Accurate and reproducible measurement of three-dimensional shoulder kinematics would contribute to better understanding shoulder mechanics, and therefore to better diagnosing and treating shoulder pathologies. Current techniques of 3D kinematics analysis use external markers (acromial cluster or scapula locator) or medical imaging (MRI or CT-Scan). However those methods present some drawbacks such as skin movements for external markers or cost and irradiation for imaging techniques. The EOS low dose biplanar X-Rays system can be used to track the scapula, humerus and thorax for different arm elevation positions. The aim of this study is to propose a novel method to study scapulo-thoracic kinematics from biplanar X-rays and to assess its reliability during abduction in the scapular plane.

This study is based on the EOS™ system (EOS Imaging, Paris, France), which allows acquisition of 2 calibrated, low dose, orthogonal radiographs with the subject standing at 30 to 40° angle of coronal rotation to the plane of one of the X-ray beams, in order to limit superimposition with the ribcage and spine. Seven abduction positions in the scapular plane were maintained by the subjects for 10 seconds, during X-ray acquisition. Between two positions, the subjects returned at rest position. Arm elevations were approximately 0, 10, 20, 30, 60, 90 and 150° (position 1 to 7). Six subjects were enrolled to perform a reproducibility study based on the 3D reconstructions of 2 experienced observers three times each. For each subject, a personalised 3D reconstruction of the scapula was created. The observer digitises clearly visible anatomical landmarks on both stereoradiographs for each arm position. These landmarks are used to make a first adjustment of a parameterised 3D model of the scapula. This provides a pre-personalised model of the subject's scapula which is then rigidly registered on each pair of X-rays until its retroprojection fits best on the contours that are visible on the X-rays. The thorax coordinate system (CS) was built following the ISB (International Society of Biomechanics) recommendations. The CS associated to the scapula was a glenoid centred CS based on the ellipse which fit on the glenoid rim on the 3D model of scapula. Scapular CS orientation and translation in the thorax CS was calculated following a Y,X,Z angle sequence for each position.

Each 3D reconstruction of the scapula was performed in approximately 30 minutes. The most reproducible rotation was upward/downward rotation (along X axis) with a 95% confidence interval (95% CI) from 2.71° to 3.61°. Internal/external rotation and anterior/posterior tilting were comprised respectively between 5.18° to 8.01° and 5.50° to 7.23° (CI 95%). The most reproducible translation was superior-inferior translation (along Y axis) with a 95% CI from 1.22mm to 2.46mm. Translation along X axis (antero-posterior) and Z axis (medio-lateral) were comprised respectively between 2.49mm to 4.26mm and 2.47mm to 3.30mm (CI 95%).

We presented a new technique for 3D functional quantitative analysis of the scapulo-thoracic joint. This technique can be used with confidence; uncertainty of the measures seems acceptable compared to the literature. Main advantages of this technique are the very low dose irradiation compared to the CT-Scan and the possibility to study arm elevation above 120°.

Purpose of the study: Kyphoplasty was introduced to reinforce the anterior column in osteoporotic vertebral fractures. It can be used for non-osteoporotic fractures. The purpose of this work was to report the clinical and radiographic results of kyphoplasty for non-ostoporotic vertebral fractures.

Material and methods: From December 2005 to August 2008, we followed prospectively 21 patients (12 M, 9 F) mean age 45 years (16–58) treated for thoracolumbar fractures by kyphoplasty in order to reinforce the anterior column. There were 23 fractures (T11 = 2, T12 = 5, L1 = 8, L2 = 4, L3 = 4) Magerl: A1 = 6, A3.1 = 7, A3.2 = 1, B1 = 2, B2 = 7. All patients were assessed preoperatively, postoperatively, and at last follow-up with a visual analogue scale (VAS) and the EIFEL function score. The sagittal CT scans passing through the pedicles and the midline were used to measure: the height of the anterior and posterior walls of the fractured vertebra and the supra and infra vertebrae as well as the kyphosis angle.

Results: Thirteen fractures were treated by kyphoplasty alone; seven by kyphoplasty combined with percutaneous osteosynthesis; three by kyphoplasty combined with open osteosynthesis with decompression because of preoperative neurological deficits. Mean follow-up was 13 months (6–28). There were no postoperative neurological or infectious complications. At last follow-up, the mean VAS was 1.25 (05) and the mean EIFEL 4 (0–12). Preoperatively, mean compression was 40.9% (6.2–81.4) for the anterior column and 16.7% (0–60.2) for the posterior column. Postoperatively the respective values were 22.8% (5.1–69.3) and 12.3% (−12 to 72.6) for a mean correction of 46.2% for the anterior column and 14.3% for the posterior column. At last follow-up, compression was respectively 26.1% and 7.9%. The vertebral kyphosis was 16.3 (6–16.3) preoperatively and 9.1 (3–4) postoperatively (mean correction 8.7). At last follow-up, vertebral kyphosis was 9.1 (1.7–28.3).

Discussion: Kyphoplasty allows satisfactory restoration of vertebral height without loosing short-term correction. For us, kyphoplasty should be associated with posterior fixation in patients with posterior injury. For neurological lesions, kyphoplasty associated with decompression and posterior fixation avoid the need for complementary anterior procedures.

Purpose of the study: Percutaneous screw fixation is widely used for the treatment of non-displaced fresh fractures of the carpal scaphoid. This screw fixation can be achieved either via a volar approach (retrograde insertion) or via a dorsal approach (antegrade insertion). The purpose of our study was to define the best approach as a function of the orientation of the fracture line (types B1 or B2 in the Herbert classification).

Material and methods: We used 12 upper limbs. For each wrist we obtained three scanner images: in maximal flexion, in the neutral position, and in maximal extension. For each scanner image, the parasagittal slice corresponding to the ideal plane for screw position was identified by digital reconstruction. On each slice, the type B1 and B2 fractures were modellised, as was the displacement of the corresponding screws introduced via the volar incision (S1) or the dorsal incision (S2). Each virtual screw was positioned as perpendicular as possible to the fracture line. For each slice corresponding to a given wrist position, we measured the angles between the fracture line (B1, B2) and the screws (S1, S2), giving four angles V1 (S1-B1), V2 (S1-B1), D1 (S2-B1), D2 (S2-B2). Thus the angle closest to 90° was considered the most satisfactory.

Results: For B2 fractures, the position of the virtual screw perpendicular to the fracture line was possible via both the volar and the dorsal incision. For B1 fractures, it was impossible to position the screw perpendicular to the fracture line, but the dorsal approach with the wrist in maximal flexion gave the best position.

Conclusion: For B2 fractures, the dorsal and volar approach allow optimal screw insertion so the choice of the incision depends on the surgeon’s experience. For B1 fractures, we recommend the dorsal approach.

Purpose of the study: Different posterior osteosynthesis techniques have been described for the treatment of unstable injury of the pelvic girdle. Bi-iliac fixation using threaded rods or plate-screw fixation has been proposed. The purpose of this work was to describe a modification of the posterior osteosynthesis using instrumentation designed for the spinal column.

Material and methods: From January 2006 to October 2008, four patients (three men, one woman, mean age 24 years, range 18–34) underwent surgery in our unit for unstable fractures of the pelvis with a trans-sacral posterior fracture line (AO classification C1.3–4). Two patients presented neurological signs including one by head trauma with hemiplegia. Two patients had an anterior fixation with an external fixator and another an anterior plate fixation. The operation was conducted via a posterior midline incision. After reduction of the fracture, the osteosynthesis was achieved with two poly-axial screws inserted in each of the iliac wings and connected by two rods and one or two cross connectors.

Results: Mean follow-up was 7.5 months (range 5–17). None of the patients developed infectious, neurological or mechanical complications postoperatively. Complete pain-free weight-bearing and walking were achieved in patients at three months. None of the patients had a horizontal or vertical misalignment callus measuring more than 5 mm. Screw analysis showed that three screws penetrated the sacroiliac joint in the first patient of the series with no clinical consequence.

Discussion: This posterior fixation technique for unstable fractures of the pelvis appears to be reliable and reproducible for type C fractures in combination with anterior fixation. It uses standard instrumentation for spinal osteosynthesis. A study with a larger population and longer follow-up is needed.

Purpose of the study: High-energy varus or valgus ankle trauma causes severe injury to the capsule and ligaments. We describe a presentation associating massive tears of the lateral/medial collateral ligaments with a transversal wound of the corresponding malleolus. This wound results from excessive tension on the skin cause by the major varus/valgus. We have defined this injury as an open and severe ankle sprain (OSAS).

Material and method: This was a retrospective analysis. We search the databases of three participating centres using the corresponding diagnostic and therapeutic codes from January 2005 to January 2009. The identified files were screened to select patients with OSAS.

Results: There were 11 cases of OSAS. Eight involved the lateral side of the ankle and three the medial side. Mean age was 41 years (range 21–45). All patients were victims of a high-energy trauma (five motorcycle accidents) and four patients had fallen from a high point. Associated injuries were tendon section (n=3), section of the deep fibular nerve (n=2), and section of the anterior tibial artery (n=1). Pneumarthrosis was the only visible anomaly on the plain x-rays of seven ankles. Diagnosis was confirmed preoperatively in all cases clinically with varus-valgus stress manoeuvres.

Conclusion: OSAS is a rare misleading injury. Confusion with a common wound is possible. The risk is to miss acute instability and thus its treatment. The diagnosis should be proposed for all transversal wounds without contusion over the malleolus with normal x-rays.

Disruption of the interosseous membrane is easily missed in patients with Essex-Lopresti syndrome. None of the imaging techniques available for diagnosing disruption of the interosseous membrane are completely dependable.

We undertook an investigation to identify whether a simple intra-operative test could be used to diagnose disruption of the interosseous membrane during surgery for fracture of the radial head and to see if the test was reproducible.

We studied 20 cadaveric forearms after excision of the radial head, ten with and ten without disruption of the interosseous membrane. On each forearm, we performed the radius joystick test: moderate lateral traction was applied to the radial neck with the forearm in maximal pronation, to look for lateral displacement of the proximal radius indicating that the interosseous membrane had been disrupted. Each of six surgeons (three junior and three senior) performed the test on two consecutive days.

Intra-observer agreement was 77% (95% confidence interval (CI) 67 to 85) and interobserver agreement was 97% (95% CI 92 to 100). Sensitivity was 100% (95% CI 97 to 100), specificity 88% (95% CI 81 to 93), positive predictive value 90% (95% CI 83 to 94), and negative predictive value 100%).

This cadaveric study suggests that the radius joystick test may be useful for detecting disruption of the interosseous membrane in patients undergoing open surgery for fracture of the radial head and is reproducible. A confirmatory study in vivo is now required.

Introduction: The incidence of contralateral, second hip fractures after a first hip fracture is as high as 20% in the elderly. Femoroplasty using an injectable and resorbable bi-phosphonate loaded bone substitute to prevent controlateral hip fracture may represent a promising preventive therapy. We aimed to evaluate the biomechanical consequences of the femoroplasty using this bone substitute.

Materials and Methods: Twelve paired human cadaveric femora from donors with a mean age of 86 years (7 women and 6 men) were randomly assigned for femoroplasty and biomechanically tested for fracture load against their native contralateral control. Anterior–posterior and lateral radiographs and DXAscan’s were made before injection. Femoroplasty were performed under fluoroscopic guidance with an injectable and resorbable bi-phosphonate loaded bone substitute. All femurs were fractured by simulating a fall on the greater trochanter by an independent observer.

Results: Mean T-score of the tested femur were −3. Bone density was comparable for each pair of femur. All the observed fractures were Kyle II throchanteric fractures. Mean fracture load was 2786 Newton in the femoroplasty group (group F) versus 2116 Newton in the control group (group C) (p< 0.001). Fracture loads were always higher in the group F: mean 41.6% (mini: 1.2%/maxi:102.1%). Effect of femoroplasty was significantly superior for women and also correlated to initial bone density (p< 0.0001).

Discussion:According to our results, femoroplasty with an injectable and resorbable bi-phosphonate loaded bone substitute can provide significant biomechanical reinforcement of the proximal femur to prevent controlateral fracture.

Purpose: To determine whether the tendon harvester can influence harvested tendon characteristics and soft tissue disruption

Summary: We compared two harvesters with regards to the length of tendon obtained and soft tissue disruption during hamstring tendon harvesting. Thirty six semitendinosus and gracilis tendons were harvested using either a closed stripper or a blade harvester in 18 paired knees from 9 human fresh cadavers. Use of the blade harvester gave longer lengths of usable tendon (p=0.002), whilst minimising the stripping of muscle (p=0.013).

Conclusion: Our results suggest that the type of harvester per se can influence the length of tendon harvested as well as soft tissue disruption. Requesting such data from the industry prior to deciding which harvester to use seems desirable.

Total shoulder arthroplasty is a well-established and widely accepted method of treatment for a variety of shoulder disorders, loosening of the glenoid prosthesis is the main complication in total shoulder arthroplasty, it is highly dependent on the quality of the glenoid cancellous bone. Very little is known about mechanical properties of this cancellous bone. The objectives of this study were to determine the mechanical properties (elastic modulus and strength) of glenoid cancellous bone in the axial, coronal and sagittal planes including regional variation using a uniaxial compression test. To our knowledge, this kind of study was not done before.

Eleven scapulas were obtained from six fresh-frozen, unembalmed human cadavers (mean age eighty-eight years). Eighty-two cubic cancellous bone specimens of 6×6×6mm3 were used for mechanical testing in the three planes. The test was a uniaxial compression along each direction, Elastic modulus and strength were determined from the stress-strain curve. Apparent density was also calculated.

The study showed significant differences in the mechanical properties with anatomic location and directions of loading. Young modulus and strength were found to be significantly higher at the posterior part of the glenoid with the weakest properties at the antero-inferior part. Cancellous bone was found to be anisotropic with higher mechanical properties in the latero-medial direction (perpendicular to the articular surface of the glenoid). The apparent density was on average equal to 0.29 g/cm3 with the higher values at the posterior and superior part of the glenoid. Good correlation between apparent density and elastic modulus was found only in the sagittal plane but not in the coronal and axial plane, the overall correlation was low (r2 = 0.22, p< 0.0001) which emphasizes the role of trabecular bone architecture in predicting mechanical properties.

The mechanical properties determined in this study provide input data for finite element method analyses and may help to assist in uncemented shoulder prosthesis design.

The lateral ligament complex is the primary constraint to posterolateral rotatory laxity of the elbow, and if it is disrupted during surgery, posterolateral instability may ensue. The Wrightington approach to the head of the radius involves osteotomising the ulnar insertion of this ligament, rather than incising through it as in the classic posterolateral (Kocher) approach. In this biomechanical study of 17 human cadaver elbows, we demonstrate that the surgical approach to the head can influence posterolateral laxity, with the Wrightington approach producing less posterolateral rotatory laxity than the posterolateral approach.

Purpose: There is no consensus concerning the ideal incongruency of the prosthetic head and the glenoid implant in total shoulder arthroplasty. Certain recent publications suggest the rate of periglenoid lucency is lower if the incongruency is greater than 5.5 mm. The purpose of this experimental in vitro work was to study the influence of changing humeral head-glenoid congruency on periglenoid bony malformations of prosthesis-bearing cadaveric scapulae and on the motion of the glenoid implants.

Material and methods: Five scapulae from subjects aged 76 to 91 years at death were harvested and implanted with five stem cemented glenoid implants with an identical curvature. Five metallic balls with different radii were used to simulate incongruency of the humeral head-glenoid implant varying from 0 (perfect congruency) to 6 mm (0.2, 4.5, and 6 mm). The protocol involved preloading at 400 N following a normal axis for the glenoid implant and then posteroanterior translation and inferosuperior translation of 2.5 mm. The force necessary to impose the translation displacement, periglenoid bony deformations, and implant displacement compared with the bony glenoid were measured with a traction-compression device using deformation gauges and two CCD cameras in compliance with a published protocol.

Results: Increasing incongruency decreased the force necessary to displace the metallic balls, decreased periglenoid bony deformations around the loaded zones and decreased the degree of prosthetic displacement facing the loaded zone.

Discussion: The limitations of this experimentation are the small number of implants tested and the subsequent lack of statistical analysis concerning the reality of the differences observed. Besides, the experimental protocol cannot reproduce the normal conditions of the prosthesis articulation. Nevertheless, these results appear to favour the idea of greater bone and prosthetic tolerance with lesser humeral head-glenoid implant congruency. This might provide an explanation for the fewer glenoid lucent lines found in vivo in similar congruency situations.

Conclusion: These results suggest that a certain degree of incongruency of total shoulder prostheses could reduce the risk of periprosthetic lucency. Ideal incongruency remains to be determined with further in vitro and in vivo studies.

Purpose: Trapezometacarpal osteoarthritis is frequent in women aged over 50 years. Surgical cure may be needed after failure of well conducted medical treatment. When the trapezeal bone stock is insufficient for implantation of a total prosthesis, total trapezectomy can relieve the pain. This procedure is generally associated with stabilization ligamentoplasty of the first ray. The purpose of this study was to demonstrate the feasibility of this procedure using an arthroscopic approach and to detail the technique and its limitations.

Material and methods: This study was conducted on twelve cadaver specimens from eleven women and one man, mean age 85 years. Radiographs were obtained to confirm the trapezometacarpal osteoarthritis. Standard arthroscopic material used for the wrist was employed (2.4 mm optic, mini-shaver). Two portals on either side of the abductor pollicis lungus tendon were used to approach the trapezometacarpal joint. Total trapezectomy was performed with the mini-shaver distal to proximal. A tendon band measuring 6 to 7 cm was fashioned from the abductor pollicis longus tendon via a proximal contraincision. This band inserted on the first metacarpal was passed through two bone tunnels bored in the base of the first and second metacarpals then fixed to the base of the second metacarpal. Operative time was noted. The quality of the bone resection was determined on postoperative radiographs and open inspection.

Results: Arthroscopic total trapezectomy with stabilisation ligamentoplasty was achieved in all cases and evaluated radiographically and at open inspection. No lesions to noble elements were observed.

Discussion: This minimally invasive technique for trapezectomy associated with stabilisation ligamentoplasty was found to be feasible but did require a certain degree of learning. We were unable to identify any procedure-related morbidity, particularly concerning the sensorial branch of the radial nerve to the thumb.

Conclusion: The results of this preliminary study are encouraging and suggest a clinical trial should be conducted to prove the advantages of this technique in terms of morbidity and socioeconomical cost.

Purpose: The Bankart procedure is widely studied in the literature. The general lack of postoperative complications is well recognised. The purpose of this work was to study patient comfort after Bankart procedures performed in the outpatient setting in order to validate the feasibility of this approach.

Material and methods: Thirty patients underwent Bankart procedure from June 2001 to 2002 performed by the same surgeon in an outpatient clinic. There were 28 men and two women, mean age 28 years. Pain was assessed with a visual analogue scale (VAS) at entry into the recovery room (P0), when leaving the recovery room (P1), on day 1 by telephone (P2), and on day 7 at consultation (P3). On day 1 and day 7, the patient was also asked if he/she preferred staying in hospital one night. The general anaesthesia protocol was the same for all patients. Intraopeartive analgesia was 20 mg nefopam (Acupan(r)) in a half-hour infusion, 2g propacetamol (Prodafalgan(r)) or paracetamol (Perfalgan(r)), and 100 mg ketoprofen (Profenid(r)) if there were no contraindications. In the recovery room, 3 mg morphine was delivered in by iv bolus until the VAS was less than 4/10 followed by oral paracetamol-codeine combination. Home treatment used 200 mg/d ketoprofen and paracetamol+codeine.

Results: The only complication was one superficial venous thrombosis of the upper limb diagnosed on day 15. There were no cases of postoperative haematoma or infection. One patient stayed one night in hospital after the procedure due to a vagal malaise which occurred at discharge; the VAS pain score was the same in this patient as in the others. Pain assessment was: D1=2 (5. 0); D2=1 (3, 0). Postoperative comfort was thus considered satisfactory. None of the patients would have preferred 24h hospitalisation.

Discussion: There has been only one series of 25 patients reporting results of patient comfort and cost of outpatient Bankart procedure. Patients underwent surgery with a scalene interblock. Three of the 25 patients preferred a 24h hospitalisation because of pain, perhaps due to the rebound pain effect after the block. Absence of drainage did not lead to any case of haematoma, confirming an earlier unpublished study of 50 consecutive patients who underwent classical hospital procedure without drainage. Only one patient had a subcutaneous haematoma that resolve favourably spontaneously. These results suggest that satisfactory patient comfort can be achieved postoperatively for outpatient procedures. We have decided to pursue this approach.

Aim: To study the contribution of humeral avulsion of the glenohumeral ligaments (HAGL) to shoulder instability. Methods: In fourteen fresh cadaver shoulders a selective cutting sequence was performed. After each section an abduction-external rotation manoeuvre with axial compression and translation was carried out to provoke dislocation. The resulting instability was graded on a scale of five, ranging from no translation to a locked dislocation. Results: Cutting of only the inferior glenohumeral ligament complex resulted at the most in increased translation, but not in subluxation. For subluxation to occur, at least the middle glenohumeral ligament needed to be cut. The entire humeral capsuloligamentous complex needed to be sectioned before subluxation or dislocation occurred. In half of the cases an additional lesion of the subscapularis or the latissimus dorsi is necessary to allow a locked antero-inferior dislocation. Conclusion: Extensive damage to the humeral side of the capsulo-ligamentous complex and, frequently, associated lesions of the subscapularis or latissimus dorsi muscles are necessary to allow dislocation. This might be the primary reason for the low incidence of HAGL observed in clinical series of shoulder instability

Aim: To study the morphology of the anterior cap-suloligamentous structures of the glenohumeral joint. Methods: Eighty non-embalmed cadaver shoulders were studied. Twenty shoulders were dissected through an anterior approach, twenty through a posterior approach. In another twenty shoulders the anteroinferior capsuloligamentous complex was examined arthroscopically through a posterior portal. In all of these sixty shoulders the functional anatomy was studied by moving the arm from its resting position along the body to maximal abduction and external rotation. Dissecting another twenty shoulders through an inferior approach completed the study of the humeral insertion of the inferior glenohumeral ligament. Results: The inferior, middle and superior glenohumeral ligament are usually only discernible by palpation, but not visually. When the capsule is ßattened out, these ligaments can no longer be discriminated macroscopically. The classic Z-like structure can be seen when examining the anterior capsule from its posterior side, but only when the shoulder is at rest, which is with the arm along the body. The functional study shows that this Z corresponds with a folding phenomenon of the capsuloligamentous ÒpouchÒ to accommodate the relative excess of length when the arm is at rest. A progressive unfolding occurs as the arm is progressively abducted and externally rotated. By creating a functional shortening, the folding mechanism provides pretensioning of the ligaments. Conclusion: At the anteroinferior part of the shoulder joint, there is a real, functional capsuloligamentous unit.

Aim: To determine whether isolated lesions of the glenoid labrum or capsulolabral lesios influence anterior and inferior shoulder stability in a cadaver model that leaves all other glenohumeral structures, including the capsule, intact. Methods: Sequential arthroscopic resection of the labrum was performed with a motorized shaver in seventeen fresh cadaver shoulders. A capsulolabral dehiscence was created arthroscopically in another eleven fresh cadaver shoulders. The capsulolabral complex was divided into five zones: from superior to posterior. Inferior and anterior stability were tested before and after each cutting step. Results: The hyper-abduction test showed progressive increase in abduction with each step, with a maximum of 120° reached after removal of the labrum in all four zones or after detachment of all zones. Resection of the superior to inferior labrum maximally resulted in subluxation, but not in dislocation of the shoulder. Capsulolabral dehiscence of all but the posterior zones resulted in a locked (6/11) or a metastable (4/11) dislocation. In the other specimens the posterior zone needed to be detached as well for a metastable dislocation. Conclusion: Purely labral lesions and limited capsulolabral detachments do not seem to be sufficient to allow the humeral head to dislocate. Labral tears can therefore be debrided without consequences for shoulder stability. In the present study, a capsulolabral detachment in the antero-inferior zone (the typical Bankart lesion) does not allow the humeral head to dislocate. This leads us to suggest that associated lesions must exist in chronic instability.

Purpose: Study experimental instability by ligament section leaving intact all the periarticular elements.

Material and methods: Sixteen fresh cadaver shoulders were studied. Dissection was achieved via an axillary approach isolating the ligaments without muscle section. Instability was classified in five stages: 0) stable, 1) drawer and sulcus, 2) subluxation: the head crossed the glenoid border but remained in the plane of the glenoid, 3) reversible dislocation: the head was dislocated by returned spontaneously into place when the arm was left to hang along the body, 4) permanent dislocation. The ligaments were sectioned in the following order: 1) betrween 7h and 5h, 2) between 5h and 2h, 3) between 1h and 11h. Instability was tested with usual manœuvres: drawer, sulcus, hyperabduction test, provoked dislocation in elevation and maximal external rotation, downward pressure in the axis of the humerus.

Results: Dissection of the ligaments produced class 1 instability in 0% of the shoulders, Section between 7h and 5h (anterior part of the inferior glenohumeral ligament) yielded class 2 instability in 12 cases, and class 3 instability in six. The hyperabduction test was positive in all shoulders. Section between 5h and 3h (middle glenohumeral ligament) produced class 3 instability in all the shoulders but never permanent dislocation. To obtain class 4 instabilty, section between 1h and 11h (superior glenohumeral ligament) was required. Section of the cuff was not necessary to obtain permanent dislocation.

Discussion: The role of the superior glenohumeral ligament in the production of shoulder instability has not been detailed to date. Closure of the rotator interval, proposed by Nobuhar and by Field, corresponds to retightening this ligament. The function of the superior glenohumeral ligament should be taken into account during the treatment of shoulder instability.