reverse prosthesis for proximal humerus fracture - kinex

| T E C H N I Q U E |

Reverse Prosthesis for ProximalHumerus Fracture, Technique and ResultsFran0ois Sirveaux, MD, PhD, Gregory Navez, MD, Olivier Roche, MD, and Daniel Mole, MDClinique de Traumatologie et d’OrthopedieSINCAL, Nancy, France

Matthew D. Williams, MDAcadiana Orthopaedic GroupLafayette, LA

| ABSTRACT

The reverse prosthesis has become a treatment of choicefor glenohumeral osteoarthritis in a cuff-deficient shoul-der. It allows restoration of mobility despite the loss ofrotator cuff function. Hemiarthroplasty, the historicalstandard of care for proximal humeral fractures requiringreplacement, fails most commonly because of tuberositynonunion. Therefore, in the face of poor bone quality,the reverse prosthesis is a logical choice for acute fracturesto improve postoperative mobility in elevation regardlessof tuberosity healing. Nevertheless, the healing of the tu-berosities may influence the recovery in active externalrotation. The purpose of this article is to report on thesuperolateral technique to implant a reverse shoulderprosthesis for acute proximal humeral fractures, empha-sizing the technique of tuberosity fixation, and to analyzepreviously published series. Considering the follow-upand complication risks, the reverse prosthesis should bereserved for fractures involving elderly patients. The deci-sion to use a reverse prosthesis for an acute fractureshould be based on a thoughtful risk-benefit analysis be-tween hemiarthroplasty and reverse shoulder arthroplasty.Keywords: shoulder arthroplasty, reverse prosthesis,acute proximal humerus fracture, tuberosity fixation,surgical technique

Reverse shoulder arthroplasty for proximal humeralfracture is not a new concept. Grammont himself

used this prosthesis for acute fractures and fracture se-

quelae early in his experience (22 cases between 1989and 1993), but the results were not published. Theresults of the reverse prosthesis for cuff tear arthropathyand after resection for tumor demonstrate that thereverse design restores active mobility in elevation de-spite a functionally incompetent rotator cuff.1,2 Pub-lished results on hemiarthroplasty for fractures clearlydemonstrate the poor results from the loss of rotatorcuff function because of tuberosity migration or non-union. Moreover, in cases of revision for a failed hemi-arthroplasty, the reverse prosthesis improves shoulderfunction.3Y6 The use of the reverse prosthesis for prox-imal humeral fracture appears logical in select elderlypatients. Only a few small series have reported the resultsof the reverse prosthesis for fracture, and the level of evi-dence is still low.7Y9

| INDICATIONS/CONTRAINDICATIONS

The current indications for a shoulder prosthesis in thesetting of acute fractures include a displaced 4-part frac-ture with or without humeral head dislocation, a headsplitting or impaction fracture involving more than 40%of the articular surface, and some 3-part fractures withmarked displacement and diminished bone stock. The re-verse prosthesis may be used in elderly patients whenthey present with poor prognostic factors for successfulhemiarthroplasty: age older than 75 years, medicalcomorbidities, poor bone quality of the tuberosities,and preoperative fatty infiltration of the rotator cuff orthe inability to complete postoperative immobilizationand rehabilitation protocols.

The reverse prosthesis is contraindicated in youngactive patients (except as a salvage procedure), incases of active infection or axillary nerve injury, andin patients with insufficient bone stock for secure fix-ation of the baseplate. Considering the risk of hemato-ma after reverse shoulder implantation regardless ofthe diagnosis, it is preferable to delay the procedure

Reprints: Fran0ois Sirveaux, MD, Phd, Clinique de Traumatologieet d’Orthopedie, SINCAL, Nancy, France (e-mail: [email protected]).

In support of their research for the preparation of this article, none ofthe authors received grants or outside funding. One or more of theauthors received payments or other benefits or a commitment or agree-ment to provide such benefits from a commercial entity (Tornier, StIsmier, France). There was no commercial entity paid or directed, oragreed to pay or direct, any benefits to any research fund, foundation,educational institution, or other charitable or nonprofit organizationwith which the authors are affiliated or associated.

