China's drug regulator has put an off-the-shelf CAR-T for systemic sclerosis on its rare-disease guidance track. The CDE notice (8 Oct) names BRL Medicine's gene-edited, donor-derived CD19 CAR-T for relapsed or refractory SSc. The pilot launched in May and promises early, drug-by-drug CDE guidance through development. It is not an approval or a priority review. The product is already IND-cleared in China. BRL's first allogeneic CD19 CAR-T data, in Cell in 2024, covered three patients, two with diffuse cutaneous SSc. The Western comparison is Cabaletta: its autologous CD19 CAR-T plans to open a ~25-patient registrational SSc-ILD cohort in Q4 2026. $CABA $XBI $IBB ✨ Made with AI

Wolwo, a Chinese drugmaker whose main product is dust-mite allergy drops, now has clearance to test a stem cell therapy in sepsis. The IND is on the CDE's implied-approval list: an umbilical cord MSC injection ("Type II") from its subsidiary Zhejiang Wolwo Stem Cell Technology, for sepsis and/or septic shock. Accepted 28 Jul, announced as cleared 11 Oct. China had cleared 186 stem cell INDs by April and has approved one stem cell drug. No cell therapy is approved for sepsis in the US or EU. The IND is the easy step. Watch for a registered trial and its first patient. $300357 SZ $XBI $IBB ✨ Made with AI

A once-daily oral PDE4 inhibitor beat Otezla head-to-head in a 607-patient Phase 3 run entirely in China. Week 16, prifemilast 20 mg once daily vs apremilast 30 mg twice daily: PASI 75 64.6% vs 35.3%, sPGA 0/1 55.6% vs 31.3%. Serious AEs 4.0% vs 2.5%, AE discontinuations 2.0% vs 1.5%. Company-reported; Newsoara-sponsored across 60 Chinese sites (NCT07712731). The molecule has crossed the Pacific twice. vTv Therapeutics licensed it as HPP737 to Shanghai's Newsoara in 2018 and expanded that licence to global rights in February for $20M upfront. Newsoara then licensed ex-Greater China to San Diego's cAMPfield, which launched in June on a $180M Series A. vTv still holds up to about $115M in milestones plus royalties. Newsoara plans a China NDA by year-end; cAMPfield is taking the drug into UC and Crohn's. $VTVT $AMGN $XBI $IBB ✨ Made with AI

Viatris is buying Pacira for $36.50 a share in cash, about $1.65B in equity value. Exparel and Zilretta are what it is paying for. PCRX-201 is the part I'd watch. PCRX-201 (enekinragene inzadenovec) is a high-capacity adenovirus carrying IL-1Ra behind an inflammation-switched promoter, injected into the knee. Phase 2 ASCEND Part A randomised about 45 patients to two doses or placebo. Pacira guided Part A topline before the end of 2026. The most advanced knee-OA gene therapy is Kolon TissueGene's TG-C, a cell-based product. It missed both co-primaries in ACTiVION-II in July. Its last Phase 3, ACTiVION-I, has slipped from October to December while the company looks into a high placebo response. Both readouts now land by year-end, one of them inside a Mylan–Upjohn company. $VTRS $PCRX $950160 KQ $XBI $IBB ✨ Made with AI

TCGX has closed a $600M fund dedicated to Asian biotechs. TCGX Asia Life Sciences Fund I was oversubscribed. It is a separate vehicle from the firm's US and Europe flagship funds, though it works closely with them, and TCGX is adding offices in Shanghai and Hong Kong to run it. The source is the firm's own release (7 Oct). It names no investors. TCGX is a US-based crossover investor, in private and public biotechs. My read: the licence-in deal is no longer the only route for Western capital into Chinese assets. A US crossover firm now has dedicated money and local teams to invest across company stages. Watch for the first disclosed deals, and whether the fund leads rounds or follows. $XBI $IBB ✨ Made with AI

ArriVent's FURVENT miss at 240mg comes down to who read the scans, and nearly all of the gap sits in the chemo arm. Median PFS, blinded central review vs investigators: Firmonertinib 240mg: 11.0 vs 11.1 months 160mg: 8.4 vs 8.3 Platinum-pemetrexed: 9.5 vs 7.1 The trial was open-label. Central review was the primary endpoint, and 240mg vs chemo came in at HR 0.75, p=0.0654. On investigator review the same comparison was HR 0.61. The 160mg arm missed on both reads. Takeda's EXCLAIM-2 chemo arm ran 9.6 months by central review in the same first-line exon 20 setting, and mobocertinib only matched it (HR 1.04). Allist sells the drug in China as furmonertinib and licensed ex-China rights to ArriVent in 2021. OS is immature; ArriVent says it is weighing its development path. $AVBP $688578 SH $TAK $XBI $IBB ✨ Made with AI

