ASCO2026:AWeekThatMovedtheField

I spent the last week of May in Chicago at ASCO 2026 the world’s largest oncology conference, drawing over 44,000 clinicians, researchers, and industry professionals globally. Every day of the meeting, session digests, magazine briefings, and late-breaking trial summaries were circulating in real time. It took me a full week after landing to compile it all.

The Theme That Framed Everything: ASCO 2026 ran under the theme “The Science and Practice of Translation: Improving Cancer Outcomes Worldwide.” Translation not just bench to bedside, but from clinical trial to community practice globally was the lens through which every major readout should be understood. For me these five late-breaking plenary abstracts anchored the meeting: PROTEUS in prostate cancer, SARC041 in dedifferentiated liposarcoma, LIBRETTO-432 in early-stage RET fusion-positive NSCLC, HARMONi-6 in squamous NSCLC, and RASolute 302 in metastatic pancreatic cancer. Together they reflected a field simultaneously breaking ceilings and asking harder questions about who the advances actually reach.

Pancreatic Cancer: The Standing Ovation That Stopped a Room

If I had to point to one moment that defined ASCO 2026, it was May 31st in the plenary hall. When Brian Wolpin, MD, MPH, of Dana-Farber Cancer Institute showed the survival slide for RASolute 302, the auditorium gave a 42-second standing ovation cheering, whistling, the works. Wolpin deadpanned: “That time was not built into my talk.”

The reason: daraxonrasib delivered a median OS of 13.2 months versus 6.7 months for chemotherapy, with a hazard ratio of 0.40 a 60% reduction in risk of death (p<0.0001), in 500 patients with previously treated metastatic PDAC. Results published simultaneously in the New England Journal of Medicine.

The biology is worth understanding. Daraxonrasib is an oral RAS(ON) multi-selective inhibitor it targets the active, GTP-bound form of RAS proteins across multiple genotypes, not just KRAS G12C. It works in both RAS-mutant and RAS-wild-type PDAC, which matters in a disease where ~90% of patients carry a broad spectrum of RAS mutations that earlier single-mutation inhibitors couldn’t touch. The FDA had already granted Breakthrough Therapy Designation.

ASCO CMO Julie Gralow called it “a grand slam.” Invited discussant Jennifer Knox called it “a game changer in pancreatic cancer.”With a five-year survival rate of approximately 3% for metastatic disease, that language is warranted. Daraxonrasib is also being evaluated in three additional Phase 3 trials frontline PDAC, NSCLC, and colorectal cancer. This is a platform, not a single indication.

NSCLC: The Bispecific Arms Race and Precision Adjuvant Therapy

HARMONi-6: Ivonescimab Makes History

HARMONi-6 became the first China-originated investigational drug ever selected for the ASCO Plenary in the society’s 61-year history. The study enrolled 532 patients with previously untreated advanced squamous NSCLC. Approximately 39% had PD-L1 TPS below 1% and 33.8% had multi-site or brain metastases. Ivonescimab a first-in-class PD-1/VEGF bispecific beat tislelizumab plus chemotherapy on overall survival: hazard ratio 0.66, statistically significant and clinically meaningful. 

The mechanistic logic: Vascular endothelial growth factor (VEGF) is not just angiogenic, it is immunosuppressive. It drives regulatory T cell infiltration and physically excludes effector T cells from the tumor microenvironment. A single molecule blocking both PD-1 and VEGF should create a more immune-permissive tumor than either agent alone and now has Phase 3 OS evidence to prove it.

Caveats are real: patients over 75 were excluded despite the average American lung cancer patient being 70, and the Chinese trial population has demographic differences from Western cohorts. The confirmatory global HARMONi-3 trial is underway, with US PDUFA November 14, 2026. That readout settles the question.

The bispecific field is crowding fast. BMS/BioNTech’s pumitamig showed a 70% ORR in early frontline NSCLC data, and Pfizer’s PF-08634404 posted 64–75% confirmed ORR regardless of PD-L1 status. Analysts see no clear differentiation yet long-term separation will likely hinge on combinations with ADCs.

LIBRETTO-432: Adjuvant Targeting in Early-Stage NSCLC

In patients with Stage II–IIIA NSCLC (non-small cell lung cancer) who had already undergone surgical resection meaning curative-intent tumor removal and whose tumors carried a RET gene fusion, adjuvant selpercatinib produced a hazard ratio of 0.172 for event-free survival. To unpack that: adjuvant therapy is treatment given after surgery to eliminate residual microscopic disease and reduce recurrence risk. A RET fusion occurs when the RET gene which encodes a receptor tyrosine kinase, a cell-surface protein that normally controls growth and survival signaling becomes abnormally fused to a partner gene, locking RET in a constitutively active state that drives uncontrolled cell proliferation. Selpercatinib is a highly selective RET kinase inhibitor that fits into RET’s active site and silences that runaway signal. A hazard ratio of 0.172 means patients on selpercatinib had an 83% lower risk of disease recurrence or death compared to placebo translating to two-year EFS (event-free survival) rates of 91.5% versus 61.1%. That is not an incremental improvement. That is a category shift.

The ADAURA analogy is exact. ADAURA was the Phase 3 trial that established osimertinib a selective inhibitor of mutant EGFR (Epidermal Growth Factor Receptor), another receptor tyrosine kinase frequently mutated in never-smoker lung cancer as adjuvant standard of care in early-stage EGFR-mutated NSCLC. ADAURA showed a landmark EFS benefit first; OS (overall survival) data confirmed the signal held. The field is drawing the same arc for LIBRETTO-432: if the biology follows the same logic, OS confirmation makes selpercatinib the new adjuvant standard for RET fusion-positive early-stage NSCLC.

The practice implication is unambiguous. Comprehensive molecular profiling at initial diagnosis testing for RET fusions, EGFR mutations, and the full panel of actionable drivers is no longer a conversation reserved for metastatic disease. You cannot deploy a targeted adjuvant therapy against a driver alteration you never looked for. Biomarker testing at resection is now a clinical imperative.

Breast Cancer: De-escalation, a New ADC, and Seven-Year Durability

OPTIMA: Most “High-Risk” Patients May Not Need Chemotherapy. OPTIMA enrolled 4,429 patients with ER-positive, HER2-negative early breast cancer all clinically high-risk, many with positive lymph nodes. Approximately 68% had low Prosigna genomic risk scores. For that group, 93.6% treated with endocrine therapy alone were alive and recurrence-free at five years, versus 94.8% with standard chemotherapy, a non-inferiority result.

The core message: lymph node count is a poor proxy for chemotherapy sensitivity when genomic data is available. Tumor biology PAM50 gene expression patterns governing proliferation, luminal differentiation, and immune infiltration outperforms anatomy as a treatment decision variable. “Prosigna now has the strongest prospective evidence of any genomic test to predict which patients with clinically high-risk hormone-receptor positive breast cancer can safely avoid chemotherapy,” said Veracyte CSO Phillip Febbo. The open question remains: premenopausal women on ovarian function suppression, where chemotherapy benefit may partly reflect ovarian suppression rather than direct cytotoxicity.

Iza-bren: A Bispecific ADC Changes the Game in Metastatic TRIPLE NEGATIVE BREAST CANCER (TNBC). A result that deserves more attention than it got in the headline rush: izalontamab brengitecan (iza-bren), a first-in-class EGFRxHER3 bispecific ADC, improved both progression-free and overall survival versus standard chemotherapy in previously treated metastatic or locally advanced TNBC extending median PFS by 5.4 months and median OS by 3.4 months. The trial showed a 71% reduction in the risk of disease progression or death.

TNBC is the subtype with the fewest targeted options. Iza-bren targets two receptors simultaneously EGFR and HER3 on a single ADC payload, exploiting co-expression patterns in triple-negative tumors. Co-developed by SystImmune and Bristol Myers Squibb, this is a potential new standard of care for pretreated metastatic TNBC.

KEYNOTE-522: Seven Years of Durability in TNBC. Final KEYNOTE-522 results confirmed that after approximately seven years, over 85% of early-stage TNBC patients who received pembrolizumab plus chemotherapy before surgery were alive, versus 77% on chemotherapy alone. An eight-point survival advantage that has held across the full follow-up window. The durability of this result matters it validates pCR as a meaningful surrogate for long-term outcomes in early TNBC.

GRAIL and the Cancer Screening Reckoning: The NHS-Galleri trial, the first and only randomized controlled trial of a multi-cancer early detection test enrolled 142,942 asymptomatic adults aged 50–79. Its primary endpoint of significantly reducing combined Stage III–IV cancer diagnoses was not met.

But the data beneath that headline are directional and accumulating. Annual Galleri testing reduced Stage IV diagnoses by 22% and 26% in the second and third screening rounds respectively, increased Stage I–II detection by 16%, and reduced emergency-presentation cancer diagnoses by 25%. The primary endpoint failed because Stage IV reductions were offset by an increase in Stage III diagnoses in the first prevalent screening round a known artifact of early screening programs, where the first screen finds cancers that were already there. The signal accumulates with repeat annual screening; three years may not be long enough for the composite endpoint to mature.

Meanwhile, PATHFINDER 2 in 35,878 North American participants showed the Galleri test increased cancer detection 6.5-fold when added to standard-of-care screenings, with 71% of newly detected cancers in Stages I–III.

The regulatory tension: Medicare’s coverage process for MCED tests has stalled precisely because no trial hit a primary endpoint. The 26% Stage IV reduction in Round 3 is GRAIL’s strongest argument but coverage without a primary endpoint hit sets a precedent CMS will not make lightly. GRAIL has extended follow-up by 6–12 months. The next data cut will be more informative than the primary result. This story is not over.

The Rest of the Plenary: Sarcoma and Prostate

SARC041 delivered the first positive Phase 3 result in dedifferentiated liposarcoma abemaciclib extended PFS from 1.5 to 9.7 months (HR 0.38). CDK4 amplification is the defining molecular feature of this tumor; this is molecularly rational in a disease with no approved targeted options.

PROTEUS showed perioperative apalutamide plus ADT around radical prostatectomy made patients nine times more likely to have near-complete pathologic response at surgery, and reduced the risk of metastasis or death by 20%. The theme is consistent with the broader meeting: treat earlier, guided by biology.

Beyond the plenary, a media briefing found that patients with stage I–III lung, breast, colorectal, or liver cancer who were on GLP-1 medications had a lower risk of progression to Stage IV. Association, not causation but the anti-inflammatory and metabolic biology is plausible, and the research agenda it opens is real.

What ASCO 2026 Tells Us

Undruggable category is shrinking. RAS was a locked door for four decades. It isn’t anymore. The “high-risk breast cancer patient who needs chemotherapy” category is also shrinking genomic testing is reading tumor biology more accurately than anatomy. Multi-cancer early detection is not defeated by one missed primary endpoint; it is maturing toward a harder evidentiary standard, and the accumulating secondary data suggest the biology is sound. And China-originated oncology data ivonescimab, iza-bren, RASolute 302’s concurrent NEJM publication is no longer peripheral to the global conversation. It is the conversation. That shift was impossible to miss on the expo floor, where biotech presence was larger, louder, and more scientifically substantive than any prior year pipelines that would have been considered early-stage curiosities five years ago are now reading out Phase 3 OS data on the ASCO Plenary stage. Some of the most intellectually rich moments of the week happened in the poster sessions. New molecules in early phase development, mechanism-of-action data that hasn’t made headlines yet, biomarker analyses that quietly reframe how we think about patient selection, and critical data points that the broader oncology community will be sitting with for months. These are the conversations that don’t make the press releases but end up shaping how the field moves. Investigators fielding questions three deep around their boards, debating signal versus noise in real time that energy was irreplaceable.

ASCO 2026 was a meeting that gave oncology genuine reasons for optimism, while asking the field to be rigorous about which reasons are earned. Not every headline was clean. Not every primary endpoint was met. But the direction of travel more precise, more biology-driven, more globally distributed is unmistakable. That balance is exactly what progress looks like.