A reconstructive algorithm for plastic surgery following extensive chest wall resection

The British Association of Plastic Surgeons (2004) 57, 295 302 A reconstructive algorithm for plastic surgery following extensive chest wall resection A. Losken a, *, V.H. Thourani b, G.W. Carlson a, G.E. Jones a, J.H. Culbertson a, J.I. Miller b, K.A. Mansour b a Division of Plastic and Reconstructive Surgery, Joseph B Whitehead Department of Surgery, Emory University School of Medicine, Atlanta, GA, USA b Division of Cardiothoracic Surgery, Joseph B Whitehead Department of Surgery, Emory University School of Medicine, Atlanta, GA, USA Received 13 March 2003; accepted 16 February 2004 KEYWORDS Resection; Chest-wall; Pleural cavity Summary Chest wall reconstruction following extensive resection is greatly facilitated by the use of vascularised flaps and prosthetic material. Plastic surgeons are often asked to assist with coverage of large chest wall defects. However, in addition to soft tissue coverage, we need to address other important issues such as the status of the pleural cavity, and the requirement for skeletal support. The purpose of this report is to analyse our experience, provide a reconstructive algorithm following the ablative procedure and review the literature. Two hundred chest wall resections were performed from 1975 to 2000. Defect location was divided into anterior ðn ¼ 73Þ; lateral ðn ¼ 36Þ; anterior-lateral ðn ¼ 36Þ; posterior-lateral ðn ¼ 19Þ; posterior ðn ¼ 22Þ and forequarter ðn ¼ 14Þ: Average number of ribs resected was four. One hundred and fifty-eight patients (79%) required chest wall reconstruction with either prosthetic material and/or flap closure. Mesh closure was required in 85 cases (43%), being highest for lateral defects (61%), and lowest for anterior defects (31%). Vascularised flaps were needed in 112 patients (56%), more common in anterior defects (79%), and less common for the posterior-lateral defects (26%). Inpatient complication rate was 27% (43/158) following reconstruction, with a mortality of 6% (10/158). Chest wall reconstruction is common following extensive resection. This includes management of the pleural cavity, skeletal support and soft tissue coverage. A better understanding of the respiratory mechanics and local thoracoabdominal anatomy is crucial for managing these complex defects. The need for skeletal support was more prevalent in lateral and posterior-lateral defects. Flap reconstruction was required more often to cover large anterior defects, with regional flaps predominating. Q 2004 The British Association of Plastic Surgeons. Published by Elsevier Ltd. All rights reserved. *Corresponding author. Address: Emory Division of Plastic Surgery, 550 Peachtree Street, Suite 84300, Atlanta, GA 30308, USA. Tel.: þ1-404-686-1230. E-mail address: albert_losken@emoryhealthcare.org Refinements in plastic and reconstructive surgical options have changed the management of complex chest wall defects. With the introduction of muscle and myocutaneous flaps came numerous indications S0007-1226/$ - see front matter Q 2004 The British Association of Plastic Surgeons. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.bjps.2004.02.004

296 A. Losken et al. for their use following extensive resection. The reconstruction now occurs immediately following the resection, which has contributed towards improving recovery and outcome. Communication between the ablative and reconstructive teams is important, and will ensure appropriate pre-operative planning regarding the extent of resection, keeping in mind the various reconstructive options. Although the indications for chest wall resection have not changed much, improvements in reconstructive techniques have streamlined the process. Historically, the major concerns were directed towards the respiratory mechanics rather than the actual pathology of the lesions. 1 The experience gained with thoracic trauma during World War II along with refinements in ancillary therapies by the 1950s, allowed more aggressive chest wall resections. However, skeletal protection and wound coverage remained an issue. It was not until the development of reliable synthetic substances for skeletal support, 2 and later the introduction of muscle and myocutaneous flaps that immediate reconstruction became a viable and attractive option. These local flaps from the trunk also provided alternatives for management of the pleural cavity such as obliteration of post pneumonectomy empyema spaces and the closure of fistulas. Decision as to the most appropriate reconstructive option depends on assessment of the defect in terms of location, depth, aetiology and a thorough understanding of chest wall mechanics and physiology. The ultimate goal is to obliterate dead space, maintain adequate chest wall stability and provide coverage while preserving form and function. We reviewed our 25-year experience with chest wall resections and reconstruction in an attempt to identify an algorithm for skeletal support and coverage based on the location and extent of defects. Clinical experience Resection Two hundred patients underwent chest wall resection between 1975 and 2000 at Emory University and Crawford Long Hospitals. The mean age was 54 ^ 15 years (range: 13 86 years) with 106 (53%) male and 94 (47%) female. Indications were the presence of primary lung cancer with extension into the chest wall (38%, n ¼ 75), primary chest wall tumors (27%, n ¼ 53), breast cancer with recurrence, metastasis or chest wall extension Table 1 The incidence of any chest wall reconstruction as a function of defect location Defect location N # Ribs resected Reconstruction % ðnþ Anterior 73 3.9 88 (64) Lateral 36 3.26 78 (28) Anterior-lateral 36 3.55 75 (27) Posterior 22 3.50 82 (18) Posterior-lateral 19 3.33 63 (12) Forequarter 14 5.5 64 (9) (22%, n ¼ 42), radiation necrosis (15%, n ¼ 29), and other (0.5%, n ¼ 1). Patients with less than two rib resections, routine pectus resections, acute sternal infections following median sternotomy or primary Eloesser procedures were not included in this series. The average number of ribs resected was 4 ^ 2 (range: 2 9 ribs). The most common defect location was anterior ðn ¼ 73Þ with an average number of 3.9 ribs resected (Table 1). Forequarter amputations with chest wall resection had the highest average number of ribs resected at 5.5. Forty-two patients (21%) had primary closure without the need for skeletal support, or without any type of flap coverage. Reconstruction One hundred and fifty-eight patients (79%) underwent chest wall reconstruction with either synthetic mesh and/or flap coverage. Chest wall reconstruction was required most commonly for defects in the anterior (88%), and posterior location (82%) (Table 1). Synthetic mesh was used for skeletal support in 85 patients (43%) (Table 2). The majority of mesh closures were with one synthetic material (prolene mesh n ¼ 42; polytetrafluroethylene (PTFE) n ¼ 1; Marlex mesh n ¼ 19; methyl methacrylate n ¼ 7 and double Vicryl mesh n ¼ 8). Combination mesh closures were performed in eight patients (Prolene/Vicryl n ¼ 3; Marlex/ methyl methacrylate n ¼ 1; Prolene/methylacrylate n ¼ 3 and Prolene/Marlex n ¼ 1), of which six were for anterior chest wall defects. Mesh closure was Table 2 Skeletal stabilisation and coverage as a function of defect location Defect location N Mesh % ðnþ Mesh with flap % ðnþ Anterior 73 31 (22) 73 (16/22) Lateral 36 61 (22) 36 (8/22) Anterior-lateral 36 42 (15) 33 (5/15) Posterior 22 40 (9) 27 (3/11) Posterior-lateral 19 58 (11) 22 (2/9) Forequarter 14 42 (6) 100 (6/6)

A reconstructive algorithm for plastic surgery following extensive chest wall resection 297 Table 3 required more often for the lateral defects (lateral 61%, anterior-lateral 42% and posterior lateral 58%) when compared to anterior and posterior defects. A local or distant flap was required to cover the mesh more commonly for the forequarter amputations (100%) and anterior defects (73%). Flap coverage was required in 112 patients, being utilised more often for the anterior defects (79%), with free flaps being most common in the forequarter group (Table 3). Twenty-seven patients (24%, n ¼ 27=112) required two flaps to achieve final chest wall closure, and one patient required three flaps. Table 4 lists the type of local or pedicled flaps used for each defect location. Sixteen free flaps were used which included latissimus dorsi ðn ¼ 7Þ; rectus abdominus ðn ¼ 8Þ and omentum ðn ¼ 1Þ: Immediate reconstruction was performed in 96% of the cases ðn ¼ 152=158Þ: Outcome Flap coverage as a function of defect location Defect location N Any flap % ðnþ Pedicled % ðnþ Free % ðnþ Anterior 73 79 (58) 81 (47) 19 (11) Lateral 36 39 (14) 86 (12) 14 (2) Anterior-lateral 36 44 (16) 100 (16) 0 (0) Posterior 22 45 (10) 100 (10) 0 (0) Posterior-lateral 19 26 (5) 80 (4) 20 (1) Forequarter 14 64 (9) 78 (7) 22 (2) Overall length of stay (LOS) was 18 ^ 16 days (range: 3 99 days), with an inpatient mortality rate of 7% ðn ¼ 13=200Þ: Average LOS in the immediate reconstruction group was 18.3 ^ 17.1 days ðn ¼ 152=158Þ; compared to 23.8 ^ 20.63 days in the delayed reconstruction group ðn ¼ 6=158Þ ðp ¼ 0:44Þ: The most common inpatient complications included pneumonia (14%, n ¼ 27), acute respiratory distress syndrome (6%, n ¼ 11) and flap loss (5%, n ¼ 10). Wound infection was present in nine