Category: 3. Business

  • Jensen Huang says Nvidia went from 95% market share in China to 0%

    Jensen Huang says Nvidia went from 95% market share in China to 0%

    Nvidia CEO Jensen Huang urged nuance when it comes to regulating China’s access to U.S. technologies that are critical to developing artificial intelligence.

    In an interview with Citadel Securities on Tuesday, he warned that what harms China can often harm the U.S., and sometimes even in worse ways.

    “Before we leap towards policies that are hurtful to other people, take a step back and maybe reflect on what are the policies that are helpful to America,” Huang said.

    His words of caution come as Nvidia processors have become hot commodities in the AI race as well as political bargaining chips in the U.S.-China trade war.

    Huang said he’d like the world to run on U.S. know-how, but noted about half the world’s AI researchers are in China.

    “I think it’s a mistake to not have those researchers build AI on American technology,” he added.

    Trying to strike a balance between his goal of maintaining U.S. tech supremacy along with access to China will require nuance rather than an all-or-nothing approach, Huang said. But that’s not the case now, as Nvidia is “100% out of China.”

    “We went from 95% market share to 0%, and so I can’t imagine any policymaker thinking that that’s a good idea, that whatever policy we implemented caused America to lose one of the largest markets in the world,” he said.

    He didn’t name names, or administrations. But the Biden administration imposed rules in 2022 to restrict the export of Nvidia’s most advanced AI chips to China, leading the company to design a processor that met the new limits.

    In April, Nvidia said the Trump administration blocked the sale of some of its AI chips to China without licenses and would require them for future sales. Then in August, the administration granted export licenses for certain Nvidia and AMD chips to China in exchange for 15% of the revenues.

    But Chinese regulators have reportedly told domestic tech companies not to buy Nvidia chips that were designed to meet U.S. export requirements.

    Meanwhile, Beijing placed strict limits on exports of rare earths, a critical input for a wide range of advanced technologies, mimicking U.S. export rules on AI chips.

    That prompted President Donald Trump to fire back with an additional 100% tariff on Chinese goods. Officials from both sides are due to resume talks this week, ahead of a planned meeting with Trump and his Chinese counterpart later this month.

    For now, Huang told Citadel that all of Nvidia’s financial forecasts assume China will remain out of the picture.

    “If anything happens in China, which I hope it will, it’ll be a bonus,” he said. “But it’s a large market. China is the second largest computer market in the world. It is a vibrant ecosystem. I think it’s a mistake for the United States to not participate. So hopefully we’ll continue to explain and inform and hold out hope for a change in policy.”

    Continue Reading

  • Week Ahead for FX, Bonds: U.S. Inflation, PMI -2-

    Week Ahead for FX, Bonds: U.S. Inflation, PMI -2-

    For September, retail sales, a key gauge of consumption, likely grew 3.0% on year, down from August’s 3.4% increase, the poll shows. Industrial production is estimated to have grown 5.3%, marginally above August’s 5.2%. Fixed-asset investment likely stayed flat in the first three quarters of the year, compared with a 0.5% rise through August. Property data due the same day are expected to show another weak month for the housing sector.

    The People’s Bank of China will also announce the country’s benchmark lending rates on Monday, which are widely expected to remain unchanged.

    Separately, China’s ruling communist elites are set to convene a meeting from Monday to Thursday to review the country’s 15th Five-Year Plan, mapping out key policy initiatives for the world’s second-largest economy for the rest of the decade. While detailed targets will be unveiled next March, economists at Morgan Stanley expect the focus to remain on “technological self-sufficiency, innovation and national security,” with limited market-moving surprises.

    Australia / New Zealand

    In Australia, attention will be focused on further communication from the Reserve Bank of Australia. While senior officials have recently signaled more interest-rate cuts, they may start rowing back those comments after data showed unemployment jumped to its highest level since late 2021.

    Even with inflation risks lingering, the rise in unemployment to 4.5% in September adds pressure on the RBA to keep lowering the official cash rate. The increase may reflect weaker government hiring and continued softness in the private sector, compounded by global trade uncertainty and China's tariff headwinds.

    A speech by RBA Gov. Michele Bullock on Friday will be a key focus in an otherwise light data week.

    In New Zealand, third-quarter inflation data on Monday will draw close attention. Policymakers appear increasingly attuned to signs of weakness, making further rate cuts all but certain.

    Indonesia

    Bank Indonesia is set to announce its policy decision on Wednesday and is widely expected to continue cutting interest rates to support growth.

    UOB economist Enrico Tanuwidjaja thinks the easing cycle is not complete, but the end is near. He expects a 25-basis-point cut to 4.50% in October, followed by another reduction in the first quarter of 2026, with rates likely to remain steady through the year after that.

    Malaysia

    Malaysia's September inflation data is likely to show a small uptick in price pressures but not enough to move the needle for the central bank.

    ANZ expects CPI to have edged up to 1.5% from 1.3% in August, driven by slightly higher utilities and transport costs. However, with the government reduction of fuel prices, transport inflation could ease in the coming months, ANZ said.

    Overall, inflation is expected to stay subdued, supported by weak global commodity prices and moderating domestic demand. ANZ doesn't anticipate Bank Negara Malaysia to cut rates again soon unless growth weakens significantly.

    South Korea

    The Bank of Korea is expected to hold rates when the monetary policy board meets on Thursday, keeping policy settings unchanged for a third consecutive session.

    Analysts have recently pushed back forecasts for the central bank to deliver a rate cut from October to November or later, citing continued financial stability risks tied to household debt and Seoul's overheated property market. Lower borrowing costs could further stoke mortgage lending, complicating the BOK's decision.

