Functional results and quality of life after shoulder girdle resections in musculoskeletal tumors János Kiss, MD,b Gergely Sztrinkai, MD,a Imre Antal, MD, PhD,a Jeno⬙ Kiss, MD, PhD,a and Miklós Szendro⬙i, MD, McS,a Budapest, Hungary
From 1981 to 2001, 91 shoulder resections were performed to treat shoulder girdle tumors (64 primary and 27 metastatic) in 90 patients (53 male and 37 female patients). The mean age was 34 years in patients with a primary tumor and 61 years in those with metastases. There were 7 partial scapulectomies, 13 total scapulectomies, 56 proximal humeral resections, 5 diaphyseal resections, 5 total shoulder girdle resections (Tikhoff-Linberg procedure), and 5 other procedures performed. Prosthesis implantation was carried out in 41 cases, autologous fibular transposition was done in 19, and massive homologous bone grafting was done in 4. Of the patients, 37 were clinically reviewed with a mean follow-up of 4.7 years (range, 1-20 years) by use of the recommendations of the Musculoskeletal Tumor Society for pain, function, position of hand, lifting ability, manual dexterity, and satisfaction. Nine patients were reviewed via a questionnaire and telephone interview. Twenty-six had died, and eighteen were lost to follow-up. The best results were achieved after partial scapulectomy and after humeral resection reconstructed with fibular transposition, when the function of the rotator cuff was preserved. After total scapulectomy and after humeral resection with the implantation of a tumor endoprosthesis, the function of the shoulder remained moderate because the rotator cuff was damaged. The overall satisfaction was generally good after all types of shoulder resections as a result of pain relief, preserved hand function, and improvement of psychological status. Patients can compensate extremely well by using the preserved joints and the contralateral upper limb; therefore, patient satisfaction does not rely on shoulder function alone. (J Shoulder Elbow Surg 2007;16:273-279.) From the aDepartment of Orthopaedics, Faculty of Medicine, Semmelweis University, and bDepartment of Orthopaedics and Trauma Surgery, Saint John’s Hospital. ⬙ Kiss, MD, PhD, Department of Orthopedics Reprint requests: Jeno and Trauma Surgery, Saint John’s Hospital, Dios Arok 1-3, 1125 Budapest, Hungary (E-mail: [email protected]
). Copyright © 2007 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2007/$32.00 doi:10.1016/j.jse.2006.08.011
B efore 1970, disarticulations or forequarter ampu-
tations of the upper extremity were performed for highly malignant sarcomas around the shoulder. Only in the case of a low-grade malignant sarcoma was limb preservation (Tikhoff-Linberg procedure) performed. In 1977, Marcove et al16 reported that the oncologic efficacy of limb preservations (modified Tikhoff-Linberg procedure) was similar to that of amputations. Subsequently, limb-preserving shoulder girdle resections for malignant or aggressive benign tumors became generally used. At our institution, we have performed upper limb amputation, disarticulation, or forequarter amputation in only 9 cases compared with 91 limb-preserving procedures. There are several successful methods for reconstruction of the shoulder girdle after tumor resections, but the function of the shoulder joint—with very few exceptions— can only be restored partially as a result of various degrees of muscle loss. The aim of our study was to evaluate the functional results achieved with shoulder girdle reconstruction procedures with special reference to the differences deriving from the extent of resection, the malignancy of the tumor, and the method of reconstruction. MATERIALS AND METHODS
From 1981 to 2001, 91 shoulder girdle resections were performed on 90 patients (53 male and 37 female patients) at the Department of Orthopaedics at Semmelweis University (Budapest, Hungary); these were done to treat primary tumors in 64 cases and metastases in 27 cases (1 patient had bilateral metastases). The histologic diagnosis of the primary and metastatic tumors is summarized in Table I. The mean age at the time of surgery was 33.6 years (range, 10-71 years) for patients with primary tumors and 61.2 years (range, 32-76 years) for those with metastases. Resections performed for musculoskeletal tumors in different regions of the human body can be compared by use of the system developed by Enneking et al8 and modified by the Musculoskeletal Tumor Society (MSTS). This system divides the skeleton into 8 regions (1, shoulder girdle; 2, elbow; 3, wrist/hand; 4, trunk; 5, pelvis; 6, hip/femur; 7, knee; and 8, ankle/foot). The shoulder girdle is divided into 5 subsections: 1, scapular body; 2, glenoid region; 3, proximal epiphysis of the humerus; 4, proximal metaphysis of the humerus; and 5, proximal two-thirds of the diaphysis of the humerus (Figure 1). In our series, the
Kiss et al
J Shoulder Elbow Surg May/June 2007
Table I Distribution of histologic diagnosis Diagnosis Primary tumor Chondrosarcoma Osteosarcoma Ewing’s sarcoma MFH Aneurysmal bone cyst Soft-tissue tumor Disappearing bone syndrome Metastasis Hypernephroma Adenocarcinoma bronchii Mamma carcinoma Adenocc. ventriculi Plasmocytoma Unknown primary tumor and/or adenocarcinoma (with unknown origin)
26 15 10 3 4 5 1 16 4 2 1 1 2
MFH, malignant fibrous histiocytoma.