patients, all in the reconstruction group. Possible risk factors for this group included mesh closure ðn ¼ 4=9Þ; delayed reconstruction ðn ¼ 1=9Þ; radiation treatment ðn ¼ 2=9Þ; hypertension ðn ¼ 5=9Þ; and forequarter amputations ðn ¼ 4=9Þ: Comparisons between those patients who had primary closure and those who required reconstruction are presented in Table 5. Discussion Plastic surgeons are becoming more involved in the immediate reconstruction of massive chest wall defects. This allows the extirpative surgeon a certain comfort level to obtain wide enough margins to eliminate all possible malignant, contaminated, irradiated or non-viable tissue. Combined efforts with the extirpative surgeons will ensure appropriate management of these patients in order to maximise outcome and improve recovery. Massive chest wall resection with immediate reconstruction has been shown to be both safe and effective. 3 Although primary closure without synthetic material or flap coverage was possible 21% of the time, the majority of chest wall defects required formal reconstruction, especially anterior and posterior defects (Fig. 1). Reconstruction of acquired thoracic defects requires attention to management of (1) the pleural cavity, (2) skeletal support and (3) soft tissue coverage. In our clinical series we focused mainly on the actual chest wall reconstruction (skeletal support and soft tissue coverage), however, the management of the pleural cavity is often essential, and has paralleled progress made in soft tissue coverage. Management of the pleural cavity Transposition of muscle flaps or vascularised tissue into the chest cavity is widely used to obliterate post-pneumonectomy empyema spaces, closure of bronchopleural or tracheoesophageal fistulas. Excellent local flap options include extrathoracic muscles (latissimus dorsi, serratus, pectoralis major) and omentum previously described at our institution. 4,16 These muscles are easily passed through a 4 5 cm chest wall defect with a two-rib Table 4 The various regional flaps used for chest wall coverage Anterior Lateral Anterior lateral Posterior Posterior lateral Forequarter N Pectoralis 20 Latissimus 9 Latissimus 7 Latissimus 5 Omentum 3 Deltoid 4 Latissimus 12 Rectus 3 Omentum 5 Serratus 4 Rectus 2 Rectus 3 Rectus 14 Serratus 3 Pectoralis 4 Rectus 3 Serratus 2 Pectoralis 2 Omentum 6 Pectoralis 1 Serratus 3 Omentum 2 Latiss 2 Trapezium 2 Serratus 4 Rectus 2 Pectoralis 1 Omentum 2 Trapezius 1

298 A. Losken et al. Table 5 Comparisons between primary closure and those requiring chest wall reconstruction Primary closure N ¼ 42 Reconstruction N ¼ 158 Chi-square or t-test Any complication 14% (6/42) 27% (43/158) P ¼ 0:126 Inpatient mortality 7% (3/42) 6% (10/158) P ¼ 0:871 Infection 0% (0/42) 6% (9/158) P ¼ 0:244 Pneumonia 7% (3/42) 15% (24/158) P ¼ 0:216 LOS 15 ^ 11.4 days 18.5 ^ 17.2 days P ¼ 0:208 LOS, length of stay. resection, location of which depends on the muscle used and its dominant blood supply. These defects are then closed over chest tubes for drainage with appropriate antibiotic usage. The maintenance of chest wall mechanics relies on both an air-tight seal within the pleural cavity as well as skeletal support over the thorax. One exception to this end is in the severely debilitated patient where sepsis is controlled by creating an Eloesser flap to drain the pleural cavity until appropriate for closure. Other options for large bronchopleural fistulas include muscle flap coverage as described above, followed by antibiotic irrigation, frequent dressing changes and secondary closure. These principles have been shown to be successful in 64% of patients with empyema and 72.5% of patients with fistula. 5 Another important issue in the management of these pleural cavity infections is the timing of the reconstruction. The success rate is higher if treatment is instituted early, prior to maturation of the empyema cavity. Thoracoplasty or total collapse of the chest wall with obliteration of remaining pleural cavities is reserved as a last resort for chronic empyema. Management of the chest wall Figure 1 Reconstructive algorithm for chest wall resection. STSG, Split thickness skin graft; MMS, methyl methacrylate sandwich; ForeQ, forequarter amputation. The management of actual chest wall defects requires similar logic with a thorough analysis of the defect and understanding of respiratory mechanics. Restoration of skeletal stability is often required to protect intrathoracic contents as well as to preserve the mechanical forces that allow respiration. Integrity of the diaphragm and accessory muscles of inspiration should always be considered when accessing the skeletal defect, and when deciding on appropriate local flap coverage if needed. Appropriate pre-operative pulmonary, cardiovascular and nutritional assessment will estimate the patients ability to tolerate certain resections. 