The concept of “basket trials” is gaining traction as a strategy for studying cancers according to driver mutations rather than by tumor type. These studies are made possible by the dramatically reduced cost of performing next-generation sequencing platforms that characterize the landscape of individual cancer genomes across a wide variety of cancer types.
“The organizing principle of basket trials is that patients have a qualifying genomic alteration regardless of tumor site. Any type of tumor can be included. The trials entail an independent analysis by tumor lineage targeted to low-prevalence genomic alterations [ie, <5% in frequency]. Once a genomic alteration has a higher prevalence, it is advisable to perform a study on a single tumor type,” explained David Hyman, MD, Memorial Sloan Kettering Cancer Center (MSKCC), New York City.
“Most of these studies are encouraged to have other cohorts for serendipitous disease discovery,” he added. Therapy for all cohorts is uniform.
One example of a basket trial being conducted at MSKCC is that of tumors with ERBB2 mutations treated with neratinib. This study has 6 “baskets” – bladder, colon, endometrial, gastric, ovarian, and “other” cancers.
“We use statistical analysis that sets high bars and allows us to enroll small numbers of patients, but we may be missing active agents by setting the bar this high,” Hyman said.
At the National Cancer Institute (NCI), “match” trials are also under way, he continued. Nine new match trials have been announced. Among these are: non-BRAF V600 and BRAF fusion; ALK/ROS1 translocation (non-lung); TSC1/2 loss of function mutations; EGFR2 T790M mutations (non-lung); and 4 others.
At the NCI, the approach is as follows: an actionable mutation is detected by genetic sequencing, and the tumor is then treated with the study agent. Patients with stable disease or complete or partial responses continue on the study drug until disease progression. Those with no response stop the drug. All patients who progress are biopsied again and sequenced and put on a new study drug if a new actionable mutation is detected.
Gynecologic cancers require special considerations for basket trials, Hyman continued. An assay is needed to detect copy number alterations. There may be financial barriers to study participation. Also, it is not clear if the primary end point should be RECIST measurable disease or detectable CA125 antigen.
“This would be ideal, but it may not be applicable,” Hyman said.
“You have to consider comorbidities associated with gynecologic cancers, and how these comorbidities might interact with cancer drugs. For example, many endometrial cancer patients have diabetes, and this would affect interaction with PI3KCa inhibitors,” Hyman explained.
MSKCC investigators use an integrated mutation profiling approach to these basket trials. Germline and mutation DNA are collected, followed by genetic profiling done by MSK-IMPACT, a new testing approach that screens for mutations in over 300 genes at once, in a CLIA lab. Results are entered into the electronic medical record regarding eligibility for a basket trial protocol.
“As we work to develop these programs, it is critically important to have informatics so we can identify patients who are eligible for these trials,” he stated.
