RESUMEN
BACKGROUND: Metastatic colorectal cancer (mCRC) is a common malignancy whose treatment has been a clinical challenge. Cancer-specific survival (CSS) plays a crucial role in assessing patient prognosis and treatment outcomes. However, there is still limited research on the factors affecting CSS in mCRC patients and their correlation. AIM: To predict CSS, we developed a new nomogram model and risk grading system to classify risk levels in patients with mCRC. METHODS: Data were extracted from the United States Surveillance, Epidemiology, and End Results database from 2018 to 2023. All eligible patients were randomly divided into a training cohort and a validation cohort. The Cox proportional hazards model was used to investigate the independent risk factors for CSS. A new nomogram model was developed to predict CSS and was evaluated through internal and external validation. RESULTS: A multivariate Cox proportional risk model was used to identify independent risk factors for CSS. Then, new CSS columns were developed based on these factors. The consistency index (C-index) of the histogram was 0.718 (95%CI: 0.712-0.725), and that of the validation cohort was 0.722 (95%CI: 0.711-0.732), indicating good discrimination ability and better performance than tumor-node-metastasis staging (C-index: 0.712-0.732). For the training set, 0.533, 95%CI: 0.525-0.540; for the verification set, 0.524, 95%CI: 0.513-0.535. The calibration map and clinical decision curve showed good agreement and good potential clinical validity. The risk grading system divided all patients into three groups, and the Kaplan-Meier curve showed good stratification and differentiation of CSS between different groups. The median CSS times in the low-risk, medium-risk, and high-risk groups were 36 months (95%CI: 34.987-37.013), 18 months (95%CI: 17.273-18.727), and 5 months (95%CI: 4.503-5.497), respectively. CONCLUSION: Our study developed a new nomogram model to predict CSS in patients with synchronous mCRC. In addition, the risk-grading system helps to accurately assess patient prognosis and guide treatment.
RESUMEN
BACKGROUND: Somatic mutations are important biomarkers for selecting an optimal targeted therapy and predicting outcomes for non-small-cell lung cancer (NSCLC) patients that are often detected from tissue samples. However, tissue samples are not always readily available from these patients. The exploration of using circulating tumor DNA (ctDNA) to identify somatic mutations offers an alternative source that should be explored. METHODS: In this retrospective study, we included 280 patients diagnosed with adenocarcinoma between 2017 and 2018 in a hospital in eastern China. Tissue or ctDNA was collected, and a wide spectrum of somatic mutations was analyzed by targeted next-generation sequencing platforms. Associations among the mutation status, biomarkers, screening methods, disease stages, and interaction with treatment with overall survival (OS) were investigated. RESULTS: We found that the EGFR L858R mutation was the most frequently identified mutation in adenocarcinoma in this population by both methods, followed by KRAS (p=3.7e-09), PIK3CA (p=5e-04), and HER2 mutations (p=6.3e-03). We observed that EGFR mutations were significantly mutually exclusive with KRAS, HER2, and MET. FGFR1 mutations were significantly more abundantly detected in the ctDNA group. We found an interaction effect between EGFR mutation and target therapies. The ability of the targeted therapy to improve OS in patients with a single EGFR mutation (HR=0.069, p=0.07) approached significance, but this was not the case for the patients with more than one EGFR mutation or without an EGFR mutation (HR=0.813, p=0.725). Furthermore, the effect of chemotherapy was more predominant in the EGFR group in comparison to the control group. CONCLUSION: These findings provide useful information on the distribution of somatic mutations via different screening methods and how this related to the optimal treatment selection in Chinese patients with NSCLC.