What extracellular vesicles are and why they matter for drug delivery

Extracellular vesicles (EVs) are nanoscale, membrane-bound particles naturally secreted by virtually all cell types, offering intrinsic biocompatibility, low immunogenicity, and the capacity to traverse biological barriers that defeat conventional synthetic nanocarriers. These properties — not replicated by liposomes or polymer nanoparticles — have made EVs the subject of intense R&D investment across oncology, neurology, and infectious disease.

EVs are broadly categorised as exosomes (endosome-derived, approximately 30–150 nm), microvesicles (plasma membrane-budded, approximately 100–1000 nm), and apoptotic bodies. A growing sub-category of EV mimetics — artificial vesicles engineered from cells or lipid materials to recapitulate EV properties — is also prominently represented in recent literature, as defined by researchers at Kyungpook National University in their 2020 review of EV-based drug delivery systems.

The core technical challenge that defines the field is threefold: how to isolate consistent vesicle populations from biological sources; how to engineer EV surfaces and luminal cargo for therapeutic specificity; and how to ensure efficient cellular uptake, intracellular trafficking, and endosomal escape once the EV reaches its target. According to WIPO‘s tracking of biologic delivery modalities, nanoscale biological carriers represent one of the fastest-growing patent categories in the pharmaceutical space.

Cargo loading strategies span endogenous cell modification — where producer cells are engineered to package specific therapeutic molecules — and exogenous post-isolation loading via electroporation, sonication, co-incubation, and click chemistry. Each approach involves distinct trade-offs in loading efficiency, cargo integrity, and scalability. Surface engineering for targeting can be achieved through genetic, chemical, or peptide-based functionalization, applied either before or after isolation.

From characterisation to clinical translation: the EV innovation arc

The EV drug delivery field has passed through four identifiable maturation stages between 2012 and 2023, moving from foundational biology into active clinical translation — a trajectory visible in both the character of published literature and the timing of patent filings retrieved in this dataset.

The foundational period (2012–2016) centred on characterisation methodology and establishing EVs as viable drug carriers. Work from Harvard Medical School (2012) on alternative characterisation methods and Tokyo Medical University (2016) on EV therapeutic system development established the “message in a bottle” delivery paradigm — the concept that EVs naturally protect their cargo from degradation while transporting it between cells.

During the translational ramp-up (2017–2019), research pivoted toward clinical obstacles. Boston University (2017) articulated the blood-brain barrier crossing advantage, while a position paper from the Finnish Red Cross Blood Service, widely cited through 2019, codified regulatory frameworks for EV-based therapeutics. Codiak BioSciences filed patents in this period staking IP on EV vaccine platforms.

The engineering maturity phase (2020–2022) produced the largest cluster of records in this dataset. Scaffold protein engineering, hybrid EV systems, and non-mammalian EV sources all entered the literature concurrently, with LYSATPHARMA GMBH filing multiple active patents across Israel and Singapore between 2018 and 2022. The most recent phase — advanced engineering and clinical convergence (2023–present) — is characterised by precision protein delivery, bioinformatics-guided scaffold discovery, and multimodal intracellular release systems.

“The largest cluster of EV engineering records falls in the 2020–2022 window — scaffold protein engineering, hybrid systems, and non-mammalian EV sources all entered the literature concurrently, signalling a field-wide shift from characterisation to construction.”

Four technology clusters defining the competitive frontier

The EV drug delivery patent and literature landscape organises into four distinct technology clusters, each addressing a different layer of the delivery problem — from how cargo gets into the vesicle, to how the vesicle finds its target, to how cargo escapes into the cell.

Cluster 1: Scaffold Protein-Mediated Cargo Loading

Scaffold protein-mediated loading leverages endogenous EV-sorting proteins as molecular anchors to tether or encapsulate therapeutic cargo at high density, enabling both luminal loading and surface display. This is the most IP-active engineering cluster in the dataset. Codiak BioSciences (2021) identified PTGFRN and BASP1 as EV-preferentially sorted scaffold proteins enabling loading of cytokines, antibody fragments, RNA-binding proteins, Cas9, and TNF superfamily members. The University of Oxford (2023) subsequently identified N-glycosylation as an EV-sorting signal and exploited PTTG1IP as a scaffold for macromolecular cargo with self-cleaving linkers for post-delivery cargo release — a bioinformatics-guided approach that signals the shift from empirical to computationally guided scaffold discovery.

A 2023 study further demonstrated that rational design of proteins to associate with lipid rafts at the plasma membrane enhances loading of structurally diverse transmembrane and peripheral membrane proteins into EVs — a generalizable strategy applicable across diverse cargo types.

