Synovial Fluid Proteomic Biomarkers in Knee Osteoarthritis: A Systematic Review and Gene Ontology Analysis

Abstract

Background

Osteoarthritis (OA) of the knee is a degenerative disorder characterized by cartilage degradation, synovial inflammation, and structural remodeling. Synovial fluid (SF) biomarkers may improve diagnosis, staging, and patient stratification.

Methods

Following PRISMA guidelines, a systematic search of PubMed and Scopus (2015-2025) was conducted. Human studies analyzing SF with proteomic techniques (LC-MS/MS, SWATH-MS, ELISA) were included. Extracted data were classified by OA stage, sample type, proteomic platform, and identified biomarkers. Functional enrichment was performed with ShinyGO, and protein-protein interaction (PPI) analysis with STRING (v12.0).

Results

Seven studies met the inclusion criteria. Reported biomarkers included Cathepsin G (CTSG), angiotensinogen (AGT), periostin (POSTN), haptoglobin (HP), complement components, and matrix-related proteins such as ADAMTS4 and LYVE-1, which are involved in extracellular matrix remodeling, inflammation, and joint tissue homeostasis. Functional annotation revealed enrichment in glycosaminoglycan binding, complement cascades, and redox pathways. PPI analysis identified central nodes including COL1A1, ACAN, COMP, and POSTN, together with matrix-degrading enzymes such as MMP1, MMP3, and MMP13, highlighting tightly connected extracellular matrix remodeling processes in OA.

Conclusion

This review highlights reproducible SF biomarkers with diagnostic and prognostic potential in knee OA. Integrated proteomic and network analysis reinforces the multifactorial nature of OA and suggests key molecular targets. Translationally, a consistent biomarker signature could support early detection, more precise staging, and personalized management, enabling biomarker-guided diagnostics, targeted therapies, and the integration of precision medicine into OA care.

Keywords

biomarkers cartilage degradation inflammation osteoarthritis proteomics synovial fluid

Introduction

Knee osteoarthritis (OA) is a leading cause of pain and disability in aging populations, affecting over 250 million people globally and representing a substantial public health and economic burden.^1
-3 OA is a chronic, progressive joint disease characterized by the degeneration of articular cartilage, subchondral bone remodeling, synovial inflammation, and osteophyte formation, ultimately resulting in joint dysfunction, reduced mobility, and psychological stress.^4,5 Despite its high prevalence, OA remains challenging to diagnose in its early stages and to monitor longitudinally, primarily due to the lack of reliable and sensitive molecular biomarkers.⁶ Currently, OA diagnosis relies predominantly on clinical symptoms and radiographic findings, which typically reflect advanced joint damage and do not capture the molecular changes occurring in early disease.⁷ Thus, there is a critical need to identify biomarkers that can detect early pathological changes, differentiate disease stages, and potentially guide personalized therapeutic strategies.⁸ The synovial fluid (SF), which bathes the intra-articular structures, represents a dynamic source of molecular information and provides a window into the local joint microenvironment.⁹ As such, SF is an ideal biofluid for the discovery of OA-specific protein biomarkers reflecting ongoing pathophysiological processes. Advances in proteomic technologies, including mass spectrometry (MS)-based approaches, antibody arrays, and aptamer-based assays (SOMAscan), have facilitated the large-scale, high-throughput identification and quantification of proteins in SF.^10,11 These methods revealed numerous proteins involved in key OA mechanisms such as inflammation, extracellular matrix (ECM) degradation, lipid metabolism, angiogenesis, and oxidative stress.^12
-15 However, reproducibility across studies remains limited, often due to differences in sample preparation, patient stratification, analytical platforms, and statistical thresholds.¹⁶

To address these limitations, we conducted a systematic review of proteomic studies analyzing human synovial fluid in patients with knee osteoarthritis. While previous systematic reviews have summarized candidate biomarkers, few studies have attempted to integrate these findings through a systems biology framework. Therefore, beyond summarizing reported biomarkers, the present review incorporates bioinformatic analyses, including Gene Ontology (GO) enrichment and protein-protein interaction (PPI) network analysis to explore the functional relationships among repeatedly reported proteins. By integrating evidence across studies with bioinformatic interpretation, this work aims to identify reproducible molecular patterns and highlight biologically relevant pathways potentially involved in OA progression.

Materials and Methods

Search Strategy

According to the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines,¹⁷ a comprehensive and systematic search was conducted using the electronic databases: Scopus and PubMed. The search was limited to studies published over the last 10 years (2015-2025), and was designed to identify original research articles evaluating proteomic biomarkers in SF samples from patients with knee OA. Boolean logic and controlled vocabulary were applied to combine key terms related to “osteoarthritis,” “synovial fluid,” “serum,” and “proteomic techniques.” Only studies in English, conducted on humans, and applying proteomic analysis methods such as liquid chromatography-tandem mass spectrometry (LC-MS/MS), sequential window acquisition of all theoretical mass spectra (SWATH-MS), data-independent acquisition (DIA), and enzyme-linked immunosorbent assay (ELISA) were included. The detailed search strategies are provided in Suppl. “Appendix A” (Suppl. File 1).

Inclusion Criteria and Study Selection

The research question guiding this review was formulated according to the PICO model as follows¹⁸: (1) Population: Adult patients with a clinical and/or radiological diagnosis of knee OA. (2) Intervention: Proteomic analysis of SF samples using methods such as mass spectrometry (LC-MS/MS, SWATH-MS, DIA), two-dimensional difference gel electrophoresis (2D-DIGE), protein arrays, or ELISA (for biomarker validation). (3) Comparison: all studies were included irrespective of the presence or absence of comparator or control groups. (4) Outcome: Identification of proteomic biomarkers specific to OA, including both quantitative and qualitative data on differentially expressed proteins. Based on this framework, the inclusion criteria were defined as follows: Original studies involving human subjects with clinical and/or radiological diagnosis of knee OA; proteomic analysis specifically conducted on synovial fluid samples; use of established proteomic techniques such as mass spectrometry, 2D-DIGE, protein arrays, or ELISA.

Studies were excluded if they met any of the following criteria: Animal or in vitro studies; Studies analyzing tissue or blood samples only, without synovial fluid data; studies using only genetic or epigenetic approaches (other omics analysis); review articles, editorials, letters, technical notes, expert opinions, or case reports; studies with fewer than 10 participants. The selection process was performed in two stages. In the first phase, three reviewers independently screened the titles and abstracts of retrieved articles. In the second phase, potentially eligible full-text articles were evaluated in detail.¹⁹ Discrepancies were resolved through discussion or by consulting a third reviewer.²⁰ The number of studies included and excluded at each stage is presented in the PRISMA flow diagram ( Fig. 1 ).

Figure 1.

PRISMA 2020 flow diagram of study selection from PubMed and Scopus databases. PRISMA 2020 flow diagram illustrating the study selection process, including identification, screening, eligibility assessment, and final inclusion of studies.

Data Extraction and Quality Assessment

The selection and data extraction of the included studies were performed independently by three reviewers (G.L.B., M.M., and E.I.P.). For each study, the following information was collected: study title, first author, year of publication, and source of publication. Key characteristics of the study design and population were recorded, including the study type (cross-sectional, cohort, case-control), number of participants, and relevant demographic or clinical features. Detailed information on the type of biological samples analyzed (synovial fluid, plasma, or tissue specimens), the number of samples included, and the characteristics of the subjects was extracted. Proteomic techniques applied in each study such as mass spectrometry (SWATH-MS, nano-LC-MS/MS, LC-MS/MS), aptamer-based SOMA scan-assays, or ELISA were noted. Where applicable, comparator groups (healthy controls, different OA stages, clinical responders vs non-responders) were identified, along with the biomarkers reported and their expression profiles. Each study was classified according to its methodological design. All included studies were observational and non-randomized. The methodological quality of the observational studies was assessed using the Newcastle-Ottawa Scale (NOS), which evaluates selection of study groups, comparability, and outcome ascertainment. NOS scores range from 0 to 9 stars. Of the seven studies included, three were classified as moderate risk of bias, three as low risk of bias, and one as high risk of bias.²¹ A detailed summary of the risk of bias assessment using NOS is presented in Table 1 .

Table 1.

Study Quality Scores and Risk of Bias Based on the Newcastle-Ottawa Scale (NOS).

Study (author)	Selection (max 4)	Comparability (max 2)	Outcome (max 3)	Total score (/9)	Risk of bias
Kaur et al.²²	3	1	2	6	Moderate
Rydén et al.¹⁰	3	1	2	6	Moderate
Ali et al.²³	3	1	2	6	Moderate
Hulme et al.²¹	4	2	3	9	Low
Brophy et al.²⁴	4	1	3	8	Low
Liao et al.²⁵	3	1	3	7	Low
Liao et al.²⁶	2	0	2	4	High

The methodological quality of the observational studies was assessed using the Newcastle-Ottawa Scale (NOS).

Data Analysis

Given the methodological heterogeneity among the included studies particularly in terms of biological sample types, proteomic techniques, and outcome parameters a qualitative synthesis of findings was conducted. Biomarkers were grouped and compared according to their biological functions, study design, and sample type (synovial fluid and serum). To identify consistent molecular patterns and shared pathways, biomarkers reported in multiple studies were examined in parallel. To explore the biological relevance of the identified biomarkers, a functional enrichment analysis was performed using ShinyGO v0.76 (http://bioinformatics.sdstate.edu/go/). This tool enabled categorization based on GO terms including biological processes, molecular functions, and cellular components and mapped biomarkers to pathways using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Enrichment was considered significant at a false discovery rate (FDR) adjusted P-value of < 0.05 (FDR < 0.05) (Suppl. File 2). In parallel, a dataset was compiled to enable standardized comparison of biomarker expression trends across studies. For each biomarker, a discrete categorical value was assigned: +1 for upregulation, −1 for downregulation, and 0 for no significant change. Based on this dataset, a heatmap was manually generated using GraphPad Prism (version 8.0.1; GraphPad Software, Inc.) to support the visual synthesis of biomarker expression patterns across the included studies. Finally, PPI networks were constructed using the STRING database (version 12.0), enabling visualization of molecular interactions among biomarkers and highlighting potential clusters or central regulatory nodes (Suppl. File 3).

Results

Overview of Included Studies

A total of seven studies met the inclusion criteria and were included in this systematic review. The included studies encompassed a range of methodological designs, including analytical cross-sectional study,^10,22,23 prospective observational cohort study,²¹ descriptive cross- sectional study,²⁴ and a case-control study.^25,26 The majority of studies focused on proteomic analysis of synovial fluid, while others examined serum. Proteomic methodologies varied considerably across the included studies ( Table 2 ). Mass spectrometry-based approaches represented the most commonly used analytical strategy and were applied in six of the seven studies, including SWATH-MS (two studies), LC-MS/MS or nano-LC-MS/MS (two studies), and other MS/MS-based proteomic platforms (two studies). An aptamer-based proteomic platform (SOMAscan) was employed in one study, while multiplex proximity extension assays (OLINK) were used in combination with mass spectrometry in another study. In addition, enzyme-linked immunosorbent assay (ELISA) was frequently used as a validation technique for selected biomarkers in four studies. Sample sizes ranged from 10 to 75 participants, often stratified by disease stage, treatment response, or comparison with healthy controls. A summary of the characteristics of the included studies, including study design, sample size, clinical context, proteomic methods, sample type, identified biomarkers, and main findings, is provided in Table 2 .

Table 2.

Summary of Included Studies Evaluating Proteomic Biomarkers in Knee OA.

Study	Type of this study	Characteristics of the subjects	Sample	Proteomic technique	Comparison	Identified biomarkers
Kaur et al.²²	Analytical cross-sectional study non-randomized	Total patients: 75 (KL Grade 2, 3, 4; 25 in each group)→ Sex (F/M): 51/24 Age range (years): 50-70 BMI range (kg/m²): 24.6-27.9 Disease progression: Increasing trend in age and BMI with OA severity, but no observed bias Knee function (WOMAC score): Significantly worsened with disease severity	SF, serum	SWATH-MS, ELISA	Early-stage OA (G2) vs Advanced OA (G3/G4)	CTSG↑, AGT↑ (SF)/↓ (serum), FAH↓, NCAM1↓, GC ↓
Rydén et al.¹⁰	Analytical cross-sectional study non-randomized	Group 1: Mild degeneration Origin: Deceased donors Condition: Visual signs of mild degeneration of cartilage and/or menisci Mean age: 71.4 years (SD 7.0) Number: 13 (7 men, 6 women) Assessment: Macroscopic evaluation of cartilage and menisci by two independent observers; consensus reached in case of discrepancies; Group 2: Healthy controls Origin: Deceased donors Condition: No visual signs of tibiofemoral osteoarthritis and no known clinical diagnosis; cartilage and menisci smooth and intact (required for inclusion) Mean age: 70.7 years (SD 11.4) Number: 12 (7 men, 5 women) Assessment: Same macroscopic evaluation as Group 1; Group 3: Advanced stage osteoarthritis Origin: Living patients undergoing total knee replacement Condition: Clinically diagnosed advanced medial osteoarthritis Mean age: 71.1 years (SD 5.6) Number: 14 (8 men, 6 women) Assessment: Tissues collected during surgery; assessment method not specified	SF	SOMAscan assay, MS, WB	mild vs HC; only in mild group; only in HC	PADI4↑, MMP1↑, SOCS3↓, HIBCH, PHF3, SLCO5A1 (HC marker)
Ali et al.²³	Cross-sectional comparative/exploratory study randomized	Group 1: Late-stage OA synovial fluid Origin: Patients undergoing total knee arthroplasty Condition: End-stage medial compartment knee osteoarthritis Number: 11 (3 men, 8 women) Age range: 55-80 years Notes: Synovial fluid collected during arthroplasty; referred to as “late-stage OA synovial”; Group 2: Early-stage OA synovial fluid Origin: Patients undergoing knee arthroscopy Condition: Degenerative meniscal tear Number: 7 (3 men, 4 women) Age range: 50-64 years Notes: Synovial fluid collected during arthroscopy; referred to as “early-stage OA synovial”; Group 3: Healthy controls (synovial fluid) Origin: Human deceased donors Condition: No evidence of tibiofemoral OA or clinical diagnosis of knee OA Number: 13 (5 men, 8 women) Age range: 19-79 years Notes: Control synovial fluid samples from donors without joint pathology	SF	MS/MS, OLINK	3 groups: comparison late-stage OA vs HC or in the early-stage OA vs HC	GPX3↑, HRG↑ (early e late), CXCL7↓, PGS2↓
Hulme et al.²¹	Prospective observational cohort study non-randomized	Group: Early-stage OA surgical patients Origin: Patients undergoing surgery (microfracture [n = 19] or osteotomy [n = 13]) Condition: Early-stage osteoarthritis Selection: Based on clinical outcome (excellent or poor response) Age: microfracture = 34 (22) [17-67]; osteotomy 46(9) [31-58]; Sex: only males included in the study Smoking status: Includes both smokers and non-smokers	SF, plasma	LC-MS/MS, ELISA	Clinical responders vs non-responders to microfracture or osteotomy Multivariable analysis linking preoperative biomarkers to Lysholm score at 12 months post-surgery	ADAMTS4↓, YWHAQ↑, LYVE-1↑
Brophy et al.²⁴	Case-control study non-randomized	Group 1: Acute ACL injury Origin: Patients with anterior cruciate ligament (ACL) tear Condition: Sampled within 30 days post-injury Phase: Immediate post-traumatic inflammatory phase Notes: Represents early response after ACL rupture Group 2: Subacute ACL injury Origin: Patients with ACL tear Condition: Sampled between 31 and 90 days post-injury Phase: Subacute phase; may reflect early degeneration or sustained inflammation. Group 3: End-stage osteoarthritis (OA) Origin: Patients undergoing total knee arthroplasty (TKA) Condition: Radiographically confirmed advanced-stage OA Phase: Represents advanced joint degeneration Notes: No healthy controls were included in this study	SF, OA cartilage, and isolated cells	nano-LC-MS/MS, WB, multiplex ELISA, qPCR, immunostaining	The comparisons were designed to: Map molecular changes over time after ACL injury Differentiate injury-induced protein responses from those associated with chronic OA. Identify biomarkers (like POSTN) that may be useful in tracking joint status and predicting the risk of post-traumatic OA development	POSTN↑
Liao et al.²⁶	Case-control study non-randomized	Group 1: Osteoarthritis (OA) – Study group Origin: Chinese PLA General Hospital (September 2014-January 2015) Number: 10 patients (4 males, 6 females) Age range: 64-72 years (mean: 67 years) Diagnosis: Knee osteoarthritis (ACR criteria) KL grade: All Kellgren-Lawrence grade 4 (end-stage OA) Treatment: Total knee arthroplasty Group 2: Rheumatoid arthritis (RA) – Internal control group Origin: Chinese PLA General Hospital (September 2014-January 2015) Number: 10 patients (4 males, 6 females) Age range: 45-69 years (mean: 60 years) Diagnosis: Rheumatoid arthritis (ACR criteria) KL grade: All grade 4 (severe joint damage) Treatment: Total knee arthroplasty Group 3: Traumatic arthritis – Standard control group Origin: Chinese PLA General Hospital (September 2014-January 2015) Number: 10 patients (7 males, 3 females) Age range: 23-47 years (mean: 29 years) Condition: Meniscal injury without cartilage damage Timing: Surgery within 7 days post-injury Treatment: Knee arthroscopy	SF was collected during surgery.	SWATH-MS, ELISA	OA vs Traumatic arthritis: identifies OA-specific biomarkers OA vs RA: differentiates OA-specific changes from autoimmune pathology RA vs Traumatic arthritis: serves as internal control	Complement C1r↑, DKK2↑in OA
Liao et al.²⁶	Descriptive cross-sectional study non-randomized	OA Group (Primary): N = 10 patients (3 men, 7 women) Age: 52-70 years (mean: 60 years) Diagnosis: Knee OA confirmed Cartilage damage graded using Outerbridge classification: 6 grade IIIB, 4 grade IIB Samples collected preoperatively during arthroscopy or arthroplasty Control Group: N = 10 patients (6 men, 4 women) Age: 13-49 years (mean: 27 years) Condition: Meniscal injury (e.g., discoid meniscus), no cartilage damage Considered non-OA controls Validation Cohort: 24 additional SF samples (18 OA, 6 controls) used for ELISA-based validation of haptoglobin	SF	2-DE, MS/MS, ELISA	OA vs HC	Haptoglobin↑ with severity of OA

Characteristics of the studies included in this systematic review, including study design, participant characteristics, biological sample type, proteomic techniques, and identified biomarkers.

Abbreviations: SF = Synovial Fluid; WB = Western Blot; MS = Mass Spectrometry; LC-MS/MS = Liquid Chromatography Tandem Mass Spectrometry; SWATH-MS =Sequential Window Acquisition of all Theoretical Mass Spectra; 2-DE = Two-Dimensional Gel Electrophoresis; ELISA = Enzyme-Linked Immunosorbent Assay; OLINK = Multiplex Proximity Extension Assay Platform; KL Grade = Kellgren-Lawrence Grade; HC = Healthy Controls; RA = Rheumatoid Arthritis; OA = Osteoarthritis.

Identified Biomarkers

Across the included studies, a broad range of proteomic biomarkers were identified in patients with knee OA, many of which were quantitatively assessed. These biomarkers are primarily involved in inflammatory signaling, ECM degradation, oxidative stress, and cartilage or synovial tissue remodeling.^27
-29 These proteins are mainly involved in extracellular matrix remodeling, inflammatory signaling, oxidative stress regulation, and cartilage or synovial tissue turnover. Examples include Cathepsin G (CTSG), angiotensinogen (AGT), periostin (POSTN), haptoglobin (HP), complement-related proteins, and matrix-associated molecules such as ADAMTS4 and LYVE-1.^28,30 Quantitative comparisons between disease stages (early vs late OA), or between OA patients and healthy or disease controls, revealed distinct expression profiles across the included studies ( Tables 2 and 3 ). For example, Kaur et al.²² reported significantly increased levels of Cathepsin G (CTSG) and Angiotensinogen (AGT) in advanced OA compared with early-stage disease, whereas Ali et al.²³ identified elevated levels of GPX3 and HRG in both early and late OA synovial fluid compared with controls. In addition, Hulme et al.²¹ reported differential expression of biomarkers such as ADAMTS4, YWHAQ, and LYVE-1 in association with clinical outcomes following cartilage repair procedures.^22,31,32 Notably, some biomarkers AGT, demonstrated divergent patterns between local (SF) and systemic (serum) compartments. Table 3 provides a comparative summary of upregulated and downregulated biomarkers across the included studies, along with concentration values, fold changes, and diagnostic metrics when available. To further facilitate visual comparison of biomarker expression trends across studies, we generated a heatmap based on the data in Table 3 ( Fig. 2 ).

Table 3.

Comparative Quantitative Expression of Proteomic Biomarkers in OA.

Study	Comparison	Upregulated biomarkers	Quantification (upregulated)	Downregulated biomarkers	Quantification (downregulated)
Kaur et al.²²	OA Grade 2 vs Grade 4	Cathepsin G (CTSG), Angiotensinogen (AGT)	CTSG: 240.5 → 423.6 ng/mL (AUC = 94.00%)AGT (SF): 13.95 → 19.10 µg/mL (AUC = 72.50%)	Fumarylacetoacetase (FAH), NCAM1, Vitamin D Binding Protein (GC)	FAH: 4.019 → 1.861 ng/mL (AUC = 75.13%)NCAM1 (SF): 231.3 → 139.8 ng/mL (AUC = 84.09%)GC: 170.8→ 233.8 ng/ml
Rydén et al.¹⁰	Mild degeneration vs Healthy Control	PADI4, BPI protein	PADI4: log₂FC = +1.34 BPI: log₂FC = +1.01	TCN2, IGFBP2, SOCS3	TCN2: log₂FC = −1.48; IGFBP2: log₂FC= −0.98; SOCS3: log₂FC= 0.084
Ali et al.²³	Early and Late OA vs Controls; Late vs Early OA	Fibronectin, Hemoglobin α-subunit, CRAC1, HRG, GPX3	Fibronectin: FC = 1.99; Hb-α: FC = 2.25; CRAC1: FC = 1.96; HRG: Early = 7.03, Late = 7.79; GPX3: Early = 8.85, Late = 5.91	PGS2, CXCL7	PGS2: Early = 0.05, Late = 0.03; CXCL7: Early = 0.091, Late = 0.061
Hulme et al.²¹	Responders vs Non-responders to surgery	YWHAQ, LYVE-1, ADAMTS4, COMP, HA, MMP-1, MMP-3, CD14, S100A13	Biomarkers associated with favorable outcome: low ADAMTS4, high YWHAQ and LYVE-1; demographic predictors: younger age, female, non-smoker	—	—
Brophy et al.²⁴	Subacute ACL Injury vs OA vs Acute ACL Injury	Periostin (POSTN)	POSTN: Subacute ACL = 1096.02 ± 76.13 ng/mL; OA = 488.17 ± 55.70 ng/ml; Acute ACL = 533.43 ± 107.21 (mean)	—	—
Liao et al.²⁵	OA vs Traumatic Arthritis; OA vs RA	Complement C1r, DKK2,	C1r: OA = 2.063.1 ± 1.238.1 ng/ml; RA = 2.726.4 ± 1.627.8 ng/ml; Traumatic = 180.6 ± 102.4 ng/ml; DKK2: OA = 186.3 ± 41.0 ng/ml; RA= 93.0 ± 34.5 ng/ml;, Traumatic= 74.2 ± 22.9 ng/ml	Glutathione S-transferase, Hyaluronan and Proteoglycan link protein	Glutathione S-transferase FC (OA) = 2.051; FC (RA) = 0.527; Hyaluronan and proteoglycan link protein FC (OA) = 1.636; FC (RA) = 0.212
Liao et al.²⁶	OA (grades IIB-IV) vs Controls	Haptoglobin (Hp)	Non-OA = 2.54 ± 0.79 μg/mL; OA IIB = 10.85 ± 2.74μg/ml; IIIB = 25.20 ± 8.50μg/ml; IV = 33.15 ± 9.17μg/ml	—	—

Comparative expression of proteomic biomarkers reported across the included studies. Quantitative data are presented when available. Abbreviations: OA = Osteoarthritis; RA = Rheumatoid Arthritis; ACL = Anterior Cruciate Ligament; SF = Synovial Fluid; FC = Fold Change; log₂FC = Log₂ Fold Change; AUC = Area under the curve.

Figure 2.

Heatmap summarizing the regulation trends of selected synovial fluid biomarkers identified across the included proteomic studies in knee OA. Upregulated proteins are shown in green (+1), downregulated proteins in red (−1), and absence of reported differential expression in white (0). Biomarkers were selected from statistically significant findings reported in the original studies and include both recurrent candidates and representative proteins identified in individual studies.

To further identify biomarkers with higher reproducibility and translational potential, we conducted a cross-study comparison to determine which proteins were more frequently reported. This approach allowed us to filter out single-study findings and focus on a set of frequently reported biomarkers that may serve as more robust candidates for future validation. Across the included studies, several candidate biomarkers were identified, including CTSG, AGT, periostin (POSTN), haptoglobin (HP), complement-related proteins, LYVE-1, YWHAQ, and ADAMTS4. However, most biomarkers were reported in only one or two studies, reflecting the exploratory nature of current proteomic research in knee OA. These molecules were associated with key OA-related processes, including inflammation,²⁷ cartilage breakdown,²⁹ oxidative defense,³³ and intracellular signaling.²⁸ Cross-study comparison revealed a limited overlap of identified biomarkers across the included studies. Only a small number of proteins were reported in more than one study, and in most cases these biomarkers were identified in only two studies. These markers are summarized in Table 4 .

Table 4.

Biomarkers Identified in More Than One Included Study.

Biomarker	Supporting studies	Sample type	Proteomic method
MMP1	Rydén et al.¹⁰; Hulme et al.²¹	SF	SOMAscan; LC-MS/MS
LYVE-1	Hulme et al.²¹; Ali et al.²³	SF	LC-MS/MS
POSTN	Ali et al.²³; Brophy et al.²⁴	SF	LC-MS/MS

Biomarkers identified in more than one included study, indicating the supporting studies, sample type, and proteomic methodologies used for their detection.

Abbreviations: OA = Osteoarthritis; SF = Synovial Fluid; MMP1 = Matrix Metalloproteinase 1; LYVE-1 = Lymphatic Vessel Endothelial Hyaluronan Receptor 1; POSTN = Periostin.

Functional Enrichment Analysis of Frequently Reported Biomarkers

To elucidate the functional landscape of the most consistently reported proteomic biomarkers in knee OA, a gene enrichment analysis was conducted using ShinyGO v0.76. The analysis included proteins reported in at least two of the included studies and encompassed three GO domains: Biological Processes (BP), Molecular Functions (MF), and Cellular Components (CC) as well as pathway-level annotation via KEGG. Within the BP category, the most enriched terms (FDR <0.05) included cartilage development and extracellular matrix organization. The most prominent processes included cartilage development involved in endochondral bone morphogenesis, extracellular matrix organization, collagen fibril organization, and skeletal system morphogenesis. These processes reflect the fundamental role of ECM remodeling and cartilage structural changes in OA pathophysiology. Key proteins contributing to these terms included structural and matrix-associated proteins such as ACAN, COL1A1, DCN, POSTN, COMP, and CILP, as well as matrix-degrading enzymes including MMP1, MMP3, MMP13, and ADAMTS4, which are well known to participate in cartilage degradation and remodeling during disease progression ( Fig. 3A ). These pathways suggest that OA pathogenesis is not only characterized by inflammatory and degradative events but also involves compensatory redox mechanisms, possibly related to synovial oxidative damage and cartilage cell stress.^5,29 The CC analysis indicated that the majority of enriched proteins were localized within extracellular compartments ( Fig. 3B ). The most significant categories included the extracellular matrix, collagen-containing extracellular matrix, external encapsulating structure, and extracellular space. Additional enrichment was observed for collagen-related structures such as fibrillar collagen trimer and banded collagen fibril, further highlighting the strong association between the identified proteins and cartilage structural components. Proteins such as COL1A1, DCN, COMP, POSTN, and ACAN were prominently represented in these categories, supporting their established roles in cartilage matrix architecture, integrity and inflammation.^13,27 The MF enrichment analysis revealed a predominance of binding and structural activities related to extracellular matrix interactions ( Fig. 3C ). Significant terms included proteoglycan binding, glycosaminoglycan binding, and extracellular matrix structural constituent. These functions are primarily associated with structural ECM proteins such as COMP, ACAN, DCN, and POSTN, which contribute to cartilage resilience and matrix stability. In addition, enzymatic activities related to proteolysis such as metalloendopeptidase activity and serine-type endopeptidase activity were enriched and corresponded to proteases including MMP1, MMP3, MMP13, and ADAMTS4, emphasizing the contribution of proteolytic enzymes to ECM degradation in OA.^34,35 At the pathway level, KEGG enrichment revealed multiple relevant signaling cascades. Among these, the ECM-receptor interaction and focal adhesion pathways were prominently enriched, reflecting altered cell-matrix communication in osteoarthritic cartilage. Additionally, inflammatory and immune-related pathways such as the IL-17 signaling pathway and AGE-RAGE signaling pathway in diabetic complications were represented, suggesting the involvement of inflammatory signaling and metabolic dysregulation in OA progression. Other enriched pathways included rheumatoid arthritis and phagosome, indicating the participation of innate immune mechanisms within the joint environment. Collectively, these findings highlight that the identified OA-related proteins are largely involved in extracellular matrix organization, proteolytic remodeling, and inflammatory signaling processes ( Fig. 3D ).^36
-38

Figure 3.

Functional enrichment analysis of frequently reported knee osteoarthritis biomarkers. (A-C) Functional enrichment analysis of frequently reported osteoarthritis biomarkers. Gene Ontology (GO) enrichment includes (A) biological processes, (B) cellular components, and (C) molecular functions. (D) KEGG pathway analysis highlighting enriched pathways related to inflammation, oxidative stress, and extracellular matrix remodeling.

Together, these results delineate a biologically coherent framework in which OA-related proteins are predominantly secreted, matrix-interacting molecules with key roles in immune signaling, oxidative defense, and cartilage remodeling.

Network-Based Visualization of OA-Related Biomarkers

To further investigate the molecular interactions among the identified OA-related proteins, a PPI network was generated using the STRING database (version 12.0). The resulting network ( Fig. 4 ) revealed a densely connected cluster centered on extracellular matrix components and proteolytic enzymes, illustrating the coordinated regulation of cartilage remodeling processes. Core nodes within the network included COL1A1, ACAN, DCN, COMP, and POSTN, which formed a central ECM-related cluster connected to matrix-degrading enzymes such as MMP1, MMP3, MMP13, and ADAMTS3. These interactions highlight the dynamic balance between structural ECM proteins and proteases responsible for cartilage breakdown in OA. Several additional proteins were integrated into this core network, including IGF1, LYVE1, and NCAM1, suggesting roles in cell signaling, tissue remodeling, and cell-matrix communication. Inflammatory mediators such as IL6 and CD14 were also connected to the network, reflecting the contribution of immune signaling pathways to OA pathogenesis. Some proteins, including DKK2, PFN1, S100A13, and FAH, appeared as more peripheral nodes with fewer interactions, suggesting more specialized or context-dependent roles within the OA molecular landscape. Overall, the interaction map illustrates how ECM structural components, matrix-degrading enzymes, and immune mediators converge into an interconnected molecular network driving cartilage degeneration and joint remodeling in knee osteoarthritis.

Figure 4.

Protein-protein interaction (PPI) network of frequently reported knee osteoarthritis biomarkers. PPI network of frequently reported osteoarthritis biomarkers. Nodes represent proteins and edges indicate predicted interactions. Key hub proteins involved in extracellular matrix remodeling, inflammation, and immune regulation are highlighted.

Discussion

This systematic review synthesizes current evidence on proteomic biomarkers identified in the synovial fluid of patients with early and late knee OA. Across the seven included studies, we identified a heterogeneous set of candidate proteins. Among the biomarkers identified there were CTSG, AGT, periostin (POSTN), haptoglobin (HP), and matrix-related proteins such as ADAMTS4 and LYVE-1. These molecules were primarily detected in synovial fluid but in some cases also in serum, suggesting their potential relevance as both local and systemic disease indicators.^10,21
-26 Notably, biomarkers such as CTSG and AGT were upregulated in late-stage OA compared to early-stage disease, pointing to their association with disease progression. Several proteins involved in extracellular matrix remodeling, including MMP1, MMP3, MMP13, ADAMTS4, and COMP, were enriched in OA synovial fluid, highlighting the active degradation and remodeling of cartilage matrix during disease progression.^20,23,24 Similarly, ADAMTS4 and MMP-1, both matrix-degrading enzymes, highlight the ongoing ECM turnover in the OA joint microenvironment.^22,27 The consistent appearance of these proteins across multiple studies and platforms strengthens their candidacy as diagnostic or prognostic biomarkers. Functional enrichment analyses using GO and KEGG databases indicated that these biomarkers are primarily involved in extracellular matrix organization, proteolytic activity, and cell-matrix interactions, with enriched pathways including ECM-receptor interaction, IL-17 signaling, and AGE-RAGE signaling pathways.^6,10 The PPI network analysis further underscored the centrality of COL1A1, ACAN, COMP, and POSTN, together with matrix-degrading enzymes such as MMP1, MMP3, and MMP13.¹⁰ Our findings are consistent with previous systematic investigations of synovial fluid proteomics in OA. Several studies highlighted the diagnostic potential of these biomarkers. For instance, AGT and CTSG showed distinct expression profiles between early and late OA stages, with receiver operating characteristic (ROC) analyses supporting their discriminatory capacity.^21,22 This aligns with previous evidence that inflammatory and ECM-related biomarkers rise with disease severity, often reflecting irreversible structural changes in the joint.^39,40 In addition, oxidative markers such as GPX3 were associated with redox imbalance in the synovial fluid, a key contributor to chondrocyte apoptosis and matrix damage in OA.^6,27 From a clinical perspective, these findings may support the development of biomarker panels for non-invasive disease monitoring. Future platforms may integrate synovial biomarker levels with imaging to provide a composite disease activity index. For example, quantification of AGT, MMP-1, and IL-6 in synovial fluid or serum may enhance sensitivity in detecting early inflammatory OA subtypes or predicting response to therapies. Multiplex assays or aptamer-based technologies (SOMAscan) may also provide rapid, high-throughput screening suitable for clinical settings.^10,21,41 Despite these promising directions, the current body of evidence is limited by several factors. The included studies varied in sample type, analytical techniques (LC-MS/MS, SWATH-MS, ELISA), and patient stratification criteria. This methodological heterogeneity likely contributes to quantitative variability and restricts generalizability. Moreover, many studies had small sample sizes and lacked validation cohorts or longitudinal follow-up.^21,24,26 Importantly, very few biomarkers were assessed across multiple biological matrices, limiting insights into tissue-specific versus systemic expression profiles. Most studies were observational and exploratory, with limited power to draw causal inferences. From a translational standpoint, the identification of synovial fluid biomarkers related to extracellular matrix remodeling, including COMP, ACAN, COL1A1, and POSTN, together with inflammatory mediators such as CTSG and IL6, highlights promising candidates for the development of integrated diagnostic panels or point-of-care assays in OA. An additional limitation relates to the heterogeneity of the patient populations included across studies. Differences in disease stage, demographic characteristics, and clinical background may influence synovial fluid composition and biomarker expression profiles. Consequently, caution is required when interpreting cross-study comparisons, as some observed differences may reflect population-specific characteristics rather than universal OA-related molecular signatures. The biomarkers identified could serve as adjuncts to imaging or clinical assessment to enhance early detection and stratify disease stages. Furthermore, combining proteomic data with transcriptomic or metabolomic profiles and machine learning models could support the development of predictive tools for OA progression or treatment efficacy.^24,41,42 A relevant limitation of this review is the limited overlap of biomarkers across studies. Most proteins were identified in single studies, while only a small number of biomarkers were reported in more than one investigation. In most cases, these markers were identified in only two studies, which limits the strength of conclusions regarding their reproducibility. In the present review, biomarkers described as “common” were operationally defined as those reported in at least two independent studies. However, this definition should be interpreted with caution. Another limitation of this review is the relatively small number of eligible studies identified after applying strict inclusion criteria. While this approach ensured methodological consistency and focus on proteomic analyses of synovial fluid, it may limit the generalizability of the findings.

Future research should focus on standardizing sample collection and analytical workflows, validating biomarker panels in diverse and longitudinal cohorts, and exploring the integration of multi-omics data for a systems-level understanding of OA. These efforts are essential to bridge the gap between proteomic discovery and clinical application, ultimately advancing precision medicine in musculoskeletal disease.

Conclusion

This systematic review highlights the potential of proteomic analysis of synovial fluid to identify molecular biomarkers associated with the pathogenesis and progression of knee osteoarthritis. Several candidate biomarkers were reported, including Cathepsin G (CTSG), angiotensinogen (AGT), periostin (POSTN), haptoglobin (HP), and matrix-associated proteins such as ADAMTS4 and LYVE-1. These molecules are involved in key biological processes relevant to osteoarthritis, including inflammatory signaling, extracellular matrix remodeling, and joint tissue homeostasis. Proteomic profiling and network-based analyses provide valuable insights into the molecular pathways underlying disease development and progression. However, most biomarkers were reported in a limited number of studies, reflecting the exploratory nature of current research and highlighting the need for further validation. Although these findings highlight promising candidate biomarkers, the limited number of available studies and methodological heterogeneity currently limit their clinical translation. Future research should focus on standardized sampling and analytical protocols, larger multicenter cohorts, and integration with multi-omics and computational approaches. These efforts will be essential to validate candidate biomarkers and support the development of reliable, minimally invasive tools for early diagnosis, disease monitoring, and personalized management of knee OA. In addition, the application of artificial intelligence and machine learning algorithms may provide new opportunities for the analysis of complex proteomic datasets and for the identification of novel biomarker signatures associated with early osteoarthritis.

Supplemental Material

sj-docx-1-car-10.1177_19476035261446251 – Supplemental material for Synovial Fluid Proteomic Biomarkers in Knee Osteoarthritis: A Systematic Review and Gene Ontology Analysis

Supplemental material, sj-docx-1-car-10.1177_19476035261446251 for Synovial Fluid Proteomic Biomarkers in Knee Osteoarthritis: A Systematic Review and Gene Ontology Analysis by Elvira Immacolata Parrotta, Giorgia Lucia Benedetto, Giovanni Cuda, Raffaele Covello, Umile Giuseppe Longo, Arianna Carnevale, Olimpio Galasso, Giorgio Gasparini and Michele Mercurio in CARTILAGE

Supplemental Material

sj-ods-2-car-10.1177_19476035261446251 – Supplemental material for Synovial Fluid Proteomic Biomarkers in Knee Osteoarthritis: A Systematic Review and Gene Ontology Analysis

Supplemental material, sj-ods-2-car-10.1177_19476035261446251 for Synovial Fluid Proteomic Biomarkers in Knee Osteoarthritis: A Systematic Review and Gene Ontology Analysis by Elvira Immacolata Parrotta, Giorgia Lucia Benedetto, Giovanni Cuda, Raffaele Covello, Umile Giuseppe Longo, Arianna Carnevale, Olimpio Galasso, Giorgio Gasparini and Michele Mercurio in CARTILAGE

Supplemental Material

sj-tsv-3-car-10.1177_19476035261446251 – Supplemental material for Synovial Fluid Proteomic Biomarkers in Knee Osteoarthritis: A Systematic Review and Gene Ontology Analysis

Supplemental material, sj-tsv-3-car-10.1177_19476035261446251 for Synovial Fluid Proteomic Biomarkers in Knee Osteoarthritis: A Systematic Review and Gene Ontology Analysis by Elvira Immacolata Parrotta, Giorgia Lucia Benedetto, Giovanni Cuda, Raffaele Covello, Umile Giuseppe Longo, Arianna Carnevale, Olimpio Galasso, Giorgio Gasparini and Michele Mercurio in CARTILAGE

Footnotes

Acknowledgements

We thank the developers of the ShinyGO and STRING databases for their freely accessible analytical platforms, BioRender and GraphPad software.

ORCID iDs

Giorgia Lucia Benedetto

Michele Mercurio

Ethical Considerations

This article is a systematic review of previously published studies. Ethical approval was obtained from the original studies.

Informed Consent

This article is a systematic review of previously published studies. Participant consent was obtained from the original studies, as stated by the respective authors. No new data involving human participants were collected or analyzed by the authors of this review.

Consent for Publication

Not applicable. This review does not include identifiable individual data or images.

Author Contributions

E.I.P., G.L.B., and M.M. contributed equally to the conception and writing of the manuscript. G.C. and R.C. provided critical revisions and conceptual guidance. U.G.L. and A.C. contributed to the data analysis and interpretation. O.G., G.G., and M.M. supervised the project and provided final approval of the manuscript. All authors read and approved the final version of the manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the following funding: Funded by the European Union – Next Generation EU – NRRP M6C2 – Investment 2.1 Enhancement and strengthening of biomedical research in the NHS (Project no. PNRR-MAD-2022-12376080 – CUP: F83C22002450001).

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

All the data generated or analyzed during this study are included in this published article and its supplementary information files. Supplementary files include full enrichment results (GO, KEGG) and PPI network tables.

Declaration of Generative AI And AI-Assisted Technologies

During the preparation of this manuscript, the authors used ChatGPT to enhance language clarity and readability. Following this, the authors carefully reviewed and revised the content as necessary and took full responsibility for the accuracy and integrity of the publication.

Supplemental Material

Supplementary material for this article is available on the Cartilage website at .

References

Hunter

Bierma-Zeinstra

Osteoarthritis. Lancet. 2019;393:1745-59.

Carlson

Rawle

Wallace

, et al. Characterization of synovial fluid metabolomic phenotypes of cartilage morphological changes associated with osteoarthritis. Osteoarthritis Cartilage. 2019;27:1174-84.

Mercurio

Gasparini

Galasso

, et al. Lateral versus medial approach for total knee arthroplasty for valgus knee deformity shows comparable functional outcomes, hip–knee–ankle angle values, and complication rates: a meta-analysis of comparative studies. Arch Orthop Trauma Surg. 2023;144:869-78.

De Filippis

Mercurio

Segura-Garcia

, et al. Defining the minimum clinically important difference (MCID) in the hospital anxiety and depression scale (HADS) in patients undergoing total hip and knee arthroplasty. Orthop Traumatol Surg Res. 2024;110:103689.

Loeser

Goldring

Scanzello

, et al. Osteoarthritis: a disease of the joint as an organ. Arthritis Rheum. 2012;641697-707.

Mobasheri

Thudium

Bay-Jensen

A-C

, et al. Biomarkers for osteoarthritis: current status and future prospects. Best Pract Res Clin Rheumatol. 2023;37101852.

Guermazi

Hayashi

Roemer

, et al. Osteoarthritis: a review of strengths and weaknesses of different imaging options. Rheum Dis Clin North Am. 2013;39567-91.

Mercurio

Galasso

Familiari

, et al. Trend of perioperative CRP (C-reactive protein) levels in non-infected total knee arthroplasty. Orthop Rev. 2022;14:36589. doi:10.52965/001c.36589.

Kraus

Blanco

Englund

, et al. Call for standardized definitions of osteoarthritis and risk stratification for clinical trials and clinical use. Osteoarthritis Cartilage. 2015;23:1233-41.

10.

Rydén

Sjögren

Önnerfjord

, et al. Exploring the early molecular pathogenesis of osteoarthritis using differential network analysis of human synovial fluid. Mol Cell Proteomics. 2024;23:100785.

11.

Deng

Tang

Yin

, et al. Clinical and omics biomarkers in osteoarthritis diagnosis and treatment. J Orthop Transl. 2025;50:295-305.

12.

Berenbaum

Osteoarthritis as an inflammatory disease (osteoarthritis is not osteoarthrosis!). Osteoarthritis Cartilage. 2013;21:16-21.

13.

Balakrishnan

Bhattacharjee

Ahmad

, et al. Differential proteomic analysis of synovial fluid from rheumatoid arthritis and osteoarthritis patients. Clin Proteom. 2014;11:1.

14.

Huang

Chen

Zhang

, et al. Inflammatory mechanisms underlying metabolic syndrome-associated osteoarthritis and potential treatments. Osteoarthritis Cartilage Open. 2025;7100614.

15.

Maouche

Boumediene

Baugé

Bioactive compounds in osteoarthritis: molecular mechanisms and therapeutic roles. IJMS. 2024;25:11656.

16.

Mathiessen

Conaghan

PG.

Synovitis in osteoarthritis: current understanding with therapeutic implications. Arthritis Res Ther. 2017;19:18.

17.

Moher

Liberati

Tetzlaff

, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6:e1000097.

18.

Schardt

Adams

Owens

, et al. Utilization of the PICO framework to improve searching PubMed for clinical questions. BMC Med Inform Decis Mak. 2007;7:16.

19.

Familiari

Madonna

Mercurio

, et al. Outcomes and complications of inlay versus onlay patellofemoral arthroplasty: a systematic review. Knee. 2023;41:124-36.

20.

Mercurio

Cofano

Colace

, et al. Outcomes, complications, and eradication rates of two-stage revision surgery for periprosthetic shoulder, elbow, hip, and knee infections: a systematic review. Prosthesis. 2024;6:1240-58.

21.

Hulme

Peffers

Harrington

GMB

, et al. Identification of candidate synovial fluid biomarkers for the prediction of patient outcome after microfracture or osteotomy. Am J Sports Med. 2021;49:1512-23.

22.

Kaur

Rana

Sharma

, et al. Proteomic insights into early detection and progression of knee osteoarthritis: unveiling molecular signatures. Arch Med Res. 2025;56:103206.

23.

Ali

Turkiewicz

Hughes

, et al. Proteomics profiling of human synovial fluid suggests increased protein interplay in early-osteoarthritis (OA) that is lost in late-stage OA. Mol Cell Proteom. 2022;21100200.

24.

Brophy

Cai

Duan

, et al. Proteomic analysis of synovial fluid identifies periostin as a biomarker for anterior cruciate ligament injury. Osteoarthritis Cartilage. 2019;27:1778-89.

25.

Liao

, et al. Proteomic analysis of synovial fluid in osteoarthritis using SWATH-mass spectrometry. Mol Med Rep. 2018;17:2827-36. doi:10.3892/mmr.2017.8250.

26.

Liao

Zhang

, et al. Proteomic analysis of synovial fluid as an analytical tool to detect candidate biomarkers for knee osteoarthritis. Int J Clin Exp Pathol. 2015;8:9975-89.

27.

Kapoor

Martel-Pelletier

Lajeunesse

, et al. Role of proinflammatory cytokines in the pathophysiology of osteoarthritis. Nat Rev Rheumatol. 2011;7:33-42.

28.

Wojdasiewicz

Poniatowski

ŁA

Szukiewicz

The role of inflammatory and anti-inflammatory cytokines in the pathogenesis of osteoarthritis. Mediators Inflamm. 2014;2014:1-19.

29.

Troeberg

Nagase

Proteases involved in cartilage matrix degradation in osteoarthritis. Biochim Biophys Acta. 2012;18241133-45.

30.

Jiang

Lin

Zhao

, et al. ADAMTS5 in osteoarthritis: biological functions, regulatory network, and potential targeting therapies. Front Mol Biosci. 2021;8703110.

31.

, et al. Renin-angiotensin system in osteoarthritis: a new potential therapy. Int Immunopharmacol. 2019;75105796.

32.

Shao

Gao

Han

, et al. Genetic insights into serum cathepsins as diagnostic and therapeutic targets in knee and hip osteoarthritis. Sci Rep. 2024;1417553.

33.

Bertaggia

Scabia

Dalise

, et al. Haptoglobin is required to prevent oxidative stress and muscle atrophy. PLoS ONE. 2014;9(6):e100745.

34.

Posey

Coustry

Hecht

JT.

Cartilage oligomeric matrix protein: COMPopathies and beyond. Matrix Biol. 2018;71-72:161-73.

35.

Shamdani

Chantepie

Flageollet

, et al. Heparan sulfate functions are altered in the osteoarthritic cartilage. Arthritis Res Ther. 2020;22:283.

36.

Xiao

Zhang

Cai

, et al. IL-17 in osteoarthritis: a narrative review. Open Life Sci. 2023;18(1):20220747.

37.

Zhou

Zhang

Shi

, et al. Activation and modulation of the AGEs-RAGE axis: implications for inflammatory pathologies and therapeutic interventions–a review. Pharmacol Res. 2024;206:107282.

38.

Gao

Cathepsin G and its role in inflammation and autoimmune diseases. Arch Rheumatol. 2018;33:498-504.

39.

De Lange-Brokaar

BJE

Ioan-Facsinay

Van Osch

GJVM

, et al. Synovial inflammation, immune cells and their cytokines in osteoarthritis: a review. Osteoarthritis Cartilage. 2012;20:1484-99.

40.

Boffa

Merli

Andriolo

, et al. Synovial fluid biomarkers in knee osteoarthritis: a systematic review and quantitative evaluation using BIPEDs criteria. CARTILAGE. 2021;13:82S-103S.

41.

Moretti

Bizzoca

Geronimo

, et al. Towards precision medicine for osteoarthritis: focus on the synovial fluid proteome. IJMS. 2022;23:9731.

42.

Fang

Zhou

, et al. Current cutting-edge omics techniques on musculoskeletal tissues and diseases. Bone Res. 2025;13:59.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.02 MB

0.04 MB

0.00 MB