Table II Viewpoints of assessment of functional results and quality of life recommended by MSTS Assessment based on MSTS recommendations
Pain (need for medication) Abduction Lifting ability (compared with contralateral limb) Fine hand movements (sensory, motor) Daily activities (work, hobby, sport, self-care) Satisfaction (with result of operation, with usability of limb) Total
0–5 0–5 0–5 0–5 0–5 0–5 0–30
and a questionnaire based on the MSTS recommendations9 (Table II).
RESULTS Of the patients, 37 (41%) were clinically reviewed and 9 (10%) were interviewed by phone or mail; 26 (29%) died, and 18 (20%) were lost to follow-up. The mean follow-up was 4.7 years (range, 1-20 years) for the entire group and was 5.6 years for patients with primary bone tumors and 2.9 years for those with metastases. Oncologic efficacy
Figure 1 Classification of shoulder girdle resections according to MSTS recommendations. S1, Partial scapulectomy; S1,2, total scapulectomy; S4, intercalary humeral resection; S3,4,(5), proximal humeral resection; S2,3,4,(5), extracapsular humeral resection; S1,2,3,4,(5), Tikhoff-Linberg procedure (total shoulder girdle resection); n, number of procedures carried out at our institution.
proximal end of the humerus was resected in 66 cases, scapulectomy was performed in 20, and total resection of the shoulder girdle (Tikhoff-Linberg procedure) was performed in 5 (Figure 1). Reconstruction after humeral resection was performed in 64 cases by use of tumor endoprostheses in 41, autologous fibular graft in 19, and massive homologous osteochondral graft in 4. Early in the study period, we carried out the operations without any replacement of the proximal humerus on 2 occasions, but instability of the limb resulted, which is an unacceptable condition for patients. We evaluated the functional results and quality of life of the patients based on physical examination, radiographs,
Of the 90 patients, 46 were alive and apparently disease-free at the time of follow-up at a mean of 4.7 years after surgery. Twenty-six patients died after a mean of 1.6 years after the intervention under known circumstances. According to the patient notes, all of these patients were satisfied with the functional results of their shoulder surgery. There were no major implant- or fibula-related complications among these patients. We did not succeed in obtaining any information about 18 additional patients during follow-up. The 2 groups of patients who died or were lost to follow-up are characterized by high proportions of patients with metastases or Ewing’s sarcoma. Of the 10 patients with Ewing’s sarcoma, 2 were symptom-free at the last checkup, with a mean of 7.9 years of follow-up; 5 died of lung or generalized metastases after a mean of 9 months after the operation. The history of the other 3 patients remained unknown. Of the 15 patients with osteosarcoma, 6 were alive and symptom-free after a mean of 5 years after the operation. Six died because of propagation of the disease at a mean of 1 year after the intervention. The remaining 3 patients were followed up for a mean of 3 years but were lost to follow-up thereafter. Of the 26 patients with chondrosarcoma, 20 were alive without symptoms at a mean of 7 years after surgery. Three died at a mean of 2 years after the
J Shoulder Elbow Surg Volume 16, Number 3
Kiss et al
operation, and the other three were followed up for only 3 to 4 years after the operation. The mean follow-up was 10 years for patients with aneurysmal bone cyst, and all were free from symptoms and complaints at the last checkup. We treated 26 patients (27 cases) for metastases. From this group, only 8 attended the follow-up. Ten patients died as a consequence of the original disease, after a mean of 22 months, and eight were lost to follow-up. Twenty-eight local recurrences of the primary or metastatic tumor were found in 13 patients. In 7 cases, the first procedure was performed at our department, which equates to an 8% recurrence rate