I grew up inspired by Attenborough's infectious enthusiasm for nature. Here my dad Colin has a cameo as Sir David narrates his work tracking swan migrations. It's a glimpse into the Pennycuick household of the 1990s, when Attenborough's Life series were required viewing! Wishing a very happy 100th.

If you have a business idea in oncology do check this out, it's a fantastic programme. Applications open now www.cancertechaccelerator.co.uk

Cancer Tech Accelerator | Accelerate Your Research

Applications are open for the 2026 cohort!

buff.ly

AlphaMissense promises powerful AI for pathogenic variant identification. Is it clinically useful? Smart validation approach here against known signatures ascopubs.org/doi/10.1200/... Found with researchbriefing.com

Could we “see” intratumoral CD4 cells in vivo with PET? That’s the promise of this novel tracer. Very cool if it works www.science.org/doi/10.1126/... Found with researchbriefing.com

Night shifts increase lung cancer risk - and there's more to it than smoking. Very interesting BioBank analysis here academic.oup.com/sleep/advanc.... Found with researchbriefing.com

Immunotherapy tox is not always bad; can lead to better outcomes. More data on this in the post-CRT NSCLC setting bmcpulmmed.biomedcentral.com/articles/10.... Found with researchbriefing.com

Impact of irae characteristics on efficacy of consolidative immunotherapy following chemoradiotherapy in locally advanced NSCLC - BMC Pulmonary Medicine

Background Consolidative PD-L1 inhibitors after concurrent chemoradiotherapy (cCRT) have become standard care in locally advanced non-small cell lung cancer (LA-NSCLC). However, the correlation…

bmcpulmmed.biomedcentral.com

This is a compelling vision of how AI agents will improve our treatment decisions in future www.nature.com/articles/s43.... Impressive progress but still a way from the clinic. Found with researchbriefing.com

Development and validation of an autonomous artificial intelligence agent for clinical decision-making in oncology - Nature Cancer

Ferber et al. present an autonomous artificial intelligence agent system for deployment of specialized medical oncology computational tools, validating their system across various clinical scenarios…

nature.com

I’ve always struggled with dermatology. Never good at identifying skin problems. Now AI can do it for us, and better www.nature.com/articles/s41... - need this in EPIC ASAP! Found with researchbriefing.com

A multimodal vision foundation model for clinical dermatology - Nature Medicine

Trained on large and multicenter datasets with different imaging modalities, a foundation model is shown to have strong performance on the full spectrum of clinically relevant tasks and to increase…

nature.com

First line ipi nivo for mNSCLC. 19% still responding at 6y. We don’t use this combo but I’m sure there are some patients who would benefit, if we could target them better www.esmoopen.com/article/S205... Found with researchbriefing.com

Cancers are complex. This study shows moving from simple biomarkers to more complex models can improve outcomes www.nature.com/articles/s41... Found with researchbriefing.com

Real-world performance analysis of a universal computational reasoning model for precision oncology in lung cancer - npj Precision Oncology

npj Precision Oncology - Real-world performance analysis of a universal computational reasoning model for precision oncology in lung cancer

nature.com

Patients with COPD do better on immunotherapy www.nature.com/articles/s41... Found with researchbriefing.com

COPD patients with non-small cell lung cancer respond better to anti-PD-(L)1 immune checkpoint inhibitors - Scientific Reports

Scientific Reports - COPD patients with non-small cell lung cancer respond better to anti-PD-(L)1 immune checkpoint inhibitors

nature.com

In genomics we tend to think in 1D sequences. Great study shows we need think in 3D www.nature.com/articles/s41... Found with researchbriefing.com

Three-dimensional genome landscape of primary human cancers - Nature Genetics

This study characterizes the three-dimensional (3D) genome architecture of 15 primary human cancer types from The Cancer Genome Atlas. The analyses identify different archetypes of enhancer usage and…

nature.com

How old you look is a better predictor of outcome than how old you are. Another smart AI application shows prognostic value of FaceAge predictor in cancer patients buff.ly/e275kVV Found with researchbriefing.com

