The technique of hemiarthroplasty in four - part fractures of the upper end
of the humerus

Kontakis G1, Bioleau P1, Walch P2
University Orthopaedic Department, Hopital de L' Archet, Nice, France
Orthopaedic Department, St. Anne Lumiere, Lyon, France

In the beginning of the 1950's, C. Neer suggested hemiarthroplasty for the treatment of the displaced fractures of the upper end of the humerus with restoration and fixation of the humeral tuberosities[31,32,33]. According to Neer, the results of the humeral head replacement in three - and four - part fractures were satisfactory or excellent in more than 90% of patients[32,33]. Several authors have mentioned the moderate or disappointing results, with painful and stiff shoulders, confirming that the prosthetic head replacement by itself can not assure the successful outcome[5,6,8,13-15,17,19,20,21,23,27,29,35,37,39,41]. There are no other articles that confirm the results of C. Neer for these types of fractures. The poor results of the hemiarthroplasty led the surgeons to revise their opinion or to turn to conservative methods for the treatment of these complex injuries[22,36,42,44,45].
The technical details have an important role for the poor and nonpredictable results after the hemiarthroplasty of shoulder for the four - part humeral fractures. Due to the rareness of these fractures, the surgeon has not adequate experience for their treatment[14,32,33]. The proper placing of the prosthesis is not so easy because there are no clear anatomical landmarks. The reduction and stabilization of the dislocated humeral tuberosities is difficult, too[14,39]. There is a high incidence of malunion, nonunion and bony resorbtion of the tuberosities [1,5,11,30,39].
A preliminary clinical study showed the inadequate and nonpredictable results of the prosthetic replacement in the dislocated four - part fractures of the upper humerus3. We observed a clear positive correlation of the qualitative anatomical restoration and the functional result. According to our studies, the poor functional results are directly correlated with the incorrect placing of the prosthesis and the poor reduction and stabilization of the tuberosities[3,5,8]. So, the efforts are focused in two directions; the first one is the planning and the development of a proper guide - apparatus, for the precise implantation of the prosthesis in the proper height and retroversion. The second one has to do with the development of a simple, understandable surgical technique for the fixation of the tuberosities[5,6,8].

Key words: upper humeral end fractures, hemiarthroplasty.

