Bio & Healthcare Weekly Analysis — 2026-11-03: Falling prices and rising volume in Lilly's third quarter, Inovio's accelerated approval, a partial CAR-T clinical hold and a halted neurodegeneration Phase 2

Briefings covered: 2026-11-03 news briefing

Frame: Value-added distribution — the hypothesis that platforms, tools and infrastructure supplied in common to many drugs and hospitals (AI drug-discovery platforms, CDMOs, research automation, diagnostics, medical AI) capture high value through recurring revenue and data accumulation; that pharmaceutical and medical-device companies holding approved therapies or devices sit at a medium-to-high level, exposed to patent cliffs, drug-pricing pressure and reimbursement negotiations; and that single-pipeline biotechs and service providers capture low value with high volatility because they depend on binary clinical outcomes and financing. When citing figures, the stage (preclinical, Phase 1, Phase 2, Phase 3, approval, reimbursement, sales) and the basis (endpoint, list price/net price, upfront/total value) are stated together.

This Week's Key Points

Eli Lilly's third-quarter revenue was US$24.4 billion, up 39% year on year, and the increase came from volume rather than price. According to the 8-K, total volume was +53% and realized prices were −15%; in the US alone, volume was +41% and realized prices −7% (mainly Zepbound). Mounjaro revenue was US$10.7 billion (+65%) and US Zepbound revenue US$5.3 billion (+48%). Full-year revenue guidance was raised from US$85–87 billion to US$91–92 billion. However, about US$0.9 billion of revenue is one-time in nature, such as milestones and adjustments to rebate estimates, per the company, and non-GAAP EPS guidance was lowered to US$35.00–36.00 to reflect US$4.2 billion of acquired IPR&D charges related to acquisitions. Quarterly revenue of the oral GLP-1 Foundayo was US$184 million, an early-launch figure.

On October 30 the FDA granted accelerated approval to Inovio's Voceryq (INO-3107) for adult recurrent respiratory papillomatosis (RRP). It is the first commercialization of a DNA medicine, based on an open-label trial in 32 adults who had two or more surgeries in the prior year. The 12-month complete response of 7/26 (27%) is a single-source secondary figure relaying the company's announcement, and the FDA label is unverified. A confirmatory trial is a condition of approval, the price is unverified, and the full launch is in the first quarter of 2027. Precigen's Papzimeos, approved in 2025, already serves the same indication, so this approval puts two products in competition.

Three setbacks in safety or efficacy in late- and early-stage trials were disclosed in the same week. Gilead disclosed a partial clinical hold on the Phase 1/2 and Phase 3 trials of KITE-753, a CD19/CD20 CAR-T made with a three-day rapid manufacturing process, to investigate a difference in the incidence of high-grade neurotoxicity (ICANS). Teva stopped its Phase 2 TOPAS-MSA trial of emrusolmin in multiple system atrophy following an interim futility recommendation by the independent data monitoring committee. A report also said the FDA placed a partial hold on a Bright Minds epilepsy trial over toxicity findings in mice, but this is a single source at headline level.

Anthropic acquired Mithril Medicine, a one-person preclinical biotech, and began in-house drug programs. The targets are two rare childhood muscular dystrophies, and the only evidence is the company's description of "promising results" at the mouse stage. The company's target is clinical entry as early as late 2028, and the acquisition price is unverified. Following OpenAI's acquisition of Orchestra Bio on October 20, the confirmed fact is that frontier labs' move into biology has widened from supplying tools to owning in-house pipelines.

In Korea, GC Biopharma and Hanmi Pharmaceutical out-licensed the Fabry disease candidate LA-GLA to BioMarin, and draft subordinate regulations on telemedicine banned remote prescribing of obesity drugs. The out-licensing deal carries a US$70 million upfront payment and up to US$475 million in total value, and the asset is in a global Phase 1/2 trial. The telemedicine amendment is at the legislative-notice stage (closing November 9), and the amended Medical Service Act takes effect on December 24.

1. Industry Implications

(1) Demand and disease: in the incretin market, volume growth is offsetting falling realized prices

Lilly's third quarter shows a structure in which a company holding approved therapies grows revenue through volume while under price pressure. Realized prices of −15% are evidence that drug-pricing pressure is flowing into actual results, and volume of +53% is evidence of demand exceeding that pressure. Both figures are in the original 8-K and are confirmed facts. Given the company's explanation that about US$0.9 billion of the quarter's revenue growth is one-time in nature, the underlying growth rate should be read as lower than 39%.

