The Chang Gung Research Database (CGRD) is a de-identified database derived from original medical records of Chang Gung Memorial Hospital (CGMH), which comprises seven medical institutes located from the northeast to southern regions of Taiwan. The volume of medical services performed in CGMH is large, and clinical and scientific studies based on the CGRD are reported to be of high quality. However, the CGRD as a useful database for research has not been analyzed before. The objective of the study was to analyze the CGRD with regard to its characteristics and coverage of Taiwan's population. We performed a nationwide cohort study using population-based data from the Taiwan National Health Insurance Research Database (NHIRD). All patients who had any medical record of outpatient visits or admission between January 1, 1997, and December 31, 2010, were included, and the sex ratio, age distribution, socioeconomic status, urbanicity, severity of illness, prevalence of specific disease, and coverage of the CGRD were analyzed. The sex ratio, age distribution, socioeconomic status, and urbanicity of the population of the CGRD are different from those of Taiwan NHIRD and medical centers in Taiwan (all the pairwise p < 0.05). The severity of comorbidities, and prevalence of specific diseases of the population of the CGRD are significantly higher than those of Taiwan NHIRD and medical centers in Taiwan for both outpatient and inpatient samples (all the pairwise p < 0.05). The overall coverage of the CGRD was 21.2% for outpatients and 12.4% for inpatients. The disease-specific coverage of the CGRD was 27-34% for outpatients and 14-21% for inpatients. The CGRD is a multi-institutional, original medical record-based research database with high overall and disease-specific coverage of Taiwan. The population of the CGRD has significantly higher severity of comorbidities, and prevalence of specific diseases than those of Taiwan NHIRD and medical centers in Taiwan.

Danshen (salvia miltiorrhiza Bunge) is widely used in traditional Chinese medicine. However, it is definite clinical effort and mechanism on breast cancer is unclear. In our study, we used the real-world database to investigate in vivo protective effort of danshen in the breast cancer patients through using population-based data from the Taiwan National Health Insurance Research Database (NHIRD). In vitro, human breast cancer cells (MCF-7 cells and MDA-MB-231 cells) were used to investigate the effect and the underlying mechanism through XTT assay, flow cytometry, glutathione peroxidase (GPX) activity assay, GSH (reduced glutathione)/GSSG (oxidized glutathione), malondialdehyde (MDA), and western blot analysis. The in vivo effect was investigated through a xenograft nude mouse model. We found that dihydroisotanshinone I (DT), a pure compound present in danshen, can inhibit the growth of breast carcinoma cells, including MCF-7 cells and MDA-MB-231 cells. Moreover, DT induced apoptosis and ferroptosis in these breast cancer cells. DT also repressed the protein expression of GPX4 (Glutathione peroxidase 4). For in vivo study, DT treatment also significantly inhibited the final tumor volume without adverse effects in a xenograft nude mouse model. In conclusion, danshen has protective efforts in breast cancer patients, which could be attributed to DT through inducing apoptosis and ferroptosis of breast cancer cells.

This study investigated preoperative neutrophil percentage-to-albumin ratio (NPAR) for predicting oral cavity squamous cell carcinoma (OSCC) survival. We retrospectively analyzed 368 patients who received curative OSCC surgery between 2008 and 2017. Receiver operating characteristic curve analyses were employed to identify the optimal NPAR cutoff (16.93), and the patients were then separated into low-NPAR and high-NPAR groups. Intergroup differences in survival were determined through Kaplan–Meier analysis and log-rank tests. Disease-free survival (DFS) and overall survival (OS) predictors were identified using Cox proportional-hazards models. A nomogram integrating independent prognostic factors was proposed to increase the accuracy of OS prediction. A high NPAR (≥16.93) was associated with worse median OS and DFS than was a low NPAR (both p < 0.001); this finding was confirmed through multivariate analyses (hazard ratio (HR) for OS = 2.697, p < 0.001; and HR for DFS = 1.671, p = 0.008). The nomogram’s favorable predictive ability was confirmed by the calibration plots and concordance index (0.784). The preoperative NPAR is thus a promising prognostic biomarker in patients with OSCC after external validation in a larger cohort. Our nomogram can facilitate clinical use of the NPAR and provides accurate individualized OS predictions.