ESTRO 2024 - Abstract Book

S2176

Clinical - Upper GI

ESTRO 2024

923

Poster Discussion

Defining the optimal radiation-induced lymphopenia metric to discern survival in esophageal cancer

Pim J.J. Damen 1,2 , Max Peters 3 , Brian Hobbs 4 , Yiqing Chen 5 , Uwe Titt 6 , Remi Nout 1 , Radhe Mohan 7 , Steven H Lin 2 , Peter S.N. van Rossum 8,2 1 Erasmus MC Cancer Institute, University Medical Center Rotterdam, Department of Radiotherapy, Rotterdam, Netherlands. 2 The University of Texas MD Anderson Cancer Center, Department of Radiation Oncology, Houston, USA. 3 Radiotherapiegroep, Department of Radiotherapy, Deventer, Netherlands. 4 The University of Texas at Austin, Department of Population Health, Austin, USA. 5 University of Texas Health Science Center, Department of Biostatistics and Data Science, Houston, USA. 6 The University of Texas MD Anderson Cancer Center, Department of Radiation Physics, Houston, USA. 7 The University of Texas MD Anderson Cancer Center, Department of Radiation PhysicsRadiation Physics, Houston, USA. 8 Amsterdam UMC, Department of Radiotherapy, Amsterdam, Netherlands Emerging evidence suggests a detrimental prognostic association between radiation-induced lymphopenia (RIL) and pathologic response, progression-free survival (PFS) and overall survival (OS) in patients with esophageal cancer. To implement this knowledge into clinical decision-making and for the design of lymphopenia-mitigating trials, it is of considerable importance to determine an optimal metric for RIL. Such a metric remains undefined in current literature, but in principle would be easy to measure, easy to compare across studies, and most impactful in terms of independently discerning survival outcomes. In esophageal cancer most studies report on lymphopenia grade 4 (i.e. absolute lymphocyte count [ALC] <0.2 x103/μL), but other metrics are regularly reported as well. The aim of this study in patients who underwent chemoradiotherapy (CRT) for esophageal cancer was to find the optimal lymphopenia metric that best distinguishes patients in terms of PFS and OS. Consecutive patients with esophageal cancer treated with concurrent CRT between January 2004 and September 2022 at a large tertiary referral center were selected. Studied lymphopenia metrics included absolute lymphocyte and neutrophil counts -and calculated derivatives- both at baseline and during CRT (i.e. baseline neutrophil-to lymphocyte ratio [NLR], ALC nadir, lymphopenia grade ³3 [<0.5 x103/μL] and grade 4 during CRT according to Common Terminology Criteria for Adverse Events [CTCAE] criteria, ALC and neutrophils per week, maximum NLR, maximum neutrophil count, and both absolute and relative ΔALC and ΔNLR values for each week during CRT). Univariable Cox proportional hazard regression models were created to study the crude association of the RIL metrics with PFS and OS. A multivariable model was developed to analyze the association of each RIL metric with PFS and OS, adjusted for potential confounders. Model performance was defined as the best for models with the lowest value of the Akaike's Information Criterion (AIC). Corrected time-dependent c-statistics were calculated for the best performing model and also for the model with grade 4 lymphopenia (as most commonly used metric in esophageal cancer in literature). The optimal RIL metric was defined as the metric providing a model with the highest corrected c-statistic. Kaplan-Meier curves for PFS and OS were generated for the best performing metric. Purpose/Objective: Material/Methods: