| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2007;31:1120-1124. doi:10.1016/j.ejcts.2007.02.021
Copyright © 2007, European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved
a Thoracic Surgery Division, Tor Vergata University School of Medicine, PoliclinicoTor Vergata University, Rome, Italy
b Department of Biochemistry and Biophysic "F. Cedrangolo", Section of Anatomic Pathology, Second University of Naples, Naples, Italy
Received 28 November 2006; received in revised form 15 February 2007; accepted 20 February 2007.
* Corresponding author. Address: Cattedra di Chirurgia Toracica, Università Tor Vergata, Policlinico Tor Vergata, Via Oxford, 81, 00133 Rome, Italy. Tel.: +39 06 20902884; fax: +39 06 20902881. (Email: pompeo{at}med.uniroma2.it).
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
Key Words: NSCLC • Chest wall invasion • Surgery • Immunohistochemistry • Prognostic factors • Survival
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
However, some patients suffer from a poor outcome despite an apparently complete surgical resection. These features have led to hypothesize that immunohistochemistry-detected microscopic tumor spread, which is missed by standard histopathology examination, can occur and affect outcome [8].
In this study we have analyzed the impact of immunohistochemistry-detected microscopic tumor spread on long-term outcome of en-bloc resection for NSCLC invading the chest wall.
|
2. Patients and methods |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
All medical records were reviewed for patient demographics, preoperative symptoms, pulmonary function values, type of surgical procedure, histology, and pathological stage. All cases were classified according to the present update of the TNM classification [9]. Chest wall invasion was studied and its depth classified into three levels based on pathology outcome and operative notes: parietal pleura, parietal pleura and soft tissue, and parietal pleura and soft tissue with bone involvement. The patients were followed for a minimum period of 5 years, with follow-up consisting of timed outpatient visits or telephone interviews.
2.2 Immunohistochemical analysis
We reconsidered specimens from en-bloc resection margins and hilar and mediastinal lymph nodes retrieved by routine systematic lymphadenectomy, previously negative (pN0) by conventional histopathologic technique. Presence of occult metastatic tumor cells was re-assessed by immunohistochemistry with AE1/AE3 anti-cytokeratin and anti-CEA monoclonal antibodies, specifically targeted at recognizing normal and neoplastic epithelial cells.
Immunohistochemical examinations were performed by a standard avidin–biotin peroxidase technique, using formalin-fixed, paraffin-embedded material [10]. Five 4-µm slices of single paraffin-embedded lymph node sections were transferred on glass slides. Monoclonal antibodies to wide spectrum cytokeratin (AE1/AE3 (dilution 1:50; Dako Corporation, Carpintera, CA) and anti-CEA (dilution 1:100; Dako Corporation, Carpintera, CA)) were employed. Hematoxylin–eosin was used as a routine counterstain. A positive staining within the section of the en-bloc resection margins and the lymph node was accepted as evidence of micrometastatic tumor cells.
Micrometastases are tumor cell implantations with a size of 2 mm or less inside the involved organ with extravasation, proliferation, and stromal reaction often detected by immunohistochemistry and not previously evidenced by traditional histopathologic staining [11].
Tumor vessel invasion was defined as evidence of neoplastic cells inside the lumen of the blood vessel wall, evaluated according to traditional staining. The specimen underwent serial cutting, documenting the presence or absence of infiltration of the vascular wall.
2.3 Statistics
For the study purpose all data were stored in a prospective database. Interdependence among factors was assessed by the Chi-square and Fischer's exact tests. p-values <0.05 were regarded as statistically significant in two-tailed tests. Calculation of survival was performed by means of the Kaplan–Meier method. All deaths occurring during hospitalization were defined as intraoperative. Groups were compared with the log-rank test. Factors significantly affecting survival at univariate analysis were analyzed multivariately with the Cox regression model [12].
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
Chest pain was the most common presenting symptom, occurring in 29 patients (61.7%), followed by coughing in 11 patients and fever in 9. Other symptoms included dyspnea in eight and malaise and pneumonia in four. Nine patients were asymptomatic. Pulmonary function tests revealed a mean forced expiratory volume in 1 s (FEV1) of 2.06 ± 0.69 l (range 1.10–3.75), representing 78 ± 21% of the predicted value.
3.2 Operative data
Surgical approach was standard postero-lateral thoracotomy with a muscle-sparing technique whenever feasible. Whenever a wide anterior or lateral chest wall resection was performed, the latissimus dorsi muscle was always preserved for subsequent reconstruction. We usually opened the chest cavity away from the site of invasion and manually assessed the degree of infiltration to establish the number of ribs to be resected. A free margin of at least 2 cm was left upon chest wall resection. Systematic and complete mediastinal lymph node dissection was performed in all cases. All patients underwent full-thickness en-bloc chest wall and pulmonary resection. Pulmonary resections entailed pneumonectomy in 4 patients (8.5%), lobectomy in 26 (55.3%), bi-lobectomy in 3 (6.3%), and wedge resection in 14 (29.7%). Resected ribs ranged from 1 to 4 (mean 2.8 ± 0.5). Mean chest wall defect was 48 ± 23 cm2 (range 23–126); in 8 instances the defect was sited anteriorly, in another 17 it was lateral, and in 22 posterior. Eighteen patients required chest wall reconstruction. In two patients a polytetrafluoroethylene soft-tissue patch (Goretex®, Gore and Associates, Flagstaff, AZ, USA) was used while in 16 patients a latissimus dorsi muscle flap was sufficient to cover the chest wall defect.
One perioperative death (2.1%) occurred due to pulmonary embolism on the fifth postoperative day. The patient had undergone pneumonectomy.
Median hospital stay was 13 days, ranging from 6 to 37 days.
Postoperative complications were observed in 16 patients (34.0%): atrial fibrillation (n = 6), prolonged air leak (n = 5), atelectasis pneumonia (n = 3), and bleeding (n = 2). One patient required a redo thoracotomy due to postoperative bleeding.
3.3 Histopathological findings
Postoperative pathological staging was T3N0 in 33 patients (70.2%), T3N1 in 9 (19.1%), and T3N2 in 5 (10.7%). Nodal hilar micrometastases were found after immunohistochemical re-assessment in 4 out of 33 N0 patients (12.1%). Interdependences between revised N-stage and main prognostic variables are summarized in Table 1
. Apart from the kind of adjuvant therapy, which was obviously influenced by the N-status and marginally by tumor vessel invasion, no other variables resulted to be significantly correlated. The presence of nodal micrometastases correlated with a higher probability of distant relapse (p
< 0.001) (Table 2
). Histology revealed squamous cell carcinoma in 29 cases (61.7%), adenocarcinoma in 15 (31.9%), and large cell carcinoma in 3 (6.4%). Mean tumor size was 4.8 ± 2.3 cm (range 2.3–12.6), with infiltration depth confined to the parietal pleura in 15 tumors (31.9%), spread to the surrounding soft tissue in 14 (29.8%), and extending to rib structures in 18 cases (38.3%). Immunohistochemical revision of the resection margins revealed microinfiltration in five patients (10.6%) and was significantly correlated with the development of local recurrence (p
< 0.005) (Table 2). Depth of tumor infiltration into the chest wall was also re-evaluated by immunohistochemistry: 5 cases judged by conventional staining to be limited to parietal pleura showed soft-tissue invasion, thus increasing the number of this subset of patients to 19 (40.4%). Nonetheless, these series were not significantly correlated with microinfiltration of the resection margins.
|
|
Eleven patients (23.4%) developed local recurrence and 14 (29.8%) developed distant metastasis. Twenty patients died of cancer: 8 after local recurrence and 12 after distant relapse. Three patients died for non-neoplastic reasons: myocardial infarction (n = 2) and traffic accident (n = 1). Eighteen patients are currently alive and disease-free, 14 with a follow-up longer than 5 years and 4 longer than 10 years. Five-year disease-free and overall survivals were 45.7% and 50.9%, respectively.
Age (
65 years), tumor maximum diameter (<5 cm), histopathology (squamous vs nonsquamous), preoperative forced expiratory volume in 1 s (2.00 l), type of resection (atypical vs lobectomy vs pneumonectomy), number of ribs (
2), depth of invasion (pleural vs soft tissue vs rib) even after immunohistochemistry review, tumor vessel invasion, and the use of postoperative adjuvant therapy (none vs chemo vs radio) did not significantly affect survival. Instead, survival was significantly influenced by N-status (p
< 0.001); in particular, N0 patients had a 5-year disease-free survival rate of 52% and an overall survival of 65%, whereas in N1 patients these rates were 38% and 42%, respectively. Survival was worsened by the presence of micrometastasis (p
< 0.0001) (Fig. 1
) and microinfiltration of the resection margins (p
< 0.0001) (Fig. 2
). Five-year disease-free and overall survivals in patients with no evidence of residual disease after immunohistochemical re-evaluation increased up to 67% and 73%, respectively. In order to eliminate the impact of nodal micrometastases, survival evaluation was also conducted on the 29 patients restaged as pN0 after immunohistochemical re-evaluation (Fig. 2); apart from resection margin microinfiltration, no other factor resulted significantly (data not shown).
|
|
|
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
In this study, nodal micrometastases were found in 12.1% N0 patients, and correlated with distant relapse. As a result overall and disease-free survivals were significantly influenced by resection margins microinfiltration and N-status, especially after immunohistochemical assessment of micrometastases.
The presence of mediastinal nodes metastases traditionally identified patients with the most unfavorable prognosis [6,7,14,18–21]. Tumors invading the chest wall without evidence of hilar or mediastinal nodes represent locally invasive but potentially resectable lesions in which surgical resection can offer a real chance for long-term survival and even cure [14,21]. In our series, 5-year survival rate of radically resected patients without evidence of lymph nodes involvement was 73%, a remarkably greater value than that observed in stage IIB patients.
Our study results confirm the relevance of the lymph node status in determining outcome and extended these findings showing that lymph node micrometastases detected by means of immunohistochemical analysis represented one of the most important prognostic factors affecting survival.
Keratin-immunoreactive tumor cells have been detected in up to 63% of intrathoracic lymph nodes from NSCLC patients [22,23]. Cote et al. [23] indicated that the detection of occult regional and systemic metastases is an important predictor of disease progression, which is usually associated with a less favorable prognosis. In a more recent study of 119 patients with T1N0M0 adenocarcinoma and bronchioloalveolar carcinoma of the lung, Goldstein et al. [24] reported ‘micrometastases’ detectable by keratin immunohistochemistry in up to 8% of N0 patients.
As far as the issue of the resection margins is concerned, Allen et al. [17] and Albertucci et al. [18] recommend full-thickness chest wall resection, even in case of tumor adhering to the parietal pleura. Conversely, McCaughan et al. [3] consider parietal pleura as an anatomic barrier to tumor invasion and are confident that extrapleural dissection may achieve adequate oncological clearance with lower operative mortality and morbidity rates. We have found microinfiltration of the resection margins to be directly correlated with a greater risk of local recurrence. In fact, the presence of microinfiltration of the resection margins is likely to be considered a significant negative predictor of long-term survival. On the contrary, neither standard histopathologic study nor the immunohistochemical re-assessment of infiltration depth provided a significant impact on survival. According to this rationale, we believe, in agreement with other authors, that en-bloc resection is to be preferred to assure complete removal of the primary tumor even in peripheral tumors adhering to the parietal pleura only. Yet, as we had previously shown in patients with poor pulmonary function [25], we continue to consider en-bloc resection a not invalidating procedure. Immunohistochemical analysis may thus help identify patients at higher risk of recurrence, who thus would benefit of more aggressive adjuvant therapy regimens and a stricter follow-up. Conversely, in our study, age, sex, FEV1, type of surgical resection, histology, tumor size (more or less than 5 cm), number of ribs, depth of chest wall invasion, and vessel invasion resulted less relevant in comparison with other studies [14,15,17,18,21], a finding which should probably be attributed to the relatively small sample size.
The main limitations of our study rely on its retrospective nature and the limited sample size. However, it must be noticed that immunohistochemistry re-assessment of all resected specimens has been performed during a limited period of time through the same technique and by the same pathologist. Moreover, although the study covers a 13-year period, the surgical procedure was performed according to the same criteria and by the same surgeon.
In conclusion, immunohistochemistry-detected microscopic tumor spread can occur following en-bloc resection for T3-chest wall NSCLC and can affect long-term outcome in some patients’ subgroups who might warrant more aggressive adjuvant treatments. Larger prospective studies are needed to add further insights to our preliminary findings.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. DiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. C. Mineo, V. Ambrogi, E. Pompeo, A. Baldi, F. Stella, P. Aurea, and M. Marino The value of occult disease in resection margin and lymph node after extrapleural pneumonectomy for malignant mesothelioma. Ann. Thorac. Surg., May 1, 2008; 85(5): 1740 - 1746. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
