The mediating role of synovitis in meniscus pathology and knee osteoarthritis radiographic progression

Meniscus pathologies (damage, extrusion) and synovitis are associated with knee osteoarthritis (KOA); however, whether synovitis mediates the relationship between meniscus pathologies and KOA radiographic progression remains unclear. We conducted an observational study in the Osteoarthritis Initiative (OAI) cohort, with a 48-month follow-up. Meniscus pathology and synovitis were measured by MRI osteoarthritis knee score (MOAKS) at baseline and 24 months, and a comprehensive synovitis score was calculated using effusion and Hoffa synovitis scores. The knee osteoarthritis radiographic progression was considered that Kellgren–Lawrence (KL) grade and joint space narrowing (JSN) grade at 48 months were increased compared to those at baseline. This study included a total of 589 participants, with KL grades mainly being KL1 (26.5%), KL2 (34.1%), and KL3 (30.2%) at baseline, while JSN grades were mostly 0 at baseline. A logistic regression model was used to analyze the relationship between meniscus pathology, synovitis, and KOA progression. Mediation analysis was used to evaluate the mediation effect of synovitis. The average age of the participants was 61 years old, 62% of which were female. The medial meniscus extrusion was longitudinally correlated with the progression of KL (odds ratio [OR]: 2.271, 95% confidence interval [CI]: 1.412–3.694) and medial JSN (OR: 3.211, 95% CI: 2.040–5.054). Additionally, the longitudinal correlation between medial meniscus damage and progression of KOA (OR: 1.853, 95% CI: 1.177–2.941) and medial JSN (OR: 1.655, 95% CI: 1.053–2.602) was significant. Synovitis was found to mediate the relationship between medial meniscus extrusion and KL and medial JSN progression at baseline (β: 0.029, 95% CI: 0.010–0.053; β: 0.022, 95% CI: 0.005–0.046) and beyond 24 months (β: 0.039, 95% CI: 0.016–0.068; β: 0.047, 95% CI: 0.020–0.078). However, we did not find evidence of synovitis mediating the relationship between meniscal damage and KOA progression. Synovitis mediates the relationship between medial meniscus extrusion (rather than meniscus damage) and KOA progression.


Database and participants
The data were obtained from the Osteoarthritis Initiative (OAI) cohort (https:// nda.nih.gov/ oai), a multicenter observational cohort study initiated by the National Institutes of Health.Briefly, the study included clinical assessment data, X-ray data, MRI data, and a biospecimen repository of 4796 participants aged 45-79 years at baseline, whose study principles and general inclusion criteria have been previously described 16 .The OAI conducts one follow-up per year on participants, with a follow-up rate of more than 90% in the initial 48 months.The study was approved by the local institutional review board, all research was performed in accordance with relevant guidelines/regulations and with the Declaration of Helsinki, and all selected participants have provided informed consent (see https:// oai.epi-ucsf.org) 17 .
The demographic data at baseline included the sex, age, race, and body mass index (BMI) of the participants.A participant's knee injury history was defined as whether their knee had been seriously injured to limit their walking ability for at least 1 week previously.The KL grade for the knee and JSN score were evaluated from fixed flexion weight-bearing X-ray, and the radiographs were read centrally at Boston University 18 .Four clinical sites acquired MRIs of the knee using a 3 T MRI system (Trio; Siemens Healthcare).The meniscus extrusion was determined with Dual Echo Steady State (DESS) and two-dimensional weighted turbo spin-echo (2D-TSE) with sagittal slices.Hoffa synovitis was identified with axial fat-suppressed (FS) turbo spin-echo (TSE) images in both sagittal and coronal planes, while effusion synovitis was determined with axial multiplanar reformations (MPR) of the DESS sequence; these images were assessed through the semi-quantitative MRI osteoarthritis knee scoring system (MOAKS).In the absence of knowledge of the clinical characteristics of all participants, MRI images were read sequentially by musculoskeletal radiologists with extensive experience 19 .The readings from the OAI database and the reading projects IDs are 22 and 65.Project 22 visited 600 participants at baseline, 12 months, and 24 months.Prior to transferring images to Boston Imaging Core Lab (BICL) and the start of taking measurements, the MR images were blinded to the OAI Release ID.The images were assessed paired and with known chronological order under the supervision of Dr. Ali Guermazi.Cartilage morphology, bone marrow lesions (BMLs), osteophytes, meniscal damage, synovitis and effusion, and extra articular features such as cysts and bursitis were scored 20 .Project 65 visited 1033 participants at baseline, 12 months, 24 months, 36 months, and 48 months.This project was performed under the supervision of Dr. Kent Kwoh from the Arthritis Research Center at the University of Arizona (previously at the University of Pittsburgh Medical Center-one of the OAI Clinical Centers).Images were selected for reading by Dr. Kwoh and his team and were prepared and sent to

Osteoarthritis radiographic progression
The radiographic follow-up data of the participants after 48 months were limited; therefore, we only observed the KOA radiographic progression within 48 months.The KL grades of radiographic KOA severity (0 = no osteoarthritis; 1 = suspicious joint space narrowing; 2 = definite osteophytes and possible joint space narrowing; 3 = multiple osteophytes, definite joint space narrowing, and mild sclerosis; 4 = large osteophytes, obvious joint space narrowing, severe sclerosis, and deformity) and the extent of medial joint space narrowing (0-3) were scored according to the Osteoarthritis Research Society International (OARSI) atlas 33,34 .The reliability of these radiographic readings was good, with intra-observer consistency of 71% (95% CI: 0.55-0.87) 34.An increase of ≥ 1 grade in KL grade from baseline to 48 months was considered progression for radiographic KOA, while any increase in the JSN score of the medial or lateral joint from baseline to 48 months was considered a progression of joint space narrowing 35 .

Statistical analysis
Demographic characteristics are represented by means, standard deviations (SDs), percentages, median, and interquartile range (IQR).The relationship between meniscal pathology/synovitis and KL/JSN progression was examined through binary logistic regression analysis.The outcomes of this analysis are presented as odds ratios (ORs) accompanied by a 95% confidence interval (CI).
The "bruceR" package (version 0.8.10) was used to analyze the mediation model.Model 4 of "bruceR" package was selected to calculate the natural indirect effect, natural direct effect, and total effect, with adjustments for age, race, BMI, varus alignment, respective compartment meniscal posterior root tear, knee injury, knee surgery and/or sex, baseline KL grade, meniscus damage, and meniscal extrusion; this simulation model was repeated for 1000 bootstraps.In this model, the total effect represents the total influence of meniscal pathology on KL/JSN progression.Furthermore, the natural indirect effect represents the population mean of causal mediating, while the natural direct effect represents the population mean of direct effects.The proportion of meniscal pathology attributed to synovitis as an effector of KOA radiographic progression was estimated by dividing the natural indirect effect by the total effect.All data analysis was realized by R software (version 4.2.1, R Foundation for Statistical Computing, Vienna, Austria).

Ethics declarations
The OAI study was approved by the institutional review boards at each OAI clinical site and the coordinating center (University of California, San Francisco) and participants provided written informed consent.This study obtained approval from the local Galician Ethics Committee (Comité Autonómico de Ética da Investigación de Galicia) with registry code 2018/129, All research was performed in accordance with relevant guidelines/ regulations and with the Declaration of Helsinki.

Participants' characteristics
The demographic characteristics of the participants in this study are shown in Table 1.The baseline mean age (SD) of all of the participants was 61.25 ± 8.7 years, among whom, 62.5% were female and 83.0% were White.The mean BMI (SD) at baseline was 29.4 ± 4.5 kg/m 2 , with 45.2% reporting a previous knee injury and 23.6% reporting a history of knee surgery.Among all participants, 26.8% had varus alignment and 2.9% had meniscal posterior root tear.A total of 47.5% of the participants had medial meniscus damage, while 14.6% had lateral meniscus damage.The proportion of KL grades 1 (26.5%), 2 (34.1%), and 3 (30.2%)at baseline were similar across all participants, and 33.1% of participants had an increase in knee KL grade from baseline to 48 months.The medial and lateral JSN mean scores (SD) at baseline were 0.87 (0.83) and 0.07 (0.32), respectively, and 34.1% of the participants had medial or lateral progression of joint space narrowing at 48 months.A total of 207 participants had a medial meniscal extrusion, and 24 participants had a lateral meniscal extrusion.Additionally, a total of 433 participants had synovitis at baseline, and the mean synovitis summary score at 24 months increased by 0.136 for all participants.A total of 152 participants had both medial meniscal extrusion and medial meniscal damage, while 17 participants had both lateral meniscal extrusion and lateral meniscal damage.Interestingly, regardless of medial or lateral sides, participants with both meniscal damage and meniscal extrusion had higher meniscus morphology scores on average compared to those with meniscal damage but without meniscal extrusion.

Direct effect
In Table 2, after multivariate adjustments (sex, age, race, BMI, varus alignment, respective compartment meniscal posterior root tear, knee injury history, knee surgery history, and KL grade), baseline medial meniscus extrusion was longitudinally correlated with KL progression at 48 months (OR: 2.504, 95% CI: 1.568, 4.046), and also with medial JSN progression at 48 months (OR: 3.336, 95% CI: 2.129, 5.228).There was no association between baseline lateral meniscus extrusion and KL progression at 48 months, but there was an association with lateral JSN progression at 48

Indirect effect
Given that both medial meniscus pathology and synovitis are associated with KL and JSN progression in the knee joint, we hypothesize that synovitis mediates the relationship between meniscus pathology and KOA radiographic progression at 48 months.In the mediation analysis, we included exposure factors (meniscus extrusion or damage), media (synovitis), results (KL progress or JSN progress), and potential confounding factors in the causal model for analysis (Fig. 2).Based on the mediation analysis, the impact of medial meniscus extrusion on knee KL progression was mediated by synovitis at baseline (0.029, 95% CI: 0.011, 0.054) and 24 months (0.040, 95% CI: 0.018, 0.069), respectively.Following further adjustment for meniscal damage (tear or maceration) in model 2, the aforementioned mediation ratio exhibited an increase.Specifically, the mediation ratio of synovitis at baseline was found to be 18.2% (Table 3), while at 24 months it rose to 24.5% (Table 4).However, we found no such mediating effect of synovitis at baseline or at 24 months on the association between meniscus damage and KL grade progression.
We also examined whether synovitis mediated the impact of meniscus extrusion on joint space narrowing progression.The results of the mediation analysis revealed a similar mediating effect.The effect of medial  www.nature.com/scientificreports/meniscus extrusion on the progression of medial joint space narrowing (JSN) at baseline (0.022, 95% CI: 0.005, 0.045) and 24 months (0.048, 95% CI: 0.021, 0.080) was found to be mediated by synovitis.This mediation ratio increased after accounting for meniscal damage (tear or maceration) in model 2, with mediation ratios of 9.1% (Table 5) and 19.4% (Table 6).Synovitis at baseline or 24 months did not mediate other association between meniscus damage and JSN progression.However, synovitis at 24 months mediated the association between concurrent meniscal damage and extrusion at baseline and KL progression (0.031, 95% CI: 0.009, 0.060) and medial JSN progression (0.038, 95% CI: 0.010, 0.072) at 48 months (Supplementary Table 1).

Discussion
In this large longitudinal observational study, we found that after adjusting for potential confounders, baseline medial meniscus pathology (extrusion and damage) was associated with KOA radiological progression over 4 years.Additionally, mediation analysis showed that synovitis at baseline mediated approximately one-fifth of the cases of medial meniscus extrusion and KOA radiological progression, while synovitis at 2 years mediated approximately one-quarter.These mediating models suggest that medial meniscus extrusion (instead of damage) affects synovitis and thus accelerates the progress of KOA.The association between meniscus extrusion and KOA radiographic progression has been reported in the previous literature 7,36 .Our study adds to this evidence to prove the longitudinal association between medial meniscus extrusion and KOA progression.A previous study revealed that the narrowing of the initial joint space on conventional X-ray films was secondary to meniscus extrusion rather than thinning of articular cartilage 37 .Indeed, medial meniscus extrusion often occurs before cartilage injury, and changes in meniscus position account for a significant portion of minimum joint space width (mJSW) changes 7,38,39 .It is worth noting that the thickness of articular cartilage is also significantly correlated with the measurement of the joint space width of the medial lesion cavity 40 .Moreover, the pathological state of meniscus damage, including tear and maceration, has been reported as a progressive factor of KOA by several studies 41,42 .Our study further demonstrates that the medial meniscus is a strong risk factor for the progression of KOA.Therefore, the pathological status of the medial meniscus constitutes a public health problem in the aging population, and efforts need to be strengthened to better understand its etiology, prevention, and treatment.
As far as we know, this is the first study to examine synovitis as a mediator of meniscus pathology and KOA radiological progression.The value of mediation analysis in observational research has been recognized for its ability to quantitatively assess potential mechanisms 43 .Our results also have potential clinical significance.Some Table 3. Analysis of the association between meniscus pathology and Kellgren-Lawrence progression mediated by the total synovitis score at baseline.Significant values are in bold.a Adjusted for sex, age, race, body mass index, varus alignment, respective compartment meniscal posterior root tear, injury, surgery and Kellgren and Lawrence (KL) grade at baseline.b Meniscal extrusion additionally adjusted for meniscal damage (tear or maceration), and meniscal damage additionally adjusted for meniscal extrusion.www.nature.com/scientificreports/previous studies have focused on the causal relationship between meniscus extrusion and cartilage or joint space, while others have focused on the potential initial role of meniscus damage in the occurrence and progression of KOA 21 ; however, it remains unclear whether the inflammatory phenotype mediates the correlation between the meniscus pathology and outcome of KOA.Our results suggest that the progression of KOA caused by medial meniscus extrusion (but not meniscal damage) is partly mediated by synovitis.There is increasing evidence that KOA radiology progression are closely related to the inflammatory phenotype 44,45 .Macrophages in synovial tissue are activated by related molecular patterns (DAMPs) such as cartilage fragments and aggrecans, secreting a large number of inflammatory factors and increasing the secretion of matrix metalloproteinases, thereby promoting the development of the inflammatory microenvironment and osteoarthritis 46 .Interestingly, several previous studies have evaluated the sagittal and coronal MR images of the knee joint and found that the extruded medial meniscus was related to intra-articular effusion 47,48 .Grainger et al. conducted a prospective study on 43 subjects and used gadolinium-enhanced MRI to evaluate synovitis.As a result, they found an association between medial meniscus extrusion and the severity of synovitis 13 .However, the small sample size represents a serious limitation of these conclusions.A recent study attempted to determine the molecules and pathways involved in meniscussynovium interactions through co-culture experiments.The findings indicated that inflammatory molecules generated by the synovium and meniscus have the potential to initiate inflammatory signals in individuals with early osteoarthritis (OA), leading to the degradation of extracellular matrix during the pathological progression and advanced stages of OA 14 .
In terms of the potential mechanism, the change in knee joint load caused by meniscus extrusion may be the initial driving factor of the disease 5 .Atkinson, H.F et al. suggested that patients with KOA may exhibit mechanical inflammation and that the low load on the medial side of the knee may be positively associated with reduction in knee inflammation after high tibial osteotomy 49 .Compared to meniscus tearing, the reduction in meniscus coverage and height due to meniscus extrusion may lead to greater changes in tibial movement 50 .Meniscus extrusion of the knee joint can be regarded as a functional meniscus defect.Medial meniscus extrusion may increase the mechanical stress of articular cartilage and subchondral bone, thereby expediting the progression of synovitis.Conversely, synovitis may facilitate the deterioration of articular cartilage and subchondral bone by generating pro-inflammatory cytokines and matrix metalloproteinases 46 , ultimately contributing to the radiological advancement of knee osteoarthritis.Antony B et al. pointed out in their study that meniscal extrusion, rather than meniscal signal or tear, is associated with changes in knee joint structure, possibly due to the less destructive nature of these pathologies, while severe meniscal damage (maceration) is associated with knee joint pain and structural changes 51 .In our study, participants with severe meniscal damage at baseline were few, Table 4. Analysis of the association between meniscus pathology and Kellgren-Lawrence progression mediated by the total synovitis score over 24 months.Significant values are in bold.a Adjusted for sex, age, race, body mass index, varus alignment, respective compartment meniscal posterior root tear, injury, surgery and Kellgren and Lawrence (KL) grade at baseline.b Meniscal extrusion additionally adjusted for meniscal damage (tear or maceration), and meniscal damage additionally adjusted for meniscal extrusion.www.nature.com/scientificreports/but still suggested an association between meniscal damage and radiographic progression of KOA.Our further investigation found that synovitis did not mediate this association.However, we cannot overlook the intraarticular inflammation caused by meniscal injury.Previous studies have found that synovitis frequently occurs in patients with traumatic meniscal injury who undergo arthroscopic meniscectomy and have no radiographic evidence of OA, and it is associated with increased pain and functional impairment 52 .Therefore, in some severe states of meniscal damage, especially traumatic meniscal damage caused by sports, we cannot ignore the role of synovitis in the progression of KOA, and more research is needed in the future to clarify this.This study raises several key points that warrant attention.We conducted mediation analysis under the guidance of the AGReMA (Mediation Analysis Reporting Guidelines) statement to evaluate the causal pathway of the longitudinal mediation of synovitis in meniscus pathology and KOA radiology progression 43 .Additionally, the current study had advantages, in that this was a follow-up study based on a large-scale population, with a population design based on an OAI cohort.Nevertheless, this study still has several potential limitations.First, we selected the more serious knee joints in OAI for analysis, which may produce selection bias in the analysis results.However, to minimize the selection bias, we included as many samples as possible.We look forward to further research with larger sample size and more time points.Second, the synovitis and meniscus pathological scores of MOAKS were semi quantitative.However, the correlation between synovitis and histology assessed by MRI is relatively accurate and powerful 53 .Third, we used the comprehensive score of three positions of the meniscus as the score of two departments.The pathological state of the meniscus may affect different regions, although this requires further research to confirm.Finally, as this was an observational study, even if some potential confounding factors are adjusted as much as possible, there may be residual confounding effects.

Table 1 .
The baseline demographic characteristics of the participants in the study.BMI body mass index, SD standard deviation, IQR interquartile range.

Table 2 .
Associations of meniscus pathology and synovitis score with knee osteoarthritis radiographic progression.Significant values are in bold.aAdjusted for sex, age, race, body mass index, varus alignment, respective compartment meniscal posterior root tear, injury, surgery and Kellgren and Lawrence (KL) grade at baseline.bMeniscal extrusion additionally adjusted for meniscal damage (tear or maceration), and meniscal damage additionally adjusted for meniscal extrusion.cNo adjusted for respective compartment meniscal posterior root tear.Figure 2. The mediation of synovitis in the relationship between medial meniscus extrusion and radiographic progression of knee osteoarthritis.Adjusted for sex, age, race, body mass index, varus alignment, respective compartment meniscal posterior root tear, meniscal damage (tear or maceration), injury, surgery, and Kellgren and Lawrence (KL) grade at baseline.Vol.:(0123456789) Scientific Reports | (2024) 14:12335 | https://doi.org/10.1038/s41598-024-63291-6

Table 5 .
Analysis of the association between meniscus pathology and joint space narrowing (JSN) progression mediated by the total synovitis score at baseline.Significant values are in bold.a Adjusted for sex, age, race, body mass index, varus alignment, respective compartment meniscal posterior root tear, injury, surgery and Kellgren and Lawrence (KL) grade at baseline.b Meniscal extrusion additionally adjusted for meniscal damage (tear or maceration), and meniscal damage additionally adjusted for meniscal extrusion.

Table 6 .
Analysis of the association between meniscus pathology and joint space narrowing (JSN) progression mediated by the total synovitis score over 24 months.Significant values are in bold.a Adjusted for sex, age, race, body mass index, varus alignment, respective compartment meniscal posterior root tear, injury, surgery and Kellgren and Lawrence (KL) grade at baseline.b Meniscal extrusion additionally adjusted for meniscal damage (tear or maceration), and meniscal damage additionally adjusted for meniscal extrusion.