Carbodiimide Derivatized Synovial Fluid for Tendon Graft Coating Improves Long-Term Functional Outcomes of Flexor Tendon Reconstruction

Abstract

Background:

Flexor digitorum profundus (FDP) tendon injury is common in hand trauma, and flexor tendon reconstruction is one of the most challenge procedures in hand surgery due to severe adhesion that exceed 25% and hinders hand function. Surface property of graft from extrasynovial tendons is inferior to the native intrasynovial FDP tendons, which has been reported as one of the major causations. Improve surface gliding ability of extrasynovial graft is needed. Thus, this study was to use carbodiimide derivatized synovial fluid and gelatin (cd-SF-gel) to modify the graft surface thus improving functional outcomes using a dog in-vivo model.

Methods:

40 flexor digitorum profundus tendons (FDP) from the second and fifth digits of 20 adult female underwent reconstruction with peroneus longus (PL) autograft after creation of tendon repair failure model for six weeks. Graft tendons were either coated with or without de-SF-gel (n=20). Animals were sacrificed 24 weeks following reconstruction and digits were collected post-sacrifice for biomechanical and histological analyses.

Results:

Adhesion score (cd-SF-Gel 3.15±1.53, control 5±1.26 (p<0.00017)), normalized work of flexion (cd-SF-gel 0.47 N-mm/degree±0.28, control 1.4 N-mm/degree±1.45 (p<0.014)), DIP motion (cd-SF-gel (DIP 17.63⁰±6.77⁰, control (DIP 7.07⁰±12.99⁰) (p<0.0015)) in treated graft all showed significant differences compared to non-treated graft. However, there was no significant difference in repair conjunction strength between two groups.

Conclusions:

Autograft tendon surface modification with cd-SF-Gel improves tendon gliding ability, reduces adhesion formation, and enhances digit function without interfering with graft-host healing.

Introduction

Flexor digitorum profundus (FDP) tenodn injury is one of the most common hand trauma, with success rates for repaired FDP tendons around 75–80%^1,2. About 5–10% of flexor tendon injury need a tendon reconstruction to restore tendon function because of large defects or when repair fails due to rupture or severe adhesion formation^3–9. Flexor tendons are classified as intrasynovial tendons as they are within a synovial fluid sheath. However, autograft tendon grafts often come from extrasynovial tendons, due to lack of intrasynovial tendon that can be harvested from the human body without severe donor morbidity^3–5. It has been revealed that extrasynovial tendons have a rougher surface and are less lubricated compared to the intrasynovial tendons^10–12, resulting in poor clinical outcomes in both animal experiments and clinical studies^13,14. Surface modifications of extrasynovial autografts has previously been studied using different lubricating materials such as hyaluronic acid (HA) and lubricin with gelatin^15–17 to improve functional outcomes. While these compounds demonstrate improved tendon gliding^15–22, the biomaterials need to be synthesized, which causes an immune-response, increased cost, and FDA regulatory approval.

HA and lubricin have been found in native synovial fluid (SF)^23–25 with reduces friction between the articular cartilage surface of joints during motion, along with anti-adhesion properties that might inhibit healing of tissues within the joint, such as ACL or cartilage, although these mechanism are not known^26,27. SF is also a source of nutrients for both cartilage and FDP tendons in zone II^23,28–31. As intrasynovial tendons, such as FDP tendons, are normally lubricated by SF, with both HA and lubricin molecules identified on the tendon surface²⁵. SF has been used in animal models as a replacement for HA and lubricin^31–33. Ikeda et al reported that a extrasynovial tendon chemically modified with cd-SF-gel significantly improves tendon surface smoothness and decreased tendon friction even over 1000 cycles compared to the non-treated tendon in dog in-vitro model¹⁸. Ji et al further verified this technique using a dog in-vivo model and found that cd-SF-gel treated extrasynovial autograft significantly decreased adhesions and improved digit function by six weeks follow-up³⁴. However, six weeks follow-up for a flexor tendon reconstructioin is relative short, and there is no report about long-term functional outcomes.

The purpose of this study is to investigate a long-term effect of cd-SF-gel on extrasynovial autograft using a dog model. We hypothesized that cd-SF-gel could improve extrasynovial tendon surface quality, decrease gliding resistance, reduce adhesions, and improve digit function. If our hypotheses are supported, we would have developed a clinically translational therapeautics for flexor tendon reconstruction.

Materials & Methods

Study Design:

Our Institutional Animal Care and Use Committee approved the use of 20 adult female mongrel dogs, with an average age of 9 months and weight of 21.8 kg. Female animals were selected based on previous experience with males being more physically active and difficult to handle for postoperative therapy. FDP tendons from the second and fifth digits were cut and repaired^17,35,36. Following tendon repair, animals were allowed full weight bearing on the surgical forepaw, resulted in repaired tendons rupturing, thus creates a clinically relevant model for tendon grafting, as previously described by Zhao et al¹⁵. Six weeks later, the repaired and ruptured FDP tendons were reconstructed with autologous peroneus longus (PL) tendons harvested from both hind legs; one PL was treated with carbodiimide derivatized SF and gelatin (cd-SF-gel) (n=20), the other was treated with saline to serve as a control (n=20). Sacrifice occurred 24 weeks after reconstruction. Choice of 24-week follow-up was based on publications of using large animal models for tendon or ligament grafting^37,38. Following scrifice, both surgical and collateral non-surgical paws were collected for testing and analysis. Study design is shown in Figure 1.

Figure 1:

Study design; Day 0 FDP repair, 6 weeks FDP Autograft surgery and therapy begins with leg held in custom made jacket until 12 weeks where jacket it removed and free active motion takes place until sacrifice and outcomes evaluated at 24 weeks.

Surgical Procedure and Surface Coating:

For both surgical procedures, FDP tendon repair and reconstruction, animals were sedated with Acepromazine 0.05–0.2 mg/kg, Atropine 0.05 mg/kg and Morphine 0.5 mg/kg, then anesthetized with Diazepam 0.6 mg/kg, Ketamine 10 mg/kg and continued on 2% Isoflurane for the duration of the surgeries, with same lab staff (RLR) and surgeons (HL, CB and CZ) performing all procedures.

For the tendon repair surgery, second and fitth FDP tendons from randomly selected forepaws were exposed and lacerated 5-mm distal to the A2 pulley and repaired using a modified Kessler technique with a core suture of 3–0 Ethibond (Ethicon, Somerville, NJ USA) and 6–0 Prolene (Ethicon, Somerville, NJ USA) for the running suture, as previously described^{15,17,35,36,39–42}.

FDP tendon reconstruction was performed six weeks post-tendon repair. First, using an 18ga needle, SF was aspirated from knee joints, PL tendons were harvested from both hind legs for autografting. The second and fifth digits were opened and examined to confirm repair rupture visually with scar formation along the flexor tendon in zone II (Figure 2). PL autograft tendons were assigned to either the second or fifth digit²². One randomly selected PL tendon served as the control, with no surface modification, while the other was coated in cd-SF-gel with the following materials and procedures: First, 0.5ml SF mixed with 1% EDC (Thermo Fisher Scientific)⁴³ and 1% NHS (Thermo Fisher Scientific) ²⁰ were mixed with 0.9% NaCl and 0.1M MES at a pH of 6.0^21,44,45.²⁰ to form cd-SF, and then cd-SF were mixed with 10% gelatin (Thermo Fisher Scientific) to form cd-SF-gel. Mixed material was immediately coated onto the PL tendon surface. The chemical reaction was accomplished by activating carboxylic groups in the SF with the water-soluble condensing agent (EDC/NHS) to form intermediate O-acylisourea, which was covalently bonded by the exposed amino groups in the collagenous tendon matrix. This reaction was conducted in the OR just before the grafting procedure was performed (Figure 3). The role of gelatin in the modification served as a primer to smooth the tendon surface so that the lubricant molecules could more effectively function²⁰.

Figure 2:

Repair failure model showed the proximal end of the FDP tendon has ruptured and retracted to the palmer level without adhesion (while arrow), but the distal end of the ruptured tendon displayed scar formation along the sheath (black arrow)

Figure 3:

Synovial fluid was aspirated from knee, chemically modified with carbodiimide derivatized technique, and applied to the flexor tendon autograft.

Proximal end of the treated autograft tendon was repaired with a Pulvertaft weave repair technique and the distal end fastened by a 4–0 FiberWire (Arthrex, Naples, FL, USA), passing through a small hole drilled through the distal phalanx and nail, then anchored with a button (Bunnell technique); distal ends were also reinforced by suturing to the FDP stump using 3–0 nylon suture⁴⁶. Custom-made jackets were used to prevent weight bearing on the reconstructed surgical forepaw.

Five days post-reconstruction, daily synergistic passive therapy was initiated for both surgical digits and continued until six weeks post-op^17,47; at which time daily active therapy was provided by removal of the jacket for 30 minutes, allowing animals to gradually bear weight. Passive therapy was continued until jackets were completely removed at week 12. After removal, animals were allowed un-inhibited cage activity until sacrifice 24 weeks post autograft. After sacrifice, surgical and non-surgical forepaws were collected for the following analysis: normalized work of flexion (nWOF), PIP/DIP joint motion at 1.5 N, adhesion scoring, graft friction, proximal/distal repair strength, and histology.

Normalized Work of Flexion (n=10):

nWOF was used to evaluate joint motion related to graft function^15,17,22. All second and fifth digits were isolated and dissected at proximal metacarpal level with preservation of extensor and flexor tendons. A K-wire was placed in the center of the distal carpal bone and reflective markers were inserted into the fixed digit. Three markers were placed proximal to metacarpophalangeal (MCP), interphalangeal (PIP), and distal Interphalangeal (DIP) joints, respectively, while the fourth marker was placed in the proximal portion of the nail. Reflective markers were used in conjunction with motion capture cameras (Motion Analysis Corporation, Santa Rosa, California) enabling measurement of each joint’s full range of motion. A clamp was placed on the autograft tendon proximal to the proximal repair and pulled at a rate of 2mm/sec to 16mm by a motor attached to a force transducer recording at 60Hz, while a 5-N weight was attached to extensor tendon to ensure full extension at the start by applying initial tension. Displacement versus loading curve was recorded and used to calculate total WOF, which then was normalized to PIP and DIP motion^15,48,49. Native second and fifth digits from non-surgical paw were also tested in the same manner for nWOF as normal.

PIP/ DIP Joint Motion (n=10):

During WOF testing, force applied to graft tendon was truncated at 1.5N where total range of motion of the PIP and DIP was recorded as a function of the digit evaluations, as described previously ^15,17,22.

Adhesion Score (n=10):

Following completion of WOF testing, digits were further dissected along their flexor sheath in zone II, and adhesions were scored based on previously described evaluation criteria of 0 (no adhesion) to 8 (severe adhesions)^17,39,50. Adhesions were blindly scored by two individuals (RLR and HL) with consensus resolving disagreements.

Graft Tendon Friction (n=10):

After adhesion scoring was complete, graft tendons were further dissected with preservation of the graft tendon, A2 pulley and proximal phalanx to evaluate graft tendon gliding function^51–53. A mechanical actuator with a potentiometer and two load transducers on the proximal and distal ends was used to pull at a rate of 2 mm/sec^51–53. Secondand fifth native FDP tendons from the non-surgical forepaw were also tested for friction as a native control.

Graft Distal/Proximal Repair Strength (n=8):

Distal and proximal repairs were then dissected for repair strength testing. Strength of both treated and control autograft repairs were tested using a servohydraulic testing machine (MTS, Minneapolis, MN) based on established protocol^13,15,42. Testing was recorded at a sample rate of 50 Hz with 20 mm/min distraction until failure^13,42. Maximal strength to failure and stiffness of the linear region of the force/displacement slope were calculated.

Histology (n=2):

Two tendons from each group were subjected to histological evaluation⁴². All tendons (distal repair, Zone II and proximal repair) were embedded longitudinally, and sections were taken as close to the middle of the samples as possible. Distal tendon-bone attachments were placed in decalcification solution for two days before being imbedded four days later for H&E staining. Proximal repair and zone II graft were both fixed for two days immediately after collection for H&E staining. All samples were imaged with an Olympus BH-2 (Olympus, Shinjuku, Tokyo, Japan) at x100 and x200 maginfication. Tendon fibers within the tendon, adhesion tissue on the surface, tendon and bone integration at the distal repair and tendon to graft integration at the proximal repair were analyzed with Dimension cell Sens 1–16. (Olympus, Shinjuku, Tokyo, Japan)

Sample Size Calculation and Statistical Analysis:

As the major evaluations were digit function, sample size justification was calculated based on previous published data, in which nWOF of non-treated and treated graft tendons at 6 weeks was 1.60N-mm/degree (SD 0.61) and 0.67N-mm/degree (SD 0.34), respectively³⁹. A sample size of 10 in the mechanical evaluation section would be sufficient to detect a significant difference of one half of a standard deviation with an 80% power at a significance level of p<0.0017. All quantitative data are presented as the mean and standard error. Quantitative data, including adhesion score, normalized work of flexion, graft gliding resistance, and repair strength and stiffness, were analyzed with repeated-measures analysis of variance because the comparison groups were from the same animals. Statistical analysis was performed with SPSS Statistics software (version 14; SPSS). Significance was set at the level of p<0.017 in all cases after appling the Bonferroni correction.

Results

cd-SF-gel Treated Graft Decreases nWOF:

All incisions healed by primary intention. All but one FDP tendon reconstruction remained intact, with a proximal repair in the treated cd-SF-gel group rupturing. nWOF showed an average of 0.47 N-mm/degree ± 0.28 for the cd-SF-gel treated graft and 1.4 N-mm/degree ± 1.45 for the control group (p<0.014), although both FDP tendon reconstructed group’s had higher nWOF than the non-surgical native digits 0.18 N-mm/degree ± 0.08 (p<0.002) (Figure 4).

Figure 4:

Normalized work of flexion (nWOF) and adhesion score.

cd-SF-gel Treated Graft Reduces Adhesion:

Adhesion score in zone II of grafts treated with cd-SF-gel averaged 3.15 ±1.53, which was significantly lower than control grafts with an average score of 5 ±1.26 (p<0.00017) (Figure 4).

cd-SF-gel Treated Graft Increases PIP and DIP Joint Motion:

There was no significant difference in PIP joint motion at 1.5 N forces among the cd-SF-gel group (19.32⁰ ± 7.46⁰), the control group (20.14⁰ ±7.61⁰), and the native group (20.80⁰ ± 6.18⁰). DIP range of motion in the cd-SF-gel group (17.63⁰ ± 6.77⁰) and the native non-surgical digits (DIP 26.49⁰ ± 10.68⁰) was significantly higher than the control (DIP 7.07⁰ ± 12.99⁰) (p<0.015 and p<0.00003); but there was no significant difference between normal and cd-SF-gel group’s (Figure 5).

Figure 5:

Joint motion under 1.5 Newton applied to the graft tendons when moving the digit 16mm.

cd-SF-gel Treated Graft Decreases Graft Tendon Friction:

Frictional force of the graft tendons treated with cd-SF-gel (0.18 N ± 0.10) showed no significant compared to the control group’s 0.29 N ± 0.12 (p<0.017). However, both graft groups had higher friction compared to native FDP tendon 0.11 ± 0.06 (p<0.010) (Figure 6).

Figure 6:

Friction testing within the tendons.

cd-SF-gel Dose Not Cause Weakning of Counjuctions Healing:

There was no significant difference between cd-SF-gel treated and control grafts proximal failure strength and stiffness for either proximal or distal repair sites.

Histology Observation:

These results are a qualitative analysis on just two samples per group. Under gross observation it appeared that the cd-SF-gel treated grafts showed a smooth shinny surface with minimal adhesions around the tendon, whereas the control group showed severe adhesions making it difficult to isolate from surrounding tissues without sharp dissection. Healing at both the proximal and distal repair sites is also seen (Figure 7).

Figure 7:

Control (top) and autograft treated with cd-SF-Gel (bottom) showed graft/host healing at proximal (while arrow) and distal (black arrow) conjunctions. The control graft showed adhesions (gray arrow); but, the treated graft showed less adhesion formation compared to control autograft (strip arrow).

Histology images at the proximal repair site showed that recipient native FDP tendon complete integration with the graft tendon in both control and treated graft tendons with adhesion on the control (Figure 8).

Figure 8:

H&E staining of the graft proximal repair. Left picture is of the control autograft (T) and native FDP tendon (G). Right picture is the cd-SF-Gel treated autograft (G) and native FDP tendon (T). Both pictures show integration of the graft and native FDP tendon. Scale bar represents 50 μm.

Graft tendons in zone II also showed more adhesion formation around the graft in the control group compared to the cd-SF-gel treated graft, which appears smooth (Figure 9). Grafted extrasynovial tendons show loose paratenon tissue, characteristic of extrasynovial tendon, compared to native intrasynovial FDP tendon which is covered with a thin layer of epitenon (Figure 9). Both control and cd-SF-gel show tenocytes within the tendon.

Figure 9:

The normal FDP tendon shows dense collagen fiber with smooth surface covered with epitenon (black arrow). Control autograft shows severed adhesion formation on the tendon surface (black arrow). The treated autograft displays extrasynovial tendon surface with paratenon (black arrow) without adhesions. The scale bars in the bottom row represents 100 μm.

Taking a look at the native distal tendon-bone junction, a clear difference is seen between calcified fibrocartilage and fibrocartilage zone. Integration of the control and cd-SF-gel graft showed no distinct fibrocartilage zone (Figure 10).

Figure 10:

Distal tendon bone conjunction of the normal digit presents fibrocartilage zone at the tendon to bone interface. However, in both autograft group’s there is no transitional fibrocartilage zone (x 100 magnification). Scale bar represents 100 μm.

Discussion

This study used a recently developed SF modification to “synovialize” grafted extrasynovial tendons to replicate intrasynovial tendon gliding properties. Long-term results in this study have demonstrated that tendon surface modification effectively reduced adhesions, increased graft gliding ability, and improve digit function. However, this study also showed that surface modification does not bring an extrasynovial autograft to native FDP gliding ability, even after 24 weeks following surgery, although it was closer to the native digit in range of motion than the unmodified extrasynovial grafts. No adverse reaction was seen in the wound healing by adding modification treatment to the reconstruction site, along with no issues with the SF aspiration site.

Tendon surface modification using chemically modified lubricants such as hyaluronic acid and lubricin has been studied^{16,17,19–22,39,54}. Although these results have shown that tendon surface treatments improved tendon gliding ability and digit function, they also delayed recipient-graft healing that caused repair site weakening¹³. In contrast, a short-term six week follow-up study with graft surface modification using chemically modified SF did not show this side effect³⁴. In our current long-term study, we also found that extrasynovial grafts treated with chemically modified SF did not interfere with conjunction healing at either the proximal tendon to tendon interface or distal tendon to bone interface 24 weeks after FDP tendon reconstruction. Healing strength at the recipient-graft conjunctions was a 114 N at 24 weeks post-tendon reconstruction compared to a previously reported six week study with was 59 N³⁴. As well as the adhesion score being 3.15 for cd-SF-g in the long-term study compared to 3.5 in the short-term study³⁴.

The major strength of this study was to use a preclinical large animal model with a relative long-term functional outcome measure to verified a clinically-translational therapy to improve flexor tendon reconstruction. This autologous SF may also eliminate inflammatory response induced by foreign proteins. All incisions including first repair surgery and second reconstruction in all animals were healed without severe inflammation and infection. Indeed, this surface modification technology may be used to apply to other fields where adhesion prevention is important such as intestinal or uterine adhesions.

Several limitations are present in this study. First, we only included one follow-up time point, at 24 weeks after FDP tendon reconstruction, as a short-term study was previously reported³⁴. Second, as this was a relatively long-term study, we only focused on functional outcome measures and did not analyze biological responses at the cellular and molecular levels, which are more relevant at early stage healing. Third, histology was performed on only two samples for qualitative observations, without quantitative analysis. Forth, we did not look at the ideal dose of cd-SF-Gel. Finally, due to limited resources of intrasynovial autograft, we did not do a direct comparison to the intrasynovial graft.

Conclusion

Reconstruction of damaged flexor tendons, belonging to intrasynovial tendons, often uses extrasynovial tendons as the graft tendon due to limited resource of intrasynovial autograft availability. However, rough surface of extrasynovial tendon compared to intrasynovial tendon results in poor functional outcomes in both experimental and clinical settings. In this long-term in-vivo preclinical large animal mode, we demonstrated that surface modification of extrasynovial autograft tendon with cd-SF-gel for flexor tendon reconstruction improves functional outcomes. Our approach could have significant clinical relevance and translational value by using patient’s own SF to “synovialize” an autologous extrasynovial graft. Future investigation of the FDA regulatory path for chemically modified SF is needed before clinical trial.