Synovial Fluid Inflammatory Profiles did not Differ between Isolated Anterior Cruciate Ligament and Multi-ligament Knee Injuries

Objective and design The purpose of this study was to compare synovial concentrations of pro- and anti-inflammatory cytokines between multiple-ligament knee injured (MLKI) and anterior cruciate ligament (ACL)-injured patients. Subjects 14 patients with MLKI and 10 patients with isolated ACL injury Methods Synovial fluid was aspirated from the operative knee at the time of surgery and the concentrations of pro- and anti-inflammatory markers in the synovial fluid were determined. Structures injured, the time between injury and surgery, and demographic factors were collected. Linear regressions were used to determine the association between injury pattern and synovial inflammatory markers when controlling for age, BMI, and time between injury and surgery. Results When adjusting for group differences in age, BMI and the time between injury and surgery, no group differences were detected (interferon gamma (p = 0.11), interleukin-1beta (IL-1b, p = 0.35), IL-2 (p = 0.28), IL-4 (p = 0.64), IL-6 (p = 0.37), IL-8 (p = 0.54), IL-10 (p = 0.25), IL-12p70 (p = 0.81), IL-13 (p = 0.31), and tumor necrosis factor-alpha (p = 0.90)). Conclusion Synovial fluid inflammatory markers did not differ between MLKI and isolated ACL injuries. MLKIs have a greater prevalence of postoperative arthrofibrosis when compared to isolated ACL injuries; however, this may be due in part to factors other than the inflammatory status of the joint.


Introduction
Arthro brosis, a brotic joint disorder characterized by excessive collagen production and adhesions, is a known potential complication after knee surgery and often results in pain and restricted range of motion. [27] Pro-in ammatory cytokines are involved in the development of arthro brosis by inducing proliferation and differentiation of broblasts to myo broblasts. [3] Myo broblasts are contractile and apoptosis of myo broblasts is necessary for wound healing. In the absence of apoptosis of myo broblasts, the joint may experience a build-up of contractile brotic scar tissue thereby limiting mobility.[16, 25,27] As such, pro-in ammatory cytokines and T-cells are thought to mediate the pathogenesis of arthro brosis. [1,16,18,19,25,27] The prevalence of postoperative arthro brosis is greater following surgical treatment of multi-ligament knee injury (MLKI) than with isolated anterior cruciate ligament (ACL) injuries. [12,13,17] It has been proposed that the increased prevalence of postoperative arthro brosis after MLKIs compared to isolated ACL injuries may be due to increased in ammation secondary to greater trauma to the joint. [9,20] The purpose of this study was to compare synovial concentrations of pro-and anti-in ammatory cytokines between patients that have a MLKI or isolated ACL injury. We hypothesized that MLKI would demonstrate a more pro-in ammatory state than those with an isolated ACL injury at the time of reconstruction.

MLKI Patients
This was an ancillary study performed in conjunction with the Surgical Timing and Rehabilitation Trial for Multiple Ligament Knee Injuries (STaR Trial; NCT03543098). The current study shared the same inclusion and exclusion criteria as the STaR Trial but was treated as a separate protocol. Eligible patients were rst approached regarding participation in the STaR Trial; however, participation in the current study was not dependent on STaR Trial participation. Those that opted not to take part in the STaR Trial could participate in the current observational study. This was an exploratory analysis to be used for future hypothesis generation and as such, an a priori sample size calculation was not performed.
Fifteen patients with a clinical and MRI-con rmed MLKI consented to participate in this IRB-approved prospective multi-center study. Participants were recruited between November, 2019 and March, 2021 from two academic medical centers with both being the only Level 1 trauma center in their respective regions. Males and females between the ages of 16 and 55 with a MLKI (de ned as a complete grade III injury of two or more ligaments) were eligible to participate. Individuals with a nerve injury or biceps or popliteus tendon rupture/avulsion were not excluded. The following exclusion criteria were employed: a history of prior knee ligament surgery of the involved knee, a torn or avulsed patellar or quadriceps tendon, any condition that precluded early post-surgical weight bearing or range of motion exercise within 7 days of surgery (such as a periarticular or long bone fracture, a skin or soft tissue injury, surgery for extensor mechanism rupture or avulsion, or vascular graft surgery), a condition that required use of an external xator to maintain reduction of the knee or soft tissue/open wound management for greater than 10 days, were unable to weight bear on the contralateral uninjured leg, a traumatic brain injury that limited their ability to participate in their postoperative care or any other condition that would preclude the ability to comply with postoperative guidelines. ACL Patients: Samples collected from the MLKI patients were compared to synovial uid samples collected from 10 patients with isolated ACL injuries that were in the placebo control group from a previous IRB-approved randomized controlled trial (NCT01692756). Patients were included if they met the following criteria: clinical con rmation of an isolated ACL injury (positive Lachman test with intraarticular effusion, normal varus-valgus test), between the ages of 14 and 32, had an isolated ACL injury (no more than a grade 1 MCL injury de ned clinically), no history of previous ACL injury (ipsilateral or contralateral knee), injury that occurred during performance of a sporting activity, and growth plates that were closed as documented by plain x-ray. Patients were excluded if they had a history of any in ammatory disease, immune compromise, chronic use of NSAIDs, history of intra-articular cortisone injection into either knee within the prior three months, or prior knee surgery. Patients with a posterior cruciate ligament injury or more than grade I injuries to either collateral ligament were excluded; however, patients were not excluded based on meniscal status.

