Sternal Resection and Reconstruction for Primary Malignant Tumors

ORIGINAL ARTICLES: Sternal Resection and Reconstruction for Primary Malignant Tumors Alain R. Chapelier, MD, PhD, Marie-Christine Missana, MD, Benoit Couturaud, MD, Elie Fadel, MD, PhD, Dominique Fabre, MD, Sacha Mussot, MD, Pierre Pouillart, MD, and Philippe G. Dartevelle, MD Department of Thoracic and Vascular Surgery and Heart-Lung Transplantation, Hôpital Marie-Lannelongue, Le Plessis Robinson, Department of Plastic Surgery, Institut Gustave Roussy, Villejuif, Departments of Plastic Surgery and Oncology, Institut Curie, Paris, France Primary malignant sternal tumors (PMST) are uncommon and most of them are sarcomas, arising, either from the bone, or soft tissues of the sternum [1, 2]. Radical resection can offer a definitive cure, but the surgical management may be difficult because of the local aggressiveness of these tumors and a high recurrence rate. Improvement of reconstruction techniques with musculocutaneous flaps [3, 4] has made coverage of wide sternal defects reliable, especially after extensive skin excision. This study reports our experience with sternectomies for PMST. Patients and Methods Presented at the Thirty-ninth Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 31 Feb 2, 2003. Address reprint requests to Dr Chapelier, Department of Thoracic and Vascular Surgery and Heart-Lung Transplantation, Hôpital Marie- Lannelongue, 133 Ave de la Résistance, 92350 Le Plessis Robinson, France; e-mail: alain.chapelier@ccml.com. Background. Primary malignant sternal tumors (PMST) are locally aggressive and their optimal surgical management still continues to evolve. Methods. From 1986 to 2002, 38 patients (25 females/13 males) underwent radical resection of PMST. This series included 33 sarcomas, 17 of which had been radiationinduced, 3 hematologic tumors, and 2 carcinomas. Seventeen were high-grade tumors. Nine patients had received preoperative chemotherapy. Twelve patients required extensive skin excision. Eight total, seven subtotal, and 23 partial sternectomies were performed. Resection was extended to the anterior chest-wall in 4 patients, lung in 4, brachiocephalic vein in 3, superior vena cava in 2, and pericardium in 1. In 36 patients, chest wall stability was obtained by Marlex (n 21) or Vicryl (n 2) mesh and polytetrafluoroethylene patch (n 13), with methylmethacrylate reinforcement in 12 patients. Soft tissue coverage was done by the pectoralis major muscles with skin advancement in 25 patients, a myocutaneous flap in 11, a breast transposition in 1, and a skin flap in 1. Omentoplasty was performed in 3 patients. Results. One patient died from pneumonia. Two patients needed a tracheostomy after total sternectomy. No flap-related complication was observed. Four local septic complications required removal of the composite prosthesis with reoperations. Local recurrence occurred in 9 patients, 7 patients having a repeat resection. Metastases developed in eight. The 5-year overall and disease-free survival was 66% and 53%, respectively. The histologic grade of sarcomas was a survival predictor (high grade versus others p 0.035). Conclusions. Wide resection of PMST is necessary to minimize local recurrence. Large sternal defects are safely reconstructed with a musculocutaneous flap. We suggest that the use of methylmethacrylate should be limited to reconstruction after total sternectomy. (Ann Thorac Surg 2004;77:1001 7) 2004 by The Society of Thoracic Surgeons Between 1986 and 2002, 38 consecutive patients underwent curative resection of PMST in our department at the Marie Lannelongue Surgical Center. There were 25 females and 13 males with a median age of 58 years old (range 18 76 years old). The tumor involved the manubrium in 19 patients, the sternal body in 15 patients, and included both parts of the sternum in four cases. Six patients had had previous resection before referral. There were 33 sarcomas: their histologic types are depicted in Table 1. Histologic grading of soft tissue sarcomas was stated according to the international system [5]. Seventeen patients had developed a radiation-induced sarcoma (RIS) of the sternum: 15 females had a history of primary breast cancer and 2 patients had been treated for a primary Hodgkin s disease. The median interval between the onset of radiotherapy and the occurrence of the radiation-induced sternal tumor was 13 years (range 7 to 25 years). This series also included two carcinomas and three hematologic tumors. In this latter group, there were one solitary plasmacytoma and two cases of Hodgkin lymphoma with a persistent sternal tumor after medical treatment. Preoperative Assessment The precise location of the tumor and the extent of sternal and chest wall involvement were determined by conventional tomography, computed tomography, and magnetic resonance imaging in recent patients. Possible invasion of lung, pericardium, brachiocephalic vein, and 2004 by The Society of Thoracic Surgeons 0003-4975/04/$30.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2003.08.053