China's drug regulator is proposing priority review for the TIL that skips high-dose IL-2. Juncell's nolgileucel (GC101) appeared on CDE's proposed priority review list on 8 October. The objection window closes 14 October. Juncell is pre-IPO; it refiled for a Hong Kong listing in June. The data behind it is MIZAR-003, an open-label, randomized pivotal Phase 2 in about 100 PD-1-failed melanoma patients at 25 Chinese centres. Company-reported at ASCO 2026: median PFS 4.3 vs 1.6 months on chemo (HR 0.43), ORR 42.0% vs 6.1%. OS is not mature. A correction to my June post on GC101: I wrote that it needs neither chemo prep nor IL-2. Patients still get lymphodepletion first, which Juncell calls low-intensity, and the trial registry lists the PD-1 antibody sintilimab after the infusion. The no-IL-2 part was right. That is the comparison that matters for Iovance. Amtagvi, the only FDA-approved TIL, needs cyclophosphamide-fludarabine lymphodepletion before infusion and high-dose IL-2 after it. GC101 swaps the IL-2 for a PD-1 antibody. One caution: 81.8% of MIZAR-003 patients had acral or mucosal melanoma, the subtypes common in China and rare in the US. The TIL to watch is the one that is easiest to give. $IOVA $XBI $IBB ✨ Made with AI

Caribou had an FDA-aligned Phase 3 design for its off-the-shelf CD19 CAR-T. On Tuesday it said it is shelving the drug, and it has hired Wedbush to find a buyer or merger partner. The trial was ANTLER-3: about 250 second-line large B-cell lymphoma patients who are transplant-ineligible and can't get autologous CAR-T, vispa-cel against investigator's choice, PFS as the primary endpoint. Behind it, the 27-patient Phase 1 subgroup that matches the pivotal dose and donor criteria, company-reported at EHA in June: 82% ORR, 67% CR, median PFS 17.1 months. CEO Rachel Haurwitz pointed to the financing environment for allogeneic CAR-T. CB-011, the BCMA program in myeloma, goes too. Cash was $113.8M on 30 June. The 8-K puts severance and trial wind-down at $15M to $19M, before lease and impairment charges. Cellectis walked away from its two allogeneic programs on 14 September, 22 days earlier. Allogene is still funded: $423.6M on 30 June, runway guided into 2029. Its ALPHA3 randomizes MRD-positive patients after first-line lymphoma treatment to cema-cel or observation, with an interim EFS update due mid-2027. In China, CARsgen said on 29 September it had dosed the first patient in its NMPA-cleared Phase 1 of CT1190B, an off-the-shelf CD19/CD20 CAR-T. What to watch now: whether anyone buys vispa-cel with ANTLER-3 attached. $CRBU $ALLO $CLLS $2171 HK $XBI $IBB ✨ Made with AI

Regend Therapeutics says patients in Hainan have paid RMB 150,000 a treatment for its lung stem cell therapy since late May 2025. That was before its placebo-controlled Phase 2 had finished. REGEND001 is the patient's own airway basal stem cells, expanded and returned by bronchoscope. Regend told 36Kr in February it had charged for more than 140 treatments across COPD, interstitial lung disease and bronchiectasis in the Boao Lecheng pilot zone, which lets hospitals bill for cell therapies without NMPA approval. The COPD use there is registered as an open-label Phase 1/2 sponsored by Ruijin Hospital (NCT06946953). The Phase 2 (NCT05638776) randomised 58 COPD patients against placebo, single-blind, and completed on 9 Jul 2025. Regend reported a significant DLCO gain over placebo at ERS 2025. No results are posted on the registry and I found no paper. The confirmatory Phase 3 (NCT07828353) went up on 18 Sep: 102 patients, blinded, placebo-controlled, DLCO at week 52. Not yet recruiting, estimated start this month, primary completion September 2028. In China's stem cell sector, the cash register can open before the controlled data arrive. $XBI $IBB ✨ Made with AI