    Goldman Sachs economists said the government's latest housing stabilization measures-tightening mortgage and property transaction rules-support the case for the BOK to hold rates in October while signaling a dovish bias for November.

    The central bank may wait for home prices to stabilize before delivering another cut, Citigroup economist Jin-Wook Kim said.

    Singapore

    Singapore will release its September inflation data on Thursday. The central bank recently said core inflation, a measure excluding private road transport and accommodation, could bottom out soon and rise gradually in 2026.

    Core inflation cooled to 0.3% on year in August from 0.5% in July. ANZ Research expects September to mark the low point of weak inflation, forecasting a 0.2% on-year rise in core prices and a 0.6% gain in headline inflation.

    Any references to days are in local times.

    Write to Jessica Fleetham at jessica.fleetham@wsj.com and Jihye Lee at jihye.lee@wsj.com

    (END) Dow Jones Newswires

    October 19, 2025 17:14 ET (21:14 GMT)

    Copyright (c) 2025 Dow Jones & Company, Inc.

    Continue Reading

  • Trodelvy Improves Survival in EGFR-Mutated Non-Small Cell Lung Cancer

    Trodelvy Improves Survival in EGFR-Mutated Non-Small Cell Lung Cancer

    Trodelvy (sacituzumab govitecan) led to a 51% reduction in the risk of progression or death compared with chemotherapy in patients with nonsquamous epidermal growth factor receptor–mutated non–small cell lung cancer that developed epidermal growth factor receptor tyrosine kinase inhibitor resistance, according to phase 3 results of the OptiTROP-Lung04 study presented during the European Society of Medical Oncology Congress 2025.

    Results showed that, at a median follow-up of 18.9 months, the median progression-free survival assessed by blinded independent central review was 8.3 months with Trodelvy and 4.3 months with chemotherapy. The 12-month progression-free survival rates were 32% and 8%, respectively. The benefit with Trodelvy was observed across all prespecified subgroups.

    The investigator-assessed median progression-free survival was 8.4 months with Trodelvy and 4.8 months with chemotherapy. The 12-month rates were 35% and 11%, respectively.

    “Trodelvy demonstrated statistically significant and clinically meaningful improvements in progression-free and overall survival compared to platinum-based chemotherapy,” lead study author Dr. Li Zhang, professor of medical oncology at Sun Yat-sen University Cancer Center in Guangzhou, China, said in an oral presentation of the data. “The results of the OptiTROP-Lung04 study support Trodelvy as a promising new treatment option for patients with EGFR-mutated non–small cell lung cancer with epidermal growth factor receptor tyrosine kinase inhibitor resistance.”

    Trodelvy is a TROP2 antibody-drug conjugate with a unique biofunctional linker that maximizes delivery of a belotecan-derivative topoisomerase I inhibitor payload to tumor cells. TROP2 is highly expressed in patients with EGFR-mutated non–small cell lung cancer, and preclinical data have shown that Trodelvy internalization and uptake are enhanced by EGFR mutations.

    Glossary

    Progression-free survival (PFS): time during and after treatment that a patient lives without cancer growing or spreading.

    Overall survival (OS): time from treatment start or diagnosis until death from any cause.

    Objective response rate (ORR): percentage of patients whose cancer shrinks or disappears after treatment.

    Disease control rate (DCR): percentage of patients whose cancer shrinks, disappears, or remains stable after treatment.

    Duration of response (DOR): length of time a treatment keeps cancer under control after it first responds.

    Investigator-assessed PFS: progression-free survival measured by the trial’s treating investigators rather than an independent review.

    Current standard options for patients who relapse on third-generation EGFR tyrosine kinase inhibitors remain platinum-based chemotherapy, but more options are needed.

    In the multicenter, open-label, phase 3 OptiTROP-Lung04 trial, 376 patients with nonsquamous stage 3B/3C or 4 non–small cell lung cancer with EGFR-sensitive mutations were randomly assigned 1:1 to receive Trodelvy at five milligrams per kilogram intravenously every two weeks or Alimta at 500 milligrams per square meter plus carboplatin area under the curve 5 or cisplatin at 75 milligrams per square meter every three weeks for up to four cycles, followed by Alimta maintenance at 500 milligrams per square meter every three weeks. Treatment was given until disease progression, intolerable toxicity, or patient request to discontinue therapy.

    To be eligible for enrollment, patients needed to have an Eastern Cooperative Oncology Group performance status of zero or one and progression after third-generation tyrosine kinase inhibitor therapy or progression after first- or second-generation tyrosine kinase inhibitors with T790M-negative mutations.

    Stratification factors included prior EGFR tyrosine kinase inhibitor therapy (third-generation in frontline versus second line versus no third-generation) or brain metastases (yes versus no).

    The primary end point was progression-free survival assessed by blinded independent central review; secondary end points were overall survival, investigator-assessed progression-free survival, objective response rate, disease control rate, duration of response, and safety.

    A total 148 patients on Trodelvy discontinued treatment due to disease progression (125 patients), patient or guardian withdrawal (12 patients), death (6 patients), side effects (2 patients), or other (3 patients). In the chemotherapy arm, 179 patients discontinued treatment due to disease progression (140 patients), patient/guardian withdrawal (16 patients), death (9 patients), side effects (5 patients), protocol deviation (2 patients), or other (7 patients). A total 69 and 102 patients in each arm, respectively, discontinued from the study due to death (67 and 101 patients) or were lost to follow-up (2 and 1 patients).

    Patient baseline characteristics were generally well balanced between the Trodelvy (188 patients) and chemotherapy arms (188 patients). The median age was 60 years and 59 years, and 31% and 27% were at least 65 years. Most had a performance status of 1 (81% and 77%), had no smoking history (77% and 72%), had stage 4 disease (97% and 98%), and at least three metastatic sites (68% and 67%). A total 18% and 19% had brain metastases, and 13% and 18% had liver metastases. The majority in each arm had exon 19 deletions (56% and 63%), had unknown T790M mutation status (59% and 60%), and received a prior third-generation EGFR tyrosine kinase inhibitor in the frontline setting (63% and 62%).