1002 CHAPELIER ET AL Ann Thorac Surg STERNAL RESECTION FOR MALIGNANT TUMORS 2004;77:1001 7 Table 1. Histologic Type of Primary Sternal Tumors Type Number Sarcomas 33 Chondrosarcoma 9 Fibrosarcoma 7 MFH 6 Osteosarcoma 4 Undifferenciated 3 Leiomyosarcoma 3 PNET 1 Hematologic tumors 3 Hodgkin s lymphoma 2 Plasmacytoma 1 Carcinomas 2 PNET peripheral neuroec- MFH malignant fibrous histiocytoma; todermic tumor. superior vena cava were assessed by computed tomography and cavography. A biopsy was performed in 23 of 32 patients undergoing first surgical resection. Every patient was routinely evaluated with cardiopulmonary tests and completely investigated to eliminate an extrathoracic metastatic lesion. Seven patients received preoperative chemotherapy for a high-grade sarcoma. Preoperative radiotherapy was administered to three patients. The closure of wide sternal defects was carefully planned by thoracic and plastic surgeons. Operative Procedure The first step was the skin excision. When the skin and overlying soft tissues were not involved, a vertical elliptical incision encompassing the biopsy site was done. In the case of invaded skin and subcutaneous tissues, previously irradiated tissues and scars, a wide resection was carried out with a margin of at least 3 cm of macroscopically normal surrounding tissues. Sternal resection included wide resection of the affected part of the sternum with a bone margin of 3 cm. Resection was started over the costal margins, sparing the unaffected lateral part of the pectoralis major muscles. A total sternectomy including the internal third of clavicles was undertaken for tumors involving both parts of the bone, for large tumors of the mid sternum and in the case of a primitive neuroectodermal tumor. A subtotal sternectomy was performed, sparing the uppermost 2 cm of the manubrium and clavicles in the case of tumors invading the sternal body (Fig 1). A partial sternectomy was carried out for tumors limited to the manubrium or the lower part of the sternum. Tumor extension into the chest cavity was evaluated. Lung and involved mediastinal structures were excised en bloc. Resection Twelve patients required wide skin excision, because the skin was invaded or ulcerated by the tumor in seven cases, and was inflammatory in five. All patients underwent a radical resection of the tumor and involved structures. Eight total, 7 subtotal, and 23 partial sternectomies were performed. Resection was extended to the anterior chest-wall in 4 patients, lung in 4 patients, brachiocephalic vein in 3 patients, superior vena cava with polytetrafluoroethylene (PTFE) revascularization in 2 patients, and pericardium in 1 patient. A mastectomy was associated in one case (Table 2). Margins were macroscopically tumor-free in all patients. Sternal Reconstruction Chest wall stability was obtained by prosthetic material sutured to the edges of the sternal defect with nonabsorbable sutures in 36 patients. We used a Marlex mesh (Bard Inc, Murray Hill, NJ) in 21 patients, a polytetrafluoroethylene patch (W. L. Gore & Assoc, Flagstaff, AZ) in 13 patients, and a polyglactin mesh (Ethnor Inc, Summerville, NJ) in 2 patients. Chest wall rigidity was reinforced with methylmethacrylate in 12 patients after a total or subtotal sternectomy. In 7 patients, the methylmethacrylate was spread on the mesh and anchored to the surrounding tissues; in 5 patients it was introduced into two 28F silicone chest tubes, which were interposed and fixed to the edges of the ribs laterally. Soft tissue coverage of the sternal defect was done by the pectoralis major (PM) muscles with skin advancement in 25 patients and by a musculocutaneous flap in 11 patients of major skin and subcutaneous excision. A unilateral PM musculocutaneous flap was done in 3 men and a bilateral PM flap was performed in 1 man after a wide skin excision extending to the cervical area. Sternal reconstruction was carried out with a latissimus dorsi (LD) musculocutaneous flap in 5 women (Fig 2). In two other patients of recurrent PMST, it was possible to preserve the previously done LD flap. A breast flap was used in an elderly woman who had undergone prior mastectomy. Omentoplasty was performed in 2 patients with a recurrent manubrial tumor, and in 1 patient after a total sternectomy for a RIS (Table 3). Postoperative Management Adjuvant chemotherapy was given to 7 patients mostly based on the high grading of the tumor. Statistical Analysis Overall and disease-free survivals were calculated from the date of operation to death, or date of last follow-up for those patients who remained alive and to date of local or systemic recurrence. They were estimated by the Kaplan-Maier product-limit method [6]. Results Mortality One patient died after partial sternectomy from bacterial pneumonia with septicemia. A paradoxical motion occurred in 2 patients after total sternectomy and rigid replacement of the sternum, and in 1 patient with a subtotal sternectomy, which was extended to the anterior chest-wall. These patients required a prolonged ventilla-

