By Deb Borfitz
Aug 4, 2009 | When the clinical feasibility of molecular targeted agents (MTAs) is being tested, “faster answers” will require better use of creative trial designs and approaches, says Timothy Yeatman, president and CSO of M2Gen.
Most MTAs developed to date “don’t really target the right population” and thus patient response can be unpredictable. Clinical improvement also tends to be “incremental” when the MTA is taken in combination with an existing, fairly effective standard-of-care drug, as is typically required in phase II–III oncology trials.
M2Gen plans to improve the situation. The company supports the partnership between H. Lee Moffitt Cancer Center & Research Institute and drug maker Merck in a personalized medicine initiative coined “Total Cancer Care.” (see, “Merck-Moffitt Partnership Breaks Down Silos”) To date, the for-profit, wholly-owned subsidiary of Moffitt Cancer Center has collected tumor tissues from more than 36,000 patients in an effort to identify the biological markers unique to each tumor.
In the coming months, the Total Cancer Care Database will allow M2Gen investigators to identify patients with the desirable phenotype before a trial accrues, rather than testing each patient for a possible biomarker once a trial is underway. Doing so, Yeatman says, “has the potential to halve the time it takes to accrue [trial participants] and double the response rate of [tested] drugs.”
A similar database, known as the International Genomics Consortium, is headquartered at the non-profit Translational Genomics Research Institute (TGen) in Phoenix. Behind TGen’s drive to find the right drug for the right patient is the dismal 10–20% success rate for oncology treatment regimes based on tumor type, says Ramesh Ramanathan, medical director of TGen Clinical Research Services at Scottsdale Healthcare.
“Analyzing the huge amounts of data generated from molecular and genomic profiling of tumors and making it fit for the individual patient is critical,” says Ramanathan, and moreover, “requires close collaboration with experienced bioinformatics scientists.”
Even in phase I studies, Ramanathan says it is increasingly important to more selectively screen would-be study participants. Typically, therapies get tested across cancer types. Genentech veered from this norm by intentionally enrolling patients with advanced basal cell cancer in phase I trials of its targeted therapy GDC0449. The rationale is that 90% or more of these patients harbor aberrant Hedgehog genes the agent is known to inhibit.
Abandoning traditional trial design rules can also help make studies of MTAs less costly to conduct and more palatable to participants, says Yeatman. In phase I studies, for instance, it may make more sense to pinpoint the molecular drug effect (the so-called “target effect”) via pharmacodynamic assays than quantify the maximum tolerable dose.
Investigators might also want to measure surrogates for biological activity rather than escalate dosage to the point of toxicity. Many MTAs do not cause major side effects even at huge doses, “so we don’t learn anything” by escalating dosage, says Ramanathan. The important question is whether or not the drug gets into a cancer cell and affects its function. That means greater reliance on MRI scans, specialized PET scans, biopsies, and blood tests to see if proteins are being secreted by the tumor.
During a typical phase II, single-arm study, “response rate” may be an inadequate endpoint because the MTA—however efficacious it proves to be—isn’t getting compared head-to-head with the standard of care, says Yeatman. A randomized phase II trial would take a lot longer but provide better data. It’s not necessarily a worthwhile tradeoff. The compromise is to start with a single-arm study, “to see if you’re even in the ballpark” and, if so, follow up with a randomized phase II trial against the standard of care.
Ramanathan notes that MTAs often don’t shrink a cancer but rather make it inactive—sometimes for a long time. “You may have an effective agent and not realize it.” For this reason, older chemotherapy drugs are now being eyed as potential targeted agents.
It’s probably best to avoid measuring clinical effects directly related to the level of target inhibition, says Yeatman. “A cancer may have multiple molecular pathways that a tumor cell can use to escape a particular drug.” Unless the cancer happens to be addicted to a particular pathway or is a genetic mutation, a highly targeted MTA may actually be counter-productive.