| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;83:402-408
© 2007 The Society of Thoracic Surgeons
Thoracic Surgery Division, Tor Vergata University School of Medicine, Rome, Italy
Accepted for publication September 1, 2006.
* Address correspondence to Dr Pompeo, Cattedra di Chirurgia Toracica, Policlinico Tor Vergata, V. le Oxford 81, Rome 00133, Italy (Email: pompeo{at}med.uniroma2.it).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
METHODS: Between March 2003 and December 2005, 22 patients with NSCLC deemed unresectable owing to signs of proximal vascular invasion at multislice computed tomography were included in a nonrandomized prospective study entailing FVP performed before curative-intent thoracotomy. Primary outocome measures were technical feasibility (graded from 1 = poor to 3 = excellent) and quality of anatomic visualization (graded as diagnostic or nondiagnostic). Patients with FVP evidence of proximal vascular invasion did not receive thoracotomy and were included in an induction chemoradiotherapy protocol.
RESULTS: Mean operative time was 31 ± 6 minutes. Technical feasibility ranged from good to excellent in 21 patients. In 1 patient, FVP was not completed owing to intrapericardial adhesions. In 7 patients, FVP disclosed no tumor spread to intrapericardial pulmonary vessels, and radical resection was carried out. The FVP findings were considered not diagnostic because of poor visualization of the right pulmonary artery in 2 patients with bulky hilar tumors. There was no operative mortality and morbidity. Mean hospital stay was 2.9 ± 1 days for patients who did not undergo thoracotomy and 6.3 ± 1 days for patients receiving FVP plus resection. Definitive pathologic staging in patients undergoing resection was pT3N1 (n = 2), pT3N0 (n = 2), pT2N0 (n = 3), and pT2N1 (n = 1).
CONCLUSIONS: Flexible videopericardioscopy proved safely feasible and allowed a better assessment of centrally located NSCLC, with radiologic evidence of proximal vascular involvement eventually resulting in increased resectability rate.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
We have undertaken a prospective nonrandomized study aimed at evaluating the usefulness of pericardioscopy performed by a flexible videobronchoscope during exploratory video-assisted thoracoscopic surgery, namely, flexible videopericardioscopy (FVP), for patients with centrally located NSCLC who were considered unresectable owing to radiologic evidence of proximal vascular involvement.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Eligibility criteria included a diagnosis of NSCLC; good performance status (Karnofsky score >90); no medical, functional, or bronchoscopic contraindication to pneumonectomy; and clinical stage IIIB with multislice computed tomography (CT) evidence of proximal vascular invasion according to previously validated criteria [7, 8], including (1) contact between a pulmonary vessel and the tumoral mass for more than 3 cm or more than 90 degrees, or both, without a recognizable fat plane, (2) distortion or indentation of the involved vessels, or (3) obliteration of pulmonary vessel lumen.
Exclusion criteria included malignant pleural effusion, imaging-based suspicion of vertebral or esophageal involvement, histologically proven N2-3 disease, M1 disease, previous induction therapy, previous FVP, and cardiac surgery or thoracic surgery on the side targeted for FVP.
In 15 patients, histologic diagnosis was obtained preoperatively through fiberoptic bronchoscopy (n = 9) or CT-guided core-needle biopsy (n = 6). In the remaining patients, biopsy specimens were obtained during video-assisted thoracoscopic surgery.
Study Design
The study was set as a prospective nonrandomized clinical investigation. The primary outcome measure was quality of anatomic visualization, which was scored as diagnostic or nondiagnostic on the basis of the possibility to provide accurate real-time assessment of the tumor local extension. In particular, anatomic visualization was defined by intraoperative assessment of three subjective indexes by two surgeons (T.C.M., E.P.): (1) pulmonary vessel recognition; (2) tumor-tissue discernment; and (3) estimation of tumor-free proximal vessels portion. Whenever all three indexes were judged satisfactorily assessed by both surgeons, the result was scored as diagnostic, whereas if at least one of the three indexes was considered inadequately assessed or consensus was not reached, the result was scored as nondiagnostic.
The sample size of 22 patients was calculated assuming achieving an adequate anatomic visualization in at least 80% of patients with a lower confidence interval of more than 60% and expecting to improve resectability rate and reduce exploratory thoracotomies. An interim analysis was performed after treatment of the first 10 patients to rule out achievement of resectability through FVP in at least 3 patients.
Complete preoperative clinical staging was performed, including total body positron emission tomography scan using fluorodeoxyglucose (FDG-PET) integrated with multislice CT and fiberoptic bronchoscopy.
Patients in whom FVP indicated a resectable disease immediately proceeded to thoracotomy during the same surgical session. Patients with findings suggestive of unresectable disease at FVP did not proceed to thoracotomy and underwent neoadjuvant chemoradiotherapy. Finally, patients in whom FVP was not feasible or nondiagnostic were asked to choose between exploratory thoracotomy or neoadjuvant chemoradiotherapy.
For the study’s purpose, we defined as unresectable a tumor mass extending to the intrapericardial tract of pulmonary vessels with a tumor-free portion of less than 1 cm for pulmonary veins and of less than 2 cm for pulmonary artery.
Radiologic Assessment
Computed tomography assessment was performed using a multidetector row helical scanner (LightSpeed 16; General Electric, Milwaukee, Wisconsin), with a single-breath–holding technique. Iodinated contrast medium was administred at a rate of 3.0 mL/s to a total dose of 120 mL (Optiray 300; Tyco Healthcare, Mansfield, MA). Collimation of 1.0 mm was used, with a pitch of 6.0 mm. Data acquisition started 40 s after beginning the injection. Total scanning time was 32 s, according to the protocol by Padhani and coworkers [14]. Images were finally processed in a dedicated workstation system for three-dimensional and multiplanar reconstructions. The thoracic surgeon and the radiologist both evaluated the quality of each CT slice; consensus was required for inclusion in the study.
Surgical Technique
Flexible videopericardioscopy was performed under general anesthesia with double-lumen tube intubation and single-lung ventilation. The patient was positioned leaning 45 degrees. A first trocar was inserted in the sixth intercostal space, on the midaxillary line. After verified absence of pleural adhesions, two additional trocars were placed in the fifth and fourth intecostal spaces on the anterior axillary line. Exploration of the entire pleural cavity, lung, and mediastinum was accomplished firstly. Next, a pinhole incision was performed with hook and electrocautery on the fibrous pericardium about 1 cm anterior to the phrenic nerve. The incision was then enlarged with endoscopic shears to allow insertion of the videobronchoscope (Fig 1), and 3-0 silk sutures were passed endoscopically on both margins of the pericardiotomy, to apply traction and widen the intrapericardial space during FVP. Accurate exploration of all vascular elements within the pericardium was then accomplished, with particular attention to the vessels suspected of tumor involvement on the basis of multislice CT findings.
|
|
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The outcome and most meaningful data of the study cohort are illustrated in Figure 3 and Table 1, respectively. Technical feasibility was satisfactory in 21 patients (95.4%) and poor in 1 patient (4.6%) owing to pericardial adhesions preventing exploration.
|
|
Finally, in the subgroup undergoing resection after induction therapy, definitive staging was pT2N0 in 3 patients and pT3N0 in 1 patient.
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Thus far, the question of evaluating proximal vascular involvement of NSCLC by minimally invasive staging procedures has been scarcely investigated. In 1999, Van Schil and coworkers [15] performed intrapericardial assessment of left-sided tumors through anterior mediastinotomy. In 2002, Loscertales and coworkers [13] reported an interesting series of 27 patients with centrally located tumors in whom pericardioscopy with exploration of proximal pulmonary vessels was accomplished using rigid instrumentation. Although we acknowledge the validity of the surgical technique proposed by Loscertales and coworkers [13], we believe that the use of a fiberoptic videobronchoscope could offer some advantages in terms of safety and effectiveness because of its small diameter and flexibility. In fact, exploration of the pericardial cavity by this instrument was easily accomplished in the majority of the patients with a low risk of vascular injury. Despite the limitation provided bilaterally by pericardial sinuses and by the Botallo’s ligament on the left pulmonary artery, we have found that the videobronchoscope can easily slide around the vessel’s wall, thus allowing visualization of areas such as the posterior aspects of the pulmonary vessels, which are difficult if not impossible to be visualized by a rigid scope. Video-assisted, magnified vision allowed an accurate assessment of tumor-free portions of pulmonary vessels, which were useful for eventual stapling.