Ann Thorac Surg CHAPELIER ET AL 2004;77:1001 7 STERNAL RESECTION FOR MALIGNANT TUMORS 1003 Fig 1. Leiomyosarcoma of the sternal body in a 49-year-old female. (A) Sagittal magnetic resonance imaging scan. (B) Subtotal sternectomy sparing the uppermost of the manubrium. (C) Reconstruction with pectoralis major muscles translation and skin advancement. (D) Final result. tory support and two needed a tracheostomy: one died 3-months later from respiratory insufficiency. Histology Soft tissue and bone margins were tumor free in all patients. Among sarcomas, the grade was high in 17 patients, intermediate in 9 patients, and low in 7 patients. Local Complications No flap-related complication was observed. Major septic complications occurred in 4 patients with methylmethacrylate reinforcement which required removal of the composite prosthesis (marlex mesh n 3, PTFE n 1). Good results with complete coverage of the sternal defect by the flap were obtained in two patients. In one patient with a marlex methacrylate sheet, the bilateral PM myocutaneous flap was preserved but an upper cutaneous dehiscence resulted in hemorrhage by ulceration of the superior vena cava. It was successfully managed by venous reconstruction and coverage with a myocutaneous LD flap. In one patient with marlex mesh reinforced with tubes, several reoperations including omentoplasty were required, because of residual prosthetic fragments. One patient was reoperated for a pneumothorax, and a

1004 CHAPELIER ET AL Ann Thorac Surg STERNAL RESECTION FOR MALIGNANT TUMORS 2004;77:1001 7 Table 2. Extent of Resection Major Skin Excision 12 Sternectomy 38 Total 8 Subtotal 7 Partial 23 Upper 13 Lower 10 Other involved structures Anterior chest wall 4 SVC system 5 Brachiocephalic vein 3 SVC trunk PTFE replacement 2 Lung 4 Pericardium 1 Breast 1 PTFE polytetrafluorethylene; minor cutaneous dehiscence occurred in one case. One delayed infection was observed after 12 months and was treated by the removal of the marlex mesh with preservation of the LD myocutaneous flap. Local and Systemic Recurrence Local recurrence occurred in 9 patients, with a mean interval of 11 months and led to death in two. Six of them had a RIS. The margins were histologically read as close in 3 patients. Seven patients underwent a repeat resection. Two had a bony recurrence (osteosarcoma n 1, undifferenciated sarcoma n 1) and needed a completion sternectomy; 5 patients had a limited soft tissue recurrence. Eight patients developed a systemic recurrence; of these, 2 patients had a single pulmonary metastase treated by a wedge resection, 4 multiple pulmonary lesions, 1 a liver lesion, and 1 an adrenal metastase. The patient with a solitary plasmacytoma developed subsequently a multiple myeloma. Survival The overall 5-year survival was 66% (Fig 3). The diseasefree survival was 59% at 5 years and 39% at 10 years. Histologic grade of sarcomas was a survival predictor (high versus others p 0.035; Fig 4). Comment SVC superior vena cava. Our results confirm that radical resection of PMST and satisfactory reconstruction of wide sternal defects can be performed with a low mortality. The presence of a PMST is based on clinical findings. Imaging studies assess the tumor extension and any infiltration of adjacent structures; magnetic resonance imaging sagittal views can compliment conventional axial computed tomography scanning. PMST are mostly sarcomas, including several histologic types and are rarely hematologic tumors (lymphomas and plasmacytomas), which may present as a localized sternal mass [1]. The incidence of sarcomas arising in a previously irradiated sternal field represents 50% of our patients in this series. The long-term risk for RIS of the chest-wall in patients treated with radiotherapy for breast cancer and Hodgkin s disease has been already reported [7 9]. Early in our experience we performed resection of PMST without prior biopsy in some patients. We believe, as others [1, 10], that a histologic diagnosis is now required to identify patients with a high-grade tumor whose primary treatment should be chemotherapy. Most of our patients had an incisional biopsy, which is preferable to a needle biopsy for a highly accurate diagnosis. Wide excision remains the key to success for local control of PMST. The skin excision must be large in the case of ulceration, previous scars, and if the tumor involves subcutaneous tissues. RIS of the sternum frequently requires a wide cutaneous excision including previously irradiated surrounding tissues. Sternal resection with clear margins of at least 3 cm is advocated to minimize the risk of local recurrence [11, 12]. Total sternectomy must be done if the entire sternum is invaded by the tumor and in the case of a PNET. In the Fig 2. Radiation-induced malignant fibrous histiocytoma of the sternum in a 67-year-old female. (A) Axial computed tomographic scan. (B) Soft tissue reconstruction with a latissimus dorsi myocutaneous flap after partial sternectomy associated with a mastectomy.

