No abstract

Background Lung adenocarcinoma (LUAD) is one of the most common types in the world with a high mortality rate. Despite advances in treatment strategies, the overall survival (OS) remains short. Our study aims to establish a reliable prognostic signature closely related to the survival of LUAD patients that can better predict prognosis and possibly help with individual monitoring of LUAD patients. Methods Raw RNA-sequencing data were obtained from Fudan University and used as a training group. Differentially expressed genes (DEGs) for the training group were screened. The univariate, least absolute shrinkage and selection operator (LASSO), and multivariate cox regression analysis were conducted to identify the candidate prognostic genes and construct the risk score model. Kaplan–Meier analysis, time-dependent receiver operating characteristic (ROC) curve were used to evaluate the prognostic power and performance of the signature. Moreover, The Cancer Genome Atlas (TCGA-LUAD) dataset was further used to validate the predictive ability of prognostic signature. Results A prognostic signature consisting of seven prognostic-related genes was constructed using the training group. The 7-gene prognostic signature significantly grouped patients in high and low-risk groups in terms of overall survival in the training cohort [hazard ratio, HR = 8.94, 95% confidence interval (95% CI)] [2.041–39.2]; P = 0.0004), and in the validation cohort (HR = 2.41, 95% CI [1.779–3.276]; P < 0.0001). Cox regression analysis (univariate and multivariate) demonstrated that the seven-gene signature is an independent prognostic biomarker for predicting the survival of LUAD patients. ROC curves revealed that the 7-gene prognostic signature achieved a good performance in training and validation groups (AUC = 0.91, AUC = 0.7 respectively) in predicting OS for LUAD patients. Furthermore, the stratified analysis of the signature showed another classification to predict the prognosis. Conclusion Our study suggested a new and reliable prognostic signature that has a significant implication in predicting overall survival for LUAD patients and may help with early diagnosis and making effective clinical decisions regarding potential individual treatment.

Limited and inefficient treatment options exist for metastatic relapsed cervical cancer (MRCC), and there are currently no reliable indicators to guide therapeutic selection. We performed deep sequencing analyses targeting 322 cancer-related genes in plasma cell-free DNA and matched white blood cells in 173 serial blood samples from 82 locally advanced CC (LACC) or MRCC patients and when possible during treatment. We identified five notable nonsynonymous mutant genes (PIK3CA, BRAF, GNA11, FBXW7 and CDH1) in the MRCC samples as the metastatic relapse significantly mutated (MSG) genes and found that MRCC patients with any detectable MSG mutations had significantly shorter progression-free survival (PFS) (P = .005) and overall survival (OS) (P = .007) times than those without detectable MSG mutations. Additionally, analyses of matched prechemotherapy and postchemotherapy plasma revealed that a reduction in the number of MSG mutations after chemotherapy was significantly associated with partial remission (PR) and stable disease (SD) (P = .007). Among the patients included in the longitudinal tracking ctDNA analysis, an increase in MSG mutations was observed earlier in response to disease progression than radiological imaging. Our results outline the mutation profiles of MRCC. We show how longitudinal monitoring with ctDNA in liquid biopsy samples provides both predictive and prognostic information during treatment.

No abstract

No abstract