 SHOULDER AND ELBOW

Reverse total shoulder replacement for nonunion of a fracture of the proximal humerus M. Zafra, P. Uceda, M. Flores, P. Carpintero From Instituto de Traumatologia Cordobes, Córdoba, Spain

Patients with pain and loss of shoulder function due to nonunion of a fracture of the proximal third of the humerus may benefit from reverse total shoulder replacement. This paper reports a prospective, multicentre study, involving three hospitals and three surgeons, of 35 patients (28 women, seven men) with a mean age of 69 years (46 to 83) who underwent a reverse total shoulder replacement for the treatment of nonunion of a fracture of the proximal humerus. Using Checchia’s classification, nine nonunions were type I, eight as type II, 12 as type III and six as type IV. The mean follow-up was 51 months (24 to 99). Post-operatively, the patients had a significant decrease in pain (p < 0.001), and a significant improvement in flexion, abduction, external rotation and Constant score (p < 0.001), but not in internal rotation. A total of nine complications were recorded in seven patients: six dislocations, one glenoid loosening in a patient who had previously suffered dislocation, one transitory paresis of the axillary nerve and one infection. Reverse total shoulder replacement may lead to a significant reduction in pain, improvement in function and a high degree of satisfaction. However, the rate of complications, particularly dislocation, was high. Cite this article: Bone Joint J 2014;96-B:1239–43.

 M. Zafra, MD, PhD, Orthopaedic Surgeon Instituto de Traumatología Cordobes, Arruzafilla Avenue S/ N 14011, Córdoba, Spain.  P. Uceda, MD, Orthopaedic Surgeon University Hospital Reina Sofía, Menendez Pidal Avenue, 14004, Córdoba, Spain.  M. Flores, MD, Orthopaedic Surgeon University Hospital Puerta del Mar, Ana de Viya avenue 21, 11009 Cádiz, Spain.  P. Carpintero, MD, PhD, Professor of Orthopaedic University Hospital Reina Sofia, Menendez Pidal Avenue, Cordoba, 14004, Spain. Correspondence should be sent to Professor P. Carpintero; email: [email protected] ©2014 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.96B9. 33157 $2.00 Bone Joint J 2014;96-B:1239–43. Received 13 September 2013; Accepted after revision 7 May 2014

Nonunion of a fracture of the proximal humerus is rare, but more common in twopart than three- or four-part fractures.1 These nonunions are usually treated surgically. Stable internal fixation may not be achievable because of a number of factors including osteoporosis, osteonecrosis, bone loss and stiffness of the shoulder, which may be due to previous surgery.1-5 Hemiarthroplasty is an option in elderly patients and good pain relief may be achieved.6 However, although adequate flexion can be restored, rotation remains limited with reduced satisfaction. This may be due to nonunion of the greater tuberosity and rotator cuff dysfunction.1,6 The good clinical and functional outcomes reported for Reverse Total Shoulder Replacement (RTSR) in other conditions of the shoulder suggest that it could also be used in patients with nonunion of a fracture of the proximal humerus, especially in elderly patients.7-11 There have been few reports detailing the use of the RTSR, which have included its application in the treatment of nonunion of fractures of the proximal third of the humerus,12,13 or specific series containing few patients with this indication.14 The aim of this multi-centre study was to evaluate prospectively the clinical results and

VOL. 96-B, No. 9, SEPTEMBER 2014

satisfaction in patients undergoing RTSR for nonunion of a fracture of the proximal third of the humerus.

Patients and Methods Between 2003 and 2011, 35 patients (28 women and seven men) with moderate to severe pain and loss of function of the shoulder, due to nonunion of a fracture of the proximal humerus, underwent RTSR. In total 20 patients with type I or II nonunions underwent osteosynthesis without RTSR, and were not included in the study. Surgery was performed at three hospitals by three different surgeons. The study had ethical approval. The mean follow-up was 51 months (24 to 99). Inclusion criteria were: (i) painful proximal humeral nonunion, Checchia et al15 type III or IV, or type I or II in patients aged > 65 years and (ii) younger patients not able to undergo internal fixation, due to the small size of the fragment of the humeral head, or previous unsuccessful internal fixation. The mean age was 69 years (46 to 83). The dominant arm was affected in 30 patients. Using Neer’s classification,16 the initial fracture was two-part in 14 patients, three-part in nine and four-part in 12 patients. In total, 25 patients were initially immobilised in a sling 1239

