Targeting mitochondrial quality control in osteoarthritis with GrpE-like 1-loaded synovial mesenchymal stromal/stem cell small extracellular vesicles

Abstract

Recent studies have demonstrated that synovial mesenchymal stem cell-derived small extracellular vesicles (EVs) engineered to deliver GrpE-like 1 activated PTEN-induced kinase 1-dependent mitophagy and restored chondrocyte homeostasis. This study revealed that interleukin-1β-challenged chondrocytes exhibited efficient cargo transfer, increased mitophagy signaling with reduced p62 levels, lower oxidative stress, and a shift toward matrix preservation, characterized by higher collagen II and aggrecan levels, and lower matrix metallopeptidase 13 and ADAM metallopeptidase with thrombospondin type 1 motif 5 levels. In a rat knee osteoarthritis model, intra-articular dosing preserved cartilage architecture and improved histological scores. Collectively, these findings suggest that EV-based delivery of mitochondrial regulators is a plausible disease-modifying strategy, rather than purely symptomatic care. Building on this evidence, this editorial distills key advances and outlines near-term research and translational priorities, including standardized EV characterization, pharmacokinetics, dosing, safety, and manufacturability. The suitability of GrpE-like 1-loaded small EVs for early-stage osteoarthritis was also evaluated.

Key Words: Osteoarthritis; Extracellular vesicles; Synovial mesenchymal stem cells; GrpE-like 1; PTEN-induced kinase 1/Parkin-mediated mitophagy

Core Tip: This editorial frames GrpE-like 1-loaded synovial extracellular vesicles as a mitochondria-centric strategy for osteoarthritis and translates bench evidence into an executable roadmap. We operationalize identity/purity panels, dual-metric dose (particles and protein), shared exposure anchors, and a mechanism-linked potency assay for lot release and comparability. Early trials should prespecify Outcome Measures in Rheumatology-Osteoarthritis Research Society International endpoints with quantitative magnetic resonance imaging cartilage thickness as an exploratory structural anchor. Transparent method registration and time-bounded regulatory messaging are emphasized to improve reproducibility, cross-study comparison, and readiness for first-in-human evaluation.

INTRODUCTION

Osteoarthritis (OA) is a leading cause of pain and disability worldwide and lacks approved disease-modifying OA drugs. The current treatment remains primarily symptomatic and does not reliably alter structural progression. The contemporary perspective recognizes OA as a heterogeneous, whole-joint disorder that involves articular cartilage, synovium, subchondral bone, and neurosensory pathways, complicating target selection and trial design[1,2]. These facts support the argument for mechanistic strategies that directly target the cellular drivers of matrix failure and joint inflammation, rather than the downstream symptoms[1].

Mitochondrial dysfunction is a prominent driver among these; chondrocytes in OA exhibit loss of membrane potential, impaired oxidative phosphorylation, and excessive reactive oxygen species, which predispose cells to catabolism and death[3]. Converging pre-clinical data suggest that defective PTEN-induced kinase 1 (PINK1)-Parkin mitophagy contributes to the accumulation of damaged mitochondria, whereas restoring this pathway reduces oxidative stress and slows structural deterioration in animal models[4-6]. This body of evidence frames mitochondrial quality control, particularly mitophagy and proteostasis, as a credible axis for disease-modifying OA drug discovery[3-5].

Parallel advances in cell-free regenerative therapeutics have positioned mesenchymal stromal/stem cell (MSC) small extracellular vesicles (sEVs) as practical delivery vehicles that can modulate joint biology with lower immunogenicity and greater manufacturing flexibility than those of live cells. Synovial MSC (SMSC) sEVs are of particular interest for cartilage repair because of their native articular niche and trophic signaling. Furthermore, they can be engineered to carry functional proteins that affect mitochondrial homeostasis[7-9]. Within the mitochondrial proteostasis network, GrpE protein homolog 1 (GRPEL1) acts as a nucleotide exchange factor for mitochondrial 70-kDa heat shock protein (mtHsp70), supporting protein import and folding. Recent structural and biophysical studies have clarified its distinct role relative to GRPEL2 and established a mechanistic basis for using GRPEL1 to stabilize organelle quality control in stressed chondrocytes[10,11]. These considerations motivated a focused evaluation of SMSC-sEVs delivery of GRPEL1 as a disease-modifying strategy for OA.