Spatial profiling of precancerous lung lesions identifies TIM3 as a potential interception target www.sciencedirect.com/science/arti.... Fantastic study. Found with researchbriefing.com

Can we improve immunotherapy responses by depleting regulatory T-cells? Encouraging data in GBM www.cell.com/immunity/ful... Found with researchbriefing.com

Can we predict patient outcomes by… asking them? Longitudinal analysis of PRO data suggests we can aacrjournals.org/clincancerre... Found with researchbriefing.com

Love this use of AI. Reviewed >600k oncology patient messages to find their main care issues. Great to focus future improvements jamanetwork.com/journals/jam... Found with researchbriefing.com

Clonal haematopoeisis; important in ageing and now associated with worse outcomes when infiltrating tumours. More great work from the TRACERx team www.nejm.org/doi/10.1056/... Found with researchbriefing.com

In scRNAseq how do you find tumour cells? This approach makes sense to me, using CNA profiles inferred from the transcriptome academic.oup.com/bib/article/... Found with researchbriefing.com

Antibiotics can affect immunotherapy efficacy via the microbiome. Should be used with caution. Helpful and detailed guide: ascopubs.org/doi/10.1200/... Found with researchbriefing.com

Adding RT to peri-op chemoIO improves MPR but not OS/PFS www.jto.org/article/S155... Found with researchbriefing.com

MetaGP - medical foundation model improves diagnostic performance of doctors buff.ly/WzRk39a Found with researchbriefing.com

This got me thinking about neoadjuvant trial design. Phase II neoadj EGFR inhibitor declared a success due to high conversion rate to resectability in unresectable stage III disease www.nature.com/articles/s41... (1/2)

Exosomes are definitely interesting for early cancer detection. AI-powered exosome detection using microfluidics seems like a smart approach buff.ly/rKqTrXu Found with researchbriefing.com

Can LLMs design proteins now? Enzymes designed with natural language prompts... incredible... buff.ly/nSvYH9Z Found with researchbriefing.com

Toward De Novo Protein Design from Natural Language

De novo protein design represents a fundamental pursuit in protein engineering, yet current deep learning approaches remain constrained by their narrow design scope. Here we present Pinal, a…

biorxiv.org

People at germline risk of autoimmune diseases also at risk of ICI tox. Makes sense - could lead to better personalisation in future buff.ly/qj0KnOh Found with researchbriefing.com

Germline prediction of immune checkpoint inhibitor discontinuation for immune-related adverse events

Introduction Immune checkpoint inhibitors (ICIs) can yield remarkable clinical responses in subsets of patients with solid tumors, but they also commonly cause immune-related adverse events (irAEs).…

jitc.bmj.com

Could protein translation defects lead to more immunogenic cancers? Could we exploit this vulnerability? Study of key translation proteins suggests so buff.ly/C6PKdF4 Found with researchbriefing.com

Here's a good question - if AI is so good at radiology interpretation, can we reduce the radiation dose? Study suggests you can, and AI copes fine with reduced quality buff.ly/GkzyUN5 Found with researchbriefing.com

DCIS recurrence is not the same as DCIS progression. Fascinating evolutionary study with important implications breast-cancer-research.biomedcentral.com/articles/10.... Found with researchbriefing.com

Evolutionary measures show that recurrence of DCIS is distinct from progression to breast cancer - Breast Cancer Research

Background Progression from pre-cancers like ductal carcinoma in situ (DCIS) to invasive disease (cancer) is driven by somatic evolution and is altered by clinical interventions. We hypothesized that…

breast-cancer-research.biomedcentral.com

Serum VEGF-A as a biomarker for bevacizumab? Intriguing data www.nature.com/articles/s41... Found with researchbriefing.com

Serum VEGF-A as a biomarker for the addition of bevacizumab to chemo-immunotherapy in metastatic NSCLC - Nature Communications

This prospective study conducted in a phase 3 trial has shown that pre-treatment serum VEGF-A was a negative biomarker for PFS in patients with advanced NSCLC treated with a combination of an…

nature.com