6 Traditionally, reconstruction of the skeletal framework was performed using bone, diced cartilage, metal sheets, autogenous rib grafts, fascia lata, Teflon, Ivalon sponges and numerous other substances. 7,8 The synthetic materials available today are diverse and provide adequate support and stability. The ideal characteristics of a prosthetic material for chest wall reconstruction include rigidity, malleability, inertness and radiolucency. The choice of prosthetic material is often based on surgeon s preference. We have had good experience with Prolene and Marlex mesh as they typically provide semi-rigid fixation and good skeletal support when sutured under tension. These materials are also used for their good in-growth and pliability (Fig. 2). Sutures

A reconstructive algorithm for plastic surgery following extensive chest wall resection 299 Figure 2 (A) Pre-operative CT scan of a 30-year-old man with osteogenicsarcoma of the left chest wall. (B) Left sided defect following en bloc chest wall resection (five ribs), with resection of the upper lobe and pericardium. (C) Skeletal reconstruction using Prolene mesh. (D) Coverage using a transverse rectus abdominus muscle (TRAM) flap. are often placed in an interrupted fashion around the ribs for added strength and support. Deschamps showed no significant difference in the postoperative outcome or complications between Prolene mesh and PTFE soft tissue patch for chest wall reconstruction. 9 Prolene mesh was often doubled or quadrupled on itself for additional support. This is similar to the double-knit prolene mesh technique proposed by Arnold et al. 10 Vicryl mesh was used in our series as temporary cover in situations where there was wound contamination or infection. However, if rigidity is considered crucial then methyl methylacrylate can be used alone or incorporated into the mesh in a sandwich fashion. When used as a sandwich with mesh, the methylacrylate can be incorporated within the mesh, leaving a rim of mesh to act as a sewing surface. This is often required following sternal resection when stability and protective coverage is essential over exposed heart or great vessels. The actual importance of rigid stability in the prevention of uncoordinated chest wall motion is unclear following chest wall resections. This flail chest phenomenon leading to pulmonary insufficiency is seen more often in the trauma situation, however, needs to be avoided at all cost. Post-operative respiratory support was not evaluated in this series, however, larger chest wall resections routinely had 3 5 days of ventilatory support. Kroll et al. evaluated the post-operative respiratory course in 101 patients with and without Marlex mesh, and found that those with Marlex mesh stabilisation required a mean of 0.8 days less mechanical ventilation. However, post-operative wound infection were more common in the Marlex mesh group in that series (5% vs. 0%), usually related to loss of the overlying flap. 11 Numerous factors influence the decision process regarding which defects required skeletal reconstruction. It has generally been accepted that a two-rib segmental chest wall resection required soft tissue coverage alone. Mesh reconstruction should be considered when the loss of skeletal support is expected. Arnold states that most patients can tolerate sternectomy or resection of 4 6 ribs at the cartilage level without experiencing flail chest or respiratory insufficiency. 10 Although the number of resected ribs is an important indicator for mesh usage, there did not appear to be a direct association between the number of ribs resected and the need for mesh reconstruction in our series. The average number of ribs resected in the mesh group and non-mesh group was similar. This is likely to be due to the presence of additional factors that influence chest wall stability and subsequently the decision process regarding the necessity for mesh reconstruction. Such factors