Data Collection and Analysis
Details of the clinical knee dislocation classi cation (KD grade), [21] the time between injury and surgery, and demographic factors (age, sex, BMI) were collected. The KD classi cation criteria are presented in Table 1. Joint aspiration was performed by the treating surgeon in the operating room after the knee had been prepped and draped aseptically through a superolateral parapatellar approach. Aspirated synovial uid was spun at 3500 RPM for 10 minutes and the supernatant was frozen at -80°C. Commercially available ELISA kits (Proin ammatory Panel 1, Meso Scale Diagnostics) were used to determine concentrations of pro-and anti-in ammatory markers in the synovial uid including interferon gamma (IFN-γ), interleukin-1 beta (IL-1b), IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, IL-13, and tumor necrosis factor alpha (TNF-a). Assays were completed per the manufacturer's instructions and were run in duplicate. For all plates, intra-assay coe cients of variance were < 9.5. For values that were below the lower limit of detection, ½ LLOD (lowest level of detection) were reported for statistical purposes. To avoid any potential bias, samples were analyzed by the biomarker assessment core laboratory at one of the participating universities.

Statistical Analyses
None of the variables assessed were normally distributed, and as such, synovial biomarkers, injury and patient factors were compared between patients with a MLKI or ACL injury using Mann-Whitney U tests.
To account for potential group differences in age, BMI, and the time since injury, biomarker values were log transformed to allow linear regression analyses to determine the association between injury pattern and synovial in ammatory markers when controlling for these patient and injury characteristics. Pearson Product Moment correlations were used to assess the associations between log transformed biomarker concentrations and age, BMI, and time between injury and surgery independent of injury type. Finally, within the MLKI group, we compared synovial biomarker concentrations between KD I and KD III injuries.
Due to the small sample size for this exploratory subgroup analysis, we did not perform statistical tests comparing the groups. Rather, effect sizes were calculated to aid in interpretation of this exploratory subgroup analysis. Effect sizes for two-sample nonparametric data were calculated by dividing the Z score by the square root of the sample size. Effect sizes < 0.30 are considered small, 0.30 and 0.49 as average, and ≥ 0.50 considered large.

Results
Synovial uid was successfully aspirated from the injured joint at the time of surgery for 14/15 MLKI patients and all 10 ACL patients ( Table 2). Of the patients with a MLKI, 9 were KD I, three were KD IIIM (medial) and 2 were KD IIIL (lateral). The MLKI group was signi cantly older with a greater BMI than the ACL group (Table 2). Unadjusted biomarker concentrations did not differ between groups, nor were group differences noted when controlling for age, BMI, and time between injury and reconstructive surgery (pvalues ranging from 0.11 to 0.90; Table 3). For the most part, time from injury, age, and BMI were not independently associated with synovial in ammatory markers with the one exception of age being signi cantly associated with IL-2 (Table 4). In the exploratory comparison between injury patterns, 8/14 (57.1%) patients were reconstructed within 3 months of injury, 3/14 (21.4%) were reconstructed between 3 and 6 months after injury, and 3/14 (21.4%) were chronic injuries that were treated more than 6 months after injury (Table 5). There were no marked differences in biomarker concentrations between KD I and KD III injuries as no large effect sizes were noted (Table 6).

Discussion
Contrary to our hypothesis, synovial markers of in ammation did not differ between patients with a MLKI or isolated ACL injury. Patients with a MLKI were signi cantly older with greater BMIs, and there tended to be greater time between injury and surgery (p = 0.11) than the isolated ACL group. Because increased age, BMI, and time between injury and surgery have been associated with reduced risk of arthro brosis, [9,14,17,20] these factors were statistically controlled for in our analyses. However, by and large these three factors were not independently associated with synovial in ammatory markers at the time of surgery (Table 3). Historically, operating too soon after ACL injury was thought to increase the risk of arthro brosis secondary to the joint being acutely in amed. [22,23] However, over the past 25 years, there has been a gradual shift to ACL reconstruction earlier after injury as multiple studies and systematic reviews have demonstrated that earlier ACL reconstruction does not increase the risk of arthro brosis. [7,13,26] The lack of difference in in ammatory pro les between MLKI and ACL injuries in the current pilot study then raises the provocative question of whether earlier surgery might also be a safe option for patients with a MLKI.
Just as intriguing as the lack of difference in in ammatory markers between MLKI and isolated ACL injuries was that the exploratory subgroup results also suggest that there may not be dramatic differences in the intraarticular in ammatory pro le between KD I and KD III injuries. If the severity of injury is not associated with a greater cytokine burden at the time of surgery, then the increased prevalence of arthro brosis after MLKI may be due, at least in part, to factors other than the in ammatory status of the joint. It is unknown whether the postoperative in ammatory burden secondary to the surgical insult and/or patient demographic or socioeconomic factors that may alter access to or willingness to comply with physical therapy may in uence the increased prevalence of arthro brosis after MLKI.
In the current study, the synovial uid concentrations of in ammatory markers did not differ between patients with an isolated ACL or MLKI patients at the time of surgery. However, the surgical insult of ACL reconstruction reinitiates the in ammatory process with synovial uid concentrations of proin ammatory cytokines increasing sharply 1 week after surgery. While cytokine concentrations lessen by 4 weeks after ACL reconstruction, they remain greater than what was observed within 10 days of the initial injury. It is unclear if this represents the joint's response to a second large in ammatory event over a relatively short period of time, or if the surgical insult itself is more traumatic than the initial injury. If the latter is true, then perhaps the additional surgical procedures associated with MLKI might result in increased postoperative in ammation and cytokine burden relative to isolated ACL reconstruction, which could contribute to the greater prevalence of arthro brosis after MKLI.
Both pre-and postoperative rehabilitation may contribute to the increased prevalence of arthro brosis after MLKI as well. Preoperatively, limited range of motion, effusion, and swelling have been previously reported to be associated with arthro brosis following ACL reconstruction. [5,16,22] It is important to note that the volume of swelling in the joint is not synonymous with increased cytokine concentrations, as the volume aspirated from the joint was previously reported to not differ between ACL patients with normal versus dysregulated in ammatory responses to injury. [10] Mechanistically, increased cytokine activity may increase brosis thereby resulting in arthro brosis whereas the volume of swelling may physically hinder the ability to fully ex or extend the knee. Additionally, a joint effusion may result in arthrogenic inhibition of the quadriceps limiting active extension. Similarly, while postoperative rehabilitation protocols after MLKI will differ based on the structures involved and procedures performed, one of the primary goals of the early postoperative rehabilitation is to restore knee range of motion and activation of the quadriceps without over-stressing the involved tissues. [15] Until recently, many rehabilitation protocols advocated immobilization of the knee after surgery for MLKI for anywhere from 1 to 6 weeks; [11] however, early initiation of range of motion has led to better range of motion outcomes without negatively affecting stability. [17] As such, reducing effusion and safely restoring range of motion and activating the quadriceps both prior to and following surgery are imperative to reduce the risk of arthro brosis no matter if treating a single or multiple ligament injury.
Patient factors may also in uence the increased prevalence of arthro brosis after MLKI. In the current study, patients with a MLKI were older with increased BMI. While both of these factors may potentially lessen the risk of arthro brosis, [9,14] other differences between MLKI and ACL patient populations may be involved. The majority of MLKIs are the result of traumatic mechanisms such as motor vehicle accidents rather than sports injuries. [2,20] Among other factors, there are differences in socioeconomic status, level of education, and insurance provider between orthopaedic trauma and sports medicine patients, with orthopaedic trauma patients also being more often male. [4,24] Male sex, lower education, low socioeconomic status, and public insurance are associated with reduced physical therapy utilization.
[6, 28] Population differences that impact access to and/or the ability to comply with postoperative rehabilitation may in uence the increased prevalence of arthro brosis after MLKI compared to isolated ACL injury.

Limitations
The results of this pilot study should be used for hypothesis generation. The sample size was too small to assess the relationship between in ammatory markers and arthro brosis. We attempted to account for potential confounders and analyzed the synovial uid data using both nonparametric methods as well as with regressions to statistically account for group differences in age, BMI, and time between injury and surgery. However, there are statistical limitations of including 3 confounding variables in our regression analyses with a relatively small sample size. Larger studies are necessary to mitigate these limitations and further elucidate the biological, rehabilitation, and patient-related factors that may contribute to the increased risk of arthro brosis following MLKI.

Conclusions
Contrary to our hypothesis, synovial markers of in ammation did not differ between MLKI and isolated ACL injuries. It has been well established that MLKI have a greater prevalence of postoperative arthro brosis when compared to isolated ACLs; however, this may be due in part to factors other than the in ammatory status of the joint. The lack of difference in in ammatory pro les between MLKI and ACL injuries raises the question of whether earlier surgery might be a safe option for MLKI. However, the biological and/or patient-related factors that contribute to increased arthro brosis after MLKI have yet to be elucidated. Future research will include examination of these markers, analysis of arthro brotic knees, and the etiologies of the stiff knee joint.

Declarations
Disclosure of potential con icts of interest

Informed consent
All patients provided written informed consent prior to participating in this study.
IRB Statement: The study was approved by the University of Kentucky and University of New Mexico IRBs and all patients provided informed consent before taking part in any study activities.
Disclaimer: None Sources of Funding/Role of Sponsor: This research was supported by the NIH National Center for Advancing Translational Sciences through grant number UL1TR001998 at the University of Kentucky and UL1TR001449 at the University of New Mexico. The content is solely the responsibility of the authors and does not necessarily represent the o cial views of the NIH, the University of Kentucky, or the University of New Mexico. The study sponsors had no involvement in the study design, collection, analysis and interpretation of data; in the writing of the manuscript; or in the decision to submit the manuscript for publication.