Recent Advances in Therapeutic Approaches for Knee Osteoarthritis: a Narrative Review

Abstract

Knee osteoarthritis (KOA) is a progressive and chronic musculoskeletal condition that continues to be the leading cause of disability worldwide. Conventional treatment approaches for the management of KOA largely focus on symptom alleviation rather than halting or reversing disease progression. However, recent advancements have highlighted the integrated interplay of mechanical stress, inflammation, cellular senescence, and chondrocyte dysfunction in the progression of KOA, in turn prompting new therapeutic strategies. Therefore, emerging interventions such as regenerative medicine, gene therapy, senolytic, platelet-rich plasma (PRP), disease-modifying osteoarthritis drugs (DMODs), and biologics have broadened the therapeutic options. Additionally, natural compounds demonstrated potential in KOA treatment with promising chondroprotective and anti-inflammatory effects. Moreover, digital technologies and clinical and molecular phenotyping enhanced early diagnosis, monitoring, and personalized management of the disease. Therefore, the current narrative review focuses on the molecular insights, clinical outcomes and prospects for the rapidly evolving landscape of current and emerging treatment approaches for the management of knee osteoarthritis (KOA).

INTRODUCTION

Osteoarthritis (OA) is a major global health concern, affecting approximately 528 million people worldwide, with knee osteoarthritis (KOA) accounting for 60-85% of all OA cases (Langworthy et al., 2024). KOA is a chronic, degenerative joint disease characterized by the gradual deterioration of synovial joints, leading to severe pain, stiffness, and loss of joint function, ultimately limiting mobility (Wu et al., 2024). The prevalence of KOA increases with age, affecting about 10% of men and 13% of women over 60 years old, and significantly diminishing quality of life (Primorac et al., 2020).

Even though the precise etiology and pathogenesis of KOA are still actively studied, its onset and progression are known to be affected by well-established and potential risk factors, including advanced age, previous joint injury, high bone mineral density, obesity, and genetic predispositions (He et al., 2020; Li et al., 2025a; Tudorachi et al., 2022).

It is now well acknowledged that knee osteoarthritis extends beyond the concept of simple, ‘wear-and-tear’, involving complex inflammatory, mechanical, and metabolic pathways leading to extensive joint deterioration (He et al., 2020). Uncontrolled signalling pathways, including NF-kB, Wnt/b-catenin, and PI3K/Akt/mTOR, facilitate chondrocyte death, degradation of extracellular matrix, and abnormal bone remodelling (Choi et al., 2019; Sun et al., 2020). Moreover, chronic inflammation involving pro-inflammatory cytokines IL-1β, TNF-α, and IL-6 causes tissue damage, while osteophyte development and subchondral bone sclerosis worsen joint function (Chen et al., 2017; Rezus et al., 2021).

KOA management is quite challenging due to the complex nature of the disease. The conventional approaches target alleviation of symptoms and improvement of joint function, utilizing non-pharmacological interventions like exercise and weight loss to pharmacological treatment and, in more severe cases, surgical joint replacement (Langworthy et al., 2024). However, novel therapeutics like senolytic agents, DMOADs, biologic agents, and natural compounds exhibited the potential of slowing down or reversing the structural damage by reducing inflammation, apoptosis, oxidative stress, and cartilage degradation (Ansari et al., 2024; Kim et al., 2025; Zhang et al., 2021)

Moreover, regenerative medicine, like mesenchymal stromal cells (MSCs), has been shown to maintain cartilage homeostasis, while inflammatory and degenerative pathways are regulated by gene therapy and CRISPR-Cas9 editing techniques (Cao et al., 2025; Jia et al., 2024; Li et al., 2025b). Concurrently, digital technologies and artificial intelligence-driven imaging are becoming the focus of the improvement of diagnostic processes, monitoring, and individualized rehabilitation programs (Feng et al., 2023).

Furthermore, emerging evidence highlights the heterogeneity of KOA, having distinct phenotypes associated with particular molecular biomarkers, imaging features, and clinical characteristics, leading KOA therapy towards precision medicine with personalized treatment approaches.

In light of this context, the current review aims to provide a comprehensive overview of emerging and personalized therapeutic options for knee osteoarthritis, focusing on the mechanistic insights along with clinical relevance, highlighting the shift in KOA management from traditional to mechanism-driven, disease-modifying and precision-based therapeutic strategies.

PATHOPHYSIOLOGY OF KNEE OSTEOARTHRITIS

KOA pathogenesis involves not only the cartilage matrix deterioration but also the synovium, ligaments, periarticular muscles, and subchondral bone remodelling (Giorgino et al., 2023; Rezus et al., 2021; Yunus et al., 2020).

Cartilage degradation, a pathological hallmark of KOA, arises from the disruption of the balance between anabolic repair mechanisms and catabolic processes (Kanakis et al., 2019). Elevated expression of matrix metalloproteinases (MMPs), ADAMTS, and aggrecanases facilitates extracellular matrix breakdown, while dysregulation of pro-inflammatory cytokines (e.g., IL-1β, TNF-α, IL-6) and signalling pathways (NF-κB, Wnt/β-catenin, and PI3K/Akt/mTOR) triggers inflammation and accelerates chondrocyte dysfunction (Choi et al., 2019; Sun et al., 2020). In addition, synovial inflammation, often associated with mechanical stress/injury, stimulates tissue damage by releasing catabolic enzymes and nitric oxide, creating a pro-inflammatory environment involving inflammatory cytokines which attract immune cells (monocytes, macrophages, and lymphocytes), perpetuating matrix destruction (He et al., 2020; Wei and Bai, 2016). Subchondral bone remodelling further accelerates disease progression through sclerosis, cystic degeneration, and osteophyte formation, which results in altered load distribution and thereby triggers cartilage damage through altered TGF-β and Wnt signalling (Chen et al., 2017; He et al., 2020).

Moreover, chondrocyte dysfunction, driven by senescence, apoptosis, mitochondrial impairment, and defective autophagy, exacerbates cartilage loss with the senescent chondrocytes releasing SASP factors (e.g., TGF-β and IL-6) that activate SMAD and STAT3 signalling and reinforce catabolic and inflammatory pathways. Mechanical stress initiates the activation of NF-kB and MAPK signalling by integrins and ion channels, whereas Wnt and TGF-β promote hypertrophy and subchondral remodelling (Rezus et al., 2021; Yunus et al., 2020; Zhang et al., 2021).

All these interdependent molecular and cellular mechanisms form a self-perpetuating cycle of inflammation, matrix degradation, and maladaptive bone remodelling, driving progressive dysfunction of the joints in knee osteoarthritis (Fig. 1).

Fig. 1.

Schematic illustration of the pathogenesis of knee osteoarthritis (KOA). Knee osteoarthritis pathogenesis illustrated the events in the disease progression. Mechanical stress/ injury or other factors trigger synovial inflammation along with the activation of macrophages, monocytes and leukocytes and the release of inflammatory cytokines (IL-1β, IL-6, TNF-α) and catabolic enzymes (MMPs, ADAMTS, NO). These, in turn, activate NF-κB, PI3K/AKT/mTOR, leading to Mitochondrial dysfunction, apoptosis, and altered autophagy in the chondrocyte. The altered chondrocyte releases SASP factors that propagate inflammation and matrix degradation. Altered TGF-β and Wnt signalling results in subchondral bone remodelling by osteophyte formation and sclerosis. These interconnected processes establish a vicious cycle with inflammation, cartilage degradation, joint space narrowing, pain and functional disability. ADAMTS: A Disintegrin and Metalloproteinase with Thrombospondin Motifs; AKT: Ak Strain Transforming; IL: Interleukin; MAPK: Mitogen-Activated Protein Kinase; MMPs: Matrix Metalloproteinases; mTOR: Mammalian Target of Rapamycin; NF-kB: Nuclear Factor-kB; NO: Nitric Oxide; SASP: Senescence-Associated Secretory Phenotype; TGF-β: Transforming Growth Factor-β; TNF-α: Tumor Necrosis Factor-α; PI3K: Phosphatidylinositol 3-Kinase.

CURRENT THERAPEUTIC PERSPECTIVES

Non-pharmacological treatments

Non-pharmacological treatments remain as the cornerstone of knee osteoarthritis (KOA) management and are recommended as first-line therapies by the international guidelines, including Osteoarthritis Research Society International (OARSI) and the American College of Rheumatology/Arthritis Foundation (Bannuru et al., 2019; Kolasinski et al., 2020). These approaches target biomechanical loading, joint strength, and patient self-efficacy, offering notable benefits with minimal adverse effects (Table 1).

Table 1.

Non-pharmacological treatment approaches for Knee osteoarthritis (KOA)

Intervention	Approach	Outcomes	Limitations	References
Patient Education	Informational sessions, counselling	Self-management, improved adherence, and health literacy	Variable	Liu et al., 2025
Exercise/Chinese therapy	Aerobic, aquatic flexibility training; Chinese therapy	Increase strength and flexibility, proprioception, neuromodulation, improved balance and reduced pain	Variable trial quality	Fransen et al., 2015; Liu et al., 2025; Mo et al., 2023; Zeng et al., 2023
Body weight loss and dietary intervention	Mediterranean diet, exercise; GLP-1 agonists (e.g., semaglutide)	Reduces weight (≥10-20%), significant symptom improvement (IDEA trial); Quality of life (FEAST trial); weight reduction, pain improvement (Semaglutide)	Long-term adherence difficult	Bliddal et al., 2024; Law et al., 2024; Messier et al., 2018
Mechanical devices	Orthoses, canes, or walking poles	Redistribute load, correct malalignment, improve stability, reduce pain, delay disease progression, improve mobility	Issues with compliance	Campbell et al., 2023
Psychological and behavioral therapies	Cognitive-behavioural therapy, mindfulness, and sleep optimisation	Reduces pain perception, enhances coping mechanisms and quality of life	Required patience and training	Lapane et al., 2021; Lin et al., 2023

Patient education

Patient education is considered the fundamental part of managing KOA as it aids in self-care, treatment adherence, decision-making, and also corrects misconceptions about KOA as an untreatable condition (Liu et al., 2025).

Exercise

Exercise is considered the most widely accepted intervention for KOA, which includes aerobic activity, resistance training, balance exercises, and different flexibility programs. Both land-based and aquatic exercises help to reduce knee pain and enhance the quality of life, along with physical functioning (Fransen et al., 2015; Liu et al., 2025). Additional studies and meta-analyses reported that exercise in addition to Traditional Chinese Therapies (TCTs), such as Tai Chi, Baduanjin, Wuqinxi, and Yijinjing, significantly reduced KOA-associated pain compared to placebo groups, reconizing Tai Chi and Baduanjin as the most effective TCTs for pain relief (Mo et al., 2023; Zeng et al., 2023).

Body weight loss and dietary intervention

Obesity is one of the major risk factors for KOA onset and progression. Excess body weight can elevate joint stress and facilitate adipokine release, resulting in joint inflammation. Studies (The IDEA trial) reported that a 10-20% reduction in body weight, along with exercise, notably improved pain, physical function, and reduced inflammatory mediators in KOA patients (Messier et al., 2018). Furthermore, it has been reported that the quality of life of obese individuals may be improved by dietary interventions, particularly Mediterranean and anti-inflammatory diets, and pharmacological agents like GLP-1 agonists (Bliddal et al., 2024; Law et al., 2024).

Mechanical devices

Use of mechanical devices such as orthoses, canes, or walking poles can redistribute joint loading, reduce mechanical stress, enhance balance, and facilitate mobility. However, these devices may be particularly advantageous in patients with advanced KOA (Campbell et al., 2023).

Psychological and behavioral therapies

Psychological interventions, including cognitive-behavioral therapy (CBT) and mindfulness-based therapies, can reduce pain catastrophizing, improve coping strategies with the disease condition, and enhance emotional resilience in individuals with chronic osteoarthritic pain (Lin et al., 2023). Additionally, sleep optimization can also improve pain perception and functional outcomes (Lapane et al., 2021).

In summary, non-pharmacological interventions are particularly safe, effective, and cost-efficient. Therefore, an individualized, multimodal program incorporating patient education, exercise, weight management, and psychological interventions could offer consistent therapeutic benefits.

PHARMACOLOGICAL THERAPIES

Pharmacological therapies are crucial in managing pain, enhancing function, and improving the quality of life in individuals with KOA. Commonly used pharmacological agents are listed in Table 2.

Table 2.

Comparison of the effectiveness of conventional pharmacological interventions in Knee osteoarthritis (KOA)

Drug	Structure	Mechanism of action	Efficacy	Limitations	References
NSAIDs	-	Inhibit cyclooxygenase (COX) enzymes and reduce prostaglandin-mediated inflammation and pain	Oral NSAIDs provide significant pain relief; topical formulations reduce localized pain with minimal systemic exposure	Oral NSAIDs: Gastrointestinal, renal, and cardiovascular risks; long-term use discouraged. Topical NSAIDs: limited penetration in deep joint tissues	Bannuru et al., 2019; Zhang et al., 2023
Opioids (e.g., Tramadol)		μ-opioid receptor agonists modulate central pain signalling	Short-term relief for refractory pain	Risk of dependence, tolerance, sedation, constipation; used as an alternative to first-line therapy	Zhang et al., 2023
Acetaminophen		Inhibit prostaglandin synthesis	Effective for mild pain; comparable outcome to NSAIDs in early KOA; well-tolerated	Limited anti-inflammatory effect; Poor efficacy in moderate-to-severe pain; risk of hepatotoxicity at high doses or with combination with others	Aminoshariae and Khan, 2015; Kolasinski et al., 2020;
Duloxetine (SNRI)		Inhibits serotonin and norepinephrine reuptake, modulating central pain pathways	Effective for patients with central sensitization or mood disorders; improves pain, function, and psychosocial outcomes; can be combined with NSAIDs or exercise therapy	Limited effect in nociceptive pain; potential side effects include nausea, fatigue, and sleep disturbances	Blikman et al., 2022; Rathbun et al., 2023
Topical capsaicin		Activates TRPV1 receptors on nociceptors, leading to desensitization and reduced pain signalling	Demonstrated clinically meaningful pain relief; some patients respond better than topical NSAIDs	Local burning sensation may limit adherence; variable response across individuals	Persson et al., 2021
Senolytics	-	Selectively remove senescent cells, reducing SASP-mediated inflammation and promoting tissue regeneration	Preclinical evidence shows reduced senescence, decreased inflammation, cartilage-protective effects, and pain reduction	Clinical efficacy is inconsistent; less effective in older patients; long-term safety under investigation	Ansari et al., 2024; Chin et al., 2023; Kapoor et al., 2023; Maurer et al., 2025; Wang et al., 2024; Yang et al., 2020; Zara-Danceanu et al., 2025; Zhang et al., 2021
Corticosteroids (intra-articular)	-	Anti-inflammatory via suppression of cytokines and immune mediators	Short-term pain and function improvement (weeks to months); limited long-term benefit	Effects diminish with repeated use; potential cartilage damage with chronic exposure	Billesberger et al., 2020; McAlindon et al., 2017
Hyaluronic acid (HA) injections		Restores synovial fluid viscosity; improves lubrication and shock absorption	Moderate improvement in pain and function; inconsistent evidence for disease modification	Response varies with OA severity and HA molecular weight; generally well-tolerated; slower onset than corticosteroids	Billesberger et al., 2020; Miller et al., 2020
Platelet-rich plasma (PRP)	-	Promote tissue repair and modulate inflammation	Potential to reduce pain and improve function; some studies show superior long-term benefit compared to HA	Mostly safe; minor local reactions; variability	Li et al., 2023; Lu et al., 2021; Tang et al., 2020

Non-steroidal anti-inflammatory drugs (NSAIDs)

Oral NSAIDs are effective for relieving KOA-related pain but may carry cardiovascular and gastrointestinal risks with long-term use (Bannuru et al., 2019). Topical NSAIDs provide localized pain relief with fewer systemic side effects and are preferred in older patients or those with comorbidities. Opioids, such as tramadol, may offer short-term relief but are associated with a higher risk of adverse events. They should be reserved for specific cases, used at the lowest effective dose, for the shortest possible duration, and always in conjunction with non-pharmacological therapies (Zhang et al., 2023).

Acetaminophen

Acetaminophen has traditionally been recommended as a first-line treatment for mild KOA due to its safety, affordability, and modest efficacy. It remains conditionally recommended in early-stage KOA, particularly in patients with contraindications to NSAIDs (Kolasinski et al., 2020). However, its effectiveness is limited in managing moderate to severe pain (Aminoshariae and Khan, 2015).

Duloxetine

Duloxetine, a serotonin-norepinephrine reuptake inhibitor (SNRI), exhibited efficacy in patients with central sensitization or coexisting mood disorders. It can improve pain, function, and psychological well-being and demonstrates notable activity when used with topical NSAIDs, bracing, or exercise therapy. However, duloxetine is particularly recommended as an adjunct therapy for persistent pain unresponsive to conventional therapies (Blikman et al., 2022; Rathbun et al., 2023).

Topical capsaicin

Capsaicin cream can alleviate pain by activating and subsequently desensitizing the TRPV1 receptor on nociceptors. Clinical studies reported variable outcomes, with many patients experiencing significant pain reduction while, in some cases, comparable or superior effects to topical ibuprofen gel (Persson et al., 2021).

Senolytics

The progression of knee osteoarthritis is closely associated with chondrocyte dysfunction resulting from cellular senescence. In case of aging and osteoarthritic cartilage, the accumulation of senescence-associated secretory phenotype (SASP) contributes to chronic inflammation, which in turn promotes tissue degeneration (Zhang et al., 2021). Senolytics can selectively eliminate senescent cells accumulating in the joint tissues, protecting chondrocyte integrity (Ansari et al., 2024). Studies reported that Dasatinib and quercetin can downregulate SASP-related factors, including IL6 and CXCL1, while Dasatinib can facilitate chondrogenesis (Maurer et al., 2025). Additionally, Senolytics like navitoclax (ABT263) and UBX0101 have demonstrated anti-inflammatory and cartilage protective effects by reducing the burden of senescent chondrocytes, decreasing SASP factor, and thereby attenuating inflammation and cartilage degradation. In addition, the elimination of senescent cells reduces inflammatory conditions and enhances cartilage regeneration (Chin et al., 2023; Yang et al., 2020). However, these effects are more pronounced in younger animals, and human trials have yielded mixed results. Innovative delivery systems, such as lipid-based nanoemulsions containing biodegradable and biocompatible vitamin E and sphingomyelin, along with the senolytic peptide NE:TUB1, demonstrated anti-senescent properties, highlighting the efficacy in attenuating osteoarthritis progression (Zara-Danceanu et al., 2025). Furthermore, Piperlongumine demonstrated anti-osteoarthritic effects in a goat model by inhibiting chondrocyte senescence and reducing inflammation (Kapoor et al., 2023). Moreover, fisetin, a flavonoid, has been found to reduce inflammation in osteoarthritic chondrocytes (Wang et al., 2024). However, despite the promise of senolytic, further research is needed to confirm their long-term safety and efficacy.

Intra-articular corticosteroid injections

Corticosteroid injections have been used for decades to reduce pain and inflammation in patients having KOA, but the efficacy varies, and the repeated use is likely associated with cartilage volume loss and diminished long-term efficacy. Therefore, their use should be limited and carefully monitored (Billesberger et al., 2020; McAlindon et al., 2017).

Hyaluronic acid (HA) injections

HA injections (viscosupplementation) can restore the viscosity of the synovial fluid, enhance joint lubrication and shock absorption. Clinical trials indicated that HA can moderately improve pain and function, although the effects are frequently delayed. However, the activity usually varies based on the severity of the disease, HA formulation, and method of administration. Hyaluronic acid is generally well-tolerated and often suitable for patients seeking long-term relief, avoiding the risks associated with the use of NSAIDs (Billesberger et al., 2020; Miller et al., 2020).

Platelet-rich plasma (PRP)

PRP therapy involves the administration of concentrated autologous blood products abundant in growth factors. Research has shown that PRP therapy can effectively reduce the expression of inflammatory cytokines such as IL-6, IL-1β, TNF-α, IL-17A, and IL-10 in the synovial fluid of KOA individuals. Moreover, PRP can also inhibit IL-8, IL-17F, and IL-4 while downregulating matrix metalloproteinases (MMPs), assisting in cartilage protection and pain relief. (Li et al., 2023; Lu et al., 2021). Additionally, other studies support that PRP may reduce OA-related symptoms, but the outcomes may vary depending on the formulation process and administration protocol. Emerging techniques like intraosseous PRP delivery demonstrated promise but require more extensive studies. Notably, PRP therapy is considered safe and well-tolerated with only minor and transient side effects (Tang et al., 2020). Therefore, adopting PRP therapy could improve the management of KOA.

NATURAL PRODUCTS AS THERAPEUTICS FOR KNEE OSTEOARTHRITIS

Natural products have attracted increasing attention for their potential in managing KOA, with various compounds exhibiting chondroprotective, anti-inflammatory, and antioxidant properties in in vitro, in vivo, and clinical settings and thereby offer pain relief, improved function, and disease-modifying effects for KOA (Table 3).

Table 3.

Mechanism of action of natural products used in KOA

Natural products	Chemical Structure	Cell type/model	Dose	Mechanism	References
Curcumin/Curcuma longa		Chondrocyte	0-10 μM	↓MMP-1, MMP-3, MMP-13, TNF-α, IL-1β, and IL-6→↑type II collagen, aggrecan and ↓OA progression	Buhrmann et al., 2021
		Male Sprague–Dawley rats	0-50 μM	↓MMP-3, MMP13→↓PI3K/AKT/mTOR→↑COL-II→↓OA-like lessions	Han et al., 2021
		Male Sprague–Dawley rats	0-10 μM	a) ↑PINK1, Parkin, LC3B, P62, Beclin1 and collagen II →↑mitophagy→↓cartilage degeneration b) ↓MMP13 and IL-1β →↓inflammation→↑chondrocyte proliferation and ↑mitochondrial function→↓cartilage degeneration	Jin et al., 2022
		Sprague–Dawley rats	0-40 mg/kg BW	a) ↓PGE2, NO, GAG→↓articular cartilage damage b) ↓ALOX5 and LTB4→↓NF-κB, COX 2, iNOS, ↓MMPs→↑type I, II collagen and aggrecan→↓OA progression	Kim et al., 2024
		Sprague-Dawley rats	0-50 mg/kg BW	↓iNOS, COX-2, 5-LOX→↓IL-1β, IL-6, TNF-α→↓MMPs (MMP-2, MMP-3, MMP-9, MMP-13→↑type II collagen and aggrecan→↓cartilage degradation→↓OA progression	Lee et al., 2025a
		Human chondrocytes	0-5 µM	a) ↓NFKB1, NFKB2, chemokines, cytokines (including TNFα, IL6, IL16, IL17R, IL34), NOS2, pro inflammatory genes (FSTL1, POSTN, IGFBP4, CHI3L1) →↓inflammation b) ↓MMP13, MMP1, ADAMTS5, MMP3, HTRA1→↓cartilage degradation c) ↓PLA2G2A, PLA2G4A, PTGS1, PTGS2, PTGES, PTGES2→↓inflammation	Sanchez-Lopez et al., 2022
		Sprague–Dawley rats	0-10 μM	a) ↓lipid ROS, MDA, Fe2+→↑GSH →↓ferroptosis→↓chondrocyte damage b) ↑SIRT5-dependent desuccinylation of ACSL4→↓ferroptosis→↓chondrocyte damage	Xu et al., 2025
		Male Sprague–Dawley rats	0-400 μg/kg BW	↓MMP-2 and MMP-9→↓miR 34a→↑E2F1, PITX1→↓apoptosis→↓Akt/mTOR→ ↑autophagy→↓OA-like lesions	Yao et al., 2021
Resveratrol		Male C57BL/6 J mice/SW1353	0-45 mg/kg.bw	↓p-JAK2, p-STAT3 →↓OA progression	Jiang et al., 2020
		Sprague–Dawley rats	0-10 mg/kg BW	↓iNOS, COX-2, 5-LOX→↓IL-1β, IL 6, IL 10, TNF-α→↓MMP13→↓OA progression	Wang et al., 2016
		Wistar albino rats	20 mg/kg/day	↓MAPK, Src kinase, STAT3 and Wnt signaling→↓inflammation	Oz et al., 2019
		OA mice	0-50 mg/kg	a) ↓COLX, MMP-13 →↑proteoglycan, lubricin, aggrecan→↑OPG and ↓RANKL→↓osteoclast differentiation→↓OA progression b) ↓CD31hiEmcnhi angiogenesis→↓VEGFA, angiopoietin-1→↓osteogenesis and vascularization of subchondral bone→↓OA progression	Xiong et al., 2021
		SW1353	0-50 μM	↓PI3K/Akt phosphorylation→↓pFoxO→↓TLR4→↓inflammation	Xu et al., 2020
Quercetin		CHON-001/ C57BL/6 mice	0-100μM	a) ↑p-AMPK, Nrf2 and Gpx4→↓ROS, Fe2+→↓ferroptosis→↓ECM degradation b) ↓IL-6, TNF-α, lipid ROS, Fe2+ and MDA→↑GSH→↓inflammation and OA symptoms	Dong et al., 2025
		Sprague-Dawley (SD) rats	0-8μM	a) ↓MMP-13→↑collagen II and aggrecan synthesis→↓OA progression b) ↓RHEB, p-mTOR, p-ULK1 and P62→↑TSC2, LC3BII→↑autophagy→↓apoptosis	Lv et al., 2022
		Sprague-Dawley rat	0-30μM	a) ↓MMP3, MMP13, INOS, COX−2, ROS→↓inflammation b) ↑SIRT1/Nrf−2/HO−1→↓ferroptosis→ ↓ECM degradation c) ↓IL−1β, TNF−α, MMP3, CTX−II and COMP→↓inflammation	Ruan et al., 2024
		Sprague–Dawley male rats	0-100 mg/kg	↓Inflammatory cytokines→↓MMP-3, MMP-13, ADAMTS4 and ADAMTS5→↑aggrecan and collagen II	Wang et al., 2023
		Sprague-Dawley rats	0-10mg/kg	↓IL-1β, TNF-α→↓TLR-4/NF-κB pathway→↓cartilage degradation	Zhang et al., 2020
Icariin		SW1353 chondrocytes	0-40 μM	a) ↓MMP-3→↑collagen II→↓cartilage degradation b) ↑LC3 II/I, Beclin 1→↓p62→↑autophagy c) ↓PI3K, Akt, and mTOR phosphorylation→↑ULK1→↑autophagy	Chen et al., 2022
		Chondrocytes/ Sprague-Dawley rats	0-20μM	↓MMP-1, MMP13, IL-1β, and IL-18→↓NLRP3, ASC, caspase-1 and GSDMD →↓pyroptosis→↓OA	Zu et al., 2019
Kaempferol		Male Sprague–Dawley rats	0-20 µM	↓IL-1β, TNF-α, iNOS→↓MMP-13 & MMP-3→↑collagen IIa1, aggrecan and SOX-9→↓OA progression	Estakhri et al., 2020
		Human chondrocyte, rat chondrocyte	0-20 µM	↓iNOS, COX2→↓MMP-1, MMP-3, MMP-13, ADAMTS4, ADAMTS5→↓p38/ERK, MAPK→↓STAT3→↓type II collagen degradation→↓OA progression	Huang et al., 2018
		ATDC5	0–50 μM	a) ↓IL-6, IL-8, and TNF-α→↓inflammation b) ↓miR-146a→↓apoptosis and caspase-3 activation→↓chondrocyte damage	Jiang et al.,2019
Boswellia serrata		Sprague–Dawley rats	0-150 mg/kg BW	↓NO, PGE2, LTB4, IL-6, TNF-α, MMP-3, MMP-13→↓Collagen type II alpha 1 and aggrecan loss	Choi et al., 2024
		Chondrocyte/ Male Sprague–Dawley rats	0-100 mg/kg BW	a) ↓Smad3, MMP-3, MMP-13, TIMP-1, TIMP-3, pro-inflammatory cytokines→↑collagen type II→↓OA progression b) ↓MMPs activation→↓caspase-3 →↓apoptosis→↓OA symptoms c) ↓COX-2, iNOS, MMPs→↓inflammation d) ↓p-IκB, p-NF-κB→↓NF-kB signal→↓OA	Kim et al., 2023
		Human chondrocytes	10–50 μg/ml	a) ↓IL6, CCL2, FSTL1, IL16, IL17, MyD88, TLR1, TLR4, TLR6→↓inflammation b) ↑FOXO1, LAMP2, RUBCNL, SQSTM1, and PLEKHM1→↑autophagy c) ↓ADAMTS1 and 5, MMP3, MMP13→↓cartilage degradation	Sanchez-Lopez et al., 2022
Star anise (Illicium verum)		SW1353 human chondrocyte/Rat	0-20 mM	a) ↓iNOS, COX-2, MMP3, MMP13, ADAMTS-5, type X collagen, and p62→↓inflammation b) ↑type II collagen, ATG7, Beclin-1, and LC3→↑autophagic flax c) ↓MAPK and NF-kB→↓inflammation and cartilage degradation	You et al., 2021
Undenatured Type II Collagen (UCII)		Sprague–Dawley rats	0-4 mg/kg BW	a) ↓TNF-α, IL-6, IL-1β, CRP, PGE2, and COMP→↓NF-κB signal→↓MMP3, RANKL, IRF-7 and MCP-1→↓inflammation→↓cartilage degradation b) ↓PGE2→↓IL-1β, IL-6, TNF-α, COX-2, MCP-1, NF-κB, MMP-3, RANKL→↓cartilage degradation	Orhan et al., 2021
		Sprague–Dawley rats	0-2 mg/kg	↓IL-10, IL-6, TNF-α, COX2→↓NF-κβ→↓MMP3 and COMP→↓cartilage degradation	Sahin et al., 2023
ASU		Synoviocytes	27.5 mg/kg	↓Pro-inflammatory cytokine→↓MMP-13, iNOS→↓inflammation	Al-Afify et al., 2018
		Sprague–Dawley rats	0-0.66 mg/kg	↓Osteophyte formation→↓cartilage deterioration	Bagi et al., 2017
Omega-3 fatty acids		Human chondrocytes/C57BL/6J mice	0-50mg/mL	a) ↓pp38, MAPK, p53, MMP3→↓disease progression b) ↓MMP13→caspase 3→↓chondrocyte apoptosis	Sakata et al., 2015
Ginger/Gingiber officinale		Sprague–Dawley rats	0-50 mg/kg/day	↓IL-1β, TNF-α, IL-6→↓Articular cartilage degradation	Luo et al., 2021
		C28I2 human chondrocyte	0-25 μg/mL	a) ↑catalase, superoxide dismutase-1, glutathione peroxidase-1, glutathione peroxidase-3, and glutathione peroxidase-4; b) ↓ROS, lipid peroxidation, Bax/Bcl-2 ratio, and caspase-3 activity→↓apoptosis	Hosseinzadeh et al., 2017
		Sprague–Dawley rats	0-100 mg/kg/day	a) ↑type II collagen and aggrecan→↓cartilage degradation b) ↓Nitric oxide synthase, cyclooxygenase-2, 5-lipoxygenase, interleukin (IL)-1β, IL-6, TNF-α, MMP-2, MMP-3, MMP-9, and MMP-13→↓disease progression	Lee et al., 2025b

↑: Induction/increase; ↓: Inhibition/decrease; ADAMTS: A Disintegrin And Metalloproteinase With Thrombospondin Motifs; COX-2: Cyclooxygenase-2; COMP: Cartilage Oligomeric Matrix Protein; iNOS: Inducible Nitric Oxide Synthase; LTB4: Leukotriene B4; MAPK: Mitogen-Activated Protein Kinase; MCP-1: Monocyte Chemoattractant Protein-1; MMP: Matrix Metalloproteinase; mTOR: Mammalian Target of Rapamycin; PGE2: Prostaglandin E2; RANKL: Receptor Activator of Nuclear Factor-Κb Ligand; ROS: Reactive Oxygen Species; TNF-α: Tumor Necrosis Factor-α; NF-kβ: Nuclear Factor Kappa β; TGF-β: Transforming Growth Factor-β; VEGFA: Vascular Endothelial Growth Factor A.

Curcumin

Curcumin, the principal bioactive compound in turmeric, exhibits potent antioxidant and anti-inflammatory activities. Several studies demonstrated the chondroprotective potential of curcumin, particularly by modulating signalling pathways involving oxidative stress, inflammation, and extracellular matrix degradation in KOA. Curcumin can suppress OA progression by increasing type II collagen and aggrecan synthesis by downregulating MMP-1, MMP-3, MMP-13, TNF-α, IL-1β, and IL-6 (Buhrmann et al., 2021). Han et al. (2021) demonstrated that curcumin inhibits PI3K/Akt/mTOR signalling by downregulating MMP-3, MMP-13 expression, reducing OA symptoms. Additionally, curcumin can promote mitochondrial homeostasis by upregulating PINK1, Parkin, LC3B, P62, Beclin1 and Collagen II, enhancing mitophagy and thereby reducing mitochondrial dysfunctions and chondrocyte apoptosis (Jin et al., 2022). Moreover, curcumin exhibited potential antioxidant and anti-inflammatory activities by reducing PGE2, NO, GAG and reducing cartilage damage by suppressing ALOX5 and LTB4, which in turn downregulate NF-κB, COX-2, iNOS, MMPs and facilitate collagen synthesis (Kim et al., 2024; Lee et al., 2025a). Additionally, another study reported that curcumin can reduce inflammation and cartilage deterioration by suppressing NF-kB, inflammatory cytokine and pro-inflammatory genes. In parallel, downregulation of metalloproteinases, ADAMTS5, MMP-3, HTRA1, PLA2G2/ PTGS/PTGES further inhibit ECM degradation (Sanchez et al., 2022). Furthermore, curcumin can increase cellular redox balance by inhibiting ferroptosis and inducing autophagy, thereby reducing OA progression (Xu et al., 2025; Yao et al., 2021). Additionally, randomized controlled trials (RCTs) and systematic reviews report modest analgesic and functional improvements compared to placebo, though further validation is needed (Hidayat et al., 2025; Wang et al., 2020).

Resveratrol

Resveratrol is a polyphenolic compound demonstrated numerous beneficial effects in OA, including suppression of inflammatory mediators (iNOS, COX-2, 5-LOX, IL-1β, IL-6, IL-10, TNF-α), s and MMP-13, while enhancing cartilage matrix synthesis and lubricin production. It also inhibits osteoclastogenesis, resulting in reduced bone remodelling associated with OA. Mechanistically, resveratrol modulates the MAPK, Src kinase, Wnt, PI3K/Akt and JAK2/STAT3 signalling cascade to reduce inflammation and cartilage damage (Jiang et al., 2020; Oz et al., 2019; Wang et al., 2016; Xiong et al., 2021; Xu et al., 2020).

Quercetin

Quercetin, a naturally occurring phytochemical, can inhibit pro-inflammatory cytokines (IL-1β, TNF-α), block NF-κB activation, and suppress MMP activity. It can reduce ECM degradation by inhibiting ferroptosis through upregulation of p-AMPK, Nrf2 and Gpx4. Additionally, it can also stimulate aggrecan and collagen II synthesis, reduce chondrocyte apoptosis, and promote autophagy by increasing TSC2, LC3BII expression. Moreover, quercetin can induce SIRT1/Nrf2 activation and PI3K/mTOR inhibition, which collectively retarding OA progression (Dong et al., 2025; Lv et al., 2022; Ruan et al., 2024; Wang et al., 2023; Zhang et al., 2020).

Icariin

Icariin can reduce inflammation, apoptosis and extracellular matrix degradation and thereby demonstrate chondroprotective effects. It can downregulate MMP-1, MMP-13, IL-1β, IL-18, and NLRP3 inflammasome activation. Moreover, by increasing the expression of LC3 II/I, Beclin 1-mediated reduction of p62, and blocking PI3K/Akt/mTOR signal, Icariin can promote autophagy and maintain cartilage integrity (Chen et al., 2022; Zu et al., 2019).

Kaempferol

Various studies demonstrated that Kaempferol can reduce the expression of inflammatory cytokines IL-1β, IL-6, IL-8, TNF-α, iNOS, COX2, and matrix metalloproteinases (MMP-1, MMP-3, MMP-13) and increase collagen II and aggrecan synthesis, demonstrating anti-inflammatory and antioxidative effects. Additionally, Kaempferol can inhibit NF-κB, MAPK, and STAT3 signalling pathways, reducing cartilage degradation. Moreover, miR-146a downregulation by Kaempferol depicted the chondroprotective potential in both in vitro and in vivo studies (Estakhri et al., 2020; Huang et al., 2018; Jiang et al., 2019).

Boswellia serrata

Boswellia serrata, can inhibit COX-2, iNOS, 5-LOX, NF-kB signalling, and downregulate the expression of NO, PGE2, LTB4, IL-6, TNF-α, MMP-3, and MMP-13, along with accelerating collagen II synthesis, demonstrating anti-inflammatory and cartilage-protective activity (Choi et al., 2024; Kim et al., 2023; Sanchez et al., 2022). Boswellia is also capable of slowing down cartilage degradation by enhancing autophagy through the inhibition of FOXO1, LAMP2, RUBCNL, SQSTM1, and PLEKHM1 (Sanchez et al., 2022). Preclinical studies reported that it is effective in reducing joint swelling, cartilage degradation, and tissue deformation. Moreover, RCTs and meta-analyses demonstrated that Boswellia could improve pain, stiffness, and function and improve the quality of life of OA individuals. It is generally well-tolerated, with only mild gastrointestinal disturbances in some cases (Kumar et al., 2025; Vaidya et al., 2025).

Star anise (illicium verum)

Star anise demonstrated anti-inflammatory and cartilage protective effects by downregulating iNOS, COX-2, MMPs, and ADAMTS-5, p62, while increasing type II collagen, ATG7, Beclin-1 and LC3 expression to enhance autophagic activity. Moreover, it also suppresses NF-κB and MAPK signalling pathways, thereby reducing OA progression (You et al., 2021).

Undenatured type II collagen (UC-II)

UC-II supplementation modulates immune and inflammatory responses in OA by inhibiting NF-κB and reducing TNF-α, IL-6, IL-1β, and PGE₂ levels. This may also downregulate MMP-3 and support cartilage preservation. Animal models show that UC-II reduces cartilage damage and osteophyte formation (Orhan et al., 2021; Sahin et al., 2023). Clinical studies report improved function and reduced pain after 40 days of supplementation (Salinas-Camargo et al., 2025). In MIA-induced OA models, higher UC-II doses improved gait and reduced inflammation. While UC-II shows promise for maintaining joint function and quality of life, more long-term, large-scale trials are needed.

Avocado-soybean unsaponifiable (ASU)

ASU exhibits anti-inflammatory and cartilage-preserving effects by upregulating aggrecan, collagen types I and II, and tissue inhibitors of metalloproteinases (TIMP-1, TIMP-3), while suppressing pro-inflammatory cytokines, iNOS, MMP-3, and MMP-13 (Al-Afify et al., 2018; Bagi et al., 2017). ASU has been shown to alleviate pain and improve quality of life in KOA patients, particularly those using NSAIDs, potentially reducing their associated gastrointestinal and cardiovascular side effects. It may be used as an adjunct therapy in the long-term treatment process (Simental-Mendia et al., 2019).

Omega-3 fatty acids

Marine-derived omega-3 polyunsaturated fatty acids (PUFAs), particularly eicosapentaenoic acid (EPA), can alleviate OA symptoms by reducing inflammation and cartilage degradation. It has been reported that EPA can inhibit the phosphorylation of p38 and downregulate the expression of MAPK, p53, MMP3, MMP-13, and caspase 3, resulting in reduced chondrocyte apoptosis, thereby improving OA (Sakata et al., 2015). Clinical evidence indicates that omega-3 supplementation can improve mobility, reduce pain, and lower mortality in patients with systemic inflammation or metabolic comorbidities (Deng et al., 2023; Huang et al., 2024).

Ginger

Ginger has been extensively studied due to having different phytoconstituents such as gingerols, shogaols, paradols, and terpenes with significant antioxidant and anti-inflammatory activities. Different studies supported that ginger can reduce IL-1 β, TNF-α, IL-6, MMP-2, MMP-3, MMP-9, and MMP-13 expression, thereby attenuating cartilage damage. It can also increase type II collagen and aggrecan and reduce IL-1β-induced oxidative stress, mitochondrial dysfunction and thereby protect chondrocyte apoptosis (Hosseinzadeh et al., 2017; Lee et al., 2025b; Luo et al., 2021). Additionally, other studies demonstrated the role of ginger in reducing muscle pain, improving stiffness, and enhancing functional outcomes. It can suppress the expression of biomarkers associated with inflammation, including IL-6, INF-γ, TNF-α, CRP, FoxP3, RORγt, and T-bet (Broeckel et al., 2025). Moreover, meta-analyses suggested that ginger is effective in reducing knee pain and improving the function of KOA individuals. Ginger is usually considered safe and well-tolerated, with minimal adverse effects (Szymczak et al., 2024).

ADVANCED AND EMERGING THERAPIES FOR KNEE OSTEOARTHRITIS

Mesenchymal stromal cells (MSCs)

Mesenchymal stromal cells (MSCs), derived from bone marrow, adipose tissue, or microfragmented adipose tissue, represent an emerging therapeutic alternative for the management of osteoarthritis because of their ability to develop into chondrocytes, modulation of inflammation and facilitate tissue regeneration (Wu et al., 2024). Clinical trials indicate that MSCs could reduce pain and mediate functional improvement of patients with non-operative KOA (Cao et al., 2025). However, the safety, efficacy and standardization of MScs are yet to be reported, and therefore further comprehensive and large-scale studies are needed before they can be adopted as a mainstream therapeutic option.

Gene therapy

Gene therapy aims to modify disease mechanisms by delivering therapeutic genes into affected joints to suppress inflammation, prevent cartilage degradation, or stimulate cartilage regeneration (Kim et al., 2025). Viral vectors such as adeno-associated viruses (AAV) and lentiviruses have been used to deliver genes, including IL-1 receptor antagonist (IL-1Ra), TGF-β, and SOX9, a chondrogenic transcription factor (Zhou et al., 2025). For example, the TGF-β1 gene-modified cell therapy Invossa™ (TissueGene-C) showed clinical efficacy in Korea. Recent long-term data presented at OARSI 2025 indicated the absence of tumor cases and a diminished need for knee replacement surgery, supporting TG-C’s potential as a disease-modifying therapy (Kolon TissueGene, Inc., 2025). Additionally, adenovirus-mediated relaxin (Ad-RLN) therapy has shown potential by reducing mRNA expression of collagen types I, III, and IV and multiple MMPs, while also decreasing TIMP-1 and TIMP-2 protein levels (Ko et al., 2019). The study suggests an anti-fibrotic effect, positioning relaxin as a potential therapy for early-stage KOA with flexion contracture.

Gene editing

Gene editing techniques, particularly by CRISPR-Cas9, offer significant potential for the modification of genes associated with inflammation and cartilage degradation. It is found that gene editing enables accurate targeting of genes such as MMP-13, IL-1β, and nerve growth factor (NGF), in animal models demonstrating diminished structural damage and inflammation in post-traumatic OA (Zhao et al., 2019). Ponta et al. (2024) reported that a non-viral CRISPR-Cas9 electroporation system achieved approximately 90% knockout efficiency of RelA (an NF-κB subunit) in primary human chondrocytes preserving cell viability and chondrogenic potential. Other notable targets include connexin 43 (Cx43), a regulator of chondrocyte senescence, and NF-κB, both of which are capable of fostering a pro-regenerative environment in osteoarthritic cartilage (Varela-Eirin et al., 2018). Jia et al. (2024) demonstrated that inflammatory signalling and ageing-related gene targeting could inhibit catabolic pathways and facilitate cartilage homeostasis. Moreover, emerging models advocate the integration of CRISPR-Cas9-based editing in conjunction with epigenome and RNA-editing technologies, expanding the therapeutic potential (Vlashi et al., 2024). Altogether, CRISPR-Cas9-based gene editing represents a novel, potentially transformative and effective approach for the development of targeted therapies aimed at halting or reversing KOA progression.

Biologic therapies

Biologic therapies have been widely used in the treatment of rheumatoid arthritis (RA), targeting to eliminate inflammatory cytokines, and have also been investigated in KOA. However, their success has been limited. TNF inhibitors demonstrated very little efficacy in KOA compared to ROA (Estee et al., 2023). Additionally, IL-1 inhibitors, such as Lutikizumab, have shown anti-inflammatory properties and cartilage and synovium-preserving effects with limited long-term benefits (Fleischmann et al., 2019).

Disease-modifying osteoarthritis drugs (DMOADs)

Several DMOADs are under investigation for their ability to halt or reverse disease progression in KOA. Sprifermin, a recombinant human fibroblast growth factor 18 (rhFGF18), has shown promise in promoting cartilage regeneration. It is reported that sprifermin increased the proportion of COL2:COL1 and promoted human OA chondrocytes proliferation, possibly by ERK signalling (Gigout et al., 2017). A FORWARD trial reported a dose-dependent increase in cartilage thickness and delayed need for joint replacement in symptomatic KOA. While encouraging, further research is necessary to establish long-term benefits and optimal dosing strategies (Eckstein et al., 2021). Dysregulation of Wnt signalling contributes to cartilage breakdown and aberrant bone remodelling. Small molecules such as lorecivivint (SM04690) downregulated Wnt signalling of CDC-like kinase 2 (CLK 2) and bispecific tyrosine phosphorylation kinase 1a (DYRK 1A), and reduced cartilage deterioration and also improved symptoms associated with KOA. However, late-phase trials have shown mixed outcomes, emphasising the need for better patient stratification and longer follow-up (Deshmukh et al., 2020; Lietman et al., 2018). Nerve Growth Factor (NGF), such as neurotrophin, can mediate pain sensitization. Tanezumab, a humanised IgG2 monoclonal antibody, can inhibit tropomyosin receptor kinase A (TrkA), abolishing interaction with NGF and thereby reducing pain (Gondal et al., 2022). Tanezumab (phase II) and fasinumab (phase III) have been reported to demonstrate significant pain-relieving ability compared to NSAIDs (Dakin et al., 2019; Hochberg et al., 2021). Another study reveals that the dose of 2.5 mg tanezumab is safer than the higher doses (Berenbaum et al., 2022) (Fig. 2).

Fig. 2.

Mechanism of disease-modifying osteoarthritis drugs (DMOADs). Key targeted signalling pathways of potential disease-modifying drugs. (Left) Sprifermin works by stimulating the FGF receptor, increasing chondrocyte proliferation along with collagen ratio and promoting cartilage regeneration. (Bottom) Lorecivivint (SM04690) inhibits CLK 2 and DYRK 1A receptors and subsequent Wnt signalling, which reduces inflammation and chondrocyte hypertrophy and reduces cartilage damage. (Right) Tanezumab/Fasinumab blocks the NGF/TrkA axis and reduces pain sensitization and inflammation. CLK 2: CDC-Like Kinase 2; COL: Collagen; DYRK 1A: Dual-Specificity Tyrosine Phosphorylation-Regulated Kinase 1A; DMOADs: Disease-Modifying Osteoarthritis Drugs; FGF: Fibroblast growth factors; NGF: Nerve Growth Factor; TrkA: Tropomyosin Receptor Kinase A.

Digital healthcare and artificial intelligence

In modern times, the therapy of knee osteoarthritis has been facilitated by the use of wearable technologies such as accelerometers and inertial measurement units, allowing continuous tracking of gait and physical activity. These devices can capture and provide quantitative data on the gait parameters, including the length, width, speed, and frequency of stride, and levels of motion of the knee and thereby, surpass the shortcomings of conventional patient-reported measures and clinical assessment (Feng et al., 2023). Moreover, wearable sensors can help physicians to monitor minute alterations in mobility and the rehabilitation process, identify suboptimal recovery processes and minimize the number of unnecessary visits to the physicians. Research indicates that this type of monitoring can effectively differentiate the healing patterns, allowing better patient adherence and enabling individualized rehabilitation, with improved functional outcomes and patient satisfaction. Additionally, remote patient monitoring utilizing these technologies ensures significantly improved rehabilitation in remote or underserved areas (Lebleu et al., 2024; Yang et al., 2023).

Additionally, in recent days, the involvement of AI technology in the analysis of MRI, X-ray data for the initial diagnosis and prognosis prediction of KOA is rapidly increasing. It has been reported that deep learning networks, such as OA-MEN, fused with ResNet and MobileNet features, achieved an average accuracy level of 84.88% while another model, OA-HybridCNN, by integrating ResNet and DensNet architectures, demonstrated higher accuracy (91.77%) and precision (92.34%) than traditional methods in the detection and classification of KOA severity based on X-ray images (Liao et al., 2025; Ren et al., 2025). Furthermore, ensemble networks based on AI and transfer learning techniques displayed enhanced consistency and reliability of the KL grading system, with significant accuracy (Pi et al., 2023). In this connection, machine and deep learning models may be utilized in MRI analysis to predict the progression of disease and identify high-risk patients based on complex multidimensional data, enabling early intervention and individualized treatment (Lv et al., 2025). Therefore, the developments of imaging technologies based on AI might play a significant role in defining precision medicine for KOA, confirming more accurate diagnosis and better predictions of disease progression.

KOA PHENOTYPES AND PERSONALIZED TREATMENT

Knee osteoarthritis has now been regarded as a complex disease with different phenotypes, each having distinct mechanical insights, including inflammatory dominant, metabolic disorder-related, biomechanical, ageing-associated, pain-driven and structural types associated with cartilage and bone (Dell’isola and Steultjens, 2018; Hunter and Deveza, 2025; Lv et al., 2021; Sanchez-Lopez et al., 2022). The advancement of molecular mechanistic approaches and imaging technologies has brought a positive outcome in the identification of knee osteoarthritis (KOA) phenotype through biomarkers from blood, urine, and synovial fluid, and technologies like MRI and CT scan. However, these phenotypes may frequently intersect and significantly influence the critical prognosis and development of targeted therapeutics. Here, we summarized the frequently studied phenotypes with their possible therapeutic measures (Table 4, Fig. 3).

Table 4.

Knee osteoarthritis phenotypes, related biomolecules, MRI/CT features and potential therapeutic strategies

Phenotypes	Pathophysiological features	Biomolecules involved	MRI/CT imaging	Treatment options	References
Inflammation driven	Cytokine activation, synovial inflammation, and immune cell infiltration	Plasma/SF: ↑IL-1β, IL-1Ra, CCL3, CCL4, IL-6, TNF-α, CD163, CD14 Plasma: ↑CCL11, IL-15, CCL11, CCL19 SF: ↑CD14, CD163, elastase, C4S, IL-6, IL-8, MMP-1/3, TIMP-1, TNF-α, VEGF	MRI: Synovial thickening (inflammation); joint effusion, alterations in the infrapatellar fat pad; bone marrow lesions	IL-1 inhibitor, TNFα inhibitor, NSAIDs, COX-2 inhibitor and metformin; Platelet-rich plasma (PRP) therapy or mesenchymal stromal cell therapies	Andia et al., 2021; Calvet et al., 2023; Fleischmann et al., 2019; Haraden et al., 2019; Lv et al., 2021;McAlindon et al., 2017; Sanchez-Lopez et al., 2022; Shin et al., 2024; Wu et al., 2024; Zhao et al., 2015
Metabolic/Obesity related	Inflammatory response, altered metabolic process	Serum/SF: ↑Leptin, adiponectin, resistin, visfatin, and chemerin Serum: ↑hs-CRP, IL-6, COMP	MRI: Diffuse cartilage thinning, bone marrow lesion, synovitis	Dietary, weight loss and exercise program; Metformin, GLp1 agonist (Semaglutide)	Bliddal et al., 2024; Calvet et al., 2023; Hunter and Deveza, 2025; Law et al., 2024
Structural phenotype (Cartilage degradation)	Cartilage degradation, inflammation	Urine: ↑CTX-II, C2C, C2M, C-Col X	MRI: Cartilage thinning, surface fracture	Hyaluronic acid, Glucosamine, Chondroitin, Sprifermin, Bisphosphonates	Bannuru et al., 2019; Householder et al., 2023; Lv et al., 2021; Roemer et al., 2023
Bone remodelling-driven phenotype	Altered bone remodelling, fractures, sclerosis, cartilage degradation	Urine: ↑CTX-I, NTX-I, TRAP5b Serum: ↑PINP, ALP, C1M	MRI/CT: Subchondral cysts, bone marrow lesion, sclerosis, bone thickening, and meniscal protrusion	Bisphosphonate, Zoledronic acid, Osteoprotegerin, Calcitonin, MIV-711	Cai et al., 2020; Lopez et al., 2022; Lv et al., 2021
Pain-driven phenotype	Altered nociceptor signalling, central sensitization	Serum/SF: ↑NGF, Bradykinin, CRPM, CGRP, hs-CRP, NGF	MRI: Cartilage defects, bone marrow lesions, effusion, and synovitis	Opioids, NSAIDs, Duloxetine, Tanezumab, CGRP inhibitor, Capsaicin	Bannuru et al., 2019; Berenbaum et al., 2020; Blikman et al., 2022; Kolasinski et al., 2020; Lv et al., 2021; Pan et al., 2019; Shin et al., 2024
Post-traumatic type	Joint injury, abnormal cartilage metabolism, and inflammation	Plasma/SF: ↑NOx, IL-1β, IL-6, IL-18, and leptin	MRI: Focal cartilage alteration, meniscal tear, subchondral sclerosis, synovial inflammation	Physiotherapy, offloading braces/orthoses, corticosteroids or hyaluronic acid	Boffa et al., 2021; Campbell et al., 2023; Lieberthal et al., 2015; Panina et al., 2017
Aging/Senescent-driven	Cellular senescence, oxidative stress, and diminished repair	Serum: ↑p16INK4A, IL-1β, IL-6, SASP	MRI: Diffuse cartilage thinning, osteophyte, mild synovial inflammation	Lifestyle modification, Sleep optimization exercise, Senolytics, Duloxetine	Ansari et al., 2024; Blikman et al., 2022; Sanchez-Lopez et al., 2022

↑: Increase; ↓: Decrease; ALP: Alkaline Phosphatase; C2C: Cartilage Collagen Type II Cleavage; C1M: Collagen Type I Degradation Marker; C2M: Collagen Type II Degradation Marker; C4M: Chondroitin-4-Sulfate; CCL: C-C Motif Chemokine Ligand; CD: Cluster of Differentiation; CGRP: Calcitonin Gene-Related Peptide; COMP: Cartilage Oligomeric Matrix Protein; Col X: Collagen X; COX-2: Cyclooxygenase-2; CTX-I: C-Terminal Crosslinked Telopeptide of Collagen Type I; CRPM: C-Reactive Protein M; CT: Computed Tomography; Hs-CRP: Highly Sensitive C-Reactive Protein; IL: Interleukin; MCP-1: Monocyte Chemoattractant Protein-1; MMP: Matrix Metalloproteinase; MRI: Magnetic Resonance Imaging; NGF: Nerve Growth Factor; NO_x: Nitrogen Oxides; NTX-I: N-Terminal Crosslinked Telopeptide of Type I Collagen; PINP: Procollagen Type I N-Propeptide; SASP: Senescence-Associated Secretory Phenotype; SF: Synovial Fluid; TIMP1: Tissue Inhibitor of Metalloproteinases-1; TNF-Α: Tumor Necrosis Factor-Α; TRAP5b: Tartrate-Resistant Acid Phosphatase 5b; VEGF: Vascular Endothelial Growth Factor.

Fig. 3.

Knee osteoarthritis phenotypes with associated biomarkers and therapeutic approaches. Overview of major phenotypes of knee osteoarthritis (KOA) characterized by distinct biomarkers (e.g., cytokines and/or signalling molecules), and possible therapeutic options like anti-inflammatory drugs, senolytic, DMODs and other relevant medications, signifying the heterogeneity of the disease. for the types described. ALP: Alkaline Phosphatase; CD: Cluster of Differentiation; CGRP: Calcitonin Gene-Related Peptide; COMP: Cartilage Oligomeric Matrix Protein; COX-2: Cyclooxygenase-2; CRPM: C-Reactive Protein Metabolite; CTX: Cross-linked C-Telopeptide; C2C: Collagenase-Generated Cleavage Neoepitope of Type II Collagen; C2M: MMP-degraded Type II Collagen; hsCRP: High-Sensitivity C-Reactive Protein; GLp1: Glucagon-Like Peptide-1; IL: Interleukin; MMP: Matrix Metalloproteinase; NOx: Nitrogen Oxides; NSAIDs: Nonsteroidal Anti-Inflammatory Drugs; NTX: N-terminal Telopeptide; p16INK4A: Protein 16, Inhibitor of CDK4, isoform A; PINP: Procollagen type I N-Propeptide; SASP: Senescence-Associated Secretory Phenotype; TIMP-1: Tissue Inhibitor of Metalloproteinases 1; TNF-α: Tumor Necrosis Factor-α; TRAP5b: Tartrate-Resistant Acid Phosphatase 5b; VEGF: Vascular Endothelial Growth Factor.

Inflammatory phenotypes

These are characterized by higher levels of inflammatory mediators such as IL-1β, IL-1Ra, IL-6, TNF-α, CCL3, CCL4, chemokines, and markers of macrophage/neutrophil activation (CD163, CD14) in synovial fluid and plasma and are frequently associated with synovial inflammation, bone marrow lesions and joint effusion (MRI) imaging (Calvet et al., 2023; Haraden et al., 2019; Lv et al., 2021; Sanchez-Lopez et al., 2022; Zhao et al., 2015). Moreover, Shin et al. (2024) reported the presence of elevated levels of inflammatory markers pCCL11, pIL-15, uCCL11, and uCCL19 denoting joint inflammation. Anti-inflammatory biologics, including IL-1 and TNF inhibitors, NSAIDs, COX-2 inhibitors, along with metformin, platelet-rich plasma (PRP) therapy or mesenchymal stromal cell therapies have been reported to be effective in reducing inflammation (Andia et al., 2021; Fleischmann et al., 2019; McAlindon et al., 2017; Wu et al., 2024).

The metabolic/obesity related phenotypes

These types are associated with altered adipokines, Leptin, adiponectin, resistin, visfatin, and chemerin, along with increased serum levels of hs-CRP, IL-6, and COMP. MRI findings typically involve diffuse cartilage thinning, bone marrow lesions, and synovitis. Dietary intervention, along with weight loss and exercise programs, Metformin, GLP-1 agonist (Semaglutide) may provide beneficial effects (Bliddal et al., 2024; Calvet et al., 2023; Hunter and Deveza, 2025; Law et al., 2024).

Pain-driven phenotype

Pain-driven phenotype has been recognized by the increased concentration of highly sensitive C-reactive protein (hs-CRP) with knee pain and functional loss in KOA. Moreover, it is associated with elevated levels of serum and synovial fluid NGF, bradykinin, CRPM, and CGRP. MRI imaging often reveals cartilage defects, bone marrow lesions, effusion, and synovitis. Therapy primarily focuses on pain modulation with Opioids, NSAIDs, Duloxetine, NGF inhibitor (Tanezumab), CGRP inhibitors and topical capsaicin (Bannuru et al., 2019; Berenbaum et al., 2020; Blikman et al., 2022; Kolasinski et al., 2020; Lv et al., 2021; Pan et al., 2019; Shin et al., 2024).

Structural phenotypes (cartilage degradation driven)

Structural phenotypes in turn are characterized by elevated urinary CTX-II, C2C, C2M, and C-Col X levels reflecting collagen turnover and cartilage degradation, while MRI demonstrates cartilage thinning and surface fissures. cartilage-reforming agents, including Sprifermin, hyaluronic acid, and chondroitin, or anti-resorptive agents, such as bisphosphonates, have shown promise to keep the integrity of cartilage (Bannuru et al., 2019; Householder et al., 2023; Lv et al., 2021; Roemer et al., 2023).

Bone remodelling types

These types are represented by altered urinary CTX-I, NTX-I, C1M, TRAP5b and elevated serum ALP, PINP and C1M levels, denoting abnormal bone formation and specific MRI traits, such as subchondral bone cysts, bone marrow lesions, meniscal protrusion, and osteophyte. Therapeutic approaches targeting bone metabolism, including bisphosphonate, zoledronic acid, osteoprotegerin, calcitonin, and MIV-711, may be utilized to reduce the turnover (Cai et al., 2020; Lopez et al., 2022; Lv et al., 2021).

Post-traumatic types

These are often connected with joint injury and mechanical stress, leading to cartilage damage along with increased levels of plasma and synovial NO_x, IL-1β, IL-6, IL-18, and leptin. MRI often shows focal cartilage alteration, meniscal tear, subchondral sclerosis and mild synovial inflammation. Biomechanical interventions like physiotherapy, offloading braces/orthoses, may be effective along with corticosteroids or hyaluronic acid (Boffa et al., 2021; Campbell et al., 2023; Lieberthal et al., 2015; Panina et al., 2017).

Ageing/Senescent-driven types

Ageing/Senescent-driven types are characterized by cellular senescence, oxidative stress, and diminished repair capacity, along with altered p16^INK4A, IL-1β, IL-6, and SASP levels in the blood. MRI imaging demonstrates cartilage thinning, osteophyte, and mild synovial inflammation. Lifestyle modification, along with sleep optimization and exercise, can be good management options. Use of Senolytics and Duloxetine may also alleviate symptoms (Ansari et al., 2024; Blikman et al., 2022; Sanchez-Lopez et al., 2022)

Overall, phenotype-based personalised, KOA management can lead to improved clinical outcomes with maximized utilization of the available resources. However, further, more long-term studies are needed to substantiate phenotype-specific biomarkers, improve classification algorithms and prove the efficacy of targeted therapies in diverse patient groups.

ADJUNCT THERAPIES

In order to enhance patient outcomes in knee osteoarthritis, several adjunct therapies, such as thermal therapy, laser therapy, electrical stimulation, and acupuncture, are commonly employed with core therapies. They may improve symptoms and enhance functional outcomes, but their overall effectiveness varies. Therefore, future research focusing on combining these therapies with regenerative medicines or gene therapy may maximise the benefits (Carvalho et al., 2024; Iijima et al., 2020).

SURGICAL APPROACHES FOR KOA

Surgical intervention is considered in KOA when conservative treatments fail to relieve symptoms or when joint damage leads to significant disability. Current surgical options include arthroscopy, osteotomies, cartilage restoration techniques, unicompartmental knee arthroplasty (UKA), and total knee arthroplasty (TKA), each offering distinct indications and long-term outcomes.

Total knee arthroplasty (TKA)

TKA is regarded as the gold standard surgical intervention for end-stage KOA, offering substantial pain relief, improved function, and enhanced quality of life, having the lowest revision rates among available surgical processes. Despite this, a significant proportion of patients (approximately 20%) experienced persistent postoperative pain (PPP), which is often associated with preoperative or early postoperative pain patterns and specific qualitative pain descriptors (e.g., cramping pain is predictive of PPP at 3 to 6 months) (Koga et al., 2024; Madry, 2022).

Arthroscopy

Arthroscopy, the commonly used technique for lavage, debridement, or partial meniscectomy, is now considered obsolete for degenerative KOA due to limited long-term efficacy. Therefore, it is primarily reserved for individuals with mechanical symptoms such as joint locking or loose bodies (Petterson et al., 2024).

High tibial osteotomy (HTO)

HTO and other similar techniques are aimed at realigning the mechanical axis of the knee to offload the affected compartment. These processes may delay the need for arthroplasty, offering noticeable functional recovery with reduced complication rates. However, their long-term effectiveness relative to arthroplasty remains under investigation (Bin et al., 2023; Peng et al., 2021).

Unicompartmental knee arthroplasty (UKA)

UKA is usually applied for patients with intact ligaments and an isolated compartment, signifying rapid recovery, good knee mobility, and higher bone preservation rates than TKA. A multicentre study of robotic-assisted medial UKA showed a 10-year survival rate of 91.7% and over 90% patient satisfaction (Bayoumi et al., 2023). Furthermore, a systematic review of more than 30,000 UKA procedures reported an 8.8% revision rate, with a mean time to revision of 6.5 ± 2.6 years (Migliorini et al., 2024). Though UKA yields satisfactory outcomes, its long-term durability is often inferior to that of TKA (Bin et al., 2023).

EMERGING AND ADJUNCT SURGICAL PROCEDURES

Genicular artery embolization (GAE)

GAE is an advanced minimally invasive procedure that aims at reducing synovial inflammation and pain. A meta-analysis demonstrated a notable reduction in Visual Analogue Scale (VAS) pain scores (up to 40 points), within 12 months following the treatment, along with high technical success rates. However, there is still a lack of long-term studies (Chlorogiannis et al., 2024).

Knee joint distraction (KJD)

The knee joint distraction process demonstrated clinical improvement in pain and function (WOMAC scores), and comparable results to controlled clinical trials. Despite this, KJD use has been limited by a high complication rate, particularly pin tract infections, reported in up to 70% of cases. Despite this, one-year outcomes suggest that KJD may be feasible and effective in routine care settings (Jansen et al., 2020; Struik et al., 2023).

Cartilage restoration techniques

Autologous Chondrocyte Implantation (ACI), Matrix-Assisted ACI (MACI), and Osteochondral Autograft/Allograft Transplantation (OAT/OCA) processes are applied to find localized cartilage defects and could be advantageous for KOA patients. Studies reported notable enhancement in clinical scores, including WOMAC, IKDC, and Lysholm, with reported failure or revision rates ranging from 8% to 10% during mid to long-term follow-up (Colombini et al., 2023; Nassar et al., 2025).

Conclusion and future directions

Knee osteoarthritis is a chronic and pathologically multifactorial disease driven by mechanical stress, ageing, inflammation and metabolic alterations. Although conventional treatments alleviate pain and improve functions, they fall short of addressing the underlying degenerative mechanism. Advances in molecular biology, regenerative medicine, MSCs, natural compounds, and gene-based therapies offer exciting potential for expanding the therapeutic horizon, shifting from symptom control to disease modification (Fig. 4). The integration of phenotype-driven therapies, advanced imaging modalities (e.g., MRI and radiography), and machine learning are enabling the development of more individualized treatment plans tailored to the unique disease profiles of patients. In addition, innovative interventions such as synovial embolization and thermal nerve ablation have shown early promise and may play a role in future therapeutic paradigms. Therefore, to ensure equitable access to these emerging treatments, future research should incorporate health economic evaluations, reimbursement policy frameworks, and strategies to promote fair access. These considerations are crucial for facilitating the widespread and ethical adoption of novel therapies across diverse patient populations.

Fig. 4.

Spectrum of therapies for knee osteoarthritis management. Spectrum of Knee osteoarthritis therapy illustrated across different stages. The X-axis depicted the disease stages, signifying early to late stages. The Y axis pointed to the therapeutic goals with symptoms modifying to disease-modifying states. The figure highlights the transition of knee osteoarthritis therapies from conservative to regenerative and surgical approaches.