Share this article on:

Sternum Sparing Thoracotomy Incisions in Lung Transplantation Surgery: A Superior Technique to the Clamshell Approach

Bittner, Hartmuth Bruno MD, PhD; Lehmann, Sven MD; Binner, Christian MD; Garbade, Jens MD; Barten, Markus MD; Mohr, Friedrich Wilhelm MD, PhD

Innovations:Technology and Techniques in Cardiothoracic and Vascular Surgery: March-April 2011 - Volume 6 - Issue 2 - p 116-121
doi: 10.1097/IMI.0b013e3182166163
Original Article

Objective: Bilateral anterior transsternal thoracotomy incision (clamshell technique) is the standard approach used for bilateral sequential lung transplantation (LTX). The morbidity and wound complications of this large incision can be considerable and costly. Muscle sparing anterior thoracotomies without division of the sternum may lead to decreasing the sequelae of wound complications. The objective of this study was to determine the rate of wound complications in the nonsternal incising lung transplant patients.

Methods: We used the single-institution–based transplant data bank, phone questionnaire, and ambulatory care unit follow-up data to investigate retrospectively the incidence of wound healing complications following muscle and sternum sparing and mammary artery protecting limited access small submammary anterior thoracotomy incisions (AT) for LTX surgery. In the need for cardiopulmonary bypass, the femoral artery and vein were cannulated.

Results: After exclusion of seven clamshell operations for LTX combined with cardiac surgery, 91 recipients (65% male), aged 19 to 68 years (mean, 54 ± 8 years), underwent 84 AT and 48 lateral thoracotomies (LT) for idiopathic pulmonary fibrosis (IPF) (48%), obstructive disease (40%), cystic fibrosis (CF) (5%), and pulmonary arterial hypertension (PAH) (7%). AT ranged from 5.5 to 26 cm (mean, 20.3 ± 4.8 cm) and LT from 12 to 25 cm (mean, 19.8 ± 2.4 cm) and was not significantly different (P = 0.37). Warm ischemic times ranged from 30 to 92 minutes (mean, 56 ± 11 minutes). Four patients required rethoracotomy for bleeding/hematoma formation. Cardiopulmonary bypass/intraop extracorporeal membrane oxygenation (ECMO) was used in 64%. Superficial wound infection and subsequent drainage/care was needed in four LTX incisions. Reoperation for lung herniation using patch repair technique for thoracic wall stabilization was required in two AT and three LT.

Conclusions: Sternum sparing and mammary artery protecting limited access submammary anterior and lateral thoracotomy incisions for LTX surgery are safe and effective. This approach avoids complications related to sternal transaction and may minimize the development of severe wound complications following LTX surgery.

From the Division of Thoracic and Cardiovascular Surgery, Heart Center Leipzig of the University of Leipzig, Leipzig, Germany.

Accepted for publication February 10, 2011.

Presented at the Annual Scientific Meeting of the International Society for Minimally Invasive Cardiothoracic Surgery, June 3–6, 2009, San Francisco, CA USA.

Address correspondence and reprint requests to Hartmuth B. Bittner, MD, PhD, Division of Thoracic and Cardiovascular Surgery, Heart Center Leipzig of the University of Leipzig, Struempell Str. 39, 04289 Leipzig, Germany. E-mail:

The ideal operative approach for lung transplantation (LTX) surgery has evolved continuously. Patterson et al1 pioneered successfully the first LTX procedures through a median sternotomy as en bloc double-lung replacement requiring a tracheal anastomosis under full cardiopulmonary bypass (CPB) support. However, tracheal anastomotic complications and poor anastomotic healing led to a shift of the airway anastomosis from the trachea to the main stem bronchi. CPB support could have been avoided for many double-lung transplantation (DLT) procedures when the diseased lungs were bilateral sequentially replaced. Bilateral sequential LTX became the procedure of choice, which is usually performed with a transsternal bilateral anterolateral thoracotomy or clamshell incision,2 which allows extensive exposure of both pleural spaces and mediastinum with access to the base of the heart for CPB. However, this approach is hampered by a number of complications related to transverse sternotomy (infection, dehiscence, pseudoartrosis, and malunion) and the division of the mammary vessels can be an additional source of bleeding. Predisposing factors for wound complications following LTX are sudden increases in intrathoracic pressure, rib or cartilage fractures, intercostal muscle weakness, surgical trauma, chronic immunosuppression therapy and steroid abuse, and in particular thoracotomy incisions with transection of the sternum (clamshell). Meyers et al3 avoided the transverse sternotomy in selected patients and performed DLT through two sequential anterolateral thoracotomies, avoiding wound problems in all 50 cases. The limited anterolateral thoracotomy incision for single left or right LTX or bilateral anterior thoracotomies without sternal division for sequential LTX became the standard approaches when this LTX program was initiated in 2003. The objective of this study was to determine the rate of wound complications in the nonsternal incising lung transplant patients.