KEY FINDINGS AND INNOVATIONS

A published study in the World Journal of Stem Cells demonstrated that SMSC-sEVs engineered to deliver GRPEL1 activate PINK1-mediated mitophagy and improve cartilage structure in arthritis models. In vitro, it demonstrated that interleukin-1β-challenged cells exhibited efficient cargo transfer, accompanied by an increase in canonical mitophagy markers and a decrease in oxidative stress. This is accompanied by a shift toward matrix preservation, as indicated by higher collagen II and aggrecan levels, and lower matrix metallopeptidase 13 and ADAM metallopeptidase with thrombospondin type 1 motif 5 levels[12]. In vivo, standardized histological assessment demonstrated improved cartilage architecture, aligning cellular rescue with tissue-level benefit under accepted pre-clinical readouts[12,13].

This study broadly aligns with Minimal Information for Studies of Extracellular Vesicles 2023 (MISEV2023) expectations in terms of reporting quality, including particle sizing, extracellular vesicle (EV) marker panels, and function-linked assays. However, there is still room to improve the transparency of negative markers, pre-analytical variables, and orthogonal dose metrics (particles, proteins, and cargo copies)[14]. Two methodological innovations elevate translational relevance. First, this program adopts a protein cargo (GRPEL1) to enable immediate mitochondrial target engagement; the detailed comparative rationale is developed later in the article. The present GRPEL1 strategy is contextualized by prior toolkits, such as EV-based protein loading via optically reversible protein-protein interactions (EXPLOR) (optically controlled protein loading) and lysosomal-associated membrane protein 2A/KFERQ motif-guided sorting, which demonstrate the feasibility of EV protein transfer[15,16]. Second, the mechanistic plausibility is independently substantiated by structural and biophysical research on mtHsp70-GRPEL1. This research suggests that GRPEL1 is situated at an upstream control point for mitochondrial proteostasis, which can converge on mitophagy and oxidative stress relief[10,11]. A comparative rationale between protein and RNA cargo is developed later in the article.

The in vivo study design strengthens external validity using Osteoarthritis Research Society International (OARSI)-compliant rat histopathology, which facilitates cross-study comparisons and dose selection[15]. The formulation literature indicates that surface-charge modulation prolongs joint residence and improves uptake of sEVs by cartilage, providing a readily testable lever for optimizing the intra-articular pharmacokinetics (PK) of GRPEL1-EVs[17]. In parallel, Good Manufacturing Practice (GMP)-aligned development of clinical-grade umbilical cord-derived MSCs (UC-MSCs) derived sEVs, together with rigorous preclinical validation, demonstrates the feasibility of constructing full chemistry, manufacturing, and controls (CMC) packages and supports a credible path toward early-phase trials once potency and safety are established[18].

METHODOLOGICAL APPRAISAL AND STRENGTH OF EVIDENCE

Chain of evidence

This study linked the engineered SMSC sEVs to mechanism-level rescue and joint-level improvement. GRPEL1 cargo delivery demonstrated PINK1 loss-of-function blunt efficacy, and histology confirmed the tissue benefits under standardized scoring. This alignment across levels, such as cargo engagement, pathway dependency, and structure, supports disease-modifying potential, while highlighting the necessity of defining the dose-exposure-response relationship before translation[12,13].

EV identity and dose reporting

Terminology follows MISEV2023: We use EVs as the generic term and sEVs as an operational descriptor when appropriate; the term exosomes is reserved only for vesicles with demonstrated endosomal biogenesis. The analytics and reporting practices in this section are generalizable across intra-articular EV programs and are not restricted to GRPEL1-sEVs, consistent with community standards for rigor and transparent reporting. Relative to MISEV2023, transparency can be improved by explicitly reporting negative markers, pre-analytical variables (such as cell passage and media conditioning), and particle-protein ratios to facilitate between-study comparisons[14]. Community platforms such as Transparent Reporting and Centralizing Knowledge in Extracellular Vesicle research (EV-TRACK) further improve interpretability through standardized method disclosure and checklist-based reporting[19]. For dose normalization, the addition of a single molecule array (Simoa)-based protein quantification enables a like-for-like comparison of yields and purity across isolation workflows, whereas single-vesicle flow cytometry refines potency metrics by resolving EV heterogeneity and cargo partitioning[20,21]. A minimum reporting set for intra-articular EV studies is provided in Table 1, which operationalizes identity/purity panels, dual-metric dose normalization, shared exposure/retention time points, mechanism-linked potency, and transparency items consistent with MISEV2023 and EV-TRACK. Comparability is supported by multiparametric single-vesicle flow cytometry and single-molecule immunoassays, enabling orthogonal confirmation of identity and potency[17,18]. Separation/characterization and dose reporting follow MISEV2023; key metadata are registered to enhance reproducibility, such as EV-TRACK IDs (from EV-TRACK) for methods and materials[14,21].