Ann Thorac Surg CHAPELIER ET AL 2004;77:1001 7 STERNAL RESECTION FOR MALIGNANT TUMORS 1005 Table 3. Sternal Reconstruction Prosthetic Material Number 36 Soft Tissue Coverage Number 38 Marlex mesh 21 PM translation with skin advancement 25 PTFE patch 13 Musculocutaneus flap 11 Polyglactin mesh 2 PM unilateral 3 PM bilateral 1 LD 7 Methylmethacrylate reinforcement 12 Breast flap 1 Mesh 7 Skin flap 1 Tubes 5 Associated omentoplasty 3 LD latissimus dorsi muscle; PM pectoralis major muscle; PTFE polytetrafluorethylene. other patients, a subtotal resection, sparing a small part of the intact manubrium with clavicles or a partial sternectomy, is carried out. Any involved structure, such as lung, mediastinal vessels, and pericardium must be en-bloc excised. Invasion of brachiocephalic veins is managed by ligation and excision; involvement of the superior vena cava may require resection and revascularization as we have done in 2 patients. Chest-wall stability after a wide sternectomy can be obtained with different prosthetic materials such as Marlex mesh or PTFE patch [13]. The respective advantages of each material have already been outlined [14, 15]. In our experience, an adequate skeletal stability could be obtained without methacrylate, in every case of partial sternectomy. We now prefer the use of a 2-mm PTFE patch, which appears to be less inclined to infection and can be easily fixed under tension to the edges of the sternal defect. However, after total sternectomy, a rigid prosthetic replacement with methylmethacrylate is recommended, in order to limit paradoxical motion. The methylmethacrylate is usually spread between two layers of Marlex mesh, as proposed by several authors [15, 16], or spread on a PTFE patch. In recent years we also used a technique of reconstruction with tubes filled with methylmethacrylate in order to limit the amount of prosthetic material and the potential risk of infection. Four patients in this series developed infection after reconstruction using both techniques of methacrylate reinforcement and required the removal of the composite prosthetic material. At that time, fibrous tissues yielded sufficient stability to the sternum, but 2 patients required reoperations. Consequently, the risks of infection already mentioned by others [1, 2] should be balanced against the functional and protective benefits of rigid prosthetic sternal reconstruction. Soft tissue coverage is the essential step of sternal reconstruction; it must be planned and is obtained with muscle or musculocutaneous flaps [3, 4, 17, 18]. PMs are the most frequently selected muscles transferred to the sternal defect with skin advancement on the front line. If necessary, section of the humeral insertion can afford a better rotation with no tension on the vascular pedicle. After a wide skin and subcutaneous excision, reconstruction of sternal and soft tissue defects can be achieved by the combined efforts of the thoracic and plastic surgeons. Musculocutaneous flaps afford an airtight closure and a well vascularized reliable sternal reconstruction. A PM musculocutaneous flap is usually done in men; it can also be used bilaterally as we have done in 1 patient with a large sternal defect extended to the neck. In women, a LD musculocutaneous flap is commonly used with skin closure or grafting of the donor site [19]. Omentoplasty can be interposed in some difficult patients, such as resection in irradiated fields and for recurrent tumors [20]. Survival in completely resected patients is related to the grade of sarcomas as previously reported by several authors [1, 21, 22]. High-grade tumors were also an adverse factor for long-term survival in this series. Local recurrence and pulmonary metastases were frequent Fig 3. Overall survival. Fig 4. Survival by histologic grade in patients with sarcomas (number 33).