INDICATIONS FOR HEMIARTHROPLASTY IN FRACTURES AND THE NEED OF THE GUIDE -
APPARATUS FOR THE PROPER PLACING OF THE PROSTHESIS
The shoulder hemiarthroplasty is usually indicated in the following upper - end humeral fractures: 1) four - part fractures in elderly patients, 2) four - part fractures-dislocations, 3) head split fractures with displacement, 4) impression fractures, with destruction of more than half of the head articular surface and 5) selected three - part fractures in elderly patients with decreased bone mass[4,12,18,31,32,34,43].
The patient's age and the bone quality are important factors for the planning of the surgical therapy. Most of the displaced fractures of the upper end of humerus occur in elderly patients, usually women, with osteoporotic bones[22,36]. In these patients, the bone quality makes difficult the fixation of the humeral tuberosities and this must be ever considered.
In the upper - end of humeral fractures, the precise reduction of the tuberosities and the anatomical restoration are necessary for the proper use of the prosthesis. This can be possible only if the prosthesis is located in a proper, steady place related to the height and the retroversion. Three reasons lead us to search for one more objective method in order to achieve an anatomical restoration of the upper end of humerus precisely and reproducibly.
1. The instability of the humeral prosthesis into the humeral.
The prosthesis may be unstable during the operation, during the trial reduction as well as during the final implantation. This impedes the right estimation of the height and the retroversion. We believe that the prosthesis should be cemented in these fractures because this is the only way for its stabilization in the proper height and orientation.
2. Inaccuracies in visual estimation.
The estimation with the eye, of the proper place of the humeral prosthesis does not ensure the correct restoration of the anatomy, and sometimes leads to the faulty placing of the prosthesis.
In the frontal plane, it is usual the humerus to be elongated more than normal. We have mentioned that every increase of the humeral length over the 10 mm, due to highly implanted prosthesis, is very destroying for the anterior flexion as well as the postoperative pain[3,4,5,9]. Certainly, the opposite situation is possible, too. This occurs more commonly when there is a metaphyseal comminution.
In the transverse plane, there is a tendency to implant the prosthesis in excess retroversion. This is due to several factors [5,8]. The implantation of the prosthesis with the proper retroversion is depended on the experience of the surgeon. Anatomical reviews showed that the retroversion is different in each individual[3,7,9,10,24,25]. Based on the results of an our study and on the studies of other authors, we consider as mean retroversion the rate of 20[o] which corresponds to the initial evaluation of C. Neer in the beginning of the 1950's decade (Neer 1955). According to several authors however the implantation of the humeral prosthesis in 35[o]-40[o] retroversion is recommended [11,12,43]. This excess retroversion of the prosthesis changes the center of the rotation of the glenohumeral joint and alters the biomechanics of the joint[7].
The surgeons tend to implant the prosthesis in excess retroversion due to their fear for the postoperative anterior dislocation. The absence of a clear landmark peripheral of the fracture, is another factor for the incorrect orientation. If we take into consideration the variable carrying angle of the forearm 38, the trans-epicondylar elbow axis is a better guide for the proper retroversion than the position of the forearm.
The bicipital groove is not a precise landmark for the orientation of the prosthesis5,16,28,40. In a recently published study the distance from the posterior edge of the bicipital groove where the lateral prosthetic fin should be placed in order to reproduce the normal retroversion was found highly variable, with a mean value 5,2±2,6 mm. If we locate the lateral fin of the prosthesis just behind the posterior edge of the groove, it results in a difference from normal retroversion ranging between -1,5[o] to 41,7[o] (mean value 20[o]), (over-retroverted prosthesis)[26] (figure 1). So, when we try to place the prosthesis in 30[o]-40[o][ retroversion we can create excess retroversion such as 50[o]-60[o].
3. Underestimation of the theoretical consequences after the inappropriate placing of the humeral prosthesis.
The theoretical consequences of the wrong placing have been underestimated until nowadays. The too high implanted prosthesis may lead to poor functional results. If the superior rotator cuff is distented excessively, the supraspinatus muscle may undergo thinning out or rupture due to squeezing between the prosthetic head and the acromion. This results in a painful arm with restricted motion. The elongation may cause excessive tension in the stay sutures of the tuberosities, which may lead in posterior migration and/or nonunion of the greater tuberosity (figure 2).
The placing of prosthesis in excessive retroversion may also leads to poor functional result. Theoretically, this may contribute in increased tension forces on the stay sutures of the tuberosities, when the humerus is internally rotated, resulting in posterior migration of the greater tuberosity and deterioration of the clinical result (figure 3).
For all these reasons, the surgeons need a more objective solution, in order to succeed an effective restoration of the proximal humerus anatomy. The use of a specific guide - apparatus (a jig), for the precise implantation of the prosthesis in the upper end humeral fractures, could minimize the subjective estimations of the surgeons with low experience.

The Aequalis fracture jig
The Aequalis fracture jig (Tornier, Inc., St. Ismier, France) is an original device aiming first to stabilize the prosthesis and second to help the surgeon to implant correctly the prosthesis in height and in retroversion. We use as landmarks the bone marks of the elbow, i.e. the medial and lateral epicondyle. This is an extramedullary guide which consists of:
1. An arm support, which keeps the elbow in constant flexion and helps the stabilization of the prosthesis.
2. A goniometer for the determination of the constant angle of the retroversion.
3. A ruler for the determination and maintaining of the constant length of the humerus.
The above mentioned components plus a prosthesis holder are connected and constitute a total.