    The interim analysis showed that for Trodelvy, the median overall survival was not reached compared with 17.4 months with chemotherapy, leading to a 40% reduction in the risk of death. The 18-month overall survival rates were 66% and 48%, respectively. Overall survival was improved with Trodelvy across all prespecified subgroups.

    Dr. Zhang also reported on subsequent anticancer treatment from the trial; 72% of patients on Trodelvy and 86% of those on chemotherapy received at least one subsequent treatment. In the Trodelvy and chemotherapy groups, respectively, these included chemotherapy (42%; 54%), specifically Alimta-based chemotherapy (37%; 13%), an EGFR tyrosine kinase inhibitor (43%; 40%), an anti-angiogenic agent (35%; 48%), immunotherapy (17%; 25%), or an antibody-drug conjugate (1%; 20%).

    When assessed via blinded independent central review, the objective response rate with Trodelvy was 61% compared with 43% with chemotherapy; the disease control rate was 87% and 80%, respectively. The median duration of response was 8.3 months with Trodelvy and 4.2 months with chemotherapy; the 12-month rates were 36% versus 8%, respectively.

    Regarding safety, treatment-related side effects occurred in 100% and 98% of Trodelvy– and chemotherapy-treated patients, respectively. Grade 3 or higher side effects occurred in 58% and 54% of patients, and serious side effects occurred in 9% and 18% of patients, respectively. Side effects that led to dose reductions and interruptions occurred in 30% and 37% of patients on Trodelvy; there were no side effects that led to discontinuation or death. In the chemotherapy arm, these rates were 23% and 33%; one patient each experienced side effects leading to discontinuation or death.

    The median duration of exposure was 9.6 months with Trodelvy and 4.9 months with chemotherapy. The most common side effects in both arms were hematologic; Trodelvy had higher incidence of stomatitis that were mostly grade 1 or 2 (any-grade 62%; grade ≥3 5%). Ocular surface toxicities also occurred in 10% of patients on Trodelvy, all grade 1 or 2. No cases of interstitial lung disease or pneumonitis were reported on the Trodelvy arm.

    Phase 3 trials are currently exploring Trodelvy alone and in combination with Tagrisso in patients with EGFR-mutant non–small cell lung cancer.

    References

    1. “Sacituzumab tirumotecan versus platinum-based chemotherapy in EGFR-mutated non-small cell lung cancer following progression on EGFR-TKIs” by Dr. Zhang, et al., presented at the 2025 ESMO Congress.
    2. “Sacituzumab tirumotecan in advanced non-small-cell lung cancer with or without EGFR mutations” by Dr. Zhao, et al., Nat Med.

    For more news on cancer updates, research and education, don’t forget to subscribe to CURE®’s newsletters here.

    Continue Reading

  • ESMO 2025: Disitamab Vedotin plus Toripalimab Versus Chemotherapy in First-Line Locally Advanced or Metastatic Urothelial Carcinoma (la/mUC) with HER2-Expression – UroToday

    1. ESMO 2025: Disitamab Vedotin plus Toripalimab Versus Chemotherapy in First-Line Locally Advanced or Metastatic Urothelial Carcinoma (la/mUC) with HER2-Expression  UroToday
    2. Loqtorzi Extends Survival in HER2-Positive Advanced Urothelial Cancer  CUREtoday.com
    3. Disitamab Vedotin/Toripalimab Combo Prolongs Survival in Frontline HER2+ Urothelial Cancer  CancerNetwork
    4. RC48-C016: DV + Toripalimab Boosts Survival in HER2+ UC  Oncodaily

    Continue Reading

  • Tiragolumab/Atezolizumab/Bevacizumab Fails to Significantly Improve PFS in Advanced/Metastatic HCC

    Tiragolumab/Atezolizumab/Bevacizumab Fails to Significantly Improve PFS in Advanced/Metastatic HCC

    Tiragolumab (MTIG7192A) plus atezolizumab (Tecentriq) and bevacizumab (Avastin) did not provide a statistically significant progression-free survival (PFS) benefit vs placebo plus atezolizumab and bevacizumab in patients with untreated locally advanced or metastatic hepatocellular carcinoma (HCC), according to findings from the phase 3 IMbrave152/SKYSCRAPER-14 study (NCT05904886) that were presented during the 2025 ESMO Congress.

    In turn, the coprimary end point of investigator-assessed PFS was missed. Patients treated with placebo plus atezolizumab and bevacizumab (n = 338) had an investigator-assessed median PFS of 8.2 months (95% CI, 7.0-9.7) vs 8.3 months (95% CI, 7.1-9.8) in patients treated with tiragolumab plus atezolizumab plus bevacizumab (HR, 0.97; 95% CI, 0.8-1.2; stratified log-rank P =.7464).

    Overall survival (OS) data remain immature and are not anticipated to reach statistical significance, according to study authors.

    “These data, now from a double-blind, placebo-controlled study of over 600 patients, confirm the activity of atezolizumab plus bevacizumab in the management of this population, as well as the safety, and moving forward, we will hopefully learn a lot from the study on how to best decide to move forward in advancing care for our patients with HCC,” study author and presenter, Richard S. Finn, MD, stated.

    Finn is a professor of medicine at the David Geffen School of Medicine, UCLA, and director of the UCLA Liver Cancer Program.

