Are basket trials the most appropriate method to capitalize on increased rates of genomic sequencing?
As the molecular landscape of cancer expands through genomic and transcriptomic sequencing, novel cancer-driving mutations have emerged as potential therapeutic targets. However, a key clinical challenge is that, unlike ERBB2 (HER-2) amplification in breast cancer and BRAF mutations in melanoma, the majority of emerging targets for therapy are rare — less than 1% to 10% — in a given cancer type. This “long tail” of molecular drivers in cancer poses a challenge for accruing patients to clinical trials, and it limits access to novel therapies for individual patients with rare cancers or rare mutations.
Basket trials have emerged as a pragmatic approach to prospectively evaluate response to targeted therapies for a rare molecular target. Several ongoing clinical trials illustrate this approach, including:
- The NCI-MATCH trial, which seeks to offer patients with any advanced cancer type an investigational therapy based molecular testing;
- ASCO’s TAPUR Study, which offers FDA–approved targeted therapies to patients with advanced cancer that has potentially actionable genomic variants; and
- Industry- and investigator-initiated trials that focus on specific pathways or markers that have been developed across academic cancer centers.
Despite the available molecular testing and trials, accrual of patients to these trials remains low and, on average, less than 10% of patients receive a matching therapy.
We propose several strategies for improving accrual to these trials. First, expanding data sharing for multihospital networks will enable identification of patients and biomarker discovery. Second, the national cancer moonshot initiative did not address or discuss how molecular testing is paid for and reimbursed. This is a major limitation and cost for investigators and industry. Third, researchers of clinical trials enrolling a rare subset of patients can consider reducing eligibility criteria to allow capture of more patients.
In conclusion, we look forward to offering patients more access to novel therapies through clinical trials.
Sameek Roychowdhury, MD, PhD, is assistant professor of medical oncology at The Ohio State University Comprehensive Cancer Center and Arthur G. James Cancer Hospital. He can be reached at firstname.lastname@example.org. Disclosure: Roychowdhury reports no relevant financial disclosures.
The fundamental principle underlying basket trials — finding a commonality between tumors so patients can be treated in the same manner — is not new and is, ultimately, a poor fit for the genomic complexity and intertumor heterogeneity uncovered by sequencing.
The basket study designation is given to trials that focus on a specific gene mutation and a cognate-targeted therapy, regardless of the tissue of origin of the tumor. These studies are sometimes called “histology-agnostic basket studies.” In contrast, traditional clinical trials test a drug or regimen against a specific cancer type. Although genomic baskets are an important advance from traditional histology-based trials, they are only a first stage in the design innovation mandated by the complicated reality that genomic sequencing has unveiled.
Traditional histology-based trials actually are basket trials (ie, gene-agnostic, histology-based baskets). Although not generally recognized as such, if one examines the scientific rationale for the basket designation — a trial that includes cancers categorized in diverse ways if they have a commonality — one realizes that traditional trials are indeed baskets. In other words, traditional trials test a basket of different genomic alterations but the same histological diagnosis. For instance, a trial of a drug regimen in lung adenocarcinoma includes a variety of different genomic alterations — such as aberrations in EGFR, BRAF and ALK, etc — in the protocol basket, as long as the diagnosis is lung cancer. It is likely that the heterogeneity of patients in such trials has limited the response rates.
Gene-based, histology-agnostic basket trials represent an important step forward in that they recognize the importance of genomic alterations in driving cancer. There are now numerous examples of gene-targeted trials with high response rates, such as those that test EGFR or ALK inhibitors in EGFR– or ALK–mutant lung cancer. The extrapolation to a basket is based on the assumption and clinical experience that many different tumor types — albeit, not all — with the same oncogenic driver will respond to the cognate inhibitor.
Because specific genomic alterations in cancers are rare, it is impractical to enroll patients in the trial and then test them for a single gene that is the focus of the basket study. Genomics must be a front-door diagnostic test and use a multigene panel so the patients can be navigated to any one of multiple basket trials or available drugs.
Genomic basket studies are an important interim step in the evolution of the precision medicine paradigm. However, a wealth of data suggests that metastatic tumors have multiple genomic alterations and are generally distinct from each other.
Therefore, treatment in trials grounded in finding a single commonality between patients is suboptimal. Next-generation trials will require individualized combination therapy and test the strategy of matching, rather than trying to fit cancers with diverse and complex molecular portfolios onto the same single-agent basket.
Razelle Kurzrock, MD, is chief of hematology and oncology, deputy director of clinical science, and director of personalized cancer therapy and clinical trials at University of California, San Diego. She can be reached at email@example.com. Disclosure: Kurzrock reports consultant fees from Actuate Therapeutics and X-biotech, and research funds from Foundation Medicine, Genentech, Guardant Health, Merck Serono, Pfizer and Sequenom. She also reports ownership interest in CureMatch Inc.