 

Figure 1. The lateral fin of the prosthesis should be located in distance d from the posterior lip of the bicipital groove in order to reproduce the retroversion in each individual. If we place the lateral fin just behind the posterior edge of the bicipital groove, then there is a difference in the real retroversion φ'.

Figure 2. If the prosthesis is located higher than the proper, it may occur (A) superior migration of the greater tuberosity or (B) pseudarthrosis between the greater tuberosity and the diaphysis or (C) both of them.

Preoperative planning
The three radiographs for the trauma of the shoulder (true anteroposterior, axillary and Y-scapular views) are necessary before any decision is taken[32]. In addition, we need a scaled radiograph of the humerus in order to determine the humeral length. For this reason, we perform a true anteroposterior radiograph of the healthy shoulder which images the whole humerus. A 10 cm ruler is adapted to the lateral surface of the arm, parallel to the bone, for the determination and correction of any magnification. The length of the humerus is estimated with the radiograph under scale of the sound humerus. The surgeon should inspect the imaging technique. The estimation of the length is made by drawing the prosthetic axis (centre of the proximal bone canal) and two perpendicular lines on it; the upper one from the highest point of the greater tuberosity and the lower one from the most medial point of the medial epicondyle (figure 5).
As retroversion angle of the humeral head, we choose the mean angle 20[o] in relation with the transepicondylar elbow axon.

Figure 3. The excess retroversion will possibly lead to detachment and posterior migration of the greater tuberosity, due to the excessive tension on the sutures when the arm is in internal rotation.

Technique of the fixation of the tuberosities and the hemiarthroplasty
We must give priority to the correct placing of the humeral prosthesis in relation with the height and the retroversion, in order to allow the anatomical restoration of the tuberosities[3,8,9]. For this reason, we use the Aequalis fracture jig[7].
The patient is placed in the beach chair position and the shoulder is kept free from the edge of the table, allowing to the surgeon to place the humerus in full extension. By this way, the humeral canal is able to be reamed vertically, without the obstruction of the coracoacromial arch. A large deltopectoral approach is preferred for the protection of the deltoid muscle. In order to achieve the best exposure it is needed patience and careful soft tissue handling.
The incision is extended from the clavicle (just under the cocaroid process) to the tendon of the pectoralis major inferiorly, which can be dissected 1-2 cm in the upper part. This surgical approach offers a very good operative field and allows the adduction - internal rotation of the arm, or the abduction -external rotation whenever is necessary intraoperatively.
A self retaining distractor is used to hold the deltoid muscle laterally and the conjoined tendon medially. The coracoacromial ligament must be preserved, because it can be used as a landmark for the length of the prosthesis and will help prevent postoperative anterosuperior instability of the prosthesis.
The manipulation of the tuberosities must be very careful in order to maintain any intact periosteal attachment to the diaphysis. Once the tendon of the long head of the biceps has been found, the fracture line between the tuberosities can be idendified. Usually, the fracture line is behind the bicipital groove. The greater or the lesser tuberosity is retracted and the rotator interval is opened as far as medial as the coracoid process. Initially we preserved the tendon of the long head of biceps, but we observed that it impedes the fixation of the tuberosities and is a reason for the postoperative pain. Nowadays, we carry out systematically a tenodesis of the long head in of the biceps after the incision of its intraarticular portion.

Placing of the sutures
With the arm in abduction and internal rotation, two pairs of strong non-absorbable sutures (Flexidene #5, Braun, Germany)-four horizontal sutures- are inserted around the greater tuberosity through the bone-infraspinatus tendon junction. A non - permanent suture of different colour around the lesser tuberosity at the bone-tendon junction, in order to retract the osteoporotic bone without harming it. The tuberosities are carefully retracted, and the humeral head fragment is removed after resection of any capsular attachments. The condition of the glenoid cavity is checked for defects or erosion and any capsular tear or labral detachmnent should be repaired to avoid postoperative instability of shoulder.
Humeral preparation and choice of the implant size
The diameter of the humerus is measured with a millimeter caliper and the size of the prosthetic head is properly chosen. This size is chosen according to the horizontal (smaller) diameter of the removed head. Generally, there is a tendency to overestimate the size of the humeral head and for this reason if we have to select between two different sizes, we choose the smaller one. The humeral canal is prepared using reamers (with hand), retracting the tuberosities gently and holding the humerus in abduction and extension. A trial implant is assembled with a stem diameter corresponding to the last reamer used, with neck angle 130[o] and the head corresponding to the diameter of the fractured head.

