Medicine and Pharmacology

Abstract: Pericytes are specialized mural cells that closely interact with endothelial cells to regulate vascular development, stability, and homeostasis through signaling pathways including PDGF-B/PDGFR-β, TGF-β, Angiopoietin-1/Tie2, and N-cadherin. In cancer, they display remarkable phenotypic and functional plasticity, becoming key regulators of tumor angiogenesis and the tumor microenvironment. Disrupted pericyte-endothelial interactions contribute to the formation of abnormal, highly permeable blood vessels, promoting hypoxia, tumor progression, immune evasion, and metastasis. Beyond their vascular role, pericytes actively communicate with tumor-associated macrophages, cancer-associated fibroblasts, and extracellular matrix components, thereby influencing epithelial-to-mesenchymal transition, pre-metastatic niche formation, and therapeutic resistance. Increasing evidence also suggests that pericytes undergo phenotypic transitions that further support tumor progression. Conversely, restoring pericyte function and promoting vessel normalization have emerged as promising strategies to improve vascular integrity, tissue oxygenation, drug delivery, and the efficacy of anti-angiogenic therapies and immunotherapy. Preclinical and early clinical studies indicate that targeting pathways regulating pericyte recruitment and function, particularly in combination with endothelial-directed therapies, may produce synergistic anti-tumor effects. This review highlights the multifaceted roles of pericytes in tumor vascular remodeling, immune regulation, metastasis, and their potential as therapeutic targets in cancer.

Abstract: Background: Leiomyosarcoma (LMS) is a rare aggressive sarcoma with substantial heterogeneity in anatomical presentation, stage, treatment, and survival. We evaluated clinical characteristics, treatment patterns, and survival, with particular emphasis on whether anatomical site independently predicted overall survival (OS). Methods: We retrospectively reviewed 240 adults with histologically confirmed LMS diagnosed between January 2015 and May 2026. OS was calculated from diagnosis to death or last contact, with administrative censoring on July 1, 2026. Kaplan–Meier and Cox proportional-hazards methods were used to evaluate age, sex, anatomical site, histologic grade, and disease stage. Multivariable models adjusted for age, sex, grade, and stage assessed the independent association of anatomical site with OS. Results: Median age was 56 years, and 152 patients (63.3%) were female. Disease was localized in 120 (50.0%), de novo metastatic in 53 (22.1%), recurrent metastatic in 35 (14.6%), and recurrent localized in 12 (5.0%). Median OS was 37.0 months. Localized disease was associated with substantially longer OS than metastatic disease (61.0 vs 13.0 months; P<.001). Grade 3 disease was associated with shorter OS than grade 1/2 disease (25.6 vs 61.1 months; P<.001). In multivariable analyses, metastatic disease remained strongly associated with inferior OS (HR, 5.59–5.94; P<.001), as did age ≥50 years (HR, 1.86–1.92; P=.006–.004) and grade 3 histology (HR, 1.74–1.92; P=.018–.007). Anatomical site was not independently associated with OS (extremity/trunk vs non-extremity: HR, 1.60; P=.063; non-uterine vs uterine: HR, 1.23; P=.41). Conclusion: Disease stage was the strongest prognostic factor, while older age and grade 3 histology were independently associated with poorer OS. Although extremity/trunk LMS showed better univariable survival, anatomical site was not independently associated with OS after adjustment.

Abstract: Triple-negative breast cancer (TNBC) remains a major therapeutic challenge, largely due to the persistence of drug-tolerant breast cancer stem cell (BCSC) populations that can with-stand cytotoxic stress and drive disease relapse. Although interleukin-1 recep-tor-associated kinase 2 (IRAK2) has been implicated in inflammatory signaling and en-do-plasmic reticulum stress responses, its functional role in the development of acquired chemotherapy resistance in BCSCs has not been fully elucidated. In this study, we inves-tigated whether IRAK2 contributes to the maintenance of a drug-tolerant phenotype in BCSC1, MDA-MB-231, and MDA-MB-468 cells. Exposure to chemotherapeutic agents in-duced IRAK2 mRNA expression in most drug-cell line combinations, while BCSC1 cells exhibited a functionally resistant phenotype following 5-fluorouracil (5-FU) treatment. Stable or doxycycline-inducible IRAK2 knockdown significantly impaired cell prolifera-tion, prevented post-treatment recovery after 5-FU exposure, reduced sphere-forming ca-pacity, and attenuated the 5-FU-induced expression of the drug transporters ABCC3 and ABCC4. Collectively, these findings identify IRAK2 as a key mediator of stress-adaptive signaling that promotes chemotherapy tolerance and stemness-associated survival mech-anisms in TNBC-derived BCSCs. IRAK2 therefore represents both a potential biomarker of emerging chemoresistance and a promising therapeutic target for combination strategies aimed at eradicating drug-tolerant cancer stem cell populations.