in our patient population. These 7 patients underwent 15 additional operations performed for recurrence. In the remaining 6 cases, the indication for the operation was local recurrence after attempted treatment at another institution. After definitive resection and limbpreserving operations in our department, 13 new recurrences were found in 4 patients. This group was characterized by a high proportion of chondrosarcomas. In addition, recurrences were seen once each in cases of Ewing’s sarcoma, osteosarcoma, malignant fibrous histiocytoma (MFH), and malignant mesenchymoma. Eight of the patients were alive at follow-up, and five had died. Functional results and quality of life
Partial scapulectomy (S1 resection). We performed 7 partial scapulectomies (S1 resections) according to Enneking’s classification8 for the treatment of tumors (5 primary and 2 metastatic) affecting the scapular body alone (Figures 1 and 2). This resection is characterized by preservation of the glenoid fossa and scapular neck, as well as a part of the rotator cuff, which can provide movement and stability of the shoulder joint. The limb function of two such patients was evaluated (Table III). On the basis of these limited data, we can say that the function of the rotator cuff can be preserved best in this group, with an overall functional result of 88 % of mean MSTS score. We did not find any complications within this group of patients. Total scapulectomy (S1,2 resection). We removed the entire scapula in 13 cases. As a result of resection of the rotator cuff and the articular socket, the active movements of the shoulder were very limited (Figure 1). We managed to review 7 of these patients (Table III). Lack of pain and the preserved fine function of the hand are enough to rate the patients’ satisfaction as good (65% of mean MSTS score). We had no complications in this group. Replacement with autologous ﬁbula after S3,4,(5) resection. We performed reconstruction by using an autologous free fibular graft after proximal humeral resection in 19 cases (Figures 1 and 3). Of the 19
Figure 2 Function after partial scapulectomy.
patients, 10 presented to follow-up (Table III). In many cases, the indication for surgery was a benign tumor with aggressive behavior, in which resection was limited and the rotator cuff could be fixed to the biologically active bone. Thus, in addition to being pain-free and having good manual dexterity, this group has acceptable shoulder movement and lifting ability. Therefore, the quality of life and satisfaction of these patients are well above those of the other groups (83% of mean MSTS score). There were no septic complications in this group. The transplanted fibula fractured in 5 cases, which were treated with immobilization. In 3 cases, proximal migration of the fibular head occurred as a result of the pull of the deltoid muscle and insufficiency of the rotator cuff. In 3 cases, the fibular head partly resolved but did not lead to any complaints or require any treatment. Replacement with tumor endoprosthesis after S3,4,(5) resection. We implanted an endoprosthesis (modular tumor prosthesis system; Protetim, Hodmezovasarhely, Hungary) after resection of the proximal third of the humerus in 36 cases (Figures 1 and 4). The diagnosis was primary bone tumor in 22 patients and metastatic disease in 14. At the time of diagnosis,
Kiss et al
J Shoulder Elbow Surg May/June 2007
Table III Summarized results based on type of resections and method of reconstruction (MSTS) MSTS type of operation and replacement
S3,4 and autologous fibula
Pain Elevation Lifting ability Hand movements Daily activities Satisfaction Mean
4.5 5 4 5 4 4 4.42
4.36 1.07 2.14 5.0 3.14 3.71 3.24
4.89 2.67 3.44 5.00 4.17 4.67 4.14
S(2),3,4 and tumor prosthesis
S1,2,3,4,(5) (Tikhoff-Linberg) and prosthesis
4.68 1.07 2.07 4.79 3.07 4.07 3.29
4.5 0.5 2.0 5.0 2.5 4 3.17
Mean 4.61 1.81 2.65 4.92 3.45 4.19
Figure 3 Function after humeral resection and autologous fibular transposition.