Positioning of the prosthesis
The exact position of the humeral prosthesis in relation with its length and retroversion is performed with the help of the Aequalis fracture jig. An average retroversion of 20[o] is selected. The trial implant is attached to the jig and it is carefully inserted to the humeral canal, holding the arm in extension and external rotation while gently retracting the tuberosities. With the jig in place the arm is put in flexion and the prosthesis is reduced into the glenohumeral joint before attempting any positioning with regard to humeral height and retroversion. After the reduction, the surgeon can position of the epicondylar guides at the elbow. This step is very important for the determination of the correct anatomic position of the prosthesis. In the case of a large big elbow, or in cases where the palpation of the epicondyles is difficult, fluoroscopy can be used to verify the position of the epicondylar guides.

Figure 4. The Aequals Fracture Jig.

Positioning of the tuberosities
After the prosthesis is stable, reduced inside the joint article and in correct height and retroversion, any adjustments of the tuberosity fragments can be easily carried out. Adjustments should always be done with the prosthesis reduced inside the joint and the arm placed in neutral rotation. The greater and lesser tuberosities are reduced alternatively, under the prosthetic head. The reduction of the greater tuberosity is the most important step. Vertically, the tip of the greater tuberosity must be level with or 5 mm below the prosthetic head. Horizontally, the greater tuberosity must overhang the fin of the prosthesis anteriorly.

Figure 5. Estimation of the humeral length using a scaled radiograph pre-operatively.

Checking of the anatomical restoration
The proper length of the humerus is confirmed using the following criteria.
1. It shouldn't be gap or overlapping between the greater tuberosity and the diaphysis.
2. The greater tuberosity should be located level with or 5 mm below the prosthetic head.
3. The supraspinatus muscle should pass free, without being squeezed, between the prosthetic head and coracoacromid ligament.
4. If there is medial metaphysial communication and bone fragments are removed, then the inferior - medial part of the prosthetic head should be placed slightly higher in relation with the humerus. Otherwise, the prosthetic head should be in contact with the medial cortex.
The evaluation of the proper retroversion is carried out with the humerus in neutral rotation and using the following criteria:
1. The prosthetic head should be exactly opposite to the glenoid cavity.
2. The greater tuberosity should overhang the fin of the prosthesis anteriorly.
3. The lateral fin of the prosthesis should be located 5 mm to 10 mm posteriorly of the bicipital groove, at the level of the surgical neck.
4. The proximal and the upper part of the bicipital groove must be aligned.

Cementation of the definitive prosthesis
If all parameters are acceptable, the definitive prosthesis can then be cemented. Before cementation two holes are drilled in the proximal part of the shaft and two strong, non-absorbable sutures are passed for the vertical fixation of the tuberosities. The humerus is hold in extension and adduction, and trial prosthesis is removed from the Jig, in order to replace it with the definitive prosthesis. The trial prosthesis and its holeder (a component of the Jig) are removed together, while the other parts of the Fracture Jig are locked in the proper height and retroversion. The implant is cemented after the diaphysis has been plugged distally with a cement restrictor. When the definitive prosthesis has been inserted into the humeral canal, the prosthesis holder is reattached to the Jig and the prosthesis is reduced into the joint.
We suggest the use of the acrylic cement for the fixation of the humeral prosthesis in fracture cases for the following reasons:
1. The patients are often elderly patients with osteoporosis, large medullary canals, minimal or no cancellous bone and very thin cortex.
2. The prosthesis can not be fixated in the proper place due to fracture of the tuberosities.
3. A press-fit prosthesis is very difficult to be placed in the proper height and can not give the proper length of the humerus. The excess of the cement is removed and cancellous bone graft are placed between the tuberosities and the diaphysis in order to succeed bone union. The graft is taken from the head which is removed.

Figure 6. For the fixation of the tuberosities, four horizontal cerclage sutures are
used and two vertical sutures as tension bands.

Fixation of the tuberosities
The four horizontal, non-absorbable sutures which were previously passed between the greater tuberosity bone and the infraspinatus tendon, are passed around the neck of the prosthesis. The restoration is applied in three steps, stabilizing at first the greater tuberosity (figure 6).
1. Two horizontal, cerclage sutures around the greater tuberosity.
We start from the greater tubercle with two horizontal sutures. Passing these sutures through the holes of the lateral fin of the prosthesis do not provide enough stabilization and for this reason we suggest to pass the sutures all round the neck of the prosthesis. By this way, the greater tuberosity is attached much better, under the prosthetic head, providing stronger stabilization in the fixation process (figure 7).
2. Two horizontal, cerclage sutures around the both tuberosities.
The next step is the reconstruction of the lesser tubercle with the other two sutures. These are passed through the subscapularis tendon, in its insertion, with direction medially to laterally. By this way, the lesser tuberosity is pulled and supported in its place under the prosthetic head (figure 8).
3. Vertical sutures in both tuberosities (tension band).
Using the two sutures, which have been passed through the opened holes of the diaphysis and applying the technique of the tension band, we pass the first suture through the subscapularis and supraspinatus tendons and the other one through infra- and supraspinatus tendons (figure 9). By this way, the tuberosities are also held in the diaphysis of the humerus.

Figure 7. The fixation of the greater tuberosity is accomplished (A) using the horizontal sutures
(B) passed around the prosthetic neck.

Figure 8. The fixation of the lesser tuberosity is accomplished (A) using the last two sutures
which (B) passed around the prosthetic neck and then are tagged.

Figure 9. The two vertical sutures are passed from the superior - anterior and the superior -
posterior cuff and function as tension bands.

Tenodesis of the tendon of the long head of biceps
After the removal of the intraarticular portion of the long head of biceps tendon, the tendon is held in the bicipital groove using a modified Kesler suture. We pass the one tip of the suture in the supraspinatus tendon and the other one in the subscapularis tendon. Then we tie the suture.
Then, the range of motion (ROM) of the shoulder is checked. We must clearly observe complete anterior elevation, external rotation 40o with the humerus at side and internal rotation 70[o] with the humerus abducted. There should be no visible movements of the tuberosities during the rotation or flexion of the humerus. Also, the prosthetic humeral head should be able to be translated (push-pull) 50%, in the anteroposterior and superioinferior direction.

Detachment and migration of the tuberosities
This is the most common complication and the most important reason which is related with the poor functional results after hemiarthroplasty for a proximal humeral fracture. This will lead to nonunion, malunion and partial or complete bone resorption[1,8,11,30]. If there is displacement of the greater tuberosity, the functional results are poor and the radiological pattern is always the same; absence of bone under the prosthetic head, and superior migration of the prosthesis under the acromion. The CT imaging reveals the greater tuberosity which is displaced posteriorly, and hidden by the neck of prosthesis. Three common mistakes are related with the migration of the greater tuberosity[6,8]:
1. The unproper placing of the prosthesis with regard to the length (high) and the retroversion (excess retroversion).
2. The uncorrect location of the tuberosities in the frontal or in the transverse or in both planes. If the greater tuberosity is fixed higher from the prosthetic head, this will lead to the subacromial impingement. If the greater tuberosity is fixed lower from the prosthetic head, there is excessive tension of the spinati tendons. In the traverse plane, if the greater tubercle is located excessively posteriorly, this will lead to impingement with the posterior lip of the scapula.
3. Poor fixation of the tuberosities. The quality of the sutures, the technique for placing them and the sequence of their reconstruction play an important role. We suggest to reconstruct first the greater tuberosity and then the lesser tuberosity as described above. The reconstruction of the greater tuberosity in the beginning helps the control of the tuberosity position and its proper placement in relation with the head. Prerequisite for the proper reconstruction of the tuberosities, is the stabilization of the prosthesis in the correct height and retroversion.

