Inside a molecular tumor board

Using genomics to guide personalized cancer treatment

Illustration by Gérard DuBois

“OK, people, mute your phones. Mute your phones, please!”

The buzz in the first-floor conference room in the Blake Wilbur Building on the medical school campus slowly subsides as people take their places around a long table and in chairs set along one wall, plopping down and opening laptops to display medical charts and presentation slides. It’s 2 p.m. on a Friday in December, and the Molecular Tumor Board is about to convene.

“Let’s get started,” says James Ford, MD, professor of medicine and director of Stanford’s Clinical Cancer Genomics Program, loudly over the chatter of voices and jovial laughter.

Stanford’s Molecular Tumor Board was launched in spring 2015 to use the molecular attributes of individual tumors and identify novel drugs that might help last-resort patients — those with advanced cancers that have resisted standard treatment. It’s one of a handful of such efforts in the country.

Since the board was created, the group has met twice a month to brainstorm about cancer cases referred to them from physicians inside and outside of Stanford. On the slate for the board on this day is a deep dive into two cases of women with cancer — one with breast cancer with an unusual combination of mutations and another with colorectal cancer.

Having an open mind is crucial. Traditional cancer care categorizes cases based on the type of tissue involved — is it in the lung, colon or breast, for example — and the stage at which the cancer was diagnosed. Has it invaded surrounding tissues or metastasized to distant sites? How large is the primary tumor? What are the patient’s clinical symptoms?

Tumor boards across the country are usually similarly organized, bringing together experts from various medical disciplines to tackle cancers of particular types; Stanford has about 13 tumor boards focused on different tumor types, including breast, head and neck, thyroid and gynecologic cancers.

By contrast, Stanford’s molecular tumor board is filled with experts who want to look under the hood of intractable tumors and identify, at a genetic level, what makes them tick. Together, oncologists, pathologists, cancer geneticists and genetic counselors, informatics experts, and medical fellows and residents comb through a labyrinth of data points, including any mutations in a tumor’s DNA sequence that could be targeted by existing drugs.

They also review patient treatment histories, which could determine whether the patient will qualify for a particular clinical trial or off-label treatment. Oncologists who wish to refer a patient to the board fill out a one-page referral form, and the patient’s primary care team is encouraged to participate in the discussion in person or in a teleconference.

This sea change in approaching cancer treatment is, in part, a result of the realization that cancers often share mutations in key genes controlling cell growth or signaling pathways. It also is a result of new technological abilities that allow rapid DNA sequencing in a fraction of the time and cost.

The National Human Genome Research Institute estimates that the cost to generate a high-quality human genome sequence was about $14 million in 2006; the same task, 10 years later, cost less than $1,500. Stanford’s Solid Tumor Actionable Mutation Panel, which combines targeted sequencing with a bioinformatics analysis of known or suspected cancer-associated mutations, delivers results to patients’ clinicians within three to four weeks — with insurance often covering the cost.

“We can now sequence hundreds and thousands of DNA sequences within a tumor in a time frame that is achievable,” Ford noted. “And this group includes people very experienced in interpreting test results with lots of subtleties. Often we will identify tens or thousands of individual mutations in one patient’s tumor. We need to decide: How do we prioritize these? Which are likely to be the driver mutations that the cancer relies on to grow? And, once we identify a potential targeted therapy, we discuss how to get that drug to the patient: Is there a clinical trial in which the patient could enroll? Might it be possible to get the drug for off-label use?”

Recommending therapies is one challenge; showing that they work and are cost-effective is yet another difficulty.

“We’ve definitely had instances in which people with widespread metastatic cancer have had tremendous responses to the targeted therapy we’ve recommended,” said Ford. “The challenge is that, with this type of personalized medicine, we’re not evaluating treatments in the standard way, with a clinical trial that enrolls many people. Instead, each patient is, in theory, unique — as determined by the combination of mutations found within their tumor.”

“We can now sequence hundreds and thousands of DNA sequences within a tumor in a time frame that is achievable.”

James Ford, MD, director of Stanford’s Clinical Cancer Genomics Program

Ford is leading a collaboration with Utah-based Intermountain Healthcare, which has the nation’s largest repository of biological samples. By linking molecular data from the database of millions of samples to the health outcomes experienced by the people who provided them, researchers hope to advance the understanding of many diseases, including cancer.

“Intermountain can capture information about an entire population, so we can compare data about even very rare mutations and combinations and learn how best to treat these patients,” Ford said.

But on this December day, Ford and his colleagues aren’t thinking about the big picture. Instead they are hyper-focused on a few select cases referred to them by oncologists at Stanford Health Care and a few neighboring medical centers.

In addition to two main patients, the team will, in rapid-fire fashion, consider seven others — some of whom had previously appeared before the board. Patient photos appear alongside their medical charts, and a familiar patient’s case is greeted with a murmur of recognition and concern.

The team quickly gets down to business with a review of the relevant scientific articles from Cancer Discovery and Nature. Rochelle Reyes, a physician assistant who works closely with Ford on the tumor board, is the presenter. She details the molecular vagaries of an important signaling protein, analyzing how the patient’s mutation is likely to affect the three-dimensional shape of the resulting protein and its ability to bind to other proteins to affect cell growth.

