The study, published last week in the New England Journal of Medicine, demonstrates that both tumor histology and molecular markers are indispensible factors when crafting precision treatment strategies. Looking specifically at patients with a range of cancers with BRAF V600 mutations, researchers led by MSK's Jose Baselga and David Hyman found that certain tumor types — like lung cancer or Langerhans cell histiocytosis — but not all appear to respond to the BRAF inhibitor Zelboraf (vemurafenib).
In the era of biomarkers and precision medicine, some proponents have predicted a future where patients will receive therapy based on the molecular markers driving their disease — in this case BRAF mutations — instead of the type of tissue or cells their tumors are rooted in (i.e. lungs, colon, etc.) "A lot of people have jumped to that conclusion," said Hyman, acting director of developmental therapeutics at MSK and the first author of the published study.
The work by Hyman's group suggests that tissue characteristics will likely remain an important consideration for personalizing cancer treatments. "A lot of this pertains to the off-label use of approved targeted therapies," he told GenomeWeb. "This is increasingly becoming an issue as we get a wider compendium of targeted therapies for specific indications, which are then commercially available for use in an off-label fashion."
Add to this the fact that more and more people are getting genomic testing through academic centers and a variety of commercial labs. "The issue is how are we going to use that data?" Hyman reflected. "Certainly, we are very enthusiastic about a precision medicine approach, but one of the things that gets lost in extrapolating data in a off-label setting is the relevance of the tumor histology. That clearly has importance."
Meanwhile, increasing understanding of the molecular underpinnings of cancer has disrupted neat histological groupings. Lung cancer is no longer just one disease. A lung tumor can have, for example, ALK rearrangements, EGFR mutations, over-expression of PD-L1, or none of these features. While these tumor features must be attacked using different treatment strategies, the small numbers of patients in these molecularly defined groups make it impossible to conduct traditional randomized-controlled trials.
Enter basket studies. Hyman defined a basket trial as one that enrolls patients regardless of tumor type, based on the presence of a genetic marker, but "with some effort to evaluate the effectiveness [of treatment] on a tumor-specific level." This offers drugmakers a way to quickly detect "early signals" of drug activity across tumor types, and also enable research into how tumor lineage influences response, the researchers wrote in NEJM.
The aim of this study was to gauge whether cancers, other than melanoma, that are characterized by BRAF V600 mutations respond to Roche/Genentech's BRAF inhibitor Zelboraf. The trial steering committee designed the study with a team from Roche, which also provided funding.
Zelboraf is already approved as a treatment for metastatic melanoma patients with BRAF V600E mutations. Approximately half of melanoma patients have BRAF V600 mutations and studies have shown that half of those patients respond to the BRAF inhibitor and many live longer than those without this marker.
BRAF V600 mutations also show up in other cancers, although at much lower rates, making it challenging to explore the efficacy and safety of Zelboraf in other tumor indications. This is precisely what Baselga's team wanted to investigate in their Phase II basket study.
From 2012 to 2014, researchers enrolled 122 cancer patients with BRAF V600 mutations into several tumor-specific cohorts, and treated them with Zelboraf. The 23 centers where enrollment took place could establish patients' BRAF mutation status by whatever method they chose. Hyman noted that a small number of patients had the Roche Cobas 4800 BRAF V600 mutation test that the US Food and Drug Administration approved to identify melanoma patients who should receive Zelboraf.
The "baskets" included patients with non-small-cell lung cancer, cholangiocarcinoma, Erdheim-Chester disease, Langerhans cell histiocytosis, anaplastic thyroid cancer, multiple myeloma, and other tumor types. Previous case studies had offered hints that a BRAF inhibitor might work well in some, but not all of these cancers.
Based on early signals in the study that colorectal cancer patients weren't responding well to Zelboraf, researchers amended the protocol and decided to give the drug in combination with the anti-EGFR therapy Erbitux (cetuximab). Hyman's team wasn't taking a shot in the dark in deciding to go with this combo, but based it on recent preclinical evidence suggesting that adding an EGFR inhibitor might help colorectal cancer patients overcome resistance to Zelboraf monotherapy.
This demonstrated the speed and flexibility of the basket approach. "This is an example of a very short bench-to-bedside transition," said Hyman, noting that there was a six-month gap between the time that the preclinical studies were published and the basket study was amended for colorectal cancer patients.
Although none of the colorectal cancer patients on Zelboraf monotherapy responded, one patient who received Zelboraf and Erbitux responded and half the patients had some tumor shrinkage.
