Hematology-oncology just got a wake-up call, and pharma should be listening.
A JAMA analysis of OpenEvidence (Patel, Jena and colleagues; 952,008 verified U.S. clinicians and 271 million queries, Jan 2024–Jun 2026) put oncology at the top of specialty use of OpenEvidence, with a use ratio of 2.00 against its share of the U.S. physician workforce, ahead of internal medicine (1.88) and family medicine (1.80). Physicians reading the paper were quick to add the fine print: OpenEvidence supplied the data, three of the seven authors work for the company, and the study measures use, not accuracy.
Oncologists are among OpenEvidence's heaviest physician users, with primary care close behind. JAMA, 952,008 US clinicians, 271 million queries.
— Katy Beckermann (@katy_beckermann) Oct 19, 2026
📊 Query share vs workforce share: oncology 2.00x, internal medicine 1.88x, family medicine 1.80x, radiology 0.24x
👥 NPs and PAs: 44.5 queries each per active month vs 37.1 for physicians
🎯 Treatment 33.1% of queries, diagnosis 13.9%
🔍 OpenEvidence supplied the data. 3 authors on staff, does not address accuracy
AI evaluations have focused on diagnosis, but clinicians ask about treatment more than twice as often. Treatment answers in LLM are where we need additional accuracy data. Tab 1. EMR Tab 2. Clinical trials Tab 3. @DeepScribeAI Tab 4. @OpenEvidence
👉 https://t.co/6Eip03nCyY
#AIinMedicine #ClinicalAI #Oncology
A JAMA study of 271M queries from 952,008 US clinicians: OpenEvidence use is mostly treatment and management (33.1%), not diagnosis (13.9%). Orthopedics sits at a 0.24 use ratio. Three of seven authors work for OpenEvidence, and the paper measures use, not accuracy. https://t.co/Ln27ObRbEW
— Felix C. Öttl, MD (@felixcoettlmd) Oct 17, 2026
That ranking shouldn’t surprise anyone who’s been paying attention. This field has led for years in online data creation, conference content and digital opinion leadership, and many of the earliest digital opinion leaders didn’t stay on the sidelines. They became industry and policy leaders: Nathan Pennell as SVP, Worldwide Head of Hematology/Oncology Medical Affairs at Bristol Myers Squibb; Jack West at Summit Therapeutics; Vinay Prasad as FDA CBER Director and Chief Medical and Scientific Officer before returning to UCSF. Their peers have followed into leadership across hematology and oncology.
What we saw at ESMO 2026 was the next chapter: digital opinion leaders sharing insights, aggregating content and using AI to publish faster than pharma and its agency stacks can keep up.
That gap showed up clearly around the INTerpath-001 embargo breach for intismeran autogene, the Moderna–Merck individualized mRNA therapy given with pembrolizumab in resected stage IIB–IV melanoma. Investors moved first: Moderna shares fell as much as 25% intraday after the early release (STAT’s Adam Feuerstein put Thursday’s drop at 17%). Physicians scrambled to reconcile a strong clinical signal, RFS HR 0.58 and DMFS HR 0.62 (LBA1, ESMO 2026; OS immature; investigational), with a market that looked irrational from the clinic floor. Melanoma oncologist Mario Balsa called the breach out directly (our INTerpath-001 leak recap).
INTerpath-001 leaked early. @myESMO lifted the embargo. RFS HR 0.58; Moderna −25%.
— KOL Pulse AI (@KolPulseAI) Oct 22, 2026
https://t.co/Czu1pUXh9j
@moderna_tx @Merck #ESMO26 https://t.co/UmYvaJ8vEk
🚨 Scientific embargoes exist for a reason: fair access to data, scientific integrity and, ultimately, patients.
— Mario Balsa (@MarioBalsaMD) Oct 23, 2026
Seeing a major #ESMO26 study disclosed ahead of schedule is disappointing.
I think ESMO takes its embargo policy seriously, and I would expect a breach of this magnitude to have consequences.
Science should set the timeline, not the stock market 😡
With INTerpath-001, someone clearly took a shortcut on the path to the data.
@OncoAlert @OncoReporte
@myESMO @_SEOM @GrupoMelanoma
Twelve more patients in a hundred stay recurrence-free at 2 years — on top of pembrolizumab alone. INTerpath-001 (LBA1, #ESMO26): resected stage IIB–IV melanoma; intismeran autogene + pembrolizumab vs pembrolizumab (n=1,137).
