- Email: [email protected]
Reconstruction after major chest wall resection: Can rigid fixation be avoided?
Central Surgical Association Reconstruction after major chest wall resection: Can rigid fixation be avoided? Wael C. Hanna, MD, MBA, Lorenzo E. Ferri, MD, PhD, Katherine M. McKendy, MD, Robert Turcotte, MD, Christian Sirois, MD, and David S. Mulder, MD, Montreal, Canada
Background. Rigid fixation is advocated as the best method to achieve good respiratory outcomes after chest wall resection at the expense of a high complication rate. The following study aims to examine the role of myocutaneous pedicled flaps, with or without soft prosthesis, in the reconstruction of small and large chest wall defects. Methods. All patients who underwent resection of chest wall tumors between 2003–2010 were identified from a prospectively entered database. Operative and postoperative outcomes were documented. Patients were stratified into 2 separate groups based on the size of the residual chest wall defect; the Small Defect (SD) group (<60 cm2) and the Large Defect (LD) group (>60 cm2). Results. Thirty-seven patients were identified over a 7-year period: 9 in the SD group and 28 in the LD group. Primary sarcoma was the most common indication for resection (57%). The mean size of the chest wall defect was 50.8 cm2 in the SD group and 149.4 cm2 in the LD group (P = .001). All patients underwent reconstruction with autologous tissue, nonrigid prosthesis, or a combination of the two. Prosthesis was used in 11% of patients in the SD group and 61% of patients in the LD group (P = .018). The rate of immediate postoperative extubation was 100% in the SD group and 89% in the LD group (P = .42). The rate of postoperative pneumonia was 7% in the LD group vs 0% in the SD group. The rate of surgical site infection was 7% in the LD group and 0% in the SD group. A subgroup analysis of the LD group demonstrated no statistical differences in any of the measured outcomes between patients in whom mesh prosthesis was used and patients in whom a myocutaneous flap alone was used. However, there was a clinical suggestion of prolonged ventilation in the subgroup where mesh was not used and of higher infection rates in the subgroup where mesh was used. Conclusion. Small chest wall defects can be reconstructed with pedicled myocutaneous flaps alone without compromising respiratory outcomes. In carefully selected patients with moderate size defects larger than 60cm2, reconstruction with pedicled myocutaneous flap alone offers similar postoperative outcomes as reconstruction with nonrigid prosthesis, at the expense of a possible need for a short period of mechanical ventilation. (Surgery 2011;150:590-7.) From the McGill University Health Centre, Montreal General Hospital, Montreal, Quebec, Canada
THE PRIMARY CONCERN after chest wall resection is the maintenance of the stability and integrity of the chest wall.1 A wide excision often results in a large chest wall defect with instability, disruption of pulmonary mechanics, and considerable respiratory complications. These sequelae are particularly noted when multiple ribs are resected or when Accepted for publication July 11, 2011. Reprint requests: Wael C. Hanna, MD, MBA, The Montreal General Hospital, 1650 Cedar Avenue, Rm L9-112, Montreal, QC H3G 1A4, Canada. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2011 Mosby, Inc. All rights reserved. doi:10.1016/j.surg.2011.07.055
590 SURGERY
the sternum is resected.2 Reconstruction also protects chest organs, prevents entrapment of the scapula, and maintains the cosmetic shape of the chest wall.3 It is generally accepted that defects less than 5 cm in diameter usually do not require reconstruction. Posterior defects less than 10 cm in diameter can also be safely closed without reconstruction, provided there is no risk of entrapment of the tip of the scapula.4 For larger defects, no consensus exists on the ideal method of reconstruction and varied results have been reported in the surgical literature.5 Rigid fixation with methacrylate or metallic prostheses has traditionally been advocated as the
Surgery Volume 150, Number 4
best method to achieve chest wall stability and prevent respiratory complications.4-6 However, it is frequently complicated by infection, pain, and rigid deformities.7 Furthermore, should an infection arise, the presence of rigid foreign material significantly complicates management. Thus, an ideal repair would be one that adequately maintains chest wall stability to prevent pulmonary complications but avoids rigid fixation. The purpose of this study is to examine the role of myocutaneous pedicled flaps, with or without soft prosthesis, in the reconstruction of small and large chest wall defects. MATERIALS AND METHODS After ethics board approval, all patients who underwent major resection of primary or metastatic chest wall tumors at a university-affiliated tertiary care institution between 2003 and 2010 were identified from a prospectively entered database. All patients with primary lung or pleural cancers invading the chest wall were excluded to eliminate potential confounding factors, as respiratory complications were a major outcome of interest. Among the patients included, 4 patients had undergone previous unrelated lung resections and 3 patients underwent 2 separate chest wall resections with reconstruction. Through a retrospective chart review, information was recorded on patient characteristics including age, sex, nature of the chest wall tumor, and presence of pulmonary comorbidity. Treatment with neoadjuvant radiation or chemotherapy was also recorded. Operative and pathology reports were reviewed for the number of ribs resected, concurrent sternal or clavicular resection, size of the residual defect, and method of reconstruction. A linear regression analysis identified a cut-off defect size as an independent predictor for reconstruction with mesh, and therefore patients were stratified into 2 groups: those who had a Small Defect area <60 cm2 (SD) and those who had a Large Defect area $60 cm2 (LD). Only nonrigid mesh composed of polypropylene (Marlex, Davol, Warwick, RI, and Prolene, Ethicon, Inc, Somerville, NJ), polyesther (Mersilene, Ethicon, Inc, Somerville, NJ), polytetrafluoroethylene (Gore-Tex and Gore-Dualmesh, W.L. Gore & Associates, Inc, Flagstaff, AZ,) and polyglactin-910 (Vicryl, Ethicon, Inc, Somerville, NJ) was used. No patients underwent rigid fixation with methyl methacrylate or metal plates. All postoperative complications were identified from the prospectively entered database. Pulmonary complications included pneumonia and
Hanna et al 591
admission to the intensive care unit (ICU) for assisted ventilation. Pneumonia was defined according to the Center for Disease Control and Prevention guidelines.8 Assisted ventilation was defined as the need for mechanical ventilation with either invasive or noninvasive means. Surgical site infection was defined by the need to open and drain the surgical site, regardless of whether bacteriology was obtained. Mesh infection was defined as a fluid collection in the vicinity of the mesh with proven bacteria on Gram stain or culture. The primary outcome of interest was the rate of pulmonary complications. Secondary outcomes included the rate of surgical site infection, mesh infection, need for reoperation, length of hospitalization, and in-hospital mortality. Data are presented as median (range) and the Fisher exact test or Mann-Whitney U test was used to determine statistical significance (P < .05). RESULTS A total of 37 patients were included in the study. Sixty-two percent (62%) were male and the median age for all patients was 49 (range, 19–75) years. Nine patients were stratified into the SD group and 28 patients in the LD group. Four patients (11%) had undergone a chest wall resection in the past, all of them in the LD group. A multidisciplinary team including thoracic, plastic, and orthopedic surgeons determined the method of reconstruction for each patient. Although this decision was based on tumor size, location on the chest wall, number of ribs to be resected, and feasibility of mobilizing autologous tissue flaps, surgeon preference and judgment played a major role in selection; Figure. In the SD group, 33% of patients were smokers but none had a diagnosis of Chronic Obstructive Lung Disease. In the LD group, 21% of patients were smokers and 11% had a diagnosis of Chronic Obstructive Lung Disease based on pulmonary function tests. Primary chest wall sarcoma was the most common indication for resection (57%), followed by metastases to the chest wall (22%) and desmoid tumors (14%). The mean size of the defect created in the chest wall was 50.8 cm2 (range, 14–59) in the SD group and 149.4 cm2 (range, 60–405) in the LD group (P = .001). The incidence of partial sternal resection was similar in both groups (33% in LD group vs 27% in SD group, P = .12). In the LD group, 50% of patients had 3 or more ribs resected, whereas none did in the SD group (P = .007). Neoadjuvant chemotherapy was given to 18% of patients in the LD group, vs none in the SD group (P = .05). Neoadjuvant
592 Hanna et al
Surgery October 2011
Fig. Illustration of the steps for surgical reconstruction after chest wall resection. A, Preoperative appearance of chest wall tumor. B, Chest wall defect after resection. C, First step of reconstruction with soft prosthesis. D, Second step of reconstruction with myocutaneous flap and final appearance.