JUST IN: The 2026 Nobel Prize in Physiology or Medicine goes to Karl Deisseroth, Peter Hegemann and Georg Nagel "for their discoveries concerning light-gated ion channels and optogenetics." Before the announcement, our newsletter ran five picks: GLP-1, CAR-T, cGAS-STING, optogenetics and phase separation. One of five hit. We also got two of the three names. Our third pick was Gero Miesenböck. Nagel and Hegemann showed in 2002–03 that channelrhodopsins from green algae are light-gated ion channels. Deisseroth put ChR2 into neurons in 2005. 🇩🇪 Both German laureates trained at the Max Planck Society. Hegemann did his PhD in Dieter Oesterhelt's lab at the MPI of Biochemistry in Martinsried, then ran his own group there until he moved to Regensburg in 1993. Nagel led a group at the MPI of Biophysics in Frankfurt from 1992 to 2004, which covers the years of the channelrhodopsin papers. The CGT angle: optogenetics now has a gene therapy under FDA review. Nanoscope's MOGENRY is an AAV opsin injected into the eye for retinitis pigmentosa, and the FDA accepted its BLA on 9 Sep. The other four picks stay on our list for 2027. $XBI $IBB ✨ Made with AI

Novartis is paying $575 million upfront for Abogen's in-vivo mRNA T-cell engager, ABO2203. Total milestones reach $7.2 billion, plus royalties. The target is CD19xCD3 for autoimmune indications, delivered via lipid nanoparticles to generate bispecific engagers directly inside the patient. The deal validates in-vivo T-cell redirection as a direct threat to ex-vivo autologous CAR-T. By having the patient's own liver or immune cells produce the CD19 engager from synthetic mRNA, Novartis bypasses leukapheresis, viral vectors, and multi-week vein-to-vein turnarounds entirely. Autologous cell therapy platforms pushing into lupus and sclerosis now face an incumbent willing to price in-vivo convenience at half a billion in cash before clinical proof of concept. Watch Phase 1 PK/PD readouts to see whether transient mRNA expression can match the deep, reset-level B-cell depletion achieved by cellular infusion. $NVS $XBI $IBB ✨ Made with AI

JUST IN: Novartis will pay Abogen $575M upfront for ABO2203, a CD19×CD3 T-cell engager that is injected as mRNA and built by the patient's own cells. Milestones can reach about $7.2B if Novartis exercises every option on Abogen's other RNA programs. Exclusive worldwide rights; closing still needs regulatory clearances. On 22 September I wrote up this drug's first autoimmune data, published in Cell: three women with refractory immune thrombocytopenia, all with complete platelet responses within a month that held at six months, no CRS, no ICANS, and no IND that I could find. Ten days later it has a buyer. In late August Novartis paused all eight autoimmune trials of rap-cel, its autologous CD19 CAR-T, after three patients died of IEC-HS, a hyperinflammatory syndrome that sets in after CRS has passed. BMS disclosed a pause of its own autoimmune CD19 CAR-T the same week. Neither has announced a restart. ABO2203 takes the cell out. Abogen's lipid nanoparticles go under the skin, whatever cells take them up make the bispecific, and the protein has a half-life of 7 to 11 days in patients, so it is dosed weekly and can be stopped. Across nine lymphoma patients at AACR and the three ITP patients in Cell there was no CRS and no ICANS; the paper's 16 autoimmune patients had nothing worse than grade 2. Both are Ruijin Hospital studies, and Abogen co-funded and co-wrote the Cell paper. Small numbers, nothing yet past six months. Novartis's president of biomedical research, Fiona Marshall, says the approach could complement existing modalities. The price says where it sits. Gilead agreed in March to pay $1.675B upfront for Ouro's BCMA×CD3 protein engager, and UCB in May $2B upfront for Candid's, both early-stage assets that also trace to Chinese companies. Novartis is paying about a third of that or less, for fewer patients and options on more programs from the same platform. What to watch: whether Novartis moves ABO2203 into a company-sponsored trial under an IND, since both trials so far are the hospital's, and what the Cell patients' platelets do now that their B cells have returned, which happened at month four or five. Novartis has not said when rap-cel's autoimmune trials resume. $NVS $GILD $XBI $IBB ✨ Made with AI

