S590
Clinical – Head & neck
ESTRO 2026
Digital Poster 2193
Definitive Radiotherapy Alone for Head and Neck Cancer in Patients Aged ≥70 Years: Predictors of Incidence in In-Field Recurrence and Prognosis. Masaaki Goto, Takeshi Fujisawa, Ren Iijima, Kei Nakai, Haruko Numajiri, Takashi Saito, Takashi Iizumi, Masatoshi Nakamura, Toshiki Ishida, Keiichiro Baba, Takuya Sawada, Masahiko Harada, Hideyuki Sakurai Department of Radiation Oncology, University of Tsukuba, Tsukuba, Japan Purpose/Objective: Definitive chemoradiotherapy is one of the standard treatments for patients with locally advanced head and neck cancer (HNC). However, elderly patients are often treated with definitive radiotherapy (dRT) alone due to their comorbidities or lower performance status, and there are few reports on contemporary clinical outcomes in this specific population, including patterns of failure and prognosis. Therefore, we conducted a retrospective study to evaluate clinical outcomes in elderly patients with HNC who received dRT alone. Material/Methods: We retrospectively reviewed patients aged ≥ 70 years with advanced squamous cell carcinoma of the oral cavity, oropharynx, and hypopharynx who received dRT alone at our institution between March 2013 and September 2023. The primary endpoint was the cumulative incidence of in-field recurrence (IFR), with death treated as a competing risk. Secondary endpoints were overall survival (OS) and planned treatment completion. Predictors of IFR were assessed using multivariable Fine–Gray subdistribution hazard models with prespecified covariates, including age, performance status (PS), primary site, clinical stage, p16 status, and overall treatment time. Median follow- up was calculated using the reverse Kaplan–Meier method. OS was estimated by the Kaplan–Meier
Conclusion: The proposed text-guided latent diffusion model generates synthetic CT images with high accuracy in image quality as well as OAR delineation consistency for HNC radiotherapy, demonstrating strong potential for application in MR-only radiotherapy workflows. References: 1.Yan Y, Kim JP, Nejad-Davarani SP, et al. Deep Learning-Based Synthetic Computed Tomography for Low-Field Brain Magnetic Resonance-Guided Radiation Therapy. Int J Radiat Oncol Biol Phys. 2025;121(3):832- 843. doi:10.1016/j.ijrobp.2024.09.0462.Wang C, Ong HH, Chiba S, Rajapakse JC. DE NOVO MOLECULE GENERATION WITH GRAPH LATENT DIFFUSION MODEL.IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS. 2024;2121-2125. doi:10.1109/ICASSP48485.2024.10447480. Keywords: Diffusion model ， Multi-sequence MRI ， Synthetic CT
method. Results:
Sixty-seven patients aged ≥ 70 years were included (median age: 79 years, range: 70–92) with a median follow-up of 44 months (95% confidence interval [CI], 27–51). Primary sites were oral cavity/oropharynx/hypopharynx in 20/22/25 patients. Clinical stage (UICC 8th) were I/II/III/IV/IVA/IVB in 5/4/9/4/39/6. In oropharyngeal patients, p16 status was positive/negative/unknown in 15/2/5. The median prescribed dose was 70 Gy in 35 fractions (range: 8– 71Gy/ 4–38fr). During follow-up, 30 IFR events (45%) occurred. Out-of-field recurrences alone occurred in nine patients (13%), all of which were distant metastases. The 3-year cumulative incidence of IFR was 47% (95% CI, 34–59). The 3-year OS was 56% (95% CI, 44–70). In multivariable analysis, performance
Made with FlippingBook - Share PDF online