— Mustafa Özdoğan, MD (@ozdogan_md) Oct 24, 2026
Early release after embargo breach — confirm at PS I (Long). @ProfGLongMIA
STAT’s Adam Feuerstein explained the selloff for the investors in the room:
For the $MRNA haters, I wrote this yesterday...
— Adam Feuerstein ✡️ (@adamfeuerstein) Oct 24, 2026
The intismeran study was positive, so why did Moderna’s stock price plunge?
When Wall Street meets a medical meeting, strange things can happen. This is especially true when investors were already expecting very big things to happen.
The situation that Moderna (the stock, not the company) finds itself in at ESMO is what’s referred to by folks who sit in front of Bloomberg terminals all day as a “sell on the news event.” Stock prices often reflect future events and possibilities, and Moderna, which added $60 billion in market value just on the August intismeran press release, alone, had already priced in a lot of the therapy’s revenue upside.
If we’re being totally honest, Moderna’s tripled stock price and $90 billion market value overpriced the most optimistic intismeran business case. By how much is debatable, but to use more Wall Street jargon, the stock had gotten way ahead of itself.
So, that’s why Moderna’s stock price plunged 17% on Thursday when the intismeran data leaked. As this newsletter was put together, the stock was down another 1% Friday. That’s actually pretty good.
“This was a positive trial, and a very important outcome for patients and the field,” Jedd Wolchok, director of the Meyer Cancer Center at Weill Cornell Medicine and an immunotherapy pioneer, told STAT on Friday. “I don’t quite understand what the reason for the disappointment was, but you know, I don’t pretend to be smarter than a biotech analyst.”
This is not to say that the INTerpath-001 study leaves all questions answered. There will be debates. But investors on Wall Street and oncologists in Madrid are not always on the same wavelength.
“Taking a step back, the data were very much in line with expectations,” said Mizuho health care strategist Jared Holz, who graciously spoke to STAT Friday while touring colleges with his kid. “From the Wall Street perch, this was always going to be a tricky set-up for the stock given a parabolic move over the past month coupled with a $100 billion market value coming into ESMO.”
Meanwhile NADINA’s three-year update kept the perioperative IO debate alive: response-adapted neoadjuvant ipilimumab plus nivolumab versus surgery then adjuvant nivolumab, with 3-year EFS 72.9% vs 50.0% (HR 0.42; ESMO 2026 2076RO, data cut 30 Mar 2026). Neoadjuvant use is not on the FDA label. Digital opinion leaders, including surgeon Eric D. Whitman (@Melanoma_doctor), who disclosed that he took part in INTerpath-001, pushed the hard comparison problem: adjuvant RFS vs neoadjuvant EFS, what counts as standard in resectable stage III, and why a trial’s comparator shapes how the field reads a “win.”
Sad that the @myESMO embargo was broken...
— Eric D Whitman, MD (@Melanoma_doctor) Oct 23, 2026
It is difficult to compare results between an adjuvant study and a neoadjuvant study (ie, S1801) in any cancer, including #melanoma. The patient populations were different and the primary endpoint (RFS vs EFS) was of necessity different for technical reasons that as @adamfeuerstein points out don't translate well to this platform. Until tomorrow's presentation we won't have all the data necessary to even try and compare.
It is also important to note that NO drugs are technically @FDA approved for neoadjuvant treatment for melanoma. It's a moot point as several studies including S1801 and NADINA have made neoadjuvant therapy standard of care in some situations.
I think it's great that the study (that I participated in) appears to have hit its primary endpoint, which typically ultimately translates into FDA approval.
Tomorrow (1030 ET) will be interesting...
thanks for your reply.
— Eric D Whitman, MD (@Melanoma_doctor) Oct 23, 2026
Actually, neoadjuvant pembro is not necessarily the standard for resectable Stage III; based on the NADINA study and on clinical judgement, ipi/nivo combo neoadjuvant therapy is often used instead of single agent pembro.
It is paradoxical but true; the more approved or "proven" (but not FDA approved) therapies there are, the harder it is to design the next study. What is the "standard" treatment you compare to? Will all experts/investigators be able to agreed to such a standard? Will the comparison remain meaningful through trial data completion?
There are also other historical comparison trials of single agent adjuvant immunotherapy for #melanoma, whose patient cohorts and endpoints may be more comparable.
Again, it all comes down to full data disclosure tomorrow (and the inevitable simultaneous @NEJM or other publication).
You can’t run oncology algorithms without being able to predict the next readout. The future is murky. Digital opinion leaders are using AI to cut through that murk: what the data actually say, where crossover muddies survival, and how follow-on therapy looks in real practice.
