2026 ASH®
Annual Meeting & Exposition December 12-15, 2026 ▪ New Orleans, Louisiana

Paper Details

Abstracts, authors, affiliations, and disclosures.

Monday, December 14, 2026
06:00 p.m. - 08:00 p.m.
MCCNO - Hall F
Access, Representation, Population Health, Technology and Procedures, Imaging, Molecular Testing, Cell Expansion, Treatment Considerations, Combination Therapy, Drug Development, Drug-Drug Interactions, Supportive Care, Therapy Sequence, Biological Therapies, Antibody Therapy, Bispecific Antibody Therapy, Chimeric Antigen Receptor, (CAR)-T Cell Therapies, Immunotherapy, Monoclonal Antibody Therapy, Non-Biological Therapies, Chemotherapy, Pharmacology, Adverse Events, Biological Processes, Immune Mechanism, Clinical Practice (Health Services and Quality), Health Economics, Patient Safety, Practice Models, Quality Improvement, Quality of Care, Value, Diseases, Immunology, Immunodeficiency, Infectious Diseases, Viral, Bacterial, Fungal, Other Pathogens, Lymphoid Malignancies, Lymphoid Leukemias, Lymphomas, Plasma Cell Disorders, Myeloid Malignancies, Acute Myeloid Malignancies, Chronic Myeloid Malignancies, Metabolic Disorders, Measurable Residual Disease, ALL, CLL, Non-Hodgkin Lymphoma, B Cell Lymphoma, Indolent Lymphoma, Aggressive Lymphoma, Research, Clinical Research, Clinical Trials, Epidemiology, Health Outcomes Research, Health Disparities Research, Patient-Reported Outcomes, Real-World Evidence, Registries, Survivorship, Study Population, Human, Adult, Elderly, Young Adult
PAPER

6288 Real-world clinical outcomes and CD20/CD19 antigen dynamics in 3l+ relapsed/refractory diffuse large B-cell lymphoma: A US EHR analysis of an understudied population

Yannis Darhi, Muhammad Ekram2, Marlene Zuraek, MD3, Wen-I (Wendy) Chang, MD4, Alexander Grigorenko, MS5, Azadeh Shohoudi, PhD6, Malini Iyengar7, Yasuhiro Oki, MD8
1 AstraZeneca, Barcelona, Spain
2 AstraZeneca, Waltham, Massachusetts, United States
3 AstraZeneca, San Francisco, California, United States
4 AstraZeneca, New York, New York, United States
5 AstraZeneca, New York, New York, United States
6 AstraZeneca, Gaithersburg, Maryland, United States
7 AstraZeneca, San Francisco, California, United States
8 AstraZeneca, San Francisco, California, United States
Background

Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma. Although many patients achieve durable remission with frontline therapy, 35-40% develop relapsed/refractory (R/R) disease. In the era of newer therapies, real-world data on the treatment landscape and outcomes in patients receiving third-line or later (3L+) therapy remain limited. This study evaluated outcomes by treatment class in US patients with R/R DLBCL and characterized the prevalence and clinical impact of CD20 loss after repeated CD20-targeted therapy.

Methods

This retrospective study used de-identified COTA electronic health record data from 13,907 US DLBCL patients treated in academic and community practices between 2002 and 2025 (data cutoff Aug 29, 2025). Patients initiating 2L+ therapy after Jan 2023 were included to reflect the post-approval era of T-cell engagers (TCE). Lines of therapy (LoT) were classified using a proprietary algorithm that combines treatments into LoT based on a set of disease-specific rules. An AI-assisted algorithm grouped more than 600 treatment combinations by therapeutic class. Outcomes included real-world overall response rate (rwORR), complete response rate (rwCRR), and overall survival (rwOS) analyzed by LoT (3L vs. 4L+) and treatment group (TCE, CAR-T, polatuzumab-based, and tafasitamab-lenalidomide).  CD19 and CD20 status for each line were defined by the closest assessment, which was abstracted from source documents, based on qualitative clinician assessment of IHC and flow cytometry. Associations between CD20 loss and response to subsequent therapy were assessed using Fisher's exact test and multivariable logistic regression adjusting for demographic and clinical characteristics. Treatment setting was evaluated as a proxy for access.

Results

Among eligible patients, 260 received 3L therapy (median age 68 yrs, 41.8% female) and 156 received 4L+ therapy (median age 66 yrs, 37.6% female). In 3L and 4L+ respectively, 40.4%/29.5% received CAR-T, 23.1%/26.9% chemotherapy-based, 8.8%/28.8% TCE, 4.6%/8.3% polatuzumab-based and 4.2%/10.3% tafasitamab-lenalidomide.

Use of novel therapies was concentrated in academic centers. Among patients treated with TCE, 69.6% in 3L and 75.6% in 4L+ were treated in academic centers. Similarly, CAR-T use was more frequent in academic centers (61.9% in 3L and 71.7% in 4L+) than community settings, suggesting care-setting differences in access to newer therapies.

Overall rwORR/rwCR was 56.5%/35.4% in 3L (N=260) vs. 45.5%/25.6% in 4L+ (N=156). In 3L, rwORR/rwCRR was 67.6%/54.3% with CAR-T (N=105) and 65.2% /30.4% with TCE (N=23) and in 4L+, 54.3% /43.5% with CAR-T (N=46) and 42.2%/20.0% with TCE (N=45). Median rwOS was 28 months in 3L (28 months with CAR-T; 12 months with TCE) versus 11 months in 4L+ (not reached with CAR-T; 11 months with TCE). Outcomes for polatuzumab-based and tafasitamab-lenalidomide were not characterizable due to small sample sizes.

CD20 loss increased with later lines of therapy, from 1.2% at 1L (125 negative/10387 tested) to 38.2% in 4L (13/34). In paired pre- and post-treatment samples, 56% (14/25) of patients treated with CD20-targeting TCEs lost CD20 expression after treatment. In unadjusted analysis, CD20 loss was associated with lower odds of response to subsequent CD20-targeted therapy in unadjusted analysis (OR=0.56, 95% CI 0.34-0.92); this association was attenuated after adjustment for baseline characteristics (OR=0.38, 95% CI 0.10-1.36). CD19 loss increased from 2.3% in 1L (38/1677) to 11.7% in 4L (7/60). Among those receiving CD19-directed CAR-T, CD19 loss was observed in 39.4% (28/71) suggesting antigen escape as a potential mechanism for relapse. Among patients with CD20 loss, subsequent CD19 loss was uncommon, observed in 9.1% (2/22) at 1L and 16.7% (6/36) at 4L+, suggesting that CD19 expression is often retained despite CD20 loss.

Conclusions

In this real-world analysis of R/R DLBCL, outcomes remained poor in patients receiving 3L and 4L+ therapy, underscoring persistent unmet need despite the availability of TCE and CAR-T therapies. Use of newer therapies was concentrated in academic centers, suggesting potential care-setting differences in access. CD20 loss increased with successive CD20-targeted therapy and was associated with lower response. In contrast, CD19 expression loss was less frequent. Further prospective evaluation of CD19-directed and antigen-guided treatment strategies is warranted in advanced R/R DLBCL.

Yannis Darhi: AstraZeneca - Current Employment