Lymphocyte Nadir Following Radiation for Soft-Tissue Sarcoma: Defining Key Dosimetric Parameters and Outlining Clinical Significance

Abstract

<h3>Purpose/Objective(s)</h3> Lymphopenia following radiotherapy (RT) is associated with worse oncologic outcomes in numerous cancers, but its incidence and effects are unknown in soft-tissue sarcomas (STS). Our goals were to determine the dosimetric parameters associated with development of lymphopenia and evaluate the impact of lymphopenia on overall survival (OS) and disease-free survival (DFS). <h3>Materials/Methods</h3> We retrospectively reviewed 51 patients who underwent either preoperative, postoperative, or definitive RT for STS at our institution between April, 2010 and May, 2021. Absolute lymphocyte count (ALC) nadir was defined as lowest lymphocyte value within one year after RT, and lymphopenia grading was by CTCAE v5.0. Dosimetric parameters were evaluated for associations with ALC nadir, including PTV size as well as mean and volumetric doses (V10, V20, V30) of body, bone, heart, and lung. <h3>Results</h3> Median follow-up was 1.9 years (range 0.3-9.7 years). Stage I cancer was present in 15.7%, Stage II in 11.8%, Stage III in 68.6%, Stage IV in 3.9% (for N1). Grade 1 disease was found in 6.5%, Grade 2 in 6.5%, Grade 3 in 78.4%, and unknown in 9.8%. Chemotherapy was delivered to 35.3%. Median RT dose was 50 Gy (range 45-78 Gy) and IMRT was used for 70.6%. Time to ALC nadir was within 6 months of RT for 86% of patients. Median ALC at baseline was 1400 cells/mm³ (range 600 to 3400 cells/mm³) and at nadir was 900 cells/mm³ (range 40 to 2400 cells/mm³). Strongest correlations with ALC nadir were body V10 (r_s = 0.435, p = 0.002), mean bone (r_s = 0.632, p = 0.001), bone V10 (r_s = 0.632, p <0.001), and bone V20 (r_s = 0.562, p <0.001). Significant but weaker correlations included mean body, body V20, body V30, and bone V30. Lung and heart parameters as well as total dose did not show significant correlations. On multivariable linear regression controlling for disease grade, surgical resection, and chemotherapy use, strongest correlations with ALC nadir were body V10 (β = 0.468, p < 0.001) and bone V10 (β = 0.479, p < 0.001). Distribution of lymphopenia was: grade 1 (17.6%), grade 2 (13.7%), grade 3 (19.6%), and grade 4 (5.9%). Grade 2-4 lymphopenia was associated with inferior OS (HR 5.82, p = 0.01) and DFS (HR 3.85, p=0.002). Lower ALC nadir was also associated with worse OS (HR = 1.95, p = 0.014) and DFS (HR = 1.15, p = 0.001). On multivariate regression, only Grade 3-4 lymphopenia remained associated with worse OS and DFS. <h3>Conclusion</h3> Lower ALC nadir and development of Grade 2-4 lymphopenia were associated with worse OS and DFS in STS. Despite the heterogenous location of STS presentation, lower ALC nadir showed strong correlations with increasing body V10 and bone V10, consistent with the known radiosensitivity of lymphocyte cell lines. Our study supports efforts to reduce treatment-related hematopoietic toxicity as a move to improve oncologic outcomes. This work also raises the hypothesis that radiation effects on lymphocyte progenitors in the bone marrow dominate those of circulating lymphocytes for treatments with minimal contribution to heart and lung.