1240

M. ZAFRA, P. UCEDA, M. FLORES, P. CARPINTERO

Fig. 1a

Fig. 1b

Radiographs showing a) Checchia type IV nonunion treated with 90º angled plate with no consolidation achieved and b) reverse total shoulder replacement showing marked metaphyseal bone deficit.

for five weeks, and ten had previously undergone internal fixation. Three had been treated using a locked plate, three with an intramedullary nail and four with Kirchner (K-) wires. The mean time from the fracture to RTSR was 23 months (3 to 190). The following were measured pre- and post-operatively: pain was assessed using Cofield’s 5-point scale,17 satisfaction was recorded, and flexion, abduction, external rotation, and internal rotation in extension (according to the vertebral level reached by the patient’s thumb: to buttock, sacrum, L3, T12 or T8) were recorded.18 The clinical outcome was assessed using the Constant score.18 Anteroposterior, axillary and lateral radiographs were performed prior to surgery, immediately after surgery and at the final follow-up. Using Checchia et al classification,15 nine nonunions were type I, eight type II, 12 type III, and six type IV (Fig. 1). Radiological examination, undertaken by a consultant radiologist, included assessment of notching on the inferior rim of the scapular neck, the presence of radiolucent lines around the glenoid or humeral component, breakage of a screw, and loosening of the glenoid or humeral component. Radiolucent lines and notching were scored using the classification of Boileau et al.12 The delto-pectoral approach was used in all patients. In patients with Checchia type I and II nonunions the decision to use a locked screw-plate or RTSR was made at the time of surgery. However, in patients aged > 65 years, and those with small humeral head fragments, or who had undergone a previous unsuccessful osteosynthesis, RTSR was planned using the following points for guidance: a) no osteotomy of the greater tuberosity was required in order to avoid detaching the external rotator muscles; b) the humeral head could be divided at the level of the anatomical neck, leaving a collar of the surgical humeral neck; c) it was possible to introduce

the humeral stem to act as an internal guide in order to stabilise the nonunion (Fig. 2); d) the use of iliac crest bone graft could be avoided; e) patients were unable to externally rotate prior to the surgery, and f) it was possible to transfer the latissimus dorsi and teres major tendons to the metaphysis using the L’Episcopo procedure,19-21 which was required for three patients with gross limitation of internal rotation. A Delta III (Depuy, Warsaw, Indiana) RTSR was used in all patients, the glenoid component being placed as low as possible in order to avoid notching,13 and the humeral component was cemented into position in 10º to 20º of retroversion. A polyethylene insert of appropriate size and length was used to extend the metaphysis (20 +3mm; 11 +6mm; and 4 retentive humeral cups). A long head of biceps tenodesis was performed, except in patients with a prior spontaneous tear of the biceps tendon. Finally, fixation of the tendon of subscapularis to bone was performed to prevent anterior dislocation of the replacement. Post-operatively the arm was immobilised in a sling for four weeks, after which it was retained for a further four weeks. Movements of the hand and wrist were allowed at both time points. Passive movements were encouraged and gradually increased. Over the following six weeks, isometric and resistance movements were gradually added to the programme. Statistical analysis. For the purposes of descriptive analysis, means and standard deviations (SD) were calculated for quantitative variables, while absolute and relative frequencies were calculated for qualitative variables. Bivariate analysis was performed using Student’s t-test for independent data and the Mann–Whitney U test when normality criteria were not fulfilled or variables were ordinal. Paired data were analysed using Student’s t-test, and the Wilcoxon test was used when normality criteria were not fulfilled or variables were ordinal. The Kruskal–Wallis H test was used to compare more than two independent groups. THE BONE & JOINT JOURNAL

REVERSE TOTAL SHOULDER REPLACEMENT FOR NONUNION OF A FRACTURE OF THE PROXIMAL HUMERUS

1241

Fig. 2b

Fig. 2a

Fig. 2c Figure 2a - Radiograph showing Checchia type I non-union; Figure 2b - MRI showing cavitation of the head fragment; Figure 2c - Radiograph after reverse total shoulder replacement; the remaining humeral head bone was left in situ, so that following osteotomy of the humeral head, the point of insertion of the external rotator muscles remained intact.

All comparisons were bilateral, and differences were considered significant at a p-value of < 0.05. Data were processed and analysed using the SPSS v.17 software package (SPSS Inc., Chicago, Illinois).