    Design and Baseline Characteristics of the IMbrave152/SKYSCRAPER-14 Study

    The IMbrave152/SKYSCRAPER-14 study is a double-blind, placebo-controlled, randomized, global trial that enrolled 669 patients between September 14, 2023, and September 23, 2024. Eligible patients were 18 years of age or older with unresectable locally advanced or metastatic HCC, an ECOG performance status of 0 or 1, and Child-Pugh A liver function.

    Patients were excluded from the study if they had received prior systemic therapy for advanced disease or experienced recurrence within 26 months of completing adjuvant treatment. Patients were randomly assigned 1:1 to receive either tiragolumab at 600 mg plus atezolizumab at 1200 mg plus bevacizumab at 15 mg/kg intravenously (IV) every 3 weeks, or placebo plus atezolizumab at 1200 mg and bevacizumab at 15 mg/kg IV every 3 weeks.

    Treatment continued until loss of clinical benefit or unacceptable toxicity, and crossover was not allowed. Imaging assessments were performed every 6 weeks, with a 1-week margin. Stratification factors included geographic region (Asia and Africa vs the rest of the world, including Japan), macroscopic vascular invasion (MVI) and/or extrahepatic spread (EHS; presence versus absence), baseline AFP level (≥ 400 ng/mL vs < 400 ng/mL), and HCC etiology (viral vs non-viral).

    The study’s primary end points were investigator-assessed PFS per RECIST 1.1 criteria and OS. Key secondary end points included objective response rate (ORR), duration of response (DOR), PFS and OS rates at select time points, safety, and patient-reported outcomes (PROs).

    The baseline characteristics were generally well balanced between the tiragolumab and the placebo arms. The median age for all patients was 44.5 years, with 82.2% of patients aged 65 years or older. The majority of patients were male (87.6% in both arms).

    The most common geographic region was Asia/Asian Pacific/Australia (tiragolumab, 52.0%; placebo, 48.1%), followed by Europe and Middle East/North America (18.7%; 17.5%). Hepatitis B was the most frequent HCC etiology (65.6%; 59.8%), whereas 76.1% of all patients had an ECOG performance status of 0 or 1. Most patients had Barcelona Clinic Liver Cancer (BCLC) stage B or C disease (75.9% overall) and a Child-Pugh score of A5 or A6 (63.5% overall).

    Other key characteristics were similar: macrovascular invasion and/or extrahepatic spread was present in 36.7% of all patients, AFP levels of 400 ng/mL or higher were seen in 25.9% of all patients, and 38.9% of the overall population had received prior local cancer therapy.

    Exploratory Subgroup Analysis

    An exploratory subgroup analysis of 909 patients showed that adding tiragolumab to atezolizumab plus bevacizumab was numerically associated with an improvement in median investigator-assessed PFS of 8.3 months vs 8.2 months (HR, 0.98; 95% CI, 0.8-1.2) compared with the atezolizumab plus bevacizumab arm.

    Treatment benefit varied across prespecified baseline characteristics. Patients with baseline AFP levels less than 400 ng/mL appeared to derive the greatest benefit (HR, 0.74; 95% CI, 0.6-1.0), with a median investigator-assessed PFS of 9.8 months vs 5.5 months in the control arm. Conversely, patients with an ECOG performance score of 1, those in the Americas excluding Japan, those with high MVI and/or EHS (HR, 1.18; 95% CI, 0.9-1.5), and those with a Child-Pugh score A8 (HR, 1.18; 95% CI, 0.8-1.7) appeared to have a diminished or unfavorable treatment effect.

    The median investigator-assessed PFS was longer in the experimental arm for nearly all subgroups, with notable exceptions including HCC etiology of hepatitis C (7.7 months vs 12.8 months) and the Child-Pugh A8 subgroup (4.7 months vs 5.7 months).

    Understanding Additional Outcomes of the IMbrave152/SKYSCRAPER-14 Study

    The ORR was 29.9% for the tiragolumab group with a complete response (CR) rate of 2.1% and a partial response (PR) rate of 27.8%. In the placebo group, the ORR was 26.0%, the CR rate was 1.8%, and the PR rate was 24.3%.

    Among patients receiving the tiragolumab combination, there were 99 responders. Of these, 33.3% of patients experienced a subsequent event. The median DOR was 15.0 months (range,13.9-not estimable [NE]).

    Among those receiving placebo, there were 88 responders. Of these, 38.6% experienced a subsequent event. The median DOR was 13.2 months (range, 10.1-NE).

    Stable disease was reported in 48.0% of patients in the tiragolumab group vs 48.8% of those in the placebo group; progressive disease was reported in 16.3% vs 18.9% of patients, respectively, and the disease control rates were 77.9% vs 74.9%, respectively.

    How Safe Is Tiragolumab Plus Atezolizumab and Bevacizumab?

    The safety profile of the tiragolumab plus atezolizumab and bevacizumab combination (n = 332) was similar to that of the placebo plus atezolizumab and bevacizumab regimen (n = 333), though patients in the tiragolumab arm experienced slightly higher rates of adverse effects (AEs).

    Nearly all patients in both arms reported AEs of any cause (tiragolumab arm, 98.8%; placebo arm, 97.6%). Grade 3/4 AEs were observed in 53.6% of patients in the tiragolumab-treated group, and grade 5 AEs were seen in 8.4% of patients in this arm. The rates of grade 3/4 and grade 5 AEs in the placebo arm were 47.7% and 7.2%, respectively.

    Serious AEs occurred in 45.8% of patients in the tiragolumab arm and 38.4% of those in the placebo arm. Drug-related AEs were reported in 92.8% and 87.1% of patients, respectively. AEs led to withdrawal of tiragolumab in 73.8% of patients receiving it, with 18.4% of those patients requiring systemic corticosteroids. AEs resulted in bevacizumab withdrawal in 64.5% of patients in the tiragolumab group and 60.7% of those in the placebo group.