Back to Top | Article Outline


The Institutional Data Use Committee at the Heart and Lung Transplantation Center of the University of Leipzig approved this study, and patients provided written and informed consent for data registration and analysis at institutional and national and international data registries (German Transplant Procurement Agency, Eurotransplant, and the Registry of the International Society for Heart and Lung Transplantation). From 2003 to 2009, 98 lung transplants were performed and seven of them by clamshell sternotomy. These patients were excluded from analysis because LTX was combined with cardiac surgery. Ninety-one recipients (65% male) who underwent LTX, aged 19 to 68 years (mean, 54 ± 8 years), underwent 84 anterior thoracotomy incisions (AT) and 49 lateral thoracotomies (LT) for IPF (48%), obstructive disease (40%), CF (5%), and PAH (7%). The standard recipient immunosuppressive protocol was a triple medication regimen consisting of intravenous glucocorticoids, per os calcineurin inhibitor, and per os mycophenolate mofetil.4 The immunosuppressive medications were started 2 hours preoperatively in all recipients. This strict orally based immunosuppressive regimen was maintained postoperatively for a long term with an immediate switch from intravenous glucocorticoids to an oral steroid application.

Back to Top | Article Outline

Donor and Recipient Surgery

We used well-established criteria for accepting donor lungs, including objective evidence of adequate gas exchange and bronchoscopic evaluation to exclude aspiration or purulent secretions.5 Standardized organ procurement and recipient implantation techniques were used for LTX. A cold crystalloid preservation solution (low-potassium dextran solution; Vitrolife, Gottenberg, Sweden) was infused via the donor pulmonary artery at low pressure in an antegrade fashion immediately after prostaglandin intrapulmonary artery injection. During the procurement, the vascular structures were divided in situ, and the trachea was dissected well proximal to the carina. With the lungs partially inflated, the trachea was divided between staple lines, and the organ was transported to the center immersed in low-potassium dextran solution.

Access for bilateral sequential LTX is usually obtained through anterolateral thoracotomy incisions (clamshell technique). The ipsilateral chest is rolled anteriorly 45 degrees, with the ipsilateral arm abducted and the upper arm supported over the head. An inframammary incision is made, and the pectoralis major muscle and medial parts of the latissimus dorsi muscle are incised. The fourth or fifth intercostal space is entered, and the incision may be extended medially to the internal mammary artery and vein or sternum, which are transected if extension across the body of the sternum may be used. This incision provides exposure to the mediastinum, pericardium, hilum, lower lobe, and middle lobe or lingual.

The standard approach for single-lung transplantation is obtained by means of a posterolateral thoracotomy. The chest is in straight lateral position, with a roll placed under the dependent axilla and the upper arm extended anteriorly. The skin is incised along the inframammary line anterior to 1 cm below the tip of the scapula and the spine posterior. The serratus anterior and the latissimus dorsi muscles are divided. Alternatively, they may be spared by retracting them anteriorly and posteriorly. The forth or fifth intercostal space is entered, or the fifth rib may be removed and the pleural space is entered through the bed of the fifth rib. Additional exposure may be obtained by resecting a 1-cm section of rib above and below, preserving the intercostal neurovascular bundle. The trapezius and rhomboid muscles may be incised posteriorly as necessary up to the paraspinous muscle, which is spared.