Volume 9, Issue 1 15

Techniques in Shoulder & Elbow Surgery 9(1):15–22, 2008 � 2008 Lippincott Williams & Wilkins, Philadelphia

Copyright @ 2008 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

for 2 to 6 days after the trauma to decrease periopera-tive bleeding.

| PREOPERATIVE ASSESSMENT

Thorough patient evaluation with history and physical ex-amination is mandatory. Considering the age of thesepatients, it is important to assess their overall health statusand address or eliminate medical comorbidities that canpreclude the safe administration of anesthesia. Associatedtraumatic lesions (ie, hip fracture, radius fracture, etc) fre-quent in these patients must be taken into account duringpreoperative planning. Evaluation of the contralateralshoulder should be completed considering the current lim-itation in internal rotation with the reverse prosthesis. Inevery case, it is critical to evaluate the axillary nerve forinjury. The deltoid must be evaluated by clinical examina-tion and, if any questions arise, with an electromyogram.The muscle should be functional and contractile at theclinical examination. Weakness of the deltoid does notrepresent a strict contraindication to a reverse prosthesis.Radiographic studies needed to assess and classify thefracture include standard x-rays and a computed tomog-raphic (CT) scan. The CT scan provides a clear illustra-tion of the position of the humeral head and tuberositiesin complex situations10,11 and allows evaluation of therotator cuff tendons and fatty infiltration of the cuffmusculature. Actually, finding a deficient rotator cuffin cases of 3- or 4-part fractures is uncommon, reportedat 0% to 5%.12Y14 In addition, the CT scan demon-strates any glenoid bone defects and allows preopera-tive planning for the position of the central peg of thebaseplate. Bilateral full-length x-rays are useful to eval-uate the amount of proximal bone loss in cases of com-minuted fractures involving the metaphysis and allowthe surgeon to template the approximate height of theprosthesis for insertion.

| SURGICAL TECHNIQUE

ApproachOur preferred method and recommendation is to use thesuperolateral approach for these cases; however, the del-topectoral approach may be used as well. It has beendemonstrated that the risk of instability is higher byusing the deltopectoral approach.15 In fractures cases,there is a potential risk of instability because of hemato-ma, detachment of the lesser tuberosity, and proximalbone loss; in these cases, the superior approach seemspreferable. The deltopectoral approach is useful whenthe fracture extends distally down the humeral shaftand an extensile approach is required.

The patient is placed in a beach-chair position. Alongitudinal incision is made, starting from the acromio-clavicular joint and running distally for 4 cm from the

lateral edge of the acromion (Fig. 1). The anterior andmiddle deltoid muscles are separated, a retractor isplaced in the subacromial space, and the fracture hema-toma is removed. A stay-suture can be applied to the dis-tal deltoid split to avoid potential damage to the axillarynerve. A deltoid split can be extended distally butrequires direct visualization and vigilance during the pro-cedure to isolate and protect the axillary nerve.16 Theanterior deltoid and the coracoacromial ligament are de-tached subperiosteally from the acromion. An anterioracromioplasty may be performed if necessary to im-prove exposure. The deltoid is retracted anteriorly, andthe subacromial bursa is removed.

Fracture ExposureThe first step is to identify the fracture fragments. Therotator interval is opened along the bicipital groove be-tween the subscapularis and the supraspinatus. The inter-val between the supraspinatus and the infraspinatus isalso identified and opened. The supraspinatus tendon isthen resected to the level of the glenoid rim (Fig. 2).The proximal portion of the long head of the biceps ten-don is resected and the humeral head fragment removed.The greater tuberosity is mobilized posteriorly, and4 mattress sutures are placedVtwo through the infraspi-natus tendon and two through the teres minorVat thetendon-bone junction (Fig. 3). The lesser tuberosity isidentified anteriorly, and 2 stay-sutures are placed

FIGURE 1. The incision starts from the acromioclavicularjoint and runs distally for 4 cm from the lateral edge ofthe acromion.

Techniques in Shoulder & Elbow Surgery16

Sirveaux et al


through the subscapularis tendon at the tendon-bonejunction. The lesser tuberosity is retracted anteriorly forglenoid exposure.

Glenoid PreparationThe glenoid is prepared first, and the baseplate insertedwithout cement as customary.17Y19

The lesser tuberosity is retracted anteriorly and thegreater tuberosity posteriorly by using 2 Homan orKolbel retractors. Because of the epiphyseal fracture, gen-tle caudal traction is usually sufficient to expose the infe-rior portion of the glenoid (Fig. 4). A 2-prong Tiemanncapsular retractor can be added inferiorly along the scap-ular neck if needed. The meridian and equator of the gle-

noid are marked with a Bovie electrocautery to establish areference for guide placement and reaming. We alignthe 6-mm drill guide with the inferior border of the gle-noid with a 10-degree inferior tilt to allow for inferiorplacement of the glenosphere (Fig. 5). The baseplate isfixed as described by Hatzidakis,19 and the sphere is im-pacted and secured with the central screw into the peg ofthe baseplate.

Humeral Stem Preparation and TrialingAt this point, the proximal humerus is dislocated later-ally and superiorly with the greater tuberosity retracted

FIGURE 3. Position of the sutures through the infraspi-natus tendon bone junction.

FIGURE 4. Glenoid exposure.

FIGURE2. Fracture exposure and supraspinatus resection.

FIGURE 5. Alignment of the drill guide with the inferiorborder of the glenoid.


Reverse Prosthesis for Proximal Humerus Fracture


posteriorly and the lesser tuberosity retracted anteriorly.Distal diaphyseal reaming is performed until the reamercomes into contact with the cortex. The metaphysealreamer is usually unnecessary in these cases secondaryto metaphyseal comminution but should be used if meta-physeal bone is present. The trial component is assem-bled and inserted into the diaphysis at the preoperativelydetermined height in 20 degrees of retroversion and im-pacted with a mallet as necessary (Fig. 6). Should proxi-mal bone loss preclude an adequate interference fit withthe trial prosthesis, a sponge may be used around thestem to stabilize the trial component. The preoperative-ly determined height of insertion is a useful guide incases of proximal bone loss but is not definite. Theprosthesis is reduced (initially using the 6-mm polyeth-ylene trial spacer), and its final height adjusted accord-ing to the tension of the deltoid. Minimal diastasisshould be present between the cup and glenospherewith axial stress. The prosthesis must be stable secondaryto deltoid tension alone before fixation of the tuberosi-ties.19 The height may be adapted, after cementation,with the trial humeral inserts (6, 9, or 12 mm) orby adding a metaphyseal extension (+9 mm). In casesof diaphyseal involvement, a long-stem prosthesis isrequired.

Final Implantation and TuberosityReattachmentThe trial component is dislocated and removed. Twoholes are drilled laterally in the proximal humeral diaph-ysis 2 cm distal to the fracture. Two nonabsorbablesutures are passed through the holes to create a sutureloop in the diaphysis. Then, the final hybrid humeralcomponent (uncemented hydroxyapatite covered meta-physis and cemented stem) is assembled. A cementrestrictor is inserted to the appropriate depth, and themedullary canal is dried. After injecting the cement, theprosthesis is inserted into the canal, through the sutureloop, in 20 degrees of retroversion. The sutures are nowsecured around the prosthesis, which will improve tuber-osity fixation in elderly patients. Excess cement is re-moved proximally at the metaphyseal junction. After thehumeral component is well fixed and the cement is dry,a final trial reduction should be performed with the in-serts to confirm stability and appropriate deltoid tension.The final polyethylene component is inserted, and theprosthesis is reduced. The sutures passing through theinfraspinatus and teres minor should be passed aroundthe neck of the prosthesis before reduction.

After reduction, the tuberosities are mobilized and re-duced temporarily around the metaphysis. The techniqueof tuberosity fixation is derived from the 6-suture config-uration technique described by Boileau et al20 in 2000for hemiarthroplasty in acute fracture. The greater tuber-osity is first fixed by tying 2 of the previously passedsutures around the neck of the prosthesis (Figs. 7A, B).The remaining 2 sutures are passed around the neck ofthe prosthesis, through the tendon bone junction of thesubscapularis, and tied in the horizontal plane (Figs.7C, D). The transosseous metaphyseal sutures are usedas a figure-of-8 to improve the fixation between both tu-berosities and the diaphysis (Fig. 8). Due to medializa-tion of the proximal humerus, there is no tension onthe rotator cuff. After tuberosity fixation, the range ofmotion, stability, and soft tissue tension are evaluatedto design the postoperative program. A meticulous anti-septic irrigation of the subacromial space must be per-formed to remove the hematoma and bone particles toprevent infection and ossifications. Considering thehigh risk of hematoma, we use a suction drain in the sub-acromial space for a minimum of 2 days. At the end ofthe procedure, the anterior deltoid must be securely reat-tached using transosseous nonabsorbable sutures to theacromion. These sutures should grasp both the superfi-cial and the deep deltoid fascia.17

RehabilitationThere are no scientific data to support one method of re-habilitation over another. Some authors recommend im-mobilization for 4 weeks in slight abduction and neutral

FIGURE 6. Insertion of the trial component with 20degrees of retroversion.


Sirveaux et al


rotation17,19 with early passive motion, whereas othersadvocate using only a simple sling. We recommendavoiding active rotation during the first 6 weeks to pro-

vide an environment conducive to tuberosity healing andto prevent tuberosity migration.

| RESULTS

Our preliminary results of a prospective study showedthat the patients treated with a reverse prosthesis in thecase of acute fracture achieved an average of 113degrees of active elevation compared with 88 degreeswith hemiarthroplasty.18 The overall results were not asgood as in patients treated with a reverse prosthesis forcuff tear arthropathy. The problem of tuberosity healingand cuff dysfunction is only one of the reasons for fail-ure of the prosthesis for fracture.

In 2006, Cazeneuve and Cristofari8 reported theirexperience of 23 cases of reverse prostheses for acutefracture. Sixteen cases were reviewed at an average fol-low-up of 86 months. The mean age of the patients was75 years (range, 58Y90 years). The prosthesis wascemented in all cases but one, and the tuberositieswere sutured in 5 cases. Four complications werereported: 1 dislocation and 1 infection (both revised),and 2 cases of reflex sympathetic dystrophy.

The mean constant score was 60 points, and active an-terior elevation was more than 120 degrees in all the casesexcept for the 2 cases requiring revision. The recovery ofactive external rotation was better in cases where the tuber-osities had been fixed. On x-rays, a scapular notch

FIGURE 7. Tuberosity fixation. A, Fixation of the greater tuberosity, B, Axial view. C, Fixation of the lesser tuberosity,D, Axial view.

FIGURE 8. Final construct with the transosseous meta-physeal suture.




was noted in 69% of the cases, and 1 metaglene had be-come loose.

Bufquin et al7 reported the largest series of reverseprostheses for fracture (43 casesV40 included with amean age of 78 years). Eleven percent of the patientshad associated traumatic fractures. The superolateralapproach was used in 20 cases and the deltopectoral ap-proach in 23 cases. The tuberosities were fixed aroundthe prosthesis in all the cases. The humeral componentwas uncemented in 37 cases. A proximal epiphysealaugment was required to improve the stability ofthe prosthesis in 15 cases. The authors recommendedimplanting the humeral component in neutral rotation,but they did not demonstrate any statistical influenceof component rotation on the clinical results. A 28%(12 cases) rate of complications was reported in thisstudy. There was 1 perioperative glenoid fracture, 5transient neurological deficits, 1 acromion fracture, 1dislocation, 1 secondary deltoid rupture, and 3 casesof reflex sympathetic dystrophy. At a mean follow-upof 22 months (range, 6Y58 months), the average activeanterior elevation was 97 degrees. The Constant scoreaveraged 44 points, and the mean active external rota-tion in abduction was 30 degrees. The clinical resultswere lower for patients older than 75 years, and the re-covery in active external rotation was better when thegreater tuberosity had healed anatomically. Radiograph-ically, the tuberosities were displaced in 53% of thecases (13.8% malunion and 38.8% nonunion), and 90%showed periprosthetic ossifications.

Recently, a small series of 15 cases, retrospectivelyreviewed with more than 2 years’ follow-up, from alarge multicenter study of reverse prostheses, has beenreported9 (mean age, 78 years). At a mean follow-upof 46 months, the mean constant score was 55 points(range, 31Y73 points), the mean active anterior eleva-tion was 107 degrees, and the mean external rotationwas 10 degrees. The recovery of active external rotationwas possible when the greater tuberosity healed, butthis was not supported by the statistical data because ofthe small number of cases (Fig. 9). These results havebeen compared with a series of elderly patients treatedwith hemiarthroplasty. The mean results in active ante-rior elevation were not significantly different betweenthe 2 groups, but the distribution of the results was dif-ferent. In the reverse group, only 1 patient had lessthan 90 degrees of active anterior elevation, but the ac-tive anterior elevation never exceeded 150 degrees,whereas in the hemiarthroplasty group, 11% had morethan 150 degrees, but 50% achieved only 90 degreesor less.

