Background: HPV-positive oropharyngeal squamous cell carcinoma (OPSCC) is a molecularly distinct disease of rising incidence that characteristically presents with advanced cervical lymph node metastases despite small primary tumours. Most molecular studies have relied on single biopsies, underestimating intratumoral heterogeneity and obscuring the evolutionary relationship between primary tumours and their metastases. Understanding this relationship is directly relevant to treatment planning, biomarker development, and ongoing de-escalation efforts.

Purpose: To characterise the evolutionary dynamics between matched primary tumours and cervical lymph node metastases in HPV-positive OPSCC through multi-region whole exome sequencing, clonal deconvolution, and phylogenetic reconstruction.

Methods: Thirty-five tumour regions and five matched normals from five HPV-positive OPSCC patients who underwent neck dissection were subjected to whole exome sequencing. Somatic variants were identified using a three-caller consensus approach with cancer cell fractions estimated by PyClone-VI. Evolutionary relationships were reconstructed using maximum parsimony phylogenetics alongside three complementary genetic similarity metrics.

Results: Primary tumours and metastases were profoundly divergent across all patients, with 81% of mutations private to individual samples and minimal subclonal sharing between compartments. Phylogenetic reconstruction consistently indicated early metastatic seeding followed by extensive independent evolution at both sites. Driver gene mutations were predominantly subclonal (65%), with chromatin remodelling and cell adhesion pathways recurrently targeted on independent branches across patients while differentiation and immune signalling alterations were enriched among truncal events. Truncal copy number alterations were more conserved than truncal SNV drivers, with 3q gain and 11q loss shared across most patients.

Conclusions: HPV-positive OPSCC undergoes early metastatic dissemination requiring only a minimal complement of founding genetic events, followed by extensive parallel diversification driven by ongoing subclonal acquisition of driver mutations. These findings highlight the limitations of single-biopsy profiling and have implications for biomarker-guided therapy and treatment de-escalation strategies.

Background:
The hypothalamus helps control hunger, body weight and metabolism, and is a major target of modern obesity drugs. Yet much of what we think we know about it comes from animal studies, especially rodents. This matters because treatments can translate broadly across species, but important details may differ in humans.

Purpose:
To build a high-resolution map of the adult human hypothalamus and use it to better understand the cells, pathways and genes involved in body-weight regulation.

Methods:
We combined single-nucleus RNA sequencing with spatial transcriptomics to create HYPOMAP, a spatio-cellular atlas of the adult human hypothalamus. This allowed us to identify individual cell populations, locate them within hypothalamic anatomy, and relate them to pathways and genes linked to obesity.

Results:
HYPOMAP includes more than 433,000 nuclei and identifies >450 distinct cell clusters across neuronal and non-neuronal populations. Major cell classes are shared across species, but we also found important human-specific differences in neuronal subtypes and receptors relevant to obesity biology and drug action. In particular, the atlas refines the organisation of melanocortin and incretin pathways, both central to appetite regulation and obesity therapeutics. Mapping human genetic signals onto the atlas showed that BMI-associated variants converge on specific neuronal populations and highlighted both established and new candidate effector genes. One example is DENND1B, where genetic, expression and spatial data together support a role in MC4R biology in the human hypothalamus.

Conclusions:
HYPOMAP provides a human reference map linking anatomy, cell identity and genetics in a brain region central to energy balance. It offers a practical framework for moving from obesity-associated genes to plausible human cell types, circuits and therapeutic hypotheses.

Background:
Axillary lymph nodes (ALN) initiate immune responses in breast cancer (BC) but how and when ALN become dysfunctional, facilitating metastasis, is unclear. The fibroblastic reticular cell (FRC) network within ALN provides structural support and mediates immune homeostasis, but we have yet to elucidate whether this network changes during BC progression or not.

Purpose:
We used an unbiased computational approach to quantify features of the immunolabelled FRC network in BC patient-derived ALN. We identified PDGFRβ as a robust, immunomarker for human FRC and used it to quantify how FRC network topology changes during BC progression and after treatment.

Methods:
Formalin-fixed paraffin-embedded ALN (n = 331) from 179 BC patients and 23 benign reactive controls were assessed for FRC network metrics, including lacunarity and branchpoints, alongside de-identified clinico-pathological data. We then integrated this data using multivariate, principal component and survival analyses.

Results:
In node-negative, treatment-naïve triple-negative BC, denser FRC networks in uninvolved nodes significantly improved survival (p = 0.0365). Conversely, similar changes seen in node-positive patients significantly worsened survival (p = 0.0407), regardless of BC subtype or treatment. In metastatic ALN, FRC network disruption grew proportionately to axillary tumour burden, and this significantly correlated with poorer outcomes (p = 0.043). Interestingly, increased FRC alignment within these metastases significantly improved survival (p = 0.0205).

Conclusions:
We have shown that changes in human ALN FRC network topology predict BC prognosis. This could improve how we risk stratify patients in future, and provide a new avenue for mechanistic, translational research.

Background: Gallbladder carcinoma (GBC) remains a highly aggressive malignancy with poor outcomes and limited therapeutic options. The absence of reliable biomarkers for early detection and prognostication further hampers effective management. Tumor budding (TB) and tumor-infiltrating lymphocytes (TILs) have emerged as promising histopathological indicators, reflecting tumor aggressiveness and host immune response in various cancers. However, their significance in GBC remains underexplored.

Methods: This retrospective observational study evaluated the association between TB, TILs, and clinicopathological parameters in histologically confirmed GBC cases that underwent surgical resection. TB and TILs were assessed in 60 patients using standardized protocols, and statistical analyses were performed to determine their correlations with key pathological features.

Results: High TB was significantly associated with advanced tumour stage and higher histological grade, indicating greater tumour aggressiveness. In contrast, high TIL density correlated with earlier stage, lower grade, and the absence of lymphovascular invasion, suggesting a favorable immune response. An inverse relationship was observed between TB and TILs, underscoring the interplay between tumor invasiveness and host immunity.

Conclusion: These findings suggest that TB and TILs serve as complementary prognostic markers in GBC, offering insights into tumor biology and potential avenues for prognostic stratification and targeted therapies. Further studies elucidating the underlying mechanisms of tumor–immune interaction in GBC are warranted to inform clinical translation.