Background: This study constructed and demonstrated a model to predict the overall survival (OS) of newly diagnosed distant metastatic cervical cancer (mCC) patients.Methods: The SEER (Surveillance, Epidemiology, and End Results) database was used to collect the eligible data, which from 2010 to 2016. Then these data were separated into training and validation cohorts (7:3) randomly. Cox regression analyses was used to identify parameters significantly correlated with OS. Harrell's Concordance index (C-index), calibration curves, and decision curve analysis (DCA) were further applied to verify the performance of this model.Results: A total of 2,091 eligible patients were enrolled and randomly split into training (n = 1,467) and validation (n = 624) cohorts. Multivariate analyses revealed that age, histology, T stage, tumor size, metastatic sites, local surgery, chemotherapy, and radiotherapy were independent prognostic parameters and were then used to build a nomogram for predicting 1 and 2-year OS. The C-index of training group and validation group was 0.714 and 0.707, respectively. The calibration curve demonstrated that the actual observation was in good agreement with the predicted results concluded by the nomogram model. Its clinical usefulness was further revealed by the DCAs. Based on the scores from the nomogram, a corresponding risk classification system was constructed. In the overall population, the median OS time was 23.0 months (95% confidence interval [CI], 20.5–25.5), 12.0 months (95% CI, 11.1–12.9), and 5.0 months (95% CI, 4.4–5.6), in the low-risk group, intermediate-risk group, and high-risk group, respectively.Conclusion: A novel nomogram and a risk classification system were established in this study, which purposed to predict the OS time with mCC patients. These tools could be applied to prognostic analysis and should be validated in future studies.

The combination of bone marrow-derived mesenchymal stem cells (BMSCs) and biological scaffolds has been demonstrated to be a promising strategy for bone regeneration. However, this method does not result in satisfactory bone regeneration, because the BMSCs are dispersed in the biological scaffolds. The current study developed a new bone regeneration system, which combines synthetic porous three-dimensional scaffolds of β-TCP/COL-I composite with cultured osteogenic sheets of BMSCs. Activity of alkaline phosphatase (ALP), a marker of bone regeneration, was assayed in vitro using enzyme-linked immunosorbent assays and quantitative real-time polymerase chain reaction. In vivo bone regeneration was assayed in male nude mice. The study samples were BMSC sheet, scaffold/scattered BMSCs, scaffold/BMSC sheet, and scaffold alone. The samples were implanted dorsally in the mice. In vitro analysis showed that β-TCP/COL-I scaffold combined with BMSC sheets significantly upregulated both gene expression and protein levels of ALP, osteocalcin, and osteopontin. Histological and micro-computed tomography showed that the only implants that demonstrated new bone formation after 4 weeks were scaffold/BMSC sheet implants. These results underscore the crucial requirement of a synergistic effect of β-TCP/COL-I scaffolds and BMSC sheets. This could be a promising novel strategy for bone tissue engineering. © 2017 Wiley Periodicals, Inc. J Biomed Mater Res Part B: Appl Biomater, 106B: 2037-2045, 2018.

This study evaluated the prognostic effects of nutritional risk scores and performance status (PS) on unresectable locally advanced esophageal cancer (LAEC) patients who were treated with definitive concurrent chemoradiotherapy (dCRT). A total of 202 LAEC patients from four different cancer centers were retrospectively reviewed. Nutritional risk and PS were measured using the Nutritional Risk Screening 2002 (NRS-2002) scores and Eastern Cooperative Oncology Group (ECOG) scales. Outcomes were clinical response rate, overall survival (OS) and progression-free survival (PFS). Multivariate analysis of predictive factors of response to dCRT and survival were performed using a logistic regression and a Cox model, respectively. The majority of patients (71.8%) had an ECOG PS score of 0-1, and 52.5% (n=106) of patients were identified as having nutritional risk (NRS-2002 ≥3) upon treatment initiation. There was no correlation between NRS-2002 scores and ECOG PS (Spearman's ρ=0.046; P=0.516). In multivariate analysis, NRS-2002 scores (P=0.002, HR 2.805, 95%CI: 1.445-5.446) and ECOG PS (P=0.015, HR 2.719, 95%CI: 1.218-6.067) were independent prognostic factors for the response to dCRT. NRS-2002 scores (OS: HR 1.530, 95%CI 1.059-2.209; P=0.023; PFS: HR 1.517, 95%CI 1.105-2.082; P=0.010) and ECOG PS (OS: HR 1.729, 95%CI 1.185-2.522; P=0.005; PFS: HR 1.678, 95%CI 1.179-2.387; P=0.004) were both independent prognostic factors for OS and PFS. In conclusions, NRS-2002 scores and ECOG PS scales both have prognostic effects on clinical response and survival in LAEC, but a significant association of NRS-2002 scores and ECOG PS were not observed.

No abstract

Background Prediction models with high accuracy rates for nonmetastatic cervical cancer (CC) patients are limited. This study aimed to construct and compare predictive models on the basis of machine learning (ML) algorithms for predicting the 5‐year survival status of CC patients through using the Surveillance, Epidemiology, and End Results public database of the National Cancer Institute. Methods The data registered from 2004 to 2016 were extracted and randomly divided into training and validation cohorts (8:2). The least absolute shrinkage and selection operator (LASSO) regression was employed to identify significant factors. Then, four predictive models were constructed, including logistic regression (LR), random forest (RF), support vector machine (SVM), and extreme gradient boosting (XGBoost). The predictive models were evaluated and compared using Receiver‐operating characteristics with areas under the curves (AUCs) and decision curve analysis (DCA), respectively. Results A total of 13,802 patients were involved and classified into training ( N = 11,041) and validation ( N = 2761) cohorts. By using the LASSO regression method, seven factors were identified. In the training cohort, the XGBoost model showed the best performance (AUC = 0.8400) compared to the other three models (all p < 0.05 by Delong's test). In the validation cohort, the XGBoost model also demonstrated a superior prediction ability (AUC = 0.8365) than LR and SVM models (both p < 0.05 by Delong's test), although the difference was not statistically significant between the XGBoost and the RF models ( p = 0.4251 by Delong's test). Based on the DCA results, the XGBoost model was also superior, and feature importance analysis indicated that the tumor stage was the most important variable among the seven factors. Conclusions The XGBoost model proved to be an effective algorithm with better prediction abilities. This model is proposed to support better decision‐making for nonmetastatic CC patients in the future.