The researchers noted that the response in this subgroup may have been impacted by the fact that a large proportion of colorectal cancer patients had received treatment with an anti-EGFR antibody. "Given the highly aggressive and chemotherapy-resistant nature of BRAF V600-mutated colorectal cancers, strategies using dual EGFR and BRAF inhibition deserve further evaluation," they wrote in the paper.
NSCLC was one of the biggest "baskets" with 20 patients. The response rate in this cohort was 42 percent, median progression-free survival was 7.3 months, and 66 percent were alive after a year. The median overall survival had not been reached in this cohort at the time of analysis. "This rate compares favorably with the 7 percent response rate reported for standard second-line docetaxel in molecularly unselected patients," Baselga and colleagues wrote in the paper.
Among 14 evaluable patients with Erdheim-Chester disease or Langerhans cell histiocytosis — characterized by the overproduction of a kind of infection-fighting white blood cell — 43 percent responded. The median duration of treatment in this group was around six months, during which time no one progressed.
Adults with these types of rare cancers don't have any approved treatment options. Baselga's group pointed out in the paper that these findings are in line with another recent case series that suggested that a BRAF inhibitor might have activity against Erdheim-Chester disease.
Researchers also reported "anecdotal responses" from patients with pleomorphic xanthoastrocytoma (a type of brain neoplasm), anaplastic thyroid cancer, cholangiocarcinoma (cancer in the bile ducts), salivary-duct cancer, ovarian cancer, and clear-cell sarcoma, and among patients with colorectal cancer who received the Zelboraf/Erbitux combination. Some of these patients treated with Zelboraf experienced tumor shrinkage but not enough to qualify as a response according to measurement criteria.
Three patients with anaplastic thyroid cancer, cholangiocarcinoma, and ovarian cancer had responses that lasted for more than a year. None of the multiple myeloma patients has responded to treatment.
In general, the safety profile of the drug in this trial was similar to studies of Zelboraf in melanoma patients. Around 20 percent of patients experienced common adverse events associated with Zelboraf, which were rash, fatigue, and joint pain.
Options in orphan diseases
"This type of study is great for evaluating rare tumor populations," Hyman said. "It's very clear that the common cancer types are overrepresented in the studies we conduct. You may not be able to get a company interested in doing research in Erdheim–Chester disease." But by doing a basket trial, a drug firm can offer patients a chance to benefit from an investigational treatment.
Since Erdheim-Chester disease was first described in 1930, only a few hundred cases have been reported in the literature. "There's never been a prospective clinical trial in Erdheim-Chester disease," Hyman noted. "One of the things I'm proud of about this study, and I hope we'll start to see more of, is when you look at the tumor types [included] there is a significant representation of very orphan diseases."
However, because few patients with these rare tumor types were enrolled, Baselga and colleagues said their findings should be interpreted with caution. "In the absence of more definitive data, which might not be forthcoming for many diseases owing to logistical impediments, these data present a challenge to clinicians who want to make treatment decisions on the basis of tumor genomic profiling," they wrote in the paper.
Meanwhile, the present basket trial is still enrolling patients. "We've enrolled significantly larger numbers of patients than reported in this manuscript," Hyman said, noting future publications will more definitively describe patients' experiences in larger "baskets."
"We'll see certain themes emerge and this type of study can help tease that out," he added. Pharmaceutical firms can then run with the "themes" or signals, moving quickly into registration studies for a drug in a new tumor type after observing efficacy in a particular "basket" of patients.
Sponsors could, for example, plan from the beginning that after a basket yields a signal, study sites could continue to enroll that cohort until there is more certainty about the drug's efficacy and safety in the subpopulation. "Then, that data could [support market] registration for the companies," Hyman said.
Within pharma and at cancer centers, basket studies are catching on. MSK is currently running between five and 10 basket studies and conducting several in collaboration with industry, Hyman said. "We're opening almost one a month at this point," he said. "It's become very attractive to the [drug] companies."
Ahead of press time, Genentech couldn't make available an expert to discuss how it will proceed based on data from this basket study. However, a company spokesperson said, "Genentech is evaluating the study results to inform future research with the goal of addressing the unmet need of people with rare cancers."
Basket trials are smaller, but they identify cancer patients that derive better-than-average responses to treatments. As a result, companies are investing in basket studies earlier in the drug development process "to search for efficacy over a wider range of tumors," Hyman reflected. "This is the way they want to be developing drugs. They want to set a high bar and there is less and less interest in doing very large studies, looking for incremental benefit to patients."