EV-NEXT: iLoF and Karolinska Institutet apply AI and photonics to improve extracellular vesicle and lipid nanoparticle characterisation

Brussels, December 9th 2024
Summary
  • EV-NEXT is a collaboration between digital health company iLoF and Karolinska Institutet to improve characterisation of extracellular vesicles and lipid nanoparticles using AI-driven photonics.
  • The project was brokered under the EIC Innovation Procurement Programme through the InnoBuyer Call for Solvers and is part of the InnoBuyer initiative funded by Horizon Europe.
  • iLoF reports a new platform feature that distinguishes EV and LNP subtypes by molecular content through multiparameter optical analysis and AI.
  • The project faced practical setbacks such as delays in LNP production and has emphasised iterative development and frequent stakeholder engagement to meet scientific requirements.
  • Next steps focus on testing different sample purities, generating real world evidence, and preparing for regulatory and scalability hurdles before broader adoption.

EV-NEXT: bridging AI photonics and nanomedicine

EV-NEXT is a pilot collaboration that pairs iLoF, a start up specialising in AI powered photonics for drug development, with Karolinska Institutet, one of Sweden's leading medical universities. The initiative was launched through the European Innovation Council's Innovation Procurement Programme. It was selected under the InnoBuyer Call for Solvers which encourages public research institutions and other public actors to work directly with startups and SMEs to co create solutions for specific unmet needs.

Why the project matters

The EV-NEXT pilot focuses on two classes of nanoscale therapeutic and diagnostic agents that are central to current biomedicine research. Extracellular vesicles or EVs are naturally secreted particles that carry proteins and nucleic acids and are studied both as biomarkers and as potential drug carriers. Lipid nanoparticles or LNPs are synthetic lipid based carriers widely used to deliver nucleic acid therapies including mRNA. Both are heterogeneous by nature which complicates quality control, regulatory approval and reproducible manufacturing. EV-NEXT aims to combine multiparameter optical fingerprints with machine learning to improve the specificity and throughput of nanoparticle characterisation. The goal is to speed development, reduce downstream costs, and improve safety monitoring during nanomedicine development.

Extracellular vesicles and lipid nanoparticles explained:Extracellular vesicles are membrane enclosed particles released by cells. They range in size, content and origin which makes them useful as diagnostic indicators but also difficult to classify. Lipid nanoparticles are engineered particles that encapsulate therapeutic payloads like mRNA. Production processes, lipid composition and payload loading create batch to batch variability which has direct implications for efficacy and safety.
iLoF's photonics and AI platform in brief:iLoF uses a photonics based method to read so called optical fingerprints from microlitre volumes of biological samples. These fingerprints are then analysed with AI and multiparameter models to classify particle types and infer molecular content without labels or consumables. The company positions the platform as non invasive, rapid and adaptable for different targets and clinical applications.

Project progress and operational lessons

According to iLoF and project participants the collaboration produced a tangible platform feature that can identify different EV and LNP subtypes based on molecular content. Dr Sara Rocha, Senior Product and Bioclinical Manager at iLoF, said that the EV-NEXT project "has allowed us to refine nanoparticle characterization, introducing a new feature to our platform that identifies different types of EVs and LNPs based on their molecular content."

The project was not without practical problems. iLoF reports delays in LNP production which slowed testing cycles. The team attributes progress to early and regular communication and an iterative development process. Frequent meetings with Karolinska Institutet helped align expectations and adjust the platform to meet the scientific requirements set by the Challenger institution.

ItemDetailsRole
PartnersiLoF and Karolinska InstitutetSolver and Challenger
ProgrammeEIC Innovation Procurement Programme, InnoBuyer Call for SolversBrokered by EIC Business Acceleration Services funded by Horizon Europe
Technical focusMultiparameter optical fingerprinting and AI to classify EVs and LNPsImproved characterisation and QC
Operational challengesDelays in LNP production, need for iterative alignmentAddressed through ongoing stakeholder engagement
Next stepsTesting different sample purities, collecting real world evidence, regulatory compliance workScaling and broader validation

Claims, evidence and the path to deployment

The EV-NEXT project promises several potential benefits that are frequently cited for advanced characterisation tools. These include lower costs, faster development timelines and better product safety through improved quality control. iLoF and allied communications suggest the platform can streamline multiparameter analysis which could translate into economic and environmental gains from more efficient manufacturing. These are plausible outcomes but they are not automatic. Independent validation, head to head comparisons with established assays, and peer reviewed data are required to substantiate specific claims about cost savings or time reductions.

In practice, moving a lab prototype into regulated use in biotechnology and pharmaceutical companies requires meeting a sequence of technical and regulatory gates. For nanoparticle based therapies this typically includes standardisation of measurement, inter lab reproducibility studies, qualification against reference materials and engagement with regulators such as the European Medicines Agency or national competent authorities. The project notes plans to gather real world evidence and to ensure compliance with regulatory standards, but broader adoption will depend on how the platform performs in multicentre validation and how it integrates with existing quality systems.

Broader context: procurement as a tool for public research needs

InnoBuyer is a mechanism within the EIC Business Acceleration Services designed to match public institutions that have unmet research needs with startups and SMEs that can deliver innovative solutions. Funded under Horizon Europe, the initiative aims to de risk early adoption of new technologies by providing a real world testbed and a structured procurement process. EV-NEXT is an example of this model applied to a life sciences use case. Such procurement driven collaborations can accelerate product development but they also expose the limits of pilot projects when scaling to industrial workflows.

What to watch next

Key indicators to follow for EV-NEXT and similar projects include published performance data, results from tests on sample sets with varying purity, external reproducibility studies, and any regulatory interactions that clarify the platform's pathway to qualification. Market uptake will depend on ease of integration into existing laboratory workflows and whether the platform can deliver consistent, validated advantages over established methods.

Dr Rocha highlights project management lessons that are applicable to other public private pilots. She advises clear communication, flexibility in project management, iterative feedback cycles and a long term vision for scalability. Those are sensible prescriptions but they do not replace the need for rigorous validation and transparent reporting when technologies are claimed to improve clinical trial readiness or manufacturing quality.

Follow up and contact

The EV-NEXT team recommends following project updates via social media channels, newsletters and direct contact. The information about the initiative has been shared in the interest of knowledge exchange and is not an official position of the European Commission or other bodies involved.