Roche's GYMINDA trial had one primary endpoint: percent change in body weight at 48 weeks, with the myostatin antibody emugrobart or a placebo added to Lilly's tirzepatide. On Monday, Chugai said an interim analysis showed the trial is unlikely to meet its prespecified goal of clinically meaningful weight loss. No new safety signals were seen. Roche has stopped development and is returning all rights to Chugai (TSE 4519), the Japanese drugmaker it majority-owns. Weight is the number obesity drugs are approved on. Veru said last year that the FDA told it extra weight loss over a GLP-1 alone is an acceptable primary endpoint for approving its muscle-preserving add-on. Roche has now dropped emugrobart in three indications this year. In March it stopped the spinal muscular atrophy and FSHD programs after the drug did not consistently improve muscle growth or function. It is the second myostatin drug Roche has let go. The first, licensed from Bristol Myers Squibb, was stopped in Duchenne muscular dystrophy in 2019. Biohaven later licensed it from BMS as taldefgrobep alfa, and it missed its Phase 3 in SMA in 2024. The myostatin antibody that got through is Scholar Rock's. On September 11 the FDA approved apitegromab, sold as Isembyld, for SMA patients aged 2 and older who are already on nusinersen or risdiplam. It is the first muscle-targeted SMA treatment, an IV infusion every four weeks, with a list price of $11,659 a vial and a net cost Scholar Rock puts at about $310,000 a year for a typical patient. Chugai now wants to take emugrobart back into SMA against it. It cites subcutaneous dosing, the drug's safety and long-term data, and a redesigned Phase 3, and says it is exploring out-licensing to a third party. Roche is still pursuing muscle in obesity, through a different pathway. In August it licensed Hanmi's HM17321, a urocortin 2 analog in Phase 1 meant to cut fat while sparing muscle, for $190 million upfront. Scholar Rock has the SMA muscle market to itself for now. Chugai's route back starts with a Phase 3 that has not begun. $RHHBY $SRRK $LLY $XBI $IBB ✨ Made with AI

53 days after announcing its NMPA IND, CARsgen says it has dosed the first patient in the China Phase 1 trial of CT1190B, its off-the-shelf CD19/CD20 CAR-T, in large B-cell lymphoma after two or more prior lines (HKEX, 29 Sep). No new data. Behind it are hospital studies shown at EHA in June, company-reported: 11 of 12 evaluable patients responded, 8 with complete responses, across large B-cell and follicular lymphoma. One grade 3 CRS, one grade 3+ ICANS. 7 of 11 responders were still responding at a median follow-up of 6.6 months. The product now moves from hospital studies into a company trial the regulator oversees, the route that can end in an approval filing. $2171 HK $XBI $IBB ✨ Made with AI

Two CAR-Ts got the same verdict from South Korea's reimbursement committee this year, for the same third-line lymphoma patients. Only one is going ahead. The verdict: eligible for coverage in adults with relapsed or refractory DLBCL or PMBCL after two or more lines, if the company accepts the agency's assessed price or a lower one. Gilead's Yescarta got it in May. Gilead turned it down and is now seeking coverage for second-line use instead. Rimqarto, the first CAR-T developed in Korea, got the same verdict on September 3. Its maker, Curocell (KOSDAQ 372320), now has up to 60 days of price talks with the National Health Insurance Service and is aiming to be listed this year. The only CAR-T covered today is Novartis's Kymriah, listed in 2022 at about 360 million won an infusion (roughly $265,000 at current rates), with the patient's share capped at about 6 million won at the time. Korean press expects Rimqarto's price to come in around Kymriah's or lower. Korea's drug regulator, the MFDS, gave Curocell more than it asked for. Curocell filed for conditional approval, which would have left a Phase 3 to run after launch. The MFDS waived the Phase 3 and granted full approval in April. Its advisers called a randomized trial unethical, because proven CAR-Ts already exist, and impractical, because third-line patients are few. Rimqarto is a CD19 CAR-T whose vector also dials down PD-1 and TIGIT, two brakes that build up on exhausted T cells. The approval rests on one single-arm Phase 2, where Curocell reports a 75.3% response rate and 67.1% complete responses. At the analysis presented in June 2025, with median follow-up of 8.5 months, grade 3 or worse CRS occurred in 8.9%, median progression-free survival was 6.0 months, and an estimated 41.1% were progression-free at a year. The stock has fallen while Curocell waits. It was around 65,000 won when the approval came in April and about 25,000 after the committee verdict this month. Korean coverage blames the uncertain start of revenue and a thin pipeline behind Rimqarto; the next program, in lupus, was cleared for a Phase 1/2 trial in August 2025. The homegrown drug has a Chinese-owned supplier. Curocell's lentiviral vector partner is ProBio, the contract manufacturing arm of Nanjing-based GenScript. ProBio says its plant is the first lentiviral vector facility to receive GMP certification from Korea's drug regulator. Kymriah has had Korea's CAR-T reimbursement list to itself since 2022. By next year it could share third-line lymphoma with a CAR-T made in Daejeon, while Gilead tries for second line. $GILD $NVS $1548 HK $XBI $IBB ✨ Made with AI