So far, the main limitation of FVP relates, in our opinion, to the presence of bulky hilar masses, which can jeopardize transformation of the virtual pericardial cavity into a real explorable space. Indeed, this transformation is achieved by tailoring the extent of the pericardial opening and applying vertical traction on the pericardial walls, maneuvers that can be blocked because of the increased rigidity of the neoplastic mass. In these instances, visualization of the intrapericardial tract of right pulmonary artery, which is partially covered by the superior vena cava, can be hampered, as happened in 2 patients of our series. Further technical improvements such as the recently reported endosonography used during videopericardioscopy [16] will probably help overcome this limitation in the near future.
One of the expected goals of FVP was to reduce exploratory thoracotomies. In our study, FVP indicated the presence of an unresectable disease in 12 of 19 patients in whom the procedure was considered diagnostic. We believe this to represent a satisfactory outcome in terms of ethical and cost-versus-benefit concerns. Furthermore, although 2 patients still received exloratory thoracotomy owing to the failure of FVP to correctly assess proximal vascular involvement, it must be taken into account that all patients had been previously considered unresectable at multislice CT.
The role of exploratory thoracotomy in locally advanced NSCLC is still a matter of debate, as it is believed by some surgeons that definitive surgical decision can be oriented only by direct palpation and accurate dissection [17–19]. However, Roeslin and coworkers [19] have found a mortality rate of 4% after exploratory thoracotomy, which also correlated with a poorer survival in comparison with the outcome of identical-stage nonsurgical patients.
In our study, only patients in whom adequate tumor-free vascular margins were clearly identified at FVP underwent thoracotomy, to minimize the risk of incomplete resection, which is associated with significantly worse 5-year survival [2, 3]. Furthermore, Pitz and coworkers [4] reported that among 89 T4 patients, 13 of the 19 surgical deaths were of patients with proximal vessel involvement, which raises questions as to the risk-to-benefit ratio of surgical therapy for this delicate patient subgroup.
An aggressive treatment strategy for stage IIIB NSCLC can entail direct inclusion of the patient in an induction therapy protocol [20]. However, we believe that the use of invasive staging methods in these patients has a rationale and can be justified. In fact, induction therapy can increase subsequent surgical morbidity and can be associated with significant toxicity [21, 22]. Moreover, marginal surgical candidates not responding to this kind of therapy may become rapidly and definitively unresectable.
In this series, all enrolled patients underwent advanced radiologic staging by total body FDG-PET and multislice CT with multiplanar reconstruction. Multislice CT was expected to yield improved details of hostile anatomic areas (namely, the aortopulmonary window or the subcarinal region) by avoiding overlapping of neighboring structures and providing better estimation of the contact between tumor mass and mediastinal structures, regardless of their axis [14, 23–25]. Nonetheless, multislice CT scan showed a 60.0% positive predictive value in assessing tumor spread to proximal vessels. This result is similar to that reported in previous studies employing conventional CT [7–11], while it is somewhat worst than that reported in one study by thin-section electron-beam CT [26] in which accuracy, sensitivity, and specificity in evaluating invasion of the pulmonary artery were 75%, 77.8%, and 71.4%, respectively. It is worth noting that, in this study, if one considers left-sided FVP only, negative predictive value was 100%, which suggests a high accuracy in detecting resectable lesions in these patients.
As far as the usefulness of magnetic resonance imaging in this setting [27, 28] is concerned, several studies have failed to demonstrate a superiority of magnetic resonance imaging versus CT in terms of diagnostic accuracy and interobserver variability [29, 30]. Other newly available endobronchial ultrasonographic procedures are being actively investigated as additional staging tools, although so far they have been mainly employed for assessment of nodal status.
Limitations
We acknowledge as an important limitation of the study the impossibility of assessing sensitivity and positive predictive value due to the lack of routine central biopsy and thoracotomy confirmation of FVP findings, which prevented us from calculating the false positive rate. Furthermore, the small sample size did not allow statistical analysis to test relationships existing between FVP results and patient- or tumor-related factors. Finally, FVP proved not feasible or gave uncertain anatomic visualization in a few instances, reducing its overall accuracy as a staging tool.
In conclusion, in this study, FVP proved safe and feasible for the majority of the patients. It allowed a better assessment of centrally located NSCLC with radiologic evidence of proximal vascular involvement, eventually resulting in increased resectability rate. Further investigation is warranted to confirm our findings and reduce the procedure-related limitations observed in our small cohort.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
