| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2000;18:301-306
© 2000 Elsevier Science NL
a Department of Thoracic Surgery, Royal Brompton Hospital, London, UK
b Department of Histopathology, Royal Brompton Hospital, London, UK
Received 25 January 2000; received in revised form 25 May 2000; accepted 21 June 2000.
Corresponding author. Consultant Thoracic Surgeon, Clinical Director, Royal Brompton Hospital, Sydney St, London SW3 6NP, UK. Tel.: +44-171-351-8558; fax: +44-171-351-8560
e-mail: p.goldstraw{at}rbh.nthames.nhs.uk
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionAppendix A Conference discussionReferences |
|---|
Key Words: Typical carcinoid • Atypical carcinoid • Neuroendocrine tumours • Bronchial adenoma • Bronchoplastic procedures
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionAppendix A Conference discussionReferences |
|---|
Carcinoids are now considered part of the spectrum of neuroendocrine tumours, which include typical carcinoid, atypical carcinoid, large cell neuroendocrine carcinoma and small cell carcinoma. However, in the recent WHO/IASLC classification of lung tumours typical and atypical carcinoids are grouped together, whilst small cell carcinoma (SCLC) is a distinct group and large cell neuroendocrine carcinoma (LCNEC) remains as a subset of large cell carcinoma [12]. This reflects the fact that typical and atypical carcinoid tumours are more closely related clinically than the other neuroendocrine tumours and share similar biological markers [13].
Both typical and atypical carcinoids are considered malignant tumours of low to intermediate grade but, because of their indolent growth pattern and localization at presentation, surgeons have increasingly followed a policy of conservative resection.
|
2. Materials and methods |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionAppendix A Conference discussionReferences |
|---|
During a period of 20 years 110 patients with pulmonary carcinoid were treated at the Royal Brompton and Middlesex Hospitals, London, by a single surgeon (P.G.).
All patients had a chest radiograph and chest computed tomography (CT). Twenty-four-hour urine was collected for 5-hydroxyindole acetic acid (5HIAA) in 16 patients. Fiberoptic bronchoscopy was performed prior to referral in 29 patients. Prior to surgery all patients underwent rigid bronchoscopy to determine the localization and size of the tumour. Tumours visualised by either bronchoscopic technique were considered as central tumours. Details of the size and site of the tumour were determined by bronchoscopy, CT, the findings at surgery and the macroscopic and microscopic features at pathological examination. Pleural lavage cytology was performed immediately after thoracotomy in 21 patients.
All resection specimens, including hilar and mediastinal lymph nodes were sent for histological examination. For this study all of the slides were reviewed by a pathologist with a special interest in lung tumours (A.N.) and classified according to the current WHO/IASLC criteria for neuroendocrine tumours. Typical carcinoids are defined as tumours of greater than 5 mm diameter with carcinoid morphology and less than two mitoses per 2 mm2 and lacking necrosis. Tumours with a mitosis rate of two or more but less than 10 per 2 mm2, with or without necrosis, were classified as atypical [12,14]. Ninety-five patients fulfilled these criteria and are the subject of this report. The other 15 cases were reclassified as LCNEC and excluded from the study. All 15 had been treated before 1991 when LCNEC was first described [12].
Patients were followed up for periods ranging from 6 months to 12 years (mean 3.9 years).
We analyzed the influence of site of tumour, carcinoid morphology, lymph node involvement and the extent of resection on the outcome of surgery.
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionAppendix A Conference discussionReferences |
|---|
Symptoms were present in 62 patients (65.3%): haemoptysis (n=26), cough (n=25), recurrent chest infection (n=20), wheeze (n=13) chest discomfort and chest pain (n=12), dyspnoea (n=9), and fever (n=9). The length of symptoms prior to presentation varied from 6 months to 6 years. Sixteen patients (16.8%) had been treated as asthmatic for up to 4 years prior to the discovery of the tumour. Thirty-three patients (34.7%) were asymptomatic at presentation.
3.2. Investigations
All patients underwent chest radiographs, chest CT and rigid bronchoscopy. The tumour could be visualized at bronchoscopy in 58 patients (61.1%), and these were classified as central tumours. The other 37 tumours (38.9%) were classified as peripheral. In the early part of the study 24-h urine collection was performed for 5-HIAA estimation. In 15 of 16 estimations it was normal, being only slightly elevated in one patient. The use of this test was thereafter discontinued.
