- Email: [email protected]
Large cell neuroendocrine carcinoma of the lung: pathological study and clinical outcome of 18 resected cases
Lung Cancer 37 (2002) 287 /292 www.elsevier.com/locate/lungcan
Large cell neuroendocrine carcinoma of the lung: pathological study and clinical outcome of 18 resected cases Julien Mazie`res a,*, Ghislaine Daste b, Laurent Molinier c, Jean Berjaud d, Marcel Dahan d, Martine Delsol b, Pierre Carles e, Alain Didier a, Jean-Marc Bachaud f a Service de Pneumologie, Hoˆpital Rangueil, 31403 Toulouse Cedex, France Laboratoire d’Anatomopathologie, Hoˆpital Purpan, 31059 Toulouse Cedex, France c Service d’Informatique Me´dicale, Institut Claudius Regaud, 31052 Toulouse Cedex, France d Service de Chirurgie Thoracique, Hoˆpital Purpan, 31059 Toulouse Cedex, France e Service de Me´decine Interne, Hoˆpital Purpan, 31059 Toulouse Cedex, France f Service d’Oncologie-Radiothe´rapie, Institut Claudius Regaud, 31052 Toulouse Cedex, France b
Received 12 December 2001; received in revised form 18 March 2002; accepted 22 March 2002
Abstract Large cell neuroendocrine carcinoma of the lung (LCNEC) has been recently redefined by the World Health Organisation (WHO) classification but the appropriate treatment remains unclear. We reviewed 18 consecutive resected cases of LCNEC. Two pathologists assessed diagnosis by applying rigorously the last WHO criteria. We reported the pathological features and the clinical outcome of this particular tumour. All patients were men with a median age of 63 years. Clinicopathologic stages corresponded to stage I (n /8), II (n /8) and IIIA (n /2). All patients were treated as non-small cell lung carcinoma (NSCLC) and underwent surgery without any adjuvant treatment except four post-operative radiotherapy for N2 or T3 disease. The evolution was pejorative for 14 patients: one patient died of post-operative complications and 13 patients relapsed with distant metastases that occurred in 10 cases within 6 months after surgery. One-year survival rate was 27% and survival rate at the end of follow-up was 22%, which were both less than expected for stage-comparable NSCLC. Survival was neither influenced by lymph node status nor by pathological or molecular findings. Among the 10 evaluable patients with metastatic disease that received palliative platin-etoposide chemotherapy only two had partial tumour regressions (20%). Our study suggests that applying to LCNEC the NSCLC standard treatment lead to poor prognosis even in localised disease with a high incidence of early metastatic spread and a low response rate to chemotherapy. This way of relapse underlies the necessity of an efficient chemotherapy in order to improve survival. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Lung carcinoma; Large cell neuroendocrine carcinoma; Neuroendocrine carcinoma; Surgery; Chemotherapy; Prognosis
1. Introduction Large cell neuroendocrine carcinomas of the lung (LCNEC) share morphological characteristics with classic large cell carcinomas (LC) and immunohisto-
* Corresponding author. Tel.: /33-5-61-32-2773; fax: /33-5-61-322957 E-mail addresses: [email protected] (J. Mazie`res), [email protected] (G. Daste), [email protected] (L. Molinier), [email protected] (J. Berjaud), [email protected] (M. Dahan), [email protected] (M. Delsol), [email protected] (P. Carles), [email protected] (A. Didier), [email protected] (J.-M. Bachaud).
chemical features with the wide spectrum of neuroendocrine (NE) tumours. Significant controversy exists over the classification of these carcinomas [1,2]. Recently, Travis et al. affirmed the usefulness of a three-grade, four-type scheme for classification of NE tumours of the lung with typical carcinoid (TC) representing a low grade tumour, atypical carcinoid (AC) representing an intermediate grade tumour, and LCNEC and small cell lung carcinoma (SCLC) representing high grade tumours [3]. The same grading is proposed in the 1999 World Health Organization (WHO) classification with rigorous criteria for each subtype even if LCNEC is still considered as a variant of LC [4].
