S18
International Journal of Radiation Oncology Biology Physics
128
Purpose/Objective(s): RTOG 0617 suggested that not all patients benefit from radiation dose escalation. We wished to develop a model to predict who might benefit. We hypothesized that 1) overall survival (OS) of a patient after radiation treatment depends on the combined effects of an individual’s baseline hazard (H), tumor control (TC), and treatment mortality (TM); 2) both TC and TM depend on a patient’s intrinsic radiosensitivity, and thus predicted by biomarkers of DNA repair genes of single nucleotide polymorphisms (SNPs). Materials/Methods: Ninety-two eligible stage III non-small cell lung cancer (NSCLC) patients who enrolled in our institutional prospective protocols were included in the study. The minimum follow-up was 25 months. A biomarker signature, which is composed of 5 DNA repair SNPs (ERCC2_rs238406, ERCC1_rs11615, ERCC1_rs3212948, XRCC4_rs9293329, XRCC4: rs2075685), and was previously reported to predict OS in 119 NSCLC patients, including these and 27 other stage III patients, was used to divide the patients into 2 groups. We hypothesized that the group with better OS was radiosensitive, and the other radio-resistant. The 2 year OS data were fitted to the following model: OS(D,t)Z[1-H(t)][1/(1+(D50/D)k)][1-1/(1+(mD50/D)k)], where the 3 terms correspond to the contributions of H, TC and TM, respectively; D is the radiation prescription tumor dose; t is the time; D50 and k are 2 fitting parameters that depend on the biomarkers; m is a parameter reflecting the tolerance dose ratio between tumor and normal tissue toxicity, and was assumed to not relate to biomarkers, but rather to factors such as the radiation technique; and H(t) Z 5% was estimated for 2 years. The patient data were first fitted to this model for all patients as a combined group to derive the value of m, and then for the radiosensitive and resistant groups to determine D50 and k of the 2 groups, respectively. Results: The patient data and OS curves are well fitted to the model, with R2 Z 0.93 and 0.98, respectively for the radiosensitive and resistant groups. The computed m is 1.46 (95% CI: 1.35, 2.42). D50 is 66.7 (63.7, 70.2) and 81.4 (77.7, 85.1) Gy for the sensitive and resistant group, respectively. K is 11.0 and 16.2 for the 2 groups, respectively. OS is much higher for the sensitive group over the resistant group at relatively low radiation doses. However, for further increases in dose above 80 Gy, the OS for the sensitive group is predicted to drop (due to increased toxicity), while OS of the resistant group rises to > 80% at the highest dose. The results suggest that the model may be used to guide individualized dose prescription to improve OS. Conclusions: We have developed a model that uses the radiation dose, biomarker signature and baseline clinical data to predict OS in patients with lung cancer. This model, if validated, might explain the results of RTOG 0617, in that only a subset of patients with radioresistant tumors may benefit from dose escalation. Author Disclosure: J. Jin: E. Research Grant; I have a R01 grant from NCI/NIH, which is not related to this study., My wife, Feng-Ming Kong, has a R01 grant from NCI/NIH, directly related to this study, My wife, Feng-Ming Kong, has a research grant from Varian Medical system. F. Honoraria; My wife, Feng-Ming Kong received Honoraria From Varian Medical system. S. Leadership; My wife, Feng-Ming Kong, is the president elect of the American Association of Women radiologists. W. Wang: None. N. Bi: None. R.K. Ten Haken: None. M. Schipper: None. R. Sadek: None. H. Zhang: None. T. Lawrence: E. Research Grant; P01 and R01 grants from NIH/NCI. S. Leadership; previous president of ASTRO. F. Kong: E. Research Grant; Feng-Ming Kong received R01 grant from NCI/NIH, Feng-Ming Kong received research grant from Varian Medical System. F. Honoraria; Feng-Ming Kong received Honoraria From Varian Medical system. S. Leadership; Feng-Ming Kong, is the president elect of the American Association of Women radiologists.