    Disclosures: Finn reported serving as a Principal Investigator for Merck/Eisai and Roche. He has received research funding from Bristol-Myers Squibb, Merck/Eisai, Pfizer, and Roche/Genentech. His roles as a speaker, consultant, or advisor involve relationships with AstraZeneca, Bristol-Myers Squibb, Chugai, Guerbet, Merck/Eisai, Novartis, Pfizer, Roche/Genentech, and Zai Labs. He is also a steering committee member for Bristol-Myers Squibb, Merck/Eisai, and Roche.

    Reference

    Finn RS, Singal AG, Cheng A, et al. IMbrave152/SKYSCRAPER-14: a phase III study of first-line tiragolumab + atezolizumab + bevacizumab vs placebo + atezolizumab + bevacizumab for patients with untreated locally advanced or metastatic hepatocellular carcinoma. Presented at: 2025 ESMO Congress; October 19, 2025; Berlin, Germany. Abstract LBA50.

    Continue Reading

  • Amazon (AMZN) Stock in Focus as BofA Names It Top Large-Cap E-Commerce Pick

    Amazon (AMZN) Stock in Focus as BofA Names It Top Large-Cap E-Commerce Pick

    Amazon.com, Inc. (NASDAQ:AMZN) is one of the AI Stocks in the Spotlight This Week. On October 17, Bank of America reiterated Amazon and Chewy as “Buy,” stating that both stocks are top ideas in e-commerce.

    “Amazon remains our top Large Cap pick for US eCommerce given projected share gains aided by a growing grocery business, margin expansion from robotics, ability to leverage Prime user base to build strong Agentic AI position, and our view that capacity additions in 2026 will drive AWS acceleration.”

    Amazon.com Inc. (AMZN) is an American technology company offering e-commerce, cloud computing, and other services, including digital streaming and artificial intelligence solutions.

    While we acknowledge the potential of AMZN as an investment, we believe certain AI stocks offer greater upside potential and carry less downside risk. If you’re looking for an extremely undervalued AI stock that also stands to benefit significantly from Trump-era tariffs and the onshoring trend, see our free report on the best short-term AI stock.

    READ NEXT:10 AI Stocks in Focus on Wall Street and 10 AI Stocks Analysts Are Watching Closely

    Disclosure: None.

    Continue Reading

  • Andrej Karpathy Say It Will Take a Decade for Agents to Actually Work

    Andrej Karpathy Say It Will Take a Decade for Agents to Actually Work

    Even in the fast-moving world of AI, patience is still a virtue, according to Andrej Karpathy.

    The OpenAI cofounder, and de facto leader of the vibe-coding boom, appeared on the Dwarkesh Podcast last week to talk about how far we are from developing functional AI agents.

    TL;DR — he’s not that impressed.

    “They just don’t work. They don’t have enough intelligence, they’re not multimodal enough, they can’t do computer use and all this stuff,” he said. “They don’t have continual learning. You can’t just tell them something and they’ll remember it. They’re cognitively lacking and it’s just not working.”

    “It will take about a decade to work through all of those issues,” he added.

    Agents are among the most talked-about innovations in AI, with many investors dubbing 2025 “the year of the agent.” While definitions vary, agents are virtual assistants capable of completing tasks autonomously — breaking down problems, outlining plans, and taking action without user prompts.

    Karpathy is a famously fast talker. So he wrote a follow-up post on X for listeners who couldn’t quite parse everything he said. On the topic of agents, he reiterated his earlier frustrations.

    “My critique of the industry is more in overshooting the tooling w.r.t. present capability,” he wrote. “The industry lives in a future where fully autonomous entities collaborate in parallel to write all the code and humans are useless.”

    He doesn’t want to live there.

    In Karpathy’s ideal future, humans and AI collaborate to code and execute tasks.

    “I want it to pull the API docs and show me that it used things correctly. I want it to make fewer assumptions and ask/collaborate with me when not sure about something. I want to learn along the way and become better as a programmer, not just get served mountains of code that I’m told works,” he wrote.

    The con of building the kind of agents that render humans useless, he said, is that humans are then useless, and AI “slop,” the low-quality content generated by AI, becomes ubiquitous.

    Karpathy isn’t the only one to raise concerns about the functionality of AI agents.

    In a post on LinkedIn last year, ScaleAI growth lead Quintin Au talked about how the errors agents make are compounded with every additional task they take on.

    “Currently, every time an AI performs an action, there’s roughly a 20% chance of error (this is how LLMs work, we can’t expect 100% accuracy),” he wrote in a post on LinkedIn. “If an agent needs to complete 5 actions to finish a task, there’s only a 32% chance it gets every step right.”

    While skeptical of the current state of AI agents, Karpathy said he isn’t an AI skeptic.

    “My AI timelines are about 5-10X pessimistic w.r.t. what you’ll find in your neighborhood SF AI house party or on your twitter timeline, but still quite optimistic w.r.t. a rising tide of AI deniers and skeptics,” he said.


    Continue Reading

  • Sirexatamab Plus Bevacizumab/Chemo Improves Outcomes in DKK1-High mCRC

    Sirexatamab Plus Bevacizumab/Chemo Improves Outcomes in DKK1-High mCRC

    Sirexatamab (DKN-01) in combination with bevacizumab (Avastin) and standard-of-care (SOC) chemotherapy demonstrated a manageable safety profile and was well tolerated vs bevacizumab plus SOC chemotherapy in patients with metastatic colorectal cancer (mCRC), according to findings from the phase 2 DeFianCe trial (NCT05480306) presented at the 2025 ESMO Congress. Notably, patients with DKK1-high tumors derived improvements in both progression-free survival (PFS) and overall survival (OS).

