- Email: [email protected]
Technical Innovations of Carinal Resection for Nonsmall-Cell Lung Cancer
Paolo Macchiarini, MD, PhD, Matthias Altmayer, MD, Tetsuhiko Go, MD, Thorsten Walles, MD, Karl Schulze, MD, Ingeborg Wildfang, MD, Axel Haverich, MD, PhD, Michael Hardin, PhD, and the Hannover Interdisciplinary Intrathoracic Tumor Task Force Group Department of General Thoracic Surgery, University of Barcelona, Barcelona, Spain; Department of Cardiothoracic and Vascular Surgery, Hannover Medical School, Hannover, Germany; Department of Thoracic Surgery, Fukuiken Saiseikai Hospital, Fukui City, Japan; Departments of Anesthesiology and Radiation Oncology, Siloah Hospital, Hannover, Germany; and Information Systems, Statistics, and Management Science, University of Alabama, Tuscaloosa, Alabama
Background. We present our perioperative management of operable nonsmall-cell lung cancer invading the tracheobronchial bifurcation and the results obtained. Methods. Fifty consecutive patients undergoing carinal surgery with radical lymphadenectomy over a 5-year period were studied. Results. Eighteen patients (36%) were N2 and had chemoradiation (48 ⴞ 6 Gy) preoperatively. Surgery included 34 carinal pneumonectomies (24 right, 10 left), 11 carinal lobectomies (n ⴝ 6) or bilobectomies (n ⴝ 5), and 5 carinal resections, with (n ⴝ 3) and without (n ⴝ 2) reconstructions. Patients were ventilated through low tidal volume controlled techniques except during airway resection and reconstruction, during which the apneic (hyper) oxygenation techniques were used. High inspiratory oxygen concentrations, multiple collapse and reexpansions, hypoperfusion of the ipsilateral lung, and fluid overload were avoided. All patients but 1 were extubated in the operating room, 7 ⴞ 5 minutes after skin closure. Operative mortality (less than 30 days) and morbidity
were 4% (n ⴝ 2) and 37% (n ⴝ 18), respectively. All resections but 1 (98%) R1 were complete. The number of resected nodes per patient was 9 ⴞ 2, and 7 (22%) of the 32 patients who had negative preoperative positron emission tomography results had micrometastatic mediastinal nodes. With a median follow-up of 38 months, actuarial 5-year and disease-free survivals were 51% and 47%, respectively. Disease-free survival was significantly affected by endobronchial extension (tracheobronchial angle invasion versus less than 0.5 cm from carina, p ⴝ 0.03) and nodal status (N0 versus N1-2, p ⴝ 0.02) in the multivariate analysis. Conclusions. Preoperative chemoradiation, carinal lobectomy, or left pneumonectomy, and radical lymphadenectomy do not worsen the therapeutic index of carinal surgery. The high incidence of micrometastatic nodes in positron emission tomography–negative patients justifies routine mediastinoscopy and radical lymphadenectomy. (Ann Thorac Surg 2006;82:1989 –97) © 2006 by The Society of Thoracic Surgeons
E
The purpose of this paper was to present our perioperative management of operable NSCLC invading the tracheobronchial bifurcation and the results obtained.
ver since its debut [1], carinal resection has been regarded as a potential curative treatment for selected bronchial nonsmall-cell lung cancer (NSCLC) invading the tracheobronchial bifurcation— but with one vexing Achilles heel: the associated high morbidity and mortality and poor long-term outcome [2]. That hurdle has, however, struggled, as the precise preoperative, intraoperative, and postoperative guidelines initially proposed by Grillo [3] and Grillo and Mathisen [4] have been subsequently confirmed by others [5–7] and long-term outcome provided [8 –10]. Unfortunately, these procedures are still done only in a few centers worldwide, probably because of the unchanged general thinking about their limited patient benefit. Accepted for publication July 10, 2006. Presented at the Forty-second Annual Meeting of The Society of Thoracic Surgeons, Chicago, IL, Jan 30 –Feb 1, 2006. Address correspondence to Dr Macchiarini, Department of General Thoracic Surgery, Hospital Clinic of Barcelona, University of Barcelona, 170 Villarel, Barcelona E-30889, Spain; e-mail: [email protected].
© 2006 by The Society of Thoracic Surgeons Published by Elsevier Inc
Patients and Methods Fifty consecutive patients with a diagnosis of operable NSCLC approaching or invading the tracheobronchial bifurcation underwent resection and reconstruction between April 1999 and December 2004. They represent 18% of the total number of pneumonectomies performed during the same period. Written informed consent was obtained from all patients, and the study was approved by the Institutional Review Board.
Eligibility Criteria Patients were aged less than 80 years; had a performance status of 0 to 1; histologic diagnosis of NSCLC approaching (⬍ 0.5 cm) or invading the tracheobronchial bifurcation to an extent that a maximal of 4 cm of airway 0003-4975/06/$32.00 doi:10.1016/j.athoracsur.2006.07.016
GENERAL THORACIC
Technical Innovations of Carinal Resection for Nonsmall-Cell Lung Cancer
GENERAL THORACIC
1990
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
Ann Thorac Surg 2006;82:1989 –97
dict postoperative lung function, stress spirometry, and transthoracic echocardiography. Patients with suspected pulmonary hypertension related to chronic obstructive pulmonary disease underwent a right heart catheterization with assessment of the hemodynamics at baseline and after regional vasodilators, including nitric oxide eventually with the endoluminal blockage of the ipsilateral pulmonary artery. This was done to avoid operation in patients with latent pulmonary hypertension or who were unresponsive to nitric oxide. Figure 1 shows the oncologic algorithm systematically used in all eligible patients; mediastinal staging included a thorax computed tomography scan, 2-deoxy-2-[18F] fluoro-D-glucose positron emission tomography (FDGPET), and mediastinoscopy. Precise delineation of airway invasion was obtained by rigid bronchoscopy, during which serial biopsies of the mucosa and submucosa lying at least 2 cm from the grossly tumoral surface were taken routinely, as well as on the contralateral tracheobronchial angle and main bronchus. Endoesophageal sonography was performed to evaluate the tumor’s extension to the esophagus, left atrium, and biopsy inferior mediastinal nodes, if present.
Preoperative Preparation
Fig 1. Algorithm of eligibility to carinal resection for nonsmall cell lung cancer. *Only patients without involvement of more than two N2 nodal levels below station 2 and N3 lymph nodes were enrolled. (PET ⫽ positron emission tomography.)
resection was required to completely remove the tumor burden; no more than two N2 nodal levels involvement below station 2 [7, 11]; no N3 lymph nodes; no distant metastasis; and normal cardiopulmonary, liver, and renal function. Preoperative findings of asymptomatic superior vena cava (SVC) invasion or limited infiltration of the muscular wall of the intrathoracic esophagus were not operative contraindications. Chronic steroid intake was gradually but quickly reduced to 5 mg/day. Patients with positive N2 nodes diagnosed preoperatively underwent neoadjuvant radiochemotherapy including two cycles of weekly carboplatinum and paclitaxel (carboplatinum AUC 2/paclitaxel 45 mg/m 2 ) simultaneous with radiotherapy (45 Gy) delivered to the tumor-burdened area and mediastinum. In the absence of progressive disease evaluated through a new staging work-up, including positron emission tomography (PET) scan, all patients underwent surgery counting those with postchemoradiation PET-positive mediastinal nodes or intraoperative mediastinal nodes positive frozen section, or both. A redo mediastinoscopy after chemoradiation therapy was not performed to avoid excessive devascularization and preserve the mobility and flexibility of the upper intrathoracic portion of the trachea.
Preoperative Assessment All patients underwent routine functional investigations including quantitative ventilation/perfusion scans to pre-
Patients were encouraged to stop smoking, and a minimum of 10 days abstinence was required. They were usually hospitalized 3 to 5 days preoperatively to become familiar with the mechanism of breathing, coughing, and swallowing, and to have twice-daily aerosol therapy.
Anesthesia Patients received total intravenous anesthesia with propofol, remifentanil, and cisatracurium as muscle relaxants, and epidural anesthesia using a continuous infusion of ropivacaine 0.2% and sufentanil 0.5 g/mL,
Fig 2. Operative view showing a median sternotomy with an Octopus (O) stabilizer retracting the heart to the right side, and the catheter (C) of apnoic oxygenation placed into the distal trachea, which allows oxygenation and undisturbed and uninterrupted placement of airway sutures. Retracting the heart to the right not only increases the surgical exposure of the affected left hilum, especially the dorsal aspect, but also improves the hemodynamic stability during the manipulation and securing of the pulmonary veins.
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
1991
after a loading dose. Electroencephalographic monitoring (Narcotrend; MT Monitor Technik, Munich, Germany) allowed adaptation of propofol infusion rates according to the depth of narcosis. Besides the usual monitoring of electrocardiogram with 5 leads, patients had continuous monitoring of arterial blood pressure and gases. Patient were ventilated initially and after airway reconstruction through a left-sided double-lumen tube (rightsided procedures) or an extralong armored endotracheal tube positioned in the right main stem bronchus or intrathoracic trachea (left-sided procedures or carinal resections alone) with low tidal volume (volume or pressure) controlled techniques. High inspiratory oxygen concentrations, multiple collapse and reexpansions, hypoxic pulmonary vasoconstriction and hypoperfusion of the ipsilateral lung, and fluid overload were avoided in all patients. Just before airway section, and during resection and reconstruction, patients were managed with the apneic (hyper)oxygenation technique [12], as follows. Patients were preoxygenated and hyperventilated with 100% oxygen (O2) for about 10 minutes before completing the dissection to reach arterial pO2 and pCO2 levels of 450 or greater and 28 to 35 mm Hg, respectively. The airway was then sectioned and resected under total apnea. After completion, hyperoxygenation was obtained by placing a small (10F) catheter accross the surgical field into the contralateral main bronchus, fixing it to the catilagenous ring (about at the 12 o’clock position) to prevent its dislocation, and connecting it to a sterile line delivering 10 to 15 L/min O2 continuously under minimal breathing pressure (0 to 1 mm Hg). Permissive hypercapnia was eventually treated. After reconstruction, patients were volume-controlled ventilated through the original tube.
Fig 3. Illustration of the secondary end-to-side anastomosis of the right residual stump on the lateral aspect of the left main bronchus. (A) A double-ended polydioxanone (PDS; Ethicon, Sommerville, New Jersey) 3-0 or 4-0 suture is started on one or the other edges of the deepest or posterior aspect of the anastomosis, picking up the full layers of the respective bronchial walls, and inserting the needles from inside the lumen. It must be left untied to allow a number of stitches to be placed to complete the entire deepest aspect of the anastomosis before the reimplanting bronchus is pulled down onto the hosting structure. (B) The double-ended suture is then pulled tight, as the reimplanting and recipient lumens are parachuted. We recommend using the exact mucosa and stitch apposition through nerve hooks to avoid pursestring and narrows effects. (C) Several concentric interrupted 3-0 to 5-0 polyglactin (Vicryl; Ethicon) sutures are placed on the remaining quadrants 3 to 4 mm apart and 3 to 4 mm from the cut edge of the airway, leaving the membranous wall of the reimplanting bronchus until last to allow balancing of any anastomotic disparity and excessive traction on it. We found that placing small Vicryl sutures (eg, 5-0) on the membranous wall reduces the manipulation risks of injury. Once placement is completed, the wall and sutures are gently approximated, placing knots outside walls.
GENERAL THORACIC
Ann Thorac Surg 2006;82:1989 –97
GENERAL THORACIC
1992
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
Ann Thorac Surg 2006;82:1989 –97
tors during clamping by fluid implementation and regulation of mean arterial pressure with norepinephrine. To keep cerebral perfusion pressure greater than 60 mm Hg, mean arterial pressure was increased at least to 60 mm Hg above the jugular bulb pressure. Cardiac output was kept within the normal range by cristalloid or colloid infusions or epinephrine, the goal being to conserve jugular bulb oxygen saturation greater than 50%. Since the SVC step was usually made before the airway opening, patients were moderately hyperventilated to decrease cerebral blood volume, and mean arterial pressure and extent of hyperventilation were adjusted to the jugular bulb blood gas values. After declamping, mannitol 20% was administered to mitigate brain edema.
Operative Techniques The basic surgical principles of carinal resection and the end-to-end primary anastomosis outlined by Grillo and associates [3, 4, 8, 15, 16] and by us [11, 17] were followed. A few points deserve, however, reemphasis. Carinal resection without sacrifice of pulmonary parenchyma and left-sided carinal pneumonectomies were usually approached through a median sternotomy. Right-sided lesions necessitating a carinal resection with sacrifice of the pulmonary parenchyma were best approached through an ipsilateral, muscle-sparing dorsal thoracotomy in the fifth intercostal space.
INCISIONS.
Dissection was limited to the anterior surface of the lower trachea and the first 2 cm of the contralateral main bronchus while preserving as much as possible any local blood supply. Airway mobilization was only to the extent needed and limited to no more than 2 cm from the proposed proximal and distal lines of transection. All patients had a complete nodal dissection, and the lymphadenectomy was made by individually securing the nourishing node vessels. For left carinal pneumonectomies, further exposure of the left hilum was obtained through the cardiac Octopus suction stabilizer (Medtronic, Minneapolis, Minnesota) applied to the apex of the heart, permitting thereby to lift the heart laterally (right) or vertically (cranially) or put it wherever needed, always under inotropic support (Fig 2); moreover, the left lower vein was, whenever possible, sectioned with the commercially available movable endostaplers.
DISSECTION.
Fig 4. Drawing of the pericardiophrenic release (lateral view). In patients requiring a left carinal resection in which tension at the anastomosis might be anticipated, one can made at the end of the operation a surgical separation from the pericardium from the diaphragm by sharp dissection or electrocautery from ventrally to dorsally and between the two phrenic nerves. A further 2 to 3 cm can be gained in the immediate postoperative period. The hole needs to be drained for few days postoperatively.
For patients with a preoperative suspicion of a SVC invasion, we also continuously monitored the jugular bulb pressure through a right internal jugular vein cathether placed retrogradely up to the jugular bulb; the cerebral oxygenation noninvasively by near-infrared spectroscopy; and the continuous cardiac output (PiCCO; Pulsion Corp, Munich, Germany) through a femoral artery cathether. A venous access into the inferior vena cava through an indwelling catheter was placed as were other lines on the dorsal foot and in both femoral veins. These monitoring devices served to control circulation during caval clamping. Based on our experience [13], and that acquired for traumatic head injury by others [14], the main targets during SVC clamping were to compensate for the cranial venous hypertension, systemic hypotension due to the impaired venous return, and reduced cerebral blood flow. Neuroprotective measures included the administration of corticosteroids about 30 minutes before clamping, and optimization of circulatory indica-
For uniformity and teaching purposes, we always used the same anastomotic principle to perform the primary end-to-end and secondary endto-side [18] anastomoses. For carinal pneumonectomy, the end-to-end reconstruction started on the deepest anastomotic part of the left half of both cartilaginous walls or left margin of the membraneous portion. For carinal lobectomies, the choice to reimplant the residual bilobe end to side either into the trachea or left main bronchus depended on the length of the residual intermedius bronchus, the status of the trachea (fibrotic, malacic, and so forth), the extent of node dissection, and the degree of traction on hilus residual vessels. Over
ANASTOMOTIC TECHNIQUE.
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
Table 1. Patients’ Demographics
Table 3. Type of Anastomosis
Demographic
Value
Sex (male vs. female) Age, years (range)
41 vs. 9 56 ⫾ 13 (22–77) 35/15
Histology (squamous/nonsquamousa) Type of preoperative tumor invasion ⬍ 0.5 cm from origin carina Ipsilateral tracheobronchial angle Carina only Postpneumonectomy stump (relapse) Side (right/left) ppFEV1 (%) Secondary pulmonary hypertension Prior surgeryb ASA score (1/2/3) Neoadjuvant chemoradiation
23 (46%) 22 (44%) 3 (6%) 2 (4%) 37/13 54 ⫾ 12 7 (14%) 7 (14%) 20/13/17 18 (36%)
a The 15 nonsquamous tumors included 9 adenocarcinomas and 6 large b cells carcinomas. Includes 2 exploratory thoracotomics, 3 lower bilobectomies, and 2 pneumonectomies stump relapses performed 22 ⫾ 18 months earlier.
ASA ⫽ American Society of Anesthesiologists Classification of physical status; ppFEV1 ⫽ predicted postoperative forced expiratory volume in 1 second.
time, however, we have preferred to perform the end-toside secondary anastomosis on the left main bronchus (Fig 3). During the end-to-side reimplantation, the residual ipsilateral pulmonary artery was always left unclamped to avoid hypoxic pulmonary vasoconstriction and hypoperfusion of the residual lobe or hyperperfusion of the contralateral lung. The anastomosis was left uncovered (right reconstructions) or partly covered with viable neighboring tissue (left reconstructions). Release maneuvers were limited to the development of the pretracheal plane, either during mediastinoscopy or through the operative field, performance of an inferior U-shaped hilar release, and for left-sided primary or secondary end-to-side anastomoses in which tension was still too high, a pericardiophrenic release with separation of the pericardium from the diaphragm between both phrenic nerves was made at the end of the operation (Fig 4). RELEASE MANEUVERS.
Postoperative Care Hyperinflation of the residual lung was avoided while satisfactory oxygenation and ventilation was obtained by Table 2. Relation Between Type of Operation and Side of the Tumor Type of Operation Carinal Carinal Carinal Carinal Carinal Total
pneumonectomy lobectomy bilobectomy resection and reconstruction resection
1993
Right Side
Left Side
24 6 5 — 2 37
10 — — 3 — 13
Number Primary end-to-end Trachea to left main bronchus Trachea to right main bronchus Secondary end-to-side Residual right upper lobe bronchus to trachea Residual bronchus intermedius to trachea Residual bronchus intermedius to left main bronchus Residual left main bronchus to right main bronchus
37 13 3 2 6 3
adequate pain relief with epidural analgesia or patientcontrolled analgesia, chest physiotherapy, and repeat aspirative fibroscopies. In patients having a pneumonectomy, the nasogastric tube was left in place for 24 to 36 hours. A temporary minimal invasive tracheotomy was performed at the end of operation to reduce the physiologic respiratory dead space and facilitate direct aspiration whenever the predicted residual ventilatory function reserve was borderline and the patient’s collaboration reduced. Patients with SVC revascularization received intravenous heparin 4 to 6 hours postoperatively until reaching a therapeutic prothrombin level and full mobilization was obtained. Then, aspirin was substituted and given for 6 months postoperatively [18].
Adjuvant Therapy All patients entering the neoadjuvant protocol received two additional courses of carboplatinum (AUC 6) and paclitaxel (175 ng/m2). Patients with intraoperative first detected positive N2 nodes received four courses of carboplatinum (AUC 2) and paclitaxel (45 mg/m2) chemotherapy and 50 Gy radiotherapy delivered to the mediastinum only.
Statistical Methods Data are presented as mean ⫾ SD of number (n) of observations. Survival was measured from the date of operation to death and included postoperative deaths. Disease-free survival (DFS) was measured from the date of surgery until the first evidence of tumor recurrence, where tumor recurrence included local-only failure, disTable 4. Relation Between Type of Operation and Mortality and Morbidity Operative Mortality
Operative Morbidity
pneumonectomy (n ⫽ 34) lobectomy (n ⫽ 6) bilobectomy (n ⫽ 5) resection-reconstruction
1 (2%) — 1 (2%) —
8 (16%) 3 (50%) 3 (60%) 3 (100%)
resection (n ⫽ 2)
— 2 (4%)
Type of Operation Carinal Carinal Carinal Carinal (n ⫽ 3) Carinal Total
1 (50%) 18 (36%)
GENERAL THORACIC
Ann Thorac Surg 2006;82:1989 –97
GENERAL THORACIC
1994
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
Ann Thorac Surg 2006;82:1989 –97
All but the very first patient were extubated in the operating room after 7 ⫾ 5 minutes from the end of skin closure. Operative mortality (less than 30 days) was 4% and included a massive myocardial infarction and a lung embolism. Complications were observed in 18 patients (36%) and were supraventricular arrhythmias (n ⫽ 7), cerebral transient ischemic attack (n ⫽ 1), cardiac luxation requiring surgical intervention (n ⫽ 1), anticoagulation-related bleeding (n ⫽ 1), and anastomotic leaks in 8 patients successfully managed either conservatively (temporary drainage of the ipsilateral cavity, n ⫽ 4) or through immediate reoperation and anastomosing (n ⫽ 3) or tissue engineered airway interposition (n ⫽ 1; Table 4) [22]. Beyond the 2 operative deaths, the 2 patients having a transient ischemic attack and cardiac luxation required mechanical ventilation postoperatively for 3 ⫾ 2 days. All patients but 1 (98%) with residual microscopic disease (R1) on the ipsilateral residual pulmonary artery stump were completely resected. The frequency of nodal involvement compared with the type of operation is shown in Table 5. The overall number of resected nodes per patient was 9 ⫾ 2, and the number of positive nodes was higher in N2 patients (4.7 ⫾ 2.9) than in N1 patients (2.7 ⫾ 3.1; p ⫽ 0.054). Among the 18 neoadjuvant chemoradiation N2 patients, 11 (61%) had no pathologic response while 7 (39%) had a histologic downstaging to N1 status (n ⫽ 2) or N0 status (n ⫽ 5). In contrast, 7 of the residual 32 patients (22%) had pathology-positive N2 nodes, despite negative preoperative PET scan. With a median follow-up time of 38 months, the projected 5-year survival and DFS were 51% and 47%, respectively. Survival was significantly affected by the need for postoperative ventilation (p ⫽ 0.013) and development of anastomotic complications (p ⫽ 0.0236) in the univariate but not in the multivariate analysis. In contrast, DFS was significantly affected by the endobronchial extension of the primary tumor (p ⫽ 0.033) and nodal status (p ⫽ 0.024) in the univariate and multivariate analysis (Table 6). Among 22 patients with tumor recurrence, one recurrence was locoregional and the others were systemic, as solitary (n ⫽ 14) or multiple (n ⫽ 7) first tumors.