Postoperative Radiation Therapy (PORT) Is Associated With Better Survival in Non-Small Cell Lung Cancer With Involved N2 Lymph Nodes Locally Advanced Non-Small Cell Lung Cancer J.L. Mikell,1,2 T.W. Gillespie,3,2 W.A. Hall,1,2 D.C. Nickleach,4,2 Y. Liu,4,2 J. Lipscomb,3,2 S.S. Ramalingam,5,2 R.S. Rajpara,1,2 S.D. Force,6,2 F.G. Fernandez,6,2 T.K. Owonikoko,5,2 R.N. Pillai,5,2 F.R. Khuri,5,2 N.K. Jegadeesh,1,2 W.J. Curran,1,2 and K.A. Higgins1,2; 1Department of Radiation Oncology, Emory University, Atlanta, GA, 2Winship Cancer Institute, Atlanta, GA, 3Rollins School of Public Health, Emory University, Atlanta, GA, 4Biostatistics and Bioinformatics Shared Resource, Emory University, Atlanta, GA, 5Department of Hematology and Oncology, Emory University, Atlanta, GA, 6Department of Surgery, Atlanta, GA Purpose/Objective(s): The use of PORT for resected non-small cell lung cancer (NSCLC) remains controversial. Limited data indicate that PORT may be beneficial for patients with involved N2 nodes. The present study evaluates this hypothesis in a large retrospective cohort of patients treated with adjuvant chemotherapy and contemporary radiation techniques. Materials/Methods: The National Cancer Data Base (NCDB), a prospectively-acquired database and joint endeavor of the Commission on Cancer of the American College of Surgeons and the American Cancer Society, was queried for patients with resected NSCLC with pathologically involved N2 (pN2) nodes treated between 2004 and 2006. Patients with positive margins, incomplete survival data, no receipt of adjuvant chemotherapy, histology other than NSCLC, and those treated with cobalt60, non-beam radiation therapy, or neoadjuvant therapy were excluded. A multivariable Cox proportional hazards model was used to assess factors associated with overall survival (OS). Inverse probability of treatment weighting (IPTW) using the propensity score was also used to reduce treatment selection bias. OS was compared between patients treated with PORT vs those not treated with PORT using the adjusted Kaplan-Meier estimator and the weighted log-rank test based on IPTW. Results: Between 2004 and 2006, 2115 patients were eligible for analysis. Nine hundred eighteen (43.4%) received PORT, while 1197 (56.6%) did not. PORT was associated with better OS (median survival time (MST) 42 months with PORT vs 38 months without, PZ.045). This effect was significant in multivariable and IPTW Cox models (HR 0.87, 95% CI 0.780.98, PZ.026, and HR 0.89, 95% CI 0.79-1.00, PZ.046, respectively). On multivariable analysis, female gender, adenocarcinoma histology, higher income, urban/rural setting vs metro area, lower T stage, 1-2 involved lymph nodes (LN) vs 3, higher number of examined LN, and younger age were associated with better OS (all P<.05). No interaction was seen between the effects of PORT and number of involved LN (PZ.615). Conclusions: PORT was associated with better survival for patients with pN2 nodes also treated with adjuvant chemotherapy. No interaction was seen between the benefit of PORT and number of involved LN. Despite the limitations of retrospective analysis and of the NCDB, including potential data miscoding, limited survival data, and inclusion of patients treated 10 years ago, these findings reinforce the benefit of PORT for N2 disease in modern practice using the largest, most recent cohort of chemotherapytreated pN2 patients to date. Author Disclosure: J.L. Mikell: None. T.W. Gillespie: None. W.A. Hall: None. D.C. Nickleach: None. Y. Liu: None. J. Lipscomb: None. S.S. Ramalingam: None. R.S. Rajpara: None. S.D. Force: None. F.G. Fernandez: None. T.K. Owonikoko: None. R.N. Pillai: None. F.R. Khuri: None. N.K. Jegadeesh: None. W.J. Curran: None. K.A. Higgins: None.
129 A Blood Biomarker Dependent Survival Model for NSCLC Patients Treated With Radiation Therapy Locally Advanced Non-Small Cell Lung Cancer J. Jin,1 W. Wang,2 N. Bi,2 R.K. Ten Haken,2 M. Schipper,2 R. Sadek,1 H. Zhang,1 T. Lawrence,2 and F. Kong1; 1Georgia Regents University, Augusta, GA, 2University of Michigan, Ann Arbor, MI
130 Comparing Risk of Radiation Pneumonitis After Intensity Modulated Radiation Therapy Versus Proton Therapy Locally Advanced Non-Small Cell Lung Cancer Z. Liao, R. Mohan, T. Xu, J.D. Cox, and S.L. Tucker; University of Texas MD Anderson Cancer Center, Houston, TX