Table I. Demographic, tumor, and treatment characteristics for all patients who underwent chest wall resection with reconstruction Variable Patient characteristics Mean age Male (%) Smoking (%) COPD (%) Previous lung resection (%) Tumor characteristics Sarcoma (%) Metastases (%) Desmoid (%) Neurofibromas (%) Treatment characteristics Neoadjuvant chemotherapy (%) Neoadjuvant radiotherapy (%) Neoadjuvant chemoradiation (%)
All patients
Small defect cohort
Large defect cohort
n = 37
n=9
n = 28
P value
49 (19–75) 62 24 8 8
45 (19–71) 56 33 0 0
47 (35–75) 64 21 11 11
.13 .29 .19 .22 .22
57 22 14 8
78 11 0 11
50 25 18 7
.21 .19 .22 .19
14 11 5
0 0 0
18 14 7
.24 .22 .13
radiotherapy was given to 14% in the LD group vs none in the SD group (P = .05, Table I). There was no significant difference between the 2 groups in terms of usage of autologous muscle flaps, which were constructed in 56% of patients in the SD group and 75% of patients in the LD
group (P = .24). When a muscle flap was constructed, pectoralis major was used in 41% of the cases and latissimus dorsi in 16%. Soft prosthesis was used much more frequently in the LD group (61% of patients) than in the SD group (11%, P = .018). One quarter of patients in the LD group
Hanna et al 593
Surgery Volume 150, Number 4
Table II. Resection and reconstruction parameters for both cohorts of patients Variable Resection parameters Mean size of chest wall defect (cm2) More than 2 ribs resected (%) Partial sternal resection (%) Total sternal resection (%) Clavicle resection (%) Lung resected (%) Diaphragm resected (%) Vertebra resected (%) Reconstruction parameters Use of soft prosthetic mesh (%) Use of muscle flap (%) Pectoralis major (%) Latissimus dorsi (%) Serratus (%) Rectus abdominis (%) Diaphragm (%)
All patients
Small defect cohort
Large defect cohort
n = 37
n=9
n = 28
P value
100.14 38 22 3 11 8 5 3
50.84 0 33 0 11 11 0 0
149.45 50 18 4 11 7 7 4
.001 .04 .21 .19 .12 .13 .31 .41
49 70 41 16 8 5 5
11 56 80 40 0 0 20
61 75 52 19 14 10 5
.018 .14 .19 .31 .43 .31 .22
Statistically significant P values in bold.
Table III. Postoperative outcomes in both cohorts of patients Variable Postoperative outcomes Immediate extubation postoperatively (%) Admission to the ICU For ventilatory support (%) For flap monitoring (%) Median number of days ventilated Median length of stay in ICU Postoperative pneumonia (%) Surgical site infection (%) Re-operation for infection or bleeding (%)
All patients
Small defect cohort
Large defect cohort
n = 37
n=9
n = 28
P value
100 0
89 32% 44 56 2 (1–7) 2 (1–9) 11 7 11
.42
92 24%
(25%) required a skin graft to cover their reconstruction, whereas none did in the SD group (P = .11, Table II). The rate of immediate extubation postoperatively was 100% in the SD group vs 89% in the LD group (P = .32). Patients remained intubated postoperatively for a variety of reasons, including prolonged operative time, magnitude of reconstruction, the need for flap monitoring, and analgesia in the absence of an adequately functioning epidural catheter. All patients were extubated successfully by the second day after the operation. One more patient was re-intubated on the third day postoperatively due to deterioration in respiratory status and remained intubated for 7 days. The rate of postoperative admission to the ICU was 0%
8 5 8
0 0 0
.31 .29 .31
in the SD group and 32% in the LD group (P = .06). Of the patients admitted to the ICU, half (56%) were admitted for monitoring of the muscle flap and did not require ventilator support. The rest (44%) were admitted for mechanical ventilation. Of those patients, 75% had undergone reconstruction via autologous tissue alone, without the use of soft prosthesis. The median number of days on the ventilator was 2,1-7 and the median length of stay in the ICU was 2 days.1-9 Two patients developed postoperative pneumonia, both in the LD group (7% vs 0% in the SD group, P = .56). Surgical site infection developed in 7% of patients in the LD group but not in any patient in the SD group (P = .56). One patient in the LD group required reoperation for mesh removal due to
594 Hanna et al
Surgery October 2011
Table IV. Sub-group analysis for postoperative outcomes in patients of the large defect cohort Variable 2
Mean size of chest wall defect (cm ) Use of muscle flap (%) Immediate extubation postoperatively (%) Admission to the ICU (%) For ventilatory support (%) For flap monitoring (%) Surgical site infection (%) Re-operation for infection or bleeding (%)
MESH
NO MESH
n = 16
n = 12
P value
161.9 75 100 19 33 67 13 19
131.72 75 75 42 60 40 0 0
.29 .25 .15 .21 .19 .13 .14 .14
MESH indicates the subgroup of patients where soft prosthesis was used for reconstruction vs NO MESH, which indicates the subgroup of patients where reconstruction was done with autologous tissue alone.
infection, while the remainder of the patients were treated conservatively either with local intervention or systemic antibiotics. Another patient in LD group required reoperation for significant bleeding, rendering the rate of reoperation in the LD group higher than in the SD group (7% vs 0% respectively, P = .56). Median length of stay in hospital for the SD group was 6 days (range, 5–10) compared to 9 days (range, 7–15) in the LD group (P = .41). Two patients in the LD group (7%) required readmission for pulmonary embolus and pneumonia respectively. There were no mortalities within the postoperative period of 30 days, Table III. A subgroup analysis of the LD cohort was done aiming to elucidate any differences in postoperative outcomes for patients in whom a prosthesis was used (MESH) vs those in whom reconstruction was done with autologous tissue alone (NO MESH), Table IV. In the NO MESH cohort, the incidence of admission to the ICU for ventilator support was higher than in the MESH group (60% vs 33%, P = .19). In contrast, the rate of surgical site infection and reoperation was higher in the MESH group. No statistical significance could be achieved in any of these parameters. Patients were seen postoperatively every 3 months for the first 2 years after operation, then at 6-month intervals thereafter, with a mean followup period of 42 months. During every visit, the patient would be examined by visual inspection of the scar, palpation of the chest wall, auscultation for lung breath sounds, and provocative coughing. Although no questionnaires were submitted to address specific postoperative outcomes, the visit notes reported no incidences of dyspnea at home, exercise limitation, oxygen requirement, chronic cough, chronic pain, nerve injury, motion impairments, or cosmetic defects as perceived by the
patient. Patients did not receive postoperative computed tomography (CT) scans on a routine basis. Of those who did get CT scans for oncological follow-up, none had reported lung hernias. DISCUSSION In our institution over the past 7 years, chest wall reconstruction has been performed exclusively using autologous muscle flaps or nonrigid mesh. This choice was part of our strategy to abandon rigid fixation due to largely disappointing results in terms of prosthesis infection, postoperative pain, and chest wall deformities. In our series, the immediate postoperative extubation rate was comparable to most series that used rigid fixation with methyl methacrylate.7,9 Those same studies demonstrate a 7–10% incidence of pleural effusions requiring drainage, whereas in our series we had none. We have also shown that nonrigid mesh prosthesis reconstruction carries a lower rate of surgical site infection (5%) than does rigid fixation (7– 20%).7,9-11 More importantly, we have documented no prosthesis infections that occurred later than 30 days after surgery, whereas delayed infection with methyl methacrylate has been reported as late as 12 months after surgery.7 It is important to note that the rate of reoperation for mesh infection in our series (3%) was considerably lower than what is usually associated with rigid fixation (8–12%).7,9 Although in our series, nonrigid reconstruction carried a slightly higher rate of postoperative pneumonia than what is historically reported for rigid fixation, the postoperative length of stay in hospital was considerably shorter.7 On long-term follow-up of 42 months, we have not recorded any incidence of chronic pain, neurologic impairment, or cosmetic or motion deformities associated with soft prosthesis reconstruction of the chest wall. This is in contrast with rigid fixation where long-term
Hanna et al 595
Surgery Volume 150, Number 4
pain, deformity, and cosmesis have been major issues with this type of reconstruction. Therefore, our data suggests that when the use of prosthetic material cannot be avoided, nonrigid prosthesis may offer more favorable short-term and long-term outcomes than what was historically demonstrated in the literature on rigid fixation. Furthermore, our data suggest that construction of a myocutaneous flap is crucial to the realization of good outcomes in patients with large residual chest wall defects. Within our cohort of patients with large defects, our surgeons elected to use a myocutaneous flap in 75% of large defects, even when a prosthesis was used. Although our data suggests that the incidence of admission to the ICU and the incidence of postoperative pneumonia may be higher when myocutaneous flap was used alone without mesh, this difference does not achieve statistical significance. As expected, the incidence of surgical site infection and reoperation is considerably lower when myocutaneous flap is used alone without mesh, again without any statistical significance. It is important to emphasize, nevertheless, that the small sample size in this subgroup analysis may prohibit the generation of any statistically significant results, even when clinical significance is present. Our data therefore suggests that, for chest wall defects larger than 60 cm2, reconstruction can be done without the use of prosthetic material, with little difference in postoperative outcomes but a possibility of prolonged ventilation requirements. It is important to note that defect size alone is not the only determinant of whether mesh should be used in reconstruction or not. The location of the chest wall defect (anterior vs posterior), the number of ribs removed, and the body habitus of the patient (thin vs obese) are factors that cannot be ignored when planning reconstruction. The present study is limited by several factors. Although information was rigorously collected from a prospectively entered database, the retrospective nature of this study predisposes the data to strong selection bias, especially regarding the choice of reconstructive technique and operative planning. The small number of patients included in this study is a reflection of the rarity of elective chest wall resection for primary tumors, but it also implies that general conclusions have to be extrapolated from a small sample size. Finally, although postoperative long-term outcomes were very well documented on follow-up, the lack of standardized questionnaires and pulmonary function test makes this data difficult to assess objectively and compare across all patients.
In conclusion, we have demonstrated that in patients with chest wall defects smaller than 60 cm2, there appears to be little benefit of reconstruction with a prosthesis, because postoperative outcomes with primary closure of the chest wall are excellent. In moderate size defects larger than 60 cm, and in carefully selected patients, reconstruction with pedicled myocutaneous flap alone offers similar postoperative outcomes as reconstruction with nonrigid prosthesis, at the expense of a possible need for a short period of mechanical ventilation.
REFERENCES 1. Nirula R, Mayberry JC. Rib fracture fixation: controversies and technical challenges. Am Surg 2010;76:793-802. 2. Mansour KA, Thourani VH, Losken A, Reeves JG, Miller JI Jr, Carlson GW, et al. Chest wall resections and reconstruction: a 25-year experience. Ann Thorac Surg 2002;73:1720-5. 3. Deschamps C, Tirnaksiz BM, Darbandi R, Trastek VF, Allen MS, Miller DL, et al. Early and long-term results of prosthetic chest wall reconstruction. J Thorac Cardiovasc Surg 1999;117:588-91. 4. Pfannschmidt J, Geisb€ usch P, Muley T, Hoffmann H, Dienemann H. Surgical resection of secondary chest wall tumors. Thorac Cardiovasc Surg 2005;53:234-9. 5. Thomas PA, Brouchet L. Prosthetic reconstruction of the chest wall. Thorac Surg Clin 2010;20:551-8. 6. Granetzny A, Abd El-Aal M, Emam E, Shalaby A, Boseila A. Surgical versus conservative treatment of flail chest. Evaluation of the pulmonary status. Interact Cardiovasc Thorac Surg 2005;4:583-7. 7. Lardinois D, M€ uller M, Furrer M, Banic A, Gugger M, Krueger T, et al. Functional assessment of chest wall integrity after methylmethacrylate reconstruction. Ann Thorac Surg 2000;69:919-23. 8. Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health care–associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008;36:309-32. 9. Weyant MJ, Bains MS, Venkatraman E, Downey RJ, Park BJ, Flores RM, et al. Results of chest wall resection and reconstruction with and without rigid prosthesis. Ann Thorac Surg 2006;81:279-85. 10. Losken A, Thourani VH, Carlson GW, Jones GE, Culbertson JH, Miller JI, et al. A reconstructive algorithm for plastic surgery following extensive chest wall resection. Br J Plast Surg 2004;57:295-302. 11. Chang RR, Mehrara BJ, Hu QY, Disa JJ, Cordeiro PG. Reconstruction of complex oncologic chest wall defects: a 10-year experience. Ann Plast Surg 2004;52:471-9.