0169-5002/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 9 - 5 0 0 2 ( 0 2 ) 0 0 0 9 9 - 5
288
J. Mazie`res et al. / Lung Cancer 37 (2002) 287 /292
LCNEC represents about 1/2% of lung malignancies and is known for having a very poor prognosis [3,5]. At the present time, it is not clear how these patients should be treated. Surgery is considered as the primary therapeutic option for TC and chemotherapy for SCLC, but the appropriate treatment remains very unclear for AC and LCNEC [6]. We reported 18 consecutive patients that underwent surgery in our institution between 1996 and 2000 for LCNEC. In order to obtain a homogenous series, we applied rigorously the WHO criteria and all the slides were submitted to a double reading. Until now, LCNEC was considered in our institution as non-small lung carcinoma (NSCLC), and all patients were treated by surgery without any adjuvant treatment except postoperative radiotherapy for N2 or T3 disease with chest involvement. Chemotherapy was only proposed in a palliative purpose when a metastatic disease was diagnosed. We reported the pathological findings and the clinical outcome of these 18 patients in order to better specify the morphological and molecular characteristics of this distinct subset of tumour and to evaluate the result of applying to LCNEC the NSCLC standard treatment.
2. Patients and methods Registries of the Department of Thoracic Surgery and of the Department of Pathology from our institution were reviewed to cull all cases of LCNEC resected between 1996 and 2000. During this period, 1118 patients underwent surgery for primary lung cancer in our institution including 46 (4%) NE tumours (6 SCLC, 9 TC, 5 AC and 26 LCNEC). The 46 NE tumours were reviewed by two pathologists. Hematoxylin /eosin stained slides, alcian blue sections and immunohistochemical stains for chromogranin A and synaptophysin were available in all cases for review. Only 18 diagnosis of LCNEC were confirmed by both pathologists according to the WHO 1999 criteria [4]: (1) a tumour with a NE morphology (organoid nesting, palisading, rosettes, trabeculae); (2) high mitotic rate: 11 or greater per 2 mm2; (3) necrosis (often large zones); (4) cytological features of a non-SCNEC: large cell size, low nuclear to cytoplasmic ratio, vesicular chromatin and frequent proeminent nucleoli, some tumours have fine nuclear chromatin and lack nucleoli, but qualify as non-SCNEC because of large cell size and abundant cytoplasm; (5) positive immunohistochemical staining for one or more NE markers including chromogranin A, synaptophysin and NCAM. Of the 26 initially so-called ‘LCNEC’, eight tumours were reclassified into combined LCNEC (squamous cell carcinoma or adenocarcinoma with a NE component, (n /3), in AC (n /1), in SCLC (n /1), in LC (n /2), in
basaloid carcinoma (n / 1). No NE tumour from other subtype was reclassified in LCNEC. We thus retained 18 confirmed LCNEC (1.6% of resected tumours during this period). For each case, we specified the mitotic activity and morphological features such as nuclear details, frequency of rosettes structures and abundance of necrosis. Mitoses were counted at a high power field (HPF) of 40 / with a field-of-view number of 20 (2 mm2) in the areas of highest mitotic activity. Necrosis consisted of either punctate foci or large infarctlike zones. Rosettes were counted at a HPF of 40 / with an average on 10 fields. Immunohistochemical stains were performed using paraffin embedded tissue and the streptavidinbiotin-peroxidase method in a Techmate 500-220 automated immunostainer. We tested the following antibodies directed against NE markers: Chromogranin A (DAK A3, 1:300, DAKO), Synaptophysin (SY38, 1:10, DAKO), NCAM (123C3, 1:50, DAKO) and against P53 (D0-7, 1:50, DAKO) and Bcl-2 (124, 1:50, DAKO). At least 20% of positive cells were recorded to quote P53 or Bcl-2 positivity. The percentage of positive nuclei to quote P53 positivity was of 20%. We thus retained 18 LCNEC that fulfilled totally the WHO criteria and exhibited both morphologic and immunohistochemical features in order to constitute a homogenous series. The pre-operative assessment included for each patient, computed tomography scanner of the lungs, adrenal glands and brain, liver ultrasonography, bone scan and bronchial endoscopy. The pathological status of the disease was assessed according to the International Staging System [7]. We reviewed the therapeutic management including surgical procedure and postoperative treatment and the clinical outcome including global survival and time to progression for each patient. Survival from the time of surgery was calculated according to the Kaplan/Meier method [8]. Survival curves according to nodal status (pN0 vs. pN1/pN2) were also obtained.
3. Results 3.1. Pathological characteristics (Table 1) By definition, all selected tumours had a NE appearance with organoid, trabecular or rosette-like patterns. The frequency of rosette-structures was variable (from less than 1 to 5 for 10 HPF). The mitotic index was high from 35 to 71 and a median of 52. In all patients, histologic NE differentiation was confirmed by the presence of positive immunohistochemical staining for chromogranine A (18 cases) or synaptophysin (15 cases). NCAM was expressed in 11 cases. Bcl-2 was positive in nine cases and p53 was over-expressed in 15 cases, often with intensity.
J. Mazie`res et al. / Lung Cancer 37 (2002) 287 /292
289
Table 1 Histological and immunohistochemical characteristics N
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Diagnosis
Histology
Immunohistochemistry
Biopsy
Surgery
M.I.
Necrosis (%)
Rosettes
Chrom
Synap
NCAM
p53 (%)
60 44 71 32 44 45 48 64 35 44 100 90 56 72 80 34 20 77
30 60 50 20 70 50 10 30 50 70 10 80 50 50 30 50 20 40
1 /3 B1 B1 B1 B1 B1 B1 B1 B1 1 /3 B1 B1 3 /5 B1 3 /5 B1 B1 3 /5
(95) (5) (80) (5) (90) (35) (50) (90) (70) (35) (15) (90) (80) (95) (90) (75) (50) (80)
Bcl-2 (%)
(70) (0) (60) (30) (5) (40) (0) (30) (0) (5) (0) (0) (50) (5) (5) (40) (20) (100)
M.I., mitotic index; Chrom, chromogranin A; Synap, synaptophysine; NCAM, neural cell adhesion molecule. Histological and immunohistological procedures and counting are detailed in Section 2.
3.2. Clinical and therapeutic characteristics (Table 2) Of the 18 patients studied, all were men. Median age was 63 with a range from 49 to 78 years. All but one had a smoking history. Three patients (17%) had a central tumour with positive bronchoscopy in two cases. The 15 other patients (83%) had a peripheral mass with negative bronchoscopy. Overall, the diagnosis of LCNEC was
established pre-operatively in only three cases (two bronchoscopy and one fine needle biopsy). In four cases, the pre-operative biopsy concluded in favour of unclassified carcinoma and in six cases, the biopsy was negative. In all cases, a complete surgical proceeding including a mediastinal lymphadenectomy was performed either by lobectomy (n /14) or by pneumonectomy (n / 4). Post-operative radiotherapy (50 /54 grays
Table 2 Clinical characteristics and outcome Age Stage (TNM)
Surgery (months)
TTP (months)
Survival or last follow-up (months)
Site of relapse
CHT (courses)
Response
69 59 54 43 64 69 51 45 70 66 63 52 68 49 78 54 67 72
L (LLL) L (RUL) L (LLL) P (R) L (RLL) L (RUL) L (RUL) P (R) L (RUL) L (RLL) L (RLL) P (R) L (ML) L (RUL) L (LLL) P (R) L (LUL) L (LUL)
5 11 / 3 / 1 3 6 4 1 2 / 3 1 3 5 8
Alive 11 20 Alive 10 Alive 4 10 9 8 11 7 Alive 8 1 10 10 12
/ Brain Liver / Bone / Brain, bone Brain, bone Bone Liver Bone Brain, bone / Liver / Brain Brain Brain
/ C/V C/V / C/V / C/V C/V C/V C/V C/V C/V / C/V / C/V C/V C/V
/ PD PR / NE / PD PD PD NE NE PD / PD / PR PD PD
T2N1M0 T3N0M0 T2N0M0 T2N0M0 T2N0M0 T2N1M0 T2N1M0 T3N2M0 T2N0M0 T2N0M0 T3N2M0 T2N1M0 T2N0M0 T2N0M0 T2N0M0 T3N0M0 T2N1M0 T2N1M0
(42)
(30) (22)
(12)
(4) (6) (4) (2) (4) (2) (3) (3) (4) (3) (4) (2) (1)
L, lobectomy; LLL, left lower lobe; LUL, left upper lobe; RLL, right lower lobe; ML, middle lobe; RUL, right upper lobe; P, pneumonectomy, R, right; FU, follow-up; TTP, time to progression; S, survival; CHT, chemotherapy; C, cis or carboplatine; V, vepeside; NE, non-evaluable; PR, partial response; PD, progressive disease.
J. Mazie`res et al. / Lung Cancer 37 (2002) 287 /292
290
in two grays daily fractions) was delivered using a 10 MeV linear accelerator on the mediastinum and/or the chest wall in four cases of tumours classified T3 and/or N2. None received adjuvant post-operative chemotherapy. Clinicopathologic stages corresponded to localised neoplasm in most of the cases (stage I [T1N0, T2N0], n /8; stage II [T1N1, T2N1, T3N0], n /8; stage IIIA [T1N2, T2N2, T3N2, T3N1], n /2). 3.3. Clinical outcome (Table 2) Follow-up was an average of 24 months. No patient was lost to follow-up. The evolution was pejorative for 14 patients (78%). A patient died of acute respiratory failure 45 days after surgery (5% perioperative mortality) and 13 patients relapsed with distant metastases (brain /7, bone /6, liver /4, adrenal glands /2). The metastatic spread was detected in 10 cases within 6 months after surgery. No loco-regional recurrence was noted. One-year survival was 27% and at the end of the follow-up, only four patients were alive without evidence of disease (22%). Survival was not significantly influenced by nodal status (P /0.9) (Figs. 1 and 2). Palliative chemotherapy was proposed for all patients in whom metastatic relapse was diagnosed. Chemotherapy regimen was either cisplatin (80 /100 mg/m2 per day, day 1) or carboplatin (Area Under the Curve from 4 to 5, day 1) every 4 weeks associated with etoposide (100/m2 per day, day 1 /3). They received an average of 3.2 courses (from 1 to 6). Among these 13 patients, three were not evaluable, eight did not show any objective tumour response and only two did show partial response (20%).
4. Discussion Multiplication of several classifications and of alternative terms has led to considerable confusion in the diagnosis and treatment of NE tumours. LCNEC must
Fig. 1. Kaplan /Meier survival curves for all patients.
Fig. 2. Kaplan /Meier survival curves related to node status (N0: 10 patients vs. N1 and N2: 8 patients).
be strictly separated from other NE tumours such as SCLC and AC and also from NSCLC with NE differentiation. The distinction between AC and LCNEC is based on Travis’s works [3] that had identified a better correlation with the clinical prognosis by considering the threshold mitotic rate at 10 per 2 mm2. In fact, LCNEC and SCLC have very high mitotic rates and also more extensive necrosis than AC [4]. Consistent with this, we report in our series a median mitotic index of 52 per 2 mm2 and the presence of necrosis in all slides. Furthermore, LCNEC cells more closely resemble those of a large cell carcinoma than a carcinoid tumour. In the NE tumours, SCLC and LCNEC have been kept separate because they are both high-grade tumours. We confirm these data with on one hand the high mitotic rate and on the other hand the high incidence of p53 and Bcl-2 abnormalities (respectively in 83 and 50% of the tumours). A high frequency of p53 mutations has already been described [9]. According to our series and to the literature, it appears that more than a variant of LC, LCNEC should be considered as a major histological type with specific features. LCNEC should be distinguished not only from AC or SCLC but also from common NSCLC, which do not show NE morphology by light microscopy but demonstrate immunohistochemical and/or ultrastructural evidence of NE differentiation [10]. These tumours should be classified according to the conventional typing herein, with NE differentiation noted [4]. Furthermore, the term ‘combined large cell neuroendocrine carcinoma’ should be used for LCNEC with components of adenocarcinoma, squamous cell carcinoma or giant cell carcinoma. Of the 26 initially called LCNEC, three tumours were reclassified in this subtype. These patients were eliminated from our study because we do not know yet if there is any difference in their biological and clinical behaviour. It will be of high interest to determine whether there are differences in survival or response to therapy among these various subtypes.
