Extracorporeal Shock Wave Therapy in Knee Osteoarthritis: A Mechanistic Narrative Review of Neurogenic Inflammation and the MMP/ADAMTS Axis

Abstract

Knee osteoarthritis (KOA) is a whole-joint disorder in which structural degeneration interacts with synovial inflammation, peripheral nociceptive sensitization, and the local neuroimmune microenvironment. Extracorporeal shock wave therapy (ESWT) improves pain and function in many clinical studies, but its disease-modifying potential remains uncertain. This narrative review integrates clinical and experimental evidence on two candidate mechanisms: suppression of neurogenic inflammation and regulation of the protease-inhibitor network formed by matrix metalloproteinases (MMPs), a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) proteases, and tissue inhibitors of metalloproteinases (TIMPs). The most consistent human evidence supports symptomatic benefit. Mechanistically, ESWT may reduce sensory nerve excitability, substance P and calcitonin gene-related peptide signaling, and neuroimmune feedback, thereby attenuating peripheral sensitization and synovial inflammation. Preclinical studies also report changes in inflammatory and oxidative-stress mediators, MMP expression, TIMP balance, and ADAMTS-related aggrecan degradation. However, evidence for structural cartilage regeneration is derived mainly from animal models, small biomarker studies, and heterogeneous protocols; long-term human imaging evidence is scarce. Shared interleukin-1β, tumor necrosis factor-α, nuclear factor-κB, and mitogen-activated protein kinase signaling provides a plausible interface between neuroimmune amplification and matrix catabolism. Overall, ESWT is best viewed as a parameter-dependent mechanobiological intervention with supported analgesic effects and plausible, but unconfirmed, matrix-modulatory activity. Standardized protocols, longitudinal biomarker and imaging studies, and pathway-directed experiments are needed before ESWT can be considered a disease-modifying therapy for KOA.

Introduction

Knee osteoarthritis (KOA) is a major cause of persistent pain, restricted mobility, and reduced quality of life in middle-aged and older adults. Its burden increases with age and is higher in women; in China, the overall prevalence is approximately 18%.^1,2 Advances in whole-joint biology, molecular profiling, and three-dimensional tissue and organ-on-a-chip models have broadened the traditional view of KOA from isolated cartilage wear to a disorder involving articular cartilage, synovium, subchondral bone, and periarticular tissues.^3–5 Neuroimmune studies further implicate sensory nerves and local immune responses in the persistence of pain and inflammation.^6,7 The incomplete correspondence between radiographic damage and pain severity further indicates that peripheral sensitization and the local neuroimmune microenvironment contribute to symptom persistence. Current management remains centered on pain relief and function. Pharmacological treatment, intra-articular injection, and surgery can be beneficial, but adverse effects, invasiveness, and limited structural disease-modifying activity constrain long-term use in some patients.^2,8–10 Non-invasive physical management therefore remains important and may include exercise therapy, therapeutic ultrasound, electrical stimulation, and short-wave diathermy, although these modalities differ in biological target, protocol, and evidential support.^8,11–13

Extracorporeal shock wave therapy (ESWT) delivers acoustic pressure waves to biological tissues and has expanded from lithotripsy to a range of musculoskeletal indications.¹⁴ Randomized trials, meta-analyses, and a recent umbrella review indicate that ESWT can reduce pain and improve knee function in KOA, with generally mild and transient adverse events.^15–19 Dose-comparison data also suggest that clinical response is parameter dependent rather than uniform across energy settings.²⁰ Despite this symptomatic evidence, the biological pathways through which ESWT acts in KOA remain incompletely defined. Its mechanical energy delivery may influence both nociceptive neuroimmune signaling and the inflammatory-cartilage microenvironment. This perspective is important because pain relief may arise relatively rapidly from changes in sensory nerve excitability and neuropeptide signaling, whereas matrix-metabolic effects would require slower changes in chondrocyte homeostasis, protease activity, and tissue structure.

This narrative review therefore evaluates two mechanistic dimensions: suppression of neurogenic inflammation and regulation of the protease-inhibitor network formed by matrix metalloproteinases (MMPs), a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) proteases, and tissue inhibitors of metalloproteinases (TIMPs). ESWT studies remain heterogeneous in device type, energy level, treatment target, pulse number, session schedule, and follow-up duration. The review consequently does not define an optimal protocol or assume that symptomatic benefit establishes cartilage regeneration. Instead, it uses a dual-pathway framework to distinguish supported clinical effects from preclinical matrix-modulatory signals and unconfirmed disease-modifying hypotheses.

Literature Search and Evidence Synthesis

This mechanism-oriented narrative review integrated clinical efficacy evidence with preclinical and mechanistic findings on ESWT in KOA. Searches were conducted in PubMed, Web of Science, Embase, and the China National Knowledge Infrastructure (CNKI) and were updated through May 2026. Search concepts included knee osteoarthritis, extracorporeal shock wave therapy, neurogenic inflammation, substance P, calcitonin gene-related peptide, matrix metalloproteinases, ADAMTS, cartilage degradation, synovial inflammation, nuclear factor-κB, and mitogen-activated protein kinase. Reference lists of relevant clinical studies, experimental studies, and reviews were also screened.

Evidence was selected according to topic relevance, directness, and mechanistic interpretability. Randomized trials, cohort studies, meta-analyses, and umbrella reviews were used to assess pain, function, safety, and human biochemical outcomes. KOA animal models and joint-related tissues, cells, or biological fluids were used to examine inflammation, matrix degradation, oxidative stress, and sensory nerve-related mechanisms. Other osteoarthritis models, non-KOA pain models, bone-repair models, and general cartilage biology were used only for mechanistic comparison or hypothesis generation. Findings from ESWT combined with electroacupuncture, low-intensity pulsed ultrasound, or other interventions were treated as auxiliary evidence and were not attributed to ESWT alone.

Because study populations, models, devices, treatment targets, parameters, and outcomes were heterogeneous, no Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)-based selection process, formal risk-of-bias assessment, or quantitative synthesis was performed. This review does not replace systematic reviews of clinical efficacy, safety, or optimal parameters. Mechanisms not directly verified in patients with KOA or KOA-specific models are described as indirect or hypothesis-generating.

Pathological Basis Relevant to ESWT Action in KOA

KOA is a whole-joint disorder in which cartilage matrix loss, synovial inflammation, subchondral bone remodeling, and pain sensitization interact.^4,21 For interpretation of ESWT, two processes are particularly relevant. MMP/ADAMTS-driven extracellular matrix (ECM) catabolism contributes to structural degeneration, whereas neurogenic inflammation links sensory nerve activation to synovial immune responses and persistent pain. Inflammatory cytokines, oxidative stress, mechanical loading, and matrix-derived damage signals can connect these processes, providing the pathological basis for the dual-pathway framework used below.

Cartilage Matrix Degradation and Abnormal Activation of the MMP/ADAMTS Axis

Articular cartilage integrity depends on chondrocyte maintenance of type II collagen and aggrecan. In KOA, abnormal loading, inflammatory cytokines, oxidative stress, and matrix damage shift chondrocytes toward catabolic and hypertrophy-like states, with reduced SOX9, type II collagen, and aggrecan and increased Runx2, type X collagen, and MMP-13.^22–26 Cartilage loss therefore reflects dysregulated cell phenotype and protease activation rather than mechanical abrasion alone.

MMPs and ADAMTS proteases are major effectors of ECM degradation. MMP-1, MMP-3, and MMP-13 degrade collagen and other matrix components, with MMP-13 serving as a key collagenolytic enzyme in osteoarthritic cartilage.^27–32 ADAMTS-4 and ADAMTS-5 preferentially cleave aggrecan, compromising water retention and load-bearing capacity before extensive collagen-network failure.^33–37 Experimental data support an important role for ADAMTS-5, although the relative contribution of ADAMTS-4 and ADAMTS-5 varies by species and inflammatory context.^38–40

TIMPs restrain the proteolytic network: TIMP-1 inhibits several MMPs, whereas TIMP-3 is a potent inhibitor of ADAMTS-4/5.^41–43 Protease expression alone does not determine matrix loss.⁴⁴ Interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) can activate nuclear factor-κB (NF-κB) and mitogen-activated protein kinase (MAPK) signaling, increase MMP and ADAMTS expression, and suppress anabolic activity.^45–48 The MMP/ADAMTS-TIMP network therefore integrates inflammatory signaling, mechanical stress, and chondrocyte phenotype and is more informative than any isolated protease measurement.

Neurogenic Inflammation and Neuroimmune Amplification

Pain severity in KOA is only partly explained by radiographic damage. The synovium, subchondral bone, meniscus, and joint capsule are innervated by nociceptive fibers whose cell bodies lie in the dorsal root ganglia (DRG). Mechanical stress, cytokines, and damage signals can activate these terminals and induce release of substance P (SP) and calcitonin gene-related peptide (CGRP), which influence vascular responses, immune-cell recruitment, and peripheral sensitization.^6,7,49 SP/neurokinin-1 receptor (NK-1R) signaling contributes to osteoarthritis-related chronic pain, while SP can activate mast cells, macrophages, and synovial cells through NF-κB, extracellular signal-regulated kinase 1/2 (ERK1/2), and p38 MAPK pathways.^50–53 CGRP has context-dependent roles in nociception, vascular regulation, and tissue remodeling, and altered DRG or synovial CGRP/receptor activity-modifying protein 1 (RAMP1) signaling has been reported in inflammatory knee models and human KOA.^54–58

These neural and inflammatory processes form a reciprocal circuit. SP, CGRP, and related signals can sustain activation of mast cells, macrophages, and synovial cells, increasing IL-1β, TNF-α, and nitric oxide (NO). These mediators in turn lower nociceptor thresholds and maintain peripheral sensitization.^6,7,49–53 The same inflammatory transcriptional programs can promote MMP and ADAMTS expression, while cartilage fragments provide additional damage signals. Neurogenic inflammation can therefore contribute to pain persistence and may also reinforce a matrix-catabolic environment, although direct human evidence linking this circuit to structural progression remains limited (Figure 1).

Figure 1.

Pathological basis relevant to ESWT action in knee osteoarthritis.

Notes: This schematic illustrates the proposed interaction between neurogenic inflammation and MMP/ADAMTS-mediated matrix catabolism in KOA. The left pathway summarizes nociceptive nerve activation, SP/CGRP release, immune and synovial cell activation, and peripheral sensitization. The right pathway summarizes chondrocyte phenotypic shift, ECM homeostasis imbalance, upregulation of matrix-degrading enzymes, and cartilage matrix degradation. Shared IL-1β, TNF-α, NF-κB, and MAPK signaling may link neuroimmune activation to matrix catabolism. Solid arrows indicate the principal pathological sequence and cross-talk, whereas dashed arrows indicate reciprocal reinforcement from the integrated pathological consequences to upstream neuroimmune and matrix-catabolic processes.

Abbreviations: ADAMTS, a disintegrin and metalloproteinase with thrombospondin motifs; CGRP, calcitonin gene-related peptide; ECM, extracellular matrix; ESWT, extracorporeal shock wave therapy; IL-1β, interleukin-1β; KOA, knee osteoarthritis; MAPK, mitogen-activated protein kinase; MMP, matrix metalloproteinase; NF-κB, nuclear factor-κB; SP, substance P; TIMP, tissue inhibitor of metalloproteinases; TNF-α, tumor necrosis factor-α.

Physical Characteristics and Biological Effects of ESWT

ESWT consists of brief acoustic pressure pulses with rapid pressure rise and high peak pressure. According to generation and energy distribution, it is commonly divided into focused shock wave (FSW) and radial pressure wave (RPW) modalities. FSW can concentrate energy at a selected depth, whereas RPW delivers maximal pressure near the surface with progressive attenuation.^59–62 These modalities are not biologically interchangeable. Device type, energy flux density or pressure setting, treatment depth, and target tissue all influence the delivered mechanical load and may contribute to inconsistent clinical and mechanistic findings.^20,60

Mechanotransduction and cavitation convert acoustic energy into cellular and microenvironmental responses. Mechanical coupling among the cell membrane, cytoskeleton, mechanosensitive structures, and ECM can induce calcium-dependent signaling relevant to inflammation, repair, and matrix metabolism.^59,62 Experimental studies also report context-dependent effects on angiogenesis, inflammatory signaling, neuronal function, and tissue repair.^63–67 KOA studies have observed changes in pain, inflammatory and oxidative-stress markers, and matrix-catabolic indicators after ESWT or ESWT-based treatment.^68–70 Because some reports used combined interventions, they support biological plausibility but cannot establish ESWT-specific effects. ESWT is therefore better understood as a parameter-dependent mechanobiological intervention than as a fixed single-target therapy.

Clinical studies generally report a favorable safety profile, with adverse events usually limited to transient local discomfort or mild soft-tissue reactions. Reproducible interpretation nevertheless requires consistent reporting of device type, focused or radial delivery, energy setting, pulse number, frequency, target site, treatment interval, session number, and follow-up duration. Without this information, symptomatic outcomes and mechanistic biomarkers cannot be compared reliably across protocols.

Potential Matrix-Metabolic Effects of ESWT Relevant to the MMP/ADAMTS Axis

Clinical Biochemical Markers and Evidence from KOA Animal Models

Human biochemical evidence remains limited. A small pre-experimental study reported reduced serum MMP-3 after ESWT in patients with KOA.⁷¹ This finding suggests a change in a circulating marker associated with cartilage catabolism but does not demonstrate local protease inhibition or structural cartilage protection. Studies of ESWT combined with low-intensity pulsed ultrasound, electroacupuncture, or electrothermal acupuncture have also reported changes in inflammatory, oxidative-stress, and matrix-metabolic outcomes.^68,70,72 Because these are composite interventions, they are supportive only at the level of hypothesis generation.

KOA animal models provide more direct tissue-level evidence. In rabbits with anterior cruciate ligament transection-induced KOA, ESWT reduced cartilage MMP-1 and MMP-3 messenger RNA and increased TIMP-1, together with lower IL-1β, TNF-α, and NO.⁷³ Other animal studies report improvements in synovial inflammation, cartilage injury, or matrix-related markers.^73,74 These findings support modulation of the local inflammatory and matrix-catabolic environment, but species, model construction, target site, and dose prevent direct translation to human structural benefit.

MMP/TIMP Balance and Inflammation-Driven Catabolism

The MMP/ADAMTS-TIMP network determines both the magnitude and pattern of ECM degradation. MMPs participate broadly in collagen and matrix breakdown, whereas ADAMTS-4/5 preferentially cleave aggrecan; TIMP-3 is particularly relevant because it inhibits both aggrecanases.^33–43 A modest change in one enzyme may have limited meaning if the wider protease-inhibitor balance remains unmeasured. ESWT-related biomarker changes should therefore be interpreted at the network level rather than as direct inhibition of a single protease.

A plausible model is that ESWT first modifies mechanical and inflammatory conditions, reducing IL-1β, TNF-α, oxidative stress, and NF-κB/MAPK activity, and thereby shifts the MMP/ADAMTS-TIMP balance toward lower catabolism.^45–48 This is an indirect microenvironmental mechanism. Demonstration of target-specific regulation would require simultaneous measurement of active proteases, TIMPs, cleavage products, and matrix synthesis, ideally with pathway blockade or rescue experiments.

ADAMTS-Related Aggrecan Degradation and Limits of Structural Evidence

Evidence for ADAMTS regulation by ESWT is less developed than evidence for MMP-related changes. In temporomandibular-joint osteoarthritis, ESWT reduced MMP-3 and MMP-13 and was associated with improved cartilage and subchondral bone structure.⁷⁵ In early knee osteoarthritis, subchondral bone-targeted ESWT reduced MMP-13 and ADAMTS-5 alongside attenuation of cartilage, synovial, and subchondral bone changes.⁷⁶ These studies support ADAMTS-related aggrecan degradation as a candidate mechanism, but human KOA studies rarely assess ADAMTS-4/5, TIMP-3, and aggrecan cleavage products together.

Structural relevance may also involve oxidative stress and chondrocyte survival. In rabbit KOA, ESWT reduced inducible nitric oxide synthase (iNOS), synovial-fluid NO, chondrocyte apoptosis, and cartilage degeneration.⁷⁷ Bone-repair and cartilage-biology studies provide indirect support for transforming growth factor-β1 (TGF-β1), vascular endothelial growth factor (VEGF), and Smad-related repair signaling.^78–80 Nevertheless, these observations do not establish cartilage regeneration in patients. Current evidence supports possible matrix-metabolic modulation; disease modification requires longitudinal human imaging, local biomarkers, and sustained structural outcomes.

Analgesic and Anti-Inflammatory Mechanisms of ESWT Through Suppression of Neurogenic Inflammation

Neuropeptides and Sensory Nerve Fibers

KOA-specific neural evidence is limited but compatible with an analgesic effect. In a rat KOA model, ESWT improved motor and pain-related outcomes.⁸¹ Other models provide indirect information: low-energy shock waves reduced inflammatory pain and hypersensitivity in experimental prostatitis, whereas limb studies in horses and sheep found only limited, transient cutaneous analgesia under particular protocols.^82,83 These differences reinforce the dependence of neural effects on tissue, species, target, and dose.

Musculoskeletal animal studies have reported reduced periosteal SP, fewer SP-immunoreactive DRG neurons, changes in unmyelinated C fibers, and altered peripheral nerve conduction or morphology after shock-wave exposure.^84–86 Such findings support the possibility that ESWT modifies small-diameter sensory input and neuropeptide release. They remain indirect for KOA and do not justify assuming structural nerve injury or denervation in treated patients.

Neuroimmune Feedback and the Inflammatory Microenvironment

Neurogenic inflammation is sustained by reciprocal signaling among sensory terminals, neuropeptides, and immune cells. SP and CGRP can facilitate mast-cell, macrophage, and synovial-cell responses, while IL-1β, TNF-α, and other mediators further increase sensory-neuron excitability and lower nociceptive thresholds.^6,7,49–53 By reducing sensory nerve excitability, neuropeptide release, or local inflammatory burden, ESWT could weaken this feedback circuit and thereby reduce pain and peripheral sensitization.

KOA animal studies report lower IL-1β, interleukin-6 (IL-6), and TNF-α in synovial and cartilage tissues after ESWT, together with reduced pain sensitization.⁸⁷ Clinical and combined-intervention studies also describe changes in NO, antioxidant activity, and matrix-related markers.^68,70 Reduced neuroimmune amplification could therefore influence both symptoms and the inflammatory signals that drive MMP/ADAMTS expression. A slowing of structural progression remains a hypothesis, however, because no human longitudinal study has shown that neuropeptide or immune changes mediate reduced cartilage loss.

Dual-Pathway Coupling, Symptomatic Evidence, and Disease-Modifying Uncertainty

Neurogenic inflammation and matrix catabolism have plausible points of convergence. In chondrocytes and synovial cells, NF-κB and MAPK signaling promote ADAMTS-5 and other catabolic mediators.^{45–48,88,89} In neuroimmune circuits, SP, IL-1β, and TNF-α activate related pathways, facilitate macrophage and synovial responses, and maintain peripheral sensitization.^6,7,49–53 ESWT-associated reductions in neural excitation and inflammatory tone could therefore diminish both nociceptive amplification and inflammation-driven protease expression. This coupling is biologically plausible but has not been demonstrated as a causal sequence within the same KOA cohort or model.

The evidence is substantially stronger for symptomatic improvement than for disease modification. Pain and function have been evaluated in randomized trials, meta-analyses, and an umbrella review, and can change over short follow-up periods.^15–19 By contrast, structural claims rely mainly on animal histology, isolated serum or synovial biomarkers, and heterogeneous treatment protocols. Serum MMP-3 reduction does not establish local cartilage regeneration, and long-term quantitative magnetic resonance imaging, cartilage-thickness, or radiographic-progression data remain scarce. Early benefit may predominantly reflect reduced neural and inflammatory sensitization, whereas MMP/ADAMTS-TIMP remodeling would be expected to develop over a longer interval; time-dependent chondroprotective effects have been reported in a rat KOA model.⁹⁰ This temporal sequence remains an inference rather than a proven treatment trajectory in patients.

The dual-pathway framework should therefore be read as an evidence-organizing model. Human KOA studies support clinical benefit and selected biochemical changes but rarely establish tissue-level causality. KOA animal studies provide mechanistic proximity, whereas other osteoarthritis and non-osteoarthritis models are useful mainly for analogy. Causal validation will require standardized ESWT protocols, pathway blockade or genetic manipulation, and synchronized assessment of pain sensitization, neuropeptides, inflammatory mediators, active MMP/ADAMTS enzymes, TIMPs, cleavage products, and structural imaging across multiple time points. The current evidence hierarchy and interpretive boundaries are summarized in Table 1, and representative clinical and experimental studies are detailed in Table 2.

Table 1.

Evidence Hierarchy and Interpretive Boundaries for ESWT-Related Neuroimmune and Matrix-Metabolic Effects in Knee Osteoarthritis

Evidence Domain	Key evidence and Mechanistic Implication	Interpretive Boundary
Clinical symptomatic efficacy	Randomized trials and meta-analyses consistently support reductions in pain and improvements in function after ESWT.15–20,69 These findings establish clinical relevance and show that treatment response is parameter dependent.	The evidence supports symptomatic benefit, but pain and functional improvement do not demonstrate cartilage regeneration or disease modification.
Human biochemical evidence in KOA	Serum MMP-3 and selected synovial inflammatory, oxidative, or matrix-metabolic markers may improve after ESWT or ESWT-based combined therapy.68,70,71 This suggests modulation of the joint catabolic microenvironment.	Small samples, circulating biomarkers, and combined interventions limit attribution to ESWT alone; biomarker change is not equivalent to structural protection.
KOA animal matrix-metabolic evidence	KOA models report lower MMP-1/3 or MMP-13/ADAMTS-5, higher TIMP-1, reduced IL-1β, TNF-α, and NO, and improved cartilage or subchondral-bone findings after ESWT.73–77 The results support regulation of inflammation, oxidative stress, and the MMP/ADAMTS–TIMP network.	The findings are preclinical and vary by species, model, target site, and dose; they cannot establish patient-level structural modification.
Neurogenic inflammation and pain modulation	KOA and non-KOA models indicate that ESWT can alter sensory nerve input and SP/CGRP-related signaling, with accompanying changes in pain behavior or DRG neuropeptide markers.81–87 This supports reduced neural excitability and neuroimmune amplification as plausible analgesic mechanisms.	Much of the neural evidence is non-KOA or animal based; structural nerve-fiber effects should not be assumed in patients receiving routine clinical protocols.
Structural modification and disease progression	Animal studies show cartilage, synovial, and subchondral-bone changes after ESWT, particularly when treatment is directed to the subchondral bone.75–77,90 These observations provide biological plausibility for matrix-modulatory effects.	Human studies rarely include longitudinal MRI, cartilage-thickness measures, cleavage products, or progression endpoints; disease-modifying efficacy remains unproven.
Shared signaling and temporal coupling	IL-1β, TNF-α, NF-κB, and MAPK may connect neuroimmune amplification with MMP/ADAMTS transcription. Analgesic effects may emerge earlier than matrix-related changes.45–48,53,88–90 This provides the rationale for the proposed dual-pathway framework.	The coupling and temporal sequence are cross-study inferences and require longitudinal, multi-endpoint, pathway-directed validation within the same model or cohort.

Notes: Evidence domains indicate source and directness rather than formal certainty grading. Interpretive boundaries identify limits of causal inference and distinguish symptomatic improvement from unconfirmed structural modification.

Abbreviations: ADAMTS, a disintegrin and metalloproteinase with thrombospondin motifs; CGRP, calcitonin gene-related peptide; DRG, dorsal root ganglion; ESWT, extracorporeal shock wave therapy; IL-1β, interleukin-1β; KOA, knee osteoarthritis; MAPK, mitogen-activated protein kinase; MMP, matrix metalloproteinase; NF-κB, nuclear factor-κB; NO, nitric oxide; SP, substance P; TIMP, tissue inhibitor of metalloproteinases; TNF-α, tumor necrosis factor-α.

Table 2.

Representative Clinical and Experimental Studies Informing the Neuroimmune and Matrix-Metabolic Effects of ESWT in Knee Osteoarthritis

Study/Model	ESWT Modality and Parameters	Target Tissue or Site	Main Outcomes	Mechanistic Domain	Evidential Limitation
Kim et al, 2015; human RCT, K–L grade II–III KOA (n=61)20	Focused electromagnetic ESWT; 1,000 shocks/session; EFD 0.040 vs 0.093 mJ/mm2; once weekly for 3 sessions	Tender medial tibial plateau	Both doses improved VAS and function through 12 weeks; medium energy produced greater improvement at selected time points	Parameter dependence and symptomatic efficacy	No sham group; modest sample and short follow-up; no biomarker or structural endpoint
Randita et al, 2024; human pre-experimental KOA (n=14)71	Piezoelectric ESWT once weekly for 6 weeks; EFD and pulse number not reported in the indexed report	Symptomatic knee; exact intra-articular target not specified	Serum MMP-3 and pain scores decreased during treatment	Human biochemical signal related to matrix catabolism	Uncontrolled, small sample; serum marker does not establish local cartilage protection
Wang et al, 2014; rabbit ACLT-induced KOA73	Focused ESWT after ACLT; detailed dose not available in the indexed English abstract	Articular cartilage and subchondral bone	Lower MMP-1/3 and IL-1β, TNF-α, and NO; higher TIMP-1; improved cartilage and subchondral-bone remodeling	MMP/TIMP balance, inflammation, and oxidative stress	Animal model; incomplete parameter reporting; no human or longitudinal imaging confirmation
Chou et al, 2019; rat early KOA induced by ACLT + MM76	Focused ESWT; 0.25 mJ/mm2; 800 impulses; target-site comparison	Medial tibial subchondral bone vs articular cartilage	Subchondral targeting improved pathology and micro-CT findings, increased TGF-β1, and reduced MMP-13 and ADAMTS-5	Target-dependent matrix and subchondral-bone modulation	Preclinical site-comparison study; cannot establish disease modification in patients
Zhao et al, 2012; rabbit ACLT-induced KOA77	Single ESWT application; 600 impulses at 1.5×105 Pa	Knee joint, synovial cavity, and articular cartilage	Reduced synovial NO, chondrocyte apoptosis, and cartilage-lesion severity at 4 and 8 weeks	Oxidative stress, chondrocyte survival, and structural injury	Single animal protocol; no patient biomarker, imaging, or functional validation
Ochiai et al, 2007; rat KOA81	Focused ESWT; 1,000 impulses/session; 3 weekly sessions; EFD not specified in the accessible report	Affected knee and associated sensory pathway	Improved motor dysfunction and pain behavior, with fewer CGRP-positive DRG neurons	Sensory nerve input and neurogenic analgesia	Preclinical behavioral and neural-marker evidence; translation to human KOA is uncertain
Hausdorf et al, 2008; normal rabbit distal-femur model84	High-energy focused ESWT; single application to the distal femur; exact dose not stated in the indexed abstract	Ventral distal femur and ipsilateral L5–L7 DRG	Reduced SP-immunoreactive neurons in the ipsilateral L5 DRG without loss of total neurons	Neuropeptide-related sensory modulation	Non-KOA model and high-energy exposure; analogical evidence only and not proof of structural nerve effects in patients

Note: This table presents representative rather than exhaustive studies and is intended to reduce narrative density. Where full treatment parameters were not available in the indexed or accessible report, this is stated explicitly. Clinical symptomatic outcomes, biochemical changes, and preclinical structural findings represent different levels of inference.

Abbreviations: ACLT, anterior cruciate ligament transection; ADAMTS, a disintegrin and metalloproteinase with thrombospondin motifs; CGRP, calcitonin gene-related peptide; DRG, dorsal root ganglion; EFD, energy flux density; ESWT, extracorporeal shock wave therapy; K–L, Kellgren–Lawrence; KOA, knee osteoarthritis; MM, medial meniscectomy; MMP, matrix metalloproteinase; NO, nitric oxide; RCT, randomized controlled trial; SP, substance P; TGF-β1, transforming growth factor-β1; TIMP, tissue inhibitor of metalloproteinases; TNF-α, tumor necrosis factor-α; VAS, visual analogue scale.

Conclusions and Perspectives

Current evidence supports ESWT as a generally safe, parameter-dependent intervention that can reduce pain and improve function in KOA. Modulation of sensory nerve excitability, SP/CGRP signaling, and neuroimmune feedback provides a plausible explanation for analgesia. Changes in inflammatory and oxidative-stress mediators and in the MMP/ADAMTS-TIMP network also suggest possible matrix-metabolic effects. These pathways may converge through IL-1β, TNF-α, NF-κB, and MAPK signaling, but the proposed coupling has not been established as a causal human mechanism.

The principal challenge is to distinguish symptomatic improvement from structural disease modification. Future studies should standardize FSW and RPW parameters, define treatment targets, and combine patient-reported and functional outcomes with synovial or blood biomarkers, quantitative imaging, and long-term follow-up. Multi-time-point studies that assess neural, inflammatory, matrix-metabolic, and structural outcomes in the same participants, supported by pathway-directed experiments in validated models, are needed to determine whether ESWT can alter KOA progression. Until such evidence is available, ESWT should be regarded as a clinically supported symptomatic therapy with plausible matrix-modulatory activity but unconfirmed disease-modifying potential.

Funding Statement

This work was supported by the Scientific Research Project of the Department of Education of Hunan Province (No. 25A0737) and the Special Project on the Popularization of Traditional Chinese Medicine Classics in Heilongjiang Province (No. ZYW2024-040).

Abbreviations

ACLT, anterior cruciate ligament transection; ADAMTS, a disintegrin and metalloproteinase with thrombospondin motifs; CGRP, calcitonin gene-related peptide; CNKI, China National Knowledge Infrastructure; DRG, dorsal root ganglia; ECM, extracellular matrix; EFD, energy flux density; ERK1/2, extracellular signal-regulated kinase 1/2; ESWT, extracorporeal shock wave therapy; FSW, focused shock wave; IL-1β, interleukin-1β; IL-6, interleukin-6; iNOS, inducible nitric oxide synthase; K–L, Kellgren–Lawrence; MAPK, mitogen-activated protein kinase; MM, medial meniscectomy; MMP, matrix metalloproteinase; NF-κB, nuclear factor-κB; NK-1R, neurokinin-1 receptor; NO, nitric oxide; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; RAMP1, receptor activity-modifying protein 1; RCT, randomized controlled trial; RPW, radial pressure wave; SP, substance P; TGF-β1, transforming growth factor-β1; TIMP, tissue inhibitor of metalloproteinases; TNF-α, tumor necrosis factor-α; VAS, visual analogue scale; VEGF, vascular endothelial growth factor.

Disclosure

The authors report no conflicts of interest in this work.