Frontiers in Oncology (Jul 2022)

Risk Prediction Model for Synchronous Oligometastatic Non-Small Cell Lung Cancer: Thoracic Radiotherapy May Not Prolong Survival in High-Risk patients

  • Chunliu Meng,
  • Fang Wang,
  • Fang Wang,
  • Jia Tian,
  • Jia Wei,
  • Xue Li,
  • Kai Ren,
  • Liming Xu,
  • Lujun Zhao,
  • Ping Wang

DOI
https://doi.org/10.3389/fonc.2022.897329
Journal volume & issue
Vol. 12

Abstract

Read online

Background and PurposeOn the basis of the promising clinical study results, thoracic radiotherapy (TRT)1 has become an integral part of treatment of synchronous oligometastatic non–small cell lung cancer (SOM-NSCLC). However, some of them experienced rapid disease progression after TRT and showed no significant survival benefit. How to screen out such patients is a more concerned problem at present. In this study, we developed a risk-prediction model by screening hematological and clinical data of patients with SOM-NSCLC and identified patients who would not benefit from TRT.Materials and MethodsWe investigated patients with SOM-NSCLC between 2011 and 2019. A formula named Risk-Total was constructed using factors screened by LASSO-Cox regression analysis. Stabilized inverse probability treatment weight analysis was used to match the clinical characteristics between TRT and non-TRT groups. The primary endpoint was overall survival (OS).ResultsWe finally included 283 patients divided into two groups: 188 cases for the training cohort and 95 for the validation cohort. Ten prognostic factors included in the Risk-Total formula were age, N stage, T stage, adrenal metastasis, liver metastasis, sensitive mutation status, local treatment status to metastatic sites, systemic inflammatory index, CEA, and Cyfra211. Patients were divided into low- and high-risk groups based on risk scores, and TRT was found to have improved the OS of low-risk patients (46.4 vs. 31.7 months, P = 0.083; 34.1 vs. 25.9 months, P = 0.078) but not that of high-risk patients (14.9 vs. 11.7 months, P = 0.663; 19.4 vs. 18.6 months, P = 0.811) in the training and validation sets, respectively.ConclusionWe developed a prediction model to help identify patients with SOM-NSCLC who would not benefit from TRT, and TRT could not improve the survival of high-risk patients.

Keywords