Table 1 Minimum reporting set for intra-articular extracellular vesicle studies (bench and early clinical).

Domain	Required elements (minimum)	How to report (units/assays)	Harmonized time-points	Quality/acceptance notes
EV identity & purity	Particle size/number; protein; ≥ 1 positive and ≥ 1 negative protein marker	NTA (particles/mL); BCA or Simoa (μg/mL); western/flow for tetraspanins; negative marker (e.g., calnexin)	N/A	Follow MISEV2023 identity/purity panels; disclose isolation workflow and co-isolates
Dose normalization	Report both particle number and protein mass	Per-joint dose: Total particles and μg; lot-to-lot CVs	N/A	Dual metrics enable cross-study comparison and QC trending
Exposure/retention (biodistribution)	Joint residence and/or tissue association	Fluorescent or radiolabel tracking; cargo-based tracers (qPCR/protein)	1 hour, 24 hours, 7 days (shared anchors)	Predefine quench/correction; report synovial fluid/tissue ROIs and decay characteristics
Potency/MoA-linked readout	Mechanism-aligned assay + viability/toxicity	Cell-based PD (e.g., PINK1/PRKN index), live/dead, cytokines	N/A	Specify acceptance ranges; link potency to dose and exposure
Transparency & rigor	Protocol registration; method completeness	EV-TRACK/EV-METRIC ID; checklist link	N/A	Improves reproducibility and peer auditability
Safety	Local/systemic AEs; labs	Joint flare, effusions, vitals; CRP; imaging if indicated	Visit-aligned (e.g., baseline, week 2, week 12)	IND oversight where applicable; note no FDA-approved EV products

EV: Extracellular vesicle; NTA: Nanoparticle tracking analysis; BCA: Bicinchoninic acid assay; N/A: Not applicable; MISEV2023: Minimal Information for Studies of Extracellular Vesicles (2023 update); CV: Coefficient(s) of variation; sEV(s): Small extracellular vesicle(s); QC: Quality control; qPCR: Quantitative polymerase chain reaction; ROI: Region of interest; MoA: Mechanism of action; PD: Pharmacodynamics; PINK1: PTEN-induced kinase 1; PRKN: Parkin E3 ubiquitin ligase; CMC: Chemistry, manufacturing, and controls; Simoa: Single molecule array; PK: Pharmacokinetics; EV-TRACK: Transparent Reporting and Centralizing Knowledge in Extracellular Vesicle Research; EV-METRIC: Transparent Reporting and Centralizing Knowledge in Extracellular Vesicle Research Reporting Completeness Score; AEs: Adverse events; CRP: C-reactive protein; IND: Investigational New Drug; FDA: Food and Drug Administration.

Cargo engineering and mechanism attribution

Establishing protein (not RNA) delivery is a significant advancement that builds on EXPLOR and lysosomal-associated membrane protein 2A/KFERQ sorting as orthogonal protein-loading pathways[15,16]. At the same time, EV-delivered non-coding RNAs are increasingly recognized as regulators of recipient-cell programs relevant to OA, including mitophagy and immunomodulation[22]. Mechanistically, the mtHsp70-GRPEL1 structure presents the independent plausibility that augmenting GRPEL1 can stabilize mitochondrial proteostasis upstream of mitophagy, consistent with PINK1-dependency data of the study[10,12]. Coupling biochemical interaction evidence with pathway readouts and loss-of-function perturbations is appropriate for causality in EV-based mechanistic studies. When feasible, mechanism-parsing controls should include RNase-treated vs untreated EVs, protease or protein-depletion controls, and, where appropriate, antagonism of candidate small RNAs in recipient cells to distinguish protein- from RNA-driven effects[14]. Accordingly, programs should predefine fit-for-purpose RNA analytics (for example, small-RNA profiling and quantitative assays for candidate microRNA/Long noncoding RNA cargo) alongside protein assays to contextualize mechanism attribution[23].