In our study, the most common recipient operation performed was a single or bilateral sequential LTX accessed through the fourth (preferred for restrictive lung disease) or fifth (for obstructive end-stage lung disease) intercostal space thoracotomy without sternal deviation. DLT was performed using a sequential single-lung implantation technique through bilateral anterolateral thoracotomies without transverse incision of the sternum. It was the surgeons' preference to perform single-lung transplantation through an LT. The ipsilateral chest was rolled anteriorly, with the ipsilateral arm abducted. An inframammary incision was made protecting the mammary gland, and the pectoralis major muscle was incised. The fourth or fifth intercostal space was entered, and the incision may be extended medially to the internal mammary artery, as needed, and laterally to the latissimus dorsi muscle. Only the medial parts of the serratus anterior muscles were divided. The fourth or fifth rib was transected (as needed medially to the costal cartilage of the third or fourth rib). For a lateral thoracotomy incision, the ipsilateral chest was rolled anteriorly 45 degrees. The pectoralis major muscle and parts of the latissimus dorsi muscle and serratus anterior muscle were divided. After the insertion of the chest retractor and after retraction of the lung, the mediastinum, pericardium, and hilum were exposed.

The need for CPB was determined based on trial occlusion of the pulmonary artery with one lung ventilation as indicated by pulmonary pressures changes, intolerance of single-lung ventilation, or increasing hemodynamic instability. Once the donor lung was present in the operating room, the recipient pneumonectomy was completed. The bronchial anastomosis was accomplished first generally followed by the vascular pulmonary artery and left atrial cuff anastomoses. De-airing was done thoroughly through the atrial cuff anastomosis.

Two 28-Fr chest tubes are placed in each pleural space one to two interspaces below the incision (one angled posterior chest tube, and the other straight chest tube were placed anterior toward the apex of the lung). The incision was closed in layers using two to three strong monofilament suture loops (1 CTX 150 cm, PDS Plus, Polydioxanon; Ethicon, St-Stevens-Woluwe, Belgium) in a figure of eight technique around the rips for the reapproximation of the intercostal space. The serratus anterior, latissimus dorsi, and pectoralis major muscles with associated fascia were closed with continuous monofilament suture (2 CTX plus 90 cm PDS, Polydioxanon; Ethicon). The subcutaneous tissue was closed with continuous 2-0 Vicryl suture (2-0 SH plus 70 cm Vicryl plus, Polyglactin; Ethicon). The skin was closed with a skin stapler.

Back to Top | Article Outline


In this retrospective analysis, data were expressed as mean ± standard deviation of n number of observations; comparisons between the two groups were made by the Student t test and χ2 analysis. The a priori level of significance was at P < 0.05. Categorical data were analyzed by χ2 tests. The Kaplan-Meier analysis was used to evaluate long-term survival. Results are expressed as means ± standard errors. Significant differences are reported for P values <0.05.

Back to Top | Article Outline


There were 98 lung transplants performed; 7 of them were excluded from the study as they were in clamshell-treated patients. Three patients underwent combined heart-lung transplantation, and in addition to LTX surgery, four patients required coronary artery bypass grafting or aortic and tricuspid valve replacement/reconstruction performed through clamshell incisions. Therefore, 91 LTX patients (65% male), aged 19 to 68 years (mean, 54 ± 8 years), were analyzed retrospectively, experiencing the same surgical protocol for single or bilateral sequential LTX. The close follow-up of the LTX patients through frequent ambulatory care unit visits and close monitoring of the immunosuppressive regimen and the medication response was 100% complete. The mean duration of observation per patient was 1.8 ± 1.7 years (median, 1.4 years; range, 0.01–6.4 years), and this study included 176.5 patient-related years of follow-up. Using the Kaplan-Meier survival analysis, the 1-, 3-, and 5-year survival rate after LTX were 70%, 60%, and 55%, respectively. The 30-day and in-hospital mortality was 19.5%. Primary graft failure (severe reperfusion injury) and cytomegalovirus infection were the main causes of in-hospital death. The causes of death are depicted in Table 1 and were unrelated to the surgical approach used.



