Sage Journals: Discover world-class research

Abstract

Background:

Osteoarthritis (OA) is associated with lost range of motion in the affected joint(s). Evidence suggests that this may be due to increased activity of posterior capsule fibroblasts, cells in turn derived from mesenchymal stromal cells (MSCs).

Objectives:

To test the hypotheses that (1) MSCs are more numerous in the posterior capsule of patients with knee flexion contracture (FC) and (2) in OA participants with knee FC, the MSC population in the posterior capsule differentiates toward a fibrotic phenotype. In order to complete these objectives, we looked for associations between capsule histologic and MSC outcomes with clinical outcomes.

Design:

Cross-sectional translational research design using data from the Ottawa Knee Osteoarthritis (OKOA) database.

Methods:

A total of 71 OKOA database participants and their relevant clinical and laboratory outcomes were included. Associations were first tested with bivariate correlation, then for p < 0.10, tested using a linear model.

Results:

No lab-based differences between FC and no-FC groups we discovered. In the posterior capsule, there was an association between knee flexion and adipogenic capacity (p = 0.001), osteogenic capacity (p < 0.001), KL grade and percent “other” (mainly neurovascular) tissue (p = 0.039), visual analog scale pain, and percent fibrous tissue (p = 0.014). For the anterior capsule, there was an association between knee flexion (p = 0.002) and extension (p = 0.005) with MSC enumeration, KL grade with MSC fibrogenic capacity (p = 0.002), and Knee Injury and Osteoarthritis Outcome Score quality of life with chondrogenic capacity (p < 0.001).

Conclusion:

Joint capsule composition, MSC enumeration, and function were associated with important clinical OA outcomes. These findings suggest that the entire joint capsule may play an important role in OA-related morbidity and progression and could represent an underappreciated target for OA treatment.

Keywords

contracture knee range of motion mesenchymal stromal cells osteoarthritis translational research

Introduction

Osteoarthritis (OA) affects more than 50% of adults over the age of 65,¹ with >1.5 million arthroplasties being performed in the UK annually for OA.² The burden of OA has, therefore, become an urgent health issue. OA is a total joint disease, known to affect cartilage, bone, as well as the synovium-lined joint capsule. In the OA-affected joint, the synovial portion of the joint capsule may proliferate and become inflamed, contributing the overall disease progression.³ Deep to the synovium, the composition of the capsule may vary depending on the location, with the posterior capsule being more fibrous (having a higher proportion of collagen I), while the anterior capsule has a higher composition of adipose tissue.^4,5 Though the role of the synovium in OA has been described,³ little is known regarding the role of the underlying capsule tissue in OA.

Many patients with OA develop a contracture, that is, a restriction in the passive range of motion (ROM) of affected joint(s).^6,7 Contractures restrict mobility, negatively impact quality of life (QoL), contribute to pain, and prevent basic activities of daily living.^4,5,8
–10 About 1/3 of people with knee OA will develop a knee flexion contracture (FC),⁶ which is a limitation in knee extension.^9,11 These individuals have worse pain, function, and joint stiffness as compared to those without knee FC and are more likely to progress to knee arthroplasty over a shorter time period.^12,13 Despite the high prevalence and detrimental effects of FC in knee OA, the underlying pathophysiology of joint contractures remains elusive. Cross-sectional MRI data suggest that the etiology may be multifactorial, including alterations in articular structures such as bone (osteophytes, bone marrow lesions), cartilage, and the presence of joint effusion.¹⁴ MRI, however, does not adequately visualize the posterior capsule,¹⁴ and its role in OA-associated FC remains unknown. A contribution from the posterior capsule to FC is supported indirectly by both animal and clinical studies. In a rat model of immobility-induced knee contracture, the posterior capsule contributed more than muscle to lost knee extension with capsular fibrosis being the likely underlying cause.⁷ Capsular shortening prevented restoration of knee extension after remobilization.¹⁵ In a rabbit study, knees that had been immobilized for 32 weeks contained up to five times more fibroblasts in the posterior capsule versus controls.¹⁶ Clinically, a shortening of the posterior joint capsule has been proposed as a contributor to knee FC.^17,18 Capsulotomy and lateral ligament division during total knee arthroplasty (TKA) improved knee ROM.^17,18 In patients with post-traumatic elbow FC, myofibroblasts and elevated mRNA synthesis for pro-fibrotic growth factors were reported in the joint capsule.^16,19 Full-genome microarray evaluating the gene expression in posterior knee capsule biopsies of OA patients showed upregulated cell and biological adhesion pathways consistent with an underlying activated connective tissue process, such as excess fibrous tissue production limiting knee extension.⁵

Mesenchymal stromal cells (MSCs) are precursors to mesenchymal-lineage cells, such as fibroblasts, that reside within cells (MSCs) are precursors to mesenchymal-lineage cells, such as fibroblasts, that reside within the joint capsule.^20
–22 In non-OA models, MSCs have been implicated in pulmonary, renal, and bone marrow fibrosis²¹ and were the primary contributor to muscle fibrosis in a mouse model of muscular dystrophy.²⁰ MSCs may also play a role in pathologic cutaneous scar formation, resulting in increased fibroblast proliferation and expression of the pro-fibrotic collagen genes.^20,23 MSCs may, therefore, contribute to pathologic tissue alterations in the joint capsule in OA, such as posterior capsule fibrosis²⁰; however, we are not aware of any study describing the phenotype of the MSC population in the knee capsule in OA, nor any evaluation of their potential correlation with OA-associated clinical outcomes, such as pain or function. Should MSCs in the posterior capsule be implicated in FC development, or should their numbers or function be associated with clinical outcomes, this could provide new avenues of research to target for prevention and treatment of OA-associated morbidity. Therefore, the objectives of this study were to test the hypotheses that (1) MSCs are more numerous in the posterior capsule of patients with knee FC, and that (2) in OA participants with knee FC, the MSC population in the posterior capsule differentiates toward a fibrotic phenotype. In order to complete these objectives, we looked for associations between capsule histologic and MSC outcomes with clinical outcomes using the Ottawa Knee Osteoarthritis (OKOA) database, a database that includes clinical as well as histological and MSC functional data in participants with severe knee OA.

Patients and methods

Participant selection and clinical evaluation

The OKOA is a primary OA database evaluating clinical and laboratory biomarkers of knee OA and ROM. OKOA participants were recruited consecutively from our outpatient orthopedic clinic at The Ottawa Hospital between August 2016 and March 2021. For inclusion, participants scheduled for arthroplasty were first screened for clinical trial NCT02861521, a trial evaluating leg length inequality correction post-arthroplasty in patients with bilateral pre-operative knee FC, and included in the OKOA if FC was present in only one knee or absent in both.²⁴ For this study, we included all 71 OKOA participants that had histologic data available. Exclusion criteria comprised history of inflammatory arthritis, previous knee surgery, metastatic cancer, or bone-affecting disorders. The OKOA was approved by the local institutional research ethics board (protocol 20140139-01H). All participants provided written consent. All subjects met the American College of Rheumatology criteria for knee OA.²⁵

Maximum knee extension was measured using a standardized protocol using a method with high inter-rater reliability.^4,9,11,26 The patient was placed in the supine position with the ankle on a foam roller, and participants’ knees were extended as fully as possible by the examiner. A goniometer was used to measure the knee extension angle with the fulcrum over the knee joint line, the upper arm directed toward the greater trochanter, and the lower arm directed toward the lateral malleolus. The inability to extend the knee to 5° constituted a knee FC.^4,6

Pain was evaluated using the visual analog scale (VAS).²⁷ Other clinical outcomes were evaluated using the Knee Injury and Osteoarthritis Outcome Score (KOOS) and included pain, symptoms, function, and QoL.²⁸

Tissue collection and cells

Posterior knee capsule tissue samples approximately 150–200 mm³ were obtained during the TKA procedure from the capsule just posterior to the mid-posterior aspect of the medial femoral condyle. Tissue size was chosen based on the maximum size that could be obtained without risking injury to the posterior neurovascular bundle. Anterior capsule was obtained from the capsular region, harvested anteriorly during the initial medial parapatellar arthrotomy incision. Bone samples were obtained from the routine distal femoral bone cuts for component fitting. Samples were collected immediately after removal during TKA and placed in low-glucose Dulbecco’s Modified Eagle’s Medium (DMEM; Gibco, Paisley, UK) for further laboratory processing.

Histology

For each participant, roughly half of the volume of each posterior and anterior tissue samples was used for histologic analysis. Samples were fixed in formalin and paraffinized. Paraffin blocks were cut into 5-μm-thick sections and stained with Masson trichrome. Digitized light microscope color images were analyzed for collagen cross-sectional whole, collagen, adipose, synovial, and “other” tissue areas using Photoshop (version CS6; Adobe, San Jose, California, USA) and ImageJ (version 1.53; US National Institutes of Health, Bethesda, MD, USA), as described previously.⁴ First, the entire sample was photographed digitally at 20× magnification; then, images were assembled using the Photoshop photomerge tool to reconstruct the total tissue sample (Supplemental Figure 1). The total area of the tissue sample was then calculated in pixels using ImageJ. Synovial tissue area was then identified on Masson trichrome-stained slides as a villous tissue containing synoviocytes with circular nuclei, selected using Photoshop’s Lasso tool. Once selected, the synovial tissue was segregated from the image and the area calculated. Collagenous tissue was then selected using the Photoshop color select tool to select blue tissue on Masson trichrome-stained slides. Erythematous structures (including blood vessels, nervous tissue, etc.) were labeled as “other.” The remaining white space within the area of the tissue sample was labeled as adipose.⁴ All areas were calculated using ImageJ. For all histologic analyses, the analyzer was blinded to patient identity and clinical information.

Tissue digestion, flow cytometry, and cell proliferation

Tissues were digested in collagenase, as described previously.²⁹ Briefly, capsule and bone samples were weighed, manually divided, then submerged in a minimum volume of StemMACS MSC expansion media (130-091-680; Miltenyi Biotec, Bisley, UK) containing 300 U/mL of collagenase (C0130-100MG; Sigma-Aldrich, Burlington, MA, USA) for digest. Tissues were digested at 37°C for 4 h. Following digest, the tissue was plated directly into six-well tissue culture plates containing StemMACS MSC expansion media. The remaining supernatant was then spun down, washed, and resuspended for cell counting. Cells from the supernatant were then plated at 4000–6000 cells/cm² in an appropriately sized flask, separate from the already-plated post-digest tissue. Post-digest tissue and cell flasks were left for culture expansion for 3 days. This timeframe was selected in order to allow cells within the plated tissue to migrate out and adhere to plastic. After 3 days, flasks were washed with PBS to remove any remaining tissue, then media replaced. At the time of the first passage, cells were counted in each of the tissue + supernatant digest flasks and summed to get the total tissue sample count. The cells from the tissue + supernatant digest flasks were then combined for further expansion.

Flow cytometry was performed for samples from a total of 33 consecutive OKOA participants (resulting in 19 for the FC group, 14 for the no-FC group). Cells were allowed to expand to passage 3 to allow for sufficient numbers for flow cytometry analysis. Following tissue culture expansion, cells were trypsinized, resuspended in PBS, and centrifuged for 5 min at 400g. The resulting pellet was resuspended in Fluorescence-activated Cell Sorting (FACS) buffer, and incubated in a 10% Fc receptor-blocking reagent solution (Miltenyi Biotec) before fluorophore-conjugated antibodies were added, as described previously.²⁹ Staining was performed for 7AAD, CD90 (Thy1-FITC; Serotec/Bio-rad, Neuried, Germany), CD73 (5′ Ecto-nucleotidase-PE; BD Biosciences, San Jose, CA, USA), CD45 (PE-Cy7; BD Biosciences), and CD31 (V450; BD Biosciences). MSCs were enumerated based on 7AAD⁻CD73⁺CD90⁺CD45⁻CD31⁻ in order to meet the International Society for Cellular Therapy minimal criteria for defining multipotent mesenchymal stromal cells.³⁰ All flow cytometry data were analyzed using FloJo software (version 10.2; BD Biosciences; Supplemental Figure 2).

Colony-forming unit-fibroblast assay

The colony-forming unit-fibroblast (CFU-F) assay was performed for the same 33 samples as used for flow cytometry, as described previously,²⁹ with some modification. Due to the small posterior capsule sample size (sized to reduce the risk of injuring the neurovascular bundle), insufficient tissue was available for specific allocation to CFU-F, which requires fixation and staining, thus precluding any further downstream experiments. For the present CFU-F protocol, cells from both the tissue and supernatant flasks for each participant (see above description) were allowed to expand for 10 days to allow the formation of visible clusters that could be counted under the microscope without staining (which could then be further expanded for downstream experiments). Clusters were counted with an overlying grid on the culture plate wells to enhance counting accuracy. Total CFU-F count was the sum of the CFU-F counts of the tissue and supernatant culture (prior to combining the cultures) and normalized to mass of the tissue from which the cells were extracted (CFU-F/g). To ensure the accuracy of this counting method, three trial samples were first counted without staining, then fixed, and stained as described previously.²⁹ Stained clusters were then counted and compared to unstained counting. There was <2% difference in the absolute cluster count between the two methods. To further ensure counting accuracy, clusters from 10 unstained samples were counted independently by two investigators, each blinded to the other’s results. Using our grid overlay method, there was an intra-class correlation coefficient inter-rater reliability of >0.9, indicating excellent agreement.

MSC differentiation

Differentiation was performed for 12 samples (n = 6 each for the FC and no-FC groups; Supplemental Figure 3). Passage 3 MSCs were induced toward osteogenesis, chondrogenesis, and adipogenesis using standard protocols²² and toward fibrogenesis, as described previously,³¹ with the cells allowed to adhere to a circular cover slide placed in the respective culture wells. Fibrogenic cultures were grown in DMEM with 10% FBS, antibiotics, L-ascorbic acid, and recombinant human CTGF (PHG0286; ThermoFisher Scientific, Ottawa, Canada). For osteogenesis and chondrogenesis, we used StemMACS OsteoDiff and ChondroDiff media, respectively (Miltenyi Biotec); adipogenic cultures were grown in DMEM with 10% FCs, antibiotics, 10% horse serum (Stem Cell Technologies, Vancouver, Canada), 0.5 mM isobutylmethylxanthine, 60 μM indomethacin, and 0.5 μM hydrocortisone (all from Sigma). Due to limited tissue sample sizes and available cells, assays were performed in duplicate for each sample, rather than triplicate.

Fibrogenic cultures were stained using Masson trichrome on day 28. Calcium deposits were stained using alizarin red on day 21 as previously described.²² Chondrogenic pellets were cut into 5-µm sections using the Leica CM1950 cryostat (Leica Biosystems, Nussloch, Germany), fixed and stained with toluidine blue. Adipogenic cultures were stained on day 21 post-induction with oil red O.²² Differentiation analysis was performed using a standardized method across each differentiated tissue. Differentiated cells from each lineage were assessed by taking standardized digital images of cover slips with adherent differentiated cells from each lineage using the Marlin F080C digital camera (Allied Vision Technologies, Exton, PA, USA) mounted on the photo tube of the microscope and magnified using a 2.5× lens, in addition to the objective lens. Camera software was AVT Smartview 1.5.1. A standardized gray box was then placed beside the standardized digital images as a fixed reference for all samples. The intensity of the differentiated cells relative to the standardized box was then calculated using the ImageJ gel staining intensity function.

Statistics

Sample size was estimated using preliminary data from the OKOA following the first year of recruitment as we were not aware of any literature upon which to otherwise base our calculation. For n = 10 in each group (data available after year 1), we used the mean #MSCs/g measured by flow cytometry to estimate our required sample size. With mean # MSCs/g of 50,000 for the FC group versus 8000 for the no FC group with a standard deviation of 80% (based on large inter-participant deviation), a sample size of 15 per group was estimated (n = 30 total) to show a statistical inter-group difference using an alpha of 0.05 and a beta of 0.8 for flow cytometry.

Differences in the FC versus no-FC participant demographic features were evaluated using the unpaired Student’s t test. Differences in histologic features, CFU-F, and flow cytometry counts between the two groups were evaluated using the Mann–Whitney U test. For our exploratory analysis, associations between histologic and MSC outcomes with clinical outcomes was first testing using bivariate correlation using the Spearman correlation coefficient for non-parametric data. Correlations with p < 0.10 were then input as the dependent variable into a linear model using generalized estimating equations (GEE), correcting for age, biological sex, and BMI covariates. Significance following GEE was set at p < 0.05. All tests were performed using IBM SPSS Statistics 21.

Results

Participants and demographics

Participant demographics are summarized in Table 1. Just over half (55%) of the population was female. Other than ranges of knee extension (reduced in the FC group vs the no-FC group; p < 0.001), and knee flexion (also reduced in the FC group vs the no-FC group; p = 0.002), there were no significant differences between the two groups.

Table 1.

Participant demographic information.

Demographic feature	All (n = 71)	Contracture (n = 38)	No contracture (n = 33)
Age (years)	68.4 ± 9.7	67.3 ± 9.9	70.0 ± 9.5
BMI (kg/m²)	32.9 ± 7.8	33.1 ± 7.7	32.7 ± 8.0
% female	55	50	61
Maximum knee extension (°)	6.0 ± 5.4	10 ± 3.9***	1.2 ± 1.6
Maximum knee flexion (°)	107.7 ± 19.0	101.5 ± 21.7**	114.9 ± 12.2
KL grade	3.4 ± 0.6	3.4 ± 0.6	3.4 ± 0.7
VAS pain (avg)	5.5 ± 2.3	5.0 ± 2.2	6.0 ± 2.3
KOOS score
Pain	42.8 ± 18.4	43.1 ± 18.5	42.4 ± 18.9
Symptoms	42.1 ± 20.0	41.1 ± 16.8	44.1 ± 25.5
Function	42.8 ± 19.6	44.9 ± 17.7	39.0 ± 22.9
QoL	25.8 ± 12.6	25.1 ± 10.8	27.4 ± 16.2

p < 0.01 contracture versus no contracture group. ***p < 0.001 contracture versus no contracture group.

BMI, body mass index; KL, Kellgren and Lawrence radiographic osteoarthritis severity grade; KOOS, Knee Injury and Osteoarthritis Outcome Score; QoL, quality of life; VAS, visual analog pain scale.

Histology

A summary of the histologic findings in the posterior and anterior capsule is shown in Table 2. Synovial CSA accounted for <3% of the capsule sample in both FC and no-FC groups. There were no significant differences in capsule tissue composition between the FC versus no-FC groups (Table 2).

Table 2.

Histologic features of posterior and anterior capsule tissue biopsies.

Tissue	All (n = 71)	Contracture (n = 38)	No contracture (n = 33)
Posterior capsule
% Fibrous	76.4 ± 23.9	73.6 ± 25.4	80.0 ± 21.9
% Adipose	15.8 ± 22.3	19.2 ± 24.1	11.8 ± 19.6
% Synovium	2.8 ± 3.4	2.8 ± 3.5	2.7 ± 3.4
% Other	3.8 ± 4.0	3.8 ± 3.7	3.8 ± 4.2
Anterior capsule
% Fibrous	54.7 ± 30.8	50.6 ± 31.0	59.3 ± 30.3
% Adipose	37.8 ± 31.3	40.1 ± 31.1	34.3 ± 31.6
% Synovium	1.8 ± 8.3	2.4 ± 11.2	1.1 ± 1.9
% Other	4.9 ± 5.2	5.5 ± 4.7	4.3 ± 5.8

MSC enumeration and functional assays

A summary of MSC enumeration and function is shown in Table 3. There were no significant differences between the FC versus no-FC group for any of the lab-based outcomes.

Table 3.

Joint capsule mesenchymal stromal cell enumeration and function.

Tissue	All (n = 33)	Contracture (n = 19)	No contracture (n = 14)
Posterior capsule
CFU-F	29,758.3 ± 54,484.3	39,439.5 ± 69,671.9	16,619.6 ± 16,221.1
Flow cytometry count	14,282.8 ± 38,062.0	18,331.4 ± 49,239.9	8788.3 ± 12,349.9
Differentiation capacity (relative intensity)
Fibrogenic	213,040.7 ± 38,944.6	201,204.8 ± 42,221.2	226,849.3 ± 32,801.7
Adipogenic	23,916.0 ± 20,593.1	16,622.4 ± 16,377.6	32,425 ± 23,120.3
Chondrogenic	339,929.4 ± 110,721.2	302,876.2 ± 138,641.5	384,393.3 ± 45,721.3
Osteogenic	247,605.5 ± 94,397.1	224,624.2 ± 76,267.2	279,779.3 ± 116,454.0
Anterior capsule
CFU-F	23,765.8 ± 75,219.8	15,961.5 ± 50,460.3	34,357.4 ± 100,955.5
Flow cytometry count	9657.6 ± 35,235.0	2961.5 ± 4978.2	18,745.3 ± 53,546.5
Differentiation capacity (relative intensity)
Fibrogenic	250,300.5 ± 47,674.5	245,053.6 ± 55,531.4	258,695.6 ± 35,761.6
Adipogenic	56,088.7 ± 45,432.2	59,177.8 ± 58,017.2	51,764.1 ± 24,231.7
Chondrogenic	349,129.1 ± 106,813.1	374,331.0 ± 95,700.9	318,886.9 ± 120,597.0
Osteogenic	294,426.0 ± 37,118.1	278,845.6 ± 27,475.0	316,238.4 ± 40,459.4

Differentiation units are relative intensity, as output by Image J software.

CFU-F, colony forming unit-fibroblast.

Exploratory analyses

Correlations between clinical outcomes and posterior capsule tissue histology and MSC enumeration and function are summarized in Table 4. There were associations between maximal knee extension and MSC osteogenic capacity, knee flexion and synovium CSA, MSC number by flow cytometry, adipogenic capacity, and osteogenic capacity. There were associations between KL grade and fibrous tissue CSA, “other” tissue CSA, and adipogenic capacity. VAS pain was associated with fibrous tissue CSA and KOOS symptoms with adipogenic capacity.

Table 4.

Correlations between clinical outcomes and posterior capsule histology or MSC outcomes.

Clinical outcome	Laboratory outcome
	Histology				Enumeration		Differentiation capacity
	% Fibrous	%Adipose	%Synovium	%Other	CFU-F	Flow cytometry	Fibrogenic	Adipogenic	Chondrogenic	Osteogenic
Clinical outcome
Knee extension (°)	−0.024	0.046	−0.025	0.004	−0.107	−0.217	−0.114	−0.269	−0.449	−0.531
	0.844	0.706	0.836	0.971	0.555	0.224	0.711	0.374	0.166	0.076
Knee flexion (°)	0.032	0.032	−0.236*	0.132	0.136	0.331	−0.301	0.557*	0.469	0.698*
	0.789	0.790	0.048	0.271	0.449	0.060	0.318	0.048	0.146	0.012
KL grade	−0.256*	0.285*	0.004	0.348**	0.077	0.045	0.259	−0.185	0.050	−0.024
	0.049	0.027	0.973	0.006	0.695	0.819	0.392	0.544	0.883	0.941
VAS pain	−0.294*	0.110	0.172	−0.006	0.001	0.016	−0.030	−0.217	−0.245	−0.161
	0.013	0.362	0.151	0.963	0.996	0.931	0.922	0.476	0.467	0.617
KOOS
Pain^α	0.115	−0.041	−0.044	0.077	0.023	0.209	0.311	−0.287	0.000	−0.144
	0.445	0.787	0.769	0.613	0.922	0.364	0.453	0.490	1.000	0.758
Symptoms^α	−0.042	0.091	−0.013	0.197	0.055	0.204	0.313	0.627	−0.024	−0.018
	0.781	0.546	0.931	0.189	0.812	0.375	0.450	0.096	0.955	0.969
Function	0.189	−0.151	0.090	−0.025	0.016	0.101	0.491	−0.527	−0.467	−0.321
	0.209	0.316	0.552	0.868	0.944	0.663	0.217	0.180	0.243	0.482
QoL^β	0.136	−0.009	−0.241	−0.131	−0.226	0.047	−0.561	0.412	0.206	−0.618
	0.450	0.960	0.176	0.467	0.457	0.878	0.190	0.359	0.658	0.191

Bold, p < 0.05. Italics, p < 0.10.

^αdata available for n=46.

^βdata available for n=33.

KL, Kellgren and Lawrence radiographic osteoarthritis severity grade; KOOS, Knee Injury and Osteoarthritis Outcome Score; VAS, visual analog pain scale.

Correlations between clinical outcomes and anterior capsule tissue composition and MSC outcomes are summarized in Table 5. There were associations between knee extension and MSC flow cytometry enumeration, knee flexion and MSC flow cytometry enumeration, KL grade and MSC fibrogenic capacity, and KOOS QoL with chondrogenic capacity.

Table 5.

Correlations between clinical and anterior capsule histology or MSC outcomes.

Clinical outcome	Laboratory outcome
	Histology				Enumeration		Differentiation capacity
	%Fibrous	%Adipose	%Synovium	%Other	CFU-F	Flow cytometry	Fibrogenic	Adipogenic	Chondrogenic	Osteogenic
Clinical outcome
Knee extension (°)	−0.127	0.056	0.084	0.104	−0.240	−0.307	0.091	0.049	0.170	−0.264
	0.290	0.645	0.486	0.387	0.178	0.082	0.766	0.879	0.618	0.407
Knee flexion (°)	0.043	0.027	0.065	−0.077	0.279	0.361*	0.091	0.049	0.170	−0.264
	0.722	0.825	0.592	0.524	0.116	0.039	0.766	0.879	0.618	0.407
KL grade	−0.035	0.077	−0.006	0.060	−0.114	−0.138	0.535	0.027	−0.280	0.039
	0.791	0.557	0.966	0.650	0.564	0.483	0.060	0.933	0.404	0.904
VAS Pain	0.186	−0.165	−0.252*	0.070	−0.002	−0.099	0.264	0.074	−0.396	0.088
	0.121	0.168	0.034	0.562	0.992	0.585	0.383	0.820	0.228	0.787
KOOS
Pain^α	−0.145	0.139	0.236	0.002	0.010	0.142	−0.216	−0.337	0.523	−0.325
	0.335	0.358	0.114	0.990	0.966	0.540	0.608	0.414	0.229	0.432
Symptoms^α	−0.162	0.163	0.240	0.034	0.129	0.227	0.241	0.602	0.364	0.491
	0.282	0.278	0.108	0.824	0.577	0.323	0.565	0.115	0.423	0.217
Function	−0.112	0.121	0.177	−0.196	0.039	0.179	−0.228	0.024	0.429	−0.238
	0.457	0.423	0.239	0.191	0.866	0.439	0.588	0.955	0.337	0.570
QoL^β	−0.136	0.168	−0.195	0.084	0.112	0.187	0.136	0.239	0.971**	−0.349
	0.451	0.349	0.277	0.641	0.716	0.540	0.748	0.606	0.001	0.443

Bold, p < 0.05. Italics, p < 0.10.

^αdata available for n=46.

^βdata available for n=33.

KL, Kellgren and Lawrence radiographic osteoarthritis severity grade; KOOS, Knee Injury and Osteoarthritis Outcome Score; MSC, mesenchymal stromal cell; QoL, quality of life; VAS, visual analog pain scale.

Following input into the GEE model (Table 6), knee flexion remained associated with posterior capsule MSC adipogenic and osteogenic activities. KL grade was associated with “other” tissue CSA, and VAS pain was associated with fibrous tissue CSA. For anterior capsule tissue, knee extension was associated with MSC enumeration by flow cytometry, knee flexion with flow cytometry MSC enumeration, KL grade with MSC fibrogenic capacity, and KOOS QoL with MSC chondrogenic capacity.

Table 6.

Correlation results of linear model—effect size of tissue and cellular outcomes on clinical outcomes.

Clinical and laboratory outcomes	Effect size and 95% CI	p-Value
Posterior capsule
Knee extension and osteogenic capacity	−1.6 × 10⁻⁵ [−3.5 × 10⁻⁵, 2.7 × 10⁻⁶]	0.094
Knee flexion and % synovium	−0.1 [−1.3, 1.0]	0.842
Knee flexion (°) and flow cytometry enumeration	7.8 × 10⁻⁵ [−4.5 × 10⁻⁵, 2.0 × 10⁻⁴]	0.212
Knee flexion (°) and adipogenic capacity	4.0 × 10⁻⁴ [1.74.5 × 10⁻⁴ , 6.2 × 10⁻⁴ ]	0.001
Knee flexion (°) and osteogenic capacity	1.1 × 10⁻⁴ [6.7.5 × 10⁻⁵ , 1.4 × 10⁻⁴ ]	<0.001
KL grade and % fibrous	−0.00 [−0.01, 0.00]	0.189
KL grade and % adipose	0.00 [−0.00, 0.01]	0.282
KL grade and % other	0.04 [0.00, 0.08]	0.039
VAS pain and % fibrous	−0.026 [−0.05,−0.01]	0.014
KOOS symptoms and adipogenic capacity	−2.4 × 10⁻⁵ [−2.0 × 10⁻⁴, 1.6 × 10⁻⁴]	0.779
Anterior capsule
Knee extension (°) and flow cytometry enumeration	−4.3 × 10⁻⁵ [−7.3 × 10⁻⁵, −1.3 × 10⁻⁵ ]	0.005
Knee flexion (°) and flow cytometry enumeration	1.5 × 10⁻⁴ [5.4 × 10⁻⁵ , 2.5 × 10⁻⁴ ]	0.002
KL grade and fibrogenic capacity	7.3 × 10⁻⁶ [2.6 × 10⁻⁶ , 1.2 × 10⁻⁵ ]	0.002
KOOS QoL and chondrogenic capacity	5.9 × 10⁻⁵ [2.6 × 10⁻⁵ , 9.2 × 10⁻⁵ ]	<0.001

Model corrected for age, biological sex, and BMI.

Bold, p < 0.05.

BMI, body mass index; CI, confidence interval; KL, Kellgren and Lawrence radiographic osteoarthritis severity grade, KOOS, Knee Injury and Osteoarthritis Outcome Score; QoL, quality of life; VAS, visual analog pain scale.

Discussion

In this cohort of patients with terminal knee OA, we investigated the histologic composition as well as numeric and functional differences in MSCs in the anterior and posterior capsule between those with knee FC and those without FC. We found no differences in these outcomes between the FC and no-FC groups. We did, however, identify that the FC group (by definition lacking in knee extension) also demonstrated significant reduced flexion as compared to the no-FC group. As well, our analyses revealed multiple novel and clinically important associations between knee joint histologic, MSC capsule composition, and key clinical and radiological OA markers.

The similar histologic and MSC outcomes between the FC and no-FC groups were surprising given previous findings of increased capsular fibrous tissue in joint capsules in animal models of acute joint immobilization^32
–34 and post-trauma,¹⁹ and a few factors may have contributed. First, when considering animal models, these are relatively homogenous with control over factors such as age, weight, biological sex, and animal strain.^32
–34 Conversely, the OKOA participant population was heterogeneous in terms of their demographic features (Table 1). Second, animal models allow assessment of the entire capsule. Though our human capsular biopsy sites were standardized, the amount of tissue harvested for histologic and cellular analysis was limited, especially at the posterior capsule due to the presence of the neurovascular bundles,³⁵ limiting the analyses to only a small proportion of the capsule, which may not have represented the entire posterior capsule tissue composition. Third, the cohort with terminal knee OA that evolved over many years markedly differed clinically from animal models with a maximum of 8 months of knee joint immobilization. Finally, additional non-capsular articular structural alterations may impact knee ROM and FC severity, including meniscal pathology, joint effusion, and/or femoral intercondylar notch osteophytes,¹⁴ some of which may be amenable to intervention (e.g., arthroscopic notchplasty³⁶) in select cases.

Interestingly, we discovered that OA participants with limited knee extension also had limited knee flexion. While other studies have reported the association between OA disease severity, pain and functional loss with knee extension,^9
–11 studies reporting outcomes associated with both extension and flexion are less frequent. Steultjens et al.³⁷ reported correlations between identical actions of the lateral and contralateral joints (e.g., flexion of the left and right hips), especially for joints in the transverse and sagittal planes. As OA progresses, functional loss may be associated with lost knee ROM in both directions.

Animal models of knee FC demonstrated fibrosis and shortening of the posterior capsule as a major contributor to the lost knee extension following 32 weeks of rigid immobility.^7,15,16 In people with post-traumatic elbow FCs, myofibroblasts and elevated mRNA synthesis for pro-fibrotic growth factors were reported in the joint capsule.^16,19 Based on these previous findings, we had anticipated discovering an increase in the number and fibrogenic potential of MSCs in the posterior capsule in participants with knee FC. Our analyses instead revealed surprising associations with MSC numbers in the anterior capsule: a negative association with lost knee extension (i.e., increasing MSCs numbers associated with increasing FC severity) and a positive association with knee flexion (i.e., increasing MSCs numbers associated with increasing flexion ability). Jones et al.³⁸ found increased MSC numbers in synovium samples from the suprapatellar pouch in patients with end-stage OA versus RA. Sekiya et al.³⁹ reported that synovial fluid MSCs numbers correlated with OA severity. As with our histologic findings, the mechanism of contracture development in our participants¹⁴ versus animal models of immobilization, as well as disease duration/chronicity in people with OA, may explain the novel and unexpected findings in our study. Biomechanics may also provide insight: ambulating on a knee that is unable to fully extend (adopting a knee-flexed position) may put increased mechanical strain on the anterior region of the joint and capsule through persistent activation of the knee extensors in stance.⁴⁰ The increase in MSC number and their fibrogenic capacity that we observed may represent an attempt by the anterior capsule to adopt the stronger tensile properties of fibrotic tissue to withstand the greater force across the anterior region of the joint⁴¹ (though we could not corroborate this with our histological data), without hindering flexion.

Our analysis showed a positive relationship between KL grade and percentage of “other” (mainly neurovascular) tissue CSA, KOOS QoL, and anterior capsule chondrogenic capacity, and a negative association between VAS pain and posterior capsule fibrous tissue CSA. Vascularization and innervation of the calcified cartilage has been described during OA progression (increasing KL score), likely contributing to pain.^42,43 While vascularization from the underlying subchondral bone is a potential etiology for pathologic cartilage neoangiogenesis, the synovium was reported to increase its expression of pro-angiogenic factor vascular endothelial growth factor (VEGF) and nerve growth-promoting factors in OA joints,⁴⁴ making it an additional potential tissue origin for vascular and nerve invasion of cartilage. Higher QoL in the setting of better chondrogenic capacity may simply reflect that preserved endogenous cartilage repair capacity is associated with slower disease progression and better QoL. While there was no difference in fibrous tissue in the posterior capsule between OA patients with FC versus without FC, the mean CSA in the FC group was lower than the no-FC group. While this may reflect the heterogeneity of our participant population, a lower VAS pain was associated with increasing fibrous tissue composition. Increased extracellular matrix in the posterior capsule could contribute to a “stiffer” and “shorter” capsule limiting knee extension.⁴ Loadbearing on regions of degenerated cartilage is a pain trigger,¹⁰ while avoiding loading such regions to reduce pain would restrict capsule elongation and consequently knee ROM (e.g., avoiding loading a degenerated anterior articular cartilage to reduce pain will restrict knee extension^10,45), which could be achieved physiologically through increasing capsule fibrous tissue deposition.⁸

Our findings have clinical implications. OA is now appreciated as a total joint disease and the synovium is a well-described contributor to disease burden and progression.⁴⁶ The novel associations between OA outcomes and capsule tissue composition deep to the synovium suggest a role for the entire joint capsule, not limited to the thin synovial lining. This may help explain why intra-articular treatments, such as joint injection, are less effective in some OA patients.⁴⁷ Should the treatment remain localized to the joint space and not accessing deeper tissue, efficiency may be limited. Future work looking at developing technology that allows localized diffusion of injectate (e.g., depo-glucocorticoids, or emerging VEGF-inhibiting antibodies⁴⁸), or direct injection into the capsule may improve the injection outcomes in knee OA. Associations between capsular MSCs and function with clinical outcomes such as KL grade and QoL suggest that capsule MSCs could represent targets for therapies, in addition to those derived from subchondral bone and synovium,^29,49 aimed at enhancing endogenous joint repair and slowing OA progression.

Limitations include the cross-sectional design, which can identify associations but not causation. The heterogeneous nature of the participant population may have impacted the ability to identify correlations between clinical and laboratory outcomes, but is in keeping with the heterogeneous nature of OA. Variability in sample size for laboratory outcomes may have also impacted the analysis, with sample size limited by database availability.

Conclusion

Joint capsule histology and MSCs were associated with important clinical outcomes of knee OA, including knee extension, flexion, KL grade, VAS pain, and QoL. These findings suggest that the entire joint capsule, not only the synovium, may play important roles in OA-related morbidity and progression and could represent an underappreciated target for OA treatment.

Supplemental Material

sj-docx-1-tab-10.1177_1759720X251321941 – Supplemental material for Capsular stem cell function and tissue composition are associated with symptoms and radiographic severity in people with knee osteoarthritis

Supplemental material, sj-docx-1-tab-10.1177_1759720X251321941 for Capsular stem cell function and tissue composition are associated with symptoms and radiographic severity in people with knee osteoarthritis by T. Mark Campbell, Robert Feibel, Jeffrey Dilworth, Odette Laneuville and Guy Trudel in Therapeutic Advances in Musculoskeletal Disease

Supplemental Material

sj-docx-2-tab-10.1177_1759720X251321941 – Supplemental material for Capsular stem cell function and tissue composition are associated with symptoms and radiographic severity in people with knee osteoarthritis

Supplemental material, sj-docx-2-tab-10.1177_1759720X251321941 for Capsular stem cell function and tissue composition are associated with symptoms and radiographic severity in people with knee osteoarthritis by T. Mark Campbell, Robert Feibel, Jeffrey Dilworth, Odette Laneuville and Guy Trudel in Therapeutic Advances in Musculoskeletal Disease

Footnotes

Acknowledgements

The authors thank Katherine Reilly and Mohamed Thabet for performing the laboratory experiments described in this report, Fernando Ortiz for assisting with flow cytometry and MSC gating strategy and Naomi Abayomi for providing measurements the tissue proportions for histology. We thank our knowledge user/patient partner Alfretta Vanderheyden for contributing to study design, application to funding, and knowledge dissemination activities.

Declarations

ORCID iD

T. Mark Campbell

Supplemental material

Supplemental material for this article is available online.

References

Loeser

. Age-related changes in the musculoskeletal system and the development of osteoarthritis. Clin Geriatr Med 2010; 26: 371–386.

National Joint Registry. National Joint Registry 20th Annual Report 2023, https://www.hqip.org.uk/wp-content/uploads/2023/10/NJR-20th-Annual-Report-2023.pdf (2024, accessed 7 July 2024).

Sellam

Berenbaum

. The role of synovitis in pathophysiology and clinical symptoms of osteoarthritis. Nat Rev Rheumatol 2010; 6: 625–635.

Campbell

Trudel

Laneuville

. Knee flexion contractures in patients with osteoarthritis: clinical features and histologic characterization of the posterior capsule. PM R 2015; 7: 466–473.

Campbell

Trudel

Wong

, et al. Genome wide gene expression analysis of the posterior capsule in patients with osteoarthritis and knee flexion contracture. J Rheumatol 2014; 41: 2232–2239.

Ritter

Lutgring

Davis

, et al. The role of flexion contracture on outcomes in primary total knee arthroplasty. J Arthroplasty 2007; 22: 1092–1096.

Matsumoto

Trudel

Uhthoff

. High collagen type I and low collagen type III levels in knee joint contracture: an immunohistochemical study with histological correlate. Acta Orthop Scand 2002; 73: 335–343.

Trudel

Laneuville

Uhthoff

. Joint contractures. Clin Orthop Relat Res 2007; 456: 2.

Campbell

McGonagle

. Flexion contracture is a risk factor for knee osteoarthritis incidence, progression and earlier arthroplasty: data from the osteoarthritis initiative. Ann Phys Rehabil Med 2021; 64: 101439.

10.

Campbell

Reilly

Laneuville

, et al. Baseline knee extension may be associated with volumetric cartilage loss in the anterolateral tibia: data from the osteoarthritis initiative. Rheumatology (Oxford, England) 2023; 62(5): 1950–1954.

11.

Campbell

Ramsay

Trudel

. Knee flexion contractures are associated with worse pain, stiffness, and function in patients with knee osteoarthritis: data from the osteoarthritis initiative. PM R 2021; 13: 954–961.

12.

Lim

Huang

, et al. Evaluation of the elderly patient with an abnormal gait. J Am Acad Orthop Surg 2007; 15: 107–117.

13.

Potter

Kirby

MacLeod

. The effects of simulated knee-flexion contractures on standing balance. Am J Phys Med Rehabil 1990; 69: 144–147.

14.

Campbell

Trudel

Conaghan

, et al. Flexion contracture is associated with knee joint degeneration on magnetic resonance imaging: data from the osteoarthritis initiative. Clin Exp Rheumatol 2022; 40: 993–998.

15.

Zhou

Trudel

Uhthoff

, et al. Range of extension correlates with posterior capsule length after knee remobilization. Med Sci Sports Exerc 2018; 50: 2401–2408.

16.

Hildebrand

Sutherland

Zhang

. Rabbit knee model of post-traumatic joint contractures: the long-term natural history of motion loss and myofibroblasts. J Orthop Res 2004; 22: 313–320.

17.

Firestone

Krackow

Davis

4th , et al. The management of fixed flexion contractures during total knee arthroplasty. Clin Orthop Relat Res 1992; 284: 221–227.

18.

Mihalko

Whiteside

. Bone resection and ligament treatment for flexion contracture in knee arthroplasty. Clin Orthop Relat Res 2003; 406: 141–147.

19.

Hildebrand

Zhang

Hart

. Myofibroblast upregulators are elevated in joint capsules in posttraumatic contractures. Clin Orthop Relat Res 2007; 456: 85–91.

20.

Uezumi

Ito

Morikawa

, et al. Fibrosis and adipogenesis originate from a common mesenchymal progenitor in skeletal muscle. J Cell Sci 2011; 124: 3654–3664.

21.

El Agha

Kramann

Schneider

, et al. Mesenchymal stem cells in fibrotic disease. Cell Stem Cell 2017; 21: 166–177.

22.

Jones

English

Henshaw

, et al. Enumeration and phenotypic characterization of synovial fluid multipotential mesenchymal progenitor cells in inflammatory and degenerative arthritis. Arthritis Rheum 2004; 50: 817–827.

23.

Cipriani

Di Benedetto

Ruscitti

, et al. Perivascular cells in diffuse cutaneous systemic sclerosis overexpress activated ADAM12 and are involved in myofibroblast transdifferentiation and development of fibrosis. J Rheumatol 2016; 43: 1340–1349.

24.

Fournier

Trudel

Feibel

, et al. Asymmetric flexion contracture is associated with leg length inequality in patients with knee osteoarthritis: data from the osteoarthritis initiative. Arch Phys Med Rehabil 2023; 104: 2067–2074.

25.

Altman

Asch

Bloch

, et al. Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association. Arthritis Rheum 1986; 29: 1039–1049.

26.

Maricar

Callaghan

Parkes

, et al. Interobserver and intraobserver reliability of clinical assessments in knee osteoarthritis. J Rheumatol 2016; 43: 2171–2178.

27.

Ornetti

Dougados

Paternotte

, et al. Validation of a numerical rating scale to assess functional impairment in hip and knee osteoarthritis: comparison with the WOMAC function scale. Ann Rheum Dis 2011; 70: 740–746.

28.

Roos

Toksvig-Larsen

. Knee injury and Osteoarthritis Outcome Score (KOOS)—validation and comparison to the WOMAC in total knee replacement. Health Qual Life Outcomes 2003; 25: 17–26.

29.

Campbell

Churchman

Gomez

, et al. Mesenchymal stem cell alterations in bone marrow lesions in hip osteoarthritis. Arthritis Rheumatol 2016; 68: 1648–1659.

30.

Dominici

Le Blanc

Mueller

, et al. Minimal criteria for defining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement. Cytotherapy 2006; 8: 315–317.

31.

Lee

Shah

Moioli

, et al. CTGF directs fibroblast differentiation from human mesenchymal stem/stromal cells and defines connective tissue healing in a rodent injury model. J Clin Invest 2015; 125: 3992.

32.

Chimoto

Hagiwara

Ando

, et al. Progression of an arthrogenic motion restriction after immobilization in a rat experimental knee model. Ups J Med Sci 2007; 112: 347–355.

33.

Trudel

Zhou

Uhthoff

, et al. Four weeks of mobility after 8 weeks of immobility fails to restore normal motion: a preliminary study. Clin Orthop Relat Res 2008; 466: 1239–1244.

34.

Zhou

Trudel

Goudreau

, et al. Knee joint stiffness following immobilization and remobilization: a study in the rat model. J Biomech 2020; 99: 109471.

35.

Wong

Trudel

Laneuville

. Intra-articular collagenase injection increases range of motion in a rat knee flexion contracture model. Drug Des Devel Ther 2018; 12: 15–24.

36.

Ferrari

Mannava

DePhillipo

, et al. Notchplasty for the arthroscopic treatment of limited knee extension. Arthrosc Tech 2017; 6: e517–e524.

37.

Steultjens

Dekker

van Baar

, et al. Range of joint motion and disability in patients with osteoarthritis of the knee or hip. Rheumatology (Oxford, England) 2000; 39: 955–961.

38.

Jones

Churchman

English

, et al. Mesenchymal stem cells in rheumatoid synovium: enumeration and functional assessment in relation to synovial inflammation level. Ann Rheum Dis 2010; 69: 450–457.

39.

Sekiya

Katano

Ozeki

. Characteristics of MSCs in synovial fluid and mode of action of intra-articular injections of synovial MSCs in knee osteoarthritis. Int J Mol Sci 2021; 22: 2838.

40.

Cifu

. Braddom’s physical medicine and rehabilitation. 6th ed. Philadelphia, PA: Elsevier, 2021.

41.

Wells

. Tissue mechanics and fibrosis. Biochim Biophys Acta 2013; 1832: 884–890.

42.

Goldring

. Changes in the osteochondral unit during osteoarthritis: structure, function and cartilage-bone crosstalk. Nat Rev Rheumatol 2016; 12: 632–644.

43.

Walsh

McWilliams

Turley

, et al. Angiogenesis and nerve growth factor at the osteochondral junction in rheumatoid arthritis and osteoarthritis. Rheumatology (Oxford) 2010; 49: 1852–1861.

44.

Mapp

Walsh

. Mechanisms and targets of angiogenesis and nerve growth in osteoarthritis. Nat Rev Rheumatol 2012; 8: 390–398.

45.

Campbell

Laneuville

Trudel

. Association of knee osteoarthritis and flexion contracture with localized tibial articular cartilage loss: data from the osteoarthritis initiative. J Rheumatol 2024; 51: 285–290.

46.

Martel-Pelletier

Barr

Cicuttini

, et al. Osteoarthritis. Nat Rev Dis Primers 2016; 2: 16072.

47.

Guermazi

Hunter

Kloppenburg

. Debate: intra-articular steroid injections for osteoarthritis—harmful or helpful? Osteoarthr Imaging 2023; 3: 100163.

48.

Sotozawa

Kumagai

Ishikawa

, et al. Bevacizumab suppressed degenerative changes in articular cartilage explants from patients with osteoarthritis of the knee. J Orthop Surg Res 2023; 18: 25.

49.

McGonagle

Baboolal

Jones

. Native joint-resident mesenchymal stem cells for cartilage repair in osteoarthritis. Nat Rev Rheumatol 2017; 13: 719–730.

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.

0.00 MB

3.96 MB

0.03 MB