In vivo readouts, PK/retention, and CMC

Using OARSI-compliant endpoints supports external validity and is consistent with pre-clinical standards for structural modification[13]. The subsequent steps should quantify biodistribution and joint retention, leveraging formulation strategies to improve intra-articular bioavailability (for instance, charge reversal) and, where suitable, cartilage-targeting designs to establish sustained intra-cartilage exposure[17,24]. On the manufacturing side, recent reports on clinical-grade UC-MSC sEVs provide a template for GRPEL1-EV CMC packages and lot-to-lot reproducibility by describing GMP-aligned processes, quality attributes, and stability programs[18].

Key gaps and risk of bias

To transform a strong mechanistic narrative into a trial-ready package, we recommend the following: (1) Potency assays related to mitophagy (standardized reporters) and reported alongside OARSI structure scores, consistent with MISEV2023[14]; (2) Explicit dose-exposure-response curves using exact particle counts and cargo copy estimates (Simoa and single-vesicle analytics) to define minimal effective and plateau doses[20,21]. General methodology and reporting details for particle/protein/cargo quantification and single-vesicle analytics are consolidated in “EV identity and dose reporting”; (3) Safety pharmacology (immunogenicity and off-target biodistribution) and durability studies under good laboratory practice (GLP)-like conditions; and (4) Early human-factors and administration work (syringeability and stability in synovial fluid) embedded within GMP-produced lots and predefined critical quality attributes[18,25]. Overall, there is robust evidence at mechanistic and pre-clinical efficacy levels, with distinct methodological levers to mitigate the risk of early-phase translation[12,14].

To complement the strengths, we explicitly note program risks and pragmatic mitigations: (1) Off-target biology of GRPEL1: As a mitochondrial nucleotide-exchange factor for mtHSP70, extra-cartilaginous uptake could perturb protein import and proteostasis; mitigation includes intra-articular dosing, joint-retentive or affinity designs, and prespecified exposure and response assessments in non-target tissues; (2) sEV heterogeneity and identity: Process and batch variability as well as co-isolates can confound readouts; adopt transparent analytics and single-particle flow cytometry where appropriate[26]; (3) Cargo loading consistency and potency: Variable loading undermines comparability; define Critical Quality Attributes and release specifications with a mechanism-linked potency assay, drawing on recent biopotency frameworks for EV therapeutics[27]; (4) Stability and logistics: Cold-chain dependence constrains access; implement a structured stability program and consider lyophilization with trehalose with predefined reconstitution criteria[28,29]; and (5) Regulatory and claims risk: EV products face heterogeneous global pathways and no current market authorizations; position development under appropriate oversight and avoid promotional disease-modification claims until supported by accepted endpoints[30].

TRANSLATIONAL ROADMAP: MANUFACTURING, NON-CLINICAL PACKAGE, AND EARLY-PHASE TRIAL DESIGN

The next steps focus on GRPEL1-sEVs specifically, illustrating how the general standards above are operationalized for this program across CMC, non-clinical safety, and early-phase trial design.

From signal to first-in-human

A credible path for GRPEL1-loaded SMSC sEVs based on three pillars: (1) A CMC package verifying identity, purity, potency, and lot-to-lot consistency; (2) A non-clinical program quantifying intra-articular biodistribution/retention and establishing local/systemic safety; and (3) An early-phase design demonstrating target engagement and identifying disease-modifying trajectories anchored to MISEV2023 and supported by EV-TRACK-style transparent reporting[14,19]. Generalizable analytics and dose-reporting standards are detailed in “EV identity and dose reporting”; the roadmap below applies that framework to GRPEL1-sEV-specific Critical Quality Attributes, release, and comparability. The bench-to-bedside framework adopted in this editorial is summarized in Figure 1, which aligns CMC/characterization, dose-exposure reporting, early readouts mapped to the OMERACT-OARSI core domain set, and transparency/oversight elements, thereby orienting the roadmap that follows. Early-phase studies will pre-register symptom endpoints (e.g., Western Ontario and McMaster Universities Osteoarthritis Index/Knee injury and Osteoarthritis Outcome Score) and at least one structural biomarker [magnetic resonance imaging (MRI) cartilage thickness/composition], with all details consolidated in Table 2.