Abstract: Background: Appendiceal epithelial lesions are rare and heterogeneous. Evidence regarding vascular endothelial growth factor A (VEGF-A) expression in localized appendiceal lesions is limited. We compared VEGF-A immunoreactivity across diagnostic subtypes and explored associations with histological features. Methods: This retrospective exploratory study included 73 histologically confirmed appendiceal lesions. Formalin-fixed, paraffin-embedded tissue was assessed for predominant VEGF-A staining intensity and percentage of positive neoplastic epithelial cells. Their product was reported as a modified immunoreactivity score (0-300). Permutation-based nonparametric tests, multiplicity correction, and diagnosis-adjusted sensitivity analyses were used. Results: VEGF-A staining was present in 71/73 cases. Median immunoreactivity scores were 205 for adenocarcinoma, 170 for hyperplastic polyp, 180 for low-grade appendiceal mucinous neoplasm, and 100 for sessile serrated lesion; the overall difference was not significant (p = 0.075). No histological-feature association remained significant after false-discovery-rate correction. Surface nuclear stratification was not significant in the global analysis (p = 0.051; q = 0.169), pairwise analysis, within-subtype analyses, or diagnosis-adjusted sensitivity analysis (p = 0.278). VEGF-A measures did not differ between acute and incidental presentations; patients with acute presentations were younger (p = 0.007). Conclusions: VEGF-A immunoreactivity was common, but no robust subtype-specific or histology-specific pattern was identified. These exploratory findings do not establish clinical utility and require confirmation in larger controlled cohorts.

Abstract: Background/Objectives: Castration-resistant prostate cancer (CRPC) is an aggressive and treatment-resistant stage of prostate cancer for which reliable molecular biomarkers and therapeutic targets remain limited. Previous CRISPR-Cas9-mediated knockout of androgen receptor (AR) and fatty acid-binding protein 5 (FABP5) identified a shared transcriptomic signature, suggesting the existence of a common regulatory network. This study aimed to clinically validate this network using multiple independent prostate cancer datasets. Methods: Five commonly dysregulated genes (FASN, FOSB, GRPR, PPARG, and CAV1), identified through comparative transcriptomic analyses, together with FABP5, were evaluated using multiple prostate cancer databases. Gene expression, clinicopathological associations, disease progression, genomic alterations, protein expression, and functional enrichment analyses were performed using integrated bioinformatics approaches. Correlation analyses with AR, KLK3, and VEGFA were also conducted. Results: FABP5, FASN, FOSB, and GRPR were significantly upregulated in prostate adenocarcinoma, whereas PPARG and CAV1 were significantly downregulated relative to normal prostate tissues. Several genes were significantly associated with Gleason score, disease progression, and established prostate cancer biomarkers. Protein-level validation using Human Protein Atlas data showed increased FABP5 and FASN expression in high-grade prostate adenocarcinoma tissues. Functional enrichment analyses identified significant enrichment of lipid metabolism, PPAR signaling, fatty acid biosynthesis, and hormone-responsive pathways. Conclusions: These findings clinically validate an experimentally identified FABP5-associated molecular network and support its role in prostate cancer progression and castration resistance. Collectively, the results identify FABP5 and its associated lipid metabolism-related genes as promising biomarkers and potential therapeutic targets for advanced prostate cancer.

Abstract: Background/Objectives: Non-clear cell renal cell carcinoma (nccRCC) comprises biologically heterogeneous renal malignancies with distinct molecular drivers, clinical behavior, and therapeutic responses. This review summarizes prognosis and systemic therapy and evaluates emerging histology-specific and biomarker-directed approaches. Methods: We analyzed overall survival by histology in the SEER 17 Registries database (2000–2022) and conducted a narrative review of trials, guidelines, and translational studies. The ICD-O-3-defined population included 43,966 adults; 43,789 with known survival time were included in survival analyses. Results: Median follow-up was 96 months. Compared with papillary RCC, adjusted HRs were 0.67 (95% CI 0.64–0.70) for chromophobe RCC, 3.73 (3.37–4.11) for collecting duct carcinoma, 0.97 (0.80–1.17) for oncocytic tumors, and 33.23 (27.47–40.20) for medullary carcinoma. VEGFR- and MET-directed therapies and immune checkpoint inhibitor combinations show activity that varies by histology. Conclusions: nccRCC should be studied and managed as a collection of distinct entities. Registry associations require cautious interpretation because of residual confounding, evolving classification, and non-proportional hazards; subtype-specific trials and molecular stratification remain priorities.

Abstract: Background: Sexual dysfunction is a common but frequently underrecognized consequence of gynecological cancer treatment. It may persist throughout survivorship and involve vaginal symptoms, dyspareunia, pelvic-floor dysfunction, hormonal changes, psychological distress, altered body image, and relationship difficulties. Despite its substantial impact on quality of life, sexual health is still inconsistently addressed in routine follow-up care. Objective: This narrative review evaluates contemporary treatment strategies for sexual dysfunction in gynecological cancer survivors and proposes a personalized framework for multidisciplinary oncosexual rehabilitation. Methods: A narrative literature search was conducted in PubMed/MEDLINE, Scopus, and Google Scholar, focusing primarily on publications from January 2016 to September 2026. Randomized controlled trials, observational and qualitative studies, systematic reviews, meta-analyses, clinical guidelines, consensus statements, and relevant narrative reviews were considered. Earlier landmark studies were retained when clinically relevant. Results: Current evidence supports a multidimensional approach to treatment. Management should begin with structured assessment of sexual symptoms, cancer and treatment history, vaginal and hormonal status, pelvic-floor function, psychological well-being, body image, relationships, and patient-defined goals. Therapeutic options include vaginal moisturizers and lubricants, selected local hormonal treatment, vaginal dilators, pelvic-floor rehabilitation, psychosexual and psychological interventions, and multidisciplinary survivorship care. Evidence for individual interventions varies, and emerging approaches such as vaginal laser therapy require further evaluation. The available literature supports combining interventions according to the predominant mechanisms and individual needs. Conclusions: Sexual dysfunction should not be regarded as an unavoidable consequence of gynecological cancer survivorship. Personalized, mechanism-based treatment integrated with multidisciplinary oncosexual rehabilitation offers a clinically meaningful approach to restoring sexual comfort, intimacy, confidence, and quality of life.

Abstract: Background: Fellowship websites are an important source of information for prospective radiology applicants, but the online transparency of oncologic imaging fellowship programs has not been systematically evaluated. Purpose: To assess applicant-relevant information on oncologic imaging fellowship websites in USA and Canada and determine whether website completeness differs by institutional characteristics. Materials and Methods: In this cross-sectional study, oncologic imaging fellowship programs were identified through structured manual web searches. Eligible programs offered at least 1 year of radiology fellowship training with oncologic imaging as a principal component and had an active official webpage. Websites were assessed for 24 prespecified criteria across four domains. Each item was coded as present or absent, and overall and domain completeness scores were calculated. Overall completeness was compared between dedicated cancer centers and other institutions using the Mann–Whitney U test. Results: Twenty programs were included: 18 in the United States and two in Canada; 13 were based at dedicated cancer centers. Median overall completeness was 70.8% (interquartile range, 59.4%–79.2%). Program and application information had the highest median completeness (90%), whereas people, outcomes, and employment had the lowest (25%). Alumni or career placement information was available on four websites (20%), current fellows on five (25%), and teaching opportunities on six (30%). Dedicated cancer-center programs had greater overall completeness than other institutions (79.2% vs 62.5%; P = .028). Conclusion: Oncologic imaging fellowship websites generally provide strong program and application information but frequently omit trainee, alumni, employment, and teaching information. Dedicated cancer-center programs had more complete websites.

Abstract: Fusion-positive rhabdomyosarcoma (FP-RMS) is driven by fusion transcription factors but retains an incomplete myogenic program. The roles of the two mammalian ISWI ATPases, SMARCA1/SNF2L and SMARCA5/SNF2H, remain poorly defined. This article proposes that they support FP-RMS through distinct but connected mechanisms. In RH30 cells, SMARCA1 loss markedly suppressed PAX3::FOXO1 and fusion-associated outputs, disrupted EGR1/Wnt- and TGF-β-linked chromatin programs, impaired differentiation and three-dimensional organization, and increased MEK-inhibitor sensitivity. SMARCA1 loss was also accompanied by marked depletion of MYCN and reductions in FUS and EP300, factors with established or emerging links to intrinsically disordered regions, transcriptional hubs, or biomolecular condensates. Together with recent evidence for PAX3::FOXO1/N-Myc transcriptional condensates, these observations motivate the hypothesis that SMARCA1-dependent chromatin competence may indirectly support the molecular environment required for fusion-associated hub organization, without implying that SMARCA1 itself nucleates condensates. By contrast, SMARCA5 was more closely associated with PI3K/mTOR signaling, cell-cycle progression, proliferation, and survival. In RH28 cells, strong SMARCA5 depletion with retained SMARCA1 reduced viability while preserving Wnt-responsive differentiation competence. Independent proximity labeling also detected SMARCA5 near selected FP-RMS fusion proteins, raising a context-specific possibility that SMARCA5 participates in selected fusion-associated molecular environments without establishing a condensate-scaffolding role. Single-cell analysis of 70,600 malignant cells from 16 RMS tumors further showed fusion-context-dependent associations of both ATPases with a high-risk transcriptional program. Together, these observations support a division-of-labor hypothesis in which SMARCA1 maintains chromatin competence for fusion-dependent and plastic cell states and may indirectly regulate fusion-associated transcriptional hubs, whereas SMARCA5 primarily supports proliferative fitness and may contribute to selected fusion-associated molecular environments. Matched perturbation, rescue, chromatin, protein, hub-imaging, single-cell, and three-dimensional studies are required to test this model.

Abstract: Purpose: Gallbladder cancer is highly prevalent in North India and associated with poor survival in advanced disease. Although chemoimmunotherapy improves outcomes in biliary tract cancers, prospective data specific to gallbladder cancer remain limited. We compared gemcitabine-cisplatin (GemCis) with or without durvalumab in patients with advanced gallbladder cancer. Methods: In this ambispective observational cohort study at a North Indian tertiary cancer center (Jan 2023–June 2025), 98 patients with unresectable/metastatic gallbladder cancer received GemCis alone (n=48) or GemCis+durvalumab (n=50) based on drug availability and patient preference after counseling. Primary endpoints were progression-free survival (PFS) and overall survival (OS). Secondary endpoints included objective response rate (ORR), disease control rate (DCR), and safety. Results: Baseline characteristics were balanced. ORR was 30.0% with durvalumab vs 18.8% with GemCis alone; DCR was 56.0% vs 39.6%. Median PFS was 6.1 vs 4.1 months (HR 0.62, 95% CI 0.45–0.88; P=0.007). Median OS was 14.2 vs 9.7 months (HR 0.69, 95% CI 0.52–0.91; P=0.012). Grade ≥3 adverse events were similar between arms. Immune-related adverse events occurred in 20% receiving durvalumab (grade ≥3: 2%), all manageable. Conclusions: In this gallbladder cancer-specific cohort, durvalumab added to GemCis was associated with improved PFS and OS without increased toxicity. These findings support further evaluation of chemoimmunotherapy for advanced gallbladder cancer in high-incidence regions.

Abstract: Breast histopathology classification is commonly evaluated at the image level even though diagnosis and clinical decisions are made at the patient level. This mismatch can inflate apparent performance when images from the same patient are treated as independent observations and can obscure whether model confidence is reliable across magnifications. We present a reliability-aware framework for patient-level classification on the BreakHis dataset. Four ImageNet-pretrained backbones (ResNet50, EfficientNet-B0, DenseNet121, and Swin-Small) were evaluated using a patient-disjoint split (58/13/11 patients for training/validation/test). ResNet50 and Swin-Small were selected for calibration; temperature scaling was estimated on the validation set, and seven nonlearned aggregation rules combined image probabilities within patients. A compact attention network then learned patient-level weights from calibrated probability, entropy, confidence, magnification, cross-magnification variability, and inter-model disagreement. In an exploratory fixed test set of 11 patients (3 benign, 8 malignant), all four image models produced 0.909 accuracy and 0.833 balanced accuracy after simple patient-level mean aggregation. The learned reliability-aware aggregator reached 1.000 accuracy and balanced accuracy on this small test set as a proof-of-concept result, while its probability-based metrics remained modest (Brier score 0.239; negative log-likelihood 0.671). These findings support the feasibility of reliability-aware patient aggregation but do not establish generalization, because confirmatory patient-aware cross-validation has not yet been completed. The principal contribution is a leakage-conscious evaluation and aggregation workflow, with final comparative claims reserved for the planned cross-validation analysis.

Abstract: Podocalyxin (PODXL) is a potential diagnostic biomarker and therapeutic target in various tumors. Development of anti-PODXL mAbs has been limited by concerns about potential reactivity with normal tissues. To minimize on-target, off-tumor toxicities and adverse effects, we developed cancer-specific mAbs (CasMabs) against PODXL and demonstrated antitumor efficacy in human tumor xenograft models. In this study, we evaluated humPcMab-60-DXd, an anti-PODXL CasMab conjugated with DXd, a deriv-ative of the topoisomerase I inhibitor exatecan. humPcMab-60-DXd retained reactivity with the pancreatic ductal adenocarcinoma cell line MIA PaCa-2 and inhibited cell pro-liferation in vitro. In the MIA PaCa-2 xenograft model, humPcMab-60-DXd showed su-perior antitumor efficacy compared with humPcMab-60. Furthermore, humP-cMab-60-DXd exhibited dose-dependent antitumor activity without reducing mice body weight. These results provide the first observation of antitumor efficacy for an an-ti-PODXL CasMab-drug conjugate, which could facilitate the clinical development of CasMabs for tumor therapy.

Abstract: Background/Objectives: Indocyanine green fluorescence angiography (ICG-FA) images bowel perfusion during colorectal cancer resection. Randomised trials have not established a consistent reduction in anastomotic leak (AL), so attention has shifted to how the test should be deployed. We examined the determinants of ICG-driven intraoperative decision change. Methods: Retrospective analysis of 379 consecutive patients (273 colorectal cancer, 106 benign disease) undergoing elective colorectal resection with primary anastomosis and intraoperative ICG-FA at six Spanish centres. The outcome was an ICG-prompted change of the intended transection line. Associations were examined by logistic regression, with permutation testing and case-mix adjustment for between-centre comparisons. Results: ICG-FA changed the operative plan in 46 patients (12.1%), identically in cancer and benign disease (12.1% versus 12.3%). No patient-level preoperative characteristic was associated with decision change (all ten candidates p &gt; 0.20); odds ratios for malignant diagnosis, rectal location and neoadjuvant therapy were within 0.03 of unity. Two factors did predict it: left-sided or rectal resection versus right hemicolectomy (15.0% versus 4.3%; OR 3.96, 95% CI 1.38-11.37) and treating centre (0-18.1%; p = 0.021), both persisting in a mutually adjusted model (adjusted OR 6.75, 2.21-20.60 and 5.25, 2.38-11.58). AL occurred in 25 patients (6.6%); neoadjuvant therapy (adjusted OR 3.73) and rectal location (2.64) were independent predictors, decision change was not (1.88, 0.65-5.41). The higher crude leak rate after revision (10.9% versus 6.0%) reflects confounding by indication and must not be read as harm. Conclusions: Whether ICG-FA changes a colorectal cancer operation is determined by anatomical segment and treating centre, not by the patient's preoperative risk profile. Patient-level selection is therefore not feasible whereas segment-based targeting is, and the between-centre variation indicates that the intervention as delivered is not uniform.

Abstract: The early and accurate detection of cancer is crucial for improving survival outcomes and reducing treatment costs. Traditional diagnostic tools such as imaging and biopsies are often invasive, expensive, or impractical for widespread screening. Electronic Nose (E-nose) technology, designed to mimic the human olfactory system, has emerged as a promising non-invasive approach for detecting volatile organic compounds (VOCs) associated with cancer metabolism. This review provides a comprehensive overview of E-nose sensor technologies, including metal oxide semiconductors, conducting polymers, quartz crystal microbalance, surface acoustic wave, colorimetric, and electrochemical sensors. We explore the biological mechanisms of VOC production and their diagnostic relevance in exhaled breath, urine, and other bodily fluids. The integration of machine learning (ML) and deep learning (DL) techniques is discussed in detail, highlighting their role in preprocessing, feature extraction, and classification of cancer-related VOC signatures. Case studies in lung, breast, and kidney cancers demonstrate the potential of E-nose systems to achieve high diagnostic accuracy. However, key limitations such as small dataset sizes, sensor variability, environmental interference, and the lack of standardization- remain significant challenges. By summarizing recent advancements and identifying areas for improvement, this review aims to support future research in translating E-nose-based cancer diagnostics into scalable, real-world clinical applications.

Abstract: Background Drug-drug interactions (DDIs) are an emerging issue in oncological patients, given the risk of increased toxicities and reduced efficacy from oncological treatments. This is particularly relevant in recurrent-metastatic (RM) head and neck cancer (HNC), where patients are often older, multimorbid, and receiving multiple home medications. While most DDIs are known for chemotherapy and targeted therapies, data on immune checkpoint inhibitors (ICIs) remain limited. Therefore, RM HNC patients represent a target population to investigate DDIs effects on ICIs. Material and Methods We conducted a retrospective, multicenter study to evaluate if polypharmacy affected safety and efficacy in RM HNC patients treated with ICIs, given alone (pembrolizumab or nivolumab) or in combination (combo) with chemotherapy (pembrolizumab-platinum-5-fluoruracil). DDIs were assessed using the Drug-PIN® platform, which provides for each patient two parameters: a Drug-PIN score, i.e. a continuous variable - the greater the number, the greater the likelihood of DDI; and a Drug-PIN level, with 5 levels ranging from green, meaning no interaction found, to red, meaning high potential for interaction. The relationship between Drug-PIN parameters and immune-related adverse events (irAEs) was tested using Student’s t-test. The correlation between Drug-PIN parameters and patients’ survival was assessed with the Kaplan-Meier method. Results We analysed 180 RM HNC (128 males) patients treated between June 2017 and July 2022 with ICIs (49 pembrolizumab alone, 89 nivolumab alone, 42 combo) at three different institutions. The median age was 65 years old (range: 33-89). Every patient took a median of 5 medications (0-15). Median Charlson Comorbidity Index score was 8 (2-14). IrAEs of any grade were observed in 55 patients (31%). Median PFS and OS were 5.9 and 6.5 months for pembrolizumab alone, 7.8 and 11.1 months for combo, 5 and 11.4 months for nivolumab. Median Drug-PIN score was 39.4 (0-214,3); Drug-PIN levels were green, light yellow, dark yellow, orange and red in 56, 27, 32, 9, 56 pts, respectively. Dichotomizing patients according to median Drug-PIN score, no difference was found in irAEs incidence (p 0.48), PFS (p. 0.82), or OS (p 0.89). No differences in irAEs incidence or in survival outcomes could be elucidated by stratifying patients according to Drug-PIN levels. Conclusions ICIs safety and efficacy in the treatment of RM HNC seem not to be affected by polypharmacy and DDIs.

Abstract: Rationale. Bioinformatics workflows increasingly need to integrate heterogeneous biomedical data, including experimental assay outputs, patient-derived model metadata, mechanistic models and molecular-pathology information. In glioblastoma (GBM), such integration can help connect treatment responses measured in patient-derived 3D cultures with molecular context and mechanistic representations of tumor metabolism, while maintaining a clear distinction between experimental observations and model-derived hypotheses. Results. We describe a standards-based workflow that integrates patient-derived GBM drug-response screens with a machine-readable QSP model, MGMT promoter methylation metadata and an explicit evidence-tier framework. MGMT-promoter methylation was confirmed in all patient-derived 3D cultures (Ge258, Ge518 and Ge904). Phenformin (PTF) showed the most pronounced single-agent activity, with a concentration-dependent reduction in viability, while 2-deoxy-D-glucose (2-DG) produced a more moderate effect. Among the higher-order combinations, MTF + BPTES + 2-DG + TMZ produced the most pronounced combination response and a statistically significant reduction in viability under the tested conditions. These findings are interpreted as treatment-response and prioritization signals rather than definitive evidence of pharmacological synergy or clinical efficacy. Availability and implementation. The workflow is designed around open and independently checkable data and model files, including raw plate exports, processed block-level tables, molecular metadata, an SBML/SED-ML/COMBINE model package, an evidence-tier table and a next-experiment protocol. Together, these components provide a reproducible framework for integrating computational predictions with experimental evidence and prioritizing testable drug-combination hypotheses.

Abstract: Background and Objectives: Consolidation durvalumab after chemoradiotherapy (CRT) is the standard of care for unresectable stage III non-small cell lung cancer (NSCLC). However, a substantial proportion of patients do not receive this treatment in real-world practice.. This study aimed to evaluate the efficacy of post-progression immune checkpoint inhibitor (ICI) therapy in patients with stage III NSCLC who did not receive durvalumab consolidation. Materials and Methods: This multicenter retrospective study included 403 patients with stage III NSCLC who received definitive CRT without durvalumab consolidation between January 2010 and December 2025. Patients were divided into ICI group (n=61) and non-ICI group (n=338) based on post-progression treatment. All patients has ECOG PS 0-1 The primary endpoint was overall survival (OS), analyzed using Kaplan-Meier and multivariate Cox regression methods. Early progression was defined as recurrence within 6 months after CRT completion. Results: The median OS was significantly longer in the ICI group compared to the non-ICI group (47.97 months [95% CI: 35.68–60.26] vs. 19.29 months [95% CI: 17.88–20.69]; p<0.001), representing a 2.5-fold survival advantage. Patients with early progression had a significantly worse prognosis compared to those with late progression (median OS: 7.06 vs. 20.01 months; p<0.001). In the ICI group, PD-L1-positive patients had numerically longer median OS than PD-L1-negative patients (54.05 vs. 46.03 months). Multivariate analysis identified post-progression ICI use (HR: 0.37, 95% CI: 0.25–0.54; p<0.001) and early progression (HR: 2.61, 95% CI: 1.91–3.57; p<0.001) as independent prognostic factors. Conclusions: Post-progression immunotherapy was associated witha significant survival benefit in patients with stage III NSCLC who did not receive durvalumab consolidation, with a median OS approaching 48 months. Early progression within 6 months after CRT is the strongest negative prognostic factor. These findings suggest that immunotherapy should be considered in this frequently encountered patient population, provided performance status is adequate. Prospective validation is needed.

Abstract: Hospital at Home (HaH) enables advanced cancer patients to receive end-of-life care in their preferred setting, yet evidence on home-based palliative sedation (PS) remains limited. This retrospective observational study included 113 patients who died in 2022 under the HaH Unit of the Instituto Valenciano de Oncología (Valencia, Spain). Sociodemographic, clinical, and PS-related variables were extracted from medical records. PS was required in 80 patients (70.8%). The median age was 76 years (IQR 19), 52.2% were male, and 77% lived in their own homes. Dyspnoea (43.7%) and delirium (23.7%) were the most frequent refractory symptoms leading to PS. Elastomeric infusers were used in 92.5% of cases. The median duration of PS was 1 day (IQR 2); 46.3% of patients died within 24 hours of initiation. A significant inverse correlation was found between the Barthel Index and PS duration (Spearman’s Rho = −0.226, p = 0.044), indicating shorter survival after sedation in patients with greater functional dependence. No significant associations were observed between PS requirement and multimorbidity, geriatric syndromes, polypharmacy, or prior opioid use. These findings support the feasibility of complex PS at home and highlight functional status as a potential determinant of PS duration, informing protocol optimization for home-based end-of-life care.

