Advancing Real-World Evidence in Diffuse Large B-cell Lymphoma: Insights From EHR-Derived Data in the UK
Author(s)
Elsie Horne, PhD1, Patrycja Pluta, DVM, PhD2, Christoph Buhl, MD2, Tamra Downing, MPH3, Harlan Pittell, PhD3, Blythe Adamson, PhD, MPH3.
1Flatiron Health UK, London, United Kingdom, 2Flatiron Health Germany, Berlin, Germany, 3Flatiron Health, New York, NY, USA.
1Flatiron Health UK, London, United Kingdom, 2Flatiron Health Germany, Berlin, Germany, 3Flatiron Health, New York, NY, USA.
OBJECTIVES: Diffuse large B-cell lymphoma (DLBCL) is the most common aggressive non-Hodgkin lymphoma, with significant morbidity and mortality. While real-world evidence has been instrumental in understanding DLBCL in the US, there is a growing need for region-specific insights to address variations in healthcare systems, treatment patterns, and clinical and biological characteristics. This study describes comprehensive, deidentified data from electronic health records (EHRs) of patients with DLBCL in the UK.
METHODS: The UK DLBCL dataset was derived from the Flatiron Health Research Database, a deidentified, longitudinal database derived via technology-enabled abstraction from EHRs from routine clinical care in the UK. We used descriptive analyses to profile disease characteristics, treatment patterns, and survival outcomes. Data completeness and quality were ensured through rigorous validation processes.
RESULTS: The dataset included 990 patients diagnosed with DLBCL from 2011-2024. Stage at diagnosis was recorded for 69% of patients, with 19%, 13%, 16%, and 52% classified as stages I, II, III, and IV, respectively. The cell of origin (COO) was documented for 77% of patients, of whom 51% had non-germinal center B-cell/activated B-cell subtype. Of the 73% of patients with extranodal disease, the most common site was gastrointestinal (14%). Cytogenetic testing results were available for 70% of patients, with MYC being the most frequent alteration (14% positive). Radiotherapy was received by 38% of patients as part of first-line treatment (1L). Of the 830 patients receiving 1L, 29% and 12% went on to receive second-line and third-line treatment, respectively. Median overall survival from initial diagnosis was 5.95 years (confidence interval 5.04-7.13), with COO and site of extranodal involvement heavily associated with impaired OS.
CONCLUSIONS: The UK DLBCL dataset provided real-world clinical characteristics, treatment patterns, and outcomes, which align with clinical expectations for the UK. These insights could inform strategies to optimize treatment pathways and enable regional and global insights into DLBCL management.
METHODS: The UK DLBCL dataset was derived from the Flatiron Health Research Database, a deidentified, longitudinal database derived via technology-enabled abstraction from EHRs from routine clinical care in the UK. We used descriptive analyses to profile disease characteristics, treatment patterns, and survival outcomes. Data completeness and quality were ensured through rigorous validation processes.
RESULTS: The dataset included 990 patients diagnosed with DLBCL from 2011-2024. Stage at diagnosis was recorded for 69% of patients, with 19%, 13%, 16%, and 52% classified as stages I, II, III, and IV, respectively. The cell of origin (COO) was documented for 77% of patients, of whom 51% had non-germinal center B-cell/activated B-cell subtype. Of the 73% of patients with extranodal disease, the most common site was gastrointestinal (14%). Cytogenetic testing results were available for 70% of patients, with MYC being the most frequent alteration (14% positive). Radiotherapy was received by 38% of patients as part of first-line treatment (1L). Of the 830 patients receiving 1L, 29% and 12% went on to receive second-line and third-line treatment, respectively. Median overall survival from initial diagnosis was 5.95 years (confidence interval 5.04-7.13), with COO and site of extranodal involvement heavily associated with impaired OS.
CONCLUSIONS: The UK DLBCL dataset provided real-world clinical characteristics, treatment patterns, and outcomes, which align with clinical expectations for the UK. These insights could inform strategies to optimize treatment pathways and enable regional and global insights into DLBCL management.
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
HSD4
Topic
Health Service Delivery & Process of Care
Disease
Oncology