The frame places companies holding approved products in a medium-to-high tier exposed to patent cliffs, pricing pressure and reimbursement negotiations. The incretin class, with quarterly revenue from a single drug class above US$10 billion, appears to be an exception that maintains bargaining power by accepting price cuts and widening volume and coverage. This is recorded as a hypothesis. For it to hold, volume growth must continue to exceed price declines after the US list price cut for Wegovy in January 2027 and the expansion of Medicare coverage of obesity drugs. Quarterly revenue of US$184 million for the oral Foundayo is early-launch, so there is not yet evidence to judge how far oral drugs replace or expand demand for injectables.

In the continuous glucose monitoring (CGM) tier, Dexcom reported third-quarter revenue of US$1.344 billion (+11%, US +14%) and raised full-year guidance to US$5.2–5.25 billion. It is a fact that metabolic-disease demand is carrying through to the device tier outside therapeutics at low-double-digit growth, but the effect of GLP-1 uptake on CGM demand cannot be separated from these figures alone.

(2) Technology supply: frontier labs' in-house pipelines have started at the preclinical stage

Anthropic's acquisition of Mithril Medicine is a research-stage transaction with no clinical entry or clinical data. According to the company, the founder used Claude to coordinate contracts and experiments across 11 external laboratories and produced mouse results within 18 months. On that description, the work is closer to AI support for running experiments and coordinating outsourcing than to AI design in which a model generates the candidate. How far the model's role extended in discovering the candidate cannot be verified from public material. The results disclosed by six partners (BMS, Generate Biomedicines and others) are also laboratory- and document-stage figures such as binding affinity, number of designs and document review time, and have not been peer reviewed.

Under the frame, frontier labs belong to the platform and tools tier that supplies many drug companies in common, but once they own in-house pipelines they also take on the binary clinical risk of individual programs. It usually takes 7–10 years from Phase 1 entry to approval, and while there are claims that AI shortens the preclinical period, there is not yet evidence that clinical timelines have shortened. Even if clinical entry happens in late 2028 as the company targets, the data needed to judge approval will come after that. No new clinical data or approval for an AI-designed drug was confirmed this period, and results from Recursion's REC-4881 update data are unverified.

(3) Clinical development and approval: an accelerated approval on small evidence and setbacks in cell therapy and neurodegeneration came in the same week

Inovio's approval adds another case showing that a rare-disease accelerated approval can be obtained from a small single-arm trial. The post-approval stages of price, reimbursement and prescription volume have not opened. It is a fact that this is the second approval in the same indication, so two companies now share a market with few patients, and revenue scale will first be confirmed in post-launch quarterly results. That a single-pipeline biotech that clears the binary risk through approval still carries the burden of a confirmatory trial and a competing product is in the same direction as the frame's low-value, high-volatility hypothesis.

The issue in the KITE-753 hold is whether shortening the manufacturing time affects the safety profile. Per the company, the Phase 1 autoimmune trial of KITE-363, which has the same construct made with the conventional process, continues. Rapid manufacturing is a process innovation meant to cut the cost and waiting time of cell therapy, and depending on the outcome of this investigation, process changes may come to require separate safety evidence. This is a reasonable inference, and its magnitude cannot be judged because the number of patients with ICANS has not been disclosed. On the same day Gilead reported product revenue excluding Veklury of US$8.2 billion (+16%) and raised full-year product revenue guidance (US$31.0–31.3 billion); at a company with multiple approved products, an individual clinical hold did not change the company-wide outlook.

Teva's halted MSA Phase 2 is another interim failure for the mechanism of inhibiting alpha-synuclein aggregation. Because the interim figures were not disclosed, it is not known whether there was no effect or insufficient statistical power. The company said it will reanalyze together with open-label extension data. It is a financing headwind for biotechs with Parkinson's and MSA programs using the same mechanism, but the effect on any individual company must be judged by the data for each compound.

(4) Policy and payment: the US eased unapproved compounding and EHR AI certification, while Korea is moving to block remote prescribing of obesity drugs

On October 28 the US HHS allowed compounding of six peptides not approved by the FDA (including BPC-157) through enforcement discretion. This is not an approval and does not change Section 503A, and the conditions include ingredients from FDA-registered manufacturers and reporting of serious adverse events. An FDA advisory committee recommended allowing them in July, while internal FDA scientists reportedly opposed it. Because these are not approved drugs, analysts see low odds of insurance reimbursement, so this market is likely to grow in cash-pay telehealth and med-spa channels. For companies whose drugs went through approval, this widens a competing channel sold without clinical evidence, and whether the policy extends to other ingredients such as compounded GLP-1s is the larger variable for the tier holding approved products.

The HTI-5 final rule removed the transparency (model card) requirements for predictive decision-support AI from EHR certification. This means responsibility for validating hospital AI tools shifts from federal certification to purchasing hospitals. Epic said it will keep its transparency tools regardless of the federal requirements, and in the same week STAT published an investigative report that Epic's selective restrictions on data access disadvantage competing developers. The inference is that the certification burden falls for the medical AI platform tier, while hospital-by-hospital validation demands and the terms of EHR data access remain variables that set the pace of adoption. Reports of the number of certification criteria removed are conflicting depending on how they are counted.

Korea's draft telemedicine amendment limits first-visit prescriptions to seven days and bans remote prescribing of obesity drugs. It is a draft at the legislative-notice stage, not the final promulgated version. If finalized, demand for obesity-drug prescriptions on Korean telemedicine platforms will be blocked within the regulated system. This runs in the opposite direction to the US widening of unapproved peptide compounding in the same week, showing that national regulations are setting the distribution channels for obesity-drug demand differently.

(5) Capital and Korea: the upfront share of a Korean out-licensing deal and the outline of rules on China licensing

GC Biopharma and Hanmi Pharmaceutical's LA-GLA out-licensing deal carries a US$70 million upfront payment, about 15% of the up to US$475 million total. The up to US$405 million in milestones is contingent and not a confirmed receipt. The asset is a subcutaneous long-acting enzyme replacement therapy in a global Phase 1/2 trial and has not reached the clinical-data stage. BioMarin said it targets patients with variants for whom the oral drug Galafold is not applicable, placing the deal in a trend where competition in rare-disease enzyme replacement therapy is moving to ease of administration. The deal terms are consistent across multiple Korean and international outlets, but all relay the companies' announcements.

On October 22 the US Treasury issued proposed rules implementing the COINS Act that restrict equity investments and joint ventures in Chinese biotech while in principle excluding sole licensing. Comments close on December 7 and the rules are not in effect. Some outlets' description of a "total ban" is an exaggeration. The inference is that if the rules are finalized as proposed, large drug companies can continue in-licensing China-origin assets through licensing structures without equity investment, while deals with joint-venture or equity structures decline.

In digital health, first-generation telehealth company Teladoc reported third-quarter revenue of US$605.1 million (−3%) and adjusted EBITDA of US$79.4 million (+14%), with BetterHelp revenue down 13%. It is improving profitability amid falling revenue, in contrast to Dexcom's revenue growth in the same week.

2. What to Watch

The price-volume balance in the incretin market. Check the same metrics (volume and realized price) in Novo Nordisk's third-quarter results on November 4, and watch whether volume growth continues to exceed price declines in quarterly results (the first-quarter 2027 reporting season in April–May 2027) after the US list price cut for Wegovy in January 2027 and the expansion of Medicare coverage of obesity drugs. If it does, the hypothesis that incretin holders are an exception to the frame strengthens; if volume growth slows, the view returns to the drug-pricing pressure hypothesis for the approved-product tier. Track oral-drug revenue in the same earnings disclosures.

IND filing for Anthropic's programs and clinical entry of partners' candidates. Confirm through ClinicalTrials.gov registrations, company disclosures and the business and risk-factor sections of the public version of the S-1. If an IND is filed before the late-2028 target or a partner's candidate enters Phase 1, frontier labs' contribution can be judged to have moved from the research stage to clinical entry. Even then, whether it is AI design or AI support must be checked separately against public material on how the candidate was discovered.

Cause and timing of the lift of the KITE-753 partial hold. Confirm through Gilead's fourth-quarter results (around February 2027), presentations at ASH in December, and ClinicalTrials.gov registration status. If the cause is linked to the rapid manufacturing process, the view strengthens that shortening cell-therapy processes will require separate safety evidence; if it is found to be unrelated to the process, the hold is confined to an issue with one program.

Voceryq price, reimbursement and confirmatory trial design. Confirm through the draft confirmatory trial protocol due by year-end, price disclosure at launch in the first quarter of 2027, and publication of the FDA label. If the efficacy figure on the label differs from the company's announcement (27%), that figure will be corrected. Post-launch prescription volume will first appear as numbers in 2027 quarterly results.

Adverse event reporting for unapproved peptide compounding and widening of its scope. Confirm through FDA final rules, adverse event reporting tallies, and policy announcements on other ingredients such as compounded GLP-1s. If the scope widens, the view strengthens that some demand for drugs that went through approval could move to cash-pay compounding channels.

Finalization of the rules on investment in Chinese biotech and the final Korean telemedicine amendment. Check in the Federal Register whether the sole-licensing exclusion survives in the Treasury's final rules after the comment deadline (December 7), and check through official promulgation whether Korea's telemedicine amendment is promulgated with the obesity-drug ban intact after the legislative notice closes on November 9 and before it takes effect on December 24. Also watch whether the structure of licensing deals for China-origin assets changes in deal announcements around ASH in December and JPM in January 2027.

3. Data Reliability Notes

Citable

• Lilly's third-quarter revenue, product revenue, volume and realized price, and guidance: the original 8-K, consistent with multiple media reports.
• Gilead's third-quarter product revenue and guidance, and Dexcom's third-quarter revenue and guidance: based on each company's 8-K.
• Voceryq accelerated approval and launch timeline: based on Reuters, consistent across multiple outlets.
• Teva's halt of TOPAS-MSA: the company announcement and Investing.com report are consistent.
• The six peptides and conditions in HHS's enforcement-discretion policy: STAT, NPR and law-firm commentary are consistent.
• LA-GLA out-licensing terms (upfront and total value): consistent across multiple Korean and international outlets, but all relay the companies' announcements.
• Announcement date and comment deadline of the Treasury's COINS Act proposed rules, and the exclusion in principle of sole licensing: based on law-firm commentary and reported remarks by Treasury officials.

Use with caution

• Lilly's one-time revenue of about US$0.9 billion: per the company. Foundayo revenue is an early-launch figure from which no trend can be drawn.
• Voceryq 12-month complete response of 27% (7/26): a single-source secondary figure relaying the company's announcement; the FDA label is unverified. Price and reimbursement are unverified.
• Anthropic: Mithril's mouse results and the results of the six partners are all per the company and partners, preclinical and laboratory-stage figures that have not been peer reviewed. Clinical entry in late 2028 is a target. The acquisition price is unverified, and reports of the S-1 filing date are conflicting.
• KITE-753 partial hold: disclosed by the company; FDA documents are unverified. The number of patients with ICANS is unverified.
• Teva's interim analysis figures are undisclosed. The Bright Minds partial hold is a single source at headline level.
• The Federal Register date of the HHS peptide policy is conflicting across sources. The effective date and details of the reporting system were not confirmed against the policy text.
• The number of certification criteria removed by HTI-5 is conflicting depending on how it is counted. The effective date is unverified.
• Korea's telemedicine amendment is a draft at the legislative-notice stage, not the final version.
• AI drug development: no new clinical data or approval for an AI-designed drug was confirmed this period. Recursion's third-quarter results and REC-4881 update data are unverified.
• Novo Nordisk's third-quarter results (November 4) and the PDUFA target date for Vertex's povetacicept (November 30) are scheduled dates, not results.

Correction to last week's view

Last week's analysis said that a proposed rule restricting US investment in Chinese biotech had not been confirmed and remained a possibility. This week the release of proposed rules on October 22 (Federal Register October 26) was confirmed, and sole licensing is excluded in principle. Last week's inference that the licensing structure for China-origin assets could change if rules were issued should, on the basis of the proposed rules, be narrowed to equity-investment and joint-venture structures. Last week also said the drug-pricing pressure hypothesis for companies holding approved products was neither confirmed nor refuted; this week Lilly's results confirmed a decline in realized prices and, at the same time, that volume growth offset it.

This article is industry analysis based on public information; it is not a recommendation to buy or sell any security and not medical advice.