1006 CHAPELIER ET AL Ann Thorac Surg STERNAL RESECTION FOR MALIGNANT TUMORS 2004;77:1001 7 failures in the follow-up. Local recurrences in this series were mostly observed in soft tissue. It outlines the importance of wide excision of skin and soft tissue, especially in the case of RIS of the sternum where the interpretation of frozen sections is particularly difficult. However, in selected patients with limited local recurrence or single metastase, a repeat resection was possible with a satisfactory long-term survival. Treatment of primary sternal lymphoma is classically nonsurgical but resection can be proposed if the diagnosis is in doubt or in patients with persistent local disease after medical treatment [1]. The benefit of sternal resection for a solitary plasmacytoma is jeopardized by the frequent development of multiple myeloma in the follow-up [23], which occurred in 1 patient. In conclusion, wide resection of PMST is necessary to minimize local recurrence. Large sternal defects are safely reconstructed with a musculocutaneous flap, especially in the patient with radiation-induced sarcomas. Because of the risk of infection, rigid replacement of the sternum with methylmethacrylate mesh should be limited to reconstruction after total sternectomy. References 1. Martini N, Huvos AG, Burt ME, et al. Predictors of survival in malignant tumors of the sternum. J Thorac Cardiovasc Surg 1996;111:96 106. 2. Incarbone M, Nava M, Lequaglie C, Ravasi G, Pastorino U. Sternal resection for primary or secondary tumors. J Thorac Cardiovasc Surg 1997;114:93 9. 3. Tobin GR, Mavroudis C, Howe WR, Gray LA. Reconstruction of complex thoracic defects with myocutaneous and muscle flaps. J Thorac Cardiovasc Surg 1983;85:219 28. 4. Morgan RF, Edgerton MT, Wanebo HJ, Daniel TM, Spotnitz WD, Kron IL. Reconstruction of full thickness chest wall defects. Ann Surg 1988;207:707 16. 5. Coindre JM, Trojani M, Contesso G, et al. Reproducibility of a histopathologic grading system for adult soft tissue sarcoma. Cancer 1986;58:306 9. 6. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457 81. 7. Brady MS, Garfein CF, Petrek JA, Brennan MF. Posttreatment sarcoma in breast cancer patients. Ann Surg Oncol 1994;1:66 72. 8. Chapelier A, Bacha E, De Montpreville VT, et al. Radical resection of radiation-induced sarcoma of the chest-wall: report of 15 cases. Ann Thorac Surg 1997;63:214 9. 9. Souba WW, McKenna RJ, Meis J, Benjamin R, Raymond AK, Mountain CR. Radiation-incuced sarcomas of the chest wall. Cancer 1986;57:610 15. 10. Soysal O, Walsh GL, Nesbitt JC, McMurtrey MJ, Roth JA, Putman JB Jr. Resection of sternal tumors: extent, reconstruction, and survival. Ann Thorac Surg 1995;60:1353 8. 11. King RM, Pairolero PC, Trastek VF, Piehler JM, Payne SW, Bernatz PE. Primary chest wall tumors: factors affecting survival. Ann Thorac Surg 1986;41:597 601. 12. Pairolero P. Discussion in Martini N, Huvos AG, Burt ME, et al. Predictors of survival in malignant tumors of the sternum. J Thorac Cardiovasc Surg 1996;111:106. 13. McCormack PM, Bains MS, Beattie EJ, Martini N. New trends in skeletal reconstruction after resection of chest wall tumors. Ann Thorac Surg 1981;31:45 52. 14. Chapelier A, Macchiarini P, Rietjens M, et al. Chest wall reconstruction following resection of large primary malignant tumors. Eur J Cardiothorac Surg 1994;8:351 7. 15. McCormack PM. Use of prosthetic materials in chest-wall reconstruction. Surg Clin North Am 1989;69:965 76. 16. Boyd AD, Shaw WW, McCarthy JG, et al. Immediate reconstruction of full thickness chest wall defects. Ann Thorac Surg 1980;32:337 46. 17. Arnold PG, Pairolero PC. Use of pectoralis major muscle flap to repair defects of the anterior chest wall. Plast Reconstr Surg 1979;63:205 13. 18. Pairolero PC, Arnold PG. Chest wall reconstruction. Ann Thorac Surg 1981;32:325 6. 19. Moelleken BR, Mathes SA, Chang N. Latissimus dorsi muscle musculocutaneous flap in chest wall reconstruction. Surg Clin North Am 1989;69:977 90. 20. Mansour KA, Anderson TM, Hester TR. Sternal resection and reconstruction. Ann Thorac Surg 1993;55:838 43. 21. Burt M, Fulton M, Wessner-Dunlap S, et al. Primary bony and cartilaginous sarcomas of chest wall; results of therapy. Ann Thorac Surg 1992;54:226 32. 22. McAfee MK, Pairolero PC, Bergstralh EJ, et al. Chondrosarcoma of the chest wall; factors affecting survival. Ann Thorac Surg 1985;40:535 41. 