Postoperative treatment
The shoulder mobilization is necessary in order to avoid the stiffness of the shoulder, due to intra- and extraarticular adhesions. We should remember that a more aggressive kinesiotherapy will put in danger the fixation of the tuberosities and may lead to poor functional results due to the detachment of the tuberosities before the bone union. In the last years, we tend to delay the rehabilitation program, avoiding the rotation of the arm for the first 6-8 weeks. We follow the principles of the Neer's rehabilitation program, adding the hydrotherapy program[2,5,32,33].
Initially, we aim to the recovery of the passive range of motion. We start the efforts the next day of the operation till the 6th - 8th postoperative week. The shoulder is immobilized with the elbow at side in neutral rotation or in an abduction brace.
The recovery of the joint mobility is succeeded with the passive exercises which must be done 5 times everyday, for 5 minutes each time, by the patient himself. The passive anterior elevation of the branchium is attempted only with pendulum motions. The internal and external rotation with the arm at side are not allowed because the fixation of the tuberosities may be in danger.
The second phase of the rehabilitation is related with the muscular strengthening, initially isometrically. The program begins 6-8 weeks postoperatively, after the healing of the tuberosities. The passive motion should be fully recovered until this period. The recovery of the muscular strength is succeeded with three types of exercises: 1)stabilization of the raised arm, 2) eccentric exercises applying progressive tension to the rotator cuff and 3) isometric exercises. The dynamic exercises against resistance using elastic bonds, strengthen the muscles of the shoulder girdle.
The third phase consists of recovering the extremes of passive ROM by stretching. Theses exercises aim to the recovery of the articular capsule's elasticity.
We have to encourage the patients to visit the swimming pool from the first phase (after healing of the operative wound) of their rehabilitation. The water is useful in all the phases of the program. In the first phase, for the support of the passive movements, in the second as resistance for the muscular strengthening and in the third phase as support and help for the distention of the tissues. The patient must continue the rehabilitation for at least 6-12 months after the operation.
The detachment and migration of the greater tuberosity or the other anatomical and surgical problems should be suspected in the following situations: 1) patients with abnormally painful shoulder in the early postoperative period, 2) if there is difference between the active elevation (residual) and the passive anterior elevation (preserved) and 3) if the range of the active motion is reduced by the time[2,5,6,8].

BIBLIOGRAPHY
1. Bigliani L.U., Flatow E.L., Cluskey M.C., Fisher R.A. Failed prosthetic replacement for displaced proximal humeral fractures. Orthop Trans. 1991; 15, 747-8.
2. Boileau P., Caligaris-Cordero B., Payeur F., Tinsi L., Argenson A. Prognostic factors after rehabilitation following shoulder arthroplasty for fracture. Rev Chir Orthop. 1999; 85, 106-16.
3. Boileau P., Chad E., Perraud M., Arcamone H., Argenson C. Restoration of humeral length and humeral offset after shoulder prosthesis for fractures. Rev Chir Orthop. 1993; 73, 150.
4. Boileau P., Huten D. Technique chirurgicale des protheses de Neer pour fractures cephalo-tuberositaires. Rev Chir Orthop. 1995; 85-8.
5. Boileau P., Walch G. Shoulder arthroplasty for proximal humeral fractures: problems and solutions. In: Walch G, Boileau P, eds. Shoulder Arthroplasty. Benin: Springer. 1999; 297-314.
6. Boileau P., Walch G. Preoperative planning of the use of the Aequalis fracture jig. In: Walch G, Boileau P, eds. Shoulder Arthroplasty. Berlin: Springer. 1999; 315-22.
7. Boileau P., Walch G. Three dimensional geometry of the proximal humerus: Implications for the surgical technique and prosthetic design. J Bone Joint Surg. Br 1997; 75, 857-65.
8. Boileau P., Walch G. Results of shoulder arthroplasty in acute fractures of the proximal humerus. In: Walch G, Boileau P, eds. Shoulder Arthroplasty. Berlin: Springer. 1999; 331-48.
9. Boileau P., Walch G., Mazzoleoni N., Urien J.P. In vitro study of humeral retrotorsion. J Shoulder Elbow Surg. 1993; 2(suppl), 33.
10. Broca P. La torsion de I' humerus et Ie tropometre, instrument destine a mesurer la torsion des os (redigee par L. Mamouvier). Revue d 'Anthrop, 2eme serie 4. 1881; 193-210, 385-423, 1577-92.
11. Compito C.A., Self E.B., Bigliani L.U. Arthroplasty and acute shoulder trauma, reasons for success and failure. Clin Orthop. 1994; 307, 27-36.
12. Craig E.V. Prosthetic replacement for four-part fractures of the proximal humerus. Tech Orthop. 1989; 3, 70-80.
13. De Anquin C.E., De Anquin C.A. Prosthetic replacement in the treatment of serious fractures of the proximal humerus. In: Kessel BL, ed. Shoulder Surgery. Berlin: Springer Veriag. 1982; 207-17.
14. Des Marchais J.E., Maurais G. Treatment of complex fractures of the proximal humerus by Neer's hemiarthroplasty. In Bateman J, Welsh R, eds. Surgery of the shoulder. Philadelphia: BC Decker. 1984; 60-3.
15. Dimakopoulos P., Potamitis N., Lambiris E. Hemiarthroplasty in the treatment of comminuted intraarticular fractures of the proximal humerus. Clin Orthop. 1997; 341, 7-11.
16. Doyle A.J., Burks R.T. Comparison of humeral head retroversion with the humeral axis/biceps groove relationship: a study in live subjects and cadavers. J Shoulder Elbow Surgery. 1998; 7, 453-7.
17. Frich L.H., Sojberg J.O., Sneppen O. Shoulder arthroplasty in complex acute and chronic proximal humerus fractures. Orthopaedics. 1991; 14, 949-54.
18. Gerber C., Hersche O., Berberat C. The clinical relevance of posttraumatic avascular necrosis of the humeral head. J Shoulder Elbow Surg. 1998; 7, 586-90.
19. Goldman R.T., Koval K.J., Cuomo F., Gallagher M.A., Zuckerman J.D. Functional outcome after humeral head replacement for acute three -and four- part proximal humeral fractures. J Shoulder Elbow Surg. 1995; 4, 81-6.
20. Green A., Barnard W.L., Imbird R.S., Providence R.l. Humeral head replacement for acute four-part proximal humerus fractures. J Shoulder Elbow Surg. 1993; 2, 249-54.
21. Hawkins R.J., Switlyk P. Acute prosthetic replacement for severe fractures of the proximal humerus. Clin Orthop. 1993; 289, 156-60.
22. Horak J., Nilson B. Epidemiology of fractures of the upper end of the humerus. Clin Orthop. 1975; 112, 250-3.
23. Huten D., Dupark J. L' arthroplastie prothetique dans les traumatisms complexes recents et anciens de I' epaule. Rev Chir Orthop. 1986; 72, 517-29.
24. Iannotti J.P., Gabriel J.P., Schneck S.L., Evans B.G., Misra S. The normal glenohumeral relationships: An anatomical study of one hundred and forty shoulders. J Bone Joint Surg. Am 1992; 74, 491-500.
25. Khral V.E., Evans F.G. Humeral torsion in man. Am J Phys Anthrop. 1945; 3, 229-53.
26. Kontakis G.M., Damilakis J., Christoforakis J., Papadakis A., Katonis P., Prassopoulos P. The bicipital groove as a landmark for the orientation of the humeral prosthesis in fracture cases. J Should Elbow Surg. 2001 in press.
27. Kraulis J., Hunter G. The results of prosthetic replacement in fracture-dislocations of the upper end of the humerus. Injury. 1977; 8, 129-31.
28. Kummer F., Pekins R., Zuckerman J.D. The use of the bicipital groove for alignment of the humeral stem in shoulder arthroplasty. J Shoulder Elbow Surg. 1998; 7, 144-6.
29. Moeckel B.H., Dines D.M., Warren R.F., Altchek D.W. Modular hemiarthroplasty for fractures of the proximal part of the humerus. J Bone Joint Surg. Am 1992; 74, 884-9.
30. Muldoon M.P., Cofield R.H. Complications of humeral head replacement for proximal humeral fractures. Instr Course Lect. 1997; 46, 15-24.
31. Neer C.S. Articular replacement of the humeral head. J Bone Joint Surg. Am 1955; 37, 215-28.
32. Neer C.S. Displaced proximal humeral fractures: 1, classification and evaluation. J Bone Joint Surg. Am 1970; 52, 1077-1089.
33. Neer C.S. Displaced proximal humeral fractures: 2, treatment of three-part and four part displacement. J Bone Joint Surg. Am 1970; 52, 1090-1103.
34. Nicholson G.P., Flatow E.L., Bigliani L.U. Shoulder arthroplasty for proximal humeral fractures. In: Friedman RJ, ed. Arthroplasty of the Shoulder. New York: Thieme. 1994; 183-93.
35. Pidhorz G., Saint-Cast Y., Grobost J., Boscher Y. Traumatismes severes recents de I' extremite superieure del; humerus. Apport de la prothese humerale. Annales Orthpediques de I' ouest. 1988; 20, 31-7.
36. Rose S.H., Melton J., Morrey B.F., Ilstrup D.M., Riggs B.L. Epidemiologic features of humeral fractures. Clin Orthop. 1982; 168, 24-30.
37. Stableforth P.G. Four-part fractures of the neck of the humerus. J Bone Joint Surg. Br 1984; 66, 104-8.
38. Steel F., Tomlinson J. The carrying angle in man. J Anat. 1989; 92, 315-7.
39. Tanner M.W., Cofield R.H. Prosthetic arthroplasty for fractures and fracture dislocations of the proximal humerus. Clin Orthop. 1983; 179, 116-28.
40. Tillet E., Smith M., Fulcher B.S., Shanklin B.S. Anatomic determination of humeral head retroversion: The relationship of the central axis of the humeral head to the bicipital groove. J Shoulder Elbow Surg. 1993; 2, 255-256.
41. Willems W.J., Lim T.E. Neer Arthroplasty for humeral fractures. Acta Orthop Scand. 1985; 56, 394-5.
42. Young T.B., Wallace W.A. Conservative treatment of fractures and fracture-dislocation of the upper end of the humerus. J Bone Joint Surg. Br 1985; 77, 353-7.
43. Zuckerman J.D., Cuomo F., Koval K.J. Proximal humeral replacement for complex fractures: indications and surgical technique. Instr Course Lect. 1997; 46, 7-14.
44. Zyto K., Kronberg M., Brostrom L.A. Shoulder function after displaced fractures of the proximal humerus. J Shoulder Elbow Surg. 1995; 4, 331.
45. Zyto K., Wallace A., Frostick S., Preston B.J. Outcome after hemiarthroplasty for three- and four- part fractures of the proximal humerus. J Shoulder Elbow Surg. 1998; 7, 85.

Mailing Address:
George M. Kondakis
6, 25th Martiou, Agios Basileios
68100 Alexandroupoli