Big data yield new hopes, challenges in cancer care
The ability to sequence the genome of thousands of cancerous tumors and compile the data in huge warehouses is giving scientists unprecedented tools for understanding the genetic roots of cancer. That capability allows them to identify who may be predisposed to develop the disease. It also gives drug companies and clinicians new targets for fighting, managing, and in a handful of cases, curing some of the estimated 200 forms of cancer.
Indeed, the explosion of cancer genomic data prompted the White House to launch its National Cancer Moonshot, a billion-dollar effort led by Vice President Joe Biden to “eliminate cancer as we know it” and “make a decade worth of advances in five years.” But while scientists are enthusiastic about their increased capabilities to make progress in the fight against what one author has called “the emperor of all maladies”—cancer now rivals heart disease as America’s No. 1 killer with 589,000 deaths a year—the breadth and depth of the data being generated are posing huge challenges for researchers, oncologists and patients.
The data are complicated to understand, and the process for data collection is unstructured, decentralized, and some argue, only gathered in silos. And much of what is collected remains unanalyzed because there is little consensus on the best ways to pursue data-driven research. Indeed, the current wave of optimism about new discoveries that could lead to
Cancer now rivals heart disease as America’s No. 1 killer, but the breadth and depth of the data being generated are posing huge challenges for researchers, oncologists and patients.
cures might ultimately be pulled out to sea by an undercurrent of unanswered questions.
The environment “is exciting, but frustrating,” says Dr. Barrett Rollins, chief scientific officer at the Dana-Farber Cancer Institute in Boston. “While we all see what big data could potentially do, we’re frustrated because we don’t yet have all the tools or fully understand how to use it.”
It’s a time of “profound turbulence,” adds Dr. Julie Vose, president of the American Society of Clinical Oncology, in her introduction to the group’s recent report, The State of Cancer Care in America.
The report underscores the struggle to make sense of the “information overload” of genomic data. It states that clinicians don’t have answers to the many questions being raised about the risks and benefits of genetic screenings and the role they play in selecting treatments.
In 2015, the Food and Drug Administration approved molecular diagnostic tests for lung and colorectal cancers, a tremendous benefit for some patients with those diagnoses. But while such tests can change the course of treatment for some variants of those diseases, for the majority of cancers, the genomic tests generate inconclusive results or reveal abnormalities for which clinical guidelines and treatments do not exist.
Multidisciplinary teams of clinicians, surgeons, radiation oncologists, geneticists and others have formed molecular tumor boards at some research institutions. They
study these abnormalities and offer guidance to clinicians about which tumor types are worth testing and at which stage of the disease. But the opportunities to use genomic testing to inform clinical practice remain few and far between. From the first sequencing of the first human genome in 2003—considered by Modern Healthcare readers as the most significant medical breakthrough of the past 40 years—cancer patients, their oncologists and the research community have believed the new technology would pave the way for understanding the genomic underpinnings of dozens of cancers.
“The opportunity space suddenly exploded,” says Adam Resnick, founding co-director of the Children’s Hospital of Pennsylvania’s new Center for Data-Driven Discovery in Biomedicine. Sequencing the genome “produced a transformation in the scientific arena,” he says.
Indeed, the advanced understanding of genomics led the drug industry to develop some new treatment options. There are now more than 100 targeted cancer drugs, which work by blocking the action of specific genes and their expressed proteins in cancerous cells, as well as immunotherapy agents, which mobilize the body’s immune system to kill cancer cells with specific genetic mutations. The mainstays of cancer treatment for decades—chemotherapy, radiation and surgery—are now being replaced in some cases by more precise approaches. Of the more than 60 cancer drugs that the FDA approved between 2005 and 2014, nearly 67% were precision medicine therapies, according to the American Society of Clinical Oncology, or ASCO. And 12 of the 15 cancer drugs approved in 2014 were targeted therapies.
Some major cancer centers on the leading edge of treatment and research now sequence the tumor of every cancer patient and use that data to choose the patient’s treatment. At the Dana-Farber Cancer Institute, for example, the results of DNA tests influence clinical care for nearly 3 out of every 4 patients. But just 19% of those patients receive an FDA-approved drug. For another 54%, the sequencing only provides insight into whether they might qualify for a clinical trial, Rollins says.
The quest now is to expand the paradigm, so that more than just a minority of cancer patients can get enrolled in trials. “The only people who get this detailed sequencing are either rich enough to travel to a major cancer center or just happen to live near one,” Rollins said. “Insurers should strongly consider reimbursing for it ... which will democratize the process.”