Table 5. Frequency of Nodal Status for Type of Operations Nodal Status Type of Operations Carinal Carinal Carinal Carinal Carinal Total
pneumonectomy (n ⫽ 34) lobectomy (n ⫽ 6) bilobectomy (n ⫽ 5) resection and reconstruction (n ⫽ 3) resection (n ⫽ 2)
N0
N1
N2
12 3 3 2 2 22
7 2 1 0 0 10
15 1 1 1 0 18
tant-only failure, or simultaneous local and distal failure. Patients who died from causes unrelated to NSCLC were censored at the time of death in the estimation of DFS. Survival and DFS were computed using the KaplanMeier method [19]. Estimates based on categorical clinical variables were compared using the log-rank test [20], and estimates based on continuous variables were compared by using Cox regression analysis [21]. Variables for which the p value was less than 0.05 in the univariate analysis were included in a Cox proportional hazards multivariate regression model. All computations were made using Statview software (SAS Institute, Cary, North Carolina).
Results Patients’ preoperative characteristics are depicted in Table 1. Among them, 18 N2 patients had neoadjuvant chemoradiation with a delivered radiation dose of 48 ⫾ 6 Gy. Table 2 correlates type of operation to tumor side. The length of resected airway was 2.8 ⫾ 0.6 cm, and airway reconstruction necessitated a total of 64 anastomoses (Table 3). For the 38 patients managed with the apneic oxygenation technique, the duration of the apneic and apneic hyperoxygenation phases were 5 ⫾ 2 and 26 ⫾ 11 minutes, respectively; none had specific-related morbidity. Twelve patients (24%) underwent concurrent truncular resection of the SVC (n ⫽ 4), partial excision of the muscular esophageal wall (n ⫽ 2), diaphragm (n ⫽ 3), chest wall (n ⫽ 2) and left atrium (n ⫽ 1). The SVC clamping time was 16 ⫾ 3 minutes.
Table 6. Multivariate Analysis for Disease-Free Survival Factor Endobronchial extension (p ⫽ 0.02) ⬍ 0.5 cm from carina Tracheobronchial angle Carina only Nodal status (p ⫽ 0.04) N0 N1 N2
Number of Patients
Number of Failures
Hazards Ratio
95% Confidence Interval
p Value
23 22 3
6 12 1
0.047 0.159 0.153
0.004 to 0.544 0.015 to 1.73 0.008 to 3.070
0.01 0.014
21 11 18
4 7 9
0.192 5.557 5.215
0.004 to 0.76 1.35 to 22.71 1.30 to 20.90
0.017 0.019
Cox proportional hazards regression was used to determine the influence of clinically significant factors on disease-free survival.
Comment The main innovations introduced in the present study were the restriction of patients’ eligibility, performance of routine preoperative PET scan, neoadjuvant chemoradiation therapy for N2 patients, more physiologic intraoperative management, and whenever feasible, lungsparing resections and innovative technical aspects. That produced lower mortality rates but equal morbidity rates compared with recent reports [9, 10, 23] and an encouraging early outcome. The presence of metastatic mediastinal nodes in patients requiring carinal resection has always been considered a potential contraindication in the past [7, 8], and more recently as well [9, 10]. Despite its potentiality to improve outcome [24], however, neoadjuvant chemotherapy has never been systematically employed in newly diagnosed N2 patients requiring carinal reconstruction, mainly because of its association with increased operative mortality [9]. This study shows that neoadjuvant chemoradiation is feasible, pathologically downstages the N2 status in less than 40% of patients, improves DFS compared with more recent studies [9 –10], and finally, permits the performance of more complex carinal reconstruction without significantly influencing either their operative morbidity or mortality. These results may be mainly explained by our practice of contraindicating surgery for patients with invasion of more than two mediastinal nodal stations, as well as positive station 2R/L nodes or N3 nodes, rather than the ability of PET scan to evaluate the mediastinum. Moreover, that almost one quarter of PET-negative patients had micometastatic positive nodes provides definitive evidence that mediastinoscopy cannot be excluded in the evaluation of the patient amendable to carinal surgery, as per any other NSCLC surgery [25], and should routinely be performed preoperatively [9, 10]. A similar consensus on the role of lymphadenectomy does not exist, it being the recommendation to perform lymph node dissection only of the immediate adjacent or clearly involved nodes [16]. Given the relatively high incidence of micrometastatic nodes undetectable at PET scan and given that the performance of radical lymphadenectomy did not increase airway or residual parenchymal morbidity, we might speculate that radical lymphadenectomy should be recommended to increase the therapeutic index of this surgery, especially after recent and growing evidence of the benefits of adjuvant chemotherapy for patients with completely resected NSCLC [26, 27]. Whether patients whose N2 status remains positive either at restaging (PET) or postoperatively (frozen section) should still be resected remains to be determined, but at least in this series, their DFS was encouragingly superior when compared with our previous N2 carinal patients without induction therapy [11]. The other main innovation, in our opinion, was to avoid intraoperative barotrauma, the predominant factor associated with adult respiratory distress syndrome (ARDS) [28] and the leading cause of mortality after carinal resection [8, 23]. To achieve this, ventilation was
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
1995
made through low tidal volume controlled techniques; and high inspiratory oxygen concentrations, multiple collapse and reexpansions, hypoxic pulmonary vasoconstriction during hypoperfusion of the ipsilateral lung parenchyma or hyperperfusion of the contralateral lung, and fluid overload were maximally avoided. All these factors reduce lung compliance [29, 30] and have been related to the occurrence of ARDS after pneumonectomy [31]. That we did not notice any ARDS postoperatively, despite radical lymphadenectomy, strongly suggests that these maneuvers may have played a fundamental role in avoiding this devastating complication. Moreover, apneic (hyper)oxygenation techniques were used, and their continuous capability to oxygenate and not ventilate the patients provides a more physiologic means of sustaining gas exchange and less volume, barotrauma, and biotrauma to the residual lung parenchyma [32, 33]. Briefly, before sectioning the airway, the patient is preoxygenated and hyperventilated with 100% oxygen only so that the 2,500 to 3,000 mL volume of the functional residual capacity is almost completely denitrogenated and any nitrogen entrainment is ceased. As the normal oxygen consumption is 200 to 250 mL per minute, the 2,500 to 3,000 mL intrapulmonary oxygen store provides adequate oxygenation for 10 to 12 minutes of total apnea. Furthermore, after the opening of the airway, a pediatric catheter is placed under visualization above the carina; and by simply delivering a small flow (10 to 15 L/min oxygen) under minimal breathing pressure (0 to 2 psi) to both lungs, the patient can “breathe” despite ongoing total apnea and can survive even apneic periods of 50 minutes, without any ill effects. In contrast to McClish and colleagues [34], we do not ventilate the lungs and use a low pressure (0 to 2 psi) and small flow (10 to 15 L/min oxygen) delivering system. This represents, in our opinion, a more physiologic gas exchange method while keeping the significant advantages of the catheter technique, for example, improved surgical exposure, minimal intraoperative intrusion of the anesthetic apparatus, and facilitated reconstruction. In the present series, a higher number of left carinal pneumonectomies and carinal lobectomies were made— both, surgically speaking, challenging issues. Although the main dispute concerning left carinal surgery has been the surgical approach [15, 35], we reinforce the advantages of median sternotomy, for example, excellent exposure of the tracheobronchial bifurcation and less incisional discomfort [11, 17] while proposing the use of heart stabilizers such as, in our experience, the Octopus device to lift the heart whenever the need is to improve exposure of the left hilum and release pleuroparietal adhesion. The only disadvantage of this incision is when a lymphadenectomy of a nodal station below 7 is needed, but with increasing experience, an additional left thoracotomy may be avoided by using the heart retractor and opening the posterior pericardium vertically on the midline. Hence, for more complex left reconstructions, median sternotomy allows the surgeon to further reduce the anastomotic tension by separating the pericardiodiaphragmatic reflections between the two phrenic nerves, a
GENERAL THORACIC
Ann Thorac Surg 2006;82:1989 –97
GENERAL THORACIC
1996
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
release maneuver that permits an upward gain of almost 2 cm. Care should be taken, however, to avoid injury of the right phrenic nerve because in almost every left carinal pneumonectomy, the left one will be sacrified. Because of the early and late issues related to carinal plus lobar resection, these operations are rarely performed, and preference has be given to perform carinal pneumonectomy if the patient can otherwise tolerate such a procedure [4, 16]. Based on our previous [18] and this experience, our data may counter this statement, from an oncologic and technical point of view. However, if reimplantation is to be performed (either bronchus intermedius or right lower lobe bronchus), the secondary end-to-side anastomosis should be to the side of the left main bronchus and not to the trachea, not only to avoid excessive tension but also because if a minimal anastomotic leak occurs, it will spontaneously heal through the collapse of the mediastinal and lung tissue over its surface if the pleural space is sufficiently well drained. In conclusion, this study provides evidence that some traditional pitfalls related to carinal resection in operable NSCLC can be avoided. While the oncologic outcome may be improved by contraindicating surgery for “massive” N2, when performing routine mediastinoscopy and radical lymphadenectomy, the operative mortality and some fatal morbidity, for example, ARDS, may be minimized by using intraoperative lung-protective ventilation techniques, and avoiding high inspiratory oxygen concentrations, multiple collapse and reexpansions, hypoxic pulmonary vasoconstriction during hypoperfusion of the ipsilateral lung parenchyma or hyperperfusion of the contralateral lung, and fluid overload. Technically speaking, left carinal pneumonectomy or more complex rightsided and lung-sparing reconstruction may be safely performed using the presented modus operandi. The authors would like to thank all members of the Hannover Interdisciplinary Intrathoracic Tumor Task Force Group who not only contributed to the accomplishment of the study protocols but also actively participated in the enrollment of the patients: Drs Hartmut Kirchner, Joachim Pieper, Markus Sosada, Jürgen Borghardt, Bernd Gaede, Felix Winkler (medical oncologists); and Drs Bernd Wolfgang Raack, Alexander Schulz, Hans-Peter Dirks, Matthias Kleckow, Werner Mall, Bern Vogel, Jost Niedermeyer, and Helmut Fabel (pneumologists). We are also thankful to all referring national and international doctors, and to Jeff Szychowski, PhD, University of Alabama Tuscaloosa, for the invaluable statistical assistance.