most of the primary bone tumors were already in stage IIB according to Enneking7—that is, the tumors had broken through the cortex and expanded beyond the anatomic compartments, and their diameter was more than 5 cm. During the removal of the humeral head, the insertion of the rotator cuff was resected; thus, its function could not be restored, or at best, it could be attached to the metal prosthesis. Of these 36 cases, 14 were available for follow-up (Table III). Although active movement was poor, the operation was considered successful in terms of pain, manual dexterity, and patient satisfaction (66% of mean MSTS score). We had complications in 12 cases in this group. One revision was performed 7 years after the original operation because of proximal migration of the prosthesis and its dislocation above the clavicle. In this case, a prosthesis-allograft composite was used during the first procedure. During the revision, a monoblock shoulder prosthesis was implanted. One septic complication occurred after prosthetic
Figure 4 Humerus proximal end tumor prosthesis (right) and massive homologous graft (left), with patient shown in lying position.
implantation after resection of an osteosarcoma. Two-stage revision was performed. In the first stage, we removed the prosthesis. Six weeks later, a vascularized fibular graft was transplanted. In 10 cases, we had further complications that did not need surgical intervention, such as proximal migration resulting from pull of the deltoid muscle and the lack of rotator cuff, distal subluxation as a result of
J Shoulder Elbow Surg Volume 16, Number 3
an axillary nerve lesion, and transient radial nerve palsy in 1 case. Tikhoff-Linberg procedure (extracapsular shoulder girdle resection [S1,2,3,4,(5) resection]). Because of the location and size of a primary tumor, the process can reach the scapula from the humerus through the joint, or vice versa (Figure 1). Therefore, in 5 cases, we performed an extracapsular shoulder girdle resection. Of these 5 patients, 4 had a primary tumor and 1 had a metastasis. Two of these patients were available for follow-up (Table III). The procedure yielded an excellent result in terms of pain and manual dexterity, but in evaluating the remaining parameters, we obtained moderate results (63% of mean MSTS score). In 1 case, the prosthesis migrated cranially and a decubitus developed above the prosthetic head. During the revision, the metaphyseal part of the modular prosthesis had to be shortened. DISCUSSION In resecting highly malignant musculoskeletal tumors, the primary goal is oncologic eradication, and only if this is realized can the need to preserve the limb and its function, as well as esthetic concerns, become considerations. Currently, we are able to perform limb preservation in 75% to 80% of tumor resections, whereas amputation is performed in 20% to 25% of malignant tumors, either as a planned preoperative decision or as a result of an intraoperative decision. The primary aim of reconstruction is to create a functional limb that is acceptable to the patient. Among shoulder girdle resections, we obtained the best results after partial removal of the scapular body, and the worst results were obtained by use of total resection, similar to those of the TikhoffLinberg procedure. In the case of partial scapulectomies (S1 resection), the shoulder joint, the deltoid muscle, and a significant part of the rotator cuff remain functionally intact. This improves the efficacy of the operation by approximately 30%, and the patients have a better quality of life (Table III). Partial scapulectomy (S1 resection) and total scapulectomy (S1,2 resection) are efficient to the same degree in terms of pain, oncologic results, and complications. In a series of patients, O’Connor et al17 reported a result of 45% of mean MSTS score after total removal of the scapula, whereas after subtotal scapulectomies, they reported a result of 77% of mean MSTS score using the MSTS evaluation. Gibbons et al12 reported on 14 subtotal scapulectomies, with an overall result of 72% of mean MSTS score (range, 17%-100%). After omission of the worst 2 results (17% and 34%), which derived from complications after multiple recurrences, their result (89% of mean MSTS score) is comparable to ours (88% of mean MSTS score).