    In the intent-to-treat population, patients who received the DKK1 inhibiting IgG4 antibody sirexatamab demonstrated an overall response rate (ORR) of 35.1% (95% CI, 25.5%-45.6%) vs 26.6% (95% CI, 18.0%-36.7%) for the control arm (P = .10). Median PFS was 9.2 months in the experimental arm vs 8.3 months in the control arm (P = .17).

    “Neither of these 2 outcomes reached statistical significance,” lead author Zev A. Wainberg, MD, said during the presentation of data. “However, looking at the DKK1-high population, defined as the upper median, the improvement started to bear fruit,” Wainberg, professor of medicine at UCLA and codirector, UCLA GI program, at UCLA Health in California continued.

    There were 50 patients with higher DKK1 in the sirexatamab arm and 38 in the control arm. The ORR in the sirexatamab arm was 38.0% (95% CI, 24.7%-52.8%) vs 23.7% (95% CI, 11.4%-40.2%) in the control arm (P = .0706). Median PFS in the experimental arm was 9.03 months vs 7.06 months (HR, 0.61; 95% CI, 0.37-1.0; P = .0255). Median OS was not reached in the experimental arm vs 14.39 months in the control arm (HR, 0.42; 95% CI, 0.19-0.91; P = .0118).

    What Was the DeFianCe Trial Design?

    After an initial safety run-in period (n = 33), a total of 188 patients with mCRC were randomly assigned 1:1 to receive either sirexatamab plus either leucovorin, 5-fluorouracil (5-FU), and irinotecan (FOLFIRI) or leucovorin, 5-FU, and oxaliplatin (FOLFOX) and bevacizumab (n = 94) vs FOLFIRI or FOLFOX and bevacizumab (n = 94).

    Patients were eligible if they had received 1 prior 5-FU–based therapy, had microsatellite stable (MSS) CRC, and did not have a BRAF V600 mutation. The stratification factors were tumor sidedness (left vs right) and prior antiangiogenesis therapy (yes vs no).

    The primary end point was investigator-assessed PFS and the secondary end points were safety, ORR, and OS. Key exploratory end points were PFS, ORR, and the OS of the DKK1-high subgroup.

    What Were the Patient Baseline Characteristics?

    The study was well-balanced for gender and region. The majority of patients were male across the arms (experimental, 68%; control, 55%). In the experimental arm, the majority of patients were from South Korea (53%) compared with 48% in the control arm. Forty-four percent of patients were from the United States in the experimental arm and 43% in the control arm. The majority of patients had left-sided tumors (both 75%).

    “This was a predominantly liver-metastatic patient population, with nearly 75% in both arms having liver metastatic disease,” Wainberg said. Further, 47% of patients in the experimental arm had RAS-mutated disease compared with 57% in the control arm. Looking at prior systemic therapy, 48% of patients in the experimental arm had received antiangiogenesis therapy compared with 51% in the control arm.

    What Was the Safety Profile?

    The overall rate of treatment-emergent adverse events (TEAEs) for sirexatamab was similar to the chemotherapy arm, suggesting that the addition of the agent does not adversely impact the safety profile of the combined agents.

    In the experimental arm, 15% of patients discontinued therapy compared with 19% in the control arm. In the patients with DKK1-high status, 4% discontinued in the treatment arm vs not applicable in the control arm. Dose reductions affected 37% of patients in the experimental arm compared with 43% in the control arm. Dose interruptions were similar, with 73% in the experimental arm and 71% in the control arm.

    “These data support continued development of sirexatamab in DKK1-high previously treated patients with mCRC,” Wainberg concluded.

    Disclosures: Dr Wainberg disclosed that he has provided consulting services to Alligator Therapeutics, Amgen, AstraZeneca, Arcus, Boehringer Ingelheim, Bristol Myers Squibb, Daiichi Sankyo Company, Eli Lilly and Company, EMD Serono, Roche AG, Genentech, Ipsen, Johnson & Johnson, Merus NV, Merck, Novartis, Novocure, Pfizer, Servier, and Verastem.

    Reference

    Wainberg ZA, Han S-W, Kim JG, et al. DeFianCe Trial: A randomized phase 2 trial of sirexatamab (DKN-01) plus bevacizumab and chemotherapy versus bevacizumab and chemotherapy as second-line therapy in advanced microsatellite stable (MSS) colorectal cancer (CRC). Presented at: 2025 European Society for Medical Oncology Congress; October 17-21, 2025; Berlin, Germany. Abstract LBA34.

    Continue Reading

  • Trastuzumab Rezetecan Improves PFS Vs SOC in HER2+ Breast Cancer | Targeted Oncology

    Trastuzumab Rezetecan Improves PFS Vs SOC in HER2+ Breast Cancer | Targeted Oncology

    The novel HER2-targeted antibody-drug conjugate (ADC) trastuzumab rezetecan (SHR-A1811) reduced the risk of disease progression or death by 78% compared with the standard combination of pyrotinib (Irene) plus capecitabine in patients with previously treated HER2-positive advanced breast cancer, according to findings from the phase 3 HORIZON-Breast01 trial (NCT05424835) presented at the 2025 ESMO Annual Congress.1

    At an overall median follow-up of about 15.5 months, the median progression-free survival (PFS) by blinded independent central review (BICR) was 30.6 months (95% CI, 16.8–not reached) in the trastuzumab rezetecan arm compared with 8.3 months (95% CI, 6.9–11.0) in the combination arm (HR, 0.22; 95% CI, 0.15–0.34; P <.0001). The 12-month PFS rates were 84.7% vs 35.5%, respectively.

    The median investigator-assessed PFS was 33.3 months vs 8.1 months, respectively (HR, 0.16; 95% CI, 0.10–0.25). Per the investigators, the 12-month PFS rate was 86.7% vs 36.0% in the experimental vs combination arms, respectively.