4-1BB agonists have a history of hurting the liver. BeOne, formerly BeiGene, is now taking one into a Phase 3 in liver cancer. Most of that history is BMS's urelumab. Two deaths in its early trials were attributed to drug-related liver toxicity, at 1 and 5 mg/kg, and the dose came down to 0.1 mg/kg, where the antibody disappointed on its own. The new trial, BGB-B2033-301, went up on the US trial registry on September 23. It plans to randomize about 492 people with liver cancer who have had one or two prior lines of treatment that included a PD-1 or PD-L1 drug, against the investigator's choice of lenvatinib or sorafenib. The primary endpoint is overall survival. BeOne's fix is to move the safety switch from the dose to the tumor. BGB-B2033 pairs two GPC3-binding arms with 4-1BB binders on an Fc engineered not to bind Fc-gamma receptors. BeOne describes GPC3 as absent from normal adult tissue outside the endometrium and placenta, and its preclinical work says the molecule can only activate 4-1BB where GPC3 is present. The Phase 1 update at ASCO in June covered 61 patients, 60 with liver cancer and 59 previously treated with a PD-1 or PD-L1 drug. At 300 mg and above, the confirmed response rate was 28.9% (11 of 38): 28.6% at 300 mg, 35.7% at 600 mg and 20.0% at 1,000 mg. Median follow-up was 4.8 months. The liver still shows up in the safety table. Grade 3 or worse treatment-related events hit 5 of 61 patients. The abstract lists six such events, and three were liver labs: ALT, AST and bilirubin rises. The one dose-limiting toxicity was an ALT rise, which resolved. Immune-mediated events were all grade 1 or 2. On response, BGB-B2033 already looks ahead of its control arm. Lenvatinib managed 12% after atezolizumab plus bevacizumab in a 50-patient Korean Phase 2. Survival is the harder test, and in liver cancer the two don't always move together. Sorafenib won FDA approval in 2007 on survival, 10.7 months against 7.9 on placebo, with partial responses in 2% of patients. The posted entry criteria read differently from the Phase 1's. The Phase 1 enrolled GPC3-expressing tumors. The key criteria BeOne posted for the Phase 3 only require the GPC3 level to be known, for stratification. I haven't found Phase 1 responses broken out by GPC3 level. The Phase 1 also doesn't look like Western liver cancer: 92% of patients were Asian and 87% had hepatitis B. The Phase 3 is planned for centers worldwide and carries an EU trial number. Elsewhere in the class, Leads Biolabs' opamtistomig, a PD-L1x4-1BB bispecific, is under priority review in China. Genmab dropped its own PD-L1x4-1BB drug, acasunlimab, mid-Phase 3 last December, citing focus and competition. The survival readout is years away, with primary completion estimated for September 2029. Until then, every BGB-B2033 update is a liver safety readout first. $ONC $9887 HK $GMAB $XBI $IBB ✨ Made with AI

