could help cancer scientists identify other patients who may benefit from drugs that had been deemed failures — and increase the precision of personalized cancer care.
"Traditionally, the exceptional responses have been anecdotal case reports that may or may not ever be repeated, and you couldn't learn a lot from them," says Lorch. "But now we have the tools to find out why these patients responded so well," he adds.
Looking for Rare Responses
Rare drug-response mutations aren't flukes or miracles; they are signposts that may lead to more accurate drug targeting.
More powerful DNA sequencing methods allow researchers to search for the equivalent of a needle in a haystack — combing the large protein-coding genome of a tumor specimen in search of an unsuspected mutation or other critical change in the DNA code.
"Patients with extraordinary responses teach us something important – and what we've learned has applications for other patients, clinical trials, and new ways of thinking," says Nikhil Wagle, MD, a medical oncologist at Dana-Farber/Brigham and Women's Cancer Center (DF/BWCC) and an associate member of the Broad Institute. In 2014, Wagle reported the research on Silva's response at a meeting of cancer researchers and published a report on a second case in the journal Cancer Discovery. He is leading new research on exceptional responders, making use of next-generation DNA sequencing platforms to test tumor DNA.
This research can help create expanded pools of patients whose cancers carry a specific mutation, and who may benefit from a drug, says Dana-Farber's Early Drug Development Center (EDDC) Director Geoffrey Shapiro, MD, PhD. "Now you know which mutations you can look for and design clinical trials for any cancer that has that mutation," explains Shapiro. "We would expect the response rate to be dramatically high."
Shapiro's center runs phase 1 clinical trials of experimental agents. "We've had occasional phase 1 patients who get good responses and stay on the trial for years," he says. "We have been collecting specimen slides and DNA from some of these patients and we are giving them to Dr. Wagle for sequencing, so he can try to find out what characteristics of the tumor dictate the response."
Unlike traditional chemotherapy treatments, which use toxic agents to kill all rapidly dividing cells — both cancer and normal cells — many newer cancer drugs are "targeted." That is, they are aimed at specific molecules involved in cancer growth and survival; these targets are not present in normal cells. Often these targets are networks of genes and proteins that cancer cells have hijacked to force the cells into uncontrolled growth.
Mutations in these genes and proteins enable the cancer to survive and spread by ignoring the normal checks and balances that govern normal cells. These mutations frequently cause growth switches to be stuck in the "on" mode, while disabling the molecular "brakes" that normally curb excessive growth.
The target of everolimus, the drug Lorch tried as a long shot to thwart Silva's thyroid cancer, is a protein encoded by the mTOR gene. The mTOR protein regulates cell growth and proliferation, and the mTOR pathway is overactive in some cancers. Consequently, drugs like everolimus that inhibit mTOR activity have been approved to treat certain cancers and are being tested in other cancer types.
Silva was the lone patient in the trial to have such a dramatic response. As a result, Lorch and Wagle analyzed the tumor's DNA for clues. At the Broad, sequencing instruments thoroughly scanned the critical parts of some 20,000 genes, yielding a long and unsorted list of mutations. Researchers then sifted through the list, looking for a mutation that might have caused a chink in the tumor's defenses.
They found one: a mutation in TSC2, one of two genes that regulate the mTOR protein pathway (TSC1 is the other regulating gene). The scientists then tested the mutations to confirm their effects. Mutations in these genes have been found in a small proportion of some cancers — including some that responded to mTOR inhibitors. Several years ago, Dana-Farber's David Kwiatkowski, MD, PhD, who discovered the TSC1 gene, and Andrew Wagner, MD, PhD, reported that mTOR inhibitors were effective in 50 percent or more of patients with a rare type of sarcoma that had TSC1 or TSC2 mutations.
So the "miracle" of Silva's reprieve from her life-threatening thyroid cancer is apparently due to the good fortune of her tumor having the TSC2 mutation. (Less fortunate was that after 18 months, her cancer developed resistance to everolimus; DNA sequencing found the resistance was caused by new mutations in the mTOR protein itself. However, Lorch has put her on other drugs that are holding the cancer at bay.)
Illustrating how such cases can shape clinical research, Kwiatkowski has designed a clinical trial of everolimus that will be open to patients with any type of cancer found to have TSC mutations. "This will help us determine how often this works," he explains, "and whether any other factors correlate with response to the drug."
Finding the Target
In the second exceptional responder case, Wagle and colleagues at the Broad and the Whitehead institutes studied tumor DNA of a man whose advanced cancer disappeared for 14 months when treated with everolimus. His response was linked to two mutations in mTOR that hadn't been seen in humans. Wagle says that these two cases, along with a third reported in 2012, are making an argument that screening cancer patients for response mutations could be an effective strategy for increasing the power of precision cancer medicine.
In fact, exceptional responder studies are "telling us that in many cases the targets against which we are aiming drugs are too broad," says Jason Luke, MD, FACP, a DF/BWCC oncologist specializing in melanoma research and early drug development.
They don't take into account the rare response mutations that can make all the difference, he argues. By analogy, a letter may not reach the intended recipient if it's addressed only to the state or city, rather than the exact street address. "We are going to have to be much more specific about what, at the molecular level, a target really is," he says.
Rivka Schwartz, a patient of Luke's, is a prime example. Schwartz, who lives in Israel with her husband, Lenny, was diagnosed in 2011 with an aggressive bladder cancer that invaded her kidney. Despite surgery, chemotherapy, and radiation, the cancer's growth had metastasized to her neck and lungs, forming a lump "the size of a tennis ball," says Lenny Schwartz.
Lenny and his brother worked the Internet and consulted with oncologists in Israel and the United States, eventually arranging for genetic testing of Rivka’s tumor. The tumor had a mutation in the FGFR gene, which makes a protein called fibroblast growth factor receptor. FGFR mutations have been found in up to 20 percent of advanced bladder cancers, like Rivka's.
When the Schwartzes learned that Dana-Farber was one of a limited number of centers in the world testing an experimental FGFR inhibitor called BGJ398, they travelled from Israel to Boston. Luke cautioned Rivka and Lenny that participating in the trial would require them to live in Boston for at least several months. They took the plunge in May 2013, when Rivka began taking the drug in pill form.
Recalls Luke: "She started feeling better, very quickly, and we could see her large neck mass immediately start to shrink. It was amazing and must have seemed like a miracle after everything she went through." CT scans showed the cancerous spots on her lungs getting smaller, "and within four to six months, the cancer had essentially gone away," Luke says.
"There is something different about her tumor beyond just the FGFR mutation," says Luke. He suggested to Wagle that he arrange for her tumor DNA to be analyzed to find out "what else is going on." So on a recent visit, the Schwartzes hand-carried a set of her specimens for sequencing.
Meanwhile, Luke says, Novartis, the maker of BGJ398, has expanded the clinical trial "to find more patients like Rivka. As far as I can tell, she has no active cancer."
Exceptional response cases are also being studied in a new initiative at the National Cancer Institute, where Barbara Conley, MD, says about 100 have been culled from clinical trial databases. Tumor DNA from these patients will be searched for response mutations, which could lead to new clinical trials.
Conley says, "You could say we're hoping to repeat the miracles."
Request a Publication
Receive by mail the current issue of a Dana-Farber publication by completing this request form.
Media Contacts
If you are a journalist and have a question about any of our stories or need more information, email media@dfci.harvard.edu or call 617-632-4090 and ask to speak to a member of the media team.
The Media Team cannot respond to patient inquiries. For more information on contacting Dana-Farber, please see Contact Us.