Some insurers are already moving in that direction. In January, Independence Blue Cross, a Blue Cross and Blue Shield affiliate based in Philadelphia, said it struck a deal with health technology firm NantHealth to offer coverage for genome sequencing. Commercial plan members with cancers that did not respond to traditional therapies gained access to the testing in March.
The steady growth of new cancer therapies based on genomics has added to an already huge problem for practicing oncologists. Determining the best course of treatment for a cancer patient has become a daunting task. A recent policy statement from ASCO called the tens of thousands of existing clinical pathways “overwhelming and unsustainable.”
The silos within which many researchers work on new cancer genomics pose another problem. As cancer centers and their molecular review boards evaluate patients, they need to form collaborations in order to generate a critical mass of data for meaningful analysis. “To understand what exactly is taking place, we have to position that patient in the comparative space of patients with similar and different forms of cancer,” Resnick says. “That’s the only way you can draw interpretations.”
Cancer researchers are beginning to coordinate across institutions. They’re also seeking ways to quickly disseminate their findings to oncologists in communities without a cancer center. Community oncologists treat about twothirds of cancer patients. But they often lack access to the latest research and related advances, according to a recent commentary in the New England Journal of Medicine by Dr. Francis Collins, director of the National Institutes of Health, and Dr. Douglas Lowy, acting director of the National Cancer Institute.
ASCO would like to enlist community oncologists in generating another big-data stream that could be used for improving cancer care—in this case, getting the most out of existing therapies. Last year ASCO launched CancerLinQ, a project that captures data about genomic tumor types, treatments and outcomes. The idea is to analyze patterns in that data to determine which treatment protocols have worked best for which patients. That information could enable more personalized treatment for future patients.
A moonshot can only be attained by amassing data and analyzing all cancer patients who come into physicians’ offices, “not just that small minority who are enrolled in randomized controlled trials,” says Kevin Fitzpatrick, CEO of the CancerLinQ project.
CancerLinQ officially rolled out in June 2015 with 15 participating oncology practices. While ASCO received more than 190 inquiries from groups wanting to participate, a spokesperson told Modern Healthcare that so far only 40 have signed contracts to join. They cite patient-privacy concerns and difficulties with the technical interface. They also worry about the workload required to input all the data.
Dana-Farber supports the effort but says its cancer center is not participating because for now “the benefits of joining are not yet worth the personnel costs,” Rollins said. He is on the steering committee of a separate project run by the American Association for Cancer Research, called Project GENIE. Seven cancer centers are participating in that project, which aims to mine data on tumors that have been sequenced, in order to identify trends and provide researchers with more of the basic science that explains how cancer happens.
Immunotherapy is another area in which big data is advancing research. In March, Johns Hopkins Medicine launched a new cancer research center to focus specifically on immunotherapy. It received $125 million in funding, including $50 million given by former New York City Mayor Michael Bloomberg and another $50 million from philanthropist and movie producer Sidney Kimmel.
Dr. Drew Pardoll, the inaugural director of the new Bloomberg-Kimmel Institute for Cancer Immunotherapy, hopes the center will generate innovative clinical trials and global collaboration with scientists and cancer centers. Focused research “will advance immunotherapies to the point where the immune system will ultimately be able to beat 100% of cancers,” Pardoll said. “The potential to control or cure even the most advanced treatment-resistant cancers has been elusive until now.”
The biggest challenges come down to amassing the incredible amount of knowledge gained about cancer and translating that knowledge into new therapies that can help patients and provide new guidelines for the oncologists who treat them.
But some experts worry that the rush to research the possibilities generated by big data places too much emphasis on developing breakthrough drugs and other therapies on the “exotic ends of the spectrum,” says Dr. Leonard Lichtenfeld, deputy chief medical officer of the American Cancer Society. Such initiatives should not divert attention from a focus on prevention, he says. That concern was shared by the deans of more than 70 public health programs, who sent a letter to Vice President Biden in March. Hype over the cancer “moonshot” may be undervaluing the roles that public health and prevention play. “History has shown that the greatest impact in reducing cancer mortality rates has come from preventing cancers,” the letter stated.
Curative treatments often appear more exciting to the public, and they are essential, but controlling cancer is also a policy and public health challenge. “We must operate on both fronts,” the letter said.