Keywords osteochondritis dissecans - shoulder - dog - arthrotomy - minimally invasive approach
Introduction
Osteochondritis dissecans (OCD) of the canine shoulder is a common cause of forelimb
lameness in dogs and constitutes a disturbance of endochondral ossification resulting
in focal thickening of the joint cartilage with subsequent dissection of a flap of
this thickened cartilage away from the underlying subchondral bone.[1 ]
[2 ] The cartilaginous flap is most frequently located in the caudocentral or caudomedial
region of the humeral head.[2 ] The exact aetiology of OCD is unknown, but there is evidence that some factors,
such as heredity, rapid growth, trauma and poor diet, are essential for its occurrence.[3 ] The disease occurs mainly in large and giant breed male dogs, with 27 to 68% of
patients having radiographic lesions bilaterally.[3 ]
[4 ]
[5 ] The onset of clinical signs (lameness and pain) generally occurs between 4 and 10
months of age.[3 ]
[4 ]
[5 ] Surgical treatment options include open arthrotomy or arthroscopic joint examination,
with the subsequent removal of the elevated cartilage flap and the curettage of the
underlying subchondral bone bed to stimulate fibrocartilage formation as well as infill
the joint surface defect. The post-surgical recovery period is approximately 2 months
compared with approximately 7 months following conservative treatment.[6 ] The prognosis following surgical recovery is good, with 75 to 91% of patients showing
no lameness at a mean interval of 3 years after surgery.[7 ]
[8 ] Arthroscopic surgery is a minimally invasive approach that gives good overall joint
visibility, allows intra-articular palpation, manipulation and resection of joint
structures and when performed by experienced surgeons can result in low morbidity
and good recovery.[7 ]
[9 ]
[10 ]
[11 ] The challenges inherent with its use are the need to invest in expensive instrumentation,
the intensive training required to perform the procedure and the technical difficulty
in removing large detached cartilage flaps from the caudolateral instrument portal.
Open shoulder arthrotomy using various modifications of three standard approaches
(craniolateral, caudolateral and caudal) provides adequate exposure of the joint surfaces
without damaging the cartilage or compromising joint function, and results in minimal
postoperative complications.[12 ]
[13 ]
[14 ]
[15 ] Generally, arthrotomy techniques requiring myotomy, tenotomy or osteotomy are considered
to be more traumatic and result in more severe postoperative pain and lameness than
those requiring only muscle separation.[15 ] In particular, a caudolateral approach with craniodorsal retraction of the teres
minor muscle (no tenotomy) has been shown to result in increased joint extension and
range-of-motion compared with a craniolateral approach with tenotomy of the infraspinatus
muscle, but there was less exposure to the articular surface compared with a craniolateral
approach.[13 ]
[16 ] In the early 1970s, Punzet was the first to describe an articular access between
the distal segment of the supraspinatus muscle and the infraspinatus tendon.[17 ] A decade later, Cheli and colleagues proposed a modification of this original technique
by positioning the shoulder in forced hyperflexion[18 ] ([Fig. 1 ]). This resulted in caudal displacement of the acromial deltoid and infraspinatus
muscles, which improved exposure of the joint surfaces without the need for myotomy
or tenotomy ([Fig. 2 ]). The original Cheli craniolateral approach was further modified by Vezzoni in 1986
to create a limited open approach for the debridement of OCD lesions of the caudal
humeral head without humeral head luxation as proposed by Cheli.[19 ]
Fig. 1 (A ) Neutral radiographic view of the shoulder of an English Setter with osteochondrosis
dissecans (OCD) showing the lesion on the caudal surface of the humeral head (white
arrows). (B ) Forced hyperflexion of the shoulder joint, as proposed by Cheli in 1984, moves the
caudal surface of the humeral head cranially. The yellow stars indicate the landmarks
for the direction of the skin incision, approximately midway between the cranial edge
of the acromion process and the proximal greater humeral tubercle. The red line indicates
the joint approach for direct access to the OCD lesion.
Fig. 2 Anatomical specimen showing how forced hyperflexion of the shoulder results in caudal
displacement of the acromial head of the deltoid and infraspinatus muscles, providing
exposure of the joint capsule without the need for myotomy or tenotomy.
The aim of this study was to describe a modification of the Cheli approach for minimally
invasive treatment of shoulder OCD in dogs. A retrospective review of clinical cases
treated with this novel surgical technique is presented in this study.
Materials and Methods
Case Selection
Medical records of dogs that underwent treatment of OCD using this modified Cheli
approach from May 2001 to May 2019 at the Clinica Veterinaria Vezzoni, Cremona, Italy,
were retrospectively searched for clinical results and complications. All surgeries
were performed by one of two surgeons (A.V., L.V.). For inclusion in the retrospective
study, a complete patient medical record with preoperative radiographs of both shoulder
joints and a minimum clinical follow-up period of 2 months was required for all dogs.
Data retrieved from the medical record included breed, gender, age at diagnosis, body
weight, limb affected, unilateral or bilateral lesions, lameness score at diagnosis,
surgery duration, intraoperative observations and intraoperative and postoperative
complications. Postoperative complications were defined as any unexpected event that
occurred after surgery. Clinical follow-up observations at various time periods greater
than 2 months were required for each dog.
Anaesthesia
All dogs were premedicated with morphine (0.15 mg/kg intramuscular [IM]) or methadone
(0.1–0.2 mg/kg IM) and acepromazine (0.01 mg/kg IM). General anaesthesia was induced
using propofol (3–6 mg/kg) administered via a peripheral venous catheter and maintained
using isoflurane (1–1.5% in 80% oxygen) after endotracheal intubation. Intraoperative
analgesia was provided by a constant rate infusion of fentanyl (10 µg/kg/hr). Cefazolin
sodium (22 mg/kg IV) was only administered at the time of anaesthetic induction. Postoperative
analgesia consisted of a single intra-articular injection of lidocaine 1% (1 mg/kg)
and morphine (0.1 mg/kg) after closure of the joint capsule.
Surgical Technique
Standard surgical instrumentation was augmented with pointed Mayo forceps, two small
Gelpi retractors with overlapping deep arms, a 4 mm blunt Langenbeck periosteal elevator,
a 1.5 mm blunt meniscal probe, a 3 mm sharp periosteal elevator, 2 mm curette and
a microfracture awl. Dogs were positioned in lateral recumbency with the surgical
limb uppermost. The shoulder joint was held in maximum hyperflexion and the elbow
joint in hyperextension using a surgical positioning device secured to the surgical
table and the distal aspect of the limb was positioned close to the abdomen ([Fig. 3 ]). The shoulder area was aseptically prepared. The acromion process and the greater
tubercle were used as anatomical landmarks. A longitudinal 1 to 1.5 cm skin incision
was made on the craniolateral aspect of the shoulder, approximately midway between
the cranial edge of the acromion process and the proximal greater tubercle ([Fig. 4A ]). The cranial margin of the acromial head of the deltoid muscle was identified below
the subcutaneous tissue layer, and the intermuscular septum between the caudal margin
of the supraspinatus muscle belly and the acromial deltoid muscle was identified and
bluntly separated to expose the joint capsule beneath ([Fig. 4B ]). Two small self-retaining Gelpi retractors were positioned perpendicularly to each
other to keep the muscles separated and gain adequate joint capsule exposure ([Fig. 4C ]). The fibrous joint capsule was incised in a proximal to distal plane to mimic the
intermuscular septum dissection. The arthrotomy was adjacent and cranial to the lateral
collateral ligament. The Gelpi retractors were then placed in the joint to retract
the joint capsule arthrotomy and expose the articular surface of the humeral head
([Fig. 4D ]). The humeral head was then inspected by gentle probing with a meniscal probe to
identify the osteochondral flap, which was exposed by shoulder hyperflexion. The flap
was carefully elevated from its peripheral attachment using a sharp periosteal elevator
and removed with grasping forceps ([Fig. 5 ]). Cartilage remnants were removed from the margins of the lesion while pushing the
humeral head distally with the blunt Langenbeck elevator ([Fig. 6 ]). The meniscal probe was then used to evaluate the margins of the OCD lesion to
determine the completeness of flap removal. Microfracture was performed when sclerotic
bone was found underneath the flap to promote vascular ingrowth and subsequent fibrocartilage
formation by migrating mesenchymal cells.[20 ]
Fig. 3 Position of the dog in lateral recumbency with the surgical limb uppermost. The shoulder
joint is held in maximum hyperflexion and the elbow joint in hyperextension using
a surgical positioning device secured to the surgical table, and the distal aspect
of the limb is positioned close to the abdomen,
Fig. 4 (A ) A longitudinal 1 to 1.5 cm skin incision is made on the craniolateral aspect of
the shoulder, midway between the cranial edge of the acromion process and the proximal
greater tubercle (stars). (B ) The cranial margin of the acromial head of the deltoid muscle is identified below
the subcutaneous tissue layer, and the intermuscular septum between the caudal margin
of the supraspinatus muscle belly and the acromial deltoid muscle is identified and
bluntly separated to expose the joint capsule beneath. (C ) Two small self-retaining Gelpi retractors are positioned perpendicularly to each
other to keep the muscles separated and gain adequate joint capsule exposure which
is incised in a proximal to distal plane to mimic the intermuscular septum dissection.
(D ) The Gelpi retractors are then placed in the joint and the joint capsule arthrotomy
is divaricated to expose the articular surface of the humeral head.
Fig. 5 The flap is carefully elevated from its peripheral attachment using a sharp periosteal
elevator (A ) and removed (B ). The margins of the osteochondritis dissecans lesion are inspected to determine
the completeness of flap removal (C ). Microfracture is performed when sclerotic bone is found underneath the flap (D ) to promote vascular ingrowth and subsequent fibrocartilage formation by migrating
mesenchymal cells (E ).
Fig. 6 When significant fibrocartilage tissue is identified on the osteochondritis dissecans
lesion bed after flap removal, curettage or microfracture is not done (A ). Cartilage remnants are removed from the margins of the lesion while pushing the
humeral head distally with a blunt Langenbeck elevator (B ).
Fig. 7 Intraoperative findings.
When significant fibrocartilage tissue was identified on the OCD lesion bed after
flap removal, saline lavage of the area was done rather than curettage or microfracture.
In patients with detached OCD flaps that were not visible, joint lavage under pressure
was performed and the tip of a 19 G venous catheter was directed medially, caudally
and cranially to mobilize and remove the detached flap. To facilitate effective flushing
of the joint at the end of the procedure, the limb was positioned in a neutral standing
position, and gentle distal traction was used to distract the joint surfaces from
each other.
The joint capsule was closed with one or two interrupted sutures using absorbable
suture material with a small diameter needle (Vicryl EP 1.5 mounted on a 10 mm 1/3
circle needle). The intermuscular septum, subcutaneous tissues and skin were closed
routinely.
Postoperative Care
Dogs were discharged the same day and received amoxicillin/clavulanic acid (12.5 mg/kg
per os) for 5 days and meloxicam (0.1 mg/kg per os) once daily for 5 days postoperatively
and then every other day for another 7 days. Restricted exercise in the form of short
controlled leash walks was advised for the first month postoperatively. The duration
and intensity of the walks were gradually increased in the second month with return
to normal activity 3 months postoperatively.
In-Clinic Follow-Up Examination
Clinical and radiographic postoperative re-examinations were done 2 to 5 months (short-term)
and in some dogs 6 to 12 months (medium-term) and > 12 months (long-term) postoperatively.[21 ] At each follow-up examination, the gait and presence and degree of lameness were
assessed via video recordings in all dogs. The presence of complications was also
noted.
Data Analysis
Descriptive statistical analysis (median, range, percentage) was used to report data.[21 ]
Results
A total of 164 shoulders in 141 dogs (23 bilateral), consisting of 103 males and 38
females, were examined and treated with our modified technique for the treatment of
OCD of the humeral head. Breeds included Border Collie (n = 23), Italian Cane Corso (n = 13), Golden Retriever (n = 10), mixed-breed (n = 9) Bernese Mountain dog (n = 8), Great Swiss Mountain dog (n = 8), Boxer (n = 7), English Setter (n = 6), German Shorthaired Pointer (n = 6), Rottweiler (n = 6) Belgian Shepherd dog (n = 5), German Shepherd dog (n = 5), Saint Bernard dog (n = 5), Great Dane (n = 4), Labrador Retriever (n = 4), Dogue de Bordeaux (n = 3), Newfoundland dog (n = 3), Australian Shepherd dog (n = 2), Pyrenean Mountain dog (n = 2) and one each of the following 12 different breeds: Beauceron, Bouvier des Flandres,
Briard, Central Asia Shepherd dog, Dalmatian, Dobermann, Lagotto, Leonberger, Pit
Bull, Rhodesian, Shar-pei and Vizsla. Median body weight was 23.4 kg (range, 14–65 kg)
and median age was 9.5 months (range, 6–27 months). Bilateral OCD lesions were seen
radiographically in 23 (16%) dogs; in 11 of these, both shoulders were operated in
the same surgical session, and in 12, the operations were staged 3 to 5 weeks apart.
In the remaining 118 dogs (84%), the modified arthrotomy approach was performed on
65 left shoulders and on 53 right shoulders.
Intraoperative Data
In all cases, our novel modified craniolateral approach allowed visual inspection
and adequate exposure of the caudal surface of the humeral head. Flaps of cartilage
still attached to the humeral head were detected in 148 of 164 shoulders (90%) and
removed ([Fig. 7 ]). Microfracture was performed in 116 of 164 (71%) shoulders, and reparative fibrocartilage
was present in 42/164 (27%). Detached cartilage flaps occurred in 4 of the 164 (2.4%)
shoulders and were removed with the assistance of saline joint lavage ([Fig. 7 ]).
In 10 patients (6%), cartilage flaps were not present on the visible subchondral OCD
bone bed and detached flaps could not be collected via joint lavage.
In another two cases (1.2%), a definitive cartilage flap was not detected at the site
of the radiographic lesion. In one of those, an 8-month-old Dogue de Bordeaux, the
cartilage flap was found and removed 4 months later. In the second case, a 7-month-old
Dogue de Bordeaux, no progression of osteochondrosis to OCD was seen. Intraoperative
complications consisting of moderate intraoperative bleeding occurred in three cases
(1.8%) and were readily stopped via electrocautery. Median surgery duration was 32 minutes
(range, 20–50 minutes).
Postoperative Outcome
The results of radiographic and clinical follow-up examinations were available in
123 of 164 (75%) dogs. Early postoperative complications included two cases (1.2%)
of mild seroma formation. At the first follow-up examination 2 to 3 months after surgery,
eight dogs (6.5%) had persistent mild lameness (grade 2 of 4). In five of these, clinical
signs disappeared within the following 3 months and the remaining two cases underwent
a second surgery. The first was a Border Collie, which had been treated for bilateral
OCD at 9 months of age and had persistent lameness in one limb. At the time of the
first surgery, both osteochondral flaps were found and removed and the lesion beds
were microfractured. Arthroscopy done 7 months postoperatively showed subtotal biceps
rupture, which was treated with tendon release, resulting in complete resolution of
lameness at the 2-month-follow-up examination. The second case was a Dogue de Bordeaux
in which an elevated flap had not been identified in the initial surgery. In this
case, a second surgery performed 2 months later allowed identification and removal
of the cartilage flap with subsequent microfracture. The results of follow-up examinations
were not available for this dog. The results of medium-term follow-up, that is, 6
to 12 months after surgery, were available in 70 cases (57%): one dog showed persistent
mild lameness and another had intermittent lameness after exercise. The results of
long-term follow-up examination, that is, > 1 year after surgery, were available in
16 cases (9.7%) and none were lame on gait evaluation.
Discussion
In this retrospective study, our minimally invasive craniolateral surgical approach
to the shoulder joint allowed adequate exposure of the humeral head including the
caudal aspect in dogs with OCD. Accessing detached flaps in the caudal portion of
the joint was difficult at first, and in two cases the flaps could not be removed.
A more reliable method was subsequently developed to mobilize and collect lodged detached
flaps by repositioning the shoulder into a neutral position and applying gentle distal
traction while flushing the caudal joint space. Hyperflexion of the shoulder joint
leads to collapse of the caudal portion of the joint and prevents lavage fluid from
mobilizing free flaps. In addition, cartilaginous flaps firmly attached to the synovial
membrane in the caudal joint recess cannot be mobilized using joint lavage. When free
flaps are seen in the caudal joint recess on preoperative radiographs, an arthroscopic
approach should be considered. In the two dogs with intraoperative bleeding, the surgical
approach was more proximal than desired and bleeding was secondary to omobrachial
vein injury. Thus, careful planning of the location of the initial incision is required
to avoid this complication. Seroma formation is the most common postoperative complication
after shoulder arthrotomy. In the present study, only two cases had mild, self-limiting,
seroma formation in the postoperative period. The low number was likely attributable
to the nature of the procedure, which minimized periarticular iatrogenic trauma and
the creation of dead space. In two patients, one with lameness caused by subtotal
biceps rupture and a second with progression of osteochondrosis to OCD, the need for
further treatment was not attributable to the surgical technique. Of the 123 cases
in which the results of follow-up examination were available, 72 (97.3%) had complete
resolution of lameness. The results of range-of-motion and palpation of the shoulder
joint and data on the ability to exercise in a normal, sustained, high-intensity fashion
were not included in this series. The long-term clinical outcome of shoulder OCD surgery
is likely related to many other factors including the location, diameter and depth
of the lesion and the presence and severity of concurrent osteoarthritis.[2 ]
[4 ]
[6 ]
[8 ]
[10 ]
[22 ]
Our surgical approach can be considered a moderately challenging, minimally invasive,
efficient and inexpensive alternative to other surgical approaches including arthroscopy
for treatment of OCD of the humeral head. The surgical instrumentation required is
commonly available and inexpensive in relation to arthroscopic equipment. The use
of Gelpi retractors was adequate for retraction of the muscles and joint capsule.
The short time required for surgery was favourable in terms of anaesthesia duration
and complications related to more prolonged surgery.
In conclusion, this technique could be considered a reliable approach for surgical
treatment of shoulder OCD even though the joint cannot be completely evaluated for
other diseases. The majority of dogs with OCD could be successfully treated. In dogs
with very large or deep lesions, the Cheli approach can be used to increase exposure
of the humeral head and allow autogenic or allogenic osteochondral grafting or placement
of a synthetic resurfacing implant. For the latter procedures, the surgical approach
is expanded; wider incision of the lateral joint capsule is performed with care not
to damage the lateral collateral ligament unduly. Fine-tuning of the shoulder flexion
and humeral rotation along with careful intra-articular placement of a small Hohmann
retractor is helpful in increasing humeral head access.
