The differentiation of benign and malignant breast nodules, particularly those categorized as Breast Imaging Reporting and Data System (BI-RADS) 3-4, remains a clinical challenge due to the subjectivity and operator dependence of conventional ultrasound assessment. This study aimed to evaluate the diagnostic value of intratumoral and peritumoral ultrasound radiomics features in distinguishing benign from malignant BI-RADS 3-4 breast nodules and to construct interpretable machine learning models.

Ultrasound images and parameters (BI-RADS classification) of breast nodules were retrospectively collected from female patients at two institutions between January 2021 and June 2024: 571 patients (880 nodules) from Institution 1 (Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine) and 333 patients (351 nodules) from Institution 2 (Shanghai Second People's Hospital). Included patients had BI-RADS 3-4 nodules confirmed by pathology or, for BI-RADS 3 nodules, stable findings on follow-up for at least 2 years. Nodules from Institution 1 were randomly divided into a training set (n = 615) and an internal validation set (n = 265) at a 7:3 ratio, while patients from Institution 2 constituted an external test set (n = 351). Radiomics features of intratumoral and peritumoral regions (3 and 5 pixels wide) were extracted using PyRadiomics. Features were screened using the independent t-test, Spearman correlation, and least absolute shrinkage and selection operator (LASSO) regression. Machine learning models-including random forest (RF), multilayer perceptron (MLP), extra trees (ET), support vector machine (SVM), logistic regression (LR), k-nearest neighbor (KNN), eXtreme Gradient Boosting (XGBoost), Adaptive Boosting (AdaBoost), Gradient Boosting Decision Tree (GBDT), and Light Gradient Boosting Machine (LightGBM)-were trained and compared using area under the curve (AUC) and other metrics. The best-performing model was then used to compare intratumoral versus peritumoral regions. SHapley Additive exPlanations (SHAP) was applied to interpret feature importance.

Across the training, internal validation, and external test sets, SVM demonstrated the most stable and balanced performance. The 3 pixels transitional zone region of interest support vector machine (EI3_SVM) model achieved a higher AUC than the tumor core region of interest (T) model in the external test set (0.875 vs. 0.787, p < 0.01), with accuracy 78.1%, sensitivity 42.2%, specificity 97.0%, and Brier score 0.166, outperforming other peritumoral models. Most models incorporating peritumoral features demonstrated AUCs that were higher than or comparable to those of the T model. SHAP analysis indicated that the high specificity was mainly driven by texture and shape features.

Integrating intratumoral and peritumoral radiomics features significantly improves the diagnostic accuracy and objectivity for differentiating benign and malignant breast nodules. This approach may aid clinical decision-making, reduce unnecessary biopsies, and support early precision diagnosis of breast cancer.

Patient-Centered Care (PCC)-that which addresses individual values, needs, and preferences-is a critical component of cancer care. Value-driven care requires evidence on outcomes that matter to patients, but Patient-Centered Outcomes (PCOs) are inconsistently defined and measured.

This scoping review maps and charts the literature on PCOs in gastrointestinal (GI) cancer care and the measures used to assess them with the aim of facilitating a shift toward more PCC.

We searched Medline, Embase, CINAHL, the Cochrane Library, and APA PsycINFO databases (2000-2025) to identify studies involving adult patients with GI cancers that reported or discussed at least one PCO, excluding survival. We summarized PCOs, measures used to assess them, and key study characteristics. Using qualitative cluster analyses, we then organized PCOs into a three-level hierarchy.

Of 1626 studies screened, 140 met inclusion criteria. Across these studies, we identified 187 PCOs and 286 measures. PCOs were grouped into six clusters: symptoms (27.4%), psychosocial (23.6%), lifestyle (23.1%), functional status (11.6%), care experience (8.5%), and healthcare utilization (5.8%). The most commonly mentioned PCO measures were three questionnaires: the EORTC QLQ-C30 (8.1%), FACT-G (4.2%), and the EQ-5D (2.5%).

A wide range of PCOs and corresponding measures have been reported in GI cancer care research, with symptom, psychosocial, and lifestyle related PCOs being the most frequently studied. Considerable heterogeneity exists in both the PCOs reported and the measures used to assess them. Identifying PCOs that matter most to patients and standardizing their measurement will be essential to advancing PCC.

