The Basket Trial: An Evolving Clinical Trial Design

Get Permission

Amanda J. Redig, MD, PhD

Pasi A. Jänne, MD, PhD

The molecular alterations that define a tumor may be as relevant, if not more so, than the histopathology that would historically have been used to classify a tumor and direct therapeutic options.

—Amanda J. Redig, MD, PhD (top), and Pasi A. Jänne, MD, PhD (bottom)

The recently published results of the CUSTOM (Molecular Profiling and Targeted Therapies in Advanced Thoracic Malignancies) trial, reviewed in this issue of The ASCO Post, describe a basket trial focused on identifying molecular biomarkers in advanced non­­–small cell lung cancer (NSCLC), small cell lung cancer, and thymic malignancies, followed by evaluation of five targeted therapies in patients grouped by molecular markers and tumor histology.1,2 The five targeted therapies included erlotinib (Tarceva; EGFR mutations); the MEK inhibitor selumetinib (KRAS, HRAS, NRAS, or BRAF mutations); MK2206, an AKT inhibitor (PIK3CA, AKT1, PTEN mutations); lapatinib (Tykerb; HER2 mutations); and sunitinib (Sutent; KIT, PDGFRA mutations).

The initial study goal was to evaluate each treatment in each of 3 histology subtypes, for a total of 15 study arms. However, as the authors noted in the final publication, they were only able to complete accrual for 2 of the 15 available arms (erlotinib in NSCLC and selumetinib in NSCLC).

Basket trials are an emerging form of clinical trial design and represent a novel way of combining traditional clinical trial methodology with emerging data from the genomic era of medicine. Historically, oncology patients have been treated with a primary focus on a tumor’s tissue of origin and histopathology: This categorization is reflected in the clinical classification of modern oncology, with separate departments, journals, and academic meetings focused on the biology and treatment strategies for a group of cancers defined by anatomic location. However, advances in both technology and molecular biology have begun to challenge the elements of this paradigm.

From Histopathology to Molecular Alterations

Some 15 years after the first reports of groundbreaking efficacy in matching a targeted therapy to a defining molecular alteration in the context of the bcr-abl driver of chronic myeloid leukemia,3 the number of targetable molecular alterations and targeted therapies has expanded dramatically. In this setting, the molecular alterations that define a tumor may be as relevant, if not more so, than the histopathology that would historically have been used to classify a tumor and direct therapeutic options. In some cases, two tumors from different anatomic locations may share more in common with each other—based upon genetic vulnerabilities to targeted therapies—than either tumor does to paired tumors from the same tissue of origin but lacking the shared driver mutation.

As next-generation sequencing has become more and more integrated into research practices and clinical decision-making within oncology—particularly within thoracic oncology, as seen in CUSTOM—a logical next step in the emerging clinical trial design is to pair targeted therapies with a targetable mutation, irrespective of cancer diagnosis. In CUSTOM, targeted therapies and actionable mutations were identified prospectively, but so-called basket trials may also function in reverse, with initial use of a targeted therapy in an unselected population and use of next-generation sequencing in patients who respond to identify putative genetic biomarkers for further prospective analysis. Perhaps most central, a basket trial approach represents a hypothesis-driven mechanism of combining precision medicine with clinical trial design, a strategy that may be of particular relevance for mutations that are not uncommon within cancer globally but may be prohibitively rare for study within a single disease center.

In recognition of the value of a basket trial approach, in 2015, the National Cancer Institute will launch the NCI-MATCH trial (Molecular Analysis for Therapy Choice), with plans to screen up to 3,000 patients with enrollment of at least 1,000 individuals to a targeted drug combination, independent of tumor histology.4 Similarly the NCI-IMPACT trial (Molecular Profiling-Based Assignment of Cancer Therapy) will randomize patients with a known mutation in a specific genetic pathway to either pathway-driven targeted therapies or a treatment not known to be pathway-specific ( identifier NCT01827384).

Strengths and Challenges

Results of the CUSTOM trial reveal key strengths of the basket trial design as well as some challenges. It is notable that a basket trial strategy can be used not only to validate a clinical target but also to identify a favorable response to targeted therapy in a small number of patients. However, at the same time, CUSTOM was unable to complete accrual to 13 of 15 planned study arms, and one of the arms that did complete accrual (selumetinib in NSCLC) did not achieve its primary endpoint.

As with any clinical trial approach, rigorous evaluation of preclinical data when designing a basket study will continue to be of central importance. Notably, the one arm of CUSTOM that did complete accrual and reach its primary endpoint was using a therapeutic agent (erlotinib) for which clear evidence of clinical efficacy (in EGFR-mutant NSCLC) was already apparent. In addition, some studies suggest that inhibition of MAPK signaling in tumors with KRAS mutations may depend upon the genomic landscape of the tumor, specifically the presence or absence of key tumor suppressors.5 Furthermore, previously published studies demonstrate efficacy for combined MEK inhibition and chemotherapy, not the MEK inhibition as monotherapy used in CUSTOM.6

Moving forward, basket trials will continue to provide a valuable way to link the emerging genomic landscape of precision medicine with the emerging toolbox of targeted therapeutic strategies. As next-generation sequencing technology and drug development strategies continue to develop, basket trials will likely become more nuanced, with randomization to trial arms driven by the presence or absence of modifying mutations in addition to a known driver mutation. However, established standards of clinical trial design will continue to apply in the genomic era. As with any other approach to translational medicine, the success of the basket trial strategy is linked to both scientific preclinical development and rigorous trial design. ■

Disclosure: Dr. Redig reported no potential conflicts of interest.


1. Lopez-Chavez A, Thomas A, Rajan A, et al: Molecular profiling and targeted therapy for advanced thoracic malignancies: A biomarker-derived, multiarm, multihistology phase II basket trial. J Clin Oncol 33:1000-1007, 2015.

2. Redig AJ, Jänne PA: Basket trials and the evolution of clinical trial design in an era of genomic medicine. J Clin Oncol 33:975-977, 2015.

3. Druker BJ, Sawyers CL, Kantarjian H, et al: Activity of a specific inhibitor of the BCR-ABL tyrosine kinase in the blast crisis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N Engl J Med 344:1038-1042, 2001.

4. National Cancer Institute: Molecular analysis for therapy choice (NCI-MATCH). Available at Accessed April 13, 2015.

5. Chen Z, Cheng K, Walton Z, et al: A murine lung cancer co-clinical trial identifies genetic modifiers
of therapeutic response. Nature 483:613-617, 2012.

6. Jänne PA, Shaw AT, Pereira JR, et al: Selumetinib plus docetaxel for KRAS-mutant advanced non-small-cell lung cancer: A randomised, multicentre, placebo-controlled, phase 2 study. Lancet Oncol 14:38-47, 2013.

Related Articles

Outcomes of the CUSTOM ‘Basket’ Trial of Molecular Profiling and Targeted Therapy in Advanced Thoracic Malignancies

In the phase II CUSTOM trial reported in the Journal of Clinical Oncology, Ariel Lopez-Chavez, MD, Anish Thomas, MD, and colleagues performed molecular profiling of tumors in patients with advanced non–small cell lung cancer (NSCLC), small cell lung cancer (SCLC), or thymic malignancies and...