S750
Clinical - Lung
ESTRO 2026
Results: Among 416 patients, 198 (48%) initiated immunotherapy. Male gender, lower age, IMPT radiotherapy, former/never smoking, and lower lung V3 (volume receiving ≥ 3 Gy) were final predictors for initiation of immunotherapy (Figure 2). The model showed fair discrimination (optimism-adjusted AUC 0.71, 95% CI 0.67–0.76) and good calibration (optimism-adjusted slope 0.84, 0.75–1.08). Interestingly, after IMPT, 100 out of 137 (73%) qualified for immunotherapy compared to 98 out of 279 (35%) after VMAT (P<0.001). There was no significant interaction between radiotherapy technique and Lung V3. AUC was similar across both techniques, with adequate fit.Figure 2. Figure 2. Association (Odds Ratios and 95% CI) of predictors with starting adjuvant immunotherapy after CRT for stage III NSCLC.
Digital Poster Highlight 99
Prognostic value of longitudinal nutritional and inflammatory scores in NSCLC patients undergoing radiotherapy: a prospective study (UMIN000056183) Zhe Chen 1 , Kengo Kuriyama 2 , Mitsuhiko Oguri 3 , Takuto Kimura 4 , Kunihiro Hiranabe 4 , Masaki Matsuda 4 , Tomoko Akita 4 , Juria Muramatsu 4 , Kan Marino 4 , Takafumi Komiyama 4 , Hiroshi Onishi 4 1 Imaging-integrative Therapeutic Radiology, University of Yamanashi, Chuo, Japan. 2 Radiation Oncology, Yamanashi Prefectural Hospital, Kofu, Japan. 3 Radiology, Shizuoka General Hospital, Shizuoka, Japan. 4 Therapeutic Radiology, University of Yamanashi, Chuo, Japan Purpose/Objective: To evaluate the prognostic value of longitudinal nutritional and inflammatory scores, including the modified Glasgow Prognostic Score (mGPS), Patient- Generated Subjective Global Assessment (PG-SGA), and body weight, in patients with non-small cell lung cancer (NSCLC) undergoing radiotherapy (RT). Material/Methods: In this prospective observational study (UMIN000056183), patients aged ≥ 18 years with histologically or clinically confirmed NSCLC and ECOG performance status 0–1 were enrolled after providing written informed consent. Stage I patients received stereotactic body radiation therapy (SBRT), while those with higher-stage disease received chemoradiotherapy or RT alone. Nutritional and inflammatory parameters were assessed before and during RT, including mGPS, PG-SGA, and body weight. The primary endpoint was overall survival (OS). Time-dependent Cox regression was used to assess the association of longitudinal changes in mGPS, PG-SGA, and body weight with OS. Results: We initially planned to enroll 250 patients, but enrollment had to be stopped early due to the COVID- 19 pandemic, which resulted in a smaller sample size. A total of 92 patients were enrolled between January 2021 and December 2023, and after excluding 4 withdrawals and 6 cases with incomplete data, 82 patients were included in the final analysis. The median age was 78 years; 89% had stage I disease, and 70% received SBRT.In the time-dependent Cox model, increasing mGPS was significantly associated with worse OS (hazard ratio [HR] 1.75, 95% confidence interval [CI]: 1.06–2.87, p = 0.028). PG-SGA showed a non-significant trend (HR 1.34, 95% CI: 0.67–2.68, p = 0.415), and body weight was not predictive of OS (HR 1.00, 95% CI: 0.95–1.06, p = 0.87). Conclusion: Longitudinal assessment of mGPS during RT independently predicted OS in NSCLC patients,
Lungs V3 and Age were standardized (ORs reflect change per SD). Age (Spline 1) corresponds to cubic spline for 55–67.5 years; Age (Spline 2) to 67.5–77 years. Conclusion: We developed a model with fair performance to predict immunotherapy eligibility post-CRT. Radiotherapy technique, lung V3 and smoking represent modifiable factors that radiation oncologists can target to optimize CRT and thereby increase the proportion of NSCLC patients eligible for adjuvant immunotherapy. This model supports personalized planning and may guide selection for proton therapy. External validation is needed to confirm clinical utility. Keywords: NSCLC, prediction, immunotherapy
Made with FlippingBook - Share PDF online