Carvykti's label puts parkinsonism at 6% of patients in its original pivotal trial, CARTITUDE-1, and 1% in the later CARTITUDE-4. In the pivotal trial of Gilead's anito-cel, 117 patients, the reported count is zero. A matched comparison of the two BCMA CAR-Ts, shown at the International Myeloma Society meeting in Glasgow this week, weighted patient-level data from anito-cel's trial, iMMagine-1, against CARTITUDE-1's published and regulatory data. Efficacy came out close: response rates of 96% for anito-cel and 98% for Carvykti, which was slightly ahead on depth of response, with no efficacy gap statistically significant. Anito-cel came out lower on neurotoxicity outside ICANS, 0% against 12.4%, on grade 3 or 4 infections, 9% against 23%, and on non-relapse mortality, 3% against 11%. The presenters listed the limits: the trials ran in different eras of supportive care, and ferritin, missing from the CARTITUDE-1 data, could not be matched. NEJM published anito-cel's Phase 1 the same week. Its 38 patients had a median of four prior lines, and all were triple-class refractory. All 38 responded, nearly 80% completely. More than half were progression-free at two years, 65% were alive at three, and no delayed neurological complications were reported. Legend's data at the meeting were about durability. In CARTITUDE-2, 10 of 20 patients treated with Carvykti after one to three prior lines were alive and progression-free five years after one infusion, with no maintenance therapy, and Legend reported no new CAR-T-related neurotoxicity with the longer follow-up. Carvykti also has a randomized win on overall survival, HR 0.55 in CARTITUDE-4. Anito-cel has no randomized data yet. The FDA's target date for anito-cel, in patients after three or more prior lines, is December 23. Carvykti's boxed warning names parkinsonism and Guillain-Barré syndrome. The test for anito-cel's label is whether it mentions either one at all. The next fight is earlier in treatment. Carvykti is approved after one prior line in lenalidomide-refractory patients. Gilead finished enrolling iMMagine-3, anito-cel against standard regimens after one to three prior lines, in the second quarter and says it could file as early as 2027. Gilead paid about $7.8 billion for Arcellx and will pay $5 a share more if anito-cel sells $6 billion from launch through 2029. If the FDA approves in December, that means averaging about $2 billion a year. Carvykti, four years after launch, sold $1.25 billion in the first half of this year. Gilead's own cell therapies, Yescarta and Tecartus, fell 14% to $417 million in the second quarter. Anito-cel would start in fourth line, with safety as its pitch. The bigger risk I see for Legend and J&J is iMMagine-3. $GILD $LEGN $JNJ $XBI $IBB ✨ Made with AI

Beijing's new five-year plan for the drug industry wants at least 25% of the world's first-in-class drugs by 2030. A footnote under the targets table says what counts: every drug in Phase 1, 2 or 3, filed for approval, or approved. Preclinical, discontinued and withdrawn programmes are excluded. So the target counts clinical-stage and approved drugs, from first-in-human on. On pipeline counts, China is already at or near the top. PharmCube and Guosheng Securities data, reported by National Business Daily, put Chinese companies at 4,751 of the 14,088 active innovative drugs in development worldwide at the end of 2025, from IND filing to marketing application. That is 33.7%, against 28.5% for US companies. MIIT's own explainer of the plan puts China second. The same data say 827 Chinese originator drugs entered the clinic for the first time in 2025, 47.4% of the global total, and that 76% of China's originator drugs are at an early stage. What the plan doesn't give is a baseline: how much of that pipeline is first-in-class today. The target the market can check sits in the same table: at least five products with more than $1 billion a year in global sales. National Business Daily counts one this week, BeOne's Brukinsa, which sold $3.9 billion in 2025. Carvykti, discovered by Legend in Nanjing and sold with J&J, also passed $1 billion last year, at about $1.9 billion. The plan doesn't define which products count. The chapter I'd read as a Western investor is on supply-chain security. A box lists the bioprocess and fill-finish products China should focus on developing: serum-free media, chromatography resins, depth, ultra- and nanofiltration membranes, high-end bioreactors, automated single-use and multi-column continuous chromatography systems, oligo synthesizers and blow-fill-seal lines. It names no companies. Most of these are categories where Cytiva, Thermo Fisher, MilliporeSigma, Sartorius and Repligen sell. For now those suppliers are still growing in China. Danaher said in July that its biotechnology business there grew again, "as local companies continued to partner with Cytiva." My read: the first-in-class target is set on a pipeline measure, China's strongest ground, and the $1 billion target is the one investors can audit. For Western tools makers, the supply-chain list sets a five-year direction. I found no purchasing rules in the plan; what I'm watching is whether any follow. $ONC $LEGN $DHR $XBI $IBB ✨ Made with AI

Lepu Medical is suing the head of the BD firm it hired to find partners for its obesity drugs outside China. Its complaint, as reported, says she was working at the same time on a rival obesity drug whose potential partners overlapped heavily with Lepu's. She is Su Zhang, global BD vice president at Hengrui's US arm from 2023 to 2024, when Hengrui did the GLP-1 NewCo deal that created Kailera. In December 2024 Lepu signed her firm, CoDevCo, as its exclusive BD agent for the obesity pipeline, for $500,000 up front plus success fees. In June 2025, Endpoints introduced Zhang as BrightGene's acting CBO while BrightGene looked for a Western partner for BGM0504, its GLP-1/GIP agonist. Chinese reporting says her talks with BrightGene began in September 2024, before Lepu signed her firm, and that she said publicly in February 2025 that she had a partnership with BrightGene. Lepu's complaint alleges breach of fiduciary duty, fraudulent concealment and trade-secret misappropriation. It seeks more than $500,000 in damages, disgorgement and an injunction. Zhang denies the allegations. Her lawyers argue that Lepu's contract is with CoDevCo, so she owes Lepu no personal duty, and that the exclusivity ran one way: Lepu could not hire another agent, and nothing barred CoDevCo from serving other companies, competitors included. The federal docket shows she removed the case from New York state court to federal court on September 4 (EDNY 2:26-cv-05495). No court has ruled on the claims. The fee fight is in Singapore. In October 2025 Lepu licensed MWN105, a GLP-1/GIP/FGF21 triple agonist, outside Greater China to Denmark's Sidera Bio: $35M upfront and near-term, up to $1.01B in milestones, plus 9.99% of Sidera. In January CoDevCo filed at the Singapore International Arbitration Centre for success fees on that deal. The contract, as reported, says CoDevCo earns its success fee in any case once a deal is concluded. Lepu has rejected the claim and filed a counterclaim. The hearing is set for February 2028. The two sides describe the same process from opposite ends. Lepu's lawyers say it could not see Zhang's emails with prospective partners, only the reports she chose to send. Her lawyers say Lepu kept every decision, from which partners to approach to whether to sign. That gap is what I would flag to any Western BD team. On a China-out deal, the seller's agent may be the only person talking to every buyer, and in a crowded class like incretins the same agent can carry more than one book. Before the first call, ask who else the seller's agent represents in the class. Chinese sellers signing an agent should check which way the exclusivity runs. $300003 SZ $688166 SH $XBI $IBB ✨ Made with AI

I went looking for what is inside YN001, the plaque-reversal drug behind a nearly RMB 800 million round in Beijing this month. It is rosuvastatin. Beijing Inno Medicine puts the statin in a liposome aimed at CD44, a receptor that is ramped up on macrophages and other cells in plaque, and in its current trial gives it by IV once a week. The company calls YN001 the world's first targeted drug for reversing atherosclerotic plaque. The Series D, a little under $120 million, was announced on September 9. YuanBio Venture Capital led it, and two Beijing state funds were among the co-investors. The evidence so far is one open-label Phase 1b/2a, presented at ESC 2025. According to the trial registry, its control arm took rosuvastatin tablets, and plaque volume was a secondary endpoint. The abstract reported the first 16 patients to finish, across both arms: coronary atheroma volume on IVUS fell 13.8% on YN001 and rose 0.6% in the control arm, and CT angiography showed a 27.8% mean fall on YN001. Inno's own release gives only a range, 10% to 50%, and it has not published numbers for all 62 patients. Testing the drug against itself as a pill is the useful part of that design: in principle, any difference comes from the delivery. Sixteen completers in an open-label study can't rule out chance. The real test has started. PURIFY-TIMI 81 is run with the TIMI Study Group at Brigham and Women's, which ran FOURIER for Amgen's Repatha. It plans to enrol 456 patients, who will get 20 mg or 40 mg of YN001 or a placebo by IV every week for 12 weeks, on top of their usual treatment, double-blind. The primary endpoint is the change in non-calcified plaque volume on CT at week 13. Sites are open in China, the US and Australia, with Canada next. The bigger claim is regulatory. Inno said in December that the FDA had cleared this study as a registrational trial with plaque regression as the surrogate endpoint, and that it expected to file for accelerated approval in a little over a year. It called that a first for a cardiovascular drug. The FDA does not publish meeting outcomes, so this rests on Inno's account, and the sites only began opening this summer. Lipid drugs can already win approval on LDL cholesterol. If a plaque measurement on CT becomes a second route, every company with a plaque or anti-inflammatory heart program will want it, and the first drug through could be a Beijing reformulation of a generic statin. $AMGN $XBI $IBB ✨ Made with AI