Introduction: Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma, accounting for approximately 30% to 40% of all cases. Although randomized clinical trial data inform treatment standards, there is limited large-scale real-world evidence (RWE) from the clinical setting in Europe, particularly from Germany. Understanding patient characteristics, real-world treatment patterns, and outcomes can better inform clinical practice and identify gaps in care. This study evaluated patient characteristics, treatment pathways, and clinical outcomes in Germany and the UK using electronic health records (EHR)-derived data.

Methods:We conducted a retrospective analysis using the Flatiron Health Research Database in the UK and Germany. This is a longitudinal, deidentified database using structured and unstructured data accessed via technology-enabled abstraction from EHRs across oncology practices in Germany and NHS trusts in the UK. The dataset undergoes rigorous quality checks, is anonymized per local regulations, and updated with 90-day recency. Patients diagnosed with DLBCL between January 1, 2011 to March 31, 2025 were included. Descriptive analyses were performed to assess staging, biomarker testing, treatment sequences, and cytogenetic profiling. Kaplan-Meier methods were used to estimate overall survival (OS) from initial diagnosis, with subgroup analysis by cell of origin (COO) and extranodal involvement.

Results: Two DLBCL patient cohorts were analyzed: 190 from Germany and 990 from the UK (median age: 68 years in both). Staging at diagnosis was reported for 83% (Germany) and 64% (UK), with stage IV disease present in 26% and 52%, respectively. COO was documented in 45% (Germany) and 77% (UK). Among those with documented COO, the non-germinal center B-cell/activated B-cell subtype was documented in 34% (Germany) and 49% (UK). Extranodal disease was present in 64% (Germany) and 73% (UK), with the gastrointestinal tract being the most common site in both cohorts. Cytogenetic testing rates were 35% (Germany) and 69% (UK). MYC rearrangement was the most frequently tested alteration, with positive rates of 24% (Germany) and 14% (UK). First-line (1L) radiotherapy was administered to 24% (Germany) and 38% (UK). Among patients receiving 1L therapy, 16% (Germany) and 30% (UK) progressed to second-line, and 6% (Germany) and 12% (UK) to third-line treatment. While seeing documentation of CAR T-cell therapy in <10% of patients in Germany and the UK, the number of CAR T-cell treated patients accounted for around 40% of the number of patients that made it to 3L in both countries. There was lower median OS among patients with non-germinal center B-cell COO and extranodal involvement.Conclusions: This study offers novel, EHR-derived insights into DLBCL management in real-world clinical settings across Germany and the UK. Differences in biomarker testing and treatment transitions reflect variations in clinical practice and data infrastructure. The observed cohort characteristics align with expectations from clinical trials and other RWE studies in Europe (eg, REALYSA, REFLECT). As additional patients accrue, more detailed analyses of survival, treatment response, and care gaps beyond 1L therapy will be possible. These findings support the integration of pan-European real-world data to inform guideline adaptation and optimize treatment pathways.

This content is only available as a PDF.
2025
Sign in via your Institution