All the patients undergoing two separate thoracotomies without sternal division were placed on the operating table in the supine position with the arms abducted, slightly elevated with flexed elbow joints and the hand pointing toward the head of the patient (Fig. 1). Sternum sparing and mammary artery protecting (the incision was extended medially to the costal cartilage but still lateral to the course of the internal mammary artery and vein) limited access submammary anterior thoracotomy incisions were performed for sequential bilateral LTX surgery (Fig. 2). This allowed expanding the skin incision laterally toward the axillary line if needed (Fig. 3). Intraoperative close surveillance was accomplished with full respiratory monitoring, one central line with continuous cardiac output and pulmonary artery pressure monitoring, one arterial line, electrocardiogram, and transesophageal echocardiography. Before the thoracotomy, a longitudinal inguinal incision was performed in all patients, exposing femoral vein and artery for the possibility of immediate CPB support. Transplantation was usually started on the worse side based on preoperative computed tomography and ventilation/perfusion scans, with the table rotated toward the contralateral side. All the anterolateral thoracotomy incisions were performed through the fourth (restrictive lung disease) or fifth intercostals space (preferred in patients with predominantly obstructive end-stage lung disease and enlarged thoracic space) without dividing the mammary vessels (Fig. 4). If harvesting of the donor lungs required more time, the contralateral chest was also entered and the dissection of the hilum started while waiting for the harvesting team. Posterolateral thoracotomies were indicated for sequential bilateral LTX in two patients when mediastinal shifting led to difficult exposure of lung hilum structures requiring turning and redraping the patient.









LTX surgery was performed in all recipients in the standard way beginning with the bronchial anastomosis (Fig. 5). The airway anastomosis was usually performed with a 4-0 running monofilament suture (PDS, Polydioxanon; Ethicon) and following accomplishment, covered with peribronchial and mediastinal tissue. Only in cases of severe size-match telescoping, anastomosing techniques were used. This led to anastomotic strictures in 15% of patients requiring bronchoscopic dilatation and stenting. Bronchial anastomotic necrosis and associated bronchopleural fistulas were not observed. Exposing and mobilization of the lung hilum and completing the LTX procedure through the pulmonary artery and pulmonary vein-atrial cuff anastomoses were safe and unrestricted by the limited thoracotomy incisions (Fig. 3). Access to the great vessels of the heart and right atrium for CPB cannula insertion and connection, however, appeared to be very limited especially during left lung implantation surgery. Because the right anterior thoracotomy incision extended medially only to the costal cartilage and not reaching the sternum, the oblique angle and remaining distance to the aorta and even to the right atrium would make a secure central cannulation procedure for CPB difficult. In the event of extracorporeal circulation-supported LTX surgery, which was required in 64%, the already exposed femoral artery and vein were used. The preparation and exposition of the femoral vessel for extracorporeal circulatory support before thoracotomy in all LTX operation is a strict requirement and safety measure of our LTX protocol and relates very well to our vast experience with peripheral access of our large minimally invasive cardiac surgery program.



After implantation of the first lung, the retractors are removed, and the operatory field is covered with a wet surgical towel without closing the chest. The table was rotated toward the other side, and the operation was completed by the contralateral LTX performed in the same way. At the end of the procedure, hemostasis is ensured on both sides; if the donor lungs are too large to fit the recipient chest cavities, volume reduction is obtained performing multiple wedge resections (usually in the lingula, the middle lobe, the ventral segment of the upper lobes, and the apical segment of the lower lobes) with GIA staplers (Ethicon Inc., Somerville, NJ USA). In the majority of cases, the donor lung volume reduction was accomplished at the back-table in the recipient operating room with the allograft entirely deflated. The staple line was protected from the occurrence of air leakage through the application of autologous glue generated from 100 mL of blood of the recipient, which was collected after skin incision.

There were 84 AT and 49 LT for IPF (48%), obstructive disease (40%), CF (5%), and PAH (7%). AT ranged from 5.5 to 26 cm (mean, 20.3, ± 4.8 cm), and LT ranged from 12 to 25 cm (mean, 19.8 ± 2.4 cm) and was not significantly different (P = 0.37). The patients were asked to measure the length of the thoracotomy wound and scar and report the result in the phone questionnaires, or to assist the length measuring of the wound during the clinic follow-up visits. Ischemic times for single LTX (322 ± 59 minutes) and sequential bilateral LTX (399 ±135 minutes) were markedly longer compared with International Society for Heart and Lung Transplantation (ISHLT) data reporting. Warm ischemic times (time from back-table donor lung preparation to accomplishment of the implantation and release of the pulmonary cross clamp) ranged from 30 to 92 minutes (mean, 56 ± 11 minutes).