J. Mazie`res et al. / Lung Cancer 37 (2002) 287 /292
Most of the studies attempting to make a recommendation about the therapeutic option for LCNEC have used several classifications and have probably included some SCLC, AC and undifferentiated large cell carcinomas. In Dresler’s series a mitotic threshold at five is used to distinguish AC from LCNEC. This means that this study includes tumours with between 5 and 10 mitoses per 2 mm2 and fall below the lower limit proposed by the WHO classification [5]. The series of Jiang included as well four combined tumours of 22 with a squamous or a glandular component [10]. In the series of Carretta, AC and LCNEC were considered together as one group and in the Japanese series of Ishida, undifferentiated carcinomas were also retained [6,11]. These inclusions do not allow the authors to make meaningful conclusions about the clinical outcome and the therapeutic management. In order to avoid any confusion, we make a special effort with the pathologic analysis and immunohistochemical staining. Thus, all tumours fulfilled rigorously the predefined pathological criteria. Furthermore, as adjuvant chemotherapy was not considered as a standard treatment for LCNEC in our institution, no patient received post-operative chemotherapy and only four patients received adjuvant radiotherapy for N2 and/or chest wall involvement. Our study confirms that LCNEC have a dramatically poor prognosis even in the early stages of the disease. One-year survival rate of 27% is less than expected for stage-comparable NSCLC [7]. According to previous studies, the fatal outcome rate in LCNEC is between 57 and 100% [12 /15]. Using the new modified criteria, the 5 and 10-year survival rate for this sub-type is of 27 and 9%, respectively [3]. For Jiang and colleagues, the clinical outcome for LCNEC was significantly worse than for stage-comparable non-small-cell carcinomas [10]. For Wick and co-workers and, more recently, for Iyoda et al., LC with NE features appear to be more clinically agressive than classic LC [16,17]. For Dresler and associates, despite the tumour’s propensity to be present at low stage, the mortality was impressive after resection (survey at 5 years was of 13%) [5]. The problem of a surgical procedure in LCNEC is that even if its usefulness is unproven in the therapeutic management, it is often difficult to assess a rigorous diagnosis on biopsy specimen, therefore, the whole operative piece is needed to confirm the diagnosis. In our series, diagnosis of LCNEC was determined pre-operatively by bronchial endoscopy or transparietal fine needle biopsy in only three patients. This can be explained in part by the peripheral situation of most of the tumours (15 of 18 in our series) as already described [6,18], and by the difficulty in assessing a definitive diagnosis with immunhistochemical study on a small sample of tumour. Given our experience, we do think that cytology with immunohistochemistry on fine needle biopsy is not sufficient to accurately diagnose these tumours and
291
most of the time, surgery remains the only way to assess a rigorous diagnosis. Most of the patients had a metastatic relapse and none had a local recurrence. The fact that metastatic spread is detected a few months after surgery (77% within 6 months) proves that LCNEC is a very aggressive tumour which has a clinical behaviour closer to SCLC than to AC. In another series including 42 LCNEC or mixed carcinomas, the metastatic spread rate was of 57% after surgery [19]. This way of relapse, similar to SCLC, underlies the necessity of an effective chemotherapy. In our series, a combination of cisplatin and etoposide resulted in only 20% partial response. We can nevertheless speculate on a potential efficacy as neoadjuvant or adjuvant therapy. Even if this response rate is probably superior in an earlier stage, metastatic LCNEC seems to be less chemosensitive than SCLC for whom an objective response from 29 to 58% is usually reported [20]. Some authors have suggested that LCNEC are chemosensitive, similar to SCLC. At the