Trauma resulting in hemarthrosis and long medial collateral ligament desmitis of the tarsocrural joint in a horse

Abstract

A horse was initially diagnosed with hemarthrosis and desmitis of the long medial collateral ligament of the right tarsus and later developed prominent enthesiophytosis at the site of insertion of the ligament’s deep portion. Hemarthrosis due to intra- or peri-articular pathology can cause recurrent lameness, even without evident external trauma.

Hemarthrosis has been reported in horses as a sequela to soft tissue or bony trauma in many joints (1–5), hemophilia (6), or as a complication of arthroscopy (7). In human athletes, acute hemarthrosis of the knee is considered to indicate serious ligamentous injury unless proven otherwise (8). Typical symptoms of hemarthrosis in the horse are reported as mild to severe lameness of acute onset with joint effusion (4,9,10). Ultrasonography of the joint cavity can be useful in characterizing the effusion (11) and arthrocentesis can confirm the diagnosis of hemarthrosis (4).

In hemophilic humans, recurrent hemarthrosis leads to chronic synovitis, pain, and reduction of range of motion (12–14). It is thought that the blood itself causes an inflammatory reaction with hemosiderin deposits (14–16). Furthermore, persistent synovitis and joint capsule distension combined with limited joint mobility may cause stretching of tendons, ligaments, and the capsule leading to further joint instability and secondary degeneration (13,14). Due to the possibility of performance-limiting sequelae from repeated hemarthrotic episodes, it is crucial to optimize treatment of hemarthrotic joints through rapid diagnosis of the hemarthrosis and any underlying bony or soft-tissue pathologies that may accompany it.

Case description

A 596-kg, 12-year-old Warmblood gelding was examined for a mild lameness of 4-days duration and swelling of his right hock. Upon examination, the right tarsus was severely swollen with significant effusion of the tarsocrural joint. Palpation and manipulation were not resented and there was no swelling in the lower limb. The horse was not lame at the walk or trot in a straight line but was reluctant to fully bear weight on the right hind leg when turned to the right [grade 2/5 lameness according to the AAEP grading scale (17)]. All vital parameters (temperature, heart rate, respiratory rate) were within normal limits.

Radiographs of the right tarsus revealed no bony abnormalities. Ultrasound examination of the tarsocrural joint revealed an increased amount of hypoechoic fluid which contained multiple areas of increased, swirling, heterogenous echogenicity. These findings were suggestive of an organizing hematoma and/or fibrin within the joint (5,11). No abnormalities were seen in the periarticular ligaments. A diagnosis of hemarthrosis was tentatively made based on the duration of clinical signs, lack of significant lameness, ultrasound evidence of organizing hematoma, a lack of radiographic lesions, and lack of changes in the periarticular ligaments. The horse was stall rested with daily hand-walking for a period of 3 wk. He was administered 2 g [3.5 mg/kg body weight (BW)] of phenylbutazone intravenously (Phenylbutazone Sterile Injection, Vétoquinol Canada, Lavaltrie, Quebec) and prescribed 2 g of phenylbutazone (Phenylbutazone Oral, Vétoquinol Canada), PO, q24h for the following 4 d.

The horse was re-examined 3 wk and 6 wk after the initial examination. The lameness had been improving, but had suddenly worsened on both occasions, with no history of renewed trauma or change in management.

On the first occasion, the tarsocrural joint was still severely effusive, and the horse was grade 4/5 lame, greatly resenting even minor flexion of the hock (such as picking up the foot for cleaning), which then made him non-weight-bearing lame (grade 5/5) (17). Radiographs of the hock were repeated and again, no bony abnormalities were found. On ultrasound, a markedly increased amount of hypoechoic joint fluid was noted, as well as periarticular edema. Thick hyperechoic strands were seen floating in the synovial fluid, which were suspected to be proliferative synovium and/or tags of fibrin. Periarticular ligaments appeared normal. Arthrocentesis of the tarsocrural joint was performed and approximately 10 mL of slightly cloudy and flocculent, amber-colored fluid of low viscosity was collected. Injection of 20 mL of 2% mepivacaine hydrochloride (Carbocaine-V; Pfizer Canada, Kirkland, Quebec) into the tarsocrural joint resulted in an improvement of the lameness to grade 2/5 after 10 min. The cytology results revealed a mild increase in nucleated cells (nucleated cell count 0.8 × 10⁹/L; reference: < 0.5 × 10⁹/L) with a small quantity of erythrocytes. A differential cell count yielded 72% neutrophils, 22% macrophages, and 6% lymphocytes. The neutrophils were non-degenerate and the macrophages were mildly foamy and infrequently contained phagocytosed cytoplasmic debris. Occasional red blood cells and hemosiderophages were seen. These results were interpreted as mild neutrophilic inflammation with scant hemorrhage. Routine aerobic and anaerobic cultures yielded no growth. As the inflammation was mild and its exact cause unknown, diagnostic arthroscopy was recommended. The owners elected conservative management, which consisted of box stall rest with gradually increasing increments of hand-walking, and a course of oral phenylbutazone (Phenylbutazone Oral; Vétoquinol Canada) at 2 g (3.5 mg/kg BW) PO q24h for 7 d.

The horse was re-examined 6 wk after initial presentation. The owners reported that the lameness had been resolving and the horse had not been visibly lame at the walk for the last week. On the day of the examination, he was visibly lame at the walk (Grade 4/5) (17) and resented flexion of the hock. Moderate effusion of the tarsocrural joint was still present, with the distension seeming more severe in the medial part of the tarsocrural joint pouch, along the medial aspect of the tarsocrural joint capsule. Radiographs of the hock were taken, and an ill-defined opacity adjacent to the distal dorsomedial aspect of the talus was seen (Figure 1). This was interpreted as possibly being an osteochondral fragment arising from the talar trochlea or calcification within the medial long collateral ligament of the tarsus. Ultrasonography of the tarsus revealed thickened synovial membrane and the presence of fibrin within the tarsocrural joint. No abnormalities of the periarticular ligaments were noted. Due to the recurrent nature of the problem, lack of improvement with rest, and the possibility of the presence of an osteochondral fragment, the horse was referred to the Western College of Veterinary Medicine for diagnostic arthroscopy.

Figure 1.

Dorsolateral to plantaromedial oblique view of the right hock 6 wk after initial presentation showing an opacity in the insertion of the deep portion of the long medial collateral ligament of the tarsus (circle).

Due to the recurrent nature of the hemarthrosis, platelet count, partial thromboplastin time (PTT), prothrombin time (PT) and fibrin degradation products (FDP) were measured to rule out hemostatic disorders before surgery. The usual pre-surgical analyses of blood gas, packed cell volume (PCV), and total protein (TP) were also carried out. All values were within normal limits.

Under anesthesia but prior to joint puncture with the arthroscope, a sample of synovial fluid was taken and submitted for cytology and culture. The fluid was serosanguinous and had reduced viscosity. Arthroscopy revealed the synovial membrane to be diffusely hyperemic and thickened. There was a copious amount of fibrin deposited along the surface of the synovium and free-floating fibrin strands were abundant within the joint cavity. A combination of manual and mechanical debridement was used to remove the fibrin and perform a partial synovectomy of the hypertrophic synovial villi. Samples of the fibrin deposits and synovium were submitted for culture. Following curettage the joint was flushed with isotonic saline solution, and the cartilage surfaces and periarticular structures visible from inside the joint were inspected. No evidence of an osteochondral fragment was found in the joint cavity. Mild fibrillation of the medial long collateral ligament of the tarsus was seen. The arthroscopic portals were closed routinely and an intravenous regional perfusion of the tarsus was performed using 3.6 g of gentamicin (Gentocin Injectable Solution (100 mg); Intervet-Schering Plough, Kirkland, Quebec) in 60 mL of sterile (0.9%) saline.

The cytology of the joint fluid collected just prior to surgery was consistent with mild suppurative inflammation (nucleated cell count 0.6 × 10⁹/L; reference < 0.5 × 10⁹/L; total protein 25 g/L; reference range: 9.2 to 31.1 g/L). A differential cell count yielded 52% neutrophils, 31% mononuclear cells, and 17% lymphocytes. The neutrophils seen were hypersegmented and non-degenerate. Numerous highly vacuolated macrophages were present with a few undergoing erythrophagocytosis. Rare eosinophils were present. Cultures of synovium and fibrin samples obtained during surgery were negative.

After surgery, the horse received 3 treatments of intravenous limb perfusion using gentamicin q36h at 3.3 mg/kg BW. Systemic treatment with ceftiofur (Excenel; Pfizer Animal Health, Kirkland, Quebec) was administered intravenously q24h at 2.2 mg/kg for 5 d. The horse was discharged with instructions for strict stall rest for 4 wk followed by gradually increasing amounts of hand-walking. Due to the fibrillation and enthesiophytosis of the medial long collateral ligament, the owners were given a guarded prognosis for full return to function.

Four months after discharge, the horse was again examined. He was grade 2/5 (17) lame at the trot and had mild effusion of the tarsocrural joint. Intra-articular injection of 100 mg methyl-prednisolone acetate [Depo-Medrol Sterile Aqueous Suspension (40 mg), Pfizer Animal Health], 100 mg of gentamicin sulfate [Gentocin Injectable Solution (100 mg), Intervet-Schering Plough], and 40 mg of hyaluronate sodium (Hyaluronic Acid; Legend Injection, Bayer Animal Health, Toronto, Ontario) was given at this time. The owners were instructed to continue stall confinement with increased hand-walking.

Eight months after discharge, the horse was again re-examined. He had been turned out in a small individual paddock for the last month and the owners had resumed riding him at a walk and light trot. On examination, he was grade 3/5 (17) lame on the right hind leg. Mild tarsocrural effusion remained. Flexion tests were not performed, but there was no reluctance when the horse was prompted to pick up his right hind foot for cleaning. Radiographs of the right tarsus were repeated. Significant enthesiophyte formation was visible on the distal dorsomedial aspect of the tallus (Figure 2). The location of the enthesiophyte corresponded to the location of the calcification seen 8 months previously, before surgery.

Figure 2.

A 125° dorsolateral to plantaromedial oblique radiograph of the right hock 10 mo after initial presentation showing prominent enthesiophytosis at the insertion of the deep portion of the long medial collateral ligament of the tarsus (circle).

Discussion

The horse in this case initially showed the hallmark signs of hemarthrosis, most likely of traumatic origin. The progression of the case was atypical in that the lameness re-occured multiple times and that there was evidence of significant joint inflammation.

Hemarthrosis in horses occur as a sequela to soft tissue or bony trauma in many joints (1–5), hemophilia (6), or a complication of arthroscopy (7). Other causes of hemarthrosis reported in humans and dogs are intra-articular neoplasms of vascular origin, such as hemangiosarcomas or developmental defects such as hamartomas (18).

Typical symptoms of hemarthrosis in the horse are mild to severe lameness of acute onset with joint effusion (4,9,10). Typically, lameness is of short duration, usually resolving within 48 h (4). Radiography shows no abnormality other than joint distension (9,19). Ultrasound can be a valuable tool in diagnosis since an older hematoma is characterized by echogenic septa separating anechogenic fluid pockets (11). Arthrocentesis will reveal pure blood in acute hemarthrosis (4), or dark-colored synovial fluid containing hemosiderophages (20). Chronic hemarthrosis can cause inflammation in itself, and although the cellular reaction may vary widely, it is often predominantly neutrophilic (20). Draining of the bloody joint fluid usually brings immediate improvement of clinical signs (4).

In humans with hemophilia, recurrent hemarthrosis leads to chronic synovitis, pain, and reduction of range of motion, which greatly reduces quality of life (12–14). The pathogenesis of the cartilage degeneration that ensues, however, is still poorly understood. It is thought that because very little blood can be actively reabsorbed from a joint by synoviocytes (12), the blood itself causes an inflammatory reaction with hemosiderin deposits and organization of intra-articular exudates (13,14). The deposits of iron in the joint seem to play a crucial role in the development of hemophilic arthropathy (15) independent of the damage caused by the ensuing synovitis. Iron promotes apoptosis of chondrocytes through the formation of oxygen metabolites, and favors the proliferation of the synovial membrane through the stimulation of inflammatory cytokines (15). The changes in chondrocyte activity found in an early hemophilic ankle were found to be characteristic of that of early osteoarthritic cartilage (12) and many of the characteristics of hemophilic arthropathy are similar to those of rheumatoid arthritis (15). Furthermore, persistent synovitis and joint capsule distension combined with limited joint mobility may produce stretching of tendons, ligaments, and the capsule leading to further joint instability and secondary degeneration (13). Hemophilic hemarthrosis is self-propagating, as the synovium hypertrophies in the presence of a large quantity of blood. The synovium develops focal areas of increased villus formation as early as 4 d after the onset of an acute episode of hemarthrosis (16). This hypertrophic synovium is richly vascularized, so the chances of having recurrent small bleeds secondary to minor injuries are increased (9,16).

Hemophilia, a hereditary disorder of secondary hemostasis, has been diagnosed in Thoroughbred, Standardbred, Quarter horse and Arabian colts and is transmitted as an X-linked recessive trait (6,21–23). As only 1 case of a horse diagnosed with hemophilia A or B has been reported to survive beyond the age of 3 y (6) it is unknown whether or not hemarthrosis would lead to chronic joint degeneration in equine patients.

Only 1 case of chronic, recurrent hemarthrosis in the carpus of a horse with recurrent severe lameness has been reported (9). Synovectomy was performed through arthroscopy and resulted in return of the horse to full work. Synovectomy is recommended for treatment of chronic synovitis in the horse (9). In the case reported here, fibrillation of the long medial collateral ligament during arthroscopy and subsequent enthesiophyte formation at the insertion of its deep component 8 months later may explain the reoccurrence of hemarthrosis and continued inflammation of the joint.

Desmitis of the collateral ligaments of the tarsus has been reported to be caused by trauma (24) or by cyclic loading during training in Standardbred racehorses (1). Because of the oblique angle of the talus under flexion, the medial collateral ligaments are under more tension than the lateral collateral ligaments during flexion, although these are both relatively loose in flexion compared to the tension they undergo during extension of the tarsus (25). The long collateral ligaments are under more tension than the short collateral ligaments during extension, which may explain why they have a higher tendency to develop desmitis than the short collateral ligaments (24,25). Based on the location of enthesiophyte formation in this case, it is likely that the deep layer of the long medial collateral ligament (dorsal tarsal ligament) was strained in addition to the fibrillation seen arthroscopically in the main intra-articular body of the ligament. The new bone formation seen on radiographs at the time of surgery was extra-articular and as such, could not be seen arthroscopically.

The degree of lameness with collateral ligament desmitis can vary from subtle to non-weight-bearing; in most cases the lameness worsens after flexion and tarsocrural effusion is sometimes present (1,26). Pain may be present on digital pressure over the affected ligament (24,27). Radiographs may reveal only soft tissue swelling or the presence of avulsion fragments, which are often buried within the ligament (1,26–28). In some cases, it is useful to repeat the radiographs 2 to 4 wk after the initial injury, as evidence of periosteal reaction at the site of injury may be visible at that time (26). Ultrasonography may identify which ligament is involved and the severity of the desmitis (28,29). Injury to the collateral ligaments of the hock usually present as diffuse thickening of the ligament rather than discrete tears (5) and in older lesions this thickening remains (29). Treatment of acute collateral ligament injury usually involves 30 d to 6 mo of rest and oral administration of non-steroidal anti-inflammatory drugs during the first 10 to 14 d (24). Cold therapy and topical dimethyl sulfoxide (DMSO) with support bandaging or splinting has been reported to improve the response to therapy (24,26). The prognosis for return to full performance is guarded as concurrent lameness problems such as degenerative joint disease often accompany the condition (1,24,28).

It is unfortunate that the desmitis of the collateral ligament was not diagnosed until 2 mo after the initial injury, despite multiple ultrasonographic examinations of the periarticular structures. The first ultrasonographic examination of the joint focused mainly on the intra-articular structures and fluid and may have missed the early signs of desmitis. In subsequent ultrasonographic examinations, it is possible that the thickening of the medial collateral ligament which is reported to be common in such injuries (28) was mild due to the intraarticular position of the fiber tears, or was mild enough to be not readily detected. Comparing circumferential and cross-sectional measurements of the thickness of the collateral ligaments of the right versus left hock may have aided in earlier diagnosis of this condition. It could be argued that earlier, more aggressive immobilization through splinting or casting could have led to optimal healing of the long medial collateral ligament and a better long-term outcome. CVJ