Nevus-associated melanoma (NAM) is histologically defined by coexisting nevus and melanoma components, yet the nevus component is frequently not visible on dermoscopy. We performed a retrospective, single-centre observational study including histologically diagnosed NAMs (2011-2024). Dermoscopic images were assessed independently by two readers. Histopathology was systematically revised to quantify nevus proportion, size, and depth. Univariate and multivariable logistic regression evaluated associations with (a) invasive vs. in situ NAM and (b) dermoscopically visible vs. non-visible nevus component. Among 340 NAMs, 36.8% were in situ and 63.2% invasive. A dermoscopic nevus was visible in 45.6%. In multivariable analysis, nevus visibility under dermoscopy was independently associated with histopathologic features such as larger nevus size, particularly 2.1-4 mm (OR 2.5, 95% CI 1.3-4.6), 4.1-10 mm (OR 3.7, 95% CI 2.0-7.0), and > 10 mm (OR 5.3, 95% CI 1.4-20.0), whereas deep nevus location (OR 0.3, 95% CI 0.2-0.5) and ulceration (OR 0.2, 95% CI 0.1-0.6) reduced visibility. Invasive NAM was independently associated with a visible nevus component under dermoscopy (OR 2.5, 95% CI 1.7-3.7), shiny white structures (OR 5.0, 95% CI 2.6-9.4), negative pigment network (OR 1.9, 95% CI 1.2-3.2), and blue-white veil (OR 8.3, 95% CI 4.5-15.4), whereas atypical pigment network and melanoma pigmentation were inversely associated with invasion. Dermoscopic nevus visibility in NAM is mainly determined by nevus size and depth, while melanoma-related changes, particularly ulceration, can obscure the nevus component. Dermoscopic markers of invasiveness should prompt timely management even when an associated nevus is suspected.

The optimal choice between sedating histamine H₁ antagonists (sAHs) and non-sedating histamine H₁ antagonists (nsAHs) for preventing infusion-related reactions (IRRs) during daratumumab subcutaneous (DARA-SC) therapy remains unclear. Here, we aimed to compare the efficacy and safety of nsAH and sAH as premedication for DARA-SC therapy. Patients with multiple myeloma or light-chain amyloidosis who received DARA-SC therapy at eight hospitals were enrolled in this prospective, multicenter, observational study. Patients were categorized into nsAH and sAH groups based on their premedication. IRRs and drowsiness were assessed using patient-reported questionnaires, including the Stanford Sleepiness Scale (SSS) and the Japanese Epworth Sleepiness Scale (JESS), at baseline, post-administration (Q2), and before bedtime (Q3). Overall, 104 patients (nsAH, n = 49; sAH, n = 55) were analyzed. No significant differences were observed in the IRR incidence between the nsAH and sAH groups at Q2 (8.2 vs. 9.1%) or Q3 (13 vs. 17%). Conversely, the incidence of new-onset drowsiness at Q2 was significantly lower in the nsAH group (13%) than in the sAH group (32%, p < 0.05). Additionally, the increase in SSS from baseline to Q2 was significantly lower in the nsAH group than in the sAH group (p < 0.05). No significant differences were observed in JESS scores. Thus, nsAH was associated with reduced early post-administration drowsiness and there were no statistically significant differences in IRR preventive effects compared to that for sAH, suggesting that nsAH may reduce sedation without a significant increase in the incidence of IRRs; however, these hypothesis-generating findings warrant verification through future prospective comparative trials.

Given the limited Japanese real-world evidence, this study aimed to evaluate the survival and safety of adjuvant nivolumab after neoadjuvant chemotherapy in patients with an incomplete pathological response to esophageal squamous cell carcinoma (ESCC).

This single-center retrospective study included 105 of 259 patients with ESCC who had an incomplete pathological response after neoadjuvant chemotherapy and esophagectomy. Recurrence patterns and immune-related adverse events (irAEs) were analyzed, and disease-free survival (DFS) and overall survival (OS) were compared between the nivolumab and non-nivolumab groups before and after propensity score matching (PSM).

Among 105 patients, 17 received adjuvant nivolumab. Median follow-up was 26.8 months; recurrence occurred in 33 patients (31.4%), and irAEs occurred in 2 nivolumab-treated patients (11.8%). Before PSM, OS and DFS were not significantly different between 2 groups. After PSM, OS remained nonsignificant (P = 0.202), whereas DFS showed a favorable trend toward improvement in the nivolumab group, with 1-year DFS rates of 75.0% versus 68.8% and 3-year DFS rates of 68.2% versus 19.3%, respectively (P = 0.067). Nivolumab was not an independent prognostic factor, although exploratory high-risk subgroups showed favorable DFS trends.

Adjuvant nivolumab showed a nonsignificant favorable trend in DFS after PSM, with late curve separation favoring the nivolumab group and manageable irAEs.

Dedicated service for Barrett's oesophagus (BO) surveillance may be more effective than conventional service, according to some single centre studies.

To determine whether the surveillance of BO through a dedicated service can improve key performance indicators (KPIs), and dysplasia detection rate (DDR) compared with conventional service in a multi centre study.

A retrospective cohort study was conducted across 6 NHS-hospitals, capturing BO surveillance data over 8 years. Factors associated with DDR were assessed by logistic regression.

There were 1037 dedicated and 976 conventional surveillance procedures (N = 2013), male: female ratio = 2.2:1; mean age = 64.4 (SD ± 12.2) years; mean maximum length of BO = 4.1 cm (range: 0-18 cm). All the KPIs and DDR were significantly higher in the dedicated service (DDR = 6.9%) than in the conventional service (DDR = 2.8%), p < 0.001. The use of narrow band imaging (NBI) and acetic acid chromoendoscopy (AAC) and sedation were high, and the complication rate was significantly lower in the dedicated service. The lesion recognition was significantly associated with DDR (OR = 6.9). Surprisingly, Seattle biopsy protocol adherence showed no correlation with DDR (OR = 0.47).

The dedicated service provides higher quality endoscopy, and a higher yield of early neoplasia. It uses more sedation, advanced imaging, and detects more lesions than conventional services. Thus, the dedicated surveillance could potentially replace time and cost consuming Seattle biopsies with targeted biopsies of visible lesions. In future, this may become easier with the use of artificial intelligence for lesion detection (CADe).

Active surveillance strategies have evolved to incorporate new technologies, such as MRI. However, trends in the adoption of this technology within the first year after cancer diagnosis as part of the initial active surveillance activity remain unclear.

Using SEER-Medicare data, this cross-sectional study examined trends in the use of various active surveillance strategies, including MRI associated biopsy, within the first year of cancer diagnosis among men diagnosed with localized prostate cancer between 2010-2011 and 2018-2019. Patients were stratified by Gleason score, age, and race/ethnicity.

Among patients without active treatment in the first year after localized prostate cancer diagnosis, the unadjusted proportion undergoing MRI associated biopsy as an active surveillance strategy increased from 0.8% in 2010-2011 to 3.16% in 2018-2019 for GS ≤ 6, and from 0.7% to 4.69% over the same period for GS ≥ 7. Patients aged 66-74 had consistently higher probabilities of undergoing MRI associated biopsy than patients aged 75+, and non-Hispanic White patients had consistently higher probabilities of undergoing MRI associated biopsy than non-Hispanic Black.

The use of new active surveillance strategies, such as MRI associated biopsy, were low in overall use but increased over time from 2010-2011 to 2018-2019, with notable differences across patient subgroups defined by age and race/ethnicity.

Avapritinib is a selective tyrosine kinase inhibitor approved for advanced PDGFRA exon 18-mutant gastrointestinal stromal tumors (GIST), including the imatinib-resistant D842V variant. However, its role in the preoperative setting for localized, resectable disease has not been defined.

We conducted a retrospective two-center exploratory case series of eight patients with localized, resectable PDGFRA-mutant gastric GIST who received preoperative avapritinib between 2020 and 2025. Radiographic and pathologic responses, treatment duration, adverse events (AEs), and surgical outcomes were evaluated.

All patients initiated avapritinib at 300 mg daily. Median treatment duration was 2 months (range, < 1-21 months). Tumor size decreased in seven patients (88%), with a median reduction of 25% and an overall range from -51% to +29%. Three patients (38%) met RECIST-based thresholds for partial response (≥ 30% reduction). Five patients underwent surgical resection, all achieving complete (0% viable tumor) or near-complete (≤ 5% viable tumor) pathologic response. Dose reductions were required in three patients due to toxicity; however, tumor regression continued in all. Treatment-related grade ≥ 3 AEs occurred in 50% of patients, including one fatal intracranial hemorrhage. Most toxicities emerged within two months, and most nonfatal toxicities improved with dose adjustment or discontinuation.

In this exploratory two-center case series, preoperative avapritinib demonstrated marked pathologic activity in localized PDGFRA-mutant GIST, including after dose reduction in some patients. However, early toxicity at the standard 300 mg dose was frequent and, in several cases, prevented patients from reaching planned surgery. These findings argue against routine preoperative avapritinib use and support limiting this approach to clinical trials or highly selected patients with a compelling surgical rationale until prospective data are available.

Proteolysis-targeting chimeras (PROTACs) are heterobifunctional molecules that hijack the ubiquitin-proteasome system to drive catalytic, sub-stoichiometric degradation of disease-associated proteins, offering a mechanistic advantage over occupancy-driven inhibitors and access to 'undruggable' targets. However, their clinical translation is constrained by high molecular weight, poor solubility, low oral bioavailability, inefficient membrane permeability, nonspecific biodistribution, off-target degradation, and the concentration-dependent 'hook effect.' Exosomes, nanoscale extracellular vesicles with innate biocompatibility, low immunogenicity, prolonged circulation, and the ability to cross barriers such as the blood-brain barrier, offer a biologically integrated platform to overcome these limitations. This review traces the evolution of PROTAC technology, delineates the challenges of conventional delivery, and evaluates the rationale for exosomal encapsulation, including cargo protection, intracellular trafficking, endosomal escape, and release kinetics. We examine natural and engineered exosomes spanning source selection, active loading strategies, and surface functionalization for tissue-specific homing and synthesize therapeutic applications across viral infections, cancer, neurodegenerative disorders, and inflammatory diseases. Proof-of-concept studies, such as camel milk-derived exosomes delivering the BRD4-targeting PROTAC ARV-825, demonstrate enhanced permeability, lower IC₅₀ values, and improved oral bioavailability. Finally, we discuss key hurdles to clinical translation: scalable production, purification, and standardization, and outline future directions for exosome-mediated targeted protein degradation.

Scalable, non-invasive tools are critically needed to improve early lung cancer detection and optimize primary care referral pathways. We evaluated Inflammacheck, a point-of-care device utilizing exhaled breath condensate (EBC) H₂O₂ and physiological parameters with machine learning for non-invasive lung cancer detection in a real-world screening population. Exhaled Hydrogen Peroxide for Early Lung Cancer Detection (ExPeL) study participants, from the UK Targeted Lung Health Check (TLHC) programme, included individuals with suspected lung cancer and low-risk ever-smoker controls. EBC was collected via Inflammacheck, measuring H₂O₂, end-tidal CO₂, humidity, temperature, and exhalation flow rate. Multivariate analyses (PCA, LDA and Mahalanobis distance) assessed intrinsic group separation. SMOTE-balanced data trained supervised machine learning models (stacked and voting ensembles), which were then evaluated on held-out test sets. In parallel, untargeted LC-MS metabolomics was performed to identify discriminatory molecular features. Analyzing 34 participants with valid EBC data, 83% of cancer cases were early-stage (I-II), reflecting a screening population. Multivariate analysis clearly separated lung cancer and controls across PCA, LDA, and Mahalanobis mapping. The voting ensemble model achieved: Accuracy 85.7%, Sensitivity 80%, Specificity 100%, Precision (PPV) 100%, ROC-AUC 0.90 and MCC 0.73. Crucially, no false positives were identified. EBC variables revealed greater dispersion in cancer patients, reflecting physiological heterogeneity missed by univariate analysis. Untargeted metabolomics identified 2132 features, with four key metabolites yielding an AUC of 0.969 for cancer discrimination. Inflammacheck effectively distinguishes early-stage lung cancer via a rapid, non-invasive breath test, findings which are highly relevant for primary care and screening triage, where non-specific symptoms and low prevalence pose challenges.