Kiss et al
Clarke et al3 reported on a series of 6 cases in which glenothoracic fusion was obtained in 1 stage after near-total scapulectomy. Functional range of movement was attained, but they noticed considerable reduction of power. The Tikhoff-Linberg procedure (S1,2,3,4) is the most extensive procedure but still preserves the upper limb. Previously, it was the only procedure used for the treatment of malignant scapular tumors, because it was considered more effective oncologically than total scapulectomy.5 Today, because of effective chemotherapy, its use is limited to malignant tumors that grow from either the scapula to the humeral head through the joint, or vice versa, but the lesion should not extend to the chest wall and must not involve major vessels or nerves.21 The prosthesis that is used to replace the proximal humerus and fixed to the lateral end of the clavicle functions only as a spacer but protects the main vascular and neurologic structures from shearing from the weight of the limb, and the stability provides pain control as well. Our finding of 60% can be explained by the total loss of shoulder movement. The results of total scapulectomy and the Tikhoff-Linberg procedure are nearly similar. Both procedures preserve fine hand movements that, together with a lack of pain and cessation of thoughts of the disease, provide an appropriate quality of life. The Tikhoff-Linberg procedure has been used since 1928,14 but there are only a few large-population surveys published. These report results similar to ours. Capanna et al2 reviewed 13 cases and reported good functional results in 4 cases, moderate results in 6, and poor results in 3. Voggenreiter et al21 evaluated 19 patients by using the MSTS recommendations and found an overall result of 72% of mean MSTS score (range, 58%-86%). After proximal humeral resections (S3,4 resection), prosthetic implantation was most commonly used in our series (41 cases). The primary indication was highly malignant bone tumors, where chemotherapy is also part of the treatment protocol. Solitary metastasis can be an optimal indication as well. The advantage of tumor prosthetic implantation is an easy reconstruction. The drawbacks are loosening, proximal migration, distal subluxation, periprosthetic fracture, and infection, but they are not very common. In 5 of our cases, the tumor invaded the shoulder joint; therefore, tumor eradication could only be achieved by a resection performed beyond the shoulder joint capsule, where the glenoid fossa was also removed (extracapsular [S2,3,4] resection). The result of these 5 extracapsular resections was 59%, and it was 68% of mean MSTS score of the intracapsular removals, suggesting that extracapsular resection does not lead to significantly worse limb function. In the great majority of operations for metastases, we implanted prostheses when the limb could be used within a short period of time, complications were less common in the early postoperative period, and the rehabilitation was shorter. In the last 2 years, we have improved stability by using a Trevira web (Implantcast GmbH, Buxtehude, Germany), which is sutured to the
Kiss et al
labrum, and the prosthesis head is rolled into the web. Since then, we have not experienced any proximal or distal migration. In 1 case, we performed a reoperation because of recurrence of disease 8 months after the original surgery, and we found connective tissue formation in the web. Summarizing our experiences with prosthesis implantation, we can state that, although shoulder joint movement remains grossly limited, the cessation of awareness of the disease and movement coordination that is relearned provide good satisfaction and quality of life. Infection occurred in 1 case, and proximal migration occurred in 2 cases. All of these complications were indications for revision. Subluxation of the prosthesis, which can be seen only radiographically, does not cause any subjective complaints. Other authors have published similar results. In a series of 134 patients, Bickels et al1 reported good and excellent results in 77% of patients with humeral endoprostheses after intra-articular resections. Rödl et al19 and Kumar et al13 achieved similar results, with scores of 79% by use of the MSTS evaluation. In the study of Ross et al,18 subluxation occurred in 16 of 24 patients and proximal dislocation occurred in 3 cases, for an overall frequency of 80%. They observed active flexion, extension, and abduction under 30°, but passive shoulder joint movement was almost completely preserved. Five patients had a mean 20% loss of elbow movement. De Wilde et al6 implanted a reverse Delta shoulder prosthesis (DePuy, Leeds, England) in 13 cases at 2 centers. At 1 institution, the resected part of the humerus was reimplanted after extracorporeal irradiation. It was fixed in an intramedullary manner by cementation of the humeral prosthetic component to facilitate restoration of humeral height. This graft allowed the reinsertion of muscles (deltoid, pectoralis, biceps), thus improving power generation postoperatively. At the other institution, no graft augmentation was used, except in 1 patient. The height of the humeral prosthetic component was assessed after resection of the tumor by measurement of the resected part. Both techniques resulted in a minimum active abduction of 60°, reaching 90° or more in most patients, and an MSTS score of 75.8%. According to both the literature data and our own results, it seems that after proximal humeral resections (S3,4 resection), the best results can be achieved by autogenous fibular transplantation, because strength and shoulder active movement can often be preserved (Table III). The primary indications are young age and benign bone tumor with aggressive behavior or low-grade malignant bone tumors, where chemotherapy is not needed and long survival is expected. The procedure requires at least partial preservation of the capsule, the rotator cuff, and the axillary nerve. In young patients, fibular transplantation can be performed for highly malignant tumors, but bony integration is largely prolonged by chemotherapy.10 Frac-
J Shoulder Elbow Surg May/June 2007
Table IV Complications Complications Postoperative and operative complications Fracture of fibular graft Prosthesis migration Infection Radial nerve palsy Subluxation (only radiologic sign) Oncologic complications Relapse
8 2 1 1 18
cases cases case case cases
tures may occur during bone remodeling even without adjuvant therapy. Fibular transposition is not recommended in case of metastases because of the shorter life expectancy. In case of fractures, which are considered to be drawbacks for fibular transplantations, after temporary immobilization, limb function remains nearly similar to that in those without complications. To reduce the number of fractures, vascularized fibular transplantation is recommended when resections exceed 6 cm.20,23 Because this is considered technically difficult, the method is not widespread. As for complications, the highest number occurred in cases of fibular transposition (Table IV). In 5 of our 19 cases, we found a fracture or crack of the transplanted bone at a mean of 11 months after surgery. Fibular head subluxation can occur as a result of a lack of stability or as a consequence of the pull of the deltoid muscle, as we found in 3 cases. Absorption of the fibular head may be seen radiographically, but it causes no symptoms. We had 1 radial nerve palsy that was successfully treated with observation. We had no septic complications in our series. In the literature, fibular transposition is recommended for shoulder arthrodesis,17 or the fibula is added to the glenoid fossa to reconstruct the joint. Wada et al22 used a vascularized fibular graft suspended on the clavicle in 8 cases. Although the patients received adjuvant therapy, all of the grafts integrated. In most of their cases, the tumor was removed extracapsularly together with the rotator cuff. They lost only 1 patient during the 5-year follow-up. The mean functional value was 79%, similar to ours (81%). Their work verifies that fibular transposition is useful in young patients, even for highly malignant tumors.22 They reported a lower rate of fractures (12%) with the use of vascularized fibular grafts than what we experienced (35%). For replacement of the resected humerus, the use of massive homologous osteochondral graft has limited indications (Figure 4). The lack of an appropriate bone bank is a limiting factor in Hungary. According to Getty and Peabody,11 a further problem is the high risk of infection. Between 1971 and
J Shoulder Elbow Surg Volume 16, Number 3
1993, Mankin et al15 implanted massive osteochondral grafts in 718 cases after limb resections. The proximal of the humerus was replaced in 55 cases, and the distal end was replaced in 14. They found an overall infection rate of 11%. Other frequent complications are collapse in the subchondral region, graft fracture, and degenerative lesions in the joint. Despite all of these problems, good function can be achieved with appropriate indications, especially in older patients and those with metastases. Years after the initial good results, however, a slow worsening of function is reported.4,15 O’Connor et al17 reported on 8 cases in which the functional result was 71 % of mean MSTS score. According to our study, an overall functional result of 72% of mean MSTS score could be achieved with homograft implantation, which is better than after prosthetic implantation but worse than after fibular transposition. Life expectancy is tumor-dependent. With metastases, it depends on the primary tumor but is usually short. Because of the limited number of cases, we did not perform any statistical analysis for the various primary tumors. With modern limb-preserving operations, there are several possibilities for reconstruction. The choice of method is determined based on tumor location, the primary or secondary character, the degree of malignancy, the type of adjuvant treatment (chemotherapy, irradiation), and the character of structures to be preserved. On the basis of our findings, we can state that each resection and replacement method is effective in reducing pain and preserving fine hand movements. However, quality of life and satisfaction are determined not only by these two factors but also by the quality of the upper limb’s active movement. The functional loss of 1 of the 6 major joints of the upper limb may be well compensated for by the function of the other 5 joints. In older patients with metastases or in younger patients with osteosarcoma or Ewing’s sarcoma, a greater awareness of the disease is typical because of prolonged therapy. Radical removal of the tumor is considered by the patients as a step on the way to healing, producing significant physical and emotional improvement. This can be used to shorten healing time and to improve the results of rehabilitation after the operation. The mean satisfaction rating of 4.2 points (84%) indicates that most of the patients adapted successfully to their new circumstances and accepted both the results of the operation and the possible associated limitations. They can live a full life, and most of them are satisfied. REFERENCES
1. Bickels J, Wittig JC, Kollender Y, Kellar-Graney K, Meller I, Malawer MM. Limb-sparing resections of the shoulder girdle. J Am Coll Surg 2002;194:422-35.
