Relationship Between Synovial Fluid Biomarkers of Articular Cartilage Metabolism and the Patient’s Perspective of Outcome Depends on the Severity of Articular Cartilage Damage Following ACL Trauma

Abstract

Anterior cruciate ligament (ACL) trauma often occurs in combination with injury to the articular cartilage of the knee, this can result in earlier radiographic evidence of post traumatic osteoarthritis (OA) of the knee compared to the contralateral, ACL intact knee; however, the biomechanical and biological mechanisms associated with the onset and progression of this disease are not understood. We sought to gain insight into the mechanisms by determining the relationship between articular cartilage injury associated with ACL trauma and the expression of synovial fluid biomarkers of articular cartilage metabolism, and to evaluate the relationship between these biomarkers and the patient’s perspective of the outcomes. Synovial fluid samples were acquired from 39 ACL injured subjects at an average of 10 weeks after injury, and 32 control subjects with normal knees (documented with clinical exam and MRI assessment). Subjects in the ACL-injured group were classified as low-risk for future OA if they displayed an International Cartilage Repair Society (ICRS) Grade 2 articular cartilage lesion or less and high-risk for future OA if they had an ICRS Grade 3A articular cartilage lesion. The patient’s perspective of the injury was evaluated with the Knee Injury and Osteoarthritis Outcomes Score (KOOS). There were no significant differences in mean concentrations of the markers of type II collagen metabolism (CPII, C2C, and C1,2C) or the aggrecan breakdown Alanine–Arginine–Glycine–Serine (ARGS) -fragment between control subjects and the subjects in the low- and high-risk groups (p-value range: 0.80–0.43). Associations between ARGS-aggrecan concentration and KOOS subscales of symptoms and pain were significantly different between the low- and high-risk groups (p 0.03 and p 0.01, respectively). Likewise, there was strong evidence in support of an association between the markers of type II collagen metabolism (C1,2C and CPII concentrations) and the KOOS subscale of pain between the low- and high-risk groups (p 0.051 and 0.077, correspondingly). In ACL injured subjects with concomitant Grade 3A articular cartilage injuries, concentrations of synovial fluid ARGS-aggrecan were directly associated with improvements in KOOS symptoms and pain. These findings suggest the possible involvement of ARGS-aggrecan in a localized tissue repair response involving an increase in aggrecan turnover following severe knee trauma.

The anterior cruciate ligament (ACL) is the most commonly disrupted knee ligament. Wide ranges of incidence rates for both ACL injury and reconstruction have been reported in the literature, but most are based on estimations performed over 15 years ago.^1–4 While the true incidence of ACL reconstruction in the United States is unknown,⁵ a recent study identified a rise in ACL reconstruction from 86,687 (32.9 per 100,000 person years) in 1994 to 129,836 (43.5 per 100,000 person years) in 2006, with increased numbers of procedures occurring in females and patients younger than 20 or older than 40 years of age.⁶

Little is known about the initial onset and progression of post-traumatic osteoarthritis (PTOA) following ACL disruption. Rates of OA following ACL injury range from 10–90% at 10–20 years after injury, with a mean of more than 50% suggested.⁷ Reconstruction attempts to restore normal biomechanics of the knee, and individuals undergoing this procedure show more normal anterior-posterior laxity and improved patient and functional outcomes compared to those who do not undergo reconstruction.⁸ However, this frequently does not protect the knee from the development of PTOA.^9–11 In a meta-analysis involving 33 clinical studies, the efficacy of ACL repair or reconstruction in slowing or preventing the progression of OA was not confirmed.¹² Recently, Frobell et al.¹³ reported no significant difference in radiographic evidence of OA at 5 years between a group of patients assigned to early ACL reconstruction versus a group assigned to rehabilitation combined with delayed ACL reconstruction. Comparing those treated with any type of reconstruction (early vs. delayed) to those treated solely with rehabilitation, there was no significant difference in tibiofemoral radiographic OA.¹³ In a systematic review and meta-analysis examining ACL injury and radiographic progression to OA, Ajuied et al.¹⁴ found that irrespective of treatment, there was greater than a threefold risk of developing OA after ACL injury compared to the contralateral, ACL intact knee. Knees treated non-operatively had a significantly higher risk of developing any grade of OA compared to those treated surgically. Yet, when analyzing progression to moderate or severe OA (Kellgren Lawrence Grade III or IV), ACL reconstructed knees had a significantly higher risk than nonoperatively managed injuries at 10 year follow-up.¹⁴

Isolated ACL tears are uncommon. They are usually associated with injury to the menisci, articular cartilage, or underlying subchondral bone. Clinical studies examining isolated ACL tears have reported a lower incidence of PTOA at long-term follow up compared to those with concomitant meniscus and/or articular cartilage injury.¹⁵ At a mean follow up of 3.9 years, Ichiba and Kishimoto¹⁶ showed statistically significant increases in OA scores in patients with ACL and articular cartilage injury compared with those that suffered isolated ACL injury. They also showed significantly higher OA scores in patients with ACL injury and meniscectomy versus ACL injury and an intact (repaired or left in situ) meniscus.¹⁶ At a mean follow up of 7.6 years following ACL reconstruction, Shel-bourne et al.¹⁷ showed that there was greater anterior instability and a higher incidence of OA changes seen on radiographs in patients with associated medial meniscectomy than in those with lateral meniscectomy or meniscus intact groups.¹⁷ When associated with meniscal injuries, Oiestad et al. found that the prevalence of OA in knees following ACL injury increased from 0–13% to between 21–48%.¹⁸

Significant effort is being put forth towards identifying biomarkers that may serve as prognostic indicators capable of identifying those at increased risk for onset and progression of PTOA. A large volume of work has focused on systemic and joint specific alterations in biomarker concentrations for type II collagen and aggrecan in idiopathic and PTOA models.^7,19–28 We have reported on the association between markers of type II collagen turnover in serum and urine and radiographic measures of tibiofemoral joint space width changes in a cohort of subjects that suffered ACL injury combined with varying severity of meniscal and/or articular cartilage damage.²² Injured subjects with an abnormal tibiofemoral joint space width had diminished quality of life, increased pain, and increased type II collagen cleavage to synthesis ratios compared to matched control subjects over the 4-year follow-up interval.²² More recently, we demonstrated that articular cartilage undergoes dramatic change soon after ACL disruption with significant increases of cartilage thickness in central regions of the tibial plateau, and significant decreases of cartilage thickness in the posterior region of the tibial plateau,²³ while Eckstein et al. have demonstrated that these changes persist for years after the index injury.²⁹ The purpose of this study was twofold. First, we wanted to examine patients who suffered an ACL disruption with varying levels of articular cartilage injury to determine the relationship between these concomitant injuries and the expression of synovial fluid biomarkers of cartilage metabolism after injury but prior to surgery compared to a group of healthy, matched controls. Second, we wanted to evaluate the relationship between these synovial fluid biomarkers and patient oriented outcome measures to determine if articular cartilage injury severity correlates with knee pain, symptoms, and functional decline. We hypothesized that subjects with greater articular cartilage damage at the time of ACL injury would express higher levels of synovial fluid biomarkers of cartilage metabolism and would report greater pain, symptoms, and functional impairment in comparison to those with ACL injury combined with less severe cartilage damage.

MATERIALS AND METHODS

This study involved ancillary analysis of data collected from a longitudinal cohort study designed to evaluate relationships between biomarkers of type-II collagen metabolism and progression of PTOA as measured by tibiofemoral joint space width changes following ACL injury and reconstruction.²² While our prior publication focused on the temporal response of biomarkers collected in serum and urine,²² this publication is focused only on analysis of biomarkers in synovial fluid samples that were obtained at a mean time of 70.1 days after injury (range: 18–155 days) and prior to surgery. The ACL injured group consisted of 39 subjects with acute, first time ACL injuries without prior history of knee injury. The control group consisted of 32 healthy subjects with no prior history of knee pain or dysfunction in either knee and a normal knee clinical examination.²² Control subjects were of similar age, sex, BMI, and activity level compared to the ACL injured subjects (Table 1). Subjects were considered to be at high risk for progressive osteoarthritic change if they suffered an ICRS Grade 3A articular cartilage lesion in any compartment. Given that cartilage injuries extending down to the calcified layer or subchondral bone have the potential to impact synovial fluid biomarker concentrations, subjects with more than a Grade 3A articular cartilage injury were excluded. Subjects were considered to be at low risk for progressive osteoarthritic change if they suffered an ICRS Grade 2 or less articular cartilage lesion in any compartment. These risk group assignments have been used previously in work examining the effect of articular cartilage injury at the time of surgery on the development of osteoarthritic changes in patients undergoing ACL reconstruction.^16,31 In addition, this approach created groups with articular cartilage lesions that extended no deeper than 50% of the overall cartilage thickness (low risk group) and greater than 50% but less than 100% of cartilage thickness, as 100% would have exposed calcified cartilage and subchondral bone (high risk group). Given that our analysis was focused specifically on the condition of the articular cartilage, we did not group subjects by meniscal injury, as has been done previously.^16,31

Table 1.

Patient Characteristics at Baseline

	Control Group (re = 32)	Low Risk Group (n = 29)	High Risk Group (n = 10)
Age, years, mean (SD)	22.8 (7)	21.7 (11)	31.1 (12)
Sex (male/female)	14/18	15/14	4/6
BMI, mean (SD)	24.1 (6)	24.8 (4)	24.7 (3)
Tegner score, mean (SD)	6.2 (1)	7.9 (1)	7.1 (2)
Associated meniscus injury; n (%)	0(0)	8(28)	4(40)

BMI, body mass index; SD, standard deviation.

Entry criteria for ACL injured participants included: Age 14–55 years, BMI 18.5–30, and at least a moderate activity level (Tegner score ≥5). They had no prior surgery or injection to either knee joint, no abnormal laxity of the posterior cruciate ligament or collateral ligaments, no abnormal rotatory laxity, no radiographic evidence of fracture or OA, normal lower limb alignment (as defined by the 2000 International Knee Documentation Committee [IKDC] Knee Examination form), less than 2/3 meniscectomy in either meniscus, and Grade 3A or less articular cartilage lesions. Similar entry criteria were employed for control group participants with the following exceptions: These subjects reported no knee pain or physical dysfunction as determined by the KOOS³² and the subjective IKDC form³³, had no history of significant trauma to any joint (defined as that requiring physician referral and/or more than 3 days modified activities of daily living), no abnormal findings with clinical knee examination, and no articular pathology or abnormal findings on MRI.

Articular cartilage lesions were identified via arthroscopic visualization, graded by one of two attending surgeons, and recorded on the 2000 IKDC Surgical Documentation form. If multiple lesions existed, the lesion with the highest grade was utilized for categorization of the subject into either the low- or high-risk groups.

Patient-oriented assessments of pain, symptoms, function in sports and recreational activities, function during activities of daily living, and knee-related quality of life scores were evaluated with utilization of the KOOS, which has been validated for the study of subjects following ACL reconstruction and to assess patients with OA.³² These were obtained at initial evaluation for control subjects and within 3 weeks prior to surgery for injured subjects.

Knee synovial fluid samples were obtained from control subjects during their initial evaluation and from ACL injured subjects via non-lavage arthrocentesis under sterile conditions. Samples were centrifuged (4000g) at 4°C for 20 min and supernatants were stored at −80°C until processing. Type II collagen synthesis (CPII epitope) and cleavage concentrations (C2C and C1,2C epitopes) were evaluated with commercially available enzyme-linked immunosorbent assay (ELISA) kits (Ibex Pharmaceuticals). These were performed by the same investigator (TT) and done in duplicate per manufacturer recommendation. The biological markers chosen for this investigation were based on previous work evaluating the earliest stages of primary OA,²⁴ prior research focused on measuring type II collagen metabolism,^22,25,34 and consultation with experts in the field of OA biomarker analysis (see Acknowledgments). Type II collagen synthesis was evaluated by measuring synovial fluid procollagen II C-propeptide (CPII). The competitive ELISA for this marker uses a monoclonal antibody that binds to type II collagen propeptide epitopes cleaved from the C-terminus of procollagen after being released into the matrix.²⁷ Consequently, CPII serves as a marker of newly formed type II collagen. The intra-assay and inter-assay coefficients of variation (CVs) for CPII were 7.22% and 7.31%, respectively. Concentrations of collagen cleavage markers were evaluated in synovial fluid via ELISA and included a collagen type II cleavage product (C2C) and a collagen type I and II cleavage product (C1,2C). The intra-assay and inter-assay CVs for C2C were <5% and 7.23%, while the intra-assay and inter-assay CVs for C1,2C were both <5%.

Using an electrochemiluminescence (ECLC) immunoassay as previously described,²⁷ Alanine–Arginine–Glycine–Serine (ARGS) levels were evaluated by the same investigator (AS). ARGS is a neoepitope of aggrecan formed via enzymatic cleavage of aggrecan at the Glu/Ala bond, resulting in the release of ARGS into the synovial fluid. Thus, ARGS is a measure of aggrecan breakdown within articular cartilage that occurs in the initial weeks following ACL injury.^28,30

All biomarker data that were used as dependent variables for analyses were log-transformed in order to ensure normality. Analyses of variance were performed in order to test for significant differences in mean levels of biomarker concentrations between the low risk, high risk, and the control groups, significant associations between biomarker concentrations and KOOS subscales, and significant differences in these associations between injury risk groups. Significant differences in biomarker versus KOOS associations between injury risk groups were shown by significant group*KOOS subscale interactions, and correlations analysis was conducted using Pearson product-moment correlation coefficient (r). Comparisons of KOOS subscale means between injury risk groups were done using non-parametric Kruskal–Wallis tests. The significance level alpha was set at 0.05, and statistical analyses were conducted using SAS ver. 9.2 (SAS Institute, Inc., Cary, NC). Our institutional review board approved the protocol prior to subject enrollment and all subjects provided written informed consent prior to participation.

RESULTS

Of the 39 ACL injured subjects, nine (23%) sustained no injury to the articular cartilage surfaces of the knee, five (13%) had Grade 1A injury, two (5%) had Grade 1B injury, 13 (33%) had a Grade 2 injury, and 10 (26%) had a Grade 3A injury. Based on these findings, 10 (26%) subjects were stratified into the high-risk group (ICRS Grade 3A articular cartilage injuries) and the remaining 29 (74%) subjects were stratified into the low risk group.

Comparison of Synovial Fluid Biomarkers Between Groups

All injured subjects with the exception of one had synovial fluid available for biomarker analysis. Of the 32 control subjects, synovial fluid could only be obtained from 27. When controlling for sex, BMI, and activity level, there was no significant difference in biomarker concentration for type II collagen synthesis (CPII), type I and II collagen breakdown (C2C and C1,2C), type II collagen metabolism (measured as the ratio of breakdown to synthesis [C2C/CPII]), or aggrecan degradation (ARGS-aggrecan) between groups (Table 2). When further controlling for time between injury and acquisition of synovial fluid, there was no significant difference in any of the established bio-markers between high and low risk groups (data not shown).

Table 2.

Log Transformed Mean Synovial Fluid Markers for Type II (C2C and CPII) and Type I and II (C2,2C) Collagen and ARGS-Aggrecan (pmol/ml SF), Comparisons Between All Groups After Adjustment for Sex, BMI, and Activity Level

	Control (n = 27)	Low Risk (n = 28)	High Risk (n = 10)	p -Value
log C2C	4.89 (0.12)	5.11 (0.08)	4.83 (0.43)	0.43
log C1,2C	4.97 (0.08)	5.19 (0.08)	4.80 (0.47)	0.80
log CPU	4.76 (0.12)	4.95 (0.11)	4.32 (0.49)	0.67
log C2C/CPII	0.13 (0.14)	0.15 (0.12)	0.51 (0.24)	0.75
log C1,2C/CPII	0.17 (0.13)	0.24 (0.09)	0.43 (0.25)	0.52
log ARGS	3.13 (0.11)	3.14 (0.13)	2.60 (0.47)	0.43

Data is expressed in mean values (Std. Error).n, available synovial fluid samples.

Group Comparisons Between Synovial Fluid Biomarkers and KOOS Patient Oriented Outcomes

Data for the five KOOS subscales were available for 30 of 32 (94%) control subjects and 36 of 39 (92%) injured subjects. The remaining five subjects without available KOOS data were not included in the statistical analysis. While controls scored significantly higher across all KOOS subscales compared to the high and low risk groups, there were no significant differences in mean KOOS subscale scores between risk groups (Table 3). Biomarkers for type II collagen synthesis, breakdown, metabolism, and the biomarker for aggrecan degradation were analyzed against the KOOS subscales for each subject group to determine if changes in bio-marker concentrations were associated with subjective complaints and functional deterioration. There were no significant differences between subject groups comparing the synovial fluid markers CPII, C2C, and C1,2C with any of the KOOS subscales (Table 4). However, there were statistically significant interactions between ARGS-aggrecan and KOOS subscales, (Table 4). Synovial fluid ARGS-aggrecan level was directly associated with reports of symptoms within the high risk group (r = 0.42), but was inversely associated with symptoms within the low risk group (r = −0.31). This difference in association between groups was shown by a significant KOOS-Symptoms by group interaction (p = 0.033). Similarly, synovial fluid ARGS-aggrecan level was directly associated with reports of pain within the high risk group (r = 0.54), but was inversely associated with pain within the low risk group (r = −0.16). This difference in association between groups was shown by a significant KOOS-Pain by Group interaction (p = 0.014). However, there was an inverse association between synovial fluid ARGS-aggrecan level and function during sport within the high risk group (r = = −0.50), while within the low risk group, synovial fluid ARGS-aggrecan level was directly associated with function during sport (r = 0.18), resulting in a significant KOOS-Sports by Group interaction (p = 0.028).

Table 3.

Patient Oriented Outcomes (Using Knee Oteoarthritis Outcome Score, or KOOS) for Control Subjects, ACL Injured Subjects With ICRS Grade 2 or Less Articular Cartilage Injury (Low Risk Group), and ACL Injured Subjects With ICRS Grade 3 Articular Cartilage Injury (High Risk group)

Group	N	Mean (SD)	Group Comparison (p-value)
KOOS: Pain
High risk	9	73 (11)	<0.0001a
Low risk	27	73 (12)	<0.0001b
Control	30	99 (2)	0.88c
KOOS: Quality of Life
High risk	9	33 (13)	<0.0001a
Low risk	27	38 (20)	<0.0001b
Control	30	99 (3)	0.77c
KOOS: AD
High risk	9	90(7)	<0.0001a
Low risk	27	85 (13)	<0.0001b
Control	30	100 (1)	0.47c
KOOS: Symptom
High risk	9	69 (10)	<0.0001a
Low risk	27	65 (15)	<0.0001b
Control	30	98(4)	0.61c
KOOS: Sports
High risk	9	39 (29)	<0.0001a
Low risk	27	48 (24)	<0.0001b
Control	30	100 (2)	0.38c

^aControl subjects versus low risk group subjects.

^bControl subjects versus high risk group subjects.

^cArticular cartilage injury subgroup comparison, High risk versus Low risk.

Table 4.

Comparing Knee Osteoarthritis Outcome Score (KOOS) Subscales With Biomarker Concentrations Between High Risk (ICRS Grade 3A) and Low Risk (ICRS Grade 2 or Less) Groups

Biomarker	KOOS*group interaction (p-value)
KOOS: Pain
log C2C	0.392
log C1,2C	0.077
log CPU	0.051
log C2C/CPII	0.123
log C1,2C/CPII	0.215
log ARGS	0.014
KOOS: Quality of life
log C2C	0.161
log C1,2C	0.620
log CPU	0.542
log C2C/CPII	0.656
log C1,2C/CPII	0.291
log ARGS	0.815
KOOS: ADL
log C2C	0.918
log C1,2C	0.992
log CPU	0.392
log C2C/CPII	0.181
log C1,2C/CPII	0.193
log ARGS	0.388
KOOS: Symptoms
log C2C	0.408
log C1,2C	0.260
log CPU	0.988
log C2C/CPII	0.294
log C1,2C/CPII	0.150
log ARGS	0.033
KOOS: Sports
log C2C	0.145
log C1,2C	0.124
log CPU	0.503
log C2C/CPII	0.419
log C1,2C/CPII	0.497
log ARGS	0.028

KOOS*group interactions, test for differences in KOOS outcome (Pain, Quality of Life, Activities of Daily living or ADL, Symptoms, and Sports Participation scales) versus biomarker association between groups.Significance, p < 0.05 marked bolded.

DISCUSSION

The goal of this investigation was to examine local, joint specific alterations in biochemical markers for type II collagen metabolism and aggrecan degradation in a population of otherwise healthy individuals following acute, traumatic ACL rupture with Grade 2 or less articular cartilage damage that extended down to less than 50% of cartilage thickness (low risk group), or Grade 3A cartilage damage that extended greater than 50% but less than 100% of the cartilage thickness (high risk group). We failed to support our hypothesis that subjects in the high risk group that had ACL injury combined with substantial articular cartilage injury would express increased levels of synovial fluid biomarkers of cartilage metabolism in comparison to those in the low risk group that experienced ACL injury combined with less severe articular cartilage injury. Instead, biomarkers for type II collagen synthesis (CPII), degradation (C2C, C1,2C), metabolism (C2C:CPII), and aggrecan breakdown (ARGS-aggrecan) did not vary between risk groups and control subjects. Since prior research has shown that synovial fluid levels of ARGS-aggrecan can vary based on the time interval from injury to acquisition of the synovial fluid sample,^28,35 biomarkers of type II collagen and aggrecan breakdown were compared between the injury groups while controlling for time between injury and acquisition of the synovial fluid sample. Again, this failed to demonstrate any significant difference between groups.

There was a significant association between ARGS-aggrecan concentration and the KOOS assessments of symptoms, pain, and function during sport between the low and high risk groups. In line with our hypothesis, as ARGS-aggrecan concentration increased in subjects with worse articular cartilage injuries, function during sport declined, and as ARGS-aggrecan increased in subjects with less severe articular cartilage injury, function during sport improved. Contrary to our hypothesis, the association between ARGS-aggrecan and patient perception of pain and symptoms was direct for the high-risk group and inverse for the low-risk group. As synovial fluid levels of ARGS-aggrecan increased in subjects with more severe articular cartilage injuries, there were corresponding improvements in reports of symptoms and pain. Conversely, as synovial fluid levels of ARGS-aggrecan increased in subjects with less severe articular cartilage injury, there was a slight worsening of symptoms and pain. To our knowledge, this represents the first investigation to suggest that synovial fluid levels of ARGS-aggrecan is significantly associated with patient reported outcomes of symptoms and function during sport and that these relationships depend on the severity of articular cartilage injury. However, given that the r-values for the associations seen in the low risk group are small, there are likely additional factors contributing to the interaction of ARGS-aggrecan with patient oriented outcomes than just damage to the articular cartilage surface. Conversely, for the high risk group, the associations between ARGS-aggrecan and pain, symptoms, and function during sport are stronger. This finding introduces the hypothesis that ARGS-aggrecan may play a more significant role in identifying early subjective complaints of knee dysfunction in patients who have sustained more severe articular cartilage injury at the time of ACL disruption.

A possible explanation for these findings could be that in cases where more severe cartilage injury occurs, the body undertakes a robust local tissue-repair response in its initial attempt to facilitate healing. This would involve not only the generation of ARGS-fragments, but also result in the formation of newly synthesized aggrecan. Our results corroborate with those of Larsson et al.^27,36 who examined the association between synovial fluid levels of ARGS-aggrecan and radiographic progression of OA. In their study of 141 subjects who had undergone meniscectomy, synovial fluid ARGS-aggrecan levels were no different than those in a reference group without meniscectomy, nor was there any difference between subjects with radiographic OA and those without. However, within what were relatively low and considered normal ARGS-aggrecan levels, they found a weak inverse association between synovial fluid ARGS-aggrecan concentrations and tibiofemoral joint space width. In a follow up study of the same cohort, they investigated whether changes in ARGS-aggrecan concentration over time were associated with progression of radiographic knee OA and patient reported outcomes as measured using the KOOS. They found that in subjects who had decreasing synovial fluid ARGS-aggrecan levels over time, the likelihood of loss of tibiofemoral joint space width increased sixfold and the likelihood of worsening of KOOS pain increased fourfold compared to subjects with increasing synovial fluid ARGS-aggrecan concentrations.^27,36 Similarly, when considering the results from our high risk group, we confirm the existence of a significant association between decreasing synovial fluid ARGS-aggrecan concentrations and worsening of pain as measured via the KOOS.

Our findings of an inverse association between ARGS-aggrecan and KOOS symptoms and pain for subjects with low grade articular cartilage injury could be related to the extent of matrix and chondrocyte damage seen in this population. The mechanical load sustained at the time of ACL injury can have a significant impact on the extent of articular cartilage damage, leading to an increase in reactive oxygen species and accelerated chondrocyte senescence.³⁷ Recent studies modeling intra-articular fractures have also shown a progressive expansion of chondrocyte death away from localized fracture lines in the 48 hours following injury.³⁸ An investigation performed by Ewers et al.³⁹ explored the effect of loading rate on matrix damage and cell death in a bovine cartilage explant model. Overall, they found that there was a greater amount of matrix damage in explants subjected to high rates of loading with chondrocyte death confined to areas adjacent to fissures in the articular surface. However, explants that were subjected to lower rates of loading showed a lesser degree of matrix damage but a more diffuse pattern and a greater overall number of chondrocyte death.³⁹ In the current study, the low risk group included 41% of those enrolled that had either normal cartilage or superficial fissures, and 33% that had lesions extending down to less than 50% of cartilage thickness. It is possible that an injury of this extent is insufficient to generate a robust tissue repair response, thus blunting the synthesis of aggrecan and other extracellular matrix proteins. Additionally, as local chondrocytes are responsible for aggrecan synthesis, a more diffuse pattern of chondrocyte death in these subjects could explain the blunted anabolic response to articular cartilage injury.

While an increase in ARGS-aggrecan concentration in “high risk” subjects is associated with improvements in symptoms and pain, this increase does not appear to translate into activity related improvement. As synovial fluid ARGS-aggrecan concentration increased, KOOS function during sport declined. Conversely, as ARGS-aggrecan concentration increased in “low risk” subjects KOOS function during sport improved. These contrasting findings could be explained by the ability of subjects to self regulate their behavior and activity level based on the extent of intra-articular injury. In subjects with more severe injury, they are likely to be less active. This inactivity could also help explain the improvement in pain and symptoms that were observed. Conversely, subjects with less severe injury may still attempt to “push through the pain” which over time results in the worsening of symptoms and pain seen in this “low risk” group.

The statistically significant relationships between the high and low risk groups established in this study were based on an adequate sample size; however, one potential limitation associated with some of the comparisons is the statistical power to establish differences in biomarker concentrations between the groups if in fact they exist. For example, post-hoc analysis revealed that in order to detect a significant (p < 0.05) difference in slopes of the KOOS-pain versus log CPII relationship between the low-and high-risk groups, with 80% power, a sample size of 22 subjects per group would be required. Similarly, in order to detect a significant (p < 0.05) difference in slopes of the KOOS pain versus log C1,2C relationship between the lowand high-risk groups a sample size of 26 per group would be required. Since the actual sample sizes were 29 subjects in the low-risk group and 10 subjects in the high-risk group, and the p-values associated with the group*KOOS pain interaction terms for log CPII and log C1,2C were 0.051 and 0.077, correspondingly, a larger sample size (specifically, for the high risk group) may have resulted in significant p-values for these analyses. To establish an equal sample size of 29 subjects per group would have required us to either acquire data over a time interval that was three times longer, or conduct a multi-center study. This was a proof-of-principle study that was focused on what happens after injury and consequently the former was not possible because of finite resources and the later was not feasible because of concerns associated with the accuracy and reliability of multiple investigators at different sites evaluating entry criteria and assessing the condition of articular cartilage through arthroscopic visualization. Despite our findings, however, there is also literature suggesting that the concentration of ARGS-aggrecan within 90 days following injury is correlated with OA progression.^28,35 Given that ARGS-aggrecan has the potential to be an anabolic as well as a catabolic marker, future investigations with a combination of markers of synthesis and degradation are warranted. If the same approach is used to assign subjects to the low- and high-risk groups, this would require a multi-center study that has established accuracy and reliability with characterizing articular cartilage and meniscal lesions both within and between investigators. Such an effort was far beyond the scope of our first-step which was a proof-of-principle study that was focused on the events that occur soon after injury.

Strengths of this study include the use of a healthy control group that was matched to our ACL-reconstructed groups by age, sex, BMI, and activity level. While some may view the utilization of synovial fluid biomarkers as a concern given the difficulty of sample accessibility, our focus on synovial fluid markers allows for evaluation of joint specific biology that is more likely to reflect local environmental changes compared to those that appear in serum or urine.

This is the first study to suggest that synovial fluid levels of ARGS-aggrecan are significantly associated with patient reported outcomes of pain, symptoms, and function during sport as measured via the KOOS in high risk subjects with articular cartilage lesions that extend greater than 50% of cartilage thickness but less than 100% where calcified cartilage and subchondral bone are exposed (Grade 3A), and substantially weaker relationships in low risk subjects that have minor cartilage lesions that only involve the articular surface and extend down to less than 50% of cartilage thickness (Grade 2 or less). The results of the present study also confirm the existence of a significant association between decreasing synovial fluid levels of ARGS-aggrecan and worsening of pain as measured via the KOOS in the high risk subjects. Synovial fluid ARGS-aggrecan may be involved in a tissue repair response involving an increase in aggrecan synthesis in the acute phase following traumatic knee injury.