2026 ASCO Annual Meeting Report

Meeting theme: The Science and Practice of Translation: Improving Cancer Outcomes Worldwide

The 2026 ASCO Annual Meeting highlighted a decisive shift towards more precise, dynamic and patient-centred cancer care. From genomically guided treatment selection and de-escalation to the growing use of circulating tumour DNA for longitudinal decision-making, the meeting demonstrated how diagnostics are increasingly shaping not only which therapies patients receive, but also when treatment should be initiated, adapted or avoided. This report from Precision Oncology Consulting summarises the clinical advances, diagnostic innovations and strategic developments most relevant to the precision oncology ecosystem.

KEY TAKEAWAYS:

• Major advances for the treatment of metastatic pancreatic ductal adenocarcinoma (PDAC). Treatment with daraxonasib showed a reduction in risk of death of 60% compared to chemotherapy in the RASolute-302 trial in metastatic PDAC patients (abstract LBA5), one of the cancer types with the poorest prognosis and no effective treatments to date, representing a potential paradigm change in the treatment of these patients.
• Genomic tests moved from companion biomarkers to primary study endpoints driving de-escalation. The OPTIMA phase III study used the Prosigna/PAM50-guided test to guide adjuvant treatment decisions in clinically high-risk, node-positive, ER+/HER2− early breast cancer, resulting in a reduction of roughly two-thirds of patients treated with chemotherapy with non-inferior outcomes (abstract 500).
• Circulating tumor DNA (ctDNA) testing is increasingly used as a longitudinal treatment decision tool. Across studies like SERENA-6 (ESR1-emergence-guided endocrine treatment switching, abstract LBA1007) , FLAME (ctDNA based “non-response” triggering escalation, abstract LBA101), and GALAXY (ctDNA based adjuvant treatment decisions, abstract 102) ctDNA is positioned as a tool to inform both what therapy to administer based on the genomic profile of the patient, and when to escalate/de-escalate or switch therapies, ahead of radiographic progression.
• The field of cancer screening keeps evolving to increase test sensitivity and uptake. The landmark NHS-Galleri trial (PATHFINDER 2, abstract LBA10509) delivered the first randomized MCED data. Even though it did not meet the primary endpoint, it achieved a 14% reduction in Stage IV diagnoses across 12 major cancers. Integrating AI and blood-based lung cancer screening (DELFI) bypassed low-dose CT (LDCT) infrastructure constraints at OSF Healthcare, lifting screening rates from 18.2% in 2020, to 42.8% in 2025, and driving detection of earlier stage cancers. Though high-risk lung cancer screening remains the gold standard, access barriers exclude rising populations like non-smokers (abstract 109)
• AI advancements in the diagnostic setting led by Tempus: Multimodal foundation models, a clinical decision-support platform that analyzes digitized H&E slides to predict NGS QNS and recommend tissue input quantity, significantly improved real-world NGS success rates (abstract 1618), and AI-enabled HRD prognostic signature (abstract 4039) in tissue and several studies pushing the sensitivity boundaries of ctDNA detection in several cancer types (abstracts 5567, 8017)
• Globalisation of early-phase research and a US start-up bottleneck. Discussions highlighted China’s surging share of early-phase oncology trials, approaching or exceeding the US in some segments, driven by faster, lower-cost execution and strong early-stage biologics activity. The cost of delay was specific: the number of unique US trial sites running phase I NSCLC studies fell 44% from 2020 to 2024and one case study cited 13 oncology trials with zero patients enrolled over 18 months, ~$15.7M in lost/unrealised value (educational session, ascopost).
• The Patient-Centred Debate: The “Cancer Medicine Paradox” highlights the over-utilization of marginal, toxic, expensive therapies near the end of life alongside severe under-access to these same drugs in early stages where they could be more efficient. Patients increasingly prioritize overall survival, quality of life, and logistical burdens over unproven surrogate clinical endpoints.