Kiss et al
2. Capanna E, Giunti A, Biagini R, Ferruzzi A. Modular endoprosthesis for humerus and Tikhoff-Linberg resection. In: Yamamuro T, editor. New developments for limb salvage in musculoskeletal tumors. Tokyo: Springer; 1989. p. 547-55. 3. Clarke A, Dewnany G, Neumann L, Wallace WA. Glenothoracic fusion. An adjunct to radical scapulectomy. J Bone Joint Surg Br 2004;86:531-5. 4. Damron TA, Rock MG, O’Connor MI. Functional laboratory assessment after oncologic shoulder joint resections. Clin Orthop Relat Res 1998:124-34. 5. Das Gupta TK. Scapulectomy: indication and technique. Surgery 1970;67:601-6. 6. De Wilde L, Sys G, Julien Y, Van Ovost E, Poffyn B, Trouilloud P. The reversed Delta shoulder prosthesis in reconstruction of the proximal humerus after tumour resection. Acta Orthop Belg 2003;69:495-500. 7. Enneking WF. A system of staging musculoskeletal neoplasms. Clin Orthop Relat Res 1986:9-24. 8. Enneking WF, Dunham W, Gebhardt M, Malawar M, Pritchard D. A system for the classification of skeletal resections. Chir Org Mov 1990;75(Suppl 1):217-40. 9. Enneking WF, Dunham W, Gebhardt MC, Malawar M, Pritchard DJ. A system for the functional evaluation of reconstructive procedures after surgical treatment of tumors of the musculoskeletal system. Clin Orthop Relat Res 1993:241-6. 10. Friedlaender GE, Tross RB, Doganis AC, Kirkwood JM, Baron R. Effects of chemotherapeutic agents on bone. I. Short-term methotrexate and doxorubicin (adriamycin) treatment in a rat model. J Bone Joint Surg Am 1984;66:602-7. 11. Getty P, Peabody T. Complication and functional outcomes of reconstruction with an osteoarticular allograft after intra-articular resection of the proximal aspect of the humerus. J Bone Joint Surg Am 1999;81:1138-46. 12. Gibbons CLMH, Bell RS, Wunder JS. Function after subtotal scapulectomy for neoplasm of bone and soft tissue. J Bone Joint Surg Br 1998;80:38-42. 13. Kumar D, Grimer RJ, Abudu A, Carter SR, Tillman RM. Endoprosthetic replacement of the proximal humerus. Long-term results. J Bone Joint Surg Br 2003;85:717-22. 14. Linberg BE. Interscapulo-thoracic resection for malignant tumors of the shoulder joint region. J Bone Joint Surg 1928;10:344-9. 15. Mankin HJ, Gebhardt MC, Jennings LC, Springfield DS, Tomford WW. Long-term results of allograft replacement in the management of bone tumors. Clin Orthop Relat Res 1996:86-97. 16. Marcove RC, Lewis MM, Huvos AG. En bloc upper humeral interscapulothoracic resection. The Tikhoff-Linberg procedure. Clin Orthop Relat Res 1977:219-28. 17. O’Connor MI, Sim FH, Chao EYS. Limb salvage for neoplasms of the shoulder girdle: intermediate reconstructive and functional results. J Bone Joint Surg Am 1996;78:1872-88. 18. Ross AC, Wilson JN, Scales JT. Endoprosthetic replacement of the proximal humerus. J Bone Joint Surg Br 1987;69:656-61. 19. Rödl RW, Gosheger G, Gebert C, Lindner N, Ozaki T, Winkelmann W. Reconstruction of the humerus after wide resection of tumours. J Bone Joint Surg Br 2002;84:1004-8. 20. Shaffer JW, Field GA, Goldberg VM, Davy DT. Fate of vascularized and nonvascularized autografts. Clin Orthop Relat Res 1985:32-43. 21. Voggenreiter G, Assenmacher S, Schmit-Neuerburg KP. TikhoffLinberg procedure for bone and soft tissue tumors of the shoulder girdle. Arch Surg 1999;134:252-7. 22. Wada T, Usui M, Isu K. Reconstruction and limb salvage after resection for malignant bone tumour of the proximal humerus: a sling procedure using a free vascularised fibular graft. J Bone Joint Surg Br 1999;81:808-13. 23. Weiland AJ, Moore JR, Daniel RK. Vascularized bone autografts: experience with 41 cases. Clin Orthop Relat Res 1983:87-95.