Extracellular Matrix-Blood Composite Injection Reduces Post-Traumatic Osteoarthritis After Anterior Cruciate Ligament Injury In the Rat

Abstract

Purpose

The objective of this study was to determine if an injection of a novel extracellular matrix scaffold and blood composite (EMBC) after anterior cruciate ligament (ACL) injury would have a mitigating effect on post-traumatic osteoarthritis (PTOA) development in rat knees.

Methods

Lewis rats underwent unilateral ACL transection and were divided into three groups: 1) no further treatment (ACLT; n = 10), 2) an intraarticular injection of EMBC on day 0 (INJ0; n = 11), and 3) an intra-articular injection of EMBC on day 14 (INJ14; n = 11). Ten animals received capsulotomy only (n = 10, SHAM group). The OARSI histology scoring of the tibial cartilage and micro-CT of the tibial epiphysis were performed after 35 days. The ratio of intact/treated hind limb forces during gait was determined using a variable resistor walkway.

Results

The OARSI cartilage degradation sum score and total degeneration width were significantly greater in the ACLT group when compared to the INJ0 (P = 0.031, and P = 0.005) and INJ14 (P =0.022 and P =0.04) group. Weight bearing on the operated limb only decreased significantly in the ACLT group (P = 0.048).

Conclusion

In the rat ACL transection model, early or delayed injection of EMBC ameliorated the significant decrease in weight bearing and cartilage degradation seen in knees subjected to ACL transection without injection. The results indicate that the injection of EMBC may slow the process of PTOA following ACL injury and may provide a promising treatment for PTOA.

Introduction

Patients suffering from an acute anterior cruciate ligament (ACL) injury are at significant risk for developing posttraumatic osteoarthritis (PTOA) with or without surgical reconstruction of the ACL.^{1; 2} Among the factors thought to initiate PTOA are the impact forces induced at time of injury,³ and altered joint kinematics and loading patterns,^{4; 5} elevated levels of inflammatory cytokines and proteases,^{6; 7} and reduced lubrication post injury.⁸ All of these processes may trigger catabolic responses in the joint that degrade the cartilage matrix and promote apoptosis.⁹ Considering that many irreversible changes occur within the first weeks after injury,^9–11 treatments to prevent cartilage damage and joint arthrosis, which are applicable at the time of injury, are of clinical interest.

Extracellular matrix (ECM) proteins, in particular intra-venous injection of type 2 collagen fragments, have been used to induce auto-immune osteoarthritis in animal models.¹² Therefore the use of ECM to prevent the development of PTOA might seem counterintuitive. However, a recent study in large animals demonstrated that direct intraarticular application of an ECM scaffold rich in collagen type 1 combined with autologous blood to form an Extracellular Matrix Blood Composite (EMBC) to enhance healing of the ACL led to less osteoarthritis.¹³ We posit that the addition of the ECM material might serve as a substrate for the matrix metalloproteinases released by synovial tissue after injury or surgery and thus protect the matrix within the articular cartilage. To our best knowledge, this is the first time an injectable ECM matrix hydrogel in combination with autologous blood has been used to treat post-traumatic osteoarthritis.

Blood contains platelets and cells capable of releasing growth factors which may have beneficial effects on cartilage health. In multiple animal models, injection of PRP mitigates inflammation and the progression of osteoarthritis.^14–16 There is also evidence of improved clinical symptoms in patients with osteoarthritis after injection of platelet rich growth factors.¹⁷ However, other studies reported no improvement in OA progression with PRP or whole blood injection.^{18; 19} In addition, the combination of blood with extracellular matrix proteins has been shown to promote a prolonged release of multiple platelet-associated growth factors.^{20; 21}

We hypothesized that the application of the EMBC either immediately or 14 days after ACLT would reduce the histologic cartilage degeneration of the medial tibial plateau, prevent trabecular bone changes in the medial compartment and reduce the development of gait asymmetry after ACL injury in an ACL transection model. Post-traumatic osteoarthritis in this model is evident within two weeks,²² and well developed by day 35.²³ Thus, euthanasia at 35 days was selected to provide sufficient time for the joint to develop post-traumatic osteoarthritis while minimizing the pain and distress of the animals.

Materials and methods

Animal Protocol

The study was approved by the Institutional Animal Care and Use Committee at Rhode Island Hospital and reported in accordance with ARRIVE guidelines for reporting animal research.²⁴ Forty two 12 week old male Lewis rats (weight 267 ± 12 g [mean ± standard deviation]) underwent gait analysis three-days after arrival to the facility. Animals were randomly assigned to four treatment groups using a random number generator (Table I). The control group received a sham operation on the right knee where only the knee capsule and synovium were incised but no ACL transection was performed (n = 10, SHAM group). In the ACL transection group (n = 10, ACLT group), the ACL of the right knees was transected without subsequent injection. The third group received the EMBC injection in the ACL transected knee immediately after the skin was closed (n = 11, INJ0 group). The delayed injection group received the same EMBC injection 14 days after surgery (n = 11, INJ14 group). Two animals were housed in each cage with free access to food and water. Animals received buprenorphine [0.03mg/kg body weight] twice daily for three days and were allowed to bear weight on limbs as tolerated. All surgeries and the injections at day 14 were performed under anesthesia as outlined below and were conducted on the right knee only while the left knees remained intact. On post-operative day 35, gait analysis was performed before animals were euthanized. Hind limbs were harvested and fixed in neutral buffered formalin. Micro-CT was then performed to determine bone density. Thereafter, limbs were decalcified and transferred into paraffin for histological analysis.

Table I.

Study outline of interventions and outcome measure assessments for each group.

EMBC – Extracellular Matrix Blood composite

SHAM – capsulotomy only without ACL transection group

ACLT – ACL transection group

INJ0 – ACL transection with immediate injection of whole blood extracellular matrix gel mix

INJ14 – ACL transection with injection of whole blood extra-cellular matrix gel mix injection at day 14 post surgery

ACL transection

Animals were anesthetized (Dexmedetomidine [0.25 mg/kg body weight] subcutaneous, Ketamine [25 mg/kg bodyweight] subcutaneous), the skin over the knee was shaved, aseptically prepared and an anteromedial arthrotomy performed. In groups assigned to undergo ACL transection, complete transection was performed with a #11 scalpel blade and verified by positive tibial anterior drawer test. The incision was closed and animals of INJ0 group received the EMBC injection before anaesthesia was discontinued (atipamezole [1 mg/kg body weight] subcutaneous).

Preparation and injection of the extracellular matrix gel/autologous whole blood composite

A hydrogel of extracellular matrix proteins was made by aseptically harvesting bovine connective tissue from the knee capsule, decellularization with a detergent solution and solubilization using pepsin in hydrochloric acid as previously described.²⁵ The resulting gel was neutralized using a buffer containing HEPES, sodium hydroxide, and calcium chloride to a pH of 7.4 and kept at 4°C until use. At time of injection, animals were anaesthetized and 50 μl of blood was drawn from the tail vein of the rat and mixed with acid-citrate-dextrose (ACD) in a ratio of 10:1. 25 μl of the anticoagulated blood was mixed with 25 μl of the neutralized hydrogel and 50 μl of the composite were injected into the operated knee through the patellar tendon using a 26 gauge needle each time while the knee was flexed. Confirmation of intraarticular administration was confirmed by noticeable and palpable SF collection.

Histological analysis

Paraffin embedded para-sagittal sections were obtained from weight bearing areas of the articular cartilage from the medial compartment of each treated joint as previously described.²⁶ Briefly, after fixation in 10% neutral buffered formalin for 3 days and subsequent Micro-CT, knees were decalcified with EDTA, embedded in paraffin and sectioned in the sagittal plane. Serial, 6 μm thick sections were taken, starting 200 μm laterally from the medial margin of the joint. Consecutive sections were stained with hematoxylin and eosin or safranin O red alternately. The location of analysis was standardized by observing the triangular shape of the anterior and posterior portions of the medial meniscus in serial sections just prior to the presence of articular cartilage of the patellar groove. The articular cartilage of the entire anterior to posterior length of the medial tibial plateau was divided into three similar sized regions for scoring.

The Osteoarthritis Research Society International (OARSI) scoring method, adapted for sagittal sections, was used to measure structural cartilage changes in the central weight bearing area of the medial tibial plateau in all samples (Fig. 1).^{27; 28} Briefly, cartilage matrix loss width, total cartilage degeneration width, significant cartilage degeneration width and osteophyte length were measured as defined below, while the cartilage degeneration and synovial reaction were graded semiquantitatively. Cartilage matrix loss was defined as areas with absent matrix, total cartilage degeneration as areas affected by any type of degenerative change including glycosaminoglycan (GAG) loss or matrix fibrillation/loss (affected area between black lines in [Fig. 1(A)] and significant cartilage degeneration as areas in which >50% of cartilage depth were affected by >50% chondrocyte loss with or without matrix loss (affected area between black lines in [Fig. 1(B)]). The cartilage matrix loss width was measured at the surface (0% depth – thick black line in [Fig. 1(Main Picture)], midzone (50% depth – thick grey line [Fig. 1(Main Picture)]) and tidemark (100% depth – thick white line [Fig. 1(Main Picture)]) and normalized to the projected total tibial plateau surface length. Osteophyte length, if present, was measured from base to top in μm (Length between two arrows in [Fig. 1(D)]). The cartilage degeneration sum score was generated by calculating the ratio of cartilage area affected by cartilage matrix loss, and/or >50% chondrocyte loss, and/or glycosaminoglycan loss (area between black lines was divided by the area between upper black line and subchondral bone in [Fig. 1(A)]).The cartilage degeneration sum score was graded 0 – 5 (0 = no degeneration, 1 = minimal [5 – 10% of projected cartilage area affected by matrix or >50% chondrocyte loss] degeneration, 2 = mild [11 – 25%] degeneration, 3 = moderate [26 – 50%] degeneration, 4 = marked [51 – 75%] degeneration, 5 = severe [>75%] degeneration) and a score for anterior, medial and posterior region was generated and combined into a single cartilage degeneration sum score. The grade of synovial membrane inflammation (Grade 0 = no changes, Grade 1 = >3–4 lining cell layers or slight proliferation of subsynovial tissue, Grade 2 = >3–4 lining cell layers and proliferation of subsynovial tissue, Grade 3 = >4 lining cell layers and proliferation plus infiltration of subsynovial tissue with inflammatory cells, Grade 4 = >4 lining cell layers and proliferation plus infiltration of subsynovial tissue with a large number of inflammatory cells) was determined in a defined area proximal to the posterior horn of the medial meniscus (in [Fig. 1(C)]). Artificial defects caused by histological processing/cutting of the paraffin blocks and were excluded from the measurements (black arrows in [Fig. 1(Main Picture)]). All samples were assessed independently by two experienced observers blinded to both limb and treatment groups. The two observers reached consensus on each section as previously been reported.²³

Fig. 1.

Histological Scoring. (Main Picture) Sagittal section through central part of the medial tibial plateau [MTP] stained using Safranin-O staining for glycosaminoglycans [GAG]. Lines at superficial [black], midzone [grey] and tidemark level [white] . Thick lines indicate defects. Black arrows indicate artificial defects caused by histological processing. (A) Total Cartilage Degeneration. Magnified section of anterior third of MTP. Area between black lines indicates total cartilage degeneration [Cartilage Matrix loss, and/or >50% chondrocyte loss, and/or Glycosaminoglycan loss]. (B) Significant Cartilage Degeneration. Magnified section of posterior third of MTP. Area between black lines indicates significant cartilage degeneration [>50% of cartilage depth is affected by %50 chondrocyte loss with or without matrix loss]. Short thick grey line indicates course of cartilage matrix loss at midzone. (C) Synovial Inflammation Score. Magnified area of synovium used to generate synovial inflammation score in all samples. Scoring was done in H&E stained slides. Current example shows 3–4 lining cells [black arrows] and proliferation of subsynovial tissue [white arrows] [= Grade 2]. (D) Osteophyte Length. Length of present osteophytes was measured from base to top in [μm].

Gait analysis

Differences in weight bearing (normalized to body weight) between affected and unaffected limbs were analyzed before surgery (day0) and at day 35 post surgery using techniques previously reported.²⁹ Briefly, paw pressure and impulse were analyzed over two gait cycles on a pressure sensing walkway (Animal Walkway System (VH3); Tekscan, Boston, MA, USA) using the associated operating system software 7.0 with a paw strike resolution of 15.5 sensels/cm². Data consisted of consecutive frames collected at 100 Hz. Maximum paw pressure and stance time were registered for each gait cycle and maximum force per limb was measured on 4 trials. Changes from day 0 to day 35 in differential weight bearing between affected and unaffected limbs were analyzed for each group individually.

Micro-CT

All hind-limbs were dissected and formalin fixed for 24 hours before micro-CT. Hind-limbs were potted in 2% agarose gel to prevent dehydration and movement during scanning. Axial scans were performed using the MicroCT40 scanner (Scanco Medical AG, Switzerland) at tube settings of 70 kVp and 114 μA with a tube diameter of 30 mm (integration time: 300 ms). Sagittal slices were reconstructed at an isotropic resolution of 30 μm. Trabecular bone of the tibial epiphyseal region was defined as region of interest (ROI), and the trabecular structure was delineated using semi-automated contouring algorithm (provided by Scanco Medical AG, Switzerland). Outlined trabecular ROI was filtered using Gaussian filter (sigma 0, support 0) and segmented with a global bone threshold of 420.3 mg HA/cm³. Trabecular bone was analyzed for bone volume fraction (BV/TV, %), density (mg HA/cm³), thickness (mm), and number (1/mm), using manufacturer provided software.

Statistical analysis

The power was determined a priori and based on the gait asymmetry measurements in the rat knee and OARSI scoring data obtained following ACLT with intra-articular lubricin injection compared to saline injection or no treatment in the rat knee^{23; 29}. Given these estimates, approximately 40 animals total (n=10 per group) would provide 82% power with respect to gait asymmetry and 99% power with respect to the OARSI score.

Generalized estimating equations were used to model assay values as a function of treatment group for all outcome measures. Normal distribution of gait analysis, micro-CT and OARSI histologic grading and staging scores were analyzed for normality using the Kolmogorov-Smirnov test. For normally distributed data, one-way analysis of variance (ANOVA) was performed with post-hoc Bonferroni testing to correct for multiple comparisons. For non-normally distributed or ordinal scaled data sets, Kruskal Wallis test with Dunn post-hoc test correction for multiple comparisons was used. An alpha of 5% was considered significant. All data are presented as means with 95% confidence intervals. All analyses were performed using intercooled STATA 13 (StataCorpLP, College Station, TX).

Results

Animal welfare

One animal of the INJ14 group died while under anesthesia. All other animals returned to normal activity and partial weight bearing on the surgical limb shortly after waking up from anesthesia. No signs of local infection were seen at any point. All animals gained approximately the same weight during the 35 day period (78.4 ± 16.8 g [mean ± standard deviation]). After euthanasia, lateral subluxation of the patella (patella derailed laterally from the femoral trochlear groove during knee flexion) was found in 4 animals of the ACLT, 5 animals of the INJ0, and 2 animals of the INJ14 group.

Histological analysis: Cartilage

All cartilage-related outcome measures are presented in Table II, while representative histological images are shown in Fig. 2.

Table II.

Histological Analysis (presented as mean and 95% confidence interval) showing results for cartilage matrix loss width at surface, midzone and tidemark level, total and significant cartilage degeneration width, cartilage degeneration sum score, synovial inflammation score and osteophyte length.

	Groups	Cartilage Matrix Loss Width Surface		Cartilage Matrix Loss Width Midzone		Cartilage Matrix Loss Width Tidemark		Total Cartilage Degeneration Width		Significant Cartilage Degeneration Width		Cartilage Degeneration Sum Score		Synovium Inflammation Score		Osteophyte Length [μm]
		Mean	95CI	Mean	95CI	Mean	95CI	Mean	95CI	Mean	95CI	Mean	95CI	Mean	95CI	Mean	95CI
Absolute Values	SHAM (n = 10)	0.65	0.57	0.01	0.0	0.01	0.0	0.68	0.60	0.1	0.03	5.0	4.0	0.9	0.4	47	0
			0.73		0.04		0.03		0.75		0.16		6.0		1.4		118
	ACLT (n = 10)	0.83	0.80	0.09	0.02	0.03	0.0	0.84	0.80	0.20	0.13	7.5	7.0	2.6	2.0	218	97
			0.85		0.17		0.07		0.87		0.27		8.0		3.2		338
	INJ0 (n = 11)	0.71	0.65	0.04	0.01	0.03	0.0	0.71	0.65	0.13	0.09	6.1	5.3	2.5	1.7	97	26
			0.77		0.08		0.05		0.77		0.17		6.9		3.4		168
	INJ14 (n = 11)	0.73	0.67	0.04	0.01	0.03	0.0	0.73	0.69	0.15	0.11	6.0	5.3	3.1	2.7	189	90
			0.78		0.06		0.06		0.78		0.19		6.7		3.5		288
Comparison Padj-Values	SHAM vs. ACLT	.001▲		0.53		1.0		.001▲		.038▲		.001▲		.001▲		.035▲
	SHAM vs. INJ0	.738		1.0		1.0		1.0		1.0		.161		.001▲		1.0
	SHAM vs. INJ14	.287		1.0		1.0		.688		.608		.276		.001▲		.119
	ACLT vs INJ0	.022▲		.485		1.0		.005▲		.348		.031▲		1.0		.242
	ACLT vs. INJ14	.096		.303		1.0		.040▲		1.0		.022▲		1.0		1.0
	INJ0 vs. INJ14	1.0		1.0		1.0		1.0		1.0		1.0		.957		.685

^▲Statistically significant difference (P< .05) between groups

95% CI – 95% confidence interval

Cartilage Matrix Loss Width Surface – cartilage matrix loss width at surface level normalized to total length of tibial plateau (0% depth)

Cartilage Matrix Loss Width Midzone – cartilage matrix loss width at midzone level normalized to total length of tibial plateau (50% depth)

Cartilage Matrix Loss Width Tidemark – cartilage matrix loss width at tidemark level normalized to total length of tibial plateau (100% depth)

Total Cartilage Degeneration Width - width of cartilage affected by any type of degenerative change – GAG loss, matrix fibrillation/loss - normalized to the total length of tibial plateau

Significant Cartilage Degeneration Width - length at which >50% of cartilage depth is affected by >50% chondrocyte loss with or without matrix loss - normalized to the total length of tibial plateau

Cartilage Degeneration Sum Score – cartilage degeneration sum score (grading 0 – 5; 0 = degeneration, 1 = minimal [5 – 10% of projected cartilage area affected by matrix or >50% chondrocyte loss] degeneration, 2 = mild [11 – 25%] degeneration, 3 = moderate [26 – 50%] degeneration, 4 = marked [51 – 75%] degeneration, 5 = severe [>75%] degeneration)

Synovium Inflammation Score - grade of synovial membrane inflammation (Grade 0 = no changes, Grade 1 = >3–4 lining cell layers or slight proliferation of subsynovial tissue, Grade 2 = >3–4 lining cell layers and proliferation of subsynovial tissue, Grade 3 = >4 lining cell layers and proliferation plus infiltration of subsynovial tissue with inflammatory cells, Grade 4 = >4 lining cell layers and proliferation plus infiltration of subsynovial tissue with a large number of inflammatory cells)

Osteophyte Length – length of present osteophytes in [μm]

SHAM – capsulotomy only without ACL transection group

ACLT – ACL transection group

INJ0 – ACL transection with immediate injection of extracellular matrix blood composite

INJ14 – ACL transection with injection of extracellular matrix blood composite at day 14 post surgery

Fig. 2.

Representative Sagittal Sections through Central Part of Medial Tibial Plateau Stained using Safranin-O Staining for GAG [40X Magnification; Black Bar = 200μm]. (A) Representative section of the SHAM group with relatively smooth superficial cartilage layer with minor GAG [white arrows] and cartilage matrix loss [black arrows] in posterior section of tibial plateau. (B) Representative section of the ACLT group with rugged cartilage surface showing GAG [white arrows], chondrocyte and matrix loss [black arrows] over a significant stretch of medial tibial plateau. (C) Representative section of the INJ0 group with smoother surface than ACLT, minor GAG loss in anterior section [white arrows], and matrix loss [black arrows] in posterior section of medial tibial plateau. (D) Representative section of the ACLT group with almost similar appearance as INJ0, minimally increased loss of GAG in central portions of medial tibial plateau [white arrows] and similar matrix loss [black arrows] in posterior section of medial tibial plateau.

Injury-induced changes

In ACLT significantly deteriorated scores were observed in 4 out of 6 assessed cartilage-related parameters, when compared to SHAM (all depicted in Fig. 3). Scores included cartilage matrix loss at the surface level, total degeneration width, significant degeneration width and degeneration sum score (P = 0.001, 0.001, 0.038 and 0.001 respectively). No significant changes were observed regarding matrix loss width at midzone or at tidemark level (P = 0.53 and 1.00, respectively).

Fig. 3.

(A) Cartilage Matrix Loss Width at Surface Level [normalized to total length of cartilage surface]. ACLT samples showed statistically significantly longer stretches of cartilage matrix loss at surface level. (B) Total Cartilage Degeneration Width. Length of glycosaminoglycan, chondrocyte, or matrix loss at surface level [normalized to total length of cartilage surface]. ACLT samples showed statistically significantly longer stretches of total degeneration at surface level. (C) Significant Cartilage Degeneration Width. Length of chondrocyte/matrix loss below 50% of total cartilage depth [normalized to total length of cartilage surface]. ACLT samples showed statistically significant longer stretches of significant degeneration compared to SHAM control samples. (D) Cartilage Degeneration Sum Score. Anterior, medial, and posterior zone was scored based on area affected by Glycosaminoglycan, chondrocyte, or matrix loss and scores combined to degeneration sum score [see methods section for grading]. ACLT samples showed statistically significantly higher scores.

Effect of treatment

In comparison to ACLT, INJ0 demonstrated significant mitigation of 3 of the 4 injury-induced changes. Those changes included cartilage matrix loss at surface level, total degeneration width and degeneration sum score (P = 0.022, 0.005 and 0.031, respectively), but not degeneration width (P = 0.348). INJ14 resulted in mitigation of 2 of the 4 injury-induced changes: total degeneration width and degeneration sum score (P = 0.040 and 0.022), but not cartilage matrix loss at surface level and significant degeneration width (P = 0.096 and 1.0).

Histological analysis: Synovitis

Grades of synovial membrane inflammation of all groups are presented in Table II.

Injury-induced changes

In ACLT a significant increase in synovial membrane inflammation was observed when compared to SHAM (P = 0.001).

Effect of treatment

In comparison to ACLT, neither INJ0 nor INJ14 demonstrated a significant alteration of synovial membrane inflammation (both P = 1.0).

Histological analysis: Osteophytes

Osteophyte lengths of all groups are presented in Table II.

Injury-induced changes

In ACLT, a significant increase in osteophyte length was observed when compared to SHAM (P = 0.035). Osteophyte formation was exclusively observed at the posterior tibial plateau.

Effect of treatment

In comparison to ACLT, neither INJ0 nor INJ14 demonstrated a significant alteration of the extent of osteophyte formation (P = 0.242 and 1.0, respectively).

Micro-CT analysis: Subchondral bone

Both subchondral bone related measures are presented in Fig. 4(A)(B).

Fig. 4.

Subchondral Bone and Gait Asymmetry Analysis. (A) Subchondral Bone Volume to Trabecular Volume Ratio (BV/TV) Difference of Intact to Surgical Side using Micro-CT. Loss in subchondral bone to trabecular volume of surgical limb compared the intact limb was seen in all ACLT groups [BV/TV; [%]]. This was highly statistically different to the SHAM [Group with Capsulotomy but no ACL transection or intervention] control group [P = 0.001]. (B) Subchondral Bone Density Difference of Intact to Surgical Side using Micro-CT. Loss in subchondral bone density of surgical limb compared the intact limb was seen in all ACLT groups [mg HA/cm³]. This was highly statistically different to the SHAM control group [P = 0.001]. (C) Changes in Differential Hind Limb Load [difference between surgical and intact hind limb load] from day 0 to day 35 [in % body weight]. Change of the differential hind limb maximal load during gait expressed in percentage of body weight from day 0 to day 35 was only significant in the ACLT group [P = 0.048], where the surgical limb was significantly unloaded compared to the intact limb at day 35.

Injury-induced changes

In ACLT, a significantly greater reduction of trabecular bone volume fraction and trabecular bone density between surgical and intact limb was observed when compared to SHAM (both P = 0.001). Additionally, the surgical limb lost 17.7% of trabecular thickness and 12.6% of trabeculae (both P = 0.001)

Effect of treatment

In comparison to ACLT, neither INJ0 nor INJ14 significantly mitigated the reduction of trabecular bone volume fraction, trabecular bone density, trabecular thickness or trabecular number (P = 1.0).

Functional analysis: Gait

Changes in differential limb load from pre- to day 35 post-surgery are depicted in Fig. 4(C).

Injury-induced changes

In ACLT, a significant increase in differential limb loads from pre- to day 35 post-surgery, corresponding to reduced loading of the surgical limb, was observed (P = 0.048). In the SHAM group, no changes in differential limb loads were observed when comparing pre- and post-surgical measurements (P = 0.464).

Effect of treatment

In the INJ0 and INJ14 groups, no significant increases in differential limb load pre- to day 35 post-surgery were detected (P = 0.438 and P = 0.184, respectively).

Discussion

In this study, rats that received ACL transection without further treatment had significant superficial cartilage matrix loss, significant cartilage degeneration and osteophyte formation; however, these changes were not seen when rats were treated with EMBC. In addition, rats that received ACL transection alone exhibited significant decreases in weight bearing on the injured leg. In contrast, rats with immediate or delayed injection of EMBC after ACL transection did not have a significant change in weight bearing five weeks after injury in comparison to sham animals (capsulotomy). These findings suggest that an injection of EMBC may ameliorate some of the structural and functional features associated with PTOA seen after ACL injury.

The chondroprotective properties of the EMBC could be related to proteins in either the hydrogel or the blood. One of the major components of the gel is type I collagen. Prior reports showed that intra-articular injections of various collagen-based materials reduced instability- and antigen-induced arthritis in rats by acting as a lubricant.^{30; 31} Naraoka et al. injected collagen tripeptide – a purified tripeptide fraction containing Gly-Xaa-Yaa sequences – weekly into ACL transected rabbit knees and reported significantly decreased collagen type II loss in the cartilage matrix of the injected knees compared to ACL transected knees without further treatment.³² A study in human osteoarthritis patients conducted by Furuzawa et al. registered significant clinical improvement in pain and stiffness scores after 48 weeks using an injection of polymerized type I collagen.³³ A potential explanation for the positive effect of the injected collagen could be that the injected collagen acts as a competitive substrate for the cartilage damaging matrix metalloproteinases released into the synovial fluid after joint injury or surgery, thus protecting the extracellular matrix proteins of the articular cartilage. Further research is needed to better understand the mechanism of action.

The study limitations include the use of the quadruped rat model, which differs significantly in size and biomechanics from the human. Quadruped biomechanics may especially play a role during gait analysis, and could be influenced by a weight shift from hind to front limbs. The analysis performed was limited to the maximum force difference between the hind limbs. Secondly, this study used the ACL transection model. While this is a model previously shown to result in PTOA in the rat,²³ it is not reflective of the human condition, where the ACL is rarely torn without additional bone, cartilage or meniscal trauma. A third limitation was that the rats were 10 weeks of age when they underwent surgery while skeletal maturity occurs at the age of 12 weeks.²⁷ This is likely to be a minor concern as the progression of OA may be slower in immature as compared to skeletally mature animals.²⁷ Another limitation was that the 35 day time period for disease progression was also relatively short.²⁷ The shorter time period may be responsible for the lack of matrix loss in the lower cartilage zones. However, cartilage damage is evident in the rat ACL transection model within two weeks,³⁴ and moderate changes were observed in our study at 5 weeks. In addition, a sham control with saline injection was not included. We chose to use a control of ACL transection alone without a saline injection as prior studies have demonstrated that control injections with saline lead to greater PTOA progression than no injection after ACL transection.²⁹ Therefore, the non-injection ACL transection group was chosen as a more stringent control. Another limitation of this study was that the rat ACL transections were not surgically repaired. It is possible that stabilization of the knee with a repair or reconstruction might also minimize the development of PTOA in the rat model. Future studies to determine if there is a synergistic effect of the mechanical stabilization with ACL repair and the addition of the EMBC are planned. Additionally, animals with a lateral patella subluxation observed during limb harvest were not excluded from the study, since lateral patella subluxation was not defined ad hoc as exclusion criterion. Post hoc analyses of the data showed no significant differences between knees with or without subluxation in any of the categories analyzed. A final limitation is that we were unable to define the exact molecular or cellular mechanism of action of the EMBC therapeutic. Now that we have determined that the use of the EMBC injectable re-establishes relatively normal knee function in vivo in terms of gait analysis, and minimizes the cartilage degradation of the joint after ACL transection, future studies to define the exact biologic processes which cause this functional result are planned.

Conclusion

The results of improved gait symmetry and lower rates of cartilage damage in the rat ACL transection model indicate that the injection of the EMBC hydrogel may decrease the pain during weight bearing and cartilage damage associated with post-traumatic osteoarthritis after ACL transection in the rat model. A delayed injection, which was added to mimic the deferred appearance of the average patient to the clinic after ACL injury,³⁵ showed potential benefits over the untreated ACL transection group. Further studies are needed to optimize the EMBC concentrations and dosing strategies are needed to optimize these outcomes. Nonetheless, the EMBC may provide a promising treatment for post-traumatic osteoarthritis.