Figure 1 Bench-to-bedside framework for intra-articular extracellular vesicles. A: Chemistry, manufacturing, and controls & characterization: Isolation workflow → identity/purity panels (tetraspanins/negative markers) → batch release (particles, μg) per Minimal Information for Studies of Extracellular Vesicles 2023; B: Dose & exposure: Normalized dose (particles and protein) and harmonized exposure/retention time-points (e.g., 1 hour, 24 hours, 7 days); C: Early clinical readouts: Endpoints mapped to the Outcome Measures in Rheumatology-Osteoarthritis Research Society International core domain set, with magnetic resonance imaging cartilage thickness as an exploratory structural biomarker informed by the Foundation for the National Institutes of Health Osteoarthritis Biomarkers Consortium; D: Transparency & oversight: Transparent Reporting and Centralizing Knowledge in Extracellular Vesicle Research/EV-METRIC badge and regulatory caution that no exosome products are United States Food and Drug Administration-approved; safety monitoring loop aligned with Investigational New Drug processes. CMC: Chemistry, manufacturing, and controls; NTA: Nanoparticle tracking analysis; MISEV2023: Minimal Information for Studies of Extracellular Vesicles (2023 update); ROI: Region of interest; OMERACT: Outcome Measures in Rheumatology; OARSI: Osteoarthritis Research Society International; WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index; KOOS: Knee injury and Osteoarthritis Outcome Score; MRI: Magnetic resonance imaging; EV-TRACK: Transparent Reporting and Centralizing Knowledge in Extracellular Vesicle Research; EV-METRIC: Transparent Reporting and Centralizing Knowledge in Extracellular Vesicle Research Reporting Completeness Score; AEs: Adverse events; IND: Investigational New Drug; FDA: Food and Drug Administration.

Table 2 Phase II signal-finding endpoint framework mapped to Outcome Measures in Rheumatology-Osteoarthritis Research Society International.

OMERACT-OARSI domain	Endpoint (suggested measure)	Time-points	Role in trial	Rationale
Pain	WOMAC pain (or KOOS pain)	Baseline, week 12, week 24	Primary	Core domain; sensitive to change
Function	KOOS-ADL (or WOMAC function)	Baseline, week 12, week 24	Key secondary	Functional relevance; limits multiplicity
Patient global	Patient global assessment	Baseline, week 12, week 24	Secondary	Complements pain/function per core set
Structural progression	MRI cartilage thickness (pre-specified compartment/ROI)	Baseline, 12-24 months (exploratory in phase II)	Exploratory (or key secondary if powered)	Supported by FNIH analyses; pre-register analysis
Safety	Local/systemic AEs; labs	Visit-aligned	Required	IND-appropriate safety set

OMERACT: Outcome Measures in Rheumatology; OARSI: Osteoarthritis Research Society International; WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index; KOOS: Knee injury and Osteoarthritis Outcome Score; ADL: Activities of daily living; MRI: Magnetic resonance imaging; ROI: Region of interest; FNIH: Foundation for the National Institutes of Health; AEs: Adverse events; IND: Investigational New Drug.

CMC and analytics

Release studies should combine orthogonal identity/impurity readouts (positive and negative EV markers and particle-to-protein ratios) with a mechanism-linked potency assay such as a PINK1-dependent mitophagy reporter in chondrocytes[14]. Quantitative dose normalization should use multiple units; single-molecule immunoassays (Simoa) enable like-for-like comparisons of EV protein cargo across isolation and scale-up, whereas single-vesicle flow cytometry resolves heterogeneity and cargo partitioning to inform potency specifications[20,21]. These analytics support a quality-by-design approach and define critical quality attributes for clinical lots[25]. Developing clinical-grade UC-MSC sEV demonstrated that the identity, potency, and stability criteria can be satisfied using current manufacturing controls. Potency was assessed using mitophagy-responsive chondrocyte readouts for mechanism-linked products such as GRPEL1-sEVs[20,31]. The intra-articular dose should be normalized across particles/proteins and accompanied by exposure/retention reporting at shared time points to facilitate cross-study comparisons[7].

Non-clinical biodistribution, durability, and safety

Before human dosing, quantify residence and exposure within joint compartments in rodents (and, when feasible, in larger species) using labeled sEVs and tissue PK[14]. Charge-reversal formulations provide a practical starting point for prolonging joint bioavailability and increasing cartilage uptake in OA models[17]. Local tolerability, off-target biodistribution, and repeat-dose safety should be characterized under GLP-like conditions, in accordance with recent quality and safety recommendations for EV therapeutics[25]. Clinical-grade UC-MSC sEV programs demonstrate the feasibility of GMP-aligned production with defined release criteria and stability plans for OA[18].

Regulatory and endpoint considerations

The Food and Drug Administration (FDA) draft guidance for “disease modification” prioritizes structural endpoints in addition to symptom alleviation, and differentiates structural progression from pain/function outcomes[32]. Therefore, early trials develop a dual strategy, including validated patient-reported outcomes for symptoms and imaging-based measures of the structure, guided by the OMERACT-OARSI core domain set (pain, function, patient global, and structural damage)[33].

Imaging strategy

Plain radiography is insensitive to early matrix changes; program-level efforts from the Foundation for the National Institutes of Health (FNIH) Biomarkers Consortium demonstrate that MRI-based structural and biochemical markers can outperform joint-space measures for detecting progression and for prognostic enrichment[34]. Imaging and structural endpoints[35,36] are consolidated in Table 2; please refer there for the unified symptom-structural plan.

Operationalizing the first trial

Starting with conservative intra-articular schedules informed by pre-clinical exposure-response and joint retention, with predefined safety gates for escalation[17]. Pre-specify go/no-go criteria that integrate safety, in vivo target engagement (mitophagy-linked biomarkers or mitochondrial quality-control signatures), and stabilization or improvement on MRI cartilage metrics over a plausible structural window[25].

COMPARATIVE LANDSCAPE, DIFFERENTIATION, AND SUBSEQUENT QUESTIONS

Positioning within the OA EV field

Engineered sEVs are transitioning from concept to clinical use. Numerous intra-articular trials have been registered, including programs using allogeneic UC-MSC sEVs for safety and feasibility assessments, as well as early efficacy evaluations, in knee OA[37,38]. Additional recruiting efforts broaden geography and operational experience, underscoring momentum toward translational proof in diverse clinical settings[39]. Parallel meta-research highlights uneven trial reporting and heterogeneous methods, emphasizing the need for rigorous CMC and standardized analytics as programs expand[40]. Complementary perspectives survey the clinical translation of EV therapeutics and converge on the same bottlenecks: Comparability, dose normalization, release potency, and regulatory navigation[41].

Beyond chondrocyte-intrinsic rescue, sEVs can modulate joint-resident and recruited immune cells that shape cartilage outcomes. MSC EVs dampen inflammation and promote M2-like macrophage polarization, a process repeatedly linked to tissue repair and relevant to OA pathophysiology[42,43]. Inflammatory priming of parent MSCs with cytokines such as tumor necrosis factor-α and interferon-γ can amplify immunoregulatory programs, in part via EV-mediated communication[44]. Conversely, synovial fibroblast EVs generated in an interleukin-1β-rich microenvironment can drive chondrocyte catabolism, underscoring the need to control disease context and source biology[45]. Mechanistically, EVs also traffic non-coding RNAs that reprogram recipient cells; microRNA and long noncoding RNA cargo have been implicated in anti-inflammatory, metabolic, and pro-repair pathways in OA-relevant settings[22,46]. Taken together, an integrated view that couples mitochondrial quality control with immune re-education and RNA-mediated signaling provides a broader mechanistic canvas for GRPEL1-EV development.

Differentiation of GRPEL1-sEVs

Most OA sEV strategies emphasize RNA payloads, anti-inflammatory proteins, or generic trophic signaling. A protein-cargo, mitochondria-centric approach that restores chondrocyte quality control represents a mechanistically distinct path with robust biological plausibility[47]. An in-press study of PINK1 dependency and structural anchoring to mtHsp70-GRPEL1 indicated a coherent, upstream lever compared with catabolic cytokine targeting or matrix anabolism alone. Furthermore, this study provides a clear mechanism-linked potency readout for lot release and pharmacodynamics[10,12].

Formulation and targeting vs the state-of-the-art

Charge reversal and cartilage-targeting sEVs have independently improved joint residence and sustained payload delivery in pre-clinical OA models, indicating that cartilage-retentive designs are becoming more practical[17,22]. In addition to passive retention, charge-reversed, cartilage-addressed sEVs for gene delivery provide a practical template for improving chondrocyte access without altering the intra-articular route and are directly portable to GRPEL1 cargoes[48].

Clinical readouts and biomarker strategy

Although early trials in the EV space frequently emphasize pain/function, structural and mechanistic readouts are necessary to claim disease modification[32]. Candidate pharmacodynamic markers for mitochondria-focused therapy include mitophagy pathway signatures derived from PINK1/Parkin biology and oxidative stress recovery in the joint tissues and synovial fluid. Current reviews outline viable analytes and sampling strategies suitable for phase I/II[49]. To triangulate target engagement and structure, they can be integrated with the MRI framework recommended for the early detection of cartilage changes in the same subjects[35,36]. Early trials should prespecify the structural and symptom domains in accordance with the OMERACT-OARSI core set. A structural anchor is provided by MRI cartilage thickness/volume, with T2/T1ρ compositional mapping as an optional pharmacodynamic outcome[50-52].

Risk and comparators

Safety considerations unique to EV products, including biodistribution, immunogenicity, and thrombogenic potential, remain manageable with GMP controls and modern analytics; however, they require explicit study-level mitigation plans and transparent reporting to satisfy regulators and the broader EV community[25,40]. Comparator selection should reflect standard nonoperative care while avoiding confounding co-interventions. In certain cases, co-formulations (for instance, with visco-supplementation) are being explored and may offer pragmatic synergy; however, they should only be pursued after monotherapy signals have been established[33,39]. Where clinically and ethically justified, co-formulations can also be evaluated in parallel or sequentially under prespecified rules (for example, a 2 × 2 factorial design to estimate main and interaction effects, or a multi-arm multi-stage/adaptive platform to add or drop arms efficiently), while preserving effect attribution relative to monotherapy. Such designs can reduce time and resource use without sacrificing interpretability, provided interactions are examined when biologically plausible and decision boundaries are defined a priori[53,54].

Actionable next experiments

The following experiments are actionable: (1) Head-to-head evaluation of retention-optimized GRPEL1-sEVs vs unmodified sEVs on joint PK and histologic rescue; (2) Definition of a dose-exposure-response surface using particle, protein, and cargo-copy units; (3) A human pharmacodynamic pilot that combine synovial fluid biomarker sampling with quantitative/MRI compositional endpoints; and (4) A comparability protocol to lock CMC as scale-up progresses from phase I to multi-site phase II[20,21]. Collectively, these steps clarify the differentiation from other EV approaches and safe progression to confirmatory testing[40,41].

CONCLUSIONS AND EDITORIAL RECOMMENDATIONS

Editorial stance

Integrating in-press evidence (protein cargo delivery, PINK1 dependence, and histological benefit) with contemporary OA biology supports a mitochondria-centric, protein-cargo EV strategy, distinct from RNA-centric or purely anti-inflammatory approaches[10,12]. Current mitophagy reviews on OA further underscore the translational value of targeting PINK1/Parkin E3 ubiquitin ligase-linked quality control in chondrocytes[49].

Reporting and outcome standards

From phase I/II onward, the Consolidated Standards of Reporting Trials (CONSORT) 2025 checklist and pre-specify outcomes are adopted in accordance with the CONSORT-Outcomes 2022 extension (estimands, multiplicity, and missingness). This will significantly improve the interpretability and comparability of EV trials[55,56]. Final endpoint specifications and their roles are summarized in Table 2.

EV analytics standards

In addition to MISEV2023, programs using EV flow cytometry should align with MIFlowCyt-EV to standardize acquisition, gating, and reporting, which are critical for potency, lot release, and cross-site reproducibility[14,57]. These practices complement the single-molecule and single-vesicle assays previously recommended for dose normalization and comparability[20,21].

Imaging and pharmacodynamics

To establish pharmacodynamic readouts of matrix rescue, quantitative MRI of cartilage thickness/volume is combined with a compositional metric (for instance, T2 or T1ρ) in early trials while aligning the protocols with FNIH-aligned initiatives[34-36]. The clinical role of cartilage T2/T1ρ mapping is summarized in contemporary guidance, which also confirms the operationalization of multi-contrast MRI in OA trials[35,58].

Manufacturing and regulatory navigation

Practical CMC roadmaps for clinical translation and commercialization of EV therapeutics highlight recurring bottlenecks, such as dose units, release potency, comparability, and change control, and provide checklists that should be finalized before phase II[59]. In parallel, a new clinical application guidance (Japan Societies for Regenerative Medicine/EVs) provides implementable safety, consent, and quality considerations for early trials[60]. These complement the broader quality/safety frameworks already cited for EV therapeutics[25].

Bottom line

GRPEL1-loaded SMSC sEVs have a defensible mechanistic anchor and a feasible translational path if the program complies with a minimum viable translational dataset (potency, dose normalization, joint PK/retention, GLP-style safety, and pre-specified dual endpoints) under modern reporting and EV-specific standards[14,19,55]. The approach is well-positioned to establish mitochondrial quality control as a tractable axis for disease modification in OA using these controls and imaging/pharmacodynamic alignment, as previously described[32,35,58].

