Osteoarthritis of the coxofemoral joint in 24 horses: Evaluation of radiography, ultrasonography, intra‐articular anaesthesia, treatment and outcome

Abstract

Background

Few case reports describe equine coxofemoral joint osteoarthritis (CFJOA).

Objectives

To evaluate diagnostic findings and outcome of horses with CFJOA and to provide a score facilitating radiographic assessment.

Study design

Retrospective case series.

Methods

History, clinical signs, ultrasonographic, radiographic and intra‐articular anaesthesia findings, treatments, outcome, and necropsy results of horses with CFJOA presented between 2002 and 2023 were collated. Radiographic findings were categorised to develop a radiographic score which was applied by two masked examiners. Intra‐ and inter‐observer reliability was determined using weighted Cohen's kappa (Cκ) and the correlation between radiographic and ultrasound findings via Spearman correlation coefficient.

Results

The study included 24 horses (median age 14 years). Most of them (20/24) were chronically lame. Frequent clinical signs included unilateral gluteal muscle atrophy (18/21), lengthening of the stride of the affected limb (13/19) and locomotion on three tracks (13/20). Both imaging modalities enabled evaluation of periarticular osteophytes (correlation coefficient r = 0.64; p = 0.003). Additionally, radiography allowed detection of irregular joint spaces, subchondral bone opacity changes and femoral head flattening/tapering. Inter‐observer (Cκ = 0.846) and intra‐observer (Cκ = 0.853 and Cκ = 0.842) agreement was excellent. If treated, mostly intra‐articular corticosteroids were administered (16/18). Nine horses were euthanised immediately or during follow‐up examination. Post‐mortem, the Ligamentum capitis ossis femoris was commonly found ruptured. All surviving horses remained lame.

Main limitations

Retrospective analysis of clinical records and subjective outcome assessment based on owner follow‐up with potential recall bias. Due to overall disease severity, associations between different grades of clinical findings, radiographic abnormalities and outcome could not be evaluated.

Conclusions

Typical clinical signs may indicate CFJOA. Standardised evaluation of ventrodorsal radiographs allows a comprehensive diagnosis. Postmortem findings suggest joint instability as a possible causative factor that may contribute to the poor prognosis and resistance to medical therapy of the disorder.

1. INTRODUCTION

Osteoarthritis (OA) of the coxofemoral joint (CFJ) in horses has been described rarely and mainly in case reports. ¹ , ² , ³ , ⁴ , ⁵ Its prevalence is unclear. It occurs secondary to developmental disorders such as congenital joint dysplasia ⁶ , ⁷ , ⁸ , ⁹ and osteochondrosis, ¹⁰ secondary to trauma or fall, which may be associated with subluxation, ¹¹ , ¹² rupture of the Ligamentum capitis ossis femoris ³ , ⁴ , ⁵ or acetabular fractures, ¹² or in rare cases secondary to femoral head necrosis. ¹ A purely degenerative aetiology has not been confirmed in horses.

Reported history details of affected horses include older horses, ⁴ moderate to severe chronic hindlimb lameness ² , ⁴ and trauma. ² Variable degrees of gluteal muscle atrophy and in some cases, a characteristic toe‐out, hock‐in stance have been described. ³ , ⁴ Horses may be painful on palpation, during flexion of the limb and unwilling to stand on it when the contralateral limb is elevated. ³ At a walk, a tendency to move on three tracks (sidewinder gait) ³ , ⁵ with the limb rotated externally and carried abducted when advanced, a low arc of foot flight ² and a reduced cranial phase of stride ² , ³ may be present.

Aids to diagnosis include intra‐articular anaesthesia, ¹³ , ¹⁴ , ¹⁵ transcutaneous ultrasonography, ¹² , ¹⁶ , ¹⁷ , ¹⁸ radiography ¹⁶ , ¹⁹ , ²⁰ and nuclear scintigraphy despite its limited sensitivity and specificity. ²¹ Computed tomography has only sporadically been described for the diagnosis of CFJ disorders in foals. ⁶ , ²²

Intra‐articular diagnostic anaesthesia remains the most valuable diagnostic tool. ²³ Arthrocentesis of the CFJ can be performed using different techniques, including a blind caudolateral approach through the trochanteric notch ²³ , ²⁴ , ²⁵ as well as more recently described approaches performed under ultrasonographic guidance from a cranio‐dorsolateral ¹³ or cranioventral direction. ¹⁵ Neither approaches nor outcome after anaesthesia have been thoroughly investigated in clinical cases of CFJOA.

Static ¹³ , ¹⁶ , ¹⁷ , ²⁶ and dynamic ¹² transcutaneous ultrasonography allows examination of the contours of the greater trochanter, the neck of the femur, a small part of the femoral head and the dorsolateral aspect of the acetabulum including the fibrocartilaginous Labrum acetabulare. Thereby, CFJOA can be identified via detection of osteophytosis at the dorsal and cranio‐lateral aspect of the joint and increased joint effusion. ¹⁶ , ²⁷

Several radiographic approaches, including a lateral, ²⁸ , ²⁹ a transrectal, ³⁰ a recumbent ³¹ as well as a standing ventrodorsal, ¹⁹ a modified ventrodorsal ¹⁶ and a lateral oblique, ²⁰ , ³² have been described. For complete assessment of the CFJ, particularly ventrodorsal approaches provide the most complete assessment of the CFJ. ¹⁶ Evaluation of radiographic changes follows general rules for assessment of OA. ³³ However, a specific panel of criteria as in humans ³⁴ and dogs ³⁵ , ³⁶ has not been defined.

The aim of this retrospective case series of horses with CFJOA is to describe the history, clinical findings, implementation and results of intra‐articular anaesthesia, ultrasonographic and radiographic findings as well as to present treatment, outcomes and results of post‐mortem examinations. A further aim is to establish a radiographic score to improve systematic evaluation of radiographic changes associated with CFJOA, to compare the diagnostic value of radiographic and ultrasonographic joint assessment and to evaluate the diagnostic value of an acetabular ratio.

2. MATERIALS AND METHODS

2.1. Horses

Medical records of all horses with CFJOA presented to the Clinic for Horses of the University of Veterinary Medicine Hannover, Foundation, Hannover, Germany between 2002 and 2023 were reviewed. Breed, sex, age as well as medical history details including the duration of lameness (‘short‐lasting’ being defined as less than 2 months and ‘long‐lasting’ as more than 2 months), incidence of trauma, previous diagnostics and treatments were recorded.

2.2. Clinical and lameness examination and exclusion of distal limb lameness

All horses underwent a clinical and orthopaedic examination. Complete examination at rest included assessment of muscle atrophy (Figure 1) and resistance and/or pain during manipulations, particularly flexion, extension, abduction, and adduction, as well as the inability to stand on the affected limb while elevating the contralateral limb. Specific gait changes were assessed at a walk including locomotion on three tracks (Video S2), reduced height of the limb flight, external rotation of the limb (Video S2) and changes in the stride length of the affected limb (Videos S1 and S2). Lameness was evaluated at the walk and, when possible, at the trot. Photographs and videos were taken during the examination. For the purpose of the present study, they were examined jointly by two examiners (FS and MH). Using the photos and videos, they assessed muscle atrophies, response to palpation and manipulation, gait abnormalities as well as lameness according to the AAEP lameness grading scale. ²⁵ Whenever photographs and videos were not available or incomplete, clinical records were used instead if available. If performed, results of flexion tests, perineural and intrasynovial anaesthesia and radiographic examination of the distal limb were recorded.

FIGURE 1.

Example of a horse with unilateral, left sided gluteal muscle atrophy (case number 15).

2.3. Ultrasonography

Transcutaneous ultrasonography of the CFJ was performed by one of two examiners (FG and MH). A dorsolateral oblique transcutaneous approach was used ¹⁶ (2002–2010: Sonoline Elegra, Siemens; 2010–2021: Logiq E9, General Electric [GE] HealthCare). Ultrasonographic images were evaluated retrospectively by two examiners together (FS and MH) for the presence of mild, moderate or severe osteophyte formation at the margin of the acetabulum and/or the femoral neck. Each CFJ was categorised as having no (Grade 0), mild (Grade 1), moderate (Grade 2) or severe (Grade 3) ultrasonographic signs of OA. A joint was classified as having mild osteophytes if the margin of the acetabulum and the femoral head appeared slightly irregular, with single, small bone proliferations. Moderate osteophyte formation was present if the margins appeared moderately irregular including several protruding bone proliferations. Severe osteophyte formation was recognised if the joint margins appeared highly irregular with numerous bone proliferations of different size.

2.4. Radiography, radiographic score and acetabular ratio development

Radiographs of the CFJ were obtained with the horse standing (20/24) using the modified ventrodorsal approach (Figure 2), ¹⁶ except for horses examined between 2002 and 2003, in which radiographs were taken in dorsal recumbency under general anaesthesia (4/24) ³¹ (2002–2010: OPTIMUS 80, Philips Medizinsysteme and Fuji IP 3A, Fuji Photo Film; 2010–2023: Gierth HF 1000, Gierth X‐Ray International GmbH and Fujifilm DR‐ID 300 CL APL Software V11.0, Fujifilm Europe GmbH). Depending on the horses' size and weight, kV and mAs were set at 100–150 and 200–263, respectively.

FIGURE 2.

Set up for the modified ventrodorsal radiographic approach in the standing patient. The ipsilateral limb is abducted ~30° from vertical, the x‐ray tube is connected with a custom‐made conical tube (40 cm length) and positioned in an almost upright and slightly caudally oriented position below the horse and the cassette holder (format 35 cm × 43 cm) is placed on the horse's croup using a long handle. ¹⁶

Radiographs of the CFJ were evaluated under consideration of previously described criteria ³³ , ³⁷ , ³⁸ which were summarised in a scoring system with: (1) osteophyte formation at the cranial and/or caudal rim of the acetabulum (separate assessment for cranial and caudal: mild, moderate, severe, or 1, 2, 3 points, respectively), (2) osteophyte formation at the cranial and/or caudal contour of the femoral head (separate assessment for cranial and caudal: mild, moderate, severe, or 1, 2, 3 points, respectively), (3) irregular width or narrowing of the joint space (no or yes, or 0 or 3 points, respectively), (4) lack of visibility of an indentation of the ligament attachment (inconsistent evaluability due to older radiographs), (5) changes in subchondral bone opacity (no or yes, or 0 or 3 points, respectively) and (6) flattening or tapering of the femoral head (no or yes, or 0 or 3 points, respectively). Subsequently, scoring was independently and randomly performed twice with an interval of 5 months by two examiners (FS and MH) using a list randomiser. ³⁹ The examiners were masked to the results of the lameness examination and the ultrasonographic findings. Depending on the scoring results, each CFJ was graded into one of the following groups: no (Grade 0 [0 to 1 points]), mild (Grade 1 [2 to 7 points]), moderate (Grade 2 [8 to 14 points]) or severe (Grade 3 [15 to 21 points]) OA.

In addition to the radiographic score, an acetabular ratio was determined on the radiographs of horses with CFJOA. It was defined as the maximum depth of the acetabulum divided by the acetabulum length similar to the established glenoid ratio for evaluating scapulohumeral joint dysplasia in Shetland ponies. ⁴⁰ The acetabular ratio was also determined on radiographs of horses without radiographically detectable CFJOA. For this purpose, CFJ radiographs of adult horses were used that were taken at the clinic between 2017 and 2023 during lameness examinations. The horses were ultimately found to have an underlying cause of lameness unrelated to the CFJ. All available radiographs were independently evaluated by the two investigators. Only radiographs without any radiographic signs of bone modelling noted by either investigator were used.

2.5. Intra‐articular anaesthesia of the CFJ

For intra‐articular anaesthesia, horses were restrained in stocks. In horses examined between 2002 and 2003, the caudolateral approach to the CFJ through the trochanteric notch between the cranial and caudal part of the greater femoral trochanter was used for arthrocentesis. ²³ , ²⁴ , ²⁵ After 2003, the cranio‐dorsolateral approach was performed under ultrasonographic guidance. ¹³ After aspiration of synovial fluid, 20 mL mepivacaine (20 mg/mL, Mepidor, Richter Pharma AG) was injected intra‐articularly. Synovial fluid was submitted for analysis. Assessment of anaesthesia was performed every 5 min within 30 min of injection at a walk and trot, if possible.

2.6. Treatment, outcome and long‐term follow‐up

Applied treatments and outcome were recorded. Long‐term follow‐up in discharged horses included a structured client telephone interview with the owners in April 2023 (Text S1). Owners were asked if horses had returned to their previous or a lower level of work, were retired or had died. If deceased, the cause of death was also noted. If necropsy was performed in horses euthanised due to CFJOA, post‐mortem macroscopic findings and the results of histological examinations of the CFJ were recorded.

2.7. Data analysis

All statistical analyses were performed in R. ⁴¹ Descriptive data analysis was implemented. The inter‐observer and intra‐observer agreement was determined using quadratic weighted Cohen's kappa κ (Cκ) ⁴² for the radiographic grading (R irr package, function kappa2). Agreement was interpreted as being slight with kappa values of 0–0.2, fair with values of 0.2–0.4, moderate with values of 0.4–0.6, good with values of 0.6–0.8 and excellent with values of 0.8–1. ⁴³ For the evaluation of the acetabular ratio and the association between radiographic and ultrasonographic findings, data were tested for normality of distribution via Shapiro Wilk test and visual inspection (Histograms). Subsequently, for the acetabular ratio, mean values of patients and controls were compared using the Welch test for two independent samples with unequal variance (R psych package, function t.test). In order to measure the strength and direction of the monotonic relationship between radiographic and ultrasonographic score, Spearman correlation analysis was conducted (R base package, function cor.test). Results were considered significant at p < 0.05.

3. RESULTS

3.1. Horses

The study included 24 horses (Table 1; Figure 3), which were 1 to 22 years old with a median age of 14 years.

TABLE 1.

Summary of signalment, history and clinical signs of 24 horses with osteoarthritis of the coxofemoral joint.

	Overall number of cases for which information is available	n	%
Breed	24/24
Warmblood		10	42
Friesian		4	17
Haflinger		3	13
Standardbred		2	8
Pony		2	8
Quarter Horse		1	4
Black Forest draught horse		1	4
Andalusian		1	4
Sex	24/24
Mare		8	33
Gelding		13	54
Stallion		3	13
Age	24/24
≤4		3	13
5–12		8	33
13–18		9	38
>18		4	17
History	24/24
Short‐lasting lameness (<2 months)		4	17
Long‐lasting lameness (>2 months)		10	42
Long‐lasting lameness (>2 months) related to trauma		10	42
Orthopaedic exam
Limb	24/24
Left		10	42
Right		13	54
Both		1	4
Gluteal muscle atrophy	21/24
Yes		18	86
No		3	14
Painful on manipulation	18/24
Yes		13	72
No		5	28
Cannot stand on ipsilateral limb	16/24
Yes		7	44
No		9	56
Lameness grade AAEP	24/24
4		16	67
5		8	33
Locomotion on three tracks	20/24
Yes		13	65
No		7	35
Reduced height of the limb flight	19/24
Yes		2	11
No		17	89
External rotation of the limb	20/24
Yes		2	10
No		18	90
Lengthened cranial phase of the stride of the affected limb	19/24
Yes		13	68
No		6	32

FIGURE 3.

Flow chart displaying the diagnostic approach, treatment and outcome of horses with coxofemoral joint osteoarthritis (CFJOA) included in the study. CFJ, coxofemoral joint. *Horse showed an improved gait pattern at the control examination, **horse showed no improved gait pattern at the control examination, ***follow‐up information gained 11.4 to 20.1 years after discharge—all horses deceased in April 2023, ****follow‐up information only 3 months after initial presentation, as the horse was presented to the clinic in January 2023.

3.2. History

All horses were presented to the clinic with a history of hindlimb lameness. The duration of lameness varied between 1 week and 11 years (median 3.5 months). In 20/24 (84%) horses, the lameness was long‐lasting. Of these 20 horses, 10 horses had a history of an initial trauma such as a fall in a pasture accident. Diagnostic anaesthesia of the distal limbs had been performed in 6/24 horses before referral, resulting in no improvement in lameness in five horses, whereas in one horse, lameness improved after diagnostic anaesthesia of the CFJ. In 6/24 horses, it was known that the referring veterinarian had already taken radiographs of the distal limb. In four of these horses, no clinical significant findings could be identified, whereas in one animal osteoarthritis of the proximal interphalangeal joint was diagnosed and in another a subchondral cyst‐like lesion in the medial femoral condyle was identified and assumed to be the cause of lameness. In two horses, nuclear scintigraphy was performed, indicating CFJ disease in one horse. The other horse displayed no increased radiopharmaceutical uptake in the area of the CFJs. In 6/24 horses, variable treatments prior to admission had been recorded which included intrasynovial treatment of the CFJ using a corticosteroid (n = 1) or hyaluronic acid (n = 1), intrasynovial treatment of the proximal interphalangeal joint using a corticosteroid (n = 1) or systemic administration of nonsteroidal anti‐inflammatory drugs, NSAIDs (n = 3).

3.3. Clinical and lameness examination and exclusion of distal limb lameness

Photographs and/or videos were available in all horses except the first four horses presented in 2002 and 2003. Results of the orthopaedic examination are presented in Table 1. Flexion tests were performed on an irregular basis and therefore minimal conclusions could be drawn. Diagnostic anaesthesia of the distal limb was performed in nine horses in addition to the previously described six horses and repeated in further two horses. In 10 horses, the lameness remained unchanged; in one horse, an improvement could be observed after a high 4‐point nerve block without further diagnostic imaging findings. Radiographs of the distal limbs were obtained in 11 horses and did not reveal clinically relevant findings.

3.4. Ultrasonography

In 23/24 horses, ultrasonographic evaluation of the CFJ was performed. In 19 of these 23 horses, ultrasonographic images were available for retrospective evaluation. At the time of examination, the examiner (FG) suspected and documented osteophyte formation in 3/4 of the remaining horses. Following retrospective grading, 1/19 horses had no ultrasonographic signs of OA, five displayed mild, nine moderate and four severe signs of OA (Figure 4). The joint appeared distended in some horses (Figure 4; Video S3). Visual assessment of osteophyte formation was subjectively improved by dynamic ultrasound, if performed (n = 7) (Video S3). Evaluation of joint effusion and dynamic ultrasound were not performed as a standard procedure.

FIGURE 4.

Examples of coxofemoral joints with no (A), mild (B), moderate (C) and severe (D) ultrasonographic changes indicating osteoarthritis (from left to right: control horse and cases number 22, 14, 19). Ac, acetabulum; FH, femoral head; GM, gluteal muscles; GT, greater trochanter. Depicted abnormal findings included joint effusion (cross) as well as osteophytes at the acetabular rim (asterisks) and the edge of the femoral head (triangle).

3.5. Radiography, radiographic score and acetabular ratio

Radiographic images were taken in all horses and available for retrospective evaluation in 20/24 horses. Written documentation described radiographic signs of CFJOA in the remaining four horses. These had all been obtained using the modified ventrodorsal approach. Assessment of the radiographs by the more experienced observer (MH) using the scoring system yielded one horse without radiographic signs of OA, seven horses with mild signs, eight with moderate signs and four with severe signs of OA (Figure 5). Cohen's kappa for inter‐observer reliability of the radiographic grading was 0.846 (p < 0.001), Cohen's kappa for intra‐observer agreement for the more experienced observer was 0.853 (p < 0.001) and for the less experienced observer 0.842 (p < 0.001), all consistent with almost perfect agreement. ⁴³

FIGURE 5.

Examples of coxofemoral joints with no (A), mild (B), moderate (C) and severe (D) radiographic changes indicating osteoarthritis (from left to right: control horse and cases number 9, 17, 16). Depicted abnormal findings include osteophytes (asterisk), irregular or narrow joint spaces (triangle), lack of visibility of an indentation for ligament attachment (square), changes in subchondral bone opacity (circle) and tapering of the femoral head (cross).

Radiographs of 20 horses were available as potential controls for the assessment of the acetabular ratio. Of these 20 horses, seven showed no radiographic signs of CFJOA and were included in the final evaluation. There was no significant difference between the acetabular ratio of the patients (M = 0.547, SD = 0.12) and the seven control horses without CFJOA (M = 0.588, SD = 0.04), t(22.99) = 1.2559; p = 0.2218.

Correlation analysis between radiographic and ultrasound scoring results showed a strong positive correlation (r = 0.64; p = 0.003).

3.6. Intra‐articular anaesthesia of the CFJ

Intra‐articular anaesthesia of the CFJ was performed in 21/24 horses. In one horse, a heavy Black Forest draught breed allowing only poor quality ultrasonographic image guidance for injection, the intra‐articular anaesthesia approach was not successful. In 15/20 horses, lameness was improved afterwards, in 4/20 it was eliminated and in one horse the lameness remained unchanged. In 14/20 horses where the information was available, improvement in lameness was noted after 5 (n = 1), 10 (n = 7), 15 (n = 4) or 20 min (n = 2). Synovia could be obtained for analysis in 17 horses. The total nucleated cell count varied between 0 and 1600/μL, with a mean of 446.2/μL. Total protein values were between 14 and 38 g/L with a mean of 24.4 g/L.

3.7. Treatment, outcome and long‐term follow‐up

Overall, 19/24 horses survived to clinic discharge (Figure 3). In 18 of these 19 horses, information was available on conservative treatments that were performed. Applied treatment options included intra‐articular administration of a corticosteroid such as betamethasone (36 mg, Celestovet, MSD Animal Health GmbH) (n = 2), often in combination with hyaluronic acid (20 mg, Hylartil, Pfizer AG or Hyonate, Boehringer Ingelheim, Pharma, GmbH & Co. KG) (n = 10) or triamcinolone acetonide (10 mg, Triam Lichtenstein, Winthrop Arzneimittel GmbH) combined with hyaluronic acid (n = 3). One horse received 5 mg stanozol (5 mg, Sungate, ACME) and one polyacrylamide hydrogel (Arthramid Vet, Contura Vet) intra‐articularly. Some horses also received systemically administered phenylbutazone (Equipalazone, MSD Animal Health GmbH) (2.2 mg/kg q12h orally), either alone (n = 1) or in combination with one of the above‐mentioned intra‐articular treatments (n = 5).

Five horses received no treatment but were euthanised immediately after diagnosis due to severe lameness and poor prognosis. Six of the horses that were discharged from the clinic were presented for re‐evaluation. Three of these six horses were euthanised after 6 weeks and one horse after 18 months due to persistent severe lameness. One horse was re‐examined after 4.5 months. The gelding still showed a moderate lameness, but the owner refused euthanasia and decided to keep the horse on pasture. One horse showed an improved gait pattern 21 months after the initial examination.

Structured client telephone interviews in April 2023 were available for 10/15 horses that survived short‐term. However, one of these 10 horses was only presented to the clinic in January 2023, accordingly it was not included into long‐term follow‐up analysis. Three months later, in April 2023, the owner reported that the initial lameness had improved but was still visible after treatment of the CFJ using betamethasone in combination with hyaluronic acid. For the remaining nine horses, the structured client telephone interview was conducted 11.4 to 20.1 years after discharge (median long‐term follow‐up 15.3 years). All nine horses were reported deceased in the meantime, had remained lame at a walk for the rest of their lives and were retired. In 5/9 horses, the owners stated that they had had to euthanise them due to persistent severe lameness caused by the CFJOA. The mean survival time after discharge of these five horses was 24 months (6 to 60 months). In 4/9 horses, the owners reported that the horses had been euthanised due to other reasons. Two of these four horses had been euthanised 15 and 17 years after discharge at the age of 31 and 20 years, respectively, due to a deteriorating general condition and weakness. The remaining two horses had been euthanised 15 years after discharge due to colic and 6 years after discharge due to a head trauma, respectively, both animals aged 23 years. The mean survival time of these four deceased horses due to reasons unrelated to the diagnosis CFJOA was 13.25 years.

3.7.1. Post‐mortem findings

Necropsy was performed in 7/9 of the euthanised horses. In gross pathological examination, four horses showed unilateral and three bilateral signs of CFJOA (n = 10 joints).

A main finding in all examined horses included extensive degeneration of articular cartilage and cartilage erosions (n = 8/10 joints) with eburnation of the subchondral bone (Figures 6 and 7). Frequently, new cartilage and bone proliferations were noted at joint margins (n = 7/10 joints) (Figures 6 and 7). Inconsistently, the joint capsule was fibrotic and thickened (n = 3/10 joints).

FIGURE 6.

Examples of coxofemoral joints with osteoarthritic changes during necropsy (case number 1 [A] and case number 19 [B]). Displayed abnormal findings include extensive degeneration of articular cartilage and cartilage with eburnation of the subchondral bone (triangle), new cartilage and bone proliferations at joint margins (asterisk) and ruptured Ligamentum capitis ossis femoris (circle).

FIGURE 7.

Example of comparison of radiographic (A) ultrasonographic (B) and postmortem (C and D) appearance of the coxofemoral joint of a single horse (case number 17). Abnormal findings shown on the radiograph and the ultrasound image include osteophytes (asterisk), irregular or narrow joint spaces (triangle), lack of visibility of an indentation for ligament attachment (square). Abnormal findings depicted on the postmortem images include new cartilage and bone proliferations at joint margins (asterisk), extensive degeneration of articular cartilage and cartilage with eburnation of the subchondral bone (cross).

In 7/10 joints (5/7 horses), the Ligamentum capitis ossis femoris was partially (n = 3/7 joints) or completely (n = 4/7 joints) ruptured unilaterally (n = 3/5 horses) or bilaterally (n = 2/5 horses) (Figures 6 and 7). In one case with partial rupture, a periarticular abscess was found. Furthermore, the ligament was torn at its attachment to the femoral head, with osteomyelitis of the subchondral bone. One of the two bilaterally affected horses displayed a complete ligament rupture on the clinically diseased side and a partial rupture on the contralateral side as the only finding. Only the primarily affected limb displayed a lameness and was examined radiographically. The second bilaterally affected horse had a complete rupture of the ligament on both sides. Furthermore, the less severely affected limb also showed minor cartilage and bone proliferations at joint margins. The horse presented with lameness of the more severely affected limb yet had an overall very stiff gait pattern. Again, no radiographs were taken of the CFJ of the less affected limb.

One of the two cases without rupture of the ligament of the femoral head showed multiple acetabular rim fractures and secondary pseudarthrosis. The other case without ligament rupture was the only horse with bilateral changes of a similar degree. It was the youngest horse (1 year old) in this case series. Additional findings in this horse included a loose cartilage‐bone fragment within the joint as well as bilateral shallowing of the acetabula indicative of hip dysplasia.

3.7.2. Histopathology

Histology revealed degeneration and loss of articular cartilage of the acetabulum and femoral head in five horses, characterised by fibrillation and proliferation of reactive chondrocytes (Figure 8). Pannus formation ⁴⁴ was present in two horses. Fibrosis of joint capsules was found in two animals, accompanied by mild lymphohistiocytic inflammation and haemorrhage as well as villous proliferation of synovial membrane and osteophyte formation. Sclerosis, necrosis and/or fibrosis of subchondral bone was observed in four horses, associated with suppurative osteomyelitis detected in one case. Ruptured ligaments showed lymphohistiocytic infiltrates, metaplastic changes (chondroid and osseous) and/or haemorrhages.

FIGURE 8.

Exemplary histology of joint lesions. Degenerative cartilage with multiple clusters of reactive chrondrocytes (arrows). Haematoxylin–eosin staining.

4. DISCUSSION

This is the first large case series reporting on history, clinical, radiographic and ultrasonographic findings as well as treatment, outcome and results of post‐mortem examinations of horses with CFJOA.

In terms of medical history, the case study shows that CFJOA occurs in horses of all ages, typically presented with long‐lasting lameness visible at a walk, which is often a result of a previous trauma. These results are consistent with previous case reports and textbook descriptions. ² , ⁴ , ⁴⁵ , ⁴⁶ A clinical feature of many horses not previously reported for this disorder, was an extended cranial phase of the stride of the affected limb, which may be a potential strategy to reduce the number of impact phases per distance. Taking into consideration other observed and anecdotally described characteristic clinical signs, such as unilateral gluteal muscle atrophy ⁴⁵ , ⁴⁶ and locomotion on three tracks, ³ , ⁵ , ⁴⁶ we conclude that conformation assessment and gait analysis may already be indicative of CFJOA in affected horses.

Nevertheless, most horses in this case series underwent either regional and/or intra‐articular anaesthesia and/or radiography to rule out pain in the distal limb. This is in line with the few case reports describing CFJOA in adult horses. ² , ³ , ⁴ In these reports, diagnostic tests for exclusion of distal limb disease were either performed ² , ⁴ or at least recommended but not performed due to financial constraints. ³ Accordingly, although a presumptive diagnosis may seem obvious based on the history and clinical signs, coming to a definitive diagnosis still appears to be a challenge and the result of a systematic orthopaedic examination including exclusion of the distal limb as the cause of lameness, even in moderate to severe cases.

For definitive diagnosis, the present case series emphasises the importance of interpreting imaging findings in combination with improvement in lameness after intra‐articular anaesthesia of the CFJ.

Although ultrasonography can be challenging in cases with limited muscle atrophy due to the depth of the relevant structures, acoustic shadowing artefact produced by the greater trochanter and potentially thick skin in heavy breeds, it was successfully applied in all horses. Regarding radiographic examination, the present case series demonstrated that the modified ventrodorsal approach is feasible with a stationary unit in all horses despite some disadvantages including the potentially increased radiation hazard risk to personnel, ⁴⁷ the potential risks of painful responses due to manipulation of the limb in horses with CFJ disorders and subsequent damage to the radiography equipment. ¹⁹ , ³²

Comparison of ultrasonographic and radiographic scores revealed good correlation regarding the visualisation of irregular joint margins. ¹⁶ , ¹⁸ In contrast to radiography, ultrasonography also enables the experienced examiner to assess joint instability and effusion. ¹² In our opinion, it can be used as a reliable modality for quick assessment of the joint to ensure the indication for radiography. ¹⁶ It is a useful diagnostic tool in an ambulatory as well as hospital setting, if the required expertise is available and radiography not accessible. ¹² Despite these advantages, however, radiography is particularly informative as it allows detailed objective assessment of several well‐defined diagnostic parameters of CFJOA, which were summarised in the present study in a radiographic score. The authors propose the score as standard to diagnose CFJOA radiographically in adult horses. It is particularly interesting since computed tomographic examination of the hip, although described in smaller equids such as ponies and foals, ⁶ is currently not yet readily available as a standard procedure in adult horses.

Due to the small number of control cases, the results of the evaluation of the acetabular ratio need to be interpreted with caution. Unlike the radiographic score, the acetabular ratio does not appear to be valuable in identifying horses with CFJOA. Nevertheless, subjective flattening or tapering of the femoral head was noted in some cases and was incorporated into the radiographic score. However, the deformation of the femoral head appears to be variable and inconsistent, suggesting that it is part of a maladaptive modelling process rather than a major feature indicative of a specific cause of the disease, such as a dysplasia might be. ⁴⁰

In this case series, intra‐articular anaesthesia was performed after ultrasonography and radiography. Noticeably, intra‐articular anaesthesia of the CFJ led to lameness improvement rather than a complete elimination in most cases. Furthermore, a positive response was commonly observed after 10 to 20 min. This confirms previous anecdotal reports of a late and incomplete improvement in lameness in horses with CFJOA. ⁴ , ⁴⁶ Possibly, one reason for this was the advanced condition of the disorder in many of the horses, which could result in delayed improvement rather than complete elimination of the associated pain. Assuming that joint instability due to discontinuity of the Ligamentum capitis ossis femoris occurs early after trauma or increases with disease progression, a mechanical lameness component is also conceivable which may not improve after intra‐articular anaesthesia. Given the severity of the condition and the size of the joint, the use of larger amounts of local anaesthetics and reassessment of lameness more than 30 min after injection may be considered, as more time may potentially be needed to efficiently desensitise areas such as the Labrum acetabulare and the Ligamentum capitis ossis femoris. The proximity of the sciatic nerve to the joint, however, needs to be considered as potential limitation for usage of larger volumes. ¹³ To conclude, intra‐articular anaesthesia remains the gold standard for allocating pain to the CFJ and thereby reliably rule out other concurrent pathologies identified on imaging.

Five horses were euthanised as a direct consequence of the severity of the disease. Despite medical interventions, 4/6 horses presented for re‐examination had to be euthanised due to persistent severe lameness, and long‐term follow‐up telephone conversations revealed that all but one horse remained lame at a walk. A further five horses were euthanised due to the disease at home. This confirms what has been suspected ⁴⁵ , ⁴⁶ : the overall short‐ and long‐term prognosis for horses with radiographically moderate to severe CFJOA is poor. It has to be added that most of the cases in the present study were already long‐lasting at admission and the suffering of the horses was high. This could be due to a biased case population of a referral centre. Another reason might be the low prevalence of CFJOA in adult horses and the challenging diagnostic approach. Therefore, early recognition of CFJOA as a potential cause of lameness, especially after a fall or external trauma, is recommended. In any patient where the distal limb has been excluded as cause of lameness, the systematic workup should be continued proximally. Even minor imaging findings in the CFJ might be of importance, as they may represent an early stage of the disease. For cases recognised early, intra‐articular analgesic and anti‐inflammatory treatment might still be considered and the following phase of temporary improvement of clinical signs could be used to improve the overall gait pattern through a targeted rehabilitation programme to establish a more adequate gluteal and biceps femoris musculature. In advanced cases, on the other hand, based on the results of the present study, unnecessary long‐term suffering can be avoided by advocating with owners for timely euthanasia.

An interesting observation during necropsy was that gross pathological findings were closely associated with joint instability in all investigated horses. In five out of seven horses partial or complete rupture of the Ligamentum capitis ossis femoris could be observed, ³ , ⁴ , ⁵ suggesting joint instability to be a key pathogenetic factor for OA of the CFJ. Due to the frequent history of trauma, it is more likely that rupture of the ligaments of the head of the femur is rather the trigger than the consequence of OA, although rupture secondary to degeneration of the ligament has to be considered. Acetabular rim fractures, which were found in one horse as well as dysplasia which was present in the only non‐adult case also cause joint incongruency and instability and thereby promote OA. ⁶ , ⁷ , ⁸ , ⁹ , ¹² Considering joint instability as an important aetiological factor for CFJOA, it may also serve as an explanation for the poor prognosis in this case series because, as such, it cannot be causally treated by medication per se. Consequently, to avoid prolonged suffering, efforts to assess joint instability in suspected cases should be intensified. Targeted and systematic use of dynamic ultrasound ¹² and the use of arthrography, computed tomography or arthroscopy ⁴⁸ are of potential benefit. Debridement of partial tears of the ligament of the femoral head may even present a treatment option, allowing a return to soundness in selected, less severe cases. ⁴⁸

Limitations of this study include the limited sample size, the retrospective analysis of partly incomplete patient records and the subjective outcome assessment based on owner follow‐up with potential recall bias. Furthermore, most included horses displayed moderate to severe rather than mild clinical signs as well as imaging findings. This must be taken into account during orthopaedic examination of future cases as well as when making statements about prognosis. It also complicates validation and application of the radiographic score, as horses with mild clinical signs were underrepresented. Accordingly, evaluation of the association between clinical and imaging findings, necropsy results and outcome could not be performed.

To conclude, CFJOA is a rare cause of moderate to severe, often trauma‐related, long‐lasting lameness in horses. It should be considered as a differential diagnosis in proximal hindlimb lameness. Characteristic clinical signs of advanced CFJOA may be indicative, and definitive diagnosis should be made based on imaging findings starting with ultrasonography followed by standing radiography and lameness improvement after subsequent intra‐articular anaesthesia. Clinicians and horse owners have to be aware of the poor prognosis of advanced CFJOA even after intra‐articular treatment. This is especially true, when arthropathy is associated with instability of the joint.

FUNDING INFORMATION

Not applicable.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interests.

AUTHOR CONTRIBUTIONS

Fay J. Sauer: Conceptualization; data curation; formal analysis; investigation; methodology; validation; visualization; writing – original draft; writing – review and editing. Maren Hellige: Conceptualization; data curation; investigation; methodology; writing – review and editing. Andreas Beineke: Investigation; writing – review and editing. Florian Geburek: Conceptualization; data curation; investigation; methodology; project administration; supervision; validation; writing – review and editing.

DATA INTEGRITY STATEMENT

Fay Sauer composed the manuscript and had full access to all data and takes responsibility for the integrity of the data and accuracy of the data analysis.

ETHICAL ANIMAL RESEARCH

Research ethics committee oversight not required by this journal: retrospective study of clinical records.

INFORMED CONSENT

Explicit owner consent for animals' inclusion in the study was not stated.

PEER REVIEW

The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer‐review/10.1111/evj.14053.

Supporting information

Text S1: Structured client telephone interview for follow‐up in April 2023.

Video S1: Video displaying an example of the clinical finding during lameness examination. The horse (case number 18) in the video has a right sided coxofemoral joint osteoarthritis. It shows lengthening of the cranial phase of the stride of the affected limb.

Video S2: Video displaying an example of the clinical lameness examination. The horse (case number 19) in the video has a right sided coxofemoral joint osteoarthritis. It shows lengthening of the cranial phase of the stride of the affected limb, an external rotation of the limb and locomotion on three tracks.

Video S3: Example video of a dynamic ultrasound examination of right coxofemoral joint. The femoral head moves within the acetabulum. There are marked joint effusion as well as mild osteophytes at the acetabular rim and the edge of the femoral head.

Sauer FJ, Hellige M, Beineke A, Geburek F. Osteoarthritis of the coxofemoral joint in 24 horses: Evaluation of radiography, ultrasonography, intra‐articular anaesthesia, treatment and outcome. Equine Vet J. 2025;57(1):101–114. 10.1111/evj.14053

DATA AVAILABILITY STATEMENT

Data sharing exemption granted by the editor for this descriptive retrospective case series.