Results At a mean follow-up of 51 months (24 to 99) all 35 patients treated with RTSR showed significant improvement in pain, flexion, abduction, external rotation and Constant scores. Internal rotation increased in most patients, though not significantly (Table I). A total of 23 patients (66%) reported feeling ‘much better’; ten (29%) felt ‘better’, and two (5%) felt ‘the same’. Of the two patients reporting no benefit, one had moderate pain and only 90º flexion and abduction and the other, who had undergone several operations prior to the RTSR, ending in the eventual removal of the prosthesis, due to infection, had moderate pain and only 45º flexion and abduction. The radiological findings are summarised in Table II. VOL. 96-B, No. 9, SEPTEMBER 2014

No statistically significant difference was observed between the Constant score at the end of follow-up and the variables, including the initial type of fracture, fracture treatment, nonunion and the presence of radiolucent lines or notching (Table III). Complications occurred in seven patients (20%). There was one transient axillary nerve palsy, which had resolved six months post-operatively. There were six dislocations (17%); four of which occurred in the immediate postoperative period. The polyethylene insert, two of which were retentive, was exchanged for a larger one to increase the offset and tension. Five years after operation, one of these patients developed loosening of the glenoid component which was revised and an iliac bone graft was added. There was improvement in pain, but not of function. Another patient suffered a dislocation one month postoperatively, had a failed attempted reduction under anaesthesia and required replacement of the polyethylene liner with a thicker one. The last dislocation occurred three

1242

M. ZAFRA, P. UCEDA, M. FLORES, P. CARPINTERO

Table I. Comparison of results before and after RTSR

Pain score Flexion (°) Abduction (°) External rotation (°) Constant score Internal Rotation

Pre-operative Mean, (SD, range)

Post-operative Mean, (SD, range)

p-value

Test

4.8 (SD 0.4, 3 to 5) 45.4 (SD 22.5, 10 to 80) 38.8 (SD 18.8, 10 to90) 4.55 (SD 5.2, 0 to 20) 23 (SD 4.2, 16 to 35) Median 1 (1 to 3)

1.77 (SD 0.7) 116.7º (SD 32.2, 50 to 110) 96.1º (SD 20.7, 70 to 160) 15.5º (SD 9.1, 0 to 40) 65.5 (SD 11, 35 to 85) Median 1 (1 to 5)

< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.1796

Wilcoxon Student´s t test Student´s t test Student´s t test Student´s t test Mann-Whitney

Table II. Radiological findings n (%) Glenoid radiolucent lines Zone 1 Notching Grade 1 Grade 2 Grade 3 Humeral radiolucent lines Zone 1 Zone 7 Zones 1 and 7 Zones 1,2,6,7 Zones 1,6,7 Zones 1,5,6,7

11(31.5) 21 (60) 9 (25.7) 11(31.4) 1 (2.85) 23 (65) 4 (11.4) 3 (8.5) 12 (34.2) 2 (5.7) 1 (2.85) 1 (2.85)

Table III. Relationship between the post-operative Constant score and clinical variables (Kruskal-Wallis) Constant value Initial fracture type Initial fracture treatment Type of nonunion Presence of radiolucent lines Presence of notching

p = 0.909 p = 0.839 p = 137 p = 0.352 p = 0.179

months post-operatively and the liner was replaced by a thicker, more retentive insert. Three years later, the same patient suffered another dislocation and Staphylococcus epidermidis was isolated, resulting in removal of the prosthesis and antibiotic treatment, leaving an excision arthroplasty.

Discussion Nonunion of a fracture of the proximal humerus can be treated in patients with good bone quality by internal fixation and iliac bone graft.22-23 However, results are unsatisfactory in around 50% of patients 2-5,24 but new locked-plating techniques, also using iliac bone graft, have led to improved results with rates of union of up to 80%.22,23,25 Nevertheless, in elderly patients with poor bone quality, in fractures with resorption of the tuberosities, when there is cavitation of the humeral head, or when internal fixation has failed, hemi- or total arthroplasty may considered. In young patients the results of hemiarthroplasty have not been shown to be different from those with internal fixation.6,9