FIELD-WIDE AGENDA FOR THE DISCIPLINE

A forward-looking agenda for EV therapeutics in OA should prioritize shared measures and regulatory convergence over fixed timelines. First, intra-articular PK, retention, and biodistribution should be quantified using commonly validated toolkits and reporting to ensure that datasets are comparable across sponsors. Foundational work has demonstrated that EV fate varies based on the cell source, route, and targeting, and practical in vivo/ex vivo fluorescence pipelines suitable for OA models are already available[31,61,62]. Second, dose and potency should be harmonized by adopting multi-metric dose units (particles, total protein, and cargo copies per vesicle) anchored to a mechanism-linked mitophagy assay for lot release and comparability, thereby aligning EV programs with the FDA potency guidance for Cellular and Gene Therapy products and the International Council for Harmonisation Q8 (Pharmaceutical Development)/Q9 (Quality Risk Management)/Q10 (Pharmaceutical Quality System) control-strategy framework[63,64]. Third, regulatory alignment should map EV CMC (identity/impurities, potency, stability, and comparability) and non-clinical packages directly to the European Medicines Agency guidelines for investigational advanced therapy medicinal products, using it as a template for first-in-human studies and scale-up while maintaining the OA expectation of symptoms and structures for disease-modifying claims under the FDA framework[32,65]. Fourth, imaging and pharmacodynamics should combine quantitative MRI of cartilage thickness/volume with one compositional metric (for instance, T2 or T1ρ) under harmonized acquisition/segmentation standard operating procedures, leveraging FNIH/Osteoarthritis Initiative evidence that these markers are more sensitive than radiography in capturing progression and enriching trials[34,50,66]. Finally, formulation and targeting, such as charge-reversed or cartilage-addressed sEVs, are prospectively linked to exposure and histological rescue to ensure that retention gains are translated into dose-normalized efficacy, rather than surrogate improvements alone[17,24].

ETHICS, TRANSPARENCY, AND REPRODUCIBILITY COMMITMENTS

A sustainable path for therapeutic EVs in OA requires explicit commitments to open, reproducible, and participant-centered science that complements the technical and regulatory roadmaps outlined above. Datasets, codes, and analysis workflows should be planned and curated according to FAIR principles to ensure that they are findable, accessible, interoperable, and reusable by both humans and machines[67]. All in vivo studies should follow ARRIVE (version 2.0) (including randomization, blinding, sample-size justification, and bias mitigation) to enhance the evidentiary value of pre-clinical work[51]. For human studies, journals and registries already require trial registration before the first enrollment and an explicit data-sharing statement within manuscripts and protocols. EV programs should adhere to or surpass these International Committee of Medical Journal Editors requirements and incorporate data-access plans into consent materials from the outset[52]. Early interventional studies should also be reported in accordance with the CONSORT extension for pilot and feasibility trials, which reduces selective reporting and clarifies decision criteria for progression to definitive trials[68]. As therapeutic EVs are situated at the interface of advanced biologics and complex analytics, alignment with International Society for Extracellular Vesicles position papers on clinical translation can help prevent common pitfalls in identity/impurity definition, mechanism-linked potency, and release testing, while maintaining consistent terminology and evidence standards across sponsors[14]. To ensure the reuse and secondary validation of EV cargo data, omics outputs (proteins, RNAs, and lipids) should be deposited in community resources, such as Vesiclepedia, with sufficient metadata to support cross-study comparisons and meta-analyses[21]. Finally, patient and stakeholder engagement should be treated as a methodological requirement, as it is responsible for co-defining outcomes, visit burden, and data-sharing preferences to improve feasibility, relevance, and trust, drawing on established frameworks for meaningful partnerships in clinical research[69].

CONCLUSION

SMSC-sEVs engineered to deliver GRPEL1 form a coherent, mitochondria-centric approach to OA that links a defined intracellular target to chondrocyte rescue and joint-level structural benefits. This study supports translational readiness if identity, impurity, and potency are secured to transparent standards, dose is normalized across orthogonal units, and joint exposure/retention is quantified with shared methods. Imaging strategies that integrate quantitative cartilage morphometry with compositional MRI provide practical pharmacodynamic anchors and validated symptom outcomes, while recent manufacturing precedents have demonstrated that clinical-grade vesicle products can meet comparability and stability expectations. GRPEL1-sEVs are well-positioned to test mitochondrial quality control as a tractable axis for disease modification in OA and to advance early clinical evaluation by paying careful attention to mechanism-linked potency, joint PK, and pre-specified structural-clinical endpoints.

ACKNOWLEDGEMENTS

We would like to thank all the professionals who contributed to the discussion and elaboration of this editorial.