Real-world evidence is increasingly interpreted alongside randomized clinical trial evidence in oncology, but reported outcomes may differ because of variation in patient selection, treatment composition, follow-up, ascertainment, documentation, and reporting. We examined whether outcomes differed between randomized trials and real-world studies of CDK4/6 inhibitors. We searched PubMed, Embase, Cochrane CENTRAL, and Web of Science through December 2025 for phase 2/3 randomized trials and real-world cohort, registry, or postmarketing studies of CDK4/6 inhibitors plus endocrine therapy for first- or second-line treatment of hormone receptor-positive, HER2-negative advanced or metastatic breast cancer. Arm proportions were pooled separately using logit-transformed random-effects models, and between-source differences were assessed using mixed-effects meta-regression. Eighty-six studies were included: 31 randomized trials (40 arms; 7464 patients) and 55 real-world studies (101 cohorts; 30 858 patients). Real-world cohorts included older patients and greater first-line and palbociclib use. After aggregate-level mixed-effects meta-regression adjustment, the model-based objective response rate was higher in real-world cohorts than in trial arms (51.8
To explore the theoretical framework of lung cancer caregivers’ skill model and lay a foundation for the follow-up practical research. Based on the grounded theory, a total of 23 participants (16 lung cancer caregivers and 7 health personnel of the oncology department) in the oncology department of a tertiary hospital in Shenzhen from April to May 2024 were recruited through purposive sampling for semi-structured interviews. The data were analyzed by coding and constant comparison. Six categories of the theme ‘care ability’ and 46 free nodes under it were extracted. The basic ability included ‘knowledge, skills’ as the prerequisite for care ability, and the expansion ability included ‘personal quality, the caregiver’s own physical and mental health, utilization of family and community resources, continuity of care’ as additional dimensions for care ability, which jointly contributed to the outcome of care. The lung cancer caregiver skill model developed in this study should be considered a preliminary, context-specific conceptual framework. Given the single-center design and the relatively small qualitative sample, it provides exploratory theoretical and practical guidance for improving the care ability of lung cancer caregivers and requires further validation in larger, multi-center, and cross-cultural studies.
Abstract Background The hemoglobin-to-red blood cell distribution width ratio (HRR) is an inexpensive index derived from routine blood tests, but its relationship with gastrointestinal (GI) cancers in the general population remains unclear. Methods We pooled NHANES 2001–2018 data (adults ≥ 20 years). GI cancer history was screened using MCQ220 and site-confirmed using site-specific follow-up items for colon, rectal, stomach, esophageal, liver, and pancreatic cancers. HRR was calculated as hemoglobin (g/dL) divided by RDW (%). Survey-weighted logistic regression estimated odds ratios (ORs) and 95% CIs for prevalent GI cancer across HRR (per 0.1-unit increase and tertiles), with restricted cubic splines and subgroup interaction tests. In parallel, eight machine-learning (ML) algorithms were trained and evaluated (70/30 split) to assess predictive performance (AUC, calibration, and decision-curve analysis), with SHAP for interpretability. Results Among 39,663 participants, 344 had site-confirmed GI cancer. Higher HRR was inversely associated with GI cancers in the fully adjusted model (per 0.1-unit increase: OR 0.86, 95% CI 0.77–0.95; P = 0.003). Participants in the highest HRR tertile had lower odds than the lowest tertile (OR 0.57, 95% CI 0.35–0.93; P = 0.026; P for trend = 0.028). Sex modified the association (P for interaction = 0.002), with an inverse association in men (OR 0.74) but not women. ML models showed similar discrimination (top AUC ≈ 0.82), and SHAP highlighted HRR as an important contributor alongside age. Conclusions Higher HRR was inversely associated with prevalent, site-confirmed GI cancers in NHANES. In machine-learning analyses, HRR functioned as an informative classification feature.
Pregnancy-associated breast cancer (PABC) presents unique clinical challenges. However, its clinicopathological characteristics and prognosis compared with non-PABC, as well as the associations between reproductive factors and molecular subtypes, remain incompletely understood. This single-center retrospective cohort study included 382 consecutive eligible women with stage I–III breast cancer, comprising 185 PABC and 197 non-PABC patients. Associations between reproductive factors and molecular subtypes were evaluated using multinomial logistic regression. Recurrence-free survival (RFS) was examined using Kaplan–Meier estimates and Cox proportional hazards models. After 1:1 propensity score matching (PSM), 129 patients were included in each group. PABC patients had higher histological grades, more advanced T stage, and a higher proportion of triple-negative breast cancer (TNBC; 18.9
Cell-free DNA is a promising source of biomarkers for early cancer detection. cfDNA fragmentation patterns and methylation profiles exhibit coordinated changes. Although several studies have explored cfDNA methylation or fragmentomics individually for breast cancer detection, systematic evaluation of integrated methylated fragmentomic features remains limited. Our study focused on identifying cfDNA methylation and methylated fragmentation markers for accurate breast cancer detection. This study included 25 breast cancer patients and 110 healthy controls. We performed comprehensive profiling of methylation and methylated fragmentation patterns of cfDNA using the cfMeDIP-seq method. Then, we used multiple algorithms to identify cfDNA methylation markers and methylated fragmentation markers, and developed a stacked model. Methylated fragment features were significantly different from non-methylated fragment features, with the proportion of ultra-short fragments in breast cancer patients was lower than that in non-cancer controls. We extracted 7 types of features from the methylation and methylated fragments, and developed a stacked machine-learning model that achieved an AUC of 1 in both the test and validation sets, indicating high sensitivity and specificity. Our study demonstrates that cfDNA methylation and methylated ultra-short fragmentomics biomarkers derived from the cfMeDIP-seq method can effectively diagnose breast cancer, indicating its potential clinical utility.