References 1. Gibbon JH. Discussion of Chamberlein JM, McNeil TM, Parnassa P, Edsall JR. Bronchogenic carcinoma: an aggressive surgical attitude. J Thorac Cardiovasc Surg 1959;38: 727– 45. 2. Jensik RJ, Faber LP, Kittle CF, Miley RW, Thatcher WC, El-Baz N. Survival in patients undergoing tracheal sleeve pneumonectomy for bronchogenic carcinoma. J Thorac Cardiovasc Surg 1982;84:489 –96. 3. Grillo HC. Carinal reconstruction. Ann Thorac Surg 1982;34: 356 –73. 4. Mathisen DJ, Grillo HC. Carinal resection for bronchogenic carcinoma. J Thorac Cardiovasc Surg 1991;102:16 –23.
Ann Thorac Surg 2006;82:1989 –97
5. Tsuchiya R, Goya T, Naruke T, Suemasu K. Resection of tracheal carina for lung cancer. J Thorac Cardiovasc Surg 1990;99:779 – 87. 6. Roviaro GC, Varoli F, Rebuffat C, et al. Tracheal sleeve pneumonectomy for bronchogenic carcinoma. J Thorac Cardiovasc Surg 1994;107:13– 8. 7. Dartevelle P, Macchiarini P, Chapelier A. 1986: tracheal sleeve pneumonectomy for bronchogenic carcinoma. Ann Thorac Surg 1995;60:1854 –5. 8. Mitchell JD, Mathisen DJ, Wright CD, et al. Resection for bronchogenic carcinoma involving the carina. Long-term results and effect of nodal status on outcome. J Thorac Cardiovasc Surg 2001;121:465–71. 9. de Perrot M, Fadel E, Mercier O, Mussot S, Chapelier C, Dartevelle P. Long-term results after carinal resection for carcinoma: does the benefit warrant the risk? J Thorac Cardiovasc Surg 2006;131:81–9. 10. Regnard JF, Perrotin P, Giovannetti R, et al. Resection for tumors with carinal involvement: technical aspects, results, and prognostic factors. Ann Thorac Surg 2005;80:1841– 6. 11. Dartevelle P, Macchiarini P. Carinal pneumonectomy for bronchogenic carcinoma. Semin Thorac Cardiovasc Surg 1996;8:414 –25. 12. Go T, Altmayer M, Richter M, Macchiarini P. Decompressing manubriectomy under apneic oxygenation to release the median thoracic outlet compartment in Bechterew disease. J Thorac Cardiovasc Surg 2003;126:867–9. 13. Macchiarini P, Dartevelle P. Resection and reconstruction of the superior vena cava. In: Pearson FG, Cooper JD, Deslauriers J, et al, eds. Thoracic surgery. 2nd ed. New York: Churchill Livingston, 2002:1774 – 80. 14. Rosner MJ, Rosner SD, Johnson AH. Cerebral perfusion pressure: management protocol and clinical results. J Neurosurg 1995;83:949 – 62. 15. Mitchell JD, Mathisen DJ, Wright CD, et al. Clinical experience with carinal resection. J Thorac Cardiovasc Surg 1999; 117:39 –53. 16. Mitchell JD. Carinal resection and reconstruction. Chest Surg Clin North Am 2003;13:315–29. 17. Dartevelle P, Macchiarini P. Techniques of pneumonectomy. Sleeve pneumonectomy. Chest Surg Clin North Am 1999;9: 407–17. 18. Di Rienzo G, Go T, Macchiarini P. Simplified technique for the secondary end-to-side anastomosis in complex tracheobronchial reconstruction. J Thorac Cardiovasc Surg 2002;124: 632–5. 19. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Ass 1958;53:457– 81. 20. Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data. 2nd ed. New York: John Wiley & Sons, 2002. 21. Cox DR. Regression models and life tables. J Roy Stat Soc Series B 1972;34:187–220. 22. Macchiarini P, Walles T, Biancosino C, Mertschning H. First human transplantation of a bioengineered airway tissue. J Thorac Cardiovasc Surg 2004;128:638 – 41. 23. Porhanov VA, Poliakov IS, Selvaschuk AP, et al. Indications and results of sleeve carinal resection. Eur J Cardiothorac Surg 2002;22:685–94. 24. Betticher DC, Hsu Schmitz SF, Totsch M, et al. Mediastinal lymph node clearance after docetaxel-cisplatin neoadjuvant chemotherapy is prognostic of survival in patients with stage IIIA pN2 non-small-cell lung cancer. A multicenter phase II trial. J Clin Oncol 2003;21:1752–9. 25. Kernstine KH. Positron emission tomography with 2-[18F]fluoro-2-deoxy-D-glucose: can it be used to accurately stage the mediastinum in non–small cell lung cancer as an alternative to mediastinoscopy? J Thorac Cardiovasc Surg 2003; 126:1700 –3. 26. Arriagada R, Bergman B, Dunant A, et al. Cisplatin-based adjuvant chemotherapy in patients with completely resected non-small-cell lung cancer. N Engl J Med 2004;350: 351–360.
27. Winton T, Livingston R, Johnson D, et al. Vinorelbine plus cisplatin vs. observation in resected non-small-cell lung cancer. N Engl J Med 2005;352:2589 –97. 28. Mathisen D. J Thorac Cardiovasc Surg 2006;131:87. 29. Fessler HE, Brower RG. Protocols for lung protective ventilation. Crit Care Med 2005;33(Suppl):S223–7. 30. Demling RH. The modern version of adult respiratory distress syndrome. Ann Rev Med 1995;46:193–202. 31. Jordan S, Mitchell JA, Quinlan GJ, Goldstraw P, Evans TW. The pathogenesis of lung injury following lung resection. Eur Respir J 2000;15:790 –9.
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
1997
32. Zander R, Mertzlufft F. Clinical use of oxygen stores: preoxygenation and apneic oxygenation. Adv Exp Med Biol 1992;317:413– 420. 33. Frumin JM, Epstein RM, Cohen C. Apneic oxygenation in man. Anesthesiology 1959;20:789 –98. 34. McClish A, Deslauriers J, Beaulieu M, et al. High-flow ventilation during major tracheobronchial reconstruction. J Thorac Cardiovasc Surg 1985;89:508 –12. 35. Maeda M, Nakamoto K, Tsubota N, Okada T, Katsura H. Operative approaches for left-sided carinoplasty. Ann Thorac Surg 1993;56:441–5.
DISCUSSION DR ROBERT J. CERFOLIO (Birmingham, AL): This is an outstanding series on a very difficult technical operation, and the fact that you have done 50 of these over this short time frame really shows that you guys are experts in this and you deserve a lot of credit. I have just a couple of quick questions for you. I notice that you said 39% of your N2 patients were downstaged, but that leads me to believe that 61% weren’t. Do you normally routinely send your lymph nodes for frozen section prior to removing the lung? I know you don’t do repeat mediastinoscopy; thus, do you send them off at the time of the thoracotomy, and if there is residual N2 disease on frozen, do you abort the operation?
make that anastomosis easier? Again, excellent presentation and outstanding work.
DR MACCHIARINI: During the initial mediastinoscopy, we sample all available lymph node stations. During surgery we do a radical lymphadenectomy/frozen section. If mediastinoscopy is positive by the first time, we usually continue as long as we do not have progressive disease.
DR FRANCIS C. NICHOLS III (Rochester, MN): I have two questions. One, the right pneumonectomy group, how many of those patients had preoperative chemotherapy and radiation therapy? The second question, and I do have the advantage of having the abstract, your overall number of resected nodes was 9 nodes per patient, so if you’re doing a radical mediastinal lymphadenectomy, how do you account for just 9 nodes?
DR CERFOLIO: No, I mean after they have had radiation and chemotherapy and you come back to resect them, do you send the nodes off for frozen to see if they have been downstaged or do you just proceed with resection? DR MACCHIARINI: We send it to frozen and proceed to resection. DR CERFOLIO: So if the nodes are still positive and there is residual N2 disease in a patient who was initially N2, you would still do a carinal pneumonectomy? DR MACCHIARINI: Exactly. DR CERFOLIO: That’s pretty aggressive. My other question is just a technical tip that perhaps with your large experience you could help me, who is less experienced. When I try to bring that right main stem up and reimplant it up into the trachea, just like you show, I have always been taught that it is nice to have a couple of centimeters between your end-to-end distal trachea to left main stem anastomosis and your new right main stem side to trachea anastomosis— but sometimes it’s hard to get that lung all the way up there, besides doing pericardial releases and all these other maneuvers. Do you have any other tricks or techniques you can teach me to help
DR MACCHIARINI: Thank you, Robert. I think that the minimum that you should have is 1 cm. If you have 1 cm, you can bring it over in the trachea. However, I would not suggest to anastomose that in the trachea because you already devascularized the upper portion, you already have a mediastinoscopy, and you already have a radical lymphadenectomy. On the other hand, simply putting on the left main bronchus, the tissues that are there are very flexible, so if you have a small leakage, you can manage that conservatively.
DR MACCHIARINI: The first question was? DR NICHOLS: Right pneumonectomy, and how many of those patients had basically radiation therapy is what I’m interested in. DR MACCHIARINI: I need to pass. I don’t know the figures. Concerning the second question, one of the eligibility criteria was to avoid operating on patients with more than two N2 levels, and this is because the previous experience by Philippe Dartevelle showed us that if you have this situation, you have 0% 5-year survival. So this might be related to the fact that we have a mean 9 plus or minus—I don’t know how many. Probably this is a statistical issue. DR MARK J. KRASNA (Baltimore, MD): I enjoyed the presentation. Rendina in Rome and our group in Maryland have described using a routine intercostal muscle flap. I’m wondering if your group now advocates using that routinely, not only when you’re doing both a bronchial and a vascular anastomosis, but even just around your carinal anastomosis. Have you been using that routinely or not? DR MACCHIARINI: No, sir.
GENERAL THORACIC
Ann Thorac Surg 2006;82:1989 –97
Background. We present our perioperative management of operable nonsmall-cell lung cancer invading the tracheobronchial bifurcation and the results obtained. Methods. Fifty consecutive patients undergoing carinal surgery with radical lymphadenectomy over a 5-year period were studied. Results. Eighteen patients (36%) were N2 and had chemoradiation (48 ⴞ 6 Gy) preoperatively. Surgery included 34 carinal pneumonectomies (24 right, 10 left), 11 carinal lobectomies (n ⴝ 6) or bilobectomies (n ⴝ 5), and 5 carinal resections, with (n ⴝ 3) and without (n ⴝ 2) reconstructions. Patients were ventilated through low tidal volume controlled techniques except during airway resection and reconstruction, during which the apneic (hyper) oxygenation techniques were used. High inspiratory oxygen concentrations, multiple collapse and reexpansions, hypoperfusion of the ipsilateral lung, and fluid overload were avoided. All patients but 1 were extubated in the operating room, 7 ⴞ 5 minutes after skin closure. Operative mortality (less than 30 days) and morbidity
were 4% (n ⴝ 2) and 37% (n ⴝ 18), respectively. All resections but 1 (98%) R1 were complete. The number of resected nodes per patient was 9 ⴞ 2, and 7 (22%) of the 32 patients who had negative preoperative positron emission tomography results had micrometastatic mediastinal nodes. With a median follow-up of 38 months, actuarial 5-year and disease-free survivals were 51% and 47%, respectively. Disease-free survival was significantly affected by endobronchial extension (tracheobronchial angle invasion versus less than 0.5 cm from carina, p ⴝ 0.03) and nodal status (N0 versus N1-2, p ⴝ 0.02) in the multivariate analysis. Conclusions. Preoperative chemoradiation, carinal lobectomy, or left pneumonectomy, and radical lymphadenectomy do not worsen the therapeutic index of carinal surgery. The high incidence of micrometastatic nodes in positron emission tomography–negative patients justifies routine mediastinoscopy and radical lymphadenectomy. (Ann Thorac Surg 2006;82:1989 –97) © 2006 by The Society of Thoracic Surgeons
E
The purpose of this paper was to present our perioperative management of operable NSCLC invading the tracheobronchial bifurcation and the results obtained.
ver since its debut [1], carinal resection has been regarded as a potential curative treatment for selected bronchial nonsmall-cell lung cancer (NSCLC) invading the tracheobronchial bifurcation— but with one vexing Achilles heel: the associated high morbidity and mortality and poor long-term outcome [2]. That hurdle has, however, struggled, as the precise preoperative, intraoperative, and postoperative guidelines initially proposed by Grillo [3] and Grillo and Mathisen [4] have been subsequently confirmed by others [5–7] and long-term outcome provided [8 –10]. Unfortunately, these procedures are still done only in a few centers worldwide, probably because of the unchanged general thinking about their limited patient benefit. Accepted for publication July 10, 2006. Presented at the Forty-second Annual Meeting of The Society of Thoracic Surgeons, Chicago, IL, Jan 30 –Feb 1, 2006. Address correspondence to Dr Macchiarini, Department of General Thoracic Surgery, Hospital Clinic of Barcelona, University of Barcelona, 170 Villarel, Barcelona E-30889, Spain; e-mail: [email protected].
© 2006 by The Society of Thoracic Surgeons Published by Elsevier Inc
Patients and Methods Fifty consecutive patients with a diagnosis of operable NSCLC approaching or invading the tracheobronchial bifurcation underwent resection and reconstruction between April 1999 and December 2004. They represent 18% of the total number of pneumonectomies performed during the same period. Written informed consent was obtained from all patients, and the study was approved by the Institutional Review Board.
Eligibility Criteria Patients were aged less than 80 years; had a performance status of 0 to 1; histologic diagnosis of NSCLC approaching (⬍ 0.5 cm) or invading the tracheobronchial bifurcation to an extent that a maximal of 4 cm of airway 0003-4975/06/$32.00 doi:10.1016/j.athoracsur.2006.07.016
GENERAL THORACIC
Technical Innovations of Carinal Resection for Nonsmall-Cell Lung Cancer
GENERAL THORACIC
1990
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
Ann Thorac Surg 2006;82:1989 –97
dict postoperative lung function, stress spirometry, and transthoracic echocardiography. Patients with suspected pulmonary hypertension related to chronic obstructive pulmonary disease underwent a right heart catheterization with assessment of the hemodynamics at baseline and after regional vasodilators, including nitric oxide eventually with the endoluminal blockage of the ipsilateral pulmonary artery. This was done to avoid operation in patients with latent pulmonary hypertension or who were unresponsive to nitric oxide. Figure 1 shows the oncologic algorithm systematically used in all eligible patients; mediastinal staging included a thorax computed tomography scan, 2-deoxy-2-[18F] fluoro-D-glucose positron emission tomography (FDGPET), and mediastinoscopy. Precise delineation of airway invasion was obtained by rigid bronchoscopy, during which serial biopsies of the mucosa and submucosa lying at least 2 cm from the grossly tumoral surface were taken routinely, as well as on the contralateral tracheobronchial angle and main bronchus. Endoesophageal sonography was performed to evaluate the tumor’s extension to the esophagus, left atrium, and biopsy inferior mediastinal nodes, if present.
Preoperative Preparation
Fig 1. Algorithm of eligibility to carinal resection for nonsmall cell lung cancer. *Only patients without involvement of more than two N2 nodal levels below station 2 and N3 lymph nodes were enrolled. (PET ⫽ positron emission tomography.)
resection was required to completely remove the tumor burden; no more than two N2 nodal levels involvement below station 2 [7, 11]; no N3 lymph nodes; no distant metastasis; and normal cardiopulmonary, liver, and renal function. Preoperative findings of asymptomatic superior vena cava (SVC) invasion or limited infiltration of the muscular wall of the intrathoracic esophagus were not operative contraindications. Chronic steroid intake was gradually but quickly reduced to 5 mg/day. Patients with positive N2 nodes diagnosed preoperatively underwent neoadjuvant radiochemotherapy including two cycles of weekly carboplatinum and paclitaxel (carboplatinum AUC 2/paclitaxel 45 mg/m 2 ) simultaneous with radiotherapy (45 Gy) delivered to the tumor-burdened area and mediastinum. In the absence of progressive disease evaluated through a new staging work-up, including positron emission tomography (PET) scan, all patients underwent surgery counting those with postchemoradiation PET-positive mediastinal nodes or intraoperative mediastinal nodes positive frozen section, or both. A redo mediastinoscopy after chemoradiation therapy was not performed to avoid excessive devascularization and preserve the mobility and flexibility of the upper intrathoracic portion of the trachea.
Preoperative Assessment All patients underwent routine functional investigations including quantitative ventilation/perfusion scans to pre-
Patients were encouraged to stop smoking, and a minimum of 10 days abstinence was required. They were usually hospitalized 3 to 5 days preoperatively to become familiar with the mechanism of breathing, coughing, and swallowing, and to have twice-daily aerosol therapy.
Anesthesia Patients received total intravenous anesthesia with propofol, remifentanil, and cisatracurium as muscle relaxants, and epidural anesthesia using a continuous infusion of ropivacaine 0.2% and sufentanil 0.5 g/mL,
Fig 2. Operative view showing a median sternotomy with an Octopus (O) stabilizer retracting the heart to the right side, and the catheter (C) of apnoic oxygenation placed into the distal trachea, which allows oxygenation and undisturbed and uninterrupted placement of airway sutures. Retracting the heart to the right not only increases the surgical exposure of the affected left hilum, especially the dorsal aspect, but also improves the hemodynamic stability during the manipulation and securing of the pulmonary veins.
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
1991
after a loading dose. Electroencephalographic monitoring (Narcotrend; MT Monitor Technik, Munich, Germany) allowed adaptation of propofol infusion rates according to the depth of narcosis. Besides the usual monitoring of electrocardiogram with 5 leads, patients had continuous monitoring of arterial blood pressure and gases. Patient were ventilated initially and after airway reconstruction through a left-sided double-lumen tube (rightsided procedures) or an extralong armored endotracheal tube positioned in the right main stem bronchus or intrathoracic trachea (left-sided procedures or carinal resections alone) with low tidal volume (volume or pressure) controlled techniques. High inspiratory oxygen concentrations, multiple collapse and reexpansions, hypoxic pulmonary vasoconstriction and hypoperfusion of the ipsilateral lung, and fluid overload were avoided in all patients. Just before airway section, and during resection and reconstruction, patients were managed with the apneic (hyper)oxygenation technique [12], as follows. Patients were preoxygenated and hyperventilated with 100% oxygen (O2) for about 10 minutes before completing the dissection to reach arterial pO2 and pCO2 levels of 450 or greater and 28 to 35 mm Hg, respectively. The airway was then sectioned and resected under total apnea. After completion, hyperoxygenation was obtained by placing a small (10F) catheter accross the surgical field into the contralateral main bronchus, fixing it to the catilagenous ring (about at the 12 o’clock position) to prevent its dislocation, and connecting it to a sterile line delivering 10 to 15 L/min O2 continuously under minimal breathing pressure (0 to 1 mm Hg). Permissive hypercapnia was eventually treated. After reconstruction, patients were volume-controlled ventilated through the original tube.
Fig 3. Illustration of the secondary end-to-side anastomosis of the right residual stump on the lateral aspect of the left main bronchus. (A) A double-ended polydioxanone (PDS; Ethicon, Sommerville, New Jersey) 3-0 or 4-0 suture is started on one or the other edges of the deepest or posterior aspect of the anastomosis, picking up the full layers of the respective bronchial walls, and inserting the needles from inside the lumen. It must be left untied to allow a number of stitches to be placed to complete the entire deepest aspect of the anastomosis before the reimplanting bronchus is pulled down onto the hosting structure. (B) The double-ended suture is then pulled tight, as the reimplanting and recipient lumens are parachuted. We recommend using the exact mucosa and stitch apposition through nerve hooks to avoid pursestring and narrows effects. (C) Several concentric interrupted 3-0 to 5-0 polyglactin (Vicryl; Ethicon) sutures are placed on the remaining quadrants 3 to 4 mm apart and 3 to 4 mm from the cut edge of the airway, leaving the membranous wall of the reimplanting bronchus until last to allow balancing of any anastomotic disparity and excessive traction on it. We found that placing small Vicryl sutures (eg, 5-0) on the membranous wall reduces the manipulation risks of injury. Once placement is completed, the wall and sutures are gently approximated, placing knots outside walls.
