Feasibility Assessment of a Real-World Comparator Cohort Using Multiple Qualifying Index Lines of Therapy (LOTs) in Third-Line or Later (3L+) Diffuse Large B-cell Lymphoma (DLBCL)
Author(s)
Michael Wallington, BA1, Yong Chen, PhD2, Fei Jie, PhD3, Monica Marie Patterson, PharmD4, Karen Repetny, PhD5, Jacob Ambrose, MS6, Magnolia Watson, MPH6, Christina Zettler7, Andrew J. Belli, MPH6, Laura L. Fernandes, PhD6, Ching-Kun Wang, MD6.
1Pfizer UK, London, United Kingdom, 2Pfizer, Collegeville, PA, USA, 3Pfizer, Bothell, WA, USA, 4Pfizer, Boston, MA, USA, 5Pfizer, Cambridge, MA, USA, 6COTA, Inc., New York, NY, USA, 7Research, COTA, Inc., New York, NY, USA.
1Pfizer UK, London, United Kingdom, 2Pfizer, Collegeville, PA, USA, 3Pfizer, Bothell, WA, USA, 4Pfizer, Boston, MA, USA, 5Pfizer, Cambridge, MA, USA, 6COTA, Inc., New York, NY, USA, 7Research, COTA, Inc., New York, NY, USA.
OBJECTIVES: Real-world (rw) data, including electronic health record (EHR) datasets, are increasingly used to contextualize clinical trial results in a contemporary treatment landscape. In clinical trials, patients qualify at enrollment; however, a patient could be eligible in lines of therapy (LOTs) prior to or following enrollment, depending on when trial assessment occurred. With retrospective EHR data typically including patients’ entire treatment history, rw studies require deliberate choice to establish optimal methods regarding which LOT serves as the comparison time zero (index) for each eligible patient. Our objective was to explore multi-LOT qualification methods for future comparative analysis.
METHODS: Eligibility criteria from ECHELON-3 were adapted and applied to COTA’s rw, EHR-based diffuse large B-cell lymphoma (DLBCL) database. Eligible adult patients had received ≥2 prior LOTs for DLBCL. Eligible patient-LOTs were identified from the following criteria: 3L+ LOT initiation between OCT/18/2017 (date of first DLBCL chimeric antigen receptor T-cell therapy (CAR-T) approval) and DEC/31/2023 to allow for sufficient follow-up; no solid tumor malignancy within two years of index treatment, and no exclusionary treatment history. Results were summarized among unique patients and qualifying patient-LOTs.
RESULTS: In total 585 unique patients had at least one qualifying index-LOT, with 880 qualifying index-LOTs identified. The median number of qualifying LOTs per patient was 1 (min=1, max=7). The majority of qualifying index-LOTs occurred at 3L (n=474; 53.9%), and there were 212 and 97 qualifying patient-LOTs in 4L and 5L, respectively, representing nearly 90% of qualifying index-LOTs between 3L-5L.
CONCLUSIONS: Among 585 unique patients, there was a 50% increase in qualifying index-LOTs when assessing each LOT independently for qualification demonstrating that choice of index LOT in retrospective rw studies in not trivial. This analysis demonstrates an ability to assess qualification across a patient journey, and future research is needed to understand the sensitivity of rw results based on the selected index date.
METHODS: Eligibility criteria from ECHELON-3 were adapted and applied to COTA’s rw, EHR-based diffuse large B-cell lymphoma (DLBCL) database. Eligible adult patients had received ≥2 prior LOTs for DLBCL. Eligible patient-LOTs were identified from the following criteria: 3L+ LOT initiation between OCT/18/2017 (date of first DLBCL chimeric antigen receptor T-cell therapy (CAR-T) approval) and DEC/31/2023 to allow for sufficient follow-up; no solid tumor malignancy within two years of index treatment, and no exclusionary treatment history. Results were summarized among unique patients and qualifying patient-LOTs.
RESULTS: In total 585 unique patients had at least one qualifying index-LOT, with 880 qualifying index-LOTs identified. The median number of qualifying LOTs per patient was 1 (min=1, max=7). The majority of qualifying index-LOTs occurred at 3L (n=474; 53.9%), and there were 212 and 97 qualifying patient-LOTs in 4L and 5L, respectively, representing nearly 90% of qualifying index-LOTs between 3L-5L.
CONCLUSIONS: Among 585 unique patients, there was a 50% increase in qualifying index-LOTs when assessing each LOT independently for qualification demonstrating that choice of index LOT in retrospective rw studies in not trivial. This analysis demonstrates an ability to assess qualification across a patient journey, and future research is needed to understand the sensitivity of rw results based on the selected index date.
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
SA42
Topic
Study Approaches
Disease
Oncology