T G Dietterich Approximate statistical test for
 T.G. Dietterich, Approximate statistical test for comparing supervised classification learning algorithms, Neural Comput. 10 (1998) 1895–1923.
10 Original Study
Analysis of Relapse Events After Definitive Chemoradiotherapy in Locally Advanced NoneSmall-Cell Lung Cancer Patients
G. Daniel Grass, Arash O. Naghavi, Yazan A. Abuodeh, Bradford A. Perez, Thomas J. Dilling
The appropriate follow-up frequency after chemoradiotherapy (CRT) for locally advanced nonesmall-cell lung cancer patients is unknown. Our data on 211 patients found that 14042-01-4 most relapses happen within a year of completing CRT, and that symptomatic relapses portended worse outcomes versus relapses identified by surveillance imaging. More intensive surveillance imaging may identify asymptomatic relapses, which are more amenable to early salvage.
Background: The appropriate follow-up frequency after definitive chemoradiotherapy (CRT) for locally advanced nonesmall-cell lung cancer patients is unknown. Although surveillance guidelines have been proposed, very few data support current recommendations. Here we analyze relapse events after CRT and investigate whether symptomatic relapses versus those detected by surveillance imaging influences outcomes. Patients and Methods: Stage III nonesmall-cell lung cancer patients treated with CRT at our institution between 2005 and 2014 were retrospectively analyzed. Relapse events were grouped into posttreatment intervals and analyzed with cumulative tables. Time to relapse and overall survival (OS) were compared between patients with relapse detection via symptomatic presentation versus surveillance imaging. Results: A total of 211 patients were identified for analysis. The median follow-up was 43 months for patients alive at the time of analysis. The median age was 63 years, and equal proportions had IIIA or IIIB disease. A total of 135 patients (64%) experienced disease relapse, and of these, 74% did so within 12 months. In those who did not experience relapse at 12 months, 16%, 6%, and < 5% expe-rienced relapse during 12 to 24, 24 to 36, and > 36 months of follow-up, respectively. In patients with relapse, 56% presented symptomatically, which led to inferior median OS compared to those identified by surveillance imaging (23 vs. 36 months; P ¼ .013). Conclusion: This study identified that most relapses occur within 1 year of completing CRT, and approximately half of these occur within 6 months. A symptomatic relapse led to inferior OS. More aggressive sur-veillance imaging may therefore identify asymptomatic relapses that are amenable to earlier salvage therapy.
Keywords: Follow-up, Surveillance, Symptoms
Nonesmall-cell lung cancer (NSCLC) is the leading cause of cancer-related death in the United States and worldwide.1-3 Approximately half of newly diagnosed NSCLC patients present
Department of Radiation Oncology, H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL
Address for correspondence: Thomas J. Dilling, MD, MS, Department of Radiation Oncology, H. Lee Moffitt Cancer Center and Research Institute, 12902 USF Magnolia Dr, Tampa, FL 33612 Fax: (813) 745-7231; e-mail contact: [email protected]
with locally advanced unresectable disease, which is primarily managed with definitive chemoradiotherapy (CRT).4 Previous prospective trials have established that platinum-based doublet therapy concurrent with radiotherapy (RT) provides superior out-comes compared to sequential regimens.5 Recently RTOG 0617, a randomized phase 3 trial, demonstrated a median overall survival (OS) of 28.7 months, a new survival benchmark for biliary system concurrent CRT approach.6 Yet despite advancements in multimodal ap-proaches, NSCLC continues to be an aggressive disease with a high risk of recurrence after CRT or surgery.
Disease relapse dynamics after definitive therapy for NSCLC have mostly been described in heterogeneous retrospective surgical series,7-9 which have informed the American College of Chest
Relapse Events After Chemoradiotherapy
Physicians (ACCP) recommendations for postoperative surveil-lance.10 The European Society for Medical Oncology (ESMO) guidelines for NSCLC currently recommend chest computed to-mography (CT) imaging biannually for those deemed to be suitable for salvage treatment; otherwise CT imaging is recommended at least once at 12 and 24 months.11 The National Comprehensive Cancer Network (NCCN) NSCLC guidelines previously recom-mended a similar follow-up schedule, but this was recently changed to a shorter surveillance interval of every 3 to 6 months for the first
3 years after definitive treatment.12 Although these recommenda-tions have been provided, there are no data from randomized trials and very few retrospective data that analyze posttreatment surveil-lance after definitive CRT.