GENERAL THORACIC
Ann Thorac Surg 2006;82:1989 –97
GENERAL THORACIC
1992
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
Ann Thorac Surg 2006;82:1989 –97
tors during clamping by fluid implementation and regulation of mean arterial pressure with norepinephrine. To keep cerebral perfusion pressure greater than 60 mm Hg, mean arterial pressure was increased at least to 60 mm Hg above the jugular bulb pressure. Cardiac output was kept within the normal range by cristalloid or colloid infusions or epinephrine, the goal being to conserve jugular bulb oxygen saturation greater than 50%. Since the SVC step was usually made before the airway opening, patients were moderately hyperventilated to decrease cerebral blood volume, and mean arterial pressure and extent of hyperventilation were adjusted to the jugular bulb blood gas values. After declamping, mannitol 20% was administered to mitigate brain edema.
Operative Techniques The basic surgical principles of carinal resection and the end-to-end primary anastomosis outlined by Grillo and associates [3, 4, 8, 15, 16] and by us [11, 17] were followed. A few points deserve, however, reemphasis. Carinal resection without sacrifice of pulmonary parenchyma and left-sided carinal pneumonectomies were usually approached through a median sternotomy. Right-sided lesions necessitating a carinal resection with sacrifice of the pulmonary parenchyma were best approached through an ipsilateral, muscle-sparing dorsal thoracotomy in the fifth intercostal space.
INCISIONS.
Dissection was limited to the anterior surface of the lower trachea and the first 2 cm of the contralateral main bronchus while preserving as much as possible any local blood supply. Airway mobilization was only to the extent needed and limited to no more than 2 cm from the proposed proximal and distal lines of transection. All patients had a complete nodal dissection, and the lymphadenectomy was made by individually securing the nourishing node vessels. For left carinal pneumonectomies, further exposure of the left hilum was obtained through the cardiac Octopus suction stabilizer (Medtronic, Minneapolis, Minnesota) applied to the apex of the heart, permitting thereby to lift the heart laterally (right) or vertically (cranially) or put it wherever needed, always under inotropic support (Fig 2); moreover, the left lower vein was, whenever possible, sectioned with the commercially available movable endostaplers.
DISSECTION.
Fig 4. Drawing of the pericardiophrenic release (lateral view). In patients requiring a left carinal resection in which tension at the anastomosis might be anticipated, one can made at the end of the operation a surgical separation from the pericardium from the diaphragm by sharp dissection or electrocautery from ventrally to dorsally and between the two phrenic nerves. A further 2 to 3 cm can be gained in the immediate postoperative period. The hole needs to be drained for few days postoperatively.
For patients with a preoperative suspicion of a SVC invasion, we also continuously monitored the jugular bulb pressure through a right internal jugular vein cathether placed retrogradely up to the jugular bulb; the cerebral oxygenation noninvasively by near-infrared spectroscopy; and the continuous cardiac output (PiCCO; Pulsion Corp, Munich, Germany) through a femoral artery cathether. A venous access into the inferior vena cava through an indwelling catheter was placed as were other lines on the dorsal foot and in both femoral veins. These monitoring devices served to control circulation during caval clamping. Based on our experience [13], and that acquired for traumatic head injury by others [14], the main targets during SVC clamping were to compensate for the cranial venous hypertension, systemic hypotension due to the impaired venous return, and reduced cerebral blood flow. Neuroprotective measures included the administration of corticosteroids about 30 minutes before clamping, and optimization of circulatory indica-
For uniformity and teaching purposes, we always used the same anastomotic principle to perform the primary end-to-end and secondary endto-side [18] anastomoses. For carinal pneumonectomy, the end-to-end reconstruction started on the deepest anastomotic part of the left half of both cartilaginous walls or left margin of the membraneous portion. For carinal lobectomies, the choice to reimplant the residual bilobe end to side either into the trachea or left main bronchus depended on the length of the residual intermedius bronchus, the status of the trachea (fibrotic, malacic, and so forth), the extent of node dissection, and the degree of traction on hilus residual vessels. Over
ANASTOMOTIC TECHNIQUE.
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
Table 1. Patients’ Demographics
Table 3. Type of Anastomosis
Demographic
Value
Sex (male vs. female) Age, years (range)
41 vs. 9 56 ⫾ 13 (22–77) 35/15
Histology (squamous/nonsquamousa) Type of preoperative tumor invasion ⬍ 0.5 cm from origin carina Ipsilateral tracheobronchial angle Carina only Postpneumonectomy stump (relapse) Side (right/left) ppFEV1 (%) Secondary pulmonary hypertension Prior surgeryb ASA score (1/2/3) Neoadjuvant chemoradiation
23 (46%) 22 (44%) 3 (6%) 2 (4%) 37/13 54 ⫾ 12 7 (14%) 7 (14%) 20/13/17 18 (36%)
a The 15 nonsquamous tumors included 9 adenocarcinomas and 6 large b cells carcinomas. Includes 2 exploratory thoracotomics, 3 lower bilobectomies, and 2 pneumonectomies stump relapses performed 22 ⫾ 18 months earlier.
ASA ⫽ American Society of Anesthesiologists Classification of physical status; ppFEV1 ⫽ predicted postoperative forced expiratory volume in 1 second.
time, however, we have preferred to perform the end-toside secondary anastomosis on the left main bronchus (Fig 3). During the end-to-side reimplantation, the residual ipsilateral pulmonary artery was always left unclamped to avoid hypoxic pulmonary vasoconstriction and hypoperfusion of the residual lobe or hyperperfusion of the contralateral lung. The anastomosis was left uncovered (right reconstructions) or partly covered with viable neighboring tissue (left reconstructions). Release maneuvers were limited to the development of the pretracheal plane, either during mediastinoscopy or through the operative field, performance of an inferior U-shaped hilar release, and for left-sided primary or secondary end-to-side anastomoses in which tension was still too high, a pericardiophrenic release with separation of the pericardium from the diaphragm between both phrenic nerves was made at the end of the operation (Fig 4). RELEASE MANEUVERS.
Postoperative Care Hyperinflation of the residual lung was avoided while satisfactory oxygenation and ventilation was obtained by Table 2. Relation Between Type of Operation and Side of the Tumor Type of Operation Carinal Carinal Carinal Carinal Carinal Total
pneumonectomy lobectomy bilobectomy resection and reconstruction resection
1993
Right Side
Left Side
24 6 5 — 2 37
10 — — 3 — 13
Number Primary end-to-end Trachea to left main bronchus Trachea to right main bronchus Secondary end-to-side Residual right upper lobe bronchus to trachea Residual bronchus intermedius to trachea Residual bronchus intermedius to left main bronchus Residual left main bronchus to right main bronchus
37 13 3 2 6 3
adequate pain relief with epidural analgesia or patientcontrolled analgesia, chest physiotherapy, and repeat aspirative fibroscopies. In patients having a pneumonectomy, the nasogastric tube was left in place for 24 to 36 hours. A temporary minimal invasive tracheotomy was performed at the end of operation to reduce the physiologic respiratory dead space and facilitate direct aspiration whenever the predicted residual ventilatory function reserve was borderline and the patient’s collaboration reduced. Patients with SVC revascularization received intravenous heparin 4 to 6 hours postoperatively until reaching a therapeutic prothrombin level and full mobilization was obtained. Then, aspirin was substituted and given for 6 months postoperatively [18].
Adjuvant Therapy All patients entering the neoadjuvant protocol received two additional courses of carboplatinum (AUC 6) and paclitaxel (175 ng/m2). Patients with intraoperative first detected positive N2 nodes received four courses of carboplatinum (AUC 2) and paclitaxel (45 mg/m2) chemotherapy and 50 Gy radiotherapy delivered to the mediastinum only.
Statistical Methods Data are presented as mean ⫾ SD of number (n) of observations. Survival was measured from the date of operation to death and included postoperative deaths. Disease-free survival (DFS) was measured from the date of surgery until the first evidence of tumor recurrence, where tumor recurrence included local-only failure, disTable 4. Relation Between Type of Operation and Mortality and Morbidity Operative Mortality
Operative Morbidity
pneumonectomy (n ⫽ 34) lobectomy (n ⫽ 6) bilobectomy (n ⫽ 5) resection-reconstruction
1 (2%) — 1 (2%) —
8 (16%) 3 (50%) 3 (60%) 3 (100%)
resection (n ⫽ 2)
— 2 (4%)
Type of Operation Carinal Carinal Carinal Carinal (n ⫽ 3) Carinal Total
1 (50%) 18 (36%)
GENERAL THORACIC
Ann Thorac Surg 2006;82:1989 –97
GENERAL THORACIC
1994
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
Ann Thorac Surg 2006;82:1989 –97
All but the very first patient were extubated in the operating room after 7 ⫾ 5 minutes from the end of skin closure. Operative mortality (less than 30 days) was 4% and included a massive myocardial infarction and a lung embolism. Complications were observed in 18 patients (36%) and were supraventricular arrhythmias (n ⫽ 7), cerebral transient ischemic attack (n ⫽ 1), cardiac luxation requiring surgical intervention (n ⫽ 1), anticoagulation-related bleeding (n ⫽ 1), and anastomotic leaks in 8 patients successfully managed either conservatively (temporary drainage of the ipsilateral cavity, n ⫽ 4) or through immediate reoperation and anastomosing (n ⫽ 3) or tissue engineered airway interposition (n ⫽ 1; Table 4) [22]. Beyond the 2 operative deaths, the 2 patients having a transient ischemic attack and cardiac luxation required mechanical ventilation postoperatively for 3 ⫾ 2 days. All patients but 1 (98%) with residual microscopic disease (R1) on the ipsilateral residual pulmonary artery stump were completely resected. The frequency of nodal involvement compared with the type of operation is shown in Table 5. The overall number of resected nodes per patient was 9 ⫾ 2, and the number of positive nodes was higher in N2 patients (4.7 ⫾ 2.9) than in N1 patients (2.7 ⫾ 3.1; p ⫽ 0.054). Among the 18 neoadjuvant chemoradiation N2 patients, 11 (61%) had no pathologic response while 7 (39%) had a histologic downstaging to N1 status (n ⫽ 2) or N0 status (n ⫽ 5). In contrast, 7 of the residual 32 patients (22%) had pathology-positive N2 nodes, despite negative preoperative PET scan. With a median follow-up time of 38 months, the projected 5-year survival and DFS were 51% and 47%, respectively. Survival was significantly affected by the need for postoperative ventilation (p ⫽ 0.013) and development of anastomotic complications (p ⫽ 0.0236) in the univariate but not in the multivariate analysis. In contrast, DFS was significantly affected by the endobronchial extension of the primary tumor (p ⫽ 0.033) and nodal status (p ⫽ 0.024) in the univariate and multivariate analysis (Table 6). Among 22 patients with tumor recurrence, one recurrence was locoregional and the others were systemic, as solitary (n ⫽ 14) or multiple (n ⫽ 7) first tumors.
Table 5. Frequency of Nodal Status for Type of Operations Nodal Status Type of Operations Carinal Carinal Carinal Carinal Carinal Total
pneumonectomy (n ⫽ 34) lobectomy (n ⫽ 6) bilobectomy (n ⫽ 5) resection and reconstruction (n ⫽ 3) resection (n ⫽ 2)
N0
N1
N2
12 3 3 2 2 22
7 2 1 0 0 10
15 1 1 1 0 18
tant-only failure, or simultaneous local and distal failure. Patients who died from causes unrelated to NSCLC were censored at the time of death in the estimation of DFS. Survival and DFS were computed using the KaplanMeier method [19]. Estimates based on categorical clinical variables were compared using the log-rank test [20], and estimates based on continuous variables were compared by using Cox regression analysis [21]. Variables for which the p value was less than 0.05 in the univariate analysis were included in a Cox proportional hazards multivariate regression model. All computations were made using Statview software (SAS Institute, Cary, North Carolina).
Results Patients’ preoperative characteristics are depicted in Table 1. Among them, 18 N2 patients had neoadjuvant chemoradiation with a delivered radiation dose of 48 ⫾ 6 Gy. Table 2 correlates type of operation to tumor side. The length of resected airway was 2.8 ⫾ 0.6 cm, and airway reconstruction necessitated a total of 64 anastomoses (Table 3). For the 38 patients managed with the apneic oxygenation technique, the duration of the apneic and apneic hyperoxygenation phases were 5 ⫾ 2 and 26 ⫾ 11 minutes, respectively; none had specific-related morbidity. Twelve patients (24%) underwent concurrent truncular resection of the SVC (n ⫽ 4), partial excision of the muscular esophageal wall (n ⫽ 2), diaphragm (n ⫽ 3), chest wall (n ⫽ 2) and left atrium (n ⫽ 1). The SVC clamping time was 16 ⫾ 3 minutes.
Table 6. Multivariate Analysis for Disease-Free Survival Factor Endobronchial extension (p ⫽ 0.02) ⬍ 0.5 cm from carina Tracheobronchial angle Carina only Nodal status (p ⫽ 0.04) N0 N1 N2
Number of Patients
Number of Failures
Hazards Ratio
95% Confidence Interval
p Value
23 22 3
6 12 1
0.047 0.159 0.153
0.004 to 0.544 0.015 to 1.73 0.008 to 3.070
0.01 0.014
21 11 18
4 7 9
0.192 5.557 5.215
0.004 to 0.76 1.35 to 22.71 1.30 to 20.90
0.017 0.019
Cox proportional hazards regression was used to determine the influence of clinically significant factors on disease-free survival.
Comment The main innovations introduced in the present study were the restriction of patients’ eligibility, performance of routine preoperative PET scan, neoadjuvant chemoradiation therapy for N2 patients, more physiologic intraoperative management, and whenever feasible, lungsparing resections and innovative technical aspects. That produced lower mortality rates but equal morbidity rates compared with recent reports [9, 10, 23] and an encouraging early outcome. The presence of metastatic mediastinal nodes in patients requiring carinal resection has always been considered a potential contraindication in the past [7, 8], and more recently as well [9, 10]. Despite its potentiality to improve outcome [24], however, neoadjuvant chemotherapy has never been systematically employed in newly diagnosed N2 patients requiring carinal reconstruction, mainly because of its association with increased operative mortality [9]. This study shows that neoadjuvant chemoradiation is feasible, pathologically downstages the N2 status in less than 40% of patients, improves DFS compared with more recent studies [9 –10], and finally, permits the performance of more complex carinal reconstruction without significantly influencing either their operative morbidity or mortality. These results may be mainly explained by our practice of contraindicating surgery for patients with invasion of more than two mediastinal nodal stations, as well as positive station 2R/L nodes or N3 nodes, rather than the ability of PET scan to evaluate the mediastinum. Moreover, that almost one quarter of PET-negative patients had micometastatic positive nodes provides definitive evidence that mediastinoscopy cannot be excluded in the evaluation of the patient amendable to carinal surgery, as per any other NSCLC surgery [25], and should routinely be performed preoperatively [9, 10]. A similar consensus on the role of lymphadenectomy does not exist, it being the recommendation to perform lymph node dissection only of the immediate adjacent or clearly involved nodes [16]. Given the relatively high incidence of micrometastatic nodes undetectable at PET scan and given that the performance of radical lymphadenectomy did not increase airway or residual parenchymal morbidity, we might speculate that radical lymphadenectomy should be recommended to increase the therapeutic index of this surgery, especially after recent and growing evidence of the benefits of adjuvant chemotherapy for patients with completely resected NSCLC [26, 27]. Whether patients whose N2 status remains positive either at restaging (PET) or postoperatively (frozen section) should still be resected remains to be determined, but at least in this series, their DFS was encouragingly superior when compared with our previous N2 carinal patients without induction therapy [11]. The other main innovation, in our opinion, was to avoid intraoperative barotrauma, the predominant factor associated with adult respiratory distress syndrome (ARDS) [28] and the leading cause of mortality after carinal resection [8, 23]. To achieve this, ventilation was
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
1995
made through low tidal volume controlled techniques; and high inspiratory oxygen concentrations, multiple collapse and reexpansions, hypoxic pulmonary vasoconstriction during hypoperfusion of the ipsilateral lung parenchyma or hyperperfusion of the contralateral lung, and fluid overload were maximally avoided. All these factors reduce lung compliance [29, 30] and have been related to the occurrence of ARDS after pneumonectomy [31]. That we did not notice any ARDS postoperatively, despite radical lymphadenectomy, strongly suggests that these maneuvers may have played a fundamental role in avoiding this devastating complication. Moreover, apneic (hyper)oxygenation techniques were used, and their continuous capability to oxygenate and not ventilate the patients provides a more physiologic means of sustaining gas exchange and less volume, barotrauma, and biotrauma to the residual lung parenchyma [32, 33]. Briefly, before sectioning the airway, the patient is preoxygenated and hyperventilated with 100% oxygen only so that the 2,500 to 3,000 mL volume of the functional residual capacity is almost completely denitrogenated and any nitrogen entrainment is ceased. As the normal oxygen consumption is 200 to 250 mL per minute, the 2,500 to 3,000 mL intrapulmonary oxygen store provides adequate oxygenation for 10 to 12 minutes of total apnea. Furthermore, after the opening of the airway, a pediatric catheter is placed under visualization above the carina; and by simply delivering a small flow (10 to 15 L/min oxygen) under minimal breathing pressure (0 to 2 psi) to both lungs, the patient can “breathe” despite ongoing total apnea and can survive even apneic periods of 50 minutes, without any ill effects. In contrast to McClish and colleagues [34], we do not ventilate the lungs and use a low pressure (0 to 2 psi) and small flow (10 to 15 L/min oxygen) delivering system. This represents, in our opinion, a more physiologic gas exchange method while keeping the significant advantages of the catheter technique, for example, improved surgical exposure, minimal intraoperative intrusion of the anesthetic apparatus, and facilitated reconstruction. In the present series, a higher number of left carinal pneumonectomies and carinal lobectomies were made— both, surgically speaking, challenging issues. Although the main dispute concerning left carinal surgery has been the surgical approach [15, 35], we reinforce the advantages of median sternotomy, for example, excellent exposure of the tracheobronchial bifurcation and less incisional discomfort [11, 17] while proposing the use of heart stabilizers such as, in our experience, the Octopus device to lift the heart whenever the need is to improve exposure of the left hilum and release pleuroparietal adhesion. The only disadvantage of this incision is when a lymphadenectomy of a nodal station below 7 is needed, but with increasing experience, an additional left thoracotomy may be avoided by using the heart retractor and opening the posterior pericardium vertically on the midline. Hence, for more complex left reconstructions, median sternotomy allows the surgeon to further reduce the anastomotic tension by separating the pericardiodiaphragmatic reflections between the two phrenic nerves, a
GENERAL THORACIC
Ann Thorac Surg 2006;82:1989 –97
GENERAL THORACIC
1996
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
release maneuver that permits an upward gain of almost 2 cm. Care should be taken, however, to avoid injury of the right phrenic nerve because in almost every left carinal pneumonectomy, the left one will be sacrified. Because of the early and late issues related to carinal plus lobar resection, these operations are rarely performed, and preference has be given to perform carinal pneumonectomy if the patient can otherwise tolerate such a procedure [4, 16]. Based on our previous [18] and this experience, our data may counter this statement, from an oncologic and technical point of view. However, if reimplantation is to be performed (either bronchus intermedius or right lower lobe bronchus), the secondary end-to-side anastomosis should be to the side of the left main bronchus and not to the trachea, not only to avoid excessive tension but also because if a minimal anastomotic leak occurs, it will spontaneously heal through the collapse of the mediastinal and lung tissue over its surface if the pleural space is sufficiently well drained. In conclusion, this study provides evidence that some traditional pitfalls related to carinal resection in operable NSCLC can be avoided. While the oncologic outcome may be improved by contraindicating surgery for “massive” N2, when performing routine mediastinoscopy and radical lymphadenectomy, the operative mortality and some fatal morbidity, for example, ARDS, may be minimized by using intraoperative lung-protective ventilation techniques, and avoiding high inspiratory oxygen concentrations, multiple collapse and reexpansions, hypoxic pulmonary vasoconstriction during hypoperfusion of the ipsilateral lung parenchyma or hyperperfusion of the contralateral lung, and fluid overload. Technically speaking, left carinal pneumonectomy or more complex rightsided and lung-sparing reconstruction may be safely performed using the presented modus operandi. The authors would like to thank all members of the Hannover Interdisciplinary Intrathoracic Tumor Task Force Group who not only contributed to the accomplishment of the study protocols but also actively participated in the enrollment of the patients: Drs Hartmut Kirchner, Joachim Pieper, Markus Sosada, Jürgen Borghardt, Bernd Gaede, Felix Winkler (medical oncologists); and Drs Bernd Wolfgang Raack, Alexander Schulz, Hans-Peter Dirks, Matthias Kleckow, Werner Mall, Bern Vogel, Jost Niedermeyer, and Helmut Fabel (pneumologists). We are also thankful to all referring national and international doctors, and to Jeff Szychowski, PhD, University of Alabama Tuscaloosa, for the invaluable statistical assistance.
References 1. Gibbon JH. Discussion of Chamberlein JM, McNeil TM, Parnassa P, Edsall JR. Bronchogenic carcinoma: an aggressive surgical attitude. J Thorac Cardiovasc Surg 1959;38: 727– 45. 2. Jensik RJ, Faber LP, Kittle CF, Miley RW, Thatcher WC, El-Baz N. Survival in patients undergoing tracheal sleeve pneumonectomy for bronchogenic carcinoma. J Thorac Cardiovasc Surg 1982;84:489 –96. 3. Grillo HC. Carinal reconstruction. Ann Thorac Surg 1982;34: 356 –73. 4. Mathisen DJ, Grillo HC. Carinal resection for bronchogenic carcinoma. J Thorac Cardiovasc Surg 1991;102:16 –23.
Ann Thorac Surg 2006;82:1989 –97
5. Tsuchiya R, Goya T, Naruke T, Suemasu K. Resection of tracheal carina for lung cancer. J Thorac Cardiovasc Surg 1990;99:779 – 87. 6. Roviaro GC, Varoli F, Rebuffat C, et al. Tracheal sleeve pneumonectomy for bronchogenic carcinoma. J Thorac Cardiovasc Surg 1994;107:13– 8. 7. Dartevelle P, Macchiarini P, Chapelier A. 1986: tracheal sleeve pneumonectomy for bronchogenic carcinoma. Ann Thorac Surg 1995;60:1854 –5. 8. Mitchell JD, Mathisen DJ, Wright CD, et al. Resection for bronchogenic carcinoma involving the carina. Long-term results and effect of nodal status on outcome. J Thorac Cardiovasc Surg 2001;121:465–71. 9. de Perrot M, Fadel E, Mercier O, Mussot S, Chapelier C, Dartevelle P. Long-term results after carinal resection for carcinoma: does the benefit warrant the risk? J Thorac Cardiovasc Surg 2006;131:81–9. 10. Regnard JF, Perrotin P, Giovannetti R, et al. Resection for tumors with carinal involvement: technical aspects, results, and prognostic factors. Ann Thorac Surg 2005;80:1841– 6. 11. Dartevelle P, Macchiarini P. Carinal pneumonectomy for bronchogenic carcinoma. Semin Thorac Cardiovasc Surg 1996;8:414 –25. 12. Go T, Altmayer M, Richter M, Macchiarini P. Decompressing manubriectomy under apneic oxygenation to release the median thoracic outlet compartment in Bechterew disease. J Thorac Cardiovasc Surg 2003;126:867–9. 13. Macchiarini P, Dartevelle P. Resection and reconstruction of the superior vena cava. In: Pearson FG, Cooper JD, Deslauriers J, et al, eds. Thoracic surgery. 2nd ed. New York: Churchill Livingston, 2002:1774 – 80. 14. Rosner MJ, Rosner SD, Johnson AH. Cerebral perfusion pressure: management protocol and clinical results. J Neurosurg 1995;83:949 – 62. 15. Mitchell JD, Mathisen DJ, Wright CD, et al. Clinical experience with carinal resection. J Thorac Cardiovasc Surg 1999; 117:39 –53. 16. Mitchell JD. Carinal resection and reconstruction. Chest Surg Clin North Am 2003;13:315–29. 17. Dartevelle P, Macchiarini P. Techniques of pneumonectomy. Sleeve pneumonectomy. Chest Surg Clin North Am 1999;9: 407–17. 18. Di Rienzo G, Go T, Macchiarini P. Simplified technique for the secondary end-to-side anastomosis in complex tracheobronchial reconstruction. J Thorac Cardiovasc Surg 2002;124: 632–5. 19. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Ass 1958;53:457– 81. 20. Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data. 2nd ed. New York: John Wiley & Sons, 2002. 21. Cox DR. Regression models and life tables. J Roy Stat Soc Series B 1972;34:187–220. 22. Macchiarini P, Walles T, Biancosino C, Mertschning H. First human transplantation of a bioengineered airway tissue. J Thorac Cardiovasc Surg 2004;128:638 – 41. 23. Porhanov VA, Poliakov IS, Selvaschuk AP, et al. Indications and results of sleeve carinal resection. Eur J Cardiothorac Surg 2002;22:685–94. 24. Betticher DC, Hsu Schmitz SF, Totsch M, et al. Mediastinal lymph node clearance after docetaxel-cisplatin neoadjuvant chemotherapy is prognostic of survival in patients with stage IIIA pN2 non-small-cell lung cancer. A multicenter phase II trial. J Clin Oncol 2003;21:1752–9. 25. Kernstine KH. Positron emission tomography with 2-[18F]fluoro-2-deoxy-D-glucose: can it be used to accurately stage the mediastinum in non–small cell lung cancer as an alternative to mediastinoscopy? J Thorac Cardiovasc Surg 2003; 126:1700 –3. 26. Arriagada R, Bergman B, Dunant A, et al. Cisplatin-based adjuvant chemotherapy in patients with completely resected non-small-cell lung cancer. N Engl J Med 2004;350: 351–360.
27. Winton T, Livingston R, Johnson D, et al. Vinorelbine plus cisplatin vs. observation in resected non-small-cell lung cancer. N Engl J Med 2005;352:2589 –97. 28. Mathisen D. J Thorac Cardiovasc Surg 2006;131:87. 29. Fessler HE, Brower RG. Protocols for lung protective ventilation. Crit Care Med 2005;33(Suppl):S223–7. 30. Demling RH. The modern version of adult respiratory distress syndrome. Ann Rev Med 1995;46:193–202. 31. Jordan S, Mitchell JA, Quinlan GJ, Goldstraw P, Evans TW. The pathogenesis of lung injury following lung resection. Eur Respir J 2000;15:790 –9.
MACCHIARINI ET AL BRONCHOGENIC CANCER CARINAL SURGERY ADVANCES
1997
32. Zander R, Mertzlufft F. Clinical use of oxygen stores: preoxygenation and apneic oxygenation. Adv Exp Med Biol 1992;317:413– 420. 33. Frumin JM, Epstein RM, Cohen C. Apneic oxygenation in man. Anesthesiology 1959;20:789 –98. 34. McClish A, Deslauriers J, Beaulieu M, et al. High-flow ventilation during major tracheobronchial reconstruction. J Thorac Cardiovasc Surg 1985;89:508 –12. 35. Maeda M, Nakamoto K, Tsubota N, Okada T, Katsura H. Operative approaches for left-sided carinoplasty. Ann Thorac Surg 1993;56:441–5.
DISCUSSION DR ROBERT J. CERFOLIO (Birmingham, AL): This is an outstanding series on a very difficult technical operation, and the fact that you have done 50 of these over this short time frame really shows that you guys are experts in this and you deserve a lot of credit. I have just a couple of quick questions for you. I notice that you said 39% of your N2 patients were downstaged, but that leads me to believe that 61% weren’t. Do you normally routinely send your lymph nodes for frozen section prior to removing the lung? I know you don’t do repeat mediastinoscopy; thus, do you send them off at the time of the thoracotomy, and if there is residual N2 disease on frozen, do you abort the operation?
make that anastomosis easier? Again, excellent presentation and outstanding work.
DR MACCHIARINI: During the initial mediastinoscopy, we sample all available lymph node stations. During surgery we do a radical lymphadenectomy/frozen section. If mediastinoscopy is positive by the first time, we usually continue as long as we do not have progressive disease.
DR FRANCIS C. NICHOLS III (Rochester, MN): I have two questions. One, the right pneumonectomy group, how many of those patients had preoperative chemotherapy and radiation therapy? The second question, and I do have the advantage of having the abstract, your overall number of resected nodes was 9 nodes per patient, so if you’re doing a radical mediastinal lymphadenectomy, how do you account for just 9 nodes?
DR CERFOLIO: No, I mean after they have had radiation and chemotherapy and you come back to resect them, do you send the nodes off for frozen to see if they have been downstaged or do you just proceed with resection? DR MACCHIARINI: We send it to frozen and proceed to resection. DR CERFOLIO: So if the nodes are still positive and there is residual N2 disease in a patient who was initially N2, you would still do a carinal pneumonectomy? DR MACCHIARINI: Exactly. DR CERFOLIO: That’s pretty aggressive. My other question is just a technical tip that perhaps with your large experience you could help me, who is less experienced. When I try to bring that right main stem up and reimplant it up into the trachea, just like you show, I have always been taught that it is nice to have a couple of centimeters between your end-to-end distal trachea to left main stem anastomosis and your new right main stem side to trachea anastomosis— but sometimes it’s hard to get that lung all the way up there, besides doing pericardial releases and all these other maneuvers. Do you have any other tricks or techniques you can teach me to help
DR MACCHIARINI: Thank you, Robert. I think that the minimum that you should have is 1 cm. If you have 1 cm, you can bring it over in the trachea. However, I would not suggest to anastomose that in the trachea because you already devascularized the upper portion, you already have a mediastinoscopy, and you already have a radical lymphadenectomy. On the other hand, simply putting on the left main bronchus, the tissues that are there are very flexible, so if you have a small leakage, you can manage that conservatively.
DR MACCHIARINI: The first question was? DR NICHOLS: Right pneumonectomy, and how many of those patients had basically radiation therapy is what I’m interested in. DR MACCHIARINI: I need to pass. I don’t know the figures. Concerning the second question, one of the eligibility criteria was to avoid operating on patients with more than two N2 levels, and this is because the previous experience by Philippe Dartevelle showed us that if you have this situation, you have 0% 5-year survival. So this might be related to the fact that we have a mean 9 plus or minus—I don’t know how many. Probably this is a statistical issue. DR MARK J. KRASNA (Baltimore, MD): I enjoyed the presentation. Rendina in Rome and our group in Maryland have described using a routine intercostal muscle flap. I’m wondering if your group now advocates using that routinely, not only when you’re doing both a bronchial and a vascular anastomosis, but even just around your carinal anastomosis. Have you been using that routinely or not? DR MACCHIARINI: No, sir.
GENERAL THORACIC
Ann Thorac Surg 2006;82:1989 –97