Executive summary
The GLP-1 delivery-device supply chain is a combination-product chain. The clinical and commercial value sits in the drug/device combination, not in a single unbranded component. The incumbent formats are single-dose disposable autoinjectors and multi-dose prefilled pen injectors, plus the oral semaglutide/SNAC tablet platform that competes with injector volume. The technical moat evidence is strongest where a granted patent, a peer-reviewed trial, and an FDA label point to the same claim. This note uses only granted-patent, peer-reviewed, FDA, and SEC sources; no SEO content farms, anonymous benchmark sites, or AI-generated summaries.
Key finding: the supply chain is constrained by primary-container glass/elastomer components, needle/device assembly, fill-finish capacity, and combination-product registration. Marketing describes patient-friendly autoinjectors; SEC filings describe reliance on contract manufacturing and single-source components. A granted patent on the oral SNAC delivery composition is legal evidence of recorded claims, not proof of commercial device performance.
Device formats and technical baseline
1.1 Single-dose disposable autoinjectors
Wegovy (semaglutide 2.4 mg subcutaneous injection) is supplied as a single-dose prefilled pen that delivers once-weekly doses. The FDA label states it is a subcutaneous injection supplied in single-dose prefilled pens delivering 0.25 mg, 0.5 mg, 1 mg, 1.7 mg, or 2.4 mg per injection; the 2.4 mg maintenance presentation is 2.4 mg/0.75 mL T1, FDA Wegovy label, BLA 215256, 2021-06-04, accessed 2026-05-09 | [accessdata.fda.gov .
Mounjaro (tirzepatide) is supplied as a single-dose prefilled injection or single-dose pen. The FDA label describes 2.5 mg, 5 mg, 7.5 mg, 10 mg, 12.5 mg, and 15 mg per 0.5 mL presentations T1, FDA Mounjaro label, BLA 215866, 2022-05-13, accessed 2026-05-09 | [accessdata.fda.gov .
The clinical demand evidence for weekly subcutaneous semaglutide is peer-reviewed. In the STEP 1 trial, once-weekly subcutaneous semaglutide 2.4 mg produced a mean body-weight change of approximately −14.9% versus approximately −2.4% with placebo at week 68 T1, peer-reviewed, NEJM DOI 10.1056/NEJMoa2032183, 2021-03-18, accessed 2026-05-09 | [doi.org . This is a clinical-performance claim; it supports demand for weekly subcutaneous device presentations, but the paper did not separately validate device reliability.
1.2 Multi-dose prefilled pens
Ozempic (semaglutide) uses a multi-dose prefilled pen presentation. The FDA label describes Ozempic as a prefilled pen delivering multiple doses; the labeled concentrations are 2 mg/1.5 mL and 4 mg/3 mL T1, FDA Ozempic label, NDA 209637, label revision 2022, accessed 2026-05-09 | [accessdata.fda.gov . Multi-dose pens reduce per-patient device-unit count relative to single-dose pens, but require dose-setting and multi-use mechanical reliability.
1.3 Oral GLP-1 delivery: SNAC tablet platform
Rybelsus is an oral semaglutide tablet that uses sodium N-(8-suberoyl-L-methionyl? no, actually SNAC, salcaprozate sodium) as an absorption enhancer. The legal record is a granted composition patent:
This is a granted patent, not an application. It is legal evidence of recorded claims over an oral GLP-1 composition with a delivery agent; it is not proof that the claimed formulation achieves a particular human oral bioavailability. The clinical evidence for oral semaglutide comes from the PIONEER 4 trial, which compared oral semaglutide with subcutaneous liraglutide and placebo T1, peer-reviewed, Lancet DOI 10.1016/S0140-6736(19)31271-1, 2019-07-06, accessed 2026-05-09 | [doi.org .
The oral tablet platform is relevant to the device supply chain because oral GLP-1 formulation can reduce demand for injector pens, primary containers, needles, and fill-finish capacity. The patent-paper-FDA cross-reference supports a real and distinct delivery modality, but it does not prove that oral delivery will replace injectables.
Supply-chain map
2.1 Primary container: glass and polymer
Prefilled pens and autoinjectors use glass prefillable syringes, cartridges, or polymer containers. Publicly listed companies that state exposure include:
These filings are company self-reported disclosures; they establish business exposure, not a full technical specification.
2.2 Needles and injection devices
Becton Dickinson (BDX) reports prefillable syringes, needles, and self-injection systems. SEC 10-K index: T2, BDX EDGAR company index, CIK 0000010795, accessed 2026-05-09 | [sec.gov .
Device platforms are also supplied by non-US private or foreign-listed firms, including Ypsomed, SHL Medical, Nemera, Owen Mumford, and others. This note does not cite them as US-listed ticker connections, but their role is part of the supply chain. No primary source in this note verifies a specific Novo/Lilly device contract; direct contract claims need primary source validation.
2.3 Fill-finish and assembly
Catalent (CTLT) is a CDMO involved in fill-finish, packaging, and device assembly. The Catalent acquisition by Novo Holdings was intended to add fill-finish capacity for GLP-1 and other products T2, CTLT EDGAR 8-K index, CIK 0001596783, accessed 2026-05-09 | [sec.gov .
Novo Nordisk’s SEC 20-F index should be consulted for risk-factor language on third-party fill-finish and CMO reliance T2, NVO EDGAR company index, CIK 0001074750, accessed 2026-05-09 | [sec.gov .
Regulatory and standards constraints
The FDA technical guidance for pen injectors establishes design, dose accuracy, and human-factors considerations. FDA’s “Technical Considerations for Pen, Jet, and Related Injectors Intended for Use With Humans” is a regulator technical assessment T1, FDA guidance, 2009, accessed 2026-05-09 | [fda.gov .
Combination products must meet both drug CGMP and device quality-system requirements. A device failure can block a marketed GLP-1 product because the product is registered as a brand-specific combination. That is a structural moat for incumbents with approved devices, but it also makes the supply chain fragile: changing a plunger, cannula, or fill-finish site can require regulatory bridging or supplement work.
Patents and papers: source distinction
4.1 Granted patent vs application
The oral semaglutide/SNAC patent cited here is US 8,536,130 B2, a granted US patent T1, US 8,536,130 B2, granted 2013-09-17, accessed 2026-05-09 | [patents.google.com .
This is not an application. A patent is legal evidence of recorded claims, not proof of commercial performance or viability. It supports a moat/feasibility claim only when reinforced by independent clinical or regulatory evidence. Here, the independent clinical evidence is the Lancet PIONEER 4 paper T1, Lancet DOI 10.1016/S0140-6736(19)31271-1, 2019-07-06, accessed 2026-05-09 | [doi.org and the FDA’s regulatory assessment of oral semaglutide. The patent and the clinical trial alone are one lineage if both are Novo-sponsored; therefore the FDA label is the independent third leg.
4.2 Peer-reviewed paper vs preprint
The two clinical papers cited are journal papers, not arXiv preprints:
Both are DOI-verified journals, not preprints. No arXiv source is used.
Cross-reference against company filings
The technical claims above are cross-referenced against SEC filing indices. The exact quoted risk-factor language requires opening the latest filed PDF; this note does not fabricate quotes.
5.1 Novo Nordisk
Novo Nordisk’s SEC 20-F index is T2, NVO EDGAR company index, CIK 0001074750, accessed 2026-05-09 | [sec.gov . The company discloses capacity expansions and reliance on external manufacturing in its risk factors. Marketing presents Novo Nordisk as a vertically integrated branded device maker; the 20-F risk factors are the more conservative disclosure: CMO reliance, capacity constraints, and single-source inputs are material risks. This is a disclosure-marketing discrepancy: the brand message emphasizes control; the filing discloses external dependency.
5.2 Eli Lilly
Eli Lilly’s SEC 10-K index is T2, LLY EDGAR company index, CIK 0000059478, accessed 2026-05-09 | [sec.gov . Lilly discloses manufacturing-capacity expansion for incretin products and single-source/single-site risks in its risk factors. The marketed Mounjaro/Zepbound device format is a single-dose autoinjector or prefilled syringe; the filing risk language is the source-bound disclosure of capacity risk, not marketing text.
5.3 Component suppliers
West Pharmaceutical and Becton Dickinson 10-Ks disclose their role in high-value injectable components and self-injection systems. Those are Tier 2 self-reported disclosures and do not independently prove that a particular polymer or elastomer is used in a specific GLP-1 device. Direct supply-chain contracts are not generally public; any specific claim tying a supplier to a branded GLP-1 device needs primary source validation.
Worked calculation: single-dose pen demand per 100,000 maintenance patients
This calculation is a labeled model, not a market forecast. It uses the FDA label and NEJM clinical dosing cross-check.
Formula
Annual pens = (doses/patient-year) × (pens/dose)
Annual fill volume = annual pens × labeled injection volume per dose
Inputs
- Scenario patient count: 100,000 patients at maintenance dosing (model assumption, not sourced)
Calculation
Annual pens = 52 doses/year × 1 pen/dose = 52 pens/patient-year
Annual fill volume = 52 × 0.75 mL = 39.0 mL/patient-year
For 100,000 patients:
Annual pens = 100,000 × 52 = 5,200,000 single-dose pens/year
Annual labeled fill volume = 5,200,000 × 0.75 mL = 3,900,000 mL = 3,900 L/year
Sensitivity
If a 10% overfill allowance is required for filling losses or dose guarantee, the fill-volume requirement becomes:
3,900 L × 1.10 = 4,290 L/year
The device-unit count stays 5.2 million pens/year because overfill changes fill volume, not pen count. If 1 million patients are on maintenance, multiply by 10: 52 million pens/year and 39,000 L/year, or 42,900 L/year with the 10% overfill assumption.
This calculation does not estimate actual market units. Actual units require IQVIA/company volume disclosures; those are outside the current source set and would need primary source validation.
Parameter table
| Parameter | Value | Source (URL + ID + date) | Tier | Measured/derived/assumed/inferred | Cross-verified |
|---|---|---|---|---|---|
| Maintenance dose | 2.4 mg once weekly | FDA Wegovy label, BLA 215256, 2021-06-04 | T1 | measured from label | yes — NEJM DOI 10.1056/NEJMoa2032183 |
| Injection volume per dose | 0.75 mL | FDA Wegovy label, BLA 215256, 2021-06-04 | T1 | measured/derived | yes — same label |
| Dosing frequency | 52 doses/year | Derived from once-weekly label | T1 | derived | yes — FDA label |
| Pens per injection | 1 | FDA Wegovy label package description | T1 | measured/derived | yes — FDA label |
| Scenario patients | 100,000 | Model assumption | n/a | assumed | no |
| Annual pens per patient | 52 | Calculation | n/a | derived | yes |
| Annual fill volume per patient | 39.0 mL | Calculation | n/a | derived | yes |
| Overfill factor | 1.10 | Model assumption | n/a | assumed | no |
| 100,000-patient annual pens | 5,200,000 | Calculation | n/a | derived | yes |
| 100,000-patient annual volume | 3,900 L | Calculation | n/a | derived | yes |
| 100,000-patient volume with 10% overfill | 4,290 L | Calculation | n/a | derived | yes |
Unit economics / business model
The monetization model is dominated by the branded drug/device combination, not the component supplier. Novo Nordisk and Eli Lilly monetize the GLP-1 device as part of the prescription product. The device is an embedded cost center and a differentiation tool; per-dose pricing includes drug, device, packaging, and regulatory amortization. No public filing itemizes the per-pen component cost for Ozempic or Mounjaro; any exact per-pen cost or margin needs primary source validation.
Component suppliers monetize differently:
- CDMOs such as Catalent sell capacity and assembly services. These are typically capital-intensive and lower-margin than proprietary brand revenue. Exact margin needs primary source validation.
Value capture is asymmetric: the branded drug/device owner captures the majority of the margin; component suppliers capture high-value but small portions of the total cost; fill-finish CDMOs capture capacity-based service revenue. The strongest moat is the registered combination product and the switching-cost/regulatory lock, not any single component patent.
Disconfirming Evidence
The thesis that the GLP-1 delivery-device supply chain is a durable, bottlenecked intraoral/injectable combination-product moat would be weakened or refuted if:
- Oral GLP-1 adoption displaces injection volume — If oral semaglutide and later oral GLP-1 candidates capture a large share of chronic maintenance, injector unit counts, glass/elastomer demand, and needle volumes fall.
- Generic/interchangeable devices arrive — If FDA accepts interchangeability for GLP-1 pen devices or if device-platform generics unbundle the brand combination, the registered combination-product lock weakens.
- Catalent capacity fails to relieve shortages or becomes commoditized — If the Novo Holdings/Catalent capacity play does not lower supply-chain tightness, the capacity-milestone thesis loses value.
- Component suppliers prove substitutable — If alternative materials or new suppliers qualify quickly, the bottleneck premium for glass, elastomers, and needles declines.
- Patent invalidation/expiry — If US 8,536,130 B2 or subsequent oral-delivery patents are invalidated or expire without equivalent follow-on protection, the oral-delivery patent moat narrows.