Welcome back to This Week in Biotech by Biotech Blueprint, edition 118, covering biotech and pharma news from September 11th to 17th, 2026.
🎙️ Biotech Blueprint brings you weekly video updates on the latest biotech and pharma news, plus in-depth podcast interviews with industry leaders. You can find us on YouTube, Spotify, and Apple Podcasts. If you want to be featured on our podcast, please get in touch.
📩 For consulting inquiries, please email me.
👥 If you haven’t already, connect with me on LinkedIn.
🛒 And make sure to check out our MERCH STORE.
VIDEO SUMMARY
THIS WEEK’S KEY TAKEAWAYS 🔑
The World Conference on Lung Cancer produced excellent relapsed small cell lung cancer data, and in the same week, the sharpest public argument yet about whether that kind of data can be believed. Two B7-H3 antibody drug conjugates cut the risk of death 54% against topotecan. But both trials ran in China only. Two sessions away, Gilead and Merck presented a frontline lung cancer study where the combination arm did slightly worse than the control overall (hazard ratio 1.07) and then showed that among Chinese enrollees the same combination cut deaths 45%. This doesn’t mean the Chinese data is wrong or fraudulent. The more likely explanation is that a China-only result is a promising signal, but companies have been buying these assets as though the signal were proof.
Then Summit presented a rebuttal at the same meeting. Their ivonescimab is the drug people have been skeptical about for exactly this reason. This week they reported updated global HARMONi data, and it held up in Western patients too. Hazard ratio 0.76 overall and 0.76 in the Western subgroup. This confirmation took an extra 14 months of follow up to get. Everyone else is still asking to be trusted. GSK committed up to $750 million to another Chinese trispecific this week, and five large drugmakers confirmed they are handing records to a House committee investigating trials run at Chinese military hospitals.
Novartis lost a fourth trial in a month, and separately lost something more durable. The FDA approved the first “radioligand equivalent,” a 505(b)(2) copy of Lutathera, eight years after the original. Novartis directs about 40% of its oncology research budget at radioligand therapy. So the first real threat to that franchise isn’t a better drug, but a company using a faster approval route to make a copy.
BIOTECH/PHARMA NEWS 🧬
🔹 In KEYNOTE-D46/EVOKE-03, Trodelvy plus Keytruda produced a hazard ratio of 1.07 against Keytruda alone in frontline PD-L1-expressing non-small cell lung cancer, meaning the combination did slightly worse than the control, and the study was stopped. An exploratory analysis then found a 45% reduction in death risk among patients enrolled in China and 37% across East Asia. The drug and protocols were the same but the results differed depending on geography. Evercore ISI’s Umer Raffat said the result shows “why [the] Street has so much emphasis on U.S. trials,” and Leerink’s Daina Graybosch warned the preponderance of evidence points to China studies producing better survival than their global counterparts. Summit presented the rebuttal at the same meeting. Updated HARMONi data showed median overall survival of 16.8 versus 14 months, hazard ratio 0.76 (95% CI 0.61 to 0.95, p=0.0151), with the Western subgroup landing at the same 0.76 and 17.5 versus 14 months. The original April 2025 analysis had missed at 0.79, so it took fourteen extra months of follow up to confirm. Meanwhile GSK committed up to $750M to Chimagen for a myeloma trispecific, and Representative John Moolenaar said five large drugmakers are cooperating with his probe of trials at Chinese military hospitals.
🔹 Two B7-H3 antibody drug conjugates posted near identical wins against topotecan in relapsed small cell lung cancer. An antibody drug conjugate uses an antibody to find the tumor and deliver a chemotherapy payload directly to it, rather than flooding the whole body. Roche and MediLink’s tambotatug pelitecan, in TAISHAN-302 (n=451), reached median overall survival of 13.3 versus 9.4 months, hazard ratio 0.46 (95% CI 0.35 to 0.62, p<0.0001), with progression free survival of 7.4 versus 2.8 months and a response rate of 59.1% versus 9.7%. GSK and Hansoh’s risvutatug rezetecan, in the China-only ARTEMIS-008 (n=461), reached 18.5 versus 10.3 months, also hazard ratio 0.46, with progression free survival of 7.2 versus 3 months and a 58.3% response rate. Grade 3 or higher treatment-related adverse events were lower than topotecan in both (46.4% versus 74.7%, and 60.9% versus 78.2%), and interstitial lung disease ran 4.9% for tam-peli. Relapsed small cell lung cancer has had essentially nothing for thirty years, so a six fold response rate is the largest step the disease has taken in a generation. Both were presented by Chinese investigators from Chinese sites, which is the whole argument in one slide.
🔹 Novartis discontinued lifonebart (VHB937), a TREM2-stabilizing antibody, after the 251-patient ASTRALS Phase 2 trial missed both primary and secondary endpoints in amyotrophic lateral sclerosis (ALS). The primary endpoint was a composite of ventilator-free survival and change on the ALS Functional Rating Scale. An Alzheimer’s study continues, and full data goes to the ALS/MND symposium in Amsterdam on December 9-11. That’s the fourth failure in a month, after pelacarsen, del-desiran and the rap-cel holds. The more structural loss came a day earlier, when the FDA approved Curium’s Bexlutry, a lutetium Lu 177 dotatate injection cleared through the 505(b)(2) pathway as the first “radioligand equivalent” to Lutathera. That pathway lets a company rely partly on the original drug’s existing safety and efficacy data rather than running the full program from scratch. Novartis has sold Lutathera without competition since January 2018. Pluvicto and Lutathera together did about $2.8B last year and absorb roughly 40% of the oncology research budget. Novartis also paid $125M to Sironax for a blood-brain barrier delivery platform.