    The median PFS benefit with trastuzumab rezetecan was upheld across all predefined patient subgroups. Of note, the benefit was sustained regardless of prior pertuzumab (Perjeta) treatment and regardless of the number of prior lines of therapy.

    The overall survival (OS) data remained immature at the time of the data cutoff; however, there was a trend toward an OS benefit with trastuzumab rezetecan. The 12-month OS rate was 96.3% with the ADC vs 88.4% with the combination (HR, 0.31; 95% CI, 0.14–0.69).

    “Notably, 50.3% of patients in the combination arm received an anti-HER2 ADC as post-study anticancer treatment, suggesting the initial trend we observed with the OS benefit [for trastuzumab rezetecan] might be meaningful,” said presenting author Erwei Song, MD, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China.

    The objective response rate (ORR) was 81.7% vs 55.9% in the ADC vs combination arms, respectively. The complete response, partial response, and stable disease rates were 4.2% vs 2.8%, 77.5% vs 53.15%, and 14.1% vs 31.0%, respectively. The disease control rate was 95% vs 86.9% and the progressive disease rate was 0.7% vs 9%, respectively. The median duration of response was 27.8 months vs 10.9 months, respectively.

    The median treatment duration was 19.5 months for the ADC, 7.1 months for pyrotinib, and 7.5 months for capecitabine, indicating the ADC was well tolerated. Regarding safety, 13.4% of the ADC arm experienced serious treatment-emergent adverse events (TEAEs) vs 11.8% of the combination arm. The rate of discontinuations due to TEAEs was 4.9% vs 1.4%, respectively.

    Song also noted that the majority of patients in the ADC arm had hematologic toxicities, with about 89.4% of patients experiencing neutrophil count decrease across all grades compared with 45.1% in the combination arm. He also noted that only 2.8% of the ADC arm experienced interstitial lung disease (ILD), including only 1 patient with grade 3 ILD.

    “Trastuzumab rezetecan exhibited a significant PFS benefit and a strong trend in OS vs pyrotinib plus capecitabine in patients with HER2-positive advanced/metastatic breast cancer previously treated with trastuzumab and a taxane,” Song said.

    What Was the Study Design of the HORIZON-Breast01 Trial?

    The open-label, multicenter phase 3 HORIZON-Breast01 study enrolled 287 patients with HER2-positive unresectable or metastatic breast cancer who had prior treatment with a taxane and trastuzumab (Herceptin) in the advanced setting.

    Patients were randomized to trastuzumab rezetecan (n = 142; 4.8 mg/kg IV on day 1 of 21-day cycles) or the combination of pyrotinib (n = 145; 400 mg once daily oral on days 1–21 of 21-day cycles) and capecitabine (1000 mg/m2 orally twice daily on days 1–14 of 21-day cycles). Treatment was administered until disease progression, patient withdrawal, unacceptable toxicity, or investigator decision.

    The primary outcome measure was PFS per BICR. Secondary end points included PFS per investigator assessment, OS, ORR, duration of response, and safety.

    What Were the Patient Characteristics in the HORIZON-Breast01 Trial?

    The study enrolled 287 eligible patients with HER2-positive breast cancer between August 4, 2022, and August 9, 2024. The data cutoff date was June 30, 2025.

    Across the overall study population, the median age was 56 years (range, 27–74), about three-fourths of patients had IHC 3+ HER2 status and the remainder had IHC 2+ and ISH+ status. Almost half of patients were hormone receptor-positive. About three-fourths of patients in the trial had visceral metastases and about 47% had more than 3 organs with tumor metastases. The ECOG performance status (PS) was evenly split between 0 and 1 in the trastuzumab rezetecan arm, while in the combination arm, 58% were ECOG PS 0 and 42% were ECOG PS 1.

    The median number of prior treatment lines across all patients was 1 (range, 1–4). Most patients had received 1 prior line of therapy at 83.8% and 76.6% in the trastuzumab rezetecan and combination arms, respectively. In the experimental arm, 39.4% of patients had primary resistance to trastuzumab compared with 44.8% of patients in the combination arm. All patients had prior taxane therapy and all except 3 patients had prior trastuzumab. Three-fourths of patients overall had prior pertuzumab. Prior trastuzumab emtansine and endocrine therapy had been received by 4.9% vs 6.9% and 29.6% vs 28.3% of the 2 arms respectively.

    In his concluding remarks, Song said “Trastuzumab rezetecan represents a promising practice-changing therapeutic alternative in this patient population.”

    REFERENCE:
    Song E, Yao H, Li H, et a. LBA19 – SHR-A1811 versus pyrotinib plus capecitabine in human epidermal growth factor receptor 2-positive (HER2+) advanced/metastatic breast cancer (BC): A multicenter, open-label, randomized, phase III study (HORIZON-Breast01). Results of a phase II study. Presented at: 2025 ESMO Congress; October 17-21, 2025; Berlin, Germany. Abstract LBA19.

    Continue Reading

  • Tislelizumab Plus Chemo/Concurrent Chemoradiotherapy Yields Survival Advantage in Locally Advanced ESCC

    Tislelizumab Plus Chemo/Concurrent Chemoradiotherapy Yields Survival Advantage in Locally Advanced ESCC

    The addition of tislelizumab-jsgr (Tevimbra) to induction chemotherapy and concurrent chemoradiotherapy (CRT) boasted higher composite complete response (cCR) rates, which translated to improved survival vs the historical control of concurrent CRT alone, in patients with esophageal squamous cell carcinoma (ESCC), according to data from the phase 2 EC-CRT-002 trial (NCT05520619) presented at the 2025 ESMO Congress.1

    Patients were randomly assigned to group A (n = 57) or group B (n = 57). Investigators also included a historical control group for comparison consisting of 55 patients from a phase 2 trial (NCT02403531) who experienced a 1-year PFS rate of 56% with concurrent CRT alone. Patients in group A received induction chemotherapy consisting of paclitaxel plus cisplatin every 3 weeks for 2 cycles; concurrent CRT with intensity modulated radiation therapy at 50.4 Gy over 28 fractions and paclitaxel plus cisplatin every week for 5 cycles; and tislelizumab at 200 mg every 3 weeks for a total of 16 cycles (induction, n = 2; concurrent, n = 2; adjuvant, n = 12). Patients in group B received the same treatment schedule with the exception that tislelizumab was only administered for a total of 4 cycles (induction, n = 2; concurrent, n = 2).

    Sidebar: Notable Findings From the EC-CRT-002 Trial

    • Two different schedules of tislelizumab plus induction chemotherapy and concurrent CRT led to cCR rates of 51% and 70% vs the historical control of 40% with concurrent CRT alone in patients with ESCC.
    • The 1-year PFS rates with the 2 tislelizumab regimens were 52% and 73% vs 56% with the control.
    • Treatment-related adverse effecrs included lymphopenia, esophagitis, anemia, leukopenia, neutropenia, pneumonitis, esophageal fistula, alanine aminotransferase/aspartate aminotransferase level elevation, and rash.

    The 1-year progression-free survival (PFS) rate in group A was 52% vs 56% with the control (P = .52). The 1-year PFS rate in group B was 73% vs 56% with the control (P = .024). The 1-year overall survival (OS) rates were 82% and 69% in group A and the control arm, respectively (P = .32). The respective 1-year OS rates in group B vs the control arm were 85% and 69% (P = .004).

    “Tislelizumab plus induction chemotherapy and concurrent CRT demonstrated superior efficacy and manageable toxicity compared [with] chemoradiotherapy alone in locally advanced ESCC,” Mian Xi, MD, lead study author and chief physician of radiation oncology at Sun Yat-Sen University Cancer Center in Guangzhou, China, said in a presentation.

    What Is the Background of the EC-CRT-002 Trial?

    Definitive CRT is the preferred standard of care (SOC) for patients with locally advanced ESCC. Findings from a prior randomized phase 2 trial (NCT02403531) led by Xi failed to show an OS benefit with the addition of induction chemotherapy to definitive CRT vs definitive CRT alone in patients with ESCC.2 However, patients who responded to induction chemotherapy displayed improved OS vs those in the CRT alone arm.

    As such, investigators theorized that combining anti–PD-1 therapy with induction chemotherapy and concurrent CRT could improve OS, resulting in a new SOC for patients with locally advanced disease.

    This served as the basis for the present multicenter, randomized, parallel-group, phase 2 trial. To be eligible for enrollment, patients had to have newly diagnosed, unresectable, locally advanced, histologically confirmed ESCC; be between the ages of 18 and 70 years; have an ECOG performance status of 0 to 2; and have no history of a concomitant or prior malignancy.

    The primary end point was PFS. Secondary end points were OS, cCR, safety, and quality of life.

    Between October 2022 and October 2024, 114 patients were randomly assigned to receive treatment. The data cutoff was July 31, 2025.

    What Did the Demographic Information Reveal About the Patients in the EC-CRT-002 Trial?

    Baseline characteristics in group A (n = 57) indicated that most patients were male (84%) and had upper (49%) or middle (37%) as opposed to distal (14%) tumors. The median age was 61 years (range, 38-70), and represented cTNM stages were II (5%), III (33%), IVA (39%), and IVB (23%). In group B (n = 57), most patients were male (75%) and had upper (42%) or middle (42%) vs distal (16%) tumors. The median age was 61 years (range, 37-70), and cTNM stages included II (4%), III (32%), IVA (39%), and IVB (26%).

    What Other Efficacy Data Were Presented From the EC-CRT-002 Trial, and What Was the Safety Profile of the Regimen?

    The cCR rates were 51% and 40% in group A vs the control arm (P = .334). Patients in group B and the control arm experienced cCR rates of 70% and 40%, respectively (P = .003).

    Treatment-related adverse effects included lymphopenia (74.6%), esophagitis (16.7%), anemia (14.9%), leukopenia (13.2%), neutropenia (8.8%), pneumonitis (2.6%), esophageal fistula (2.6%), alanine aminotransferase/aspartate aminotransferase level elevation (1.8%), and rash (0.9%).

    Exploratory analysis revealed that patients with PD-L1 combined positive scores (CPS) of 1 or greater had improved PFS vs those with PD-L1 CPS below 1 (P = .036). Moreover, patients with higher levels of CD8 positivity had improved PFS (P = .017). Patients who tested negative for circulating tumor DNA also experienced improved survival (P < .001).

    “[The] benefit of adjuvant immunotherapy after definitive CRT remains uncertain,” Xi concluded.

    Disclosures: Xi had no disclosures to declare.

    References

    1. Xi M, Chen B, Liu S, et al. Tislelizumab combined with induction chemotherapy and concurrent chemoradiotherapy in locally advanced esophageal squamous cell carcinoma: a multicenter, randomized, phase II trial (EC-CRT-002). Presented at: 2025 ESMO Congress; October 17-21, 2025; Berlin, Germany. Abstract LBA82.
    2. Liu S, Luo L, Zhao L, et al. Induction chemotherapy followed by definitive chemoradiotherapy versus chemoradiotherapy alone in esophageal squamous cell carcinoma: a randomized phase II trial. Nat Commun. 2021;12(1):4014. doi:10.1038/s41467-021-24288-1

    Continue Reading