opposite, Lai et al. suggest an insensitivity to chemotherapy due to the high frequency of expression of the MDR 1 gene in the sub-group of lung cancer with NE markers [21]. Very few studies discuss the usefulness of adjuvant chemotherapy. For Dresler, adjuvant chemotherapy does not improve survival but because it was a retrospective study, criteria for post-operative treatment were not reported and treatment protocols were not standardised [5]. Carretta et al. report only one patient who received chemotherapy and died as a result of disease recurrence 2 months after surgery [6]. For Moertel et al. a response rate of 67% can be obtained with the platin-etoposide regimen in 18 patients with anaplastic NE carcinoma but the primary site was lung in only three patients [22]. There is no reliable prognostic factor that can help us evaluate the risk of recurrence after surgery. Even the lymph node involvement does not seem not to be a prognostic factor in our series as it had been reported for AC and to the contrary of non-small cell carcinoma. Some genetic and molecular markers have been described as prognosis indicators and may help to identify high-risk cases in which adjuvant chemotherapy could be recommended [9]. Furthermore, mitotic rate or molecular abnormalities such as p53 or Bcl-2 do not seem to be correlated with clinical outcome as the four patients who were still alive at the end of the follow-up did not show different histological or molecular characteristics from other patients. Because of such a small population, a statistical analysis cannot be applied to show correlation. Even if LCNEC is a distinct subset of lung tumours, we do not know yet if we have to consider it as a SCLC with a two-stage classification (localised or diffuse) and a chemotherapy-based treatment or as a non-small cell carcinoma with a surgical treatment. The staging system
292
J. Mazie`res et al. / Lung Cancer 37 (2002) 287 /292
used in NSCLC seems to be inappropriate for LCNEC, as we could not distinguish the survival rates between the different stages. Our population is too small to allow definitive assertions but we can conclude that applying to LCNEC the NSCLC standard treatment leads to a poor prognosis. LCNEC is a particular subset of tumour that should not be treated with protocols designed for non-small cell carcinomas. In fact, surgery and adjuvant radiotherapy seems to be efficient in the loco-regional control of the disease as no loco-regional relapse was noted. The very high mortality rate is entirely due to metastatic relapse. This finding underlines the necessity of an adjuvant effective chemotherapy. The low response rate of the platin-etoposide combination suggests that new chemotherapy regimens should be evaluated in large-scale trials to assess the optimal treatment of LCNEC.
Acknowledgements The authors thank Bill Mazie`res and Maite´ Turmon for their help in reviewing the manuscript.
References [1] Gould VE, Linnoila RI, Memoli VA, Warren WH. Neuroendocrine cells and neuroendocrine neoplasms of the lung. Pathol Annu 1983;18:287 /330. [2] Travis WD, Linnoila RI, Tsokos MG, et al. Neuroendocrine tumours of the lung with proposed criteria for large-cell neuroendocrine carcinoma. An ultrastructural immunohistochemical, and flow cytometric, study of 35 cases. Am J Surg Pathol 1991;15:529 /53. [3] Travis WD, Rush W, Flieder DB, et al. Survival analysis of 200 pulmonary neuroendocrine tumours with clarification of criteria for atypical carcinoid and its separation from typical carcinoid. Am J Surg Pathol 1998;22:934 /44. [4] Travis WD, Colby TV, Corrin B, Shimosato Y, Brambilla E. Histological Typing of Lung and Pleural Tumors. In: Travis WD, Colby TV, Corrin B, Shimosato Y, Brambilla E, editors. World Health Organisation. International histological classification of tumours, 3rd ed.. Berlin, Heidelberg, New York: Springer, 1999. [5] Dresler CM, Ritter JH, Patterson GA, Ross E, Bailey MS, Wick MR. Clinical-pathologic analysis of 40 patients with large cell neuroendocrine carcinoma of the lung. Ann Thorac Surg 1997;63:180 /5.
[6] Carretta A, Ceresoli GL, Arrigoni G, et al. Diagnostic and therapeutic management of neuroendocrine lung tumours. A clinical study of 44 cases. Lung Cancer 2000;29:217 /25. [7] Mountain CF. Revisions in the International System for Staging Lung Cancer. Chest 1997;111:1710 /7. [8] Kaplan EM, Meier P. Non parametric estimations from incomplete observations. J Am Stat Assoc 1958;53:457 /81. [9] Brambilla E, Negoescu A, Gazzeri S, et al. Apoptosis-related factors p53, Bcl2, and Bax in neuroendocrine lung tumours. Am J Pathol 1996;149:1941 /52. [10] Jiang SX, Kameya T, Shoji M, Dobashi Y, Shinada J, Yoshimura H. Large cell neuroendocrine carcinoma of the lung: a histologic and immunohistochemical study of 22 cases. Am J Surg Pathol 1998;22:526 /37. [11] Ishida T, Kaneko S, Tateishi M, et al. Large cell carcinoma of the lung. Prognostic implications of histopathologic and immunohistochemical subtyping. Am J Clin Pathol 1990;93:176 /82. [12] Hammond ME, Sause WT. Large cell neuroendocrine tumours of the lung. Clinical significance and histopathologic definition. Cancer 1985;56:1624 /9. [13] Neal MH, Kosinski R, Cohen P, Orenstein JM. A typical endocrine tumours of the lung: a histologic, ultrastructural, and clinical study of 19 cases. Hum Pathol 1986;17:1264 /77. [14] Mooi WJ, Van Zandwijk N, Dingemans KP, Koolen MG, Wagenvoort CA. The ‘grey area’ between small cell and nonsmall cell lung carcinomas. Light and electron microscopy versus clinical data in 14 cases. J Pathol 1986;149:49 /54. [15] Piehl MR, Gould VE, Warren WH, et al. Immunohistochemical identification of exocrine and neuroendocrine subsets of large cell lung carcinomas. Pathol Res Pract 1988;183:675 /82. [16] Wick MR, Berg LC, Hertz MI. Large cell carcinoma of the lung with neuroendocrine differentiation. A comparison with large cell pulmonary tumours. Am J Clin Pathol 1992;97:796 /805. [17] Iyoda A, Hiroshima K, Toyozaki T, Haga Y, Fujisawa T, Ohwada H. Clinical characterization of pulmonary large cell neuroendocrine carcinoma and large cell carcinoma with neuroendocrine morphology. Cancer 2001;91:1992 /2000. [18] Jung KJ, Lee KS, Han J, et al. Large cell neuroendocrine carcinoma of the lung: clinical, CT, and pathologic findings in 11 patients. J Thorac Imaging 2001;16:156 /62. [19] Rusch VW, Klimstra DS, Venkatraman ES. Molecular markers help characterize neuroendocrine lung tumours. Ann Thorac Surg 1996;62:798 /809 (discussion 809-10). [20] Greco FA, Hainsworth JD. Practical approaches to the treatment of patients with extensive stage small cell lung cancer. Semin Oncol 1994;21:3 /6. [21] Lai SL, Goldstein LJ, Gottesman MM, et al. MDR1 gene expression in lung cancer. J Natl Cancer Inst 1989;81:1144 /50. [22] Moertel CG, Kvols LK, O’Connell MJ, Rubin J. Treatment of neuroendocrine carcinomas with combined etoposide and cisplatin. Evidence of major therapeutic activity in the anaplastic variants of these neoplasms. Cancer 1991;68:227 /32.
Large cell neuroendocrine carcinoma of the lung: pathological study and clinical outcome of 18 resected cases Julien Mazie`res a,*, Ghislaine Daste b, Laurent Molinier c, Jean Berjaud d, Marcel Dahan d, Martine Delsol b, Pierre Carles e, Alain Didier a, Jean-Marc Bachaud f a Service de Pneumologie, Hoˆpital Rangueil, 31403 Toulouse Cedex, France Laboratoire d’Anatomopathologie, Hoˆpital Purpan, 31059 Toulouse Cedex, France c Service d’Informatique Me´dicale, Institut Claudius Regaud, 31052 Toulouse Cedex, France d Service de Chirurgie Thoracique, Hoˆpital Purpan, 31059 Toulouse Cedex, France e Service de Me´decine Interne, Hoˆpital Purpan, 31059 Toulouse Cedex, France f Service d’Oncologie-Radiothe´rapie, Institut Claudius Regaud, 31052 Toulouse Cedex, France b
Received 12 December 2001; received in revised form 18 March 2002; accepted 22 March 2002
Abstract Large cell neuroendocrine carcinoma of the lung (LCNEC) has been recently redefined by the World Health Organisation (WHO) classification but the appropriate treatment remains unclear. We reviewed 18 consecutive resected cases of LCNEC. Two pathologists assessed diagnosis by applying rigorously the last WHO criteria. We reported the pathological features and the clinical outcome of this particular tumour. All patients were men with a median age of 63 years. Clinicopathologic stages corresponded to stage I (n /8), II (n /8) and IIIA (n /2). All patients were treated as non-small cell lung carcinoma (NSCLC) and underwent surgery without any adjuvant treatment except four post-operative radiotherapy for N2 or T3 disease. The evolution was pejorative for 14 patients: one patient died of post-operative complications and 13 patients relapsed with distant metastases that occurred in 10 cases within 6 months after surgery. One-year survival rate was 27% and survival rate at the end of follow-up was 22%, which were both less than expected for stage-comparable NSCLC. Survival was neither influenced by lymph node status nor by pathological or molecular findings. Among the 10 evaluable patients with metastatic disease that received palliative platin-etoposide chemotherapy only two had partial tumour regressions (20%). Our study suggests that applying to LCNEC the NSCLC standard treatment lead to poor prognosis even in localised disease with a high incidence of early metastatic spread and a low response rate to chemotherapy. This way of relapse underlies the necessity of an efficient chemotherapy in order to improve survival. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Lung carcinoma; Large cell neuroendocrine carcinoma; Neuroendocrine carcinoma; Surgery; Chemotherapy; Prognosis
1. Introduction Large cell neuroendocrine carcinomas of the lung (LCNEC) share morphological characteristics with classic large cell carcinomas (LC) and immunohisto-
* Corresponding author. Tel.: /33-5-61-32-2773; fax: /33-5-61-322957 E-mail addresses: [email protected] (J. Mazie`res), [email protected] (G. Daste), [email protected] (L. Molinier), [email protected] (J. Berjaud), [email protected] (M. Dahan), [email protected] (M. Delsol), [email protected] (P. Carles), [email protected] (A. Didier), [email protected] (J.-M. Bachaud).
chemical features with the wide spectrum of neuroendocrine (NE) tumours. Significant controversy exists over the classification of these carcinomas [1,2]. Recently, Travis et al. affirmed the usefulness of a three-grade, four-type scheme for classification of NE tumours of the lung with typical carcinoid (TC) representing a low grade tumour, atypical carcinoid (AC) representing an intermediate grade tumour, and LCNEC and small cell lung carcinoma (SCLC) representing high grade tumours [3]. The same grading is proposed in the 1999 World Health Organization (WHO) classification with rigorous criteria for each subtype even if LCNEC is still considered as a variant of LC [4].
0169-5002/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 9 - 5 0 0 2 ( 0 2 ) 0 0 0 9 9 - 5
288
J. Mazie`res et al. / Lung Cancer 37 (2002) 287 /292
LCNEC represents about 1/2% of lung malignancies and is known for having a very poor prognosis [3,5]. At the present time, it is not clear how these patients should be treated. Surgery is considered as the primary therapeutic option for TC and chemotherapy for SCLC, but the appropriate treatment remains very unclear for AC and LCNEC [6]. We reported 18 consecutive patients that underwent surgery in our institution between 1996 and 2000 for LCNEC. In order to obtain a homogenous series, we applied rigorously the WHO criteria and all the slides were submitted to a double reading. Until now, LCNEC was considered in our institution as non-small lung carