Collaborative data sharing: The International Acute Myeloid Leukemia Consortium (INTERACT).

Abstract

169 Background: Outcomes for children and adolescents/young adults (AYAs) with acute myeloid leukemia (AML) remain suboptimal and are decidedly poor for patients with subtypes defined by high-risk features, including canonical genetic alterations and relapsed/refractory presentation. Due to their rarity, these subtypes are difficult to study within the existing clinical trials paradigm. Data sharing efforts aim to increase our understanding of rare cancer subtypes by pooling data from multiple sources. The INTERnational Acute myeloid leukemia ConsorTium (INTERACT) was established to pool and share data on pediatric and AYA patients with AML. Methods: In 2019, a data dictionary was created through an iterative, consensus-driven process that analyzed pre-existing clinical trials case report forms as the initial basis. By developing this standardized representation for data on patients with AML, data could be harmonized and integrated. A memorandum of understanding was signed in 2021, formally establishing INTERACT. An executive committee was convened to establish scientific priorities, facilitate ongoing data dictionary work, usher agreements with data contributors and review data requests submitted by investigators. In January 2024, the initial set of AML data was released on the Pediatric Cancer Data Commons (PCDC) data portal. Results: As of April 2024, the data dictionary includes 157 standardized data elements that were used to harmonize clinical data on 3,413 patients who participated in seven clinical trials conducted by four pediatric oncology cooperative groups across the United States, Europe, and Asia. Elements in the data dictionary include demographics, initial disease characteristics, genetic alterations, and survival status. Aggregate data can be freely explored using the publicly-available PCDC data portal (portal.pedscommons.org) and investigators seeking line-level access to data may submit a project request to the INTERACT executive committee for consideration. Conclusions: International collaborative data-sharing efforts will advance our understanding of pediatric and AYA AML by reducing barriers that constrain our study of AML to smaller cohorts of data. We hope that an early outcome of data sharing will be to increase our understanding of optimal treatments for patients with relapsed/refractory disease. Future goals of INTERACT include identifying data from additional studies to include, establishing relationships with new data contributors, and establishing working groups to advance the consortium’s scientific and clinical trials goals.