Duquin et al,6 in a study of 67 patients with nonunion of a proximal humeral fracture, treated 54 with hemiarthroplasty and 13 with a total shoulder replacement. The mean active elevation and external rotation of the shoulder improved from 46º and 26º to 104º and 50º, respectively (p < 0.001). However, the overall results proved satisfactory in less than half of the patients. The presence of a rotator cuff tear at the time of the arthroplasty and nonunion of the tuberosities were associated with poor results. Anatomical or near anatomical union of the tuberosities was a significant factor in achieving greater active forward elevation (p < 0.02). RTSR has been increasingly used, over the last few years, especially in patients aged > 65 years, to treat a number of shoulder disorders,26 including rotator cuff deficiency,8-10,27,28 painful pseudoparesis due to irreparable rotator cuff dysfunction,11 failed hemiarthroplasty,9,29 post-traumatic arthritis, and acute fractures.7 Short- and medium-term clinical and functional outcomes have so far been good, with survival rates of 90% to 95%.8,26 After ten years, however, survival rates drop to a mean of 80%,30 particularly due to loosening of the glenoid component. Therefore, RTSR may be more appropriate for patients aged > 65 years. The literature on the use of RTSR for fracture complications has generally focused on generic problems,12,13 and there have been few reports describing the outcome following surgery for specific indications such as osteonecrosis and nonunion or mal-union, the latter being considerably more common.31 Clinical and functional outcomes are worse than those reported for other conditions, and complications such as infection and dislocation are more frequent.1,10,12 There are a variety of problems associated with shoulders that have already undergone surgery including stiffness of the joint, distortion of the anatomy and a damaged rotator cuff.6 Of the nine complications recorded in the present series, six were dislocations. Two of these patients subsequently had further complications with loosening of the glenoid component and deep infection. Meticulous re-attachment of the scapularis to the bone was performed whenever possible, cementing the humeral component into position in 10º to 20º retroversion, and attempted increase of the deltoid offset and tension by increasing the size and thickness of the polyethylene insert, were all undertaken to lessen the risk of dislocation. We did not use bone graft as others have proposed.14 THE BONE & JOINT JOURNAL

REVERSE TOTAL SHOULDER REPLACEMENT FOR NONUNION OF A FRACTURE OF THE PROXIMAL HUMERUS

The findings of this study suggest that RTSR may be used in patients with nonunion of a proximal humeral fracture, when the outcome following internal fixation is unlikely to be successful. The results would appear to be better than those reported for hemiarthroplasty although the mean follow-up was only 24 months, and there is a high risk of dislocation post-operatively. Therefore, caution should be exercised when opting for this procedure, particularly in elderly patients. The authors would like to acknowledge consultant radiologist R. Montero for his contribution towards the radiological examination in this study. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by S. P. H. Hughes and first proof edited by G. Scott.

References 1. Norris TR, Turner JA, Bovill D. Nonunion of the upper humerus: an analysis of the etiology and treatment in 28 cases. In: Post M, Morrey BF, Hawkins RJ, editors. Surgery of the shoulder. St. Louis: Mosby, 1990:63–67. 2. Duralde XA, Flatow EL, Pollock RG, et al. Operative treatment of nonunions of the surgical neck of the humerus. J Shoulder Elbow Surg 1996;5:169–180. 3. Galatz LM, Iannotti JP. Management of surgical neck nonunions. Orthop Clin North Am 2000;31:51–62. 4. Healy WL, Jupiter JP, Kristiansen TK, White RR. Nonunion of the proximal humerus: a review of 25 cases. J Orthop Trauma 1990;4:424–431. 5. Nayak NK, Schickendantz MS, Regan WD, Hawkins RJ. Operative treatment of nonunion of surgical neck fractures of the humerus. Clin Orthop Relat Res 1995;313:200–205.

