ESTRO 2023 - Abstract Book

S586

Monday 15 May 2023

ESTRO 2023

Of 600 pts, 545 pts were included: OPSCCp16+ (n=282) and non-HPV-driven (n=263 (OPSCCp16- (n=100); oral cavity (n=20); hypopharynx (n=65); larynx (n=78))). Pts were excluded in cases of insufficient tumor tissue in the FFPE block (n=35) and due to missing volume data (n=20). For non-HPV-driven tumors, large total tumor volume (GTVTot) was a poor prognostic factor for LRF (Figure 1). The hazard ratio (HR) for LRF for the large volume-group was 3.2 (95 %CI: 1.7-5.7) compared to small volume. Number of CSC increased the ability to distinguish a group with lower risk of LRF, with HR=2.5 (1.2-4.9) for intermediate vs. low CSC and HR=2.7 (1.4-5.4) for high vs. low CSC number. Total volume was not prognostic for pts with OPSCCp16+ (Figure 2). The number of CSCs was able to identify groups with different risks of LRF. Pts with high number of CSCs had a higher risk (HR=4.4 (1.6-11.7)) of LRF compared to low CSC. For both OPSCCp16+ and non-HPV-driven tumors, a similar pattern was reflected in the risk of T- and N-site failure. Higher risk of failure for non-HPV-driven tumor was seen for pts with larger volume as well as with higher number of CSC. For OPSCCp16+ tumors, the absolute CSC number distinguished groups with higher and lower risk of failure (HR=3.0 (1.2-7.7) for N-site failure for high vs low CSC number).

Conclusion Tumor volume was prognostic for the non-HPV-driven tumors only. In contrast to volume, the level of CSC was prognostic in OPSCCp16+, which may benefit in selection of pts for individualized treatment. MO-0716 Falls in lymphocyte counts during radiotherapy predict cisplatin benefit in oropharynx cancer J. Price 1,2 , D. Thomson 1,2 , G. Price 2 , H. Mistry 3 , C. West 4 , R. Prestwich 5 , Z. Iyizoba-Ebozue 5 1 The Christie NHS Foundation Trust, Clinical Oncology, Manchester, United Kingdom; 2 The University of Manchester, Division of Cancer Sciences, Manchester, United Kingdom; 3 The University of Manchester, Division of Pharmacy & Optometry, Manchester, United Kingdom; 4 Division of Pharmacy & Optometry, Division of Cancer Sciences, Manchester, United Kingdom; 5 Leeds Cancer Centre, Clinical Oncology, Leeds, United Kingdom Purpose or Objective For patients with oropharynx squamous cell carcinoma (OPSCC) treated with (chemo) radiotherapy (RT), the pre-RT absolute lymphocyte count (ALC) is prognostic and predicts benefit from the addition of platinum chemotherapy. This study aimed to test the prognostic & predictive value of ALC dynamics during RT Materials and Methods A retrospective observational study of consecutive OPSCCs treated by curative intent RT +/- concurrent chemo (n = 964) with external, independent validation from a separate institution (n = 608). The primary endpoint was 5-year overall survival (OS). ALCs were recorded from RT start to day 84 & a mathematical model, previously developed to analyse chemo-induced lymphocyte falls, was used to analyse the ALC time-series data. The model is semi-mechanistic & summarises the entire time-series via 3 parameters: MTT (mean transit time, the time taken for lymphocytes to mature & appear in blood) Circ0 (the pre-RT ALC) & feedback strength (relating to the rate of recovery of ALCs). Previously determined prognostic factors were also used in a multivariable Cox proportional hazards analysis to assess the prognostic importance of ALC parameters & their interaction with cisplatin use Results Patient characteristics are shown in the Table. In the discovery cohort, ALCs fell during both RT and chemo-RT. Feedback strength (describing lymphocyte fall & recovery) was prognostic for patients receiving cisplatin chemo-RT (HR: 1.16, 95% CI: 1.04-1.29, p=0.008) but not RT alone (HR: 1.05, 95% CI: 0.95–1.15, p=0.328): patients receiving chemo-RT with low feedback values (i.e. whose time-series demonstrated larger ALC falls & slower recovery) had improved 5-year OS. The cohort was stratified into groups with differing 5-year OS estimates (Fig A). The finding was replicated in an external