Background: Over the recent years, machine learning methods have been increasingly explored in cancer prognosis because of the appearance of improved machine learning algorithms. These algorithms can use censored data for modeling, such as support vector machines for survival analysis and random survival forest (RSF). However, it is still debated whether traditional (Cox proportional hazard regression) or machine learning-based prognostic models have better predictive performance. Objective: This study aimed to compare the performance of breast cancer prognostic prediction models based on machine learning and Cox regression. Methods: This retrospective cohort study included all patients diagnosed with breast cancer and subsequently hospitalized in Fudan University Shanghai Cancer Center between January 1, 2008, and December 31, 2016. After all exclusions, a total of 22,176 cases with 21 features were eligible for model development. The data set was randomly split into a training set (15,523 cases, 70%) and a test set (6653 cases, 30%) for developing 4 models and predicting the overall survival of patients diagnosed with breast cancer. The discriminative ability of models was evaluated by the concordance index (C-index), the time-dependent area under the curve, and D-index; the calibration ability of models was evaluated by the Brier score. Results: The RSF model revealed the best discriminative performance among the 4 models with 3-year, 5-year, and 10-year time-dependent area under the curve of 0.857, 0.838, and 0.781, a D-index of 7.643 (95% CI 6.542, 8.930) and a C-index of 0.827 (95% CI 0.809, 0.845). The statistical difference of the C-index was tested, and the RSF model significantly outperformed the Cox-EN (elastic net) model (C-index 0.816, 95% CI 0.796, 0.836; P=.01), the Cox model (C-index 0.814, 95% CI 0.794, 0.835; P=.003), and the support vector machine model (C-index 0.812, 95% CI 0.793, 0.832; P<.001). The 4 models’ 3-year, 5-year, and 10-year Brier scores were very close, ranging from 0.027 to 0.094 and less than 0.1, which meant all models had good calibration. In the context of feature importance, elastic net and RSF both indicated that TNM staging, neoadjuvant therapy, number of lymph node metastases, age, and tumor diameter were the top 5 important features for predicting the prognosis of breast cancer. A final online tool was developed to predict the overall survival of patients with breast cancer. Conclusions: The RSF model slightly outperformed the other models on discriminative ability, revealing the potential of the RSF method as an effective approach to building prognostic prediction models in the context of survival analysis.

Obesity has been well studied in relation to breast cancer survival. However, the associations of post-diagnosis obesity and late outcomes (≥5 years after diagnosis) have been much less studied. A total of 4062 5-year disease-free patients were recruited from the Shanghai Breast Cancer Survival Study, a longitudinal study of patients diagnosed during 2002-2006. Cox proportional hazard model with restricted cubic spline were used to evaluate the potential non-linear associations of post-diagnosis body mass index (BMI) and waist-to-hip ratio (WHR) with late all-cause mortality and late recurrence. While no significant association was observed for post-diagnosis BMI or WHR with late recurrence; a U-shaped association was observed for the two measures with late all-cause death. Women with BMI of 25.0 kg/m² or WHR of 0.83 were at the lowest risk of late all-cause mortality, whereas those with BMI beyond the range of 22.1–28.7 kg/m² or WHR beyond the range of 0.81–0.86 had a higher risk. ER, stage or menopausal status did not modify the effect of post-diagnosis BMI or WHR on the outcomes. In conclusion, post-diagnosis BMI and WHR, as indicators of overall and central obesity respectively, were associated with late all-cause mortality in U-shaped pattern among long-term breast cancer survivors.

Background and Aim: A healthy body composition can improve the prognosis of breast cancer survivors. The study aimed to describe the body composition profile of breast cancer survivors and find out whether a short-term (3 months) wearable device-based lifestyle intervention had an effect on patients’ body weight and body composition. Methods: A before-and-after study was conducted on patients with stage I–III postoperative breast cancer, aged 18–70 years. Body composition was analyzed at baseline, and then patients went for a health education program. A wearable activity tracker and a goal of calorie consumption based on each individual’s weight were provided to each participant, and they were required to be equipped for 90 days. After 3 months, body composition was analyzed again. Results: Of 113 patients who completed the study, 65.49% showed a normal body mass index (BMI) at baseline assessment, 71.68% had a body fat percentage of more than 30%, and 41.59% had less skeleton muscle mass. During the intervention, the daily step count was 8,851.28 ± 2,399.31, and 59.21% reached the set goal calorie consumption. After a 3-month intervention, the patients had a significant reduction in body weight, fat mass, BMI, body fat percentage, and visceral fat area, but not in protein mass and skeleton muscle mass. Patients of different age, molecular classification, and therapy benefited from the intervention. Conclusion: Wearable technology with body composition analysis and health education for breast cancer survivors may help reduce weight and improve body composition even in a short time. Clinical Trial Registration: http://www.chictr.org.cn/showproj.aspx?proj=40672, identifier ChiCTR1900024258.

Intraoperative dexemdetomidine (DEX) with or without loading dose is well-established to improve postoperative pain control in patient-controlled analgesia (PCA). This study was designed to compare the pro-analgesia effect between the 2 in patients received general anesthesia. Seventy patients shceduced abdominal surgery under general anesthesia were randomly assigned into 3 groups which were maintained using propofol/remifentanil/Ringer solution (PRR), propofol/remifentanil/dexmedetomidine with (PRDw) or without (PRDo) a loading dose of dexmedetomidine before induction. PRDw/o patients displayed a greater Romsay sedation score measured immediately after surgery. When compared with PRR patients, those from the PRDw/o group had an increased time to first request of postoperative morphine and decreased 24 hours total morphine consumption. No significant difference was observed between patients from the PRDw and PRDo groups with respect to these parameters. The present study suggests that the administration of a DEX loading dose does not affect the pro-analgesic effect of intraoperative use of DEX on morphine-based PCA.

Evidence of the risk factors associated with early-onset colorectal neoplasm from prospective population-based studies is limited. We enrolled 17,293 participants younger than 50 years from the Shanghai colorectal cancer (CRC) screening program cohort. Face-to-face interviews were performed by trained primary care physicians using a standardized questionnaire to collect the information on potential risk factors at baseline entry. Furthermore, 124 cases of early-onset colorectal neoplasm, including six CRC cases and 118 colorectal adenoma (CRA) cases, were detected between 2012 and 2016. Multivariable logistic regression models and restricted cubic spline (RCS) were used to evaluate the risk factors associated with early-onset colorectal neoplasm. We found that sex, body mass index (BMI), and family history of CRC were associated with the early onset of colorectal neoplasm. The RCS model showed a positive dose–response and linear association between BMI and risk of early-onset colorectal neoplasm among young participants (p-overall = 0.19, p-nonlinear = 0.97). The findings indicated that it was beneficial for normal people younger than 50 years to start opportunistic CRC screening. As for those at high risk, increased surveillance is strongly recommended. Further close follow-up is required for research on the underlying causes of early-onset CRC.