💥 The discovery of novel #companion #diagnostics for #oncology needs to be systematised and accelerated.
My ex-EBI colleague Michael Menden has published a fantastic new tool (#OncoBird) in Nature Communications to get us closer to this. Using data from oncology randomised clinical trials (#RCTs), they are able to discover predictive response #biomarkers for #personalised treatment.
If these biomarkers can then be regulated+packaged as companion diagnostics and deployed+distributed through diagnostics-as-a-service (#DaaS) infrastructure, this could tremendously benefit #cancer #patients.
Also, I wonder if OncoBird would become more powerful using DNA methylation (#epigenetic) data? i.e. can we capture both genetic alterations and tumour subtype information using a single DNA methylation assay?
Any #pharma partners with oncology pipelines (and oncology RCT molecular data / samples), please get in touch!
URL to the paper: https://lnkd.in/er7TQrEf
**********************************************************
A quick summary of the paper for the geeky ones 🤓 :
💡 There is a massive need to develop actionable biomarkers to personalise oncology treatment.
🔬 The authors have created the OncoBird framework, which analyses biomarker data (inc. genetic data) from RCTs to yield tumour subtype-specific biomarkers with treatment benefits. In other words, given genetic data from a tumour (hypothesis-free) and previous knowledge about tumour subtypes (hypothesis-driven, prior cancer biology knowledge), we want to identify biomarkers that predict higher chances of survival with a specific treatment vs alternative treatment (i.e. develop predictive response biomarkers).
🏥 They demonstrate the utility of this framework using data from the FIRE-3 trial:
--> This includes metastatic colorectal cancer (mCRC) patients who were treated with FOLFIRI and either #cetuximab (Eli Lilly and Company / Merck Group) or #bevacizumab (Genentech).
--> Besides re-discovering known predictive biomarkers (like RAS or BRAF mutations associated with cetuximab resistance), OncoBird uncovers new ones (like ARFRP1 amplifications associated with longer survival in cetuximab vs bevacizumab in the CMS2 tumour subtype).
😍 This framework can be generalised to any RCT with molecular data, and allows for interpretability and transparency, which is key for clinical decision making.
Huw Ricketts
Richard Watts
Tanya Aneichyk
James Hadfield
Martin Widschwendter
Maria C. M. Orr
Karl Wishart
Developing safe innovative digital health products | Digital Health Consultant | CMO | Health Education | GP | Democratising healthcare | Medical Director | GMC (🇬🇧) and IMC (🇪🇺) Registered
1wI’ve only just had a chance to listen to this Daniel Elías Martín Herranz. I missed the live show but caught up on Spotify https://open.spotify.com/episode/6BPu8D0omuOmk3etfMsHYr?si=ZC66QKUcQ0y-Xy5FX2sWSg&t=3229&context=spotify%3Ashow%3A5pZb65BCLauYxbJrbzDQfP I particularly enjoyed two elements: 1- the debate about the chicken and egg (what comes first) of biomarker vs therapeutics 2- Steve Roest ‘s noble attempts to get the pronunciation of Dani’s name perfect each time