[A]lthough it was feasible to enroll a large number of patients and perform molecular profiling analyses at a high success rate in an innovative basket trial, the CUSTOM design seems to be unfeasible in its current form given the rarity of the selected genetic abnormalities in the populations under study.
—Ariel Lopez-Chavez, MD (top), Anish Thomas, MD (bottom), and colleagues
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 assigned those in any of five actionable mutation groups to matched targeted therapy.1 Enrollment and molecular profiling were rapid, but sufficient numbers of evaluable patients could not be enrolled into most treatment groups due to the low frequency of target mutations.
CUSTOM was designed as a “basket” trial—a novel type of investigation based on the hypothesis that a molecular marker can predict response to a targeted therapy independent of tumor histology. The study may be the first completed basket clinical trial to evaluate targeted agents against specific molecular aberrations simultaneously across different histologic subtypes of disease.
Giuseppe Giaccone, MD, PhD, of Georgetown University and National Cancer Institute, is the corresponding author for the Journal of Clinical Oncology article. Dr. Lopez-Chavez, of Oregon Health and Science University, and Dr. Thomas, of the National Cancer Institute, contributed equally to the article.
In the study, 647 patients with advanced thoracic malignancy underwent molecular profiling for oncogenic drivers. The population consisted of 481 patients with NSCLC (74%), 68 (10.5%) with SCLC, and 98 (15.2%) with thymic malignancy. Enrollment required only 20 months (February 2011 to December 2012). Molecular profiling was successful in 569 (88%).
Patients were assigned to standard-of-care therapies or one of five biomarker-matched treatment groups: erlotinib for EGFR mutations; the MEK inhibitor selumetinib for KRAS, NRAS, HRAS, or BRAF mutations; the AKT inhibitor MK2206 for PIK3CA, AKT, or PTEN mutations; lapatinib (Tykerb) for ERBB2 mutation or amplification; and sunitinib for KIT or PDGFRA mutation or amplification. It had been planned to generate up to 15 treatment groups by dividing the 5 biomarker-matched groups according to tumor type (NSCLC, SCLC, thymic malignancy). The primary endpoint of each treatment group was a response rate of 40%.
Among those with successful profiling, the prespecified mutations were found in 257, with 23 having multiple abnormalities; 313 were wild-type or unknown for the mutations of interest. Of the patients with mutations, 212 were considered screen failures and not enrolled in targeted treatment groups. Previous use of erlotinib and the early closure of the selumetinib arm accounted for 68% of all screen failures. A total of 602 patients received standard-of-care therapy or were enrolled in other clinical trials and were followed until death.
Frequency of Mutations and Response to Treatment
Completion of accrual to 13 of the proposed 15 treatment groups was considered unfeasible due to low frequencies of target mutations overall or among SCLC and thymic malignancy subgroups.
EGFR mutations were detected in 88 (22.1%) of 398 patients with NSCLC, 1 (2%) of 51 patients with SCLC, and 1 (1.1%) of 92 patients with thymic malignancy. A total of 16 evaluable patients were allocated to erlotinib treatment, consisting of 15 with NSCLC and 1 with thymic malignancy. Response was observed in 9 (60%) of 15 with NSCLC, with no response seen in the patient with thymic malignancy.
KRAS mutations were detected in 91 (24.9%) of 366 patients with NSCLC and 2 (4.1%) of 49 with SCLC. In nine evaluable patients with NSCLC, response was observed in one patient (11%). No response was seen in one evaluable patient with SCLC.
ERBB2 mutation was detected in 8 (2.8%) of 284 patients with NSCLC and none of 40 patients with SCLC or 85 with thymic malignancy. Of seven evaluable patients who received lapatinib, consisting of six with NSCLC and one with SCLC, none had objective response.
PIK3CA mutations were found in 11 (3.9%) of 285 patients with NSCLC, 4 (8.5%) of 47 with SCLC, and 2 (2.4%) of 85 patients with thymic malignancy. PIK3CA amplification was found in 2 (11.1%) of 18 patients with NSCLC. Mutations in AKT1 were found in 1 (0.4%) of 283 NSCLC patients and 1 (2.2%) of 45 with SCLC. PTEN mutations were found in 8 (4.4%) of 181 patients with NSCLC and 2 (9.5%) of 21 with SCLC. Among seven evaluable patients, consisting of four with NSCLC, two with SCLC, and one with thymic malignancy, no responses to MK2206 were observed.
KIT mutations were found in none of 269 patients with NSCLC, 1 (2.6%) of 38 with SCLC, and 4 (4.7%) of 85 patients with thymic malignancy. PDGFRA mutations were found in 1 (1.2%) of 85 patients with thymic malignancy and in none of 103 NSCLC patients or 23 SCLC patients. PDGFRA amplifications were found in 5 (12.8%) of 39 NSCLC patients and in none of 3 SCLC patients or 7 thymic malignancy patients. Among three evaluable patients, consisting of two with NSCLC and one with thymic malignancy, no responses to sunitinib were observed.
Among other genetic abnormalities, rearrangements in ALK were found in 29 (8.7%) of 335 patients with NSCLC and in none of 19 SCLC patients or 86 with thymic malignancy.
Survival in NSCLC by Molecular Group
Survival differences according to molecularly defined subgroups were observed among patients with NSCLC: patients with EGFR mutations had a median survival of 3.51 years, followed by 2.94 years in those with ALK rearrangements, 2.3 years in those with KRAS mutations, 2.17 years in those with other genetic abnormalities, and 1.85 years in those with no actionable mutation.
The investigators concluded: “This basket trial design was not feasible for many of the arms with rare mutations, but it allowed the study of the genetics of less common malignancies.”
They noted: “[A]lthough it was feasible to enroll a large number of patients and perform molecular profiling analyses at a high success rate in an innovative basket trial, the CUSTOM design seems to be unfeasible in its current form given the rarity of the selected genetic abnormalities in the populations under study. New basket trial designs should consider including a larger number of institutions and an adaptive design to successfully conduct such studies.” ■
Disclosure: The study was supported by the Cancer Therapy Evaluation Program at the National Cancer Institute under a collaborative research and development agreement with the study drug manufacturers AstraZeneca (selumetinib), Genentech/OSI Pharmaceuticals (erlotinib), GlaxoSmithKline (lapatinib), and Merck (MK2206). Funding for the companion research study Personalized Cancer Medicine Registry was provided by Novartis. For full disclosures of the study authors, visit jco.ascopubs.org.
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.
Amanda J. Redig, MD, PhD, of Dana-Farber Cancer Institute, and Pasi A. Jänne, MD, PhD, of Harvard Medical School and Dana-Farber Cancer Institute, discuss basket trials and the evolution of clinical trial design.
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...