🔹 The FDA opened the Expedited Investigational New Drug Pilot on September 15 under “Operation Trialblazer,” offering 8 to 10 pairings between drugmakers and qualified academic centers or contract research organizations, with applications due Oct 30. The new feature is rolling review, where the agency evaluates pieces of an application as they arrive rather than waiting for a complete package, aiming to surface clinical hold issues early. Starting a trial in the US can take up to 2 years, against under 70 days in Australia, and Chinese trial registrations have gone from about 1,000 a year in 2010 to more than 5,000 in 2024 while the American figure sat near 3,500. The same week, STAT reported the FDA has secured 600 of the more than 2,200 hires it’s trying to make, with hundreds stuck in pre-onboarding because the centralized human resources center DOGE created is itself understaffed. Rolling review requires reviewers looking at submissions continuously rather than in one batch, which takes staff the agency doesn’t currently have.
🔹 The FDA approved ISEMBYLD (apitegromab-mstn) on Sept 13 for spinal muscular atrophy patients two and older who are already receiving an SMN2-targeted therapy. Spinal muscular atrophy is caused by a genetic defect that kills the motor neurons that signal muscles to contract, and every existing drug works on the neuron. This one works on the muscle instead, by blocking latent myostatin, a protein that limits muscle growth. It’s strictly an add-on. In the 188-patient SAPPHIRE trial the recommended 10 mg/kg dose produced a 2.2 point advantage on the Hammersmith Functional Motor Scale-Expanded (nominal p=0.0121), with 34.2% of treated patients gaining at least three points versus 13.5% on placebo. Fractures occurred in 9% versus 2%. Barclays downgraded the stock anyway and shares fell about 2.5% on the week. The skepticism is because an add-on layered on top of nusinersen or risdiplam has to justify a third payer conversation for a two point motor gain, and Scholar Rock hasn’t disclosed a price.
🔹 North Immunology raised $180M from Bain Capital, Janus Henderson and Deep Track Capital and will go public by merging into Aethlon Medical, with the startup’s investors holding 95.25% of the combined company and trading as NRTX from the first quarter of 2027. The asset, NOR-101, is a dual antibody hitting both IL-13 and IL-18 in atopic dermatitis, with Phase 1a beginning as it lists. Deep Track built the company itself rather than backing a founder. The same day, cardiometabolic developer Marea Therapeutics took over Lisata Therapeutics, a solid tumor biotech left stranded by a collapsed merger. Roughly two dozen biotechs have gone public this way in 2026. A structure that used to signal distress now carries crossover money and preclinical assets, which says the appetite for biotech risk has returned and the appetite for the IPO process has not.
🔹 Skyhawk Therapeutics’ once daily oral RNA splicing modulator SKY-0515 produced a 0.94-point improvement at 15 months in Huntington’s disease, against a 0.65-point decline in a matched natural history cohort. That’s a 1.59-point separation on a scale where 1.2 points is the accepted threshold for clinical meaningfulness. Higher doses cut mutant huntingtin protein by more than 60% and lowered PMS1 messenger RNA by 25%, and the separation reached significance at nine months and held at every later timepoint. No serious adverse events were reported. The weak part is the comparison. An external control drawn from a database isn’t a placebo arm, and Huntington’s progresses slowly enough that fifteen months of natural decline is a small number to beat. The Phase 3 FALCON program has finished enrolling 144 patients in Australia and New Zealand and is recruiting about 600 more globally. Roche abandoned tominersen in July and uniQure’s AMT-130 is at the FDA, so a once-daily pill would change the economics of the entire field.
🔹 Connect Biopharma’s IL-4 receptor alpha antibody rademikibart missed its primary endpoint in acute asthma exacerbations and the stock fell nearly 40%. What the company salvaged is specific. Lung function improved significantly at day 7, which is the endpoint it intends to make primary in Phase 3, alongside a 66% reduction in treatment failures and a 50% reduction in emergency department and unscheduled visits. Safety matched placebo. The strategy is to stop competing with Dupixent on chronic asthma control, where it would lose, and claim the acute exacerbation window instead, where no biologic is approved and speed of onset is the only thing that matters. That’s a coherent repositioning, and it rests on a secondary endpoint from a failed study, which the FDA will treat accordingly. The company is requesting alignment on a registrational program in both asthma and chronic obstructive pulmonary disease.
Have a great rest of your week and thanks for reading Biotech Blueprint!
DISCLAIMER: This publication is for informational and educational purposes only and does not constitute investment, legal, medical, or tax advice, a solicitation, or an offer to buy/sell any security. Information is believed reliable but no warranty is made as to accuracy or completeness; views may change without notice. Do your own research and consult qualified professionals.
DISCLOSURE & CONFLICTS: The author may hold positions in securities mentioned and may change positions at any time without notice. No compensation is received from companies mentioned, and there are no known material business relationships unless explicitly stated. Content may reference clinical data and regulatory events; always consult primary sources and a licensed clinician for medical decisions. Past performance is not indicative of future results.
LIABILITY: Use of this content is at your own risk. The author assumes no responsibility for any losses arising from reliance on this material.
Front cover image source: Getty Images
The confirmation took an extra 14 months of follow up. No deal process tolerates 14 months, so the asset changes hands on the signal and the acquirer books the geography question as a discount rather than a doubt. That is not a lapse in rigor. It is the pricing. Whoever waits for the confirmed version pays for a confirmed version, and by then it belongs to someone else.