In its first post-pilot cancer analysis, the Cancer Genome Atlas Research Network has published a detailed genomic picture of colorectal cancer that suggests several potential new genomically-guided treatment strategies for the disease.
The group published its "comprehensive molecular characterization of human colon and rectal cancer" in Nature today, highlighting a number of frequent mutations that the researchers believe are targetable with either currently available drugs or treatments in mid- and late-stage development. Several of these mutations have not been identified in colorectal cancer before, while others have been suggested in previous studies, but their potential may be bolstered by the TCGA results, Raju Kucherlapati, the principle investigator told PGx Reporter this week.
The researchers found two genes, ERBB2 and IGF2, were amplified in a significant proportion of subjects analyzed, suggesting that about five percent of patients might be candidates for treatment with either Roche/Genentech's Herceptin, in the case of ERBB2; or with the 20 percent for IGF2-targeted therapies currently in development, Kucherlapati said. In the latter category, several companies, including Bristol-Myers Squibb, are studying small molecule IGF inhibitors. Others, like Amgen, are developing antibody agonists to IGF and exploring other inhibitory strategies.
On a larger scale, the study also found that almost all of the tumors analyzed showed deregulation of a single pathway, MYC. "It was known that MYC was amplified in some [colorectal] tumors, about 10 to 15 percent. But our result shows 95 percent have MYC deregulation," Kucherlapati said.
"This means drugs that are capable of inhibiting MYC might be very effective in these tumors, and there are some … particularly bromodomain inhibitors currently in development to inhibit MYC action, he added.
Overall, the results exceeded the group's expectations. "This is a cancer that has been studied extensively — so we didn't know when we started whether we would find significant new observations," Kucherlapati said. "I'm particularly pleased by the fact that a cancer type for which there are limited therapeutic options now opens up a whole plethora."
He added that the results bode well for a transformation of colorectal cancer treatment within a few years after further confirmatory studies.
For the study, investigators analyzed 276 samples using a variety of genomic techniques from genome sequencing to miRNA profiling and found that colon and rectal cancers are much more alike than not. The researchers identified 32 recurrent somatic mutations in the cohort. Excluding hypermutated cancers, overall patterns of mutation and gene expression were virtually "indistinguishable between colon and rectal carcinomas," the researchers reported.
The group highlighted MYC, RAS, and PI3K pathways, as well as IGF2, ERBB2, ERBB3, AKT, and MTOR as potential drug targets.
Amplification of ERBB2 showed up in about five percent of the tumor samples the group studied.
"Certainly these five percent could be tested with either Herceptin or other drugs in development for that target," Kucherlapati said.
The researchers also found that more than 20 percent of the tumors had either amplification or overexpression at the RNA level of a gene called IGF2, which Kucherlapati said was a more novel finding.
Though not a new finding, the researchers confirmed the role of BRAF mutations in colorectal cancers, finding about 15 percent of their cohort had mutations in BRAF. Kucherlapati said that BRAF-targeting drugs, like Roche's Zelboraf, have not performed well in colon caner, but he added that recent studies have suggested a combination strategy to inhibit both BRAF and EGFR simultaneously might do better.
"From a pharmacogenetics point of view," he said, "testing patients for BRAF mutations and [enrolling] them into combination drug trials would be a promising possibility because both [PGx targets] have approved [treatments on the market]."
On a larger scale, the study also assessed pathway activity, confirming the known implication of both the RAS and PI3K pathways in colorectal cancer. Importantly, though, he said, the TCGA study revealed that many tumors have both these pathways simultaneously deregulated, meaning any one drug affecting only one of the pathways is unlikely to be effective.
"People really need to be thinking about combination therapy for those patients," he said.
The multi-platform approach also revealed deregulated activity of MYC across virtually all the tumors the group profiled, Kucherlapati noted. Inhibitors of MYC activity are showing initial promise, he said, and could potentially offer a strong treatment option.
Finally, Kucherlapati said the group's results also suggested that current approaches measuring microsatellite instability as a proxy for hypermutated colorectal cancer — which is associated with lower response to chemotherapy but higher overall survival — may not be accurate.
Though these tests are not widespread, Kucherlapati said the group's results suggest better methods are necessary, especially if testing for hypermutation becomes more widely used.
"One of the things this paper has shown is that there are a bunch of tumors that have high mutation frequency but don't have this microsatellite instability … So one of the suggestions of the [paper] is that microsatellite instability is not necessarily a good surrogate for high mutation frequency, and we should consider different ways of measuring mutation frequency in these tumors because it has significant implications for outcomes and treatment for these patients," he said.
Even though some of the mutations or amplifications the group found were relatively less prevalent — only about five percent for ERBB2 for example — Kucherlapati said the therapeutic implications are still very promising.
"I don't think the numbers matter," he said. "Especially if there is already a drug on the shelf, even if a small number of patients respond well, this could be a very effective treatment for that group."
"Right now for colon and rectal cancer, after surgery, the only therapies are chemotherapeutics and the two approved targeted drugs Avastin (Genentech) and Erbitux (Merck)," he said.
"Our results showed that there are potentially many more targets that can be utilized, and luckily drugs are already in late stage development for many of these pathways."
The colorectal cancer analysis is the third the TCGA has published, following ovarian cancer last year (PGx Reporter 6/29/2011), and marks the departure from the project's pilot-phase. Kucherlapati said the team expects to next release its analyses of lung cancer and breast cancer.