Why oncology drug development will remain tissue-dependent, for now
The start of two large-scale trials enrolling patients whose tumors share a common molecular signature regardless of the tissue of origin reflects tremendous progress in both the understanding of cancer biology and the availability of effective targeted therapies. But oncology drug development is still very far away from breaking free from the ingrained system of classifying tumors according to the tissue of origin.
The reasons are many, including the difficulty of conducting and interpreting results from studies enrolling patients with different tumor types for which the standard of care is different, and the need to demonstrate a consistent effect across a sufficient number of tissue types that would enable regulators to conclude there is a positive benefit-risk profile in an indication defined by molecular signature alone.
Developing or approving drugs based on molecular or genetic signatures is further complicated by the fact that driver mutations are not always present in all tumor cells and may change as tumors evolve, and the likelihood that most cancers will require combination therapies.
In addition, data from some studies suggest that differences in the molecular context or microenvironments of tumors that share a driver mutation but originate in different organs affect whether and how they respond to targeted therapy.
Two studies launched at this year's American Society of Clinical Oncology meeting in Chicago may provide some of the information drug developers and regulators would need to move closer to a molecular-based pathway for cancer drug development.
Both ASCO and NCI announced the launch of so-called "basket trials" that aim to determine how targeted cancer therapies perform in a broad range of cancers that have the genetic abnormality the products target. These exploratory trials will look for efficacy signals across cancer types for which the products are not yet approved.
Cancer companies that spoke to BioCentury said basket trials such as these can be an efficient way of screening for activity of targeted agents in a variety of tissue-defined cancers.
They may also provide insights into how often and why driver mutations are sensitive to molecular context, the mechanisms behind exceptional responses and how oncologists interpret and use patients' genomic information.
However, the trials themselves can only hint at new uses for the therapies tested because the studies are unblinded, without a comparator, in patients