23. Burt M, Karpeh M, Ukoha O, et al. Medical tumors of the chest wall. Solitary plasmacytoma and Ewing s sarcoma. J Thorac Cardiovasc Surg 1993;105:89 96. DISCUSSION DR STEPHEN C. YANG (Baltimore, MD): Do you find that the margins are important? What kind of margins do you use for your tumors? DR CHAPELIER: Safe margins are very important. Soft tissue margins must be clear, at least 3 cm, especially in the case of radiation-induced sarcoma, because the interpretation of frozen sections is very difficult. Macroscopic bone margins are at least 3 cm to minimize the risk of local recurrence. DR YANG: The patients you had to reconstruct the superior vena cava (SVC), were you able to predict this preoperatively? DR CHAPELIER: We could predict superior vena cava invasion preoperatively in the first patient who had a second recurrence of a radiation-induced fibrosarcoma involving the manubrium. Magnetic resonance imaging and superior phlebography revealed involvement of the confluence between the right brachiocephalic vein and the SVC. The second patient had a very large malignant fibrous histiocytoma including both parts of the sternum and we could suspect SVC involvement on the computed tomographic (CT) scan. DR YANG: Would you consider this a contraindication for resection? DR CHAPELIER: Invasion of SVC by a primary malignant sternal tumor is indeed very rare but would not be considered as a contraindication for resection in selected patients, because we can easily reconstruct SVC with a PTFE graft. DR JOSEPH B. SHRAGER (Philadelphia, PA): Doctor Chapelier, I am not sure what the message is that we are supposed to be getting about stability of the anterior chest wall. Are you telling us that in your experience you don t need in any of these cases to place any sort of a stiff or hard material? I know you made the

Ann Thorac Surg CHAPELIER ET AL 2004;77:1001 7 STERNAL RESECTION FOR MALIGNANT TUMORS 1007 argument against methylmethacrylate. So you are finding that the Gore-Tex (W. L. Gore & Associates, Flagstaff, AZ) alone, if it is stretched tight, gives enough stability in all these patients? DR CHAPELIER: Regarding the risk of infection, we would now reserve rigid skeletal reconstruction for total sternectomy. Otherwise it seems to us that when a small part of the sternum can be spared, the use of a PTFE patch with an excellent muscle or myocutaneous flap is enough to reconstruct safely sternal defects. DR SHRAGER: Was there anything common about the patients, shared by the patients, who had instability? I think you mentioned there were 3 patients with instability. Were they all done with just myocutaneous flaps with nothing else? DR CHAPELIER: The 3 patients were women; 2 had a total sternectomy with a rigid replacement, and 1 had a subtotal sternectomy extended to the anterior chest wall with a PTFE patch only. The large anterior defect could explain the postoperative instability in these patients. DR ROBERT B. LEE (Jackson, MS): My question follows that of the previous discussant, and perhaps is answered by the length of time of your study. A comment preceeds my question. I think most thoracic surgeons would perform a complete sternectomy using muscle flaps for reconstruction rather than rigid material, such as methylmethacrylate, with good success. Clearly, your methylmethacrylate group was problematic and your morbidity was related to rigid materials used for reconstruction. My question for you is, as your series has matured have you abandoned the use of methyl methacrylate entirely? I personally do not use synthetic material of any type. I think that with the appropriate use of muscle flaps you are almost always going to achieve the stability you need with very low complication rate. DR CHAPELIER: Thank you for your question. Again, if a small part of the sternum is preserved, reconstruction can be performed without methylmethacrylate with enough stability. We do agree that soft tissue coverage with muscle or musculocutaneous flap is the main step of the reconstruction as we have demonstrated in our presentation. However, in the case of total sternectomy, we would still continue to advocate reinforcement with methylmethacrylate to avoid flail chest and prolonged ventillatory support. DR LUKE GASIOROWSKI (Grand Forks, ND): You had several cases of infection. Did you use any kind of antibiotic prophylaxis depending on the type of material that you used in the reconstruction? DR CHAPELIER: Yes, for every patient, we used a current prophylaxis with cefamandol at the doses of 1.5 g after anesthetic induction, then 750 mg after 4 hours, and again 750 mg at the 12th hour.