“This particular mutant variant is very unique,” Reyes notes, calling out a deletion that might drive tumor growth by causing the pathway to be abnormally active. “But if we try to block this pathway, this second variant protein might take over. Could PARP inhibitors be used in this context? It’s unclear.”

After about 10 minutes of review, the floor opens for discussion, mediated by Ford, who begins by handing out a printed page listing current information about clinical trials at Stanford and elsewhere. “I think there are several potential options for this patient,” he said, noting that he’d normally recommend a particular clinical trial, but “she’s not eligible because of these other mutations.”

More suggestions fly, fast and furiously around the table as the team considers other clinical trials. “We have a trial opening in the next few months testing an inhibitor that might work, but colon cancers are not eligible,” Ford sighs. “That’s too bad, because that sounds kind of perfect.”

Eventually, the team decides to refer the patient to the MATCH trial, or Molecular Analysis for Therapy Choice, run by the National Cancer Institute. Similar to Stanford’s tumor board, the MATCH trial — of which Ford is a co-director — seeks to assign patients, many with uncommon cancers or combinations of mutations, to targeted treatment arms based on a molecular analysis of their tumors.

“Some of these genetic changes are extraordinarily rare,” Ford said, “so some of these trials recruit participants from around the country. The first version of MATCH had 36 different arms, each focused on one specific mutation, regardless of tumor type, which was matched to treatment with a particular drug. Now we’re launching a new version of MATCH in which we’ll test the effect of drug combinations against specific mutation combinations. This is the next iteration in genomic medicine. Each tumor has so many genetic changes that it’s likely we are going to need to tackle multiple targets to be successful.”

The next case, presented by Paolo Ocampo, MD, PhD, a clinical fellow in molecular genetic pathology, highlights details from a patient whose tumor has evaded all standard treatments. Unfortunately, the mutations found in the tumor defy the standard methods of subcategorization that often aid treatment decisions for breast cancer. There appear to be two main drivers of tumor growth, but the patient also has a third mutation that often results in resistance to treatments targeting those drivers.

“Most importantly, is she eligible for PUMA?” muses Ford, referring to a trial testing the effectiveness of a drug called neratinib in tumors with similar mutations as the patient’s. “We have a trial that includes these mutations — that would be a great trial for her.”

“She’s just started treatment locally, but she’s coming to see us soon,” says Reyes.

“This is the next iteration in genomic medicine. Each tumor has so many genetic changes that it’s likely we are going to need to tackle multiple targets to be successful.”

Ford, who is also a professor of medicine

Quickly, the board addresses several other cases: breast cancer with brain metastases that must be stabilized for the patient to qualify for enrollment in a clinical trial, a head and neck squamous cell cancer with an interesting gene rearrangement, pancreatic cancer with a mutation that excludes the patient from many clinical trials, and a rectal cancer with a rare mutation. “We need something for ATM mutations,” says Ford in frustration. “I don’t know what to give this guy.”

Finally, they consider the case of a young patient whom many in the room recognize. They’ve discussed her before, but her cancer is growing again. “My guess is that her cancer is progressing because only a portion of her cells are susceptible,” says Ford, of the previous treatment. The team brainstorms new approaches before breaking up for the afternoon.

“The board serves as a way for us all to share what we’re learning about new studies and new drugs,” Reyes said. “Sometimes the answer for a particular patient is obvious, or maybe someone in the room has a tidbit of information the rest of us don’t have.”

A note documenting the board’s discussion is included in each patient’s chart for the referring oncologist.

“Our hope is that this type of approach will be successful not just for metastatic patients who don’t have options, but that what we learn here might also help to bring this personalized oncology to newly diagnosed patients to increase cure rates and decrease the toxicities that often accompany traditional therapies,” Ford said.

“While the immediate objective of the Molecular Tumor Board is to assimilate evidence that informs treatment decisions for individual patients with advanced disease, the data generated through this process also allows us to advance precision medicine more broadly,” said Christina Curtis, PhD, associate professor of medicine and of genetics at Stanford and co-director of the board. 

“In particular, it provides an important source of real-world evidence as to the relationship between the molecular profile of a patient’s tumor and their response to a given treatment. While each patient is unique, as is their cancer genome, we can learn important patterns from that data by aggregating this information across many patients. Ultimately, this — in conjunction with other data — may lead to new, more effective, therapeutic strategies.”

In 2017, Ford and Lincoln Nadauld, MD, PhD, executive director of Intermountain Precision Genomics, studied 72 patients with metastatic cancer: Half had undergone genomic testing followed by targeted treatments based on their cancers’ mutations, and the other half were treated with standard therapies. The study showed that the group who received targeted treatments experienced a longer period before their disease progressed and did not have greater health care costs than the control group.

Ford lingers after the meeting to talk with a colleague. “I really hope we get feedback on these cases,” he says. Feedback is critical to the tumor board’s mission to stay ahead of every patient’s cancer, every time.

“We’re always planning out the long game,” Reyes later said. “We want to know not just what to put the patient on now but what we’re going to put them on next, and then what we could put them on after that. So the instant a patient needs to start a new treatment we have already planned that transition. The whole point of the tumor board is to get the most promising treatment to patients as quickly as possible.”

Author headshot

Krista Conger

Krista Conger is a Senior Science Writer in the Office of Communications. Email her at kristac@stanford.edu.

Email the author