1243

12. Boilaeu P, Watkinson D, Hatzidakis AM, Hovorka L. The Grammont reverse shoulder prosthesis: results in cuff tear arthritis, fracture sequelae and revision arthroplasty. J Shoulder Elbow Surg 2006;15:527–540. 13. Wall B, Nové-Josserand L, O’Connor DP, Edwards TB, Walch G. Reverse total shoulder arthroplasty: a review of results according to etiology. J Bone Joint Surg [Am] 2007;89-A:1476–1485. 14. Martinez AA, Bejarano C, Carbonel I, et al. The treatment of proximal humerus nonunions in older patients with the reverse shoulder arthroplasty. Injury 2012;43(Suppl):S3–S6. 15. Checchia SL, Doneux P. Miyazaki AN, et al. Classification of non-unions of the proximal humerus. Int Orthop 2000;24:217–220. 16. Neer CS 2nd. Displaced proximal humeral fractures. Part I: Classification and evaluation. J Bone Joint Surg [Am] 1970;52-A:1077–1089. 17. Cofield RH. Total shoulder arthroplasty with the Neer prosthesis. J Bone Joint Surg [Am] 1984;66-A:899–906. 18. Constant CR, Murley AH. A clinical method of functional assessment of the shoulder. Clin Orthop Relat Res 1987;214:160–164. 19. L’Episcopo JB. Tendon transplantation in obstetrical paralysis. Am J Surg 1934;25:122–125. 20. Boileau P, Chuinard C, Roussanne Y, Neyton L, Trojani C. Modified latissimus dorsi and teres major transfer through a single delto-pectoral approach for external rotation deficit of the shoulder: as an isolated procedure or with a reverse arthroplasty. J Shoulder Elbow Surg 2007;16:671–682. 21. Gerber C, Pennington SD, Lingenfelter EJ, Sukthankar A. Reverse Delta-III total shoulder replacement combined with latissimus dorsi transfer. J Bone Joint Surg [Am] 2007;89-A:940–947. 22. Cadet ER, Yin BY, Schulz B, Ahmad CS, Rosenwasser MP. Proximal humerus and humeral shaft nonunions. J Am Acad Orthop Surg 2013;21:538–547. 23. Spitzer AB, Davidovitch RI, Egos KA. Use of a hybrid locking plate for complex metaphyseal fractures and nonunions about the humerus. Injury 200940–240. 24. Scheck M. Surgical treatment of nonunions of the surgical neck of the humerus. Clin Orthop Relat Res 1982;167:255–259.

6. Duquin TR, Jacobson JA, Sanchez-Sotelo J, Sperling JW, Cofied RH. Unconstrained shoulder arthroplasty for treatment of proximal humeral nonunions. J Bone Joint Surg [Am] 2012;94-A:1610–1617.

25. Ring D, McKee MD, Perey BH, Jupiter JB. The use of a blade plate and autogenous cancellous bone graft in the treatment of ununited fractures of the proximal humerus. J Shoulder Elbow Surg 2001;10:501–507.

7. Cazeneuve JF, Cristofari DJ. [Grammont reversed prosthesis for acute complex fracture of the proximal humerus in an elderly population with 5 to 12 years followup]. Rev Chir Orthop Reparatrice Appar Mot 2006;92:543–548 [article in French].

26. Smith CD, Guyver P, Bunker TD. Indications for reverse shoulder replacement: a systematic review. J Bone Joint Surg [Br] 2012;94-B:577–583.

8. Frankle M, Levy JC, Pupello D, et al. The reverse shoulder prosthesis for glenohumeral arthritis associated with severe rotator cuff deficiency: a minimum two-year follow-up study of sixty patients surgical technique. J Bone Joint Surg [Am] 2006;88A(Suppl Pt. 2):178–190. 9. Levy J, Frankle M, Mighell M, Pupello D. The use of the reverse shoulder prosthesis for the treatment of failed hemiarthroplasty for proximal humeral fracture. J Bone Joint Surg [Am] 2007;89-A:292–300. 10. Sirveaux F, Favard L, Oudet D, et al. Grammont inverted total shoulder arthroplasty in the treatment of glenohumeral osteoarthritis with massive rupture of the cuff. J Bone Joint Surg [Br] 2004;86-B:388–395. 11. Werner CML, Steinmann PA, Gilbart M, Gerber C. Treatment of painful pseudoparesis due to irreparable rotator cuff dysfunction with the Delta III reverse-ball-andsocket total shoulder prosthesis. J Bone Joint Surg [Am] 2005;87-A:1476–1486.

VOL. 96-B, No. 9, SEPTEMBER 2014

27. Middleton C, Uri O, Phillips S, et al. A reverse shoulder arthroplasty with increased offset for the treatment of cuff-deficient shoulders with glenohumeral arthritis. Bone Joint J 2014;96-B:936–942. 28. Giuseffi SA1, Streubel P, Sperling J, Sanchez-Sotelo J. Short-stem uncemented primary reverse shoulder arthroplasty: clinical and radiological outcomes. Bone Joint J 2014;96-B:526–529. 29. Abdel MP, Hattrup SJ, Sperling JW, et al. Revision of an unstable hemiarthroplasty or anatomical total shoulder replacement using a reverse design prosthesis. Bone Joint J 2013;95-B:668–672. 30. Guery J, Favard L, Sirveaux F, et al. Reverse total shoulder arthroplasty. Survivorship analysis of eighty replacements followed for five to ten years. J Bone Joint Surg [Am] 2006;88-A:1742–1747. 31. Ogawa K, Matsumura N, Yoshida A. Modified osteotomy for symptomatic malunion of the humeral great tuberosity. J Orthop Trauma 2014;Mar 21 Epub.