Cluster 2: Surface Functionalization for Targeted Delivery

Surface functionalization covers chemical, peptide, and genetic strategies — applied either post-isolation or via producer-cell engineering — to confer active targeting to specific cell types or tissues. The National University of Singapore (2021) demonstrated covalent ligation of EGFR-targeting peptides and nanobodies to EV surfaces via protein ligases, enabling targeted paclitaxel delivery in EGFR-positive lung cancer xenografts. Separately, the Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry (Russian Academy of Sciences, 2021) used a nanobody-CD206 fusion on the EV surface to selectively redirect cargo delivery to antigen-presenting cells — dendritic cells and macrophages — with implications for immunotherapy and vaccine applications.

Cluster 3: Hybrid EV Systems and Fusogenic Engineering

Hybrid systems combine EVs with liposomes or incorporate fusogenic viral proteins to overcome cargo-loading limitations and endosomal entrapment — the principal barrier to cytosolic drug delivery. University Medical Center Utrecht (2022) described the rational combination of EV biological homing properties with synthetic nanoparticle engineering versatility to address immune clearance and poor targeting of either platform alone. Kyoto University (2022) constructed hybrid EVs by fusing liposomes with insect-cell EVs expressing PD-1 and baculoviral fusogenic gp64, demonstrating enhanced uptake and endosomal membrane fusion in HeLa cells. The most advanced example in this dataset is the VEDIC and VFIC systems (2023), which combine a self-cleaving mini-intein and VSV-G fusogen to achieve endosomal escape and cytosolic release of protein cargo both in vitro and in vivo.

Cluster 4: Non-Mammalian and Alternative Source EVs

A distinct and rapidly growing cluster exploits EVs from bacteria, plants, and red blood cells — sources offering scalability and cost advantages unavailable from mammalian cell culture. Shanghai Zhongye Hospital (2022) reviewed bacterial EV biogenesis and their application as drug delivery platforms leveraging inherent immunostimulatory and targeting properties. Newland Biotechnology (2022) reviewed plant-derived EVs as oral and transdermal delivery vehicles for anti-inflammatory and anticancer applications, noting scalable extraction from edible sources. Mahidol University (2021) examined red blood cell-derived EVs exploiting CD47-mediated macrophage evasion for encapsulation and systemic delivery of siRNA, miRNA, and mRNA — a particularly promising approach for nucleic acid therapeutics where immune clearance is a critical obstacle. Standards bodies including ISO are beginning to address characterisation standards for non-mammalian EV preparations.

Application domains: oncology, CNS, inflammation, and beyond

Oncology is the most extensively represented application domain across retrieved records, but the EV drug delivery field spans at least five distinct clinical areas — each leveraging different aspects of EV biology.

Oncology

Cancer applications encompass tumor-targeted chemotherapy delivery, nucleic acid-based gene silencing, immunomodulation, and cancer vaccines. Tongji University School of Medicine (2023) comprehensively covered EV roles across tumor biogenesis and therapy. ILIAS Biologics Inc. (2022) detailed delivery of miRNA, siRNA, mRNA, and chemotherapeutics into tumor-targeted exosomes via both active and passive mechanisms. The National Center for Nanoscience and Technology in Beijing (2022) specifically addressed EV-loaded CRISPR/Cas gene-editing systems for cancer gene therapy. On the patent side, Codiak BioSciences holds pending claims in Israel and Singapore covering EV-based delivery of tumor antigens, adjuvants, and immune modulators for cancer and graft-versus-host disease (GvHD).

Central Nervous System Disease

EVs’ intrinsic capacity to traverse the blood-brain barrier has generated a substantial body of CNS-focused work. The University of Tennessee Health Science Center (2022) reviewed EV strategies for CNS drug penetration including clinical trial status, while Ghent University (2021) surveyed cell-type-specific EVs and engineered targeting ligands for brain delivery via peripheral administration routes. Despite mechanistic validation, CNS delivery via EVs remains clinically unproven at scale — making it, according to the strategic analysis in this dataset, the highest-value unmet clinical application in the field. Research published in journals tracked by Nature has highlighted BBB-crossing EVs as a priority area for neurodegenerative disease research.

Inflammatory & Autoimmune Disease

Southeast University, Nanjing (2020) reviewed unmodified and engineered EV-based therapies for inflammation-linked disorders. LYSATPHARMA GMBH holds an active patent in Israel (2022) claiming platelet-lysate-derived EV preparations for inflammatory, neurodegenerative, cardiovascular, autoimmune, and dermatologic diseases, as well as GvHD and transplant rejection — one of the broadest indication claims in the dataset.

Regenerative Medicine

Catholic Kwandong University (2021) and Dankook University (2021) framed the use of stem-cell and mesenchymal stem cell (MSC)-derived EVs in tissue repair, wound healing, and orthopaedic applications. The regenerative medicine domain is notable for its reliance on unmodified or minimally engineered EVs, where the natural paracrine signalling content of MSC-derived vesicles provides therapeutic benefit without extensive engineering.