DISCUSSION Dr Frederick Luchette (Maywood, IL): Your results in this difficult patient population are to be commended. Although I am not a thoracic surgeon, rather just a general surgeon, a few questions do come to mind. Your outcome measures were limited to immediate postoperative complications, such as ventilator days, pneumonia, mesh infection, need for reoperation, and surgical site infection. Despite a mean follow-up of
596 Hanna et al
42 months, with no incidence of chronic pain, nerve injury, motion impairments, or cosmetic defects, you did not comment on the patient’s long-term pulmonary function status, which I think is a critical variable. And I recognize it’s a retrospective study. One would assume that with any size defect, but particularly the large defects, there will be development of some degree of paradoxical chest wall motion, even with a soft mesh. Can you give us any information on their pulmonary status? Were they impaired? Did they have restriction and loss of exertion? Although the incidence of mesh infection was low, this is typically a catastrophic complication for the patient. How do you manage this situation? If you have to take out the mesh, what do you do? Finally, nearly half the patients in your study with a large defect had 1 pulmonary comorbidity, that being smokers, COPD, and prior pulmonary resection. Were these the patients who required prolonged intubation in the ICU, or do you have any information on your selection criteria with these parameters? Dr Wael Hanna (Montreal, QC): In terms of the first question, pulmonary function tests would be possible to get in a prospective study, where you plan to look at that after your surgery. And there have been a couple of studies with methacrylate, where they did look at PFTs preop and postoperation. And the problem with that was twofold. One, most patients, because of pain in the chest wall, could not comply with the measurements of the PFT in the early postoperative period. Second, most pulmonologists did not know what to do with the PFTs when the big part of the chest wall was reconstructed like that, because they didn’t have enough data on whether PFTs would actually be indicative of what their pulmonary function is like or not. The second confounding factor in this is that historically, most of these patients who got chest wall reconstruction got lung resection as well. Not in our study, but in most of the studies in literature, a lot of these patients had a lung removed. And, so, that is a major confounder in terms of your pulmonary function tests. So, I think the thoracic surgery community agrees that PFTs are not the ideal measure to look at pulmonary function post-reconstruction of the chest wall. What are good measures is oxygen requirement at home, if you move to a state where you actually are so impaired that you need oxygen at home, or limitation of daily activities, as in you cannot go up a couple of flights of stairs or you are restricted in daily motion because of shortness of breath. In terms of mesh infection, it is true that it is a highly morbid situation, especially when methacrylate or metallic prosthesis is used, because there is no way of treating that infection with antibiotics when you have a rigid prosthesis. But, when you have a soft prosthesis, especially something like Vicryl or Gore-Tex, we do see a lot of times that these people do respond to antibiotics.
Surgery October 2011
And you saw that our reoperation rate in the large defect group was much lower than in our postoperative infection rate. And this is because we do not have to go back to the operating room to remove the mesh very often, because in a lot of these meshes you can sterilize with IV antibiotics. Obviously, it’s a prolonged course of IV antibiotics. And if there’s an abscess, you have to drain it. However, most of the patients did not go back to the operating room, which is a big distinction from the methacrylate age. In terms of pulmonary comorbidities, it is true that the patients who had COPD were the ones who are more likely to require ventilation. Sometimes the anesthetist would actually say, ‘‘This is a patient with bad pulmonary function preop. I am going keep them intubated for a couple of days.’’ And, so, that was a confounding factor. However, we did not reach any statistical significance in terms of the rate of extubation between mesh and nonmesh in our large defect cohort. So, it did play a role. The mean duration in the ICU was 2 days in our population. So, it wasn’t dramatic. However, I do recognize that it might have played a role in that. Dr Gerald Larson (Louisville, KY): I think a couple things are significant. They have shown that reconstruction with myocutaneous flaps offers the same postoperative outcome as when a soft mesh, such as Mylex or Prolene mesh is used. And I think, very significant, their results have held up for at least 42 months, which is a significant length of time to evaluate these patients. I do have 3 questions for the authors. One, in 75% of your patients in the large defect group, closure was obtained with a myocutaneous flap only. So, presumably you had to close the other 25%, probably seven or so, with mesh only. What was your technique for that? And how did you get a good cosmetic result and presumably skin coverage? My second question is: why not use a combination of a mesh and flap on all patients, since your results seem to be quite equivalent, or is the risk of infection such a great concern for you? And you touched on this briefly, but I think it’s worth re-emphasizing. When is the mesh absolutely needed? Dr Wael Hanna (Montreal, QC): In terms of the first question, closure, in our study, at least, was largely the decision of the time and place, at the time of resection, what tissues would be available. The ideal way to close is to mobilize enough muscle present in the chest wall, so latissimus dorsi or serratus or pectoralis major, mobilize it enough to reapproximate the muscle, then close with skin, without having to rotate a pedicled flap. And for the 25% you’re talking about. These are closed without any mesh, without any flap, just the local muscles mobilized enough to be reapproximated without any tension. The rest, when you cannot do that, you have to fill the defect with something. And that was really up to the thoracic surgeon at that time to decide whether there was mesh or not. Our surgeons tended to decide, for bigger defects, that they would use a mesh, and that was put. But in 75%
Surgery Volume 150, Number 4
of the time, on top of the mesh, we still mobilized a myocutaneous flap to cover the mesh because of the theoretical perception that a flap on top of the mesh would provide a good oxygenated environment and would decrease the rate of infection. In terms of the second question, why not reconstruct everyone with a mesh and a flap? It is that theoretical risk of infection. It is true that, in our study, we did not reach any statistical significance in terms of postoperative surgical site infection, but you have to keep in mind that our numbers are low. It is as good as it gets, but it’s still 37 patients. So over the next 20 years, if you have 100 patients, you may reach a certain statistical significance in terms of infection of the prosthesis. We also don’t know what’s going to happen beyond 48 months or 42 months, which is our mean follow-up. So, if these are benign tumors, like desmoids, that do not have any immediate mortality from their disease, they could come back 10 years later with a mesh infection, but we still won’t know that. We know it is a risk for methacrylate, but we don’t know that for soft mesh. So, if we could avoid reconstruction with a prosthesis, we would do it, especially if the results of mesh vs no mesh are equivalent. Dr Steven Steinberg (Columbus, OH): Did any of your patients who were extubated immediately after surgery then subsequently have to be reintubated? One of the things we have noticed in our institution is that, particularly in the patients with significant chronic lung disease, they are pushed to get them extubated in the operating room, but then some of them wind up coming back to the ICU later because of respiratory failure.
Hanna et al 597
More importantly, I would like to ask you a question about one of your conclusions. In the large defect group, you’ve concluded there’s really no difference in outcome as far as early extubation in the group that got mesh vs that group that didn’t. However, the gross numbers, I think there was 25% failure of extubation in the autologous tissue group vs about 8% in the mesh group. Your numbers may not be big enough to get a statistically significant difference. That seems to me like it might be clinically significant. I wonder if you would discuss that. Dr Wael Hanna (Montreal, QC): To answer your first question, yes, indeed, we did have 1 patient who was extubated preemptively, got reintubated on postop day 1, and remained intubated for 3 days in the ICU. But that was the only patient we had. The rest, I guess we were too cautious with and had to go to the ICU for monitoring of the flap anyway; so, they remained intubated. So, we didn’t have the problem of extubating patients earlier than we needed to. The second point is very well made. We do have the clinical significance, I think. And this is what started the whole project. This is why we decided to look at that. And we would love to say that if you do not use mesh, then you have a higher chance of remaining intubated in the ICU, except that the statistics won’t allow us to do that. But, I do agree that if we looked at bigger numbers, we may have a certain statistical significance. And this is why it is important not to categorically eliminate the role of mesh reconstruction in chest wall defects. But, I guess we have to look at that on a patient-by-patient basis and decide at the time of the reconstruction. But, I do agree that there is certainly clinical significance of impaired respiratory dynamics, at least in the early postoperative period.
Background. Rigid fixation is advocated as the best method to achieve good respiratory outcomes after chest wall resection at the expense of a high complication rate. The following study aims to examine the role of myocutaneous pedicled flaps, with or without soft prosthesis, in the reconstruction of small and large chest wall defects. Methods. All patients who underwent resection of chest wall tumors between 2003–2010 were identified from a prospectively entered database. Operative and postoperative outcomes were documented. Patients were stratified into 2 separate groups based on the size of the residual chest wall defect; the Small Defect (SD) group (<60 cm2) and the Large Defect (LD) group (>60 cm2). Results. Thirty-seven patients were identified over a 7-year period: 9 in the SD group and 28 in the LD group. Primary sarcoma was the most common indication for resection (57%). The mean size of the chest wall defect was 50.8 cm2 in the SD group and 149.4 cm2 in the LD group (P = .001). All patients underwent reconstruction with autologous tissue, nonrigid prosthesis, or a combination of the two. Prosthesis was used in 11% of patients in the SD group and 61% of patients in the LD group (P = .018). The rate of immediate postoperative extubation was 100% in the SD group and 89% in the LD group (P = .42). The rate of postoperative pneumonia was 7% in the LD group vs 0% in the SD group. The rate of surgical site infection was 7% in the LD group and 0% in the SD group. A subgroup analysis of the LD group demonstrated no statistical differences in any of the measured outcomes between patients in whom mesh prosthesis was used and patients in whom a myocutaneous flap alone was used. However, there was a clinical suggestion of prolonged ventilation in the subgroup where mesh was not used and of higher infection rates in the subgroup where mesh was used. Conclusion. Small chest wall defects can be reconstructed with pedicled myocutaneous flaps alone without compromising respiratory outcomes. In carefully selected patients with moderate size defects larger than 60cm2, reconstruction with pedicled myocutaneous flap alone offers similar postoperative outcomes as reconstruction with nonrigid prosthesis, at the expense of a possible need for a short period of mechanical ventilation. (Surgery 2011;150:590-7.) From the McGill University Health Centre, Montreal General Hospital, Montreal, Quebec, Canada
THE PRIMARY CONCERN after chest wall resection is the maintenance of the stability and integrity of the chest wall.1 A wide excision often results in a large chest wall defect with instability, disruption of pulmonary mechanics, and considerable respiratory complications. These sequelae are particularly noted when multiple ribs are resected or when Accepted for publication July 11, 2011. Reprint requests: Wael C. Hanna, MD, MBA, The Montreal General Hospital, 1650 Cedar Avenue, Rm L9-112, Montreal, QC H3G 1A4, Canada. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2011 Mosby, Inc. All rights reserved. doi:10.1016/j.surg.2011.07.055
590 SURGERY
the sternum is resected.2 Reconstruction also protects chest organs, prevents entrapment of the scapula, and maintains the cosmetic shape of the chest wall.3 It is generally accepted that defects less than 5 cm in diameter usually do not require reconstruction. Posterior defects less than 10 cm in diameter can also be safely closed without reconstruction, provided there is no risk of entrapment of the tip of the scapula.4 For larger defects, no consensus exists on the ideal method of reconstruction and varied results have been reported in the surgical literature.5 Rigid fixation with methacrylate or metallic prostheses has traditionally been advocated as the
Surgery Volume 150, Number 4
best method to achieve chest wall stability and prevent respiratory complications.4-6 However, it is frequently complicated by infection, pain, and rigid deformities.7 Furthermore, should an infection arise, the presence of rigid foreign material significantly complicates management. Thus, an ideal repair would be one that adequately maintains chest wall stability to prevent pulmonary complications but avoids rigid fixation. The purpose of this study is to examine the role of myocutaneous pedicled flaps, with or without soft prosthesis, in the reconstruction of small and large chest wall defects. MATERIALS AND METHODS After ethics board approval, all patients who underwent major resection of primary or metastatic chest wall tumors at a university-affiliated tertiary care institution between 2003 and 2010 were identified from a prospectively entered database. All patients with primary lung or pleural cancers invading the chest wall were excluded to eliminate potential confounding factors, as respiratory complications were a major outcome of interest. Among the patients included, 4 patients had undergone previous unrelated lung resections and 3 patients underwent 2 separate chest wall resections with reconstruction. Through a retrospective chart review, information was recorded on patient characteristics including age, sex, nature of the chest wall tumor, and presence of pulmonary comorbidity. Treatment with neoadjuvant radiation or chemotherapy was also recorded. Operative and pathology reports were reviewed for the number of ribs resected, concurrent sternal or clavicular resection, size of the residual defect, and method of reconstruction. A linear regression analysis identified a cut-off defect size as an independent predictor for reconstruction with mesh, and therefore patients were stratified into 2 groups: those who had a Small Defect area <60 cm2 (SD) and those who had a Large Defect area $60 cm2 (LD). Only nonrigid mesh composed of polypropylene (Marlex, Davol, Warwick, RI, and Prolene, Ethicon, Inc, Somerville, NJ), polyesther (Mersilene, Ethicon, Inc, Somerville, NJ), polytetrafluoroethylene (Gore-Tex and Gore-Dualmesh, W.L. Gore & Associates, Inc, Flagstaff, AZ,) and polyglactin-910 (Vicryl, Ethicon, Inc, Somerville, NJ) was used. No patients underwent rigid fixation with methyl methacrylate or metal plates. All postoperative complications were identified from the prospectively entered database. Pulmonary complications included pneumonia and
Hanna et al 591
admission to the intensive care unit (ICU) for assisted ventilation. Pneumonia was defined according to the Center for Disease Control and Prevention guidelines.8 Assisted ventilation was defined as the need for mechanical ventilation with either invasive or noninvasive means. Surgical site infection was defined by the need to open and drain the surgical site, regardless of whether bacteriology was obtained. Mesh infection was defined as a fluid collection in the vicinity of the mesh with proven bacteria on Gram stain or culture. The primary outcome of interest was the rate of pulmonary complications. Secondary outcomes included the rate of surgical site infection, mesh infection, need for reoperation, length of hospitalization, and in-hospital mortality. Data are presented as median (range) and the Fisher exact test or Mann-Whitney U test was used to determine statistical significance (P < .05). RESULTS A total of 37 patients were included in the study. Sixty-two percent (62%) were male and the median age for all patients was 49 (range, 19–75) years. Nine patients were stratified into the SD group and 28 patients in the LD group. Four patients (11%) had undergone a chest wall resection in the past, all of them in the LD group. A multidisciplinary team including thoracic, plastic, and orthopedic surgeons determined the method of reconstruction for each patient. Although this decision was based on tumor size, location on the chest wall, number of ribs to be resected, and feasibility of mobilizing autologous tissue flaps, surgeon preference and judgment played a major role in selection; Figure. In the SD group, 33% of patients were smokers but none had a diagnosis of Chronic Obstructive Lung Disease. In the LD group, 21% of patients were smokers and 11% had a diagnosis of Chronic Obstructive Lung Disease based on pulmonary function tests. Primary chest wall sarcoma was the most common indication for resection (57%), followed by metastases to the chest wall (22%) and desmoid tumors (14%). The mean size of the defect created in the chest wall was 50.8 cm2 (range, 14–59) in the SD group and 149.4 cm2 (range, 60–405) in the LD group (P = .001). The incidence of partial sternal resection was similar in both groups (33% in LD group vs 27% in SD group, P = .12). In the LD group, 50% of patients had 3 or more ribs resected, whereas none did in the SD group (P = .007). Neoadjuvant chemotherapy was given to 18% of patients in the LD group, vs none in the SD group (P = .05). Neoadjuvant
592 Hanna et al
Surgery October 2011
Fig. Illustration of the steps for surgical reconstruction after chest wall resection. A, Preoperative appearance of chest wall tumor. B, Chest wall defect after resection. C, First step of reconstruction with soft prosthesis. D, Second step of reconstruction with myocutaneous flap and final appearance.
