🥳🎆 Kicking of 2026 with our work showing intra-tumoral bacteria 🦠 affect outcomes of Radiotherapy anti-PD-L1 now out in @NatureCancer with sister paper by @nataliesilvermd! Even when trials "fail" – we can learn from them! nature.com/articles/s43018-0…. -- w/ @GeneCollector, @imrtlee, @TylerAlban1, @DrHaddadRobert, the Pfizer team and more! (1/N)

One win away

Arguably the most boring step in genomics is the first one: normalization. Settled science. Scale log. Move on. Except that here's been a huge blind spot in the field. And it matters for AIxBio. A 🧵about what I think may be one of the most important papers I've written. 1/

🧬 Can tumor genomics be used to personalize radiation dose? Grateful to present results from our phase II trial on dose de-escalated radiation therapy for ATM altered tumors at #ASCO2026 Great to see ☢️ #Radiation in the early-phase #PrecisionOncology space!

75% of patients received 30 Gy with FMISO PET based de-escalation for HPV-associated OPSCC. 30 Gy. Need the randomized data but hard to believe this won’t be the way. Incredible!

Thrilled to share our new paper out today in @NatureCancer: Tumor irradiation promotes antigen dressing of dendritic cells to enhance CAR T cell persistence and efficacy in lung metastases. A single dose of radiotherapy makes CAR T cells work against extensive metastatic lung cancer — even when the target antigen is also on normal tissues. nature.com/articles/s43018-0… 🧵 (1/14)

No QoL difference with IMPT vs IMRT in oropharyngeal cancer in the TORPEdO 🇬🇧 trial. How to explain the differences w/ @SJFrankMD 🇺🇸 trial? Planning? Patients? Crossover in the 🇺🇸 trial? Real absence of difference? Cc @EmmaHall71 @

Absolutely terrific work by @DrSymYoung to report the long-term outcomes of our INRT experience from two prospective trials (INRT-AIR and DARTBOARD). Short version: with long-term follow-up (median 5.2 years for INRT-AIR, 3 years for DARTBOARD), we've seen zero solitary elective nodal recurrences. Longer version: ESTRO has highlighted novel approaches to managing the elective neck in HNSCC, and I believe the future will be very different than the present. Current ENI fields deliver the majority of the integral dose to patients and contribute substantially to critical structures (swallowing and xerostomia OARs). Minimizing ENI dose and volume may meaningfully improve the short- and especially long-term tolerance of radiotherapy. A few more thoughts on our INRT paradigm below:

Protons cannot do magic (😱I know ). They save you low dose exposure which is less relevant for re-irradiation. In the high dose region, depending on anatomical site, VMAT/IMRT gives you often better conformality and high dose gradients for re-irradiation.

Really exciting work by @HeadNeckMD and @jdschoenfeld1 -- The way to de-escalate HPV-related cancers, even advanced tumors is monitoring treatment response and selecting those that respond! Nice proof of concept this can be done with ctDNA --- Looking forward to seeing this space develop @CJTsaiMDPhD @imrtlee @sueyom @seanmmcbride @BillDiplas

The future is here!!! First ever plasma liquid biopsy capable of profiling the tumor microenvironment with spatial transcriptomic insight. Out today in @Nature!

Very thoughtful editorial on some possible missteps in science communication during Covid. Clearly we need new and more better strategies of public communication in an era of declining trust l.

📣 We are excited and thrilled to announce APOLLO, a healthcare system-scale multimodal temporal foundation model for virtual patient representations. Trained on 25 billion clinical events from 7.2 million patients across 33 years and 28 modalities, APOLLO learns a unified atlas of medicine. Turning labs, notes, pathology images, medications, and diagnoses into coherent, computable longitudinal trajectories. APOLLO is disease-agnostic by design, a single model that learns the shared structure underlying human health and disease across every specialty, modality, and stage of care. The possibilities are enormous: earlier risk prediction, treatment response modeling, clinical trial matching, biomarker discovery, and a new generation of agentic systems built on rich patient representations. Read the pre-print: arxiv.org/pdf/2604.18570 Read our blog post about the work: linkedin.com/pulse/apollo-mu… 👏 🎉Huge congratulations to Andrew Zhang , @TongDing99, Sophia J. Wagner, and the rest of the team.

1/n Setidegrasib the first-in-human, first-in-class, KRAS G12D-targeted protein degrader #TPD Our KRAS G12D degrader study is now published in the New England Journal of Medicine @NEJM nejm.org/doi/full/10.1056/NE… A new way to target KRAS G12D - one of the most common oncogenic drivers across cancers.

1/n Now published in @NatureMedicine : The phase 2 #POLAR trial of pembrolizumab olaparib maintenance in metastatic pancreatic cancer. @CpcrMsk @MSKCancerCenter @MSK_DeptOfMed @PICIatMSK rdcu.be/e9WmJ

🧵 Just published in @TheLancet: TORPEdO – the first phase 3 RCT designed specifically to test whether IMPT (proton beam) improves late function & QoL vs modern IMRT in oropharyngeal SCC. Short answer: It doesn’t. Long answer (with the numbers that matter) 👇