3.3. Localization and size of tumour
The tumour was in the right lung in 56 cases (61%) and in the left lung in 36 cases. Three cases had multiple nodules; in the right lung in one case, the left lung in one and in both lungs in one case.
Of the 37 tumours designated as peripheral, seven were in the right upper lobe, five in the middle lobe, nine in the right lower lobe, nine in the left upper lobe (of which six were in the lingular segments) and seven in the left lower lobe. The 55 central tumours had arisen in the right main bronchus in six cases, the upper lobe bronchus in 11 cases (seven on the right and four on the left), the intermediate bronchus in nine cases, the middle lobe bronchus in seven cases, the lingular bronchus in six cases and in the lower lobe bronchus in 19 cases (11 on the right and eight on the left).
The size of the tumour varied from 0.5–9.3 cm (mean 3.4 cm). Although the largest tumour was of typical type, atypical tumours were on average larger than typical tumours (mean 4.1 vs. 3.0 cm, range 3.1–7.8 vs. 0.5–9.3 cm).
3.4. Pathology
Eighty-one tumours were considered to be typical carcinoids and 14 were classed as atypical tumours. Fifteen other cases, initially thought to be carcinoid tumours, were reclassified as LCNEC and were excluded from further study.
Of the 81 patients with a final diagnosis of typical carcinoid, 29 had had bronchoscopic biopsies prior to referral. Of these, 16 were considered by the local pathologist to show the features of a typical carcinoid, one was considered an atypical carcinoid, six were thought to show non-small cell carcinoma (NSCLC) and six were not diagnostic. Two further patients had had fine-needle aspiration biopsies. In both the diagnosis was considered to be carcinoid, one typical and one atypical. Prior to surgery we undertook rigid bronchoscopy and biopsy in 38 of these patients. Twenty-nine were correctly diagnosed at that time as being typical carcinoid tumours, seven biopsies were not diagnostic, one case was diagnosed as a plasma cell granuloma and one case was considered to show a well-differentiated adenocarcinoma. Biopsies were taken at operation for frozen section in 28 of these patients, 15 were diagnosed as typical carcinoid, one a plasma cell granuloma, one a sclerosing haemangioma, one as an epithelial tumour not otherwise specified, one as small cell carcinoma, one as NSCLC, five as adenocarcinoma, and in three no diagnosis was possible. All cases were subsequently revised to a diagnosis of typical carcinoid after review of the resection specimens. In retrospect, the principal reasons for the incorrect diagnosis were misinterpretation of cellular pleomorphism as a malignant feature or the presence of glandular architecture causing the neuroendocrine nature of the tumour to be overlooked.
Of 14 patients finally classified as atypical carcinoid, bronchoscopic biopsy prior to referral was available in nine patients, and was interpreted by the local pathologist as showing typical carcinoid in four patients, atypical in two, adenocarcinoma in one, calcification alone in one, and no diagnosis in one patient. Two patients underwent fine-needle aspiration biopsy prior to referral, one interpreted as showing typical carcinoid and one atypical. A further five patients underwent bronchoscopic biopsy after referral, of which one was considered atypical carcinoid, three were considered typical and one was considered to show adenocarcinoma. At operation biopsies were taken for frozen section in three patients, of which two were diagnosed as typical carcinoid, and one was thought to be NSCLC.
Of the 81 tumours classified as typical carcinoids, lymph nodes were positive in 11, all in an N1 position. In three the nodal involvement was considered to be due to direct invasion. In the 14 patients with atypical carcinoid tumours four had nodal deposits, two in N1 nodes and two with N2 disease.
3.5. Surgical procedures
Ninety-two patients underwent 96 operations. Surgery was not performed on three patients; one refused surgery; one had multiple, bilateral nodules, up to 15 in each lung and one had serious cardiac pathology. Lobectomy or bilobectomy was performed in 57 patients, lobectomy and sleeve resection in 23 and bronchial sleeve resection alone in one case. Pneumonectomy was performed in three patients. In two patients pneumonectomy was necessary as the situation was complicated by severe damage to the obstructed lung. In the other patient the intraoperative biopsies were interpreted on frozen section as lung cancer. One more patient underwent completion pneumonectomy for recurrence having undergone lobectomy for a node negative, typical carcinoid tumour 18 years earlier. Three patients with typical carcinoid tumours were treated by endobronchial resection. In two of these cases there was recurrence after 2 and 3 months requiring lobectomy. The third patient is disease free at 4 years. Three patients underwent segmentectomy and four underwent wedge resection or the local excision of pulmonary nodules. One patient having undergone wedge excision for a typical carcinoid developed a recurrence at 6 months requiring completion lobectomy. One patient underwent excision of multiple pulmonary nodules. Lymph node sampling was performed in all cases, but systematic nodal dissection was not performed [15]. There were no deaths related to surgery.
Table 1 provides a breakdown of the operative procedures according to nodal status and cell type.
|
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionAppendix A Conference discussionReferences |
|---|
They affect both sexes with almost equal incidence [7]. The mean age of our patients at presentation was 51 years, which is in line with other reports in which the average age varies between 16 and 72 years [8]. In our study patients having atypical carcinoid tumours were 6 years older at presentation when compared to those with typical tumours, a finding not previously noted in the literature. Patients may present with a variety of pulmonary symptoms or their tumours may be found on a routine chest radiograph [1,8]. It should be of concern to all chest physicians that 17% of our patients had been treated as asthmatic for up to 4 years prior to the discovery of the underlying tumour [1]. The influence of smoking on the pathogenesis of carcinoid tumours is still debated [6]. Two thirds of our patients were smokers or ex-smokers, as others have shown [11].
We found CT to be the most useful investigative procedure, showing tumour in all cases. The tumour was only visible on the chest radiograph in 68.4% cases. In two thirds of our patients the tumour was central and visible at bronchoscopy, in line with other studies [5,21]. We have not experienced troublesome bleeding following bronchoscopic biopsy, although others [5] found bleeding to occur in two thirds of their patients, and some authors advise against biopsy when carcinoid is suspected for fear of serious bleeding [10,17].
Our study emphasizes the difficulty in obtaining an accurate diagnosis in carcinoid tumours. Of 38 patients who had had bronchoscopic biopsy prior to referral a diagnosis of carcinoid was suggested in only 61% of cases and in only 47% was the correct type of carcinoid suggested. More worrying in 18% of cases we were told that the diagnosis was NSCLC. Our own rigid bronchoscopic biopsies were more accurate, no doubt due to the larger samples obtained. Carcinoid was diagnosed in 77% of biopsies and the correct type was specified in 70% of cases. Even so in 5% an incorrect diagnosis of NSCLC was made. Intraoperative frozen section biopsy of peripheral tumours was open to similar errors with 55% suggesting carcinoid, the correct identification of type in 48%, but 26% of such biopsies were interpreted as NSCLC. It is not too surprising that pathologists should find difficulty in differentiating typical from atypical carcinoid tumours as such features as increased mitotic rate and necrosis may not be present in a biopsy sample. More worrying is the misdiagnosis of typical and atypical carcinoid as carcinoma. In one of our patients pneumonectomy was performed after frozen section biopsies had been reported as NSCLC. Perhaps had we known the true histology of a typical carcinoid tumour we might have been able to perform a more conservative resection. It is important that pathologists examining frozen section specimens do not interpret cellular pleomorphism as a feature of malignancy or glandular architecture as an indicator of adenocarcinoma. If neither mitoses nor necrosis are seen in such pleomorphic tumours, a diagnosis of carcinoid tumour must be considered. The surgeon must be aware of the clinical and radiographic features when interpreting the frozen section report and bear in mind the possibility of such tumours being of carcinoid type. Others have previously found that frozen section examination of bronchoscopic biopsies can prove misleading [7]. The correct diagnosis is often only made when an experienced, specialist pathologist has examined the specimen in detail using electron microscopy and immunohistochemistry [16,20,21].
We agree with others [6] that mediastinoscopy is of little value in the management of carcinoid tumours. In the seven cases in which we performed mediastinoscopy it was negative, a finding confirmed at thoracotomy. In addition, in carcinoid tumours there appears to be no prognostic significance attached to the presence of nodal disease and hence no purpose to the detection of such disease. However as many such cases are referred as lung cancer, and as the CT features may suggest mediastinal involvement mediastinoscopy may reassure the surgeon as to operability.
There is continued debate as to the prognostic value of pleural lavage cytology in NSCLC [22]. Pleural lavage cytology was negative in all of the 21 cases in which it was performed, and none of these patients subsequently relapsed. Thus we did not find pleural lavage to be of prognostic value in carcinoid tumours.
The policy for surgical resection of carcinoid tumour is more controversial [17], being particularly related to whether these tumours, especially typical carcinoid, should be regarded as benign or malignant [8]. This has been complicated by the presence of differing classification system for neuroendocrine tumours, with some authors arguing to abandon the terms typical and atypical, preferring instead to consider such tumours as varying grades of neuroendocrine carcinoma [12]. Despite the recognition that these tumours are indeed malignant, their indolent growth [6,11] has encouraged surgeons to accept a policy of conservative resection for both typical and atypical carcinoids. Whilst some authors argue that carcinoid tumours showing atypical features should be treated by wider excision, even pneumonectomy [3,10], most have recommended conservative resection whenever this does not preclude complete resection. Never the less a proportion of cases undergo pneumonectomy, and this has been between a third and a half of the patients in some series [6,7,11,17,21,23]. In some cases pneumonectomy was clearly necessary as the distal lung parenchyma was destroyed due to long-standing obstruction, but this only accounts for half of the cases undergoing pneumonectomy in series in which this factor is identified [23].
In our series the presence of atypical histological features did not necessitate more extensive resection and the presence of nodal disease did not preclude lobectomy or sleeve resection. Central tumours involving the main bronchus and descending bronchus could be resected by sleeve resection, with or without lobectomy. The only cases correctly diagnosed that required pneumonectomy were associated with destruction of the distal lung parenchyma.
We believe that our results validate a policy of conservative resection in all cases in which functioning lung is present. We had one patient who experienced a recurrence after wedge resection. Although this case was salvaged by further resection and is now alive and disease-free 6 years after second line surgery we would be hesitant to recommend such a limited resection for carcinoid tumours.
Some authors have gone further and recommend endoscopic resection of central carcinoid tumours [20,24]. In our experience, there was recurrence in two of the three cases in which endoscopic treatment was performed, requiring lobectomy. We cannot recommend endoscopic treatment but feel that it may be of value when patients are unfit for surgery or as a temporary measure to relieve symptoms, to prepare patients for surgery and to assess the reversibility of distal lung damage in borderline cases [25].
Most authors [5,6,14,16,18,19] recommend nodal dissection. We found nodal involvement in only 16% of cases, although this incidence was higher in those tumours showing atypical features (28.5%) in whom nodal disease was in the mediastinum in half of the cases. Whilst one instinctively feels that the resection of mobile, involved lymph nodes is of therapeutic value, lymph node status has been shown to have any impact on either survival or recurrence rates [7,11]. In our series the recurrence-free survival after complete resection was 96.7%. Of the 15 tumours found to have nodal involvement, eight were less than 3.0 cm in diameter. Other authors have suggested that the probability of nodal involvement is related to the size of the tumour [6,7,11].
Local relapse has been associated with inadequate surgical resection, and has been salvaged in all cases by further, wider resection. It does not appear to be related to nodal status. Distant relapse is commoner in atypical cases but can occur rarely in typical cases.
Our policy has been to discontinue follow-up 5 years after the resection of carcinoid tumours, considering these patients to have been cured. In this review however we had one patient who was found to have local recurrence 18 years after surgery, and another in whom distant metastases were found 11 years following surgery. These cases argue for life-long surveillance of patients following resection of carcinoid tumours, and emphasize the need for extraordinarily long follow-up to establish the results of surgery. As the mean follow-up in this series is less than 4 years this report can best be considered as an interim analysis.
|
5. Conclusion |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionAppendix A Conference discussionReferences |
|---|
Our results show that typical and atypical carcinoids can be treated by conservative resection with good results, avoiding pneumonectomy where the distal lung is functional. However, in order for this policy to be successful, there must be close co-operation between surgeon and pathologist to avoid the potential pitfalls in the interpretation of preoperative and perioperative biopsies.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
Appendix A Conference discussion |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionAppendix A Conference discussionReferences |
|---|
Mr El Jemal: As we mentioned, we performed three bronchoscopic resections, two of them by laser, but we had two recurrences. This treatment can be used for temporary resolution of the condition of the patient and when preparing the patient for surgery. We cannot recommend it as an alternative treatment in place of surgery.
Dr A. Lerut (Leuven, Belgium): You performed eight segmental or wedge resections. And in your conclusion you state that lobectomy is the minimum. On what evidence is this statement based? I mean if you don't find it necessary to perform a pneumonectomy for positive nodes, on what basis do you state that minimal lobectomy is the minimum while you yourself did eight segmental or wedge resections?
Mr El Jemal: Based on our experience with eight minor resections, including excision of nodules or wedge resection, we experience one recurrence. So we cannot recommend it as treatment alternative. It can be, of course, done. But in our experience with this group of patients, we would not recommend it as a treatment alternative to lobectomy.
|
References |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. Discussion5. ConclusionAppendix A Conference discussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