Table I. Demographic, tumor, and treatment characteristics for all patients who underwent chest wall resection with reconstruction Variable Patient characteristics Mean age Male (%) Smoking (%) COPD (%) Previous lung resection (%) Tumor characteristics Sarcoma (%) Metastases (%) Desmoid (%) Neurofibromas (%) Treatment characteristics Neoadjuvant chemotherapy (%) Neoadjuvant radiotherapy (%) Neoadjuvant chemoradiation (%)
All patients
Small defect cohort
Large defect cohort
n = 37
n=9
n = 28
P value
49 (19–75) 62 24 8 8
45 (19–71) 56 33 0 0
47 (35–75) 64 21 11 11
.13 .29 .19 .22 .22
57 22 14 8
78 11 0 11
50 25 18 7
.21 .19 .22 .19
14 11 5
0 0 0
18 14 7
.24 .22 .13
radiotherapy was given to 14% in the LD group vs none in the SD group (P = .05, Table I). There was no significant difference between the 2 groups in terms of usage of autologous muscle flaps, which were constructed in 56% of patients in the SD group and 75% of patients in the LD
group (P = .24). When a muscle flap was constructed, pectoralis major was used in 41% of the cases and latissimus dorsi in 16%. Soft prosthesis was used much more frequently in the LD group (61% of patients) than in the SD group (11%, P = .018). One quarter of patients in the LD group
Hanna et al 593
Surgery Volume 150, Number 4
Table II. Resection and reconstruction parameters for both cohorts of patients Variable Resection parameters Mean size of chest wall defect (cm2) More than 2 ribs resected (%) Partial sternal resection (%) Total sternal resection (%) Clavicle resection (%) Lung resected (%) Diaphragm resected (%) Vertebra resected (%) Reconstruction parameters Use of soft prosthetic mesh (%) Use of muscle flap (%) Pectoralis major (%) Latissimus dorsi (%) Serratus (%) Rectus abdominis (%) Diaphragm (%)
All patients
Small defect cohort
Large defect cohort
n = 37
n=9
n = 28
P value
100.14 38 22 3 11 8 5 3
50.84 0 33 0 11 11 0 0
149.45 50 18 4 11 7 7 4
.001 .04 .21 .19 .12 .13 .31 .41
49 70 41 16 8 5 5
11 56 80 40 0 0 20
61 75 52 19 14 10 5
.018 .14 .19 .31 .43 .31 .22
Statistically significant P values in bold.
Table III. Postoperative outcomes in both cohorts of patients Variable Postoperative outcomes Immediate extubation postoperatively (%) Admission to the ICU For ventilatory support (%) For flap monitoring (%) Median number of days ventilated Median length of stay in ICU Postoperative pneumonia (%) Surgical site infection (%) Re-operation for infection or bleeding (%)
All patients
Small defect cohort
Large defect cohort
n = 37
n=9
n = 28
P value
100 0
89 32% 44 56 2 (1–7) 2 (1–9) 11 7 11
.42
92 24%
(25%) required a skin graft to cover their reconstruction, whereas none did in the SD group (P = .11, Table II). The rate of immediate extubation postoperatively was 100% in the SD group vs 89% in the LD group (P = .32). Patients remained intubated postoperatively for a variety of reasons, including prolonged operative time, magnitude of reconstruction, the need for flap monitoring, and analgesia in the absence of an adequately functioning epidural catheter. All patients were extubated successfully by the second day after the operation. One more patient was re-intubated on the third day postoperatively due to deterioration in respiratory status and remained intubated for 7 days. The rate of postoperative admission to the ICU was 0%
8 5 8
0 0 0
.31 .29 .31
in the SD group and 32% in the LD group (P = .06). Of the patients admitted to the ICU, half (56%) were admitted for monitoring of the muscle flap and did not require ventilator support. The rest (44%) were admitted for mechanical ventilation. Of those patients, 75% had undergone reconstruction via autologous tissue alone, without the use of soft prosthesis. The median number of days on the ventilator was 2,1-7 and the median length of stay in the ICU was 2 days.1-9 Two patients developed postoperative pneumonia, both in the LD group (7% vs 0% in the SD group, P = .56). Surgical site infection developed in 7% of patients in the LD group but not in any patient in the SD group (P = .56). One patient in the LD group required reoperation for mesh removal due to
594 Hanna et al
Surgery October 2011
Table IV. Sub-group analysis for postoperative outcomes in patients of the large defect cohort Variable 2
Mean size of chest wall defect (cm ) Use of muscle flap (%) Immediate extubation postoperatively (%) Admission to the ICU (%) For ventilatory support (%) For flap monitoring (%) Surgical site infection (%) Re-operation for infection or bleeding (%)
MESH
NO MESH
n = 16
n = 12
P value
161.9 75 100 19 33 67 13 19
131.72 75 75 42 60 40 0 0
.29 .25 .15 .21 .19 .13 .14 .14
MESH indicates the subgroup of patients where soft prosthesis was used for reconstruction vs NO MESH, which indicates the subgroup of patients where reconstruction was done with autologous tissue alone.
infection, while the remainder of the patients were treated conservatively either with local intervention or systemic antibiotics. Another patient in LD group required reoperation for significant bleeding, rendering the rate of reoperation in the LD group higher than in the SD group (7% vs 0% respectively, P = .56). Median length of stay in hospital for the SD group was 6 days (range, 5–10) compared to 9 days (range, 7–15) in the LD group (P = .41). Two patients in the LD group (7%) required readmission for pulmonary embolus and pneumonia respectively. There were no mortalities within the postoperative period of 30 days, Table III. A subgroup analysis of the LD cohort was done aiming to elucidate any differences in postoperative outcomes for patients in whom a prosthesis was used (MESH) vs those in whom reconstruction was done with autologous tissue alone (NO MESH), Table IV. In the NO MESH cohort, the incidence of admission to the ICU for ventilator support was higher than in the MESH group (60% vs 33%, P = .19). In contrast, the rate of surgical site infection and reoperation was higher in the MESH group. No statistical significance could be achieved in any of these parameters. Patients were seen postoperatively every 3 months for the first 2 years after operation, then at 6-month intervals thereafter, with a mean followup period of 42 months. During every visit, the patient would be examined by visual inspection of the scar, palpation of the chest wall, auscultation for lung breath sounds, and provocative coughing. Although no questionnaires were submitted to address specific postoperative outcomes, the visit notes reported no incidences of dyspnea at home, exercise limitation, oxygen requirement, chronic cough, chronic pain, nerve injury, motion impairments, or cosmetic defects as perceived by the
patient. Patients did not receive postoperative computed tomography (CT) scans on a routine basis. Of those who did get CT scans for oncological follow-up, none had reported lung hernias. DISCUSSION In our institution over the past 7 years, chest wall reconstruction has been performed exclusively using autologous muscle flaps or nonrigid mesh. This choice was part of our strategy to abandon rigid fixation due to largely disappointing results in terms of prosthesis infection, postoperative pain, and chest wall deformities. In our series, the immediate postoperative extubation rate was comparable to most series that used rigid fixation with methyl methacrylate.7,9 Those same studies demonstrate a 7–10% incidence of pleural effusions requiring drainage, whereas in our series we had none. We have also shown that nonrigid mesh prosthesis reconstruction carries a lower rate of surgical site infection (5%) than does rigid fixation (7– 20%).7,9-11 More importantly, we have documented no prosthesis infections that occurred later than 30 days after surgery, whereas delayed infection with methyl methacrylate has been reported as late as 12 months after surgery.7 It is important to note that the rate of reoperation for mesh infection in our series (3%) was considerably lower than what is usually associated with rigid fixation (8–12%).7,9 Although in our series, nonrigid reconstruction carried a slightly higher rate of postoperative pneumonia than what is historically reported for rigid fixation, the postoperative length of stay in hospital was considerably shorter.7 On long-term follow-up of 42 months, we have not recorded any incidence of chronic pain, neurologic impairment, or cosmetic or motion deformities associated with soft prosthesis reconstruction of the chest wall. This is in contrast with rigid fixation where long-term
Hanna et al 595
Surgery Volume 150, Number 4
pain, deformity, and cosmesis have been major issues with this type of reconstruction. Therefore, our data suggests that when the use of prosthetic material cannot be avoided, nonrigid prosthesis may offer more favorable short-term and long-term outcomes than what was historically demonstrated in the literature on rigid fixation. Furthermore, our data suggest that construction of a myocutaneous flap is crucial to the realization of good outcomes in patients with large residual chest wall defects. Within our cohort of patients with large defects, our surgeons elected to use a myocutaneous flap in 75% of large defects, even when a prosthesis was used. Although our data suggests that the incidence of admission to the ICU and the incidence of postoperative pneumonia may be higher when myocutaneous flap was used alone without mesh, this difference does not achieve statistical significance. As expected, the incidence of surgical site infection and reoperation is considerably lower when myocutaneous flap is used alone without mesh, again without any statistical significance. It is important to emphasize, nevertheless, that the small sample size in this subgroup analysis may prohibit the generation of any statistically significant results, even when clinical significance is present. Our data therefore suggests that, for chest wall defects larger than 60 cm2, reconstruction can be done without the use of prosthetic material, with little difference in postoperative outcomes but a possibility of prolonged ventilation requirements. It is important to note that defect size alone is not the only determinant of whether mesh should be used in reconstruction or not. The location of the chest wall defect (anterior vs posterior), the number of ribs removed, and the body habitus of the patient (thin vs obese) are factors that cannot be ignored when planning reconstruction. The present study is limited by several factors. Although information was rigorously collected from a prospectively entered database, the retrospective nature of this study predisposes the data to strong selection bias, especially regarding the choice of reconstructive technique and operative planning. The small number of patients included in this study is a reflection of the rarity of elective chest wall resection for primary tumors, but it also implies that general conclusions have to be extrapolated from a small sample size. Finally, although postoperative long-term outcomes were very well documented on follow-up, the lack of standardized questionnaires and pulmonary function test makes this data difficult to assess objectively and compare across all patients.
In conclusion, we have demonstrated that in patients with chest wall defects smaller than 60 cm2, there appears to be little benefit of reconstruction with a prosthesis, because postoperative outcomes with primary closure of the chest wall are excellent. In moderate size defects larger than 60 cm, and in carefully selected patients, reconstruction with pedicled myocutaneous flap alone offers similar postoperative outcomes as reconstruction with nonrigid prosthesis, at the expense of a possible need for a short period of mechanical ventilation.
REFERENCES 1. Nirula R, Mayberry JC. Rib fracture fixation: controversies and technical challenges. Am Surg 2010;76:793-802. 2. Mansour KA, Thourani VH, Losken A, Reeves JG, Miller JI Jr, Carlson GW, et al. Chest wall resections and reconstruction: a 25-year experience. Ann Thorac Surg 2002;73:1720-5. 3. Deschamps C, Tirnaksiz BM, Darbandi R, Trastek VF, Allen MS, Miller DL, et al. Early and long-term results of prosthetic chest wall reconstruction. J Thorac Cardiovasc Surg 1999;117:588-91. 4. Pfannschmidt J, Geisb€ usch P, Muley T, Hoffmann H, Dienemann H. Surgical resection of secondary chest wall tumors. Thorac Cardiovasc Surg 2005;53:234-9. 5. Thomas PA, Brouchet L. Prosthetic reconstruction of the chest wall. Thorac Surg Clin 2010;20:551-8. 6. Granetzny A, Abd El-Aal M, Emam E, Shalaby A, Boseila A. Surgical versus conservative treatment of flail chest. Evaluation of the pulmonary status. Interact Cardiovasc Thorac Surg 2005;4:583-7. 7. Lardinois D, M€ uller M, Furrer M, Banic A, Gugger M, Krueger T, et al. Functional assessment of chest wall integrity after methylmethacrylate reconstruction. Ann Thorac Surg 2000;69:919-23. 8. Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health care–associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008;36:309-32. 9. Weyant MJ, Bains MS, Venkatraman E, Downey RJ, Park BJ, Flores RM, et al. Results of chest wall resection and reconstruction with and without rigid prosthesis. Ann Thorac Surg 2006;81:279-85. 10. Losken A, Thourani VH, Carlson GW, Jones GE, Culbertson JH, Miller JI, et al. A reconstructive algorithm for plastic surgery following extensive chest wall resection. Br J Plast Surg 2004;57:295-302. 11. Chang RR, Mehrara BJ, Hu QY, Disa JJ, Cordeiro PG. Reconstruction of complex oncologic chest wall defects: a 10-year experience. Ann Plast Surg 2004;52:471-9.