Abstract: Background/Objectives: Intravesical bacillus Calmette-Guérin (BCG) adverse events are often combined into a single outcome, although local symptoms, systemic toxicity, severe toxicity, and treatment-limiting intolerance have different clinical implications. We evaluated whether diabetes mellitus, recent urinary tract infection (UTI), age, baseline renal function, and baseline bladder capacity were differentially associated with these toxicity phenotypes. Methods: This retrospective exploratory cohort included 67 adults with intermediate- or high-risk non-muscle-invasive bladder cancer who received intravesical BCG at Bamenda Regional Hospital and Nkwen Baptist Hospital, Cameroon, from November 2020 to November 2025. Local-only toxicity required local symptoms without systemic toxicity. Clinically consequential intolerance comprised dose reduction, temporary interruption, permanent discontinuation, or documented poor tolerance. Five prespecified host factors were entered simultaneously into separate Jeffreys-prior penalized logistic regressions. Odds ratios (ORs) were scaled per 10 years of age, 0.1 mg/dL creatinine, and 50 mL bladder capacity; percentile intervals were obtained from 200 bootstrap resamples. Results: Mean age was 64.9 years; 14/67 (20.9%) had diabetes and 14/67 (20.9%) had a UTI within the preceding six months. Local toxicity occurred in 42 (62.7%), including 35 (52.2%) with local-only toxicity. Systemic toxicity occurred in 10 (14.9%), Grade III-IV toxicity in 6 (9.0%), and consequential intolerance in 19 (28.4%); 14 (20.9%) interrupted treatment and 3 (4.5%) permanently discontinued. Diabetes was inversely associated with local-only toxicity (adjusted OR 0.13, bootstrap 95% interval 0.02-0.53) but showed an imprecise positive association with systemic toxicity (OR 1.80, 0.28-17.05). Higher creatinine showed a positive, imprecise association with systemic toxicity (OR 1.67 per 0.1 mg/dL, 0.84-3.87), while older age showed an imprecise association with Grade III-IV toxicity (OR 2.81 per decade, 0.69-127.98). No prespecified factor consistently predicted consequential intolerance. Conclusions: Separating toxicity phenotypes revealed non-parallel host-factor associations that would be obscured by an any-toxicity endpoint. The estimates are exploratory and require prospective validation, but support phenotype-specific surveillance and standardized recording of treatment modification during BCG therapy.

Abstract: Motivation: Two independent Heidelberg drug screens recently identified five compounds with strong antimeningioma activity spanning four mechanisms: pan-class-I HDAC inhibition (panobinostat, romidepsin), proteasome inhibition (carfilzomib), microtubule stabilisation (ixabepilone), and translation elongation inhibition (omacetaxine). Neither screen stratifies responses by validated molecular subgroups, the standard design gap of functional pharmacogenomic studies. We sought to close this gap in silico by building a testable patient-selection framework returnable to the screening group. Results: We developed a convergent two-classifier analytical protocol that scores five mechanism-matched target-gene programs plus the StM 2026 panobinostat HDAC8→TGFβ→EMT resistance axis, assessing their expression across two independently derived meningioma classification systems: Nassiri 2021 Nature 4-group (Immunogenic/MG2/hypermetabolic/proliferative, N = 121) and Choudhury/Bi-lab 2022 Nature Genetics 3-group (Merlin-intact/Immune-enriched/Hypermitotic) projected via a leave-one-out cross-validated Ridge bridging classifier (LOOCV κ = 0.894, accuracy 93.0%). H1 (classifier concordance): the literature-registered 4→3 mapping (MG1↔Immune-enriched, MG2↔Merlin-intact, MG3∪MG4↔Hypermitotic) achieved Cohen's κ = 0.568 (p = < 1 × 10⁻¹⁶), ranking first among 12 enumerated alternative pairings (Δκ vs best wrong = +0.302; the most adversarial singleton-swap placebo gave κ = 0.050). H2 (subgroup stratification): all five programs differed significantly across Nassiri subgroups after Benjamini–Hochberg FDR (α = 0.10) applied across the full program × subgroup family (Kruskal–Wallis pBH = 1.7 × 10⁻⁵ to 0.084), with grade-adjusted OLS confirming subgroup F-statistics remain significant after conditioning on WHO grade. H3 (NF2 CNA-loss proxy): program z-scores associated with NF2 copy-number-loss status (nf2_cna_loss_proxy), with all five programs passing family-wide BH-FDR at α = 0.10. H4 (external replication) and H5 (recurrence-free survival) are declared not evaluable on currently open public data because replication cohorts lack subgroup metadata (H4) and time-to-recurrence variables are not deposited (H5); both are locked as pre-specified analyses awaiting co-author data sharing.