Key studies selected for the Plenary session:

• Practice changing results for previously treated, metastatic pancreatic ductal adenocarcinoma (PDAC). The RASolute 302 trial (abstract LBA5), showed a 60% reduction in risk of death in the RAS G12-mutant population receiving daraxonasib vs the ones receiving investigator’s-choice chemotherapy. Beyond the implications for PDAC, daraxonasib is a multiselective RAS(ON) inhibitor, positioning pan-RAS inhibition as a platform rather than a single-indication
• The results of the largest phase III trial in localized high-risk prostate cancer, PROTEUS (LBA1) were selected to open the plenary session. The results showed that combined perioperative apalutamide and androgen deprivation therapy (ADT) reduces risk of metastasis or death of 20% compared to perioperative placebo + ADT.
• Promising results for advanced dedifferenciated liposarcoma from the SARC041 study (LBA2). This Phase 3 study showed median 6 fold improvement in progression free survival PFS for abemaciclib vs placebo PFS 9.7 vs 1.5 months (HR 0.38–0.39; 95% CI 0.25–0.59; P<.001), although OS superiority was not reached. Importantly, abemaciclib’s CDK4 selectivity (vs CDK6) as enabling continuous dosing with less neutropenia.
• Two studies in non-small cell lung cancer (NSCLC) completed the plenary session. The LIBRETTO-432 study in stage IB-IIIA RET-fusion positive patients (LBA3) administered adjuvant selpercatinib (CNS-penetrant oral RET TKI) for up to 3 years after definitive therapy, and showed an 83% reduction in recurrence/progression.death (HR 0.172, 95% CI 0.058–0.509, P=.0003), positioning RET inh as effective in non metastatic NSCLC, next to osimertinib (EGFR inh) and alextinib (ALK inh). The HARMONi-6 trial (LBA4) showed that patients receiving ivonescimab and chemotherapy had a reduced risk of death by 34% compared to patients receiving tislelizumab and chemotherapy (ITT HR 0.66, 95% CI 0.50–0.87, P=.0017), 24-month OS 64.7% vs 48.6%, with curves widening over time. This is the First regimen to show OS superiority over an active PD-1 inhibitor in first-line squamous NSCLC, and the first China-originated drug in an ASCO plenary.

Highlighted clinical studies presented by tumor type:

NOVEL DRUGS:

INDUSTRY ANNOUNCEMENTS

• Tempus AI: Received FDA companion diagnostic approval for its tumor-only xT CDx assay, and announced upcoming clinical availability of xH, a WGS based test for patients with hematologic malignancies.  It expanded its Next platform across six solid tumors and deployed Tempus Lens AI, an agentic AI platform for oncology drug development.
• Guardant Health: Presented 38 abstracts, highlighted abstract 3077 validating Guardant360 Liquid CDx as a pan-cancer companion diagnostic, abstract 3070 showing its use to guide ALK inhibitors in NSCLC, and abstract TPS10632 evaluating the performance of Shield for colorectal cancer screening.
• Caris Life Sciences: Showcased 32 multi-omic studies across 10 tumor types, involving more than 60 institutions.
• Foundation Medicine: Showcased 14 abstracts detailing copy-number losses, serial ctDNA tracking, and its proprietary HRDsig assay on FoundationOne CDx.
• Natera: Presented 35 abstracts establishing Signatera for the Treatment-on-MRD (TOMR) in colorectal cancer based on the results from the GALAXY study, and launched the multimodal Annotation digital platform.
• GRAIL: Unveiled updates from the NHS-Galleri trial and PATHFINDER 2 study, demonstrating reduced stage IV diagnoses using its Galleri multi-cancer early detection test, but not meeting the main endpoint of the study.
• Veracyte: Demonstrated milestone phase III data from the OPTIMA (Prosigna) and ENZAMET (Decipher Prostate) trials to precisely guide treatment intensification or chemotherapy avoidance.
• Myriad Genetics: Presented its whole-genome Precise MRD assay (1,000 variants) with gastric cancer data (MOSTAR-SCREEN-3, Abstract 4081) alongside Prolaris + AI digital pathology.
• Pharma: Pfizer executed a $10.5B oncology collaboration with Innovent Biologics (China). Johnson & Johnson shared 20+ abstracts, including PROTEUS data and multiple myeloma updates for MajesTEC-9 and CEPHEUS (Abstract 7513).

KEY TECHNOLOGICAL ADVANCEMENTS

• AI Foundation Models: Tempus presented data on multimodal foundation models trained on 2.5 million longitudinal records including clinical data, digitalized medical images and genomic and transcriptomic data. A proof-of-concept analysis of EGFR-mutant NSCLC treated with Osimertinib could stratify the patients in responders and non responders (HR 4.536) (C-index of 0.802 for overall survival (P < .001)) (see announcement).
• Spatial Atlases: Tissue diagnostics transitioned to AI-driven whole-slide image analysis. Two examples presented were (1) the Lunit SCOPE IO platform combunes AI-derived tumor microenvironment deatures with clinicopathological features to stratify microsatellite stable CRC patients based on risk of recurrence (abstract 3646) (2) spatial atlas quantifying ADC targets across several cancer types identified TROP2 as the most consistent target (abstract 3004)
• Ultra-Sensitive ctDNA detection: Natera presented the results of incorporating the phased-variant technology to their WGS ctDNA detection test, detecting ctDNA below 1 part per 10 million (sub-ppm) to detect MRD at ultra-low tumor fractions. They demonstrated that patients with MRD clearance defined by their test showed 0% recurrence rate (0/4 patients) (abstract 3060). Besides, they presented the results of a blood-only MRD detection test (Latitude) as highly concordant with tissue informed tests (abstract 3042). Myriad showcased its Precise MRD assay (WGS-based, tracking up to 1,000 variants) in the MONSTAR-SCREEN-3 trial, achieving a 97% baseline detection rate to accurately stratify recurrence risk. (industry expert theater). Guardant Health’s InfinityAI platform integrates genomic sequencing with methylation profiling to isolate ultra-low frequency tumor fragments from background hematopoietic clonal variations (see announcement).
• AI-driven ctDNA detection to monitor response to treatment for low shedding tumors: Natera showed clinical validity for their AI-based analysis of ctDNA levels in low shedding tumors, opening the door to the use of ctDNA in tumors where it was difficult to detect ctDNA to date. (abstract 4522).

KEY TUMOR BIOLOGY FINDINGS

• Overcoming “undruggable” targets : Daraxonrasib (RMC-6236) reframes RAS-addicted cancers by targeting the active, GTP-bound state across G12/G13/Q61 variants and wild-type RAS, blocking hyperactive signaling in pancreatic ductal adenocarcinoma (PDAC) (abstract LBA5)
• Mutational landscape and response to treatment: KRAS G12C inhibitor outcomes depend heavily on co-mutational landscapes; STK11 or KEAP1 alterations create an immunologically “cold” phenotype (abstract e20640).
• Importance of the exposome: Dr. Laura Mezquita highlighted how the environmental exposome (PM2.5/radon) triggers chronic tissue inflammation, promoting lung cancer growth in dormant EGFR or KRAS mutation carriers (e517088).
• Liquid Biopsy as a tool to detect clonal evolution: The SERENA-6 trial demonstrated that the detection of emergent ESR1 clones in ctDNA, before radiographic progression, and switching to camizestrant provides clear survival advantages (OS, HR = 0.39) (abstract LBA1007). The CROWN exploratory translational analysis reveals that long-term lorlatinib resistance is driven by bypass-pathway activation rather than secondary on-target ALK mutations (abstract 8502). Meanwhile, the FLAME trial defined persistent week-3 EGFR ctDNA as an aggressive biological state requiring early chemotherapy intensification (abstract LBA101).

CGP FOCUSED ABSTRACTS FROM LEADING DEVELOPERS:

MRD FOCUSED ABSTRACTS FROM LEADING DEVELOPERS: