Clinique vétérinaire

du Dr Bardet

Arthroscopy of the Elbow in Dogs
Part I: The Normal Arthroscopic Anatomy Using the Craniolateral Portal

From a Referral Surgical Practice, Neuilly sur Seine, France.
The technique of arthroscopy of the elbow joint using the craniolateral portal is described in dogs.

The technique of arthroscopy of the elbow joint using the craniolateral portal is described in dogs. The technique was developed on ten cadavers and six experimental dogs. It allows a direct visualization of the cranial bony structures of the elbow of the cranial recess as well as of the humeroulnar joint and anconeal process through the supra-trochlear foramen. The technique appears safe and reliable. Postoperative neurovascular injuries were not observed.

Introduction

The first mention of arthroscopy of the elbow in the human orthopaedic literature was by Michael Burman in 1931 (1-4). At this time he reported on arthroscopy performed with a 3 mm diameter arthroscope in cadavers. He stated that the elbow was "unsuitable for examination" and that the "anterior puncture of the elbow is out of question". In 1971 Watanabe developed the 1.7 mm arthroscope to use in small joints (2). In 1979 Johnson reported a clinical study of elbow arthroscopy (3). Elbow arthroscopy is now a well-established procedure both for diagnostic and therapeutic purposes in human medicine (4-6). The procedure can be used for the diagnosis of intraarticular lesions, the removal of loose fragments and foreign bodies, irrigation of the joint, debridement of an infected joint, excision of osteophytes, synovectomy, capsular release and excision of the radial head (4-6).

Elbow arthroscopy in the dog was first reported by van Ryssen in normal dogs using a medial approach (7). A subsequent study compared the radiographic, arthroscopic and CT scan findings in dogs presented with clinical signs of elbow lameness (8). It was shown that arthroscopy of the elbow allows the diagnosis of elbow lesions before the development of osteoarthritis changes, thus avoiding arthrotomy in doubtful cases.

The goal of this paper is to describe the normal arthroscopic anatomy of the elbow in dogs and the craniolateral portal for elbow arthroscopy.

Materials and Methods

Arthroscopic instrumentation

The instrument used was a 2.7 mm, 30 degree fore oblique arthroscope with a 3.5 mm diameter arthroscopic sleeve. Light was supplied with a xenon light source. The arthroscopic procedure was visualized via a camera mounted in the scope and a monitor. The joint was distended and irrigated with Lactate Ringer's solution. Photographic documentation was made with a colour printer. If required the procedure was recorded on a videotape.

Cadaver studies

Elbow arthroscopy requires a complete understanding of the regional anatomy. A thorough knowledge of both the superficial and intraarticular structures is essential for the procedure to be safe and effective. In order to avoid injury to the nearly neurovascular structures, preliminary studies on the anatomy of the elbow were performed on 10 canine cadavers weighing 28-40 kg. The bony structures, joint capsule (Figs. 1, 2) muscles and neurovascular structures were studied.

Fig. 1: Schematic drawing of the elbow (lateral view of the left elbow) [modified from Anderson (9)]. Small arrow: needle and egress cannula portal: larger arrow: craniolateral puncture site.
Fig. 2: Schematic drawing of the elbow (caudal view of the left elbow) [modified from Anderson (9)]. Arrow: puncture site of needle and egress cannula

Experimental study

An experimental study on the arthroscopic anatomy of the elbow to locate puncture sites was performed in six canine elbows of normal patients weighing 25-38 kg.

The dogs were anaesthetized and positioned in lateral recumbency. The upper limb was prepared aseptically and draped, allowing full mobility of the front limb. The elbow was punctured with a 19-gauge needle, the puncture site being located medially to the lateral epicondylar ridge and cranial to the olecranon tuberosity pointing to-ward the olecranon fossa (9) (Fig. 2). Lactate Ringer's solution (3-10 ml) was injected, confirming the distension of the joint capsule. The needle was then removed and replaced by the 3 mm egress cannula'. The craniolateral portal was next located with a 19-gauge needle cranial and distal to the lateral humeral condyle (10) (Fig. 3), the lateral humeral condyle being identified by deep palpation. The ideal location is 1.2 cm cranial and 1 cm distal to the lateral humeral epicondyle. After incising the skin and joint capsule with an #11 scalpel blade using a blunt trocar, the arthroscopic sleeve was inserted into the joint. When the joint capsule had been penetrated, the blunt trocar was replaced by the arthro-scope. Systematic examination of the elbow joint was then performed starting with medial humeral condyle, to the medial collateral ligament, the medial coronoid process, the humeral trochlea, the radial head, the capitulum, the olecranon fossa, and the anconeal process through the olecranon fossa.

Fig.3: Nerves and arteries around the elbow joint: lateral view of the right elbow (10)

Results

Normal arthroscopic anatomy

In all joints, the following structures could be identified, moving the telescope from medial to lateral and then through the olecranon fossa: the capitulum (Figs. 4-7), medial collateral ligament (Figs. 4-5), medial coronoid process (Figs. 4-5), radial head (Fig. 4, 7), lateral humeral condyle (Fig. 4, 7), the olecranon fossa (Fig. 8), anconeal process (Fig. 8), and synovial membrane (Figs. 4, 6, 8). In two elbows, the annular ligament appeared slightly distal to the radial head (Fig. 9). In two elbows penetration through the supratrochlear foramen of the distal humerus was impossible.

Visualization of the medial coronoid process was allowed when the arthroscope progressed from a superficial position (Fig. 6) to a deeper position (Figs. 4, 5) with the light-post of the arthroscope at a two o'clock position for the left elbow (Figs. 4-6) and at 10.00 o'clock position for the right elbow (Fig. 7). The cranial recess of the humeroulnaroradial joint was improved by increasing the intraarticular joint volume by turning the egress cannula off (Figs. 6 and 7).

Fig. 4: Cranial arthroscopic view of the right elbow through the craniolateral portal.

1. Medial humeral condyle (capitulum);
2. Radial head
3. Medial coronoid process;
4. Medial collateral process;
5. Buble.
Fig. 5: Cranial arthroscopic view of the right elbow through the craniolateral portal, close-up view. Anatomical structures are identified as in Figure 4.
Fig.6: Arthroscopic view of the right elbow.

1. Medial humeral condyle;
2. Lateral humeral condyle;
3. Radial head;
4. Synovial villi.
Fig. 7: Craniolateral arthroscopic view of the cranial compartment of the left elbow

1. Medial humeral condyle;
2. Lateral humeral condyle;
3. Radial head;
4. Joint capsules.
Fig. 8: Arthroscopic view of the humeroulnar joint through the supratrochlear foramen using the craniolateral portal.

1. Olecranon fossa;
2. Anconeal process;
3. Olecranon ligament;
4. Joint capsule;
5. Joint cavity.
Fig. 9: Arthroscopic view of the proximal radius (1) and annular ligament (2).

Complications and iatrogenic damage

In initial educational procedures, dissection of the joints of the cadavers revealed scarification of the cranial border of the radial head in two of ten elbows. The trocar had passed cranial to pronator teres muscle or through the insertion of the pronator teres muscle. Signs of neurovascular damage were not seen.

Technical problems during the procedure, such as haemorrhage, obstruction of the view by synovial villi, insufficient exposure of the articular cartilage or severe iatrogenic were not observed.

Periarticular fluid accumulation was seen in two of the six dogs. Fluid accumulation resolved within 24-48 hours. None of the dogs was lame nor showed signs of pain during postoperative physical examinations.

Discussion

Arthroscopy is an emerging technique in small animal surgery. Arthroscopy of the elbow joint is also a new technique in human medicine (4-6). Although historically it was thought that arthroscopy of the elbow was too dangerous to be of clinical use, evolution in techniques and equipment, especially over the past decade, has proved it to be an acceptable alternative for the treatment of many lesions of the elbow.

In the veterinary literature the normal anatomy of the elbow joint in dogs was described using a medial approach (7). The approach only allows the visualization of the medial compartment of the elbow including the medial humeral condyle, medial collateral ligament, medial coronoid process, middle and caudal part of the radial head, lateral humeral condyle, lateral coronoid process and the ventral aspect of the anconeal process. The craniolateral portal allows access to the cranial humerora-dioulnar joint cavity with direct visualization of both humeral condyles, radial head, medial coronoid process, collateral ligaments, synovial membrane and the cranial recess of the elbow joint. In most dogs, when the arthroscope is properly positioned, the supratrochlear foramen gives access to the humeroulnar joint. The ideal position of the trocar sleeve, as described, facilitates visualization of the humeroradioulnar joint spaces but penetration through the supratrochlear foramen is more difficult. In two cases the portal was too low to give access to this foramen. When observation of the humeroulnar joint is the primary goal, the portal should be located higher, at the level of the lateral humeral epicondyle. It allows access to the humeroradioulnar joint.

A technique for elbow arthroscopy has been described in the horse (11) and is used to treat osteochondritis and sub-chondral bone cysts.

Arthroscopy of the human elbow has only been developed over the last decade. Proper examination of the elbow joint required several portals over the anterolateral, anteromedial, postero-lateral and direct posterior portals (4). Diagnostic arthroscopy is indicated when the elbow is chronically painful. The most common arthroscopic surgical procedures are: the removal of loose bodies, chondroplasty of osteochondritis dissecans, abrasion arthroplasty for degenerative joint disease, arthrolysis of contractures, evaluation of the ulnar collateral ligament. Less common surgical arthroscopic procedures of the elbow are: synovectomies in rheumatoid arthritis, treatment of synovial chondromatosis and pigmented villonodular synovitis, plicae excision, treatment of septic arthritis, diagnosis of posterolateral rotatory instability as well as treatment of some elbow fractures (4-6).

Because of direct access over the lesions in dogs the medial approach appears advantageous in treating the fragmented medial coronoid process (8). Since arthroscopy allows the diagnosis of elbow lesions before the development of osteoarthritic changes, exploratory arthrotomy is not necessary. However the medial approach shows limitations because of the inability to visualize the cranial aspect of the elbow joint. The cranial approach offers this advantage. It allows a direct inspection of both the cranial aspect of the medial coronoid lesions (12), removal of loose bodies in the cranial recess of the elbow (12) and treatment of fractures of the medial humeral condyle (13). The same craniolateral approach allows direct inspection of the humeroulnar joint through the supratrochlear foramen in eight of the ten cadavers and in all six experimental dogs. This approach appears useful for all conditions of the dorsocaudal compartment of the elbow, without using a third caudolateral or true caudal approach as is used in human patients.

Acknowledgement The author gratefully recognizes the aid of Karl Storz GmbH + Co, D-78532 Tuttlingen, Germany for supporting the publication of the coloured pictures.

What is it?

Figure 1
Figure 2

Number 28— Question

As an exercise in diagnosis V.C.O.T. publishes a radiograph, or image "puzzle" in each issue. The diagnosis and a short description are published on page 48. The author of this section is: H. Dobson, BVM & S, DVSc, MRCVS, Cert E. 0. Radiologist — Ontario Veterinary College, University of Guelph.

History and signalment: A four-year-old Holstein cow was presented with a right hind lameness of two weeks duration. No specific abnormalities were identified on physical examination. A series of radiographs consisting of dorsoplantar, lateral, dorsolateral-plantaromedial oblique and dorsomedial-plantarolateral oblique projections of the distal extremity was made. The dorsoplantar and dorsomedial-plantarolateral projections are illustrated in Figures 1 and 2.

REFERENCES

1. Boe S. Arthroscopy of the elbow: Diagnosis and extraction of loose bodies. Acta Orthop Scand 1986; 54: 637.

2. Shaffer B. Parisien PS. Elbow arthroscopy. Surg Rounds Orthop 1989: 3: 113.

3. Johnson LL. Diagnostic and Surgical Arthroscopy: The Knee and Other Joints. 2nd ed. St. Louis: Mosby 1981: 129-36.

4. Andrews JR, Soffer SR. Elbow Arthroscopy. St. Louis: Mosby 1994:1-11.

5. O'Driscoll SW, Morrey BF. Elbow arthroscopy. J Bone Joint Surg 1992; 74 A: 84.

6. Poehling GG. Arthroscopy of the elbow. J Bone Joint Surg 1994; 76A: 1265.

7. Van Ryssen B. Van Bree H, Simsoens P. Elbow arthroscopy in clinically normal dogs. Am J Vet Res 1993; 1: 191.

8. Van Ryssen B. Van Bree P. Elbow arthroscopy. ECVS proceedings, June 1995.

9. Anderson WD. Anderson BG. Atlas of Canine Anatomy. Philadelphia: Lea & Febiger. 1994: 860-2.

10. Evans HE, Christensen GC, Millers. Part 11 of the article will be published in the next Anatomy of the Dog, 2nd ed. Philadelphia: issue. Saunders 1979: 989.

11. Nixon AJ. Arthroscopic approaches and intraarticular anatomy of the equine elbow. Vet Surg 1990; 19: 191-8.

12. Bardet JF. Arthroscopy of the elbow in dogs: the cranial portals in the diagnosis J. F. Bardet, Dr. vet., MSc, Dipl. E. C. V. S. and treatment of the lesion of the coronoid Referral Surgical Practice process. V.C.O.T. (in press). 32, rue Pierret 92200 Neuilly-sur-Seine.

13. Bardet JF. Traitement de deux cas de fracture du condyle humeral medial sous con- France (role arthroscopique. Act Vet 1996; 1360: Phone 33/1/46 41 05 93 14-7. Fax 33/1/40 88 32 67

Arthroscopy of the Elbow in Dogs
Part II: The Cranial Portals in the Diagnosis and Treatment of the Lesions of the Coronoid Process

The cranial portals of the 38 elbows in 34 dogs allowed proper evaluation and treatment of the lesions of the medial coronoid process. A classification in 7 types of lesions of the medial coronoid process was established. All dogs were treated successfully by either removal of the FCP or by a proximal ulnar sliding osteotomy.

Summary
The craniolateral portal of the elbow is described. The technique was applied and evaluated in 34 dogs (38 elbows). A detailed description and classification of the fragmented coronoid process (FCP) is given. All patients with FCP were treated successfully using a second craniomedial portal. Complications were not observed. The technique appears to be safe and reliable and could also be used for other procedures such as removal of loose bodies, treatment of osteochondritis dissecans of the medial condyle, reduction of selective humeral condylar fractures, excision of osteophytes and other diagnostic purposes. It also eliminates the need for exploratory arthrotomies of the elbow joint in dogs.

Introduction

Elbow arthroscopy in the dog was first reported by van Ryssen et al. in normal dogs using a medial approach (7). A subsequent study compared radiographic, arthroscopic and CT scan findings in dogs presented with clinical signs of elbow lameness (8). It was shown that arthroscopy of the elbow allows the diagnosis of elbow lesions before the development of osteoarthritic changes, thus avoiding arthrotomy in doubtful cases.

The diagnostis of fragmented coronoid process (FCP) in young dogs (five to seven months) is usually based on clinical signs and radiographic evaluation (9-14). Because of the presence of osteoarthritis in mature animals, the diagnosis of fragmented process is easier than in young dogs (9-14). Since the fragments lie hidden between the head of the radius and the coronoid process, FCP can rarely be seen on conventional radiographs, regardless of the projections used. Very rarely is a fragment attached to the coronoid process so loosely that it can be seen as a separated ossicle. The most significant radiographic sign of FCP is osteophyte formation on the proximal margin of the anconeal process (12-14). These osteophytes do not appear until the dog is seven to eight months old. Techniques for maximizing the diagnosis of FCP have been described, including linear tomography (15), arthrography (16), computer tomography (17, 18) and arthroscopy (7, 8).

The goal of this paper is to describe the cranial portals for elbow arthroscopy, as well as the most common findings in dogs suffering from fragmented coronoid process.

Materials and Methods

Arthroscopic instrumentation

The instrument used was a 2.7 mm. 30 degree fore oblique arthroscope with a 3.5 mm diameter arthroscopic sleeve. Light was supplied with a xenon light source. The arthroscopic procedure was visible on a monitor using a camera. The joint was distended and irrigated with Lactate Ringer's solution. Photographic documentation was made with a color printer. The procedure was recorded on videotape, as required.

Clinical study

A retrospective study was done on 34 dogs with 38 elbows showing clinical and/or radiographic signs of fragmented coronoid process (FCP). There were 20 right elbows and 18 left elbows. Four dogs had bilateral elbow involvement. Twelve breeds were represented (Table 1). Labrador (13 dogs), German Shepherd dogs (6) and Rottweiler (5) were most commonly affected by FCP. There were 24 males and 10 females. The mean age was 18 months (range: five months to seven years).

The technique of arthroscopy is described in part I of the paper (V.C.O.T. 1996: 10: 1-5). When an instrument portal was required to use a probe or a grasping forceps, a craniomedial portal was created from inside out. After the craniolateral portal had been created and the cranial aspect of the elbow joint had been inspected arthroscopically, the telescope was removed from the sheath and, with the sheath still in place and the elbow flexed at 90 degrees, a blunt 2.5 mm rod was passed through the sheath and directed medially and cranially to the medial collateral ligament. In order to protect the radial nerve and allow easy access to the medial coronoid process by the grasping forceps, the rod was located at the humeroulnar junction. The rod was then advanced through the craniomedial portion of the joint capsule and sub-cutaneous tissue until it tented the skin. A scalpel blade was then used to make a skin incision over the top of the rod.

The rod was then advanced through the skin so that a cannula could be placed over it from the craniomedial side and into the joint. This craniomedial portal is an excellent working portal for instrumentation and manipulation. A grasping forceps was then introduced into the joint through this craniomedial portal.

In dogs in which the coronoid process was not fragmented and showed a step between the level of the radial head and of the medial coronoid process, a proximal ulnar sliding osteotomy was performed (21); a postoperative arthroscopy of the elbow was done.

At the end of the procedure one skin suture was used to close each of the three portals. All of the dogs were al-lowed to ambulate the same day and were discharged the day after the operation.

Table 1: List of the breeds affected by FCP
Table 2: Arthroscopic classification of lesions of the medial coronoid process in dogs

Results

All of the elbows suspected of having a fragmented coronoid process were subdivided in one of the following seven groups:

I — Fragmentation of part of the medial aspect of the medial coronoid process (Fig. .1): two cases
II — Abrasion of the cartilaginous part of the medial coronoid process (Fig. 2): one case
III — Free medial coronoid process: 13 cases (Fig. 3)
IV — Fissure of the medial coronoid process: six cases (Fig. 4)
V — Elevation of the medial coronoid process in relation to the radial head: 10 cases (Fig. 5)
VI — Osteophytic coronoid process without surelevation and fissuration (Fig. 6): three cases
VII — Joint mouse (Fig. 7): three cases

Fig. I: Fragmented left medial coronoid process: Type I fragmentation.

1. Capitulum;
2. Radial head;
3. Medial coronoid process;
4. Fragmented medial coronoid process.
Fig. 2: Type II cartilaginous erosion of the lateral rim of the medial coronoid process.

1. Capitulum;
2. Radial head;
3. Medial coronoid process;
4. Cartilaginous erosion.
Fig. 3: Type III fragmented loose coronoid process.

1. Capitulum;
2. Radial head;
3. Loose fragmented medial coronoid process.
Fig. 4: Type IV fissure of the medial coronoid process.

1. Capitulum;
2. Radial head;
3. Fissured medial coronoid process;
4. Medial collateral ligament;
5. Hyperaemic synovial villi.
Fig. 5: Type V anomaly of the medial coronoid process.

1. Capitulum;
2. Radial head;
3. Medial coronoid process;
4. Medial collateral ligament;
5. Hyperaemic synovial villi.
Fig. 6: Type VI anomaly.

1. Capitulum;
2. Radial head;
3. Osteophytic medial coronoid process;
4. Medial collateral ligament;
5. Hypertrophic synovial villi.
Fig. 7: Joint mouse in the cranial pouch of the right elbow.

1. Distal humerus;
2. Osteophytic radial head;
3. "mouse";
4. Hypertrophic synovial membrane.

In the three patients with loose fragments in the cranial compartment of the elbow joint (Fig. 7), the loose body was grasped using a small forceps (Fig. 8). The medial coronoid process was inspected (Fig. 8) and when fragmented removed (Fig. 9). In dogs with a large fragmented medial coronoid process it was difficult to see the fragmentation (Fig. 10). In that instance the craniomedial portal was helpful. The rod was inserted through the arthroscopic sheath and cranial to the medial collateral ligament (Fig. 11). Using the rod, the fragmented coronoid process was freed. The cannula was then placed over it and into the joint (Fig. 12). The loose coronoid was then grasped through the craniomedial portal (Fig. 13) before a final examination of the joint took place (Fig. 14). In every case of FCP the loose fragment was removed successfully using those two cranial portals. If an abrasion of the medial margin of the medial coronoid process was present, all of the abnormal cartilage was curetted out. In type V, the step between the articular cartilage of the radial head and that of the medial coronoid process was obvious (Figs. 5, 15). In all of the cases, the articular cartilage of the medial humeral condyle appeared fibrillated and hondroinalacic (Figs. 5, 15). Postoperative arthroscopy, following the proximal ulna sliding osteotomy, showed a decrease of the height of the step between the articular surface of the radius and radial head (Fig. 16). In all ten eases the osteotomy had healed eight to twelve weeks after the operation (Fig. 17).

Fig. 8: Excision of a joint mouse from the cranial compartment of the elbow joint.

1. Capitulum;
2. "mouse";
3. Grasping forceps.
Fig. 9: Craniolateral view of the elbow after removal of the joint mouse.

1. Distal humerus;
2. Medial coronoid process;
3. Hyperaemic synovial villi.
Fig. 10: Craniomedial portal view of the fragmented coronoid process of the right elbow.

1. Capitulum;
2. Radial head;
3. Fragmented coronoid process;
4. Medial collateral ligament.
Fig. 11: Insertion of the rod cranial to the medial collateral ligament;

1. Chondromalacic capitulum;
2. Loose medial coronoid process;
3. Rod.
Fig. 12: The craniomedial portal from inside out.

1. Capitulum;
2. Radial head;
3. Medial fragmented coronoid process;
4. Medial collateral ligament;
5. Craniomedial cannula;
6. Hypertrophic synovial villi.
Fig. 13: Introduction of the grasping forceps through the craniomedial cannula.

1. Capitulum;
2. Medial coronoid process;
3. Loose fragmented coronoid process in the jaws of the grasping forceps;
4. Grasping forceps.
Fig. 14: Postoperative view of the elbow joint after removing the loose fragmented coronoid process.

1. Capitulum;
2. Radial head;
3. Medial coronoid process;
4. Hyperaemic synovial villi.
Fig. 15: Type V before the proximal ulnar sliding osteotomy. Note the step between the level of the articular cartilage of the coronoid process (3) and that of the radial head, (2) as well as the superficial erosion of the articular cartilage of the capitulum (1).
Fig. 16: Postoperative view of the elbow joint following the ulnar osteotomy. Note that the step between the medial coronoid process (3) and that of the radial head (2) has decreased.

In one patient the elbow was normal in spite of the clinical suspicion of fragmented coronoid process. In one German Sheperd dog an immune-mediated polyarthritis was diagnosed on the basis of the synovial fluid analysis and synovial biopsies. In three patients a "mouse" was found in the cranial cavity of the humeroradioulnar joint in addition to a loose fragment coronoid process.

Fig. 17A: Preoperative view of the left elbow in an 11-month-old Labrador Retriever of Figures 15 and 16. Note the absence of secondary changes of osteoarthritis.
Fig. 17B: Postoperative view showing the proximal ulnar sliding osteotomy. Note the caudal displacement of the proximal ulnar segment at the Osteotomy site.
Fig. 17C: 14 months postoperative "follow-up" view of the elbow showing the absence of osteoarthritis of the elbow and the healed osteotomy.

Discussion

Arthroscopy is an emerging technique in small animal surgery. Arthroscopy of the elbow joint is also a new science in human medicine (4-6). Although historically it was thought that arthroscopy of the elbow was too dangerous to be of clinical use, an evolution in techniques and equipment, especially aver the past decade, has proved it to be an acceptable alternative for the treatment of many lesions about the elbow.

Dogs younger than seven to eight months may present clinical signs suggestive of FCP but without radiographic abnormalities consistent with the condition. These patients are reevaluated in four to five weeks because clinical lameness often precedes radiographic changes (10, 12). As degenerative changes progress, osteophyte formation on the medial coronoid process, the cranial surface of the radial head and the medial humeral epicondyles becomes apparent (10-12, 22-35). The development of osteophytes in these locations is not unique to FCP; osteophytes occur in the same location with osteochondrosis of the medial humeral condyle and un-united anconeal process. Osteophyte formation can be minimal when the me-dial coronoid process is fissured (23, 24). In these instances, the diagnosis of fragmented coronoid process remained a challenge. Linear tomography has been occasionally used as a diagnostic tool (14). In a comparison of radiologic imaging techniques of the fragmented coronoid process it was shown that there was a significant difference between the diagnostic abilities of plain-film radiography, xeroradiography or linear tomography of the cubital joint (17). The combination of plain-film radiography and linear tomography provided an improvement in accuracy, approaching that of computed tomography (17). Computed tomography had the highest accuracy (86.7%), sensitivity (88.2%) and negative predictive value (84.6%) (16). Since the diagnosis of FCP may be difficult to confirm arthrotomy has been recommended (8-12, 24-34).

However arthrotomy alone induces numerous pathological alterations in the canine stifle with long-standing de-generative changes in all ligaments (35). Furthermore exploratory arthrotomy in dogs with suggestive clinical and/or radiographic changes of FCP, but not having the disease should be eliminated. Arthroscopy of the elbow joint allows direct visualization of the medial coronoid process, as well as the evidence of fissures, osteophytes, joint mice, cartilage abnormalities and grading of the osteoarthritis.

This direct visualization of the medial coronoid process was helpful in classifying the different types of abnormalities. The medial coronoid process appeared abnormal in 23 of the 38 elbows. In ten patients, the articular surface of the medial coronoid process was above that of the radial head (type V); in all of these patients the articular cartilage of the medial humeral condyle appeared abnormal: fibrillated, chondromalacic or with a "kissing" lesion. When the FCP was loose the abnormal biomechanics of the joint could not be evaluated but from Wind's work (30, 31), the abnormal biomechanics of the elbow joint is suspected. From this classification seven major classes of abnormalities appear: the first one where the level of the articular cartilage of the medial coronoid process is at the same level as that of the radial head (Types III, IV, VI) and a second class where the medial coronoid appears elevated (Types III and V). This classification may support both theories of the pathogenesis of FCP related to either an osteochondrosis aetiology (22, 25) or a mechanical cause (8, 30, 31).

Olson described FCP as a result of osteochondrosis (11, 12). Wind has demonstrated that FCP appears as a result of an abnormal mechanic of the humercubital joint (30, 31). Type II and type III may result from an excessive pressure on the medial coronoid process on the capitulum. Type V lesions, even if the medial coronoid process is not fragmented or fissured appears to induce a pathology of the medial humeral condyle consistent with the clinical signs. The proximal ulnar sliding osteotomy (21) allows a caudal shift of the proximal ulna at the osteotomy site (Fig. 17), secondary to the forces applied on the olecranon by the triceps muscle pull. The craniolateral postoperative arthroscopic view obviates the decrease of the step between the levels of the articular surface of the medial coronoid process and radial head (Fig. 16). This correlates with the long-term clinical improvement of these patients (21).

The type I lesion obviates the absence of a step between the medial coronoid process and radial head; the articular cartilage of the medial humeral condyle is also normal: there is a minimal associated synovitis: the lesion is very localized and does not involve the entire lateral aspect of the medial coronoid process. Because of the lack of evidence of abnormal biomechanics of the elbow an aetiology of osteochondritis is probable.

The craniolateral portal may be disadvantageous in the full evaluation of the medial compartment of the elbow joint since only the cranial part of the medial coronoid process is visualized. This portal does not give access to en-tire articular surface of the semilunar notch and does not allow a proper inspection of the articular surface of the entire medial humeral condyle. A prospective comparison of both medial and craniolateral portals would be required. However this portal is advantageous for viewing any pathology of the humeral condyles and allows an easy access to the olecranon fossa and olecranon, using only one portal. This portal also allows easy removal of loose bodies (three of 36 FCP) in the cranial pouch of the elbow joint and a full inspection of the synovial membrane.

In conclusion, the craniolateral portal provided a safe access to the cranial aspect of the elbow joint in dogs. The anconeal process and olecranon fossa could be seen through the olecranon foramen in most cases. All of the fragmented coronoid processes were re-moved successfully through a craniomedial portal. Direct visualization of the medial coronoid allows detection of the fragmented coronoid process before secondary radiographic changes of osteoarthritis occur. This portal also appears advantageous when a loose body in the cranial pouch is present.

REFERENCES

1. Boe S. Arthroscopy of the elbow: Diagnosis and extraction of loose bodies. Acta Orthop Scand 1986; 54: 637.

2. Shaffer B. Parisien PS. Elbow arthroscopy. Surg Rounds Orthop 1989; 3: 113.

3. Johnson LL. Diagnostic and Surgical Arthroscopy: The Knee and Other Joints. 2nd ed. St Louis: Mosby 1981; 129-36.

4. Andrews JR, Soffer SR. Elbow Arthroscopy. St Louis: Mosby 1994; 1-11.

5. O'Driscoll SW. Morrey BF. Elbow arthroscopy. J Bone Joint Surg 1992; 74A: 84.

6. Poehling GG. Arthroscopy of the elbow. J Bone Joint Surg 1994; 75A: 1265.

7. Van Ryssen B, Van Bree H, Simsoens P. Elbow arthroscopy in clinically normal dogs. Am J Vet Res 1993; 1: 191.

8. Van Ryssen B, Van Bree P. Elbow arthroscopy. ECVS proceedings. June 1995.

9. Wind AP. Elbow dysplasia. In: Textbook of Small Animal Surgery. Slatter D (ed). Philadelphia: Saunders 1993; 1966-77.

10. Lewis DD. Parker RB, Hager DM. Fragmented coronoid process of the canine elbow. Comp Cont Educ 1989; 11: 703.

11. Olsson SE. The early diagnosis of fragment-ed coronoid process and osteochondritis dissecans of the canine elbow joint. J Am Anim Hosp Assoc 1983; 19: 616.

12. Henry WB. Radiographic diagnosis and surgical management of fragmented medial coronoid process in dogs. J Am Vet Med Assoc 1984; 184: 799.

13. Voorhout G. Hazewinkel AW. Radiographic evaluation of the canine elbow joint with special reference to the medial humeral condyle and the medial coronoid process. Vet Radiol 1987; 28: 158.

14. Miyabayashi T, Takiguchi M, Shrader SC, Biller DS. Radiographic anatomy of the medial coronoid process of dogs. J Am Anim Hosp Assoc 1995; 31: 125.

15. Fox SM, Roberts RE. Linear tomography in diagnosing fragmented coronoid processes in canine elbows. Comp Cont Educ 1987; 9: 60.

16. Lowry JE, Carpenter LG, Park RD. Radio-graphic anatomy and technique for arthrography of the cubital joint in clinically normal dogs. J Am Vet Med Assoc 1993; 203: 72.

17. Carpenter LG, Schwartz PD, Lowry JE. Comparison of radiologic imaging techniques for the diagnosis of fragmented medial coronoid process of the canine elbow. J Am Vet Med Assoc 1993; 203: 78.

18. Bradden TD, Stickle RL, Dejardin LM, Mostosky UV. The use of computed tomography in fragmented coronoid disease: A case report. VCOT 1994; 7: 40.

19. Anderson WD, Anderson BG. Atlas of Canine Anatomy. Philadelphia: Lea & Febiger 1994; 860-2.

20. Evans HE, Christensen GC, Miller. Anatomy of the dog. 2nd ed. Philadelphia: Saunders. 1979, 989.

21. Bardet JF, Bureau S. Traitement de l'apophyse coronoide fragmentée par arthrotomie et osteotomie ulnaire: etude retrospective de 103 cas. Prat Med Chir 1996; 31: 451-63.

22. Olsson SE. Osteochondrosis in the dog. In: Current Veterinary Therapy. VI. Kirk RW (ed). Philadelphia: Saunders 1977; 880-6.

23, Gröndalen J. Arthrosis with special reference to the elbow joint of rapidly growing dogs. II. Occurrence, clinical and radiographic findings. Nord Vet Med 1979; 31: 69-75.

24. Gröndalen J. Arthrosis in the elbow joint of young rapidly growing dogs. II. Ununited medial coronoid process of the ulna and osteochondritis dissecans of the humeral condyle. Surgical procedure for correction and postoperative investigation. Nord Vet Med 1979; 31: 520-7.

25. Gröndalen J, Gröndalen T. Arthrosis in the elbow joint of young rapidly growing dogs. V. A pathoanatomical investigation. Nord Vet Med 1981; 33: 1-16.

26. Denny HR, Gibbs C. The surgical treatment of osteochondritis dissecans and ununited coronoid process in the canine elbow joint. J Small Anim Pract 1980; 21(6): 323-31.

27. Tirgari M. Clinical, radiographical and pathological aspects of ununited medial coronoid process of the elbow joint in dogs. J Small Anim Pract 1980; 21(11): 595-608.

28. Berzon JL, Quick CB. Fragmented coronoid process: Anatomical, clinical and radiographic considerations with case analysis. JAM-IA. 1980; 16(2): 241-51.

29. Bennett D, Duff SRI, Kene RO et al. Osteochondritis dissecans and fragmentation of the coronoid process in the elbow joint of the dog. Vet Rec 1981; 109(5): 329-36.

30. Wind AP. Incidence and radiographic appearance of fragmented coronoid process. California Vet 1982; 6(1): 19-25.

31. Wind AP, Packard ME. Elbow incongruity and developmental diseases in the dog: Part I. JAAHA. 1986; 22(6): 711-24.

32. Wind AP, Packard ME. Elbow incongruity and developmental diseases in the dog. Part II. JAAHA, 1986; 22(6): 725-30.

33. Fox SM, Bloomberg MS, Bright RM. Developmental anomalies of the canine elbow. JAAHA. 1983; 19(5): 605-15.

34. Goring RL, Bloomberg MS. Selected developmental abnormalities of the canine elbow: Radiographic evaluation and surgical management. Comp Con Educ Pract Vet 1983; 5(3): 178-88.

35. Järvinen M, Jozsa L. Effect of anterior cruciate reconstruction with patellar tendon or prosthetic ligament on the morphology of other ligaments of the knee joint. Clin Orthop Rd l Res 1995; 311: 176.

J. F. Bardet, Dr. vet., MSc, Dipl. E.C.V.S. Referral Surgical Practice - 32, rue Pierret 92200 Neuilly sur Seine France