DISCUSSION Dr Frederick Luchette (Maywood, IL): Your results in this difficult patient population are to be commended. Although I am not a thoracic surgeon, rather just a general surgeon, a few questions do come to mind. Your outcome measures were limited to immediate postoperative complications, such as ventilator days, pneumonia, mesh infection, need for reoperation, and surgical site infection. Despite a mean follow-up of
596 Hanna et al
42 months, with no incidence of chronic pain, nerve injury, motion impairments, or cosmetic defects, you did not comment on the patient’s long-term pulmonary function status, which I think is a critical variable. And I recognize it’s a retrospective study. One would assume that with any size defect, but particularly the large defects, there will be development of some degree of paradoxical chest wall motion, even with a soft mesh. Can you give us any information on their pulmonary status? Were they impaired? Did they have restriction and loss of exertion? Although the incidence of mesh infection was low, this is typically a catastrophic complication for the patient. How do you manage this situation? If you have to take out the mesh, what do you do? Finally, nearly half the patients in your study with a large defect had 1 pulmonary comorbidity, that being smokers, COPD, and prior pulmonary resection. Were these the patients who required prolonged intubation in the ICU, or do you have any information on your selection criteria with these parameters? Dr Wael Hanna (Montreal, QC): In terms of the first question, pulmonary function tests would be possible to get in a prospective study, where you plan to look at that after your surgery. And there have been a couple of studies with methacrylate, where they did look at PFTs preop and postoperation. And the problem with that was twofold. One, most patients, because of pain in the chest wall, could not comply with the measurements of the PFT in the early postoperative period. Second, most pulmonologists did not know what to do with the PFTs when the big part of the chest wall was reconstructed like that, because they didn’t have enough data on whether PFTs would actually be indicative of what their pulmonary function is like or not. The second confounding factor in this is that historically, most of these patients who got chest wall reconstruction got lung resection as well. Not in our study, but in most of the studies in literature, a lot of these patients had a lung removed. And, so, that is a major confounder in terms of your pulmonary function tests. So, I think the thoracic surgery community agrees that PFTs are not the ideal measure to look at pulmonary function post-reconstruction of the chest wall. What are good measures is oxygen requirement at home, if you move to a state where you actually are so impaired that you need oxygen at home, or limitation of daily activities, as in you cannot go up a couple of flights of stairs or you are restricted in daily motion because of shortness of breath. In terms of mesh infection, it is true that it is a highly morbid situation, especially when methacrylate or metallic prosthesis is used, because there is no way of treating that infection with antibiotics when you have a rigid prosthesis. But, when you have a soft prosthesis, especially something like Vicryl or Gore-Tex, we do see a lot of times that these people do respond to antibiotics.
Surgery October 2011
And you saw that our reoperation rate in the large defect group was much lower than in our postoperative infection rate. And this is because we do not have to go back to the operating room to remove the mesh very often, because in a lot of these meshes you can sterilize with IV antibiotics. Obviously, it’s a prolonged course of IV antibiotics. And if there’s an abscess, you have to drain it. However, most of the patients did not go back to the operating room, which is a big distinction from the methacrylate age. In terms of pulmonary comorbidities, it is true that the patients who had COPD were the ones who are more likely to require ventilation. Sometimes the anesthetist would actually say, ‘‘This is a patient with bad pulmonary function preop. I am going keep them intubated for a couple of days.’’ And, so, that was a confounding factor. However, we did not reach any statistical significance in terms of the rate of extubation between mesh and nonmesh in our large defect cohort. So, it did play a role. The mean duration in the ICU was 2 days in our population. So, it wasn’t dramatic. However, I do recognize that it might have played a role in that. Dr Gerald Larson (Louisville, KY): I think a couple things are significant. They have shown that reconstruction with myocutaneous flaps offers the same postoperative outcome as when a soft mesh, such as Mylex or Prolene mesh is used. And I think, very significant, their results have held up for at least 42 months, which is a significant length of time to evaluate these patients. I do have 3 questions for the authors. One, in 75% of your patients in the large defect group, closure was obtained with a myocutaneous flap only. So, presumably you had to close the other 25%, probably seven or so, with mesh only. What was your technique for that? And how did you get a good cosmetic result and presumably skin coverage? My second question is: why not use a combination of a mesh and flap on all patients, since your results seem to be quite equivalent, or is the risk of infection such a great concern for you? And you touched on this briefly, but I think it’s worth re-emphasizing. When is the mesh absolutely needed? Dr Wael Hanna (Montreal, QC): In terms of the first question, closure, in our study, at least, was largely the decision of the time and place, at the time of resection, what tissues would be available. The ideal way to close is to mobilize enough muscle present in the chest wall, so latissimus dorsi or serratus or pectoralis major, mobilize it enough to reapproximate the muscle, then close with skin, without having to rotate a pedicled flap. And for the 25% you’re talking about. These are closed without any mesh, without any flap, just the local muscles mobilized enough to be reapproximated without any tension. The rest, when you cannot do that, you have to fill the defect with something. And that was really up to the thoracic surgeon at that time to decide whether there was mesh or not. Our surgeons tended to decide, for bigger defects, that they would use a mesh, and that was put. But in 75%
Surgery Volume 150, Number 4
of the time, on top of the mesh, we still mobilized a myocutaneous flap to cover the mesh because of the theoretical perception that a flap on top of the mesh would provide a good oxygenated environment and would decrease the rate of infection. In terms of the second question, why not reconstruct everyone with a mesh and a flap? It is that theoretical risk of infection. It is true that, in our study, we did not reach any statistical significance in terms of postoperative surgical site infection, but you have to keep in mind that our numbers are low. It is as good as it gets, but it’s still 37 patients. So over the next 20 years, if you have 100 patients, you may reach a certain statistical significance in terms of infection of the prosthesis. We also don’t know what’s going to happen beyond 48 months or 42 months, which is our mean follow-up. So, if these are benign tumors, like desmoids, that do not have any immediate mortality from their disease, they could come back 10 years later with a mesh infection, but we still won’t know that. We know it is a risk for methacrylate, but we don’t know that for soft mesh. So, if we could avoid reconstruction with a prosthesis, we would do it, especially if the results of mesh vs no mesh are equivalent. Dr Steven Steinberg (Columbus, OH): Did any of your patients who were extubated immediately after surgery then subsequently have to be reintubated? One of the things we have noticed in our institution is that, particularly in the patients with significant chronic lung disease, they are pushed to get them extubated in the operating room, but then some of them wind up coming back to the ICU later because of respiratory failure.
Hanna et al 597
More importantly, I would like to ask you a question about one of your conclusions. In the large defect group, you’ve concluded there’s really no difference in outcome as far as early extubation in the group that got mesh vs that group that didn’t. However, the gross numbers, I think there was 25% failure of extubation in the autologous tissue group vs about 8% in the mesh group. Your numbers may not be big enough to get a statistically significant difference. That seems to me like it might be clinically significant. I wonder if you would discuss that. Dr Wael Hanna (Montreal, QC): To answer your first question, yes, indeed, we did have 1 patient who was extubated preemptively, got reintubated on postop day 1, and remained intubated for 3 days in the ICU. But that was the only patient we had. The rest, I guess we were too cautious with and had to go to the ICU for monitoring of the flap anyway; so, they remained intubated. So, we didn’t have the problem of extubating patients earlier than we needed to. The second point is very well made. We do have the clinical significance, I think. And this is what started the whole project. This is why we decided to look at that. And we would love to say that if you do not use mesh, then you have a higher chance of remaining intubated in the ICU, except that the statistics won’t allow us to do that. But, I do agree that if we looked at bigger numbers, we may have a certain statistical significance. And this is why it is important not to categorically eliminate the role of mesh reconstruction in chest wall defects. But, I guess we have to look at that on a patient-by-patient basis and decide at the time of the reconstruction. But, I do agree that there is certainly clinical significance of impaired respiratory dynamics, at least in the early postoperative period.