Infectious Disease & Vaccination

Post-pandemic momentum has driven direct application of EV engineering to infectious disease vaccination. The Istituto Superiore di Sanità (2022) demonstrated EV-based delivery of SARS-CoV-2 antigens to elicit CD8+ T cell responses, directly linking EV engineering to pandemic preparedness. Codiak BioSciences holds a pending Singapore patent extending its vaccine delivery claims to infectious disease and autoimmune disease indications. According to clinical trial registries tracked by NIH, EV-based vaccine candidates are among the most actively enrolled biologic delivery modalities in early-phase trials.

The IP landscape: who holds the key patents

Among patent records specifically retrieved in this dataset, two commercial assignees account for the entirety of granted and active patent filings — a concentration that has significant freedom-to-operate implications for any organisation entering the EV drug delivery space.

LYSATPHARMA GMBH (Germany) is the most patent-prolific assignee in this dataset, with at least 5 active or previously active patent records across Israel and Singapore jurisdictions. All are directed to human platelet-lysate-derived EV pharmaceutical preparations, with filing dates spanning 2018–2022 — indicating a sustained and geographically distributed IP strategy focused on a single, well-defined source material.

Codiak BioSciences, Inc. (USA, Cambridge MA) holds at least 2 pending patent records in Israel and Singapore covering EV-based vaccine delivery platforms with broad disease indication coverage. These claims are built on the PTGFRN/BASP1 scaffold technology identified in Codiak’s 2021 academic publication — a deliberate platform IP approach that creates freedom-to-operate constraints for any organisation seeking to use similar scaffold proteins for EV cargo loading.

On the academic side, the innovation landscape is broadly distributed with no single dominant institution. Notable geographic clusters include South Korea (Kyungpook National University, ILIAS Biologics, Konkuk University), the Netherlands (Radboud University Medical Center, University Medical Center Utrecht), Japan (Tokyo Medical University, Kyoto University, Osaka University), China (Tongji University, Jiangsu University, National Center for Nanoscience and Technology Beijing), and the United Kingdom (University of Oxford). The United States spans commercial platform development at Codiak BioSciences through to clinical translation at Michigan State University, Harvard Medical School, and Stanford University. Global patent filing trends are tracked by EPO, whose annual Patent Index consistently shows biologics delivery as a high-growth category.

Five emerging directions shaping EV drug delivery through 2026

Based on records published in 2022–2023, five directional signals are visible in the dataset — each representing a distinct technical or commercial opportunity for organisations active in the EV drug delivery space.

1. Bioinformatics-Driven Scaffold Discovery

The PTTG1IP platform from the University of Oxford (2023) exemplifies a shift from empirical to computationally guided identification of EV-sorting motifs. Specifically, N-glycosylation was identified as an EV-sorting signal through bioinformatic analysis, then validated as a design principle for next-generation loading scaffolds. This approach is expected to accelerate identification of additional high-efficiency sorting proteins beyond PTGFRN, BASP1, and PTTG1IP.

2. Cytosolic Protein Delivery via Endosomal Escape Engineering

The VEDIC and VFIC systems (2023) — combining a self-cleaving mini-intein with VSV-G fusogen — represent the most advanced published solution to the endosomal escape problem. This challenge has historically been the Achilles’ heel of EV-based systems: without reliable cytosolic release, protein therapeutics are degraded rather than delivered. Investment in fusogenic protein engineering (VSV-G, gp64-based hybrids) and intein-based self-cleavage represents the highest-leverage technical challenge for the next two to three years.

3. Plant-Derived and Bacterial EVs for Scalable, Low-Cost Delivery

Bacterial outer membrane vesicles and plant-derived EVs can be produced at scales and costs unachievable with mammalian cell culture, potentially enabling oral and topical formulations. IP in this space remains comparatively open relative to exosome-based platforms, making it an attractive entry point for organisations seeking to avoid existing scaffold protein IP estates. Multiple records identify GMP manufacturing and scalability as near-term commercialisation bottlenecks for mammalian EV platforms — a constraint that non-mammalian sources directly address.

4. EV-Based Antiviral and Vaccine Platforms

Post-pandemic momentum has positioned EVs as a versatile antigen-display platform competitive with mRNA-LNP systems. The Istituto Superiore di Sanità (2022) demonstrated EV-based delivery of SARS-CoV-2 antigens to elicit CD8+ T cell responses, while the Codiak BioSciences vaccine patent family extends claims to infectious disease and autoimmune disease indications — establishing EVs as a potential platform for next-generation vaccine development.

5. Lipid Raft Engineering for Enhanced Cargo Loading

A 2023 study introduced rational protein trafficking to lipid rafts as a generalizable strategy to increase loading density across diverse cargo types. By designing proteins to associate with lipid rafts at the plasma membrane, researchers demonstrated enhanced loading of structurally diverse transmembrane and peripheral membrane proteins into EVs — a mechanistic advance with broad applicability across the scaffold protein and hybrid systems clusters.

“CNS delivery is the highest-value unmet clinical application in EV drug delivery: EV-based blood-brain barrier traversal is mechanistically validated but clinically unproven at scale, and patent density in this dataset remains low — a white-space opportunity.”