In the hemiarthroplasty group, the constant scorewas strongly influenced by tuberosity healing. Themean constant score was 41.9 points in cases of

FIGURE 9. An 80-year-old female, 4-part fracture (A).One-year follow-up x-rays (B, C). Note the quality oftuberosity healing around the prosthesis. Clinical results inelevation, active external rotation, and internal rotation(D1, D2, D3, and D4).


Sirveaux et al


nonunion or malunion of the greater tuberosity com-pared with 59 points when the greater tuberosity healed(P G 0.01). The same results were observed for activeanterior elevation and active external rotation (75 vs116 degrees, and 25 vs 14 degrees, respectively). Ifwe considered active elevation in both groups withnonunion or malunion of the greater tuberosity, theresults would be in favor of the reverse prosthesis.When the greater tuberosity did not heal, the meanConstant score was 55 points for the reverse prosthesisand 41 points for the hemiarthroplasty (P = 0.12). Themean active anterior elevation was 75 degrees in thehemiarthroplasty group and 116 degrees in the reversegroup (P = 0.01). The active external rotation was notstatistically different, 16 degrees in the hemiarthroplastygroup and 13 degrees in the reverse group. This studydemonstrates that good results and near-normal rangeof motion can be achieved in elderly patients treatedwith hemiarthroplasty for acute fractures. However,if good fixation and healing of the tuberosities are notachieved, poor functional outcomes can be expected.The reverse prosthesis provides improved clinical resultsin these cases.

The reverse shoulder prosthesis must be consideredas an alternative to hemiarthroplasty in select elderlypatients. Indeed, shoulder hemiarthroplasty performedfor acute fracture is more demanding than one per-formed for other indications,12 and hemiarthroplastyperformed by a high-volume surgeon or high-volumehospital is more likely to result in a better outcome.21,22

The reverse prosthesis is somewhat more forgiving be-cause the recovery in elevation can be expected despitetuberosity nonunion or malunion as demonstrated bythis study. However, the healing of the greater tuberosityis still necessary to recover active external rotation. Theresults of the reverse prosthesis are clearly influencedby technical factors, and the surgeon must consider thehigh risk of complications which can compromise post-operative shoulder function and the ultimate level of in-dependence of these patients. In addition, the reverseprosthesis is a constrained prosthesis, and the current lit-erature raises concerns regarding the durability of thefixation of the prosthesis in the long term.23 Indeed,Guery et al24 showed that the survival rate of the reverseis better in cuff tear arthropathy than other etiologies(rheumatoid arthritis, trauma, and revision). The indica-tion for a reverse prosthesis in acute fractures shouldbe based on realistic expectations of tuberosity non-union around a hemiarthroplastyVwhich would resultin a poor outcome with hemiarthroplasty. The decisionto use a reverse prosthesis in an elderly patient withan acute fracture should be based on a thoughtful risk-benefit analysis between hemiarthroplasty and reverseshoulder arthroplasty.

| REFERENCES

1. Sirveaux F, Favard L, Oudet D, et al. Grammont invertedtotal shoulder arthroplasty in the treatment of glenohumeralosteoarthritis with massive rupture of the cuff. Results of amulticentre study of 80 shoulders. J Bone Joint Surg Br.2004;86:388Y395.

2. De Wilde L, Sys G, Julien Y, et al. The reversed Deltashoulder prosthesis in reconstruction of the proximalhumerus after tumour resection. Acta Orthop Belg.2003;69:495Y500.

3. De Wilde L, Mombert M, Van Petegem P, et al. Revisionof shoulder replacement with a reversed shoulder prosthe-sis (Delta III): report of five cases. Acta Orthop Belg. 2001;67:348Y353.

4. Paladini P, Collu A, Campi E, et al. The inverse prosthesisas a revision prosthesis in failures of shoulder hemiar-throplasty. Chir Organi Mov. 2005;90:11Y21.

5. Jouve F, Wall B, Walch G. Revision of shoulderhemiarthroplasty with reverse prosthesis. In: Walch G,Boileau P, Mole D, eds. Reverse Shoulder Arthroplasty,Clinical ResultsVComplicationsVRevision. Montpellier:Sauramps Medical; 2006:217Y227.

6. Gohlke F, Rolf O. Revision of failed fracture hemiarthro-plasties to reverse total shoulder prosthesis through thetranshumeral approach: method incorporating a pectoralis-majorYpedicled bone window. Oper Orthop Traumatol.2007;19:185Y208.

7. Bufquin T, Hersan A, Hubert L, et al. Reverse shoulderarthroplasty for the treatment of three- and four-partfractures of the proximal humerus in the elderly: aprospective review of 43 cases with short term follow-up.J Bone Joint Surg Br. 2007;89:516Y520.

8. Cazeneuve JF, Cristofari DJ. Grammont reversed prosthe-sis for acute complex fracture of the proximal humerus inan elderly population with 5 to 12 years follow-up. RevChir Orthop Reparatrice Appar Mot. 2006;92:543Y548.

9. Sirveaux F, Navez G, Favard L, et al. Reverse prosthesisfor acute proximal humerus fracture, the multicentric study.In: Sirveaux F, ed. Reverse Shoulder Arthroplasty. ClinicalResults, Complications, Revision. Montpellier: SaurampsMedical; 2006:73Y80.

10. Dines DM, Tuckman D, Dines J. Hemiarthroplasty forcomplex four-part fracture of the proximal humerus:technical considerations and surgical technique. Univ PaOrthop J. 2002;15:29Y36.

11. Mora Guix JM, Gonzalez AS, Brugalla JV, et al. Proposedprotocol for reading images of humeral head fractures. ClinOrthop Relat Res. 2006;448:225Y233.

12. Robinson CM, Page RS, Hill RM, et al. Primary hemiar-throplasty for treatment of proximal humeral fractures. J BoneJoint Surg Am. 2003;85-A:1215Y1223.

13. Parsch D, Wittner B. Prevalence of rotator cuff defects indislocated fractures of the humerus head in elderly patients.Unfallchirurg. 2000;103:945Y948.




14. Mighell MA, Kolm GP, Collinge CA, et al. Outcomes ofhemiarthroplasty for fractures of the proximal humerus.J Shoulder Elbow Surg. 2003;12:569Y577.

15. Nove-Josserand L, Walch G, Wall B. Instability of thereverse prosthesis. In: Walch G, Boileau P, Mole D, eds.Reverse Shoulder Arthroplasty, Clinical Results, Complica-tions, Revision. Montpellier: Sauramps Medical; 2006:247Y260.

16. Webb M, Funk L. An anterosuperior approach for proximalhumeral fractures. Tech Shoulder Elbow Surg. 2007;7:77Y81.

17. Seebauer L. Reverse prosthesis through a superiorapproach for cuff tear arthropathy. Tech Shoulder ElbowSurg. 2006;7:13Y26.

18. Sirveaux F, Mole D, Boileau P. The reversed prosthesis. In:Warner JJ, Iannotti JP, Flatow E, eds. Complex andRevision Problems in Shoulder Surgery. Philadelphia:Lippincott Williams and Wilkins; 2006:497Y511.

19. Hatzidakis AM, Norris TR, Boileau P. Reverse shoulderarthroplasty, indications, technique and results. TechShoulder Elbow Surg. 2005;6:135Y149.

20. Boileau P, Walch G, Krishnan S. Tuberosity osteosynthesisand hemiarthroplasty for four-part fractures of the proximalhumerus. Tech Shoulder Elbow Surg. 2000;1:96Y109.

21. JainN, Pietrobon R, Hocker S, et al. The relationship betweensurgeon and hospital volume and outcomes for shoulderarthroplasty. J Bone Joint Surg Am. 2004;86-A:496Y505.

22. Kralinger F, Schwaiger R, Wambacher M, et al. Outcomeafter primary hemiarthroplasty for fracture of the head ofthe humerus. A retrospective multicentre study of 167patients. J Bone Joint Surg Br. 2004;86:217Y219.

23. Guery J, Favard L, Sirveaux F, et al. Reverse total shoulderarthroplasty. Survivorship analysis of eighty replacementsfollowed for five to ten years. J Bone Joint Surg Am. 2006;88:1742Y1747.


Sirveaux et al


reverse prosthesis for proximal humerus fracture - kinex

Documents