300 A. Losken et al. include defect location and history of radiation. Location of the chest wall defect did appear to influence the need for skeletal stabilisation, with mesh reconstruction being required more often for the lateral defects. The lack of sternal or spinal stability in that location renders the patient more prone to flail chest deformities following chest wall resection. Pancoast tumors typically required multi-rib resection, but mesh was rarely used in this location because of the scapular support and their anatomic orientation within the upper chest cage. Similarly, radiated chest wall defects often tolerate extensive skeletal resection given the inherent stability and fixation of the previously irradiated chest wall as a result of radiation fibrosis stiffening the soft tissues. Muscle flaps alone often provided enough stabilisation for large radiated defects without causing flail segments. The smaller defects (less than 5 cm) and those located posterior beneath the scapula and above the fourth rib could usually be closed with soft tissue only, ignoring the skeletal component. Forequarter amputations with chest wall resections were performed through a modified transmediastinal approach as described previously by the senior author. 12 These defects were often extensive and required coverage of mesh with a deltocervical flap. When mesh was used for anterior and forequarter amputation defects these were more likely to require flap coverage in our series. Soft tissue coverage is the final issue that needs to be addressed following chest wall resections. While superficial defects of the chest wall are easily closed with skin grafts, full thickness defects are more challenging. Small full thickness defects can be closed primarily. Vascularised soft tissue flaps are often required to close larger defects, control infection, obliterate dead space, and provide coverage of synthetic material. Coverage for major chest wall defects has provided some of the earliest descriptions of muscle or other vascularised flaps used in reconstructive surgery. Traditionally, soft tissue coverage was performed using skin flaps such as the Bakamjian deltopectoral flap. 13 However, with better understanding of the pattern of blood supply to local muscle and musculocutaneous flaps based on the Mathes and Nahai classification system, these became the preferred coverage by the 1970s. 14 The greater omentum is often available as a salvage procedure and provides wellvascularised tissue to areas of extensive radiation damage or infection where other local flaps have failed or are insufficient. It was described by Jurkiewicz in 1977, for the coverage of the anterior chest wall. 15 Usually transferred on the right gastroepiploic artery, it provides reliable tissue to cover a large surface area or obliterate dead space, however, not without risks of possible intraabdominal complications. 16 Although Tansini, in 1906, originally described the latissimus dorsi musculocutaneous flap for coverage of an anterior chest wall defect after radical mastectomy, it was not until the latter half of the century that this flap became a workhorse for chest wall coverage. 17 19 It can be transferred as a muscle or myocutaneous flap on the dominant thoracodorsal pedicle, and is ideally suited for anterior and anterolateral defects, but was used successfully in all the defect locations in our series. Previous thoracotomy or axillary incisions need to be taken into account for possible interruption of this dominant pedicle. Muscles located on the lateral thoracic wall, or the greater omentum became more appropriate in these situations. Other muscle units that were used in our series for soft tissue coverage included the rectus, pectoralis and serratus muscle (Fig. 3). These have all played a major role in the management of sternal wound infections and have been well described in the past for chest wall reconstruction. 20,21 Regional fasciocutaneous flaps such as the subscapula-pubic, deltopectoral or random flaps have all been described for chest wall coverage with fair results. Flap reconstruction was required more frequently for the anterior chest wall defects in our series, where numerous local flaps made effective coverage a viable option. The availability of reconstructive options with well-vascularised tissue enables the extirpative surgeon to take wide and appropriate resections to ensure successful long-term outcomes. Free tissue transfer was infrequent, and reserved for situations where regional flaps were unavailable, insufficient or had previously failed. Only four patients had delayed chest wall reconstruction usually to allow the wound time to declare itself, or the infectious process to subside. The vast majority of chest wall reconstructions were performed at the time of resection without any difficulty, reducing LOS in a cost conscious healthcare society. Although the complication rate was higher in those patients who required reconstruction, this was attributed to the extent of resection required, as well as the additional reconstructive morbidities expected following these procedures. Back wounds represent another challenge to the reconstructive surgeon, and a complete discussion of this topic is beyond the scope of this review. However, the trapezius muscle flap are useful for upper third defects, latissimus dorsi muscle of reverse latissimus dorsi muscle flap for middle third defects, and gluteus maximus muscle flap coverage for lower third defects. The authors have recently

A reconstructive algorithm for plastic surgery following extensive chest wall resection 301 flaps. Although the proposed algorithm is helpful as a guide, the complexity of each individual patient needs to be taken into account, and numerous variables will influence the decision process regarding the most appropriate management. Close interaction with the thoracic surgeon is important, and both teams must completely understand the ablative goals and reconstructive options to avoid unnecessary morbidity and ensure a successful outcome. References Figure 3 (A) Pre-operative CT of a 50-year-old man with malignant fibrous hystiocytoma of the chest wall. (B) Large chest wall resection with Prolene mesh closure. Methyl methacrylate was added for additional support and protection. (C) Coverage with a vertical rectus abdominus myocutaneous flap. had good results with mobilising bilateral paraspinous muscle flaps for the coverage of middle and lower third defects. 22 Chest wall reconstruction is both safe and affective in the immediate setting with the majority of defects being closed using either synthetic mesh or regional muscle or myocutaneous 1. Hedblom CA. Tumors of the bony chest wall. Ann Surg 1933; 528:98. 2. Usher FC, Fries JG, Ochsner JL, et al. Marlex mesh, a new plastic mesh for replacing tissue defects. Arch Surg 1959; 137:78. 3. Mansour KA, Thourani VH, Losken A, Reeves JG, Miller Jr JR, Carlson GW, Jones GE. Chest wall resection and reconstruction: a 25-year experience. Ann Thorac Surg 2002;73: 1720 26. 4. Miller JI, et al. Single-stage complete muscle flap closure of the post-pneumonectomy empyema space: a new method and possible solution to a disturbing complication. Ann Thorac Surg 1984;38:227. 5. Arnold PG, Pairolero PC. Intrathoracic muscle flaps: a 10- year experience in the management of life threatening infections. Plast Reconstr Surg 1989;84:92. 6. Azarow KS, Molloy M, Seyfer AE, Graeber GM. Preoperative evaluation and general preparation for chest-wall operations. Surg Clin N Am 1989;69:899. 7. Southwick MW, Economou SG, Otten JW. Prosthetic replacement of chest wall defects. Arch Surg 1956;901:72. 8. LeVeen HH, Barbario JR. Tissue reaction to plastics used in surgery with special reference to Teflon. Ann Surg 1949;74: 129. 9. Deshamps C, Tirnaksiz BM, Darbandi R, et al. Early and long term results of prosthetic chest wall reconstruction. J Thorac Cardiovasc Surg 1999;117:588 92. 10. Arnold PG, Pairolero PC. Chest wall reconstructions: an account of 500 consecutive cases. Plast Reconstr Surg 1996; 98(5):804. 11. Kroll SS, et al. Risks and benefits of using Marlex mesh in chest wall reconstruction. Ann Plast Surg 1993;31:303. 12. Mansour KA, Powell RW. Modified technique for radical transmediastinal forequarter amputation and chest wall resection. J Thorac Cardiovasc Surg 1978;76(3):358 63. 13. Bakamjian VY. A two staged method for pharyngeal reconstruction with a primary pectoral skin flap. Plast Reconstr Surg 1965;36:173. 14. Mathes SJ, Nahai F. Classification of the vascular anatomy of muscle: experimental and clinical correlation. Plast Reconstr Surg 1981;67:177. 15. Jurkiewicz MJ, Arnold PG. The omentum: an account of its use in reconstruction of the chest wall. Ann Surg 1977;185:548. 16. Hultman CS, Culbertson JH, Jones GE, Losken A, et al. Thoracic reconstruction with the omentum: indications, complications and results. Ann Plast Surg 2001;46:242 9. 17. Tansini I. Sopra il mio nuovo processo di amputazione della mammella. Gazz Med Ital Torino 1906;57:141. 18. McGraw JB, Peniz JO, Baker JW. Repair of major defects

302 A. Losken et al. of the chest wall and spine with the latissimus dorsi myocutaneous flap. Plast Reconstr Surg 1978;62:197. 19. Bostwick J, Nahai F, Wallace JG, Vasconez LO. Sixty latissimus dorsi flaps. Plast Reconstr Surg 1979;63:31. 20. Arnold PG, Pairolero PC. Use of pectoralis muscle flaps to repair defects of the anterior chest wall. Plast Reconstr Surg 1979;63:205. 21. Jones G, Jurkiewicz MJ, Bostwick J, Wood R, Bried JT, Culbertson J, Howell R, Eaves F, Carlson G, Nahai F. Management of the infected median sternotomy wound with muscle flaps. The Emory 20-year experience. Ann Surg 1997;225(6):766 76. 22. Hultman CS, Jones GE, Losken A, Seify H, Schaefer TG, Carlson GW. Salvage of infected spinal stabilisation devices using paraspinous muscle flaps: an anatomic and clinical study. In preparation.