There’s also the cost of overreach from the podium. In the KRASCENDO-1 discussion, Benjamin Besse showed a ChatGPT-generated slide titled “The Last Dance of KRAS G12C – The Farewell Tour of Sotorasib and Adagrasib.” Divarasib’s wins over the first-generation inhibitors are real: PFS HR 0.58 and OS HR 0.56 (LBA5, ESMO 2026; divarasib is investigational). But the control arm detail matters.
About 91% of patients in the control arm (156 of 171) were slated for sotorasib (Lumakras) and only about 9% (15 of 171) for adagrasib (Krazati), both of which hold FDA accelerated approval in previously treated KRAS G12C NSCLC, and crossover was not permitted, a point Besse himself criticized. Calling "The Last Dance" on both drugs may be a bigger claim than the data support.
🆙#ESMO26 #LCSM Presidential Symposium II Discussion🌟
— Hidehito HORINOUCHI (@HHorinouchi) Oct 25, 2026
🔥 "The Last Dance of KRAS G12C"
🎙️ @BenjaminBesseMD
🔗 https://t.co/XAhF0PTCbx
@OncoAlert @Larvol @myESMO @IASLC @KRASKickers https://t.co/nifj1h551w https://t.co/RdhPAPudZf
🔥 KRASCENDO-1 at #ESMO26: @BenjaminBesseMD takes the stage!
— Alfredo Addeo MD (@Alfdoc2) Oct 25, 2026
As usual, a provocative, funny, and absolutely captivating discussion
Following the impressive PFS and OS results with divarasib, Benjamin challenged us to think beyond the numbers.
🎯 Key messages:
🔹 Divarasib demonstrates clear superiority over sotorasib/adagrasib, including a meaningful OS benefit.
🔹 Is this the farewell tour for first-generation KRAS G12C inhibitors? 🎸
🔹 The next frontier: moving KRAS G12C inhibition into first-line combinations with immunotherapy ± chemotherapy.
🔹 Clinical trials must evolve: reconsider crossover, better capture chronic toxicity, and broaden eligibility to reflect real-world patients.
My take: Great science deserves great discussions. @BenjaminBesseMD once again managed to combine sharp scientific criticism, humour, and a clear message ! 👏
#ESMOAmbassadors #ESMO26 #KRASG12C #Divarasib #NSCLC #LungCancer
The X community added the nuance as fast as the applause landed, from the control-arm split to the crossover design (Day 3 roundup):
#ESMO26 LBA5
— Masahiro TORASAWA, MD. PhD. (@M_Torasawa) Oct 25, 2026
🚨 Krascendo-1 (Phase 3)
Divarasib vs sotorasib/adagrasib in KRAS G12C-mutant NSCLC after platinum chemotherapy and immunotherapy (n=338).
💊 91% of controls received sotorasib (only 9% adagrasib).
✅ mPFS: 9.8 vs 5.7 mo (HR 0.58; P<0.0001)
✅ mOS: 17.7 vs 12.0 mo (HR 0.56; P=0.0005)
✅ ORR: 58.4% vs 32.3%
🛡️ Safety
• G3–4 TRAEs: 24.6% vs 34.5%
• TRAE-related discontinuation: 4.8% vs 11.7%
• Median treatment duration: 11.3 vs 5.0 mo
Impressive head-to-head PFS and OS benefits, with fewer severe TRAEs despite longer treatment exposure.
#LCSM
The last dance for KRAS inhibitors. Divarasib . Nice discussion. Systemic cross over should be reconsiderd when PFS is the primar end point. It will be intr to see future data for this drug against chemo plus IO . @OncoAlert @myESMO #esmo26 https://t.co/xOGzuBin5C
— Dr Amol Akhade (@SuyogCancer) Oct 25, 2026
Three forces are colliding: the digital opinion leader, the AI toolkit (from OpenEvidence at the bedside to ChatGPT on the discussant slide), and AI as a channel to physicians. Together they will upend medical communications in pharma sooner than most teams expect.
Hematology-oncology is the tip of the spear. The future is fast.
Pharma and agencies: the floor and the feed moved faster than you at ESMO26. The next move is simple: follow KOL Pulse and work with us if you want to keep up.
If not, catch me if you can.
Sources: JAMA / OpenEvidence use study · ESMO 2026 hub · Days 1–2 · Day 3 · INTerpath-001 · NADINA · KRASCENDO-1 · DOL to industry and FDA roles