Four patients required rethoracotomy for bleeding/hematoma formation. CPB/intraop ECMO was used in 64%. None of the patients were closed in delayed fashion. Superficial wound infection and subsequent drainage/care was needed in four LTX incisions. A wound infection was defined as a collection of pus in the wound. There was no systemic disturbance, and the complication was completely resolved by simple aspiration, discharge of pus, and frequent dressing changes. A major wound infection, which is defined as one that makes the patient ill, was not observed. Reoperation for lung herniation using patch repair technique for thoracic wall stabilization was required in five patients. This was considered a serious wound complication. The development of these incisional hernias was unrelated to wound infection, and the formation was discovered after primary healing of the thoracotomy incision >8 weeks after transplantation. In two of these patients with underlying cystic fibrosis, the tissue might have been inadequate for strong closure. One patient reported longer periods of severe coughing. In the other patient, the cause for fascial separation was not apparent.

Back to Top | Article Outline


LTX has become the standard of care for most causes of end-stage lung disease and is an accepted therapeutic option for patients who are otherwise refractory to medical therapy. Based on marked improvements in surgical technique, anesthesia management, attenuation of the ischemia and reperfusion injury, and advancement in immunosuppression medications, LTX is now carried out routinely with low operative mortality. The standard surgical approach for DLT consists of bilateral anterior thoracotomies with transection of the sternum (clamshell incision) or thoracosternotomy leading to an open surgical field with nearly unlimited exposure. This exposure allows performing sequential LTX of two lungs (DLT) safely. The technique initially described by the Toronto Lung Transplant Group came from heart-lung transplantation and was performed on patients who underwent CPB, patients with routine cardiac arrest, extended dissection of the posterior mediastinum for the construction of a single tracheal, pulmonary artery, and common left atrial anastomosis. This operation carried a number of complications related to the surgical technique itself: bleeding, heart failure, and primary graft failure. Furthermore, the interruption of the bronchial circulation had an impact on tracheal healing resulting in an increased incidence of airway problems.6 The surgical technique for DLT underwent a number of modifications since it was first described by Cooper et al in 1987.7 Noirclerc et al8 described the feasibility of bilateral sequential bronchial anastomosis instead of tracheal anastomosis. Eventually, Pasque et al2 described the technique of bilateral sequential LTX that progressively became the procedure of choice when both lungs have to be transplanted. According to this technique, two separate single-lung transplants are performed in sequence through a clamshell incision. CPB is not routinely used, but when required, it can be safely instituted because the anterior mediastinum has already been dissected and the great vessels can be quickly prepared. However, the clamshell incision for itself may be associated with significant morbidity. Macchiarini et al9 evaluated the influence of sternotomy or clamshell incisions for double-lung or heart-lung transplantation on the lungs and chest wall. The clamshell incision was associated with significant more revisions, higher incidence of postoperative chronic pain, and worse postoperative mechanical properties. The overall 5-year survival of 57% was not influenced by the type of incision in 37 clamshell and 33 sternotomy patients.

In many instances, the sternum has been closed only with two or three loops of stainless steel sternal wire. Although it is generally satisfactory to prevent distraction of the ends, there is a tendency toward angulation and anterior displacement of the distal sternum (sternal overriding) and sternal malunion.10 Sternal overriding occurred in 12% or 32% of clamshell patients.9 The use of sternal plating can decrease the incidence of sternal malunion encountered with transverse sternotomy and attenuating the frequently observed severe pain experience associated with delayed or decreased functional recovery. Brown et al11 report a prevalence of 36% for sternal disruption in transverse bilateral thoracosternotomy for LTX in their experience, and they cite disruption rates of 20% to 60% at institutions worldwide.

An additional serious problem is that of deep sternal wound infection in the patient after transverse sternotomy. This is a life-threatening problem in an already immunosuppressed patient requiring frequent operative and bedside wound debridement with additional antibiotics and a prolonged hospital stay. Meyers et al3 estimated the prevalence for all sternal closure complications following sternal transection (thoracosternotomy or clamshell incision) in the range of 34%. For this reason, a less invasive surgical approach able to completely solve the problems related to transection of the sternum12 with the same operative results and reduced morbidity has been proposed.

Lung transplant recipients in our program had a markedly reduced wound healing and complication rate. The incidence of deep wound infections was 0%. Two of 86 lung transplant recipients had a superficial wound complication requiring ambulatory treatment and a drainage procedure at the bedside. However, serious thoracotomy wound problems were observed in a very limited number of patients with limited access and sternum sparing thoracotomy incisions requiring redo surgery. Lung transplant patients in five developed symptomatic lung herniation. They were taken back to the operating room electively undergoing thoracic wall restabilization through a patch plastic procedure. Notably, there was no associated mortality with the limited thoracotomy incisions.

In conclusion, the bilateral sequential anterolateral thoracotomy approach is less invasive than the clamshell incision and is certainly safe. We preserve the internal mammary vessels and avoid sternal splitting and dissection of the anterior mediastinum, which can be associated with deep sternal wound infection and life-threatening mediastinitis. Even if CPB is suddenly required and the two separate anterolateral thoracotomies can be easily converted into a clamshell incision, we still avoid splitting the sternum and subsequently institute safely extracorporeal circulation through femoral vessels cannulation.

Back to Top | Article Outline


1. Patterson GA, Cooper JD, Goldman B, et al. Technique of successful clinical double-lung transplantation. Ann Thorac Surg. 1988;45:626–633.
2. Pasque MK, Cooper JD, Kaiser LR, et al. Improved technique for bilateral lung transplantation: rationale and initial clinical experience. Ann Thorac Surg. 1990;49:785–791.
3. Meyers BF, Sundaresan RS, Guthrie T, et al. Bilateral sequential lung transplantation without sternal division eliminates posttransplantation sternal complications. J Thorac Cardiovasc Surg. 1999;117:358–364.
4. Bittner HB, Barten MJ, Binner C, et al. Preoperative introduction and maintenance immunosuppression therapy of oral-only tacrolimus, mycophenolate mofetil and steroids reduce acute rejection episodes after lung transplantation. Eur J Cardiothorac Surg. 2010;38:268–276.
5. Egan TM, Kaiser LR, Cooper JD. Lung Transplantation. Curr Probl Sur. 1989;26:675–751.
6. Patterson GA, Todd TR, Cooper JD, et al. Airway complications after double lung transplantation. Toronto Lung Transplant Group. J Thorac Cardiovasc Surg. 1990;99:14–21.
7. Cooper JD, Pearson FG, Patterson GA, et al. Technique of successful lung transplantation in humans. J Thorac Cardiovasc Surg. 1987;93:173–181.
8. Noirclerc MJ, Metras D, Vaillant A, et al. Bilateral bronchial anastomosis in double lung and heart-lung transplantation. Eur J Cardiothorac Surg. 1990;4:314–316.
9. Macchiarini P, Le Roy Ladurie F, Cerrina F, et al. Clamshell or sternotomy for double lung or heart-lung transplantation? Eur J Cardiothorac Surg. 1999;15:333–339.
10. Gandy KL, Moulton MI. Sternal plating to prevent malunion of transverse sternotomy in lung transplantation. Ann Thor Surg. 2008;86:1384–1385.
11. Brown RP, Esmore DS, Lawson C. Improved sternal fixation in the transsternal bilateral thoracotomy incision. J Thorac Cardiovasc Surg. 1996;112:137–141.
12. Taghavi S, Birsan T, Seitelberger R, et al. Initial experience with two sequential anterolateral thoracotomies for bilateral lung transplantation. Ann Thorac Surg. 1999;67:1440–1443.
Back to Top | Article Outline


Traditionally, lung transplantation has been performed through a median sternotomy, clam shell thoracotomy or single or bilateral posterolateral thoracotomies. Having the capability of sufficient access to cannulate for cardiopulmonary bypass and sufficient exposure to explant the native lung and implant the transplanted lung are considered necessary for efficiency and safety. The different traditional lung transplant access approaches have varying degrees of chest wall pain and inherent physiololgical deficits, impacting the recovery and outcomes from the procedure.

As with the proven beneficial findings of other thoracotomy indications, the authors demonstrate that the smaller chest access incision, as small as 6 cm in their series, and perhaps the resultant less rib spreading, and the mammary artery preservation and the lack of transverse sternotomy appears to result in a low chest wound complication rate without significantly affecting the conduct of the procedure, as it relates to operative bypass times. These findings provide further support to the already demonstrated attributes of the small, nonsternotomy incision for lung transplantation.


Lung transplantation; Small incisions; Sternum sparing; Limited thoracotomy

© 2011 Lippincott Williams & Wilkins, Inc.