Exosome and Extracellular Vesicle Drug Delivery — The CMC Regulatory Path Through CBER

Exosomes and other extracellular vesicles occupy one of the most jurisdictionally ambiguous positions in FDA regulatory science: are they biological products regulated by CBER as cell-derived therapeutics, or are they drug delivery vehicles regulated by CDER because the therapeutic payload is a small molecule or nucleic acid? The answer depends on primary mode of action analysis performed case by case, and it determines which center reviews the IND, which guidance framework governs the CMC package, and which characterization standard — MISEV2018’s consensus criteria or a small-molecule drug product framework — the sponsor must satisfy. Sponsors who file without first resolving this jurisdictional question are filing into regulatory uncertainty that a pre-IND meeting exists specifically to resolve.

Exosome and EV drug product CMC packages fail at FDA review not because the vesicle biology is poorly understood, but because the primary mode of action classification was never explicitly analyzed and documented, and because the MISEV2018 consensus characterization panel — tetraspanin markers, negative markers, particle concentration and size distribution — was treated as a research-grade nicety rather than the regulatory characterization standard FDA reviewers now expect as baseline.

CDER vs. CBER Jurisdiction — The Primary Mode of Action Analysis That Determines Which Center Reviews Your EV Drug Product IND

FDA’s Office of Combination Products applies primary mode of action (PMOA) analysis to determine jurisdiction for products that combine a biological vesicle carrier with a therapeutic payload: when the EV itself is the biologically active therapeutic entity — as in native exosomes carrying an endogenous cargo intended to exert a biological effect — CBER’s cell and gene therapy framework typically applies, with characterization and manufacturing expectations analogous to other cell-derived biologics. When the EV is engineered purely as a delivery vehicle for a separate small-molecule or nucleic acid payload with no independent biological activity attributed to the vesicle itself, CDER’s drug product framework may apply instead, more analogous to the classification logic already established for liposomal and polymer nanoparticle carriers. This is not a determination a sponsor should make unilaterally and discover was wrong at IND filing; a Type B pre-IND meeting requesting FDA’s classification determination, with the PMOA rationale documented in the briefing package, is the standard mechanism for resolving jurisdiction before manufacturing and characterization strategy is locked around the wrong regulatory framework.

MISEV2018 Characterization as the Regulatory Baseline — Tetraspanin Markers, Negative Controls, and the Particle-to-Protein Ratio That Establishes Purity

The International Society for Extracellular Vesicles’ MISEV2018 guidelines establish the consensus characterization panel that FDA reviewers now expect as the baseline evidence of EV identity and purity, regardless of which center has jurisdiction: nanoparticle tracking analysis (NTA) confirming modal particle diameter in the 80–150 nm range for small EVs with D90 at or below approximately 200 nm, positive identification of at least two tetraspanin markers (commonly CD9, CD63, CD81) by Western blot or flow cytometry confirming EV membrane identity, and negative marker testing — Calnexin absence confirms the preparation is free of endoplasmic reticulum-derived cellular debris, a critical purity indicator distinguishing genuine EVs from cell lysate contamination. Beyond marker identity, the particle-to-protein ratio (commonly targeted at 1×108 particles per microgram of protein or higher) is the quantitative purity metric that distinguishes a well-purified EV preparation from one still carrying substantial co-isolated soluble protein contamination, and a characterization package presenting NTA size data and tetraspanin positivity without a particle-to-protein ratio and without Calnexin-negative confirmation has demonstrated the preparation contains vesicles of roughly the right size, without demonstrating the preparation is free of the co-purifying contaminants that MISEV2018 specifically flags as the most common EV purity failure mode.

Cargo Loading Quantification and RNase Protection — The ddPCR and Nuclease Challenge Standard for siRNA-Loaded EV Products

For EV products engineered to carry an siRNA or other nucleic acid payload, two characterization elements go beyond MISEV2018’s identity panel and directly address therapeutic performance: cargo copy number quantification by droplet digital PCR (ddPCR), with typical loaded EV preparations carrying on the order of 5–50 siRNA copies per vesicle depending on loading method and target, and RNase protection assessment, in which the EV preparation is challenged with RNase A in the absence of a permeabilizing detergent, and a well-encapsulated cargo shows RNase protection of 70% or greater — meaning the majority of the loaded siRNA is genuinely encapsulated within the vesicle lumen rather than merely surface-adsorbed, since surface-adsorbed cargo is degraded rapidly in circulation and does not reach the intended intracellular target. A specification stating an average cargo loading number derived only from total EV-associated siRNA (measured after simple ultracentrifugation wash steps) without an RNase protection challenge has quantified total cargo association without distinguishing genuinely encapsulated, protected payload from surface-adsorbed material that will not survive systemic administration.

The XGene EV CMC Regulatory Architecture

The XGene EV CMC Regulatory Architecture is a structured regulatory and characterization strategy for exosome and extracellular vesicle drug products navigating CBER/CDER jurisdictional ambiguity.

1. PMOA Jurisdictional Analysis — Document the primary mode of action rationale and request FDA’s classification determination through a Type B pre-IND meeting before manufacturing and characterization strategy is finalized around an assumed center. 2. MISEV2018 Characterization Panel — Build NTA particle size distribution, tetraspanin-positive/Calnexin-negative marker testing, and particle-to-protein ratio as the mandatory baseline identity and purity package, regardless of which center reviews the IND. 3. Cargo Loading and RNase Protection Quantification — For engineered cargo-loaded products, establish ddPCR-based copy number quantification alongside an RNase protection challenge assay demonstrating genuine encapsulation versus surface adsorption. 4. Center-Appropriate CMC Package Assembly — Build the 3.2.S/3.2.P (or equivalent CBER) submission package aligned to the confirmed jurisdictional framework, not a hybrid package hedging between both.

The output is a submission-ready characterization and CMC package aligned to a confirmed regulatory jurisdiction, built on the MISEV2018 consensus standard FDA reviewers expect as baseline evidence of EV identity and purity.

An EV drug product program that builds its characterization package around vesicle biology alone, without first resolving CBER/CDER jurisdiction through a documented PMOA analysis, is building toward a review process it cannot predict — and the jurisdictional question left unresolved at IND filing does not become easier to answer once clinical manufacturing is underway.

For your exosome or EV drug product program, can you confirm today whether your regulatory strategy documents a primary mode of action analysis supporting your assumed CBER or CDER jurisdiction, and whether your characterization package includes the full MISEV2018 panel — tetraspanin-positive markers, Calnexin-negative confirmation, and a particle-to-protein ratio — rather than particle size and concentration data alone?