Baseline Lymphocyte Count Outperforms Dosimetry and Inflammatory Markers for Predicting Severe Radiation-Induced
Konrad Stawiski1, Zuzanna Nowicka2, Jacek Burzyński3
1Department of Biostatistics and Translational Medicine, Medical University of Lodz, Lodz, Poland; Department of Brachytherapy and General Oncology/Department of Radiotherapy, Copernicus Memorial Hospital in Lodz, Lodz, Poland; Department of Radiation Oncology, Dana-Farber Cancer Institute, Boston, MA, USA.
Purpose:
To determine which baseline and treatment-planning factors best predict severe radiation-induced lymphopenia during rectal cancer chemoradiation and whether an early on-treatment absolute lymphocyte count (ALC) measurement refines risk.
Methods And Materials:
We studied 179 curative-intent rectal cancer patients treated with long-course fluoropyrimidine-based chemoradiation from 2018 to 2024. Baseline models compared baseline ALC, age, sex, 23 complete blood count parameters, 6 inflammatory markers, and 600 dosimetric variables; a day-14 ALC update was also evaluated.
Results:
Severe lymphopenia (nadir ALC <0.5 × 10^9/L) occurred in 44% of evaluable patients. Baseline ALC was the dominant predictor; risk showed a fourfold incidence gradient across baseline quartiles, from 67% in Q1 to 17% in Q4, and a three-variable nomogram using ALC, age, and sex achieved a corrected C-index of 0.746. Female sex independently increased risk (adjusted odds ratio 3.44, 95% CI 1.69-7.27). No platelet index, erythrocyte parameter, inflammatory marker, or dosimetric variable improved prediction; published pelvic bone marrow V10 and Dmean constraints were met in 98% and 100% of patients, respectively, while V20 compliance was 86%, limiting dose-constraint discrimination in these standardized fields. ALC fell 71% to a nadir at day 33, and only 19% of patients recovered to at least 1.0 × 10^9/L by 3 months. Week-2 ALC improved cross-validated area under the curve from 0.737 to 0.827. Severe lymphopenia was not associated with pathological complete response, tumor regression, or T-downstaging (all p>0.20). An unadjusted nadir-based survival association (hazard ratio 0.22) disappeared after time-dependent correction (hazard ratio 0.93).
Conclusions:
A routine baseline ALC, refined by a day-14 checkpoint, outperformed dosimetric and inflammatory profiling for severe lymphopenia risk assessment. The nomogram appears to identify toxicity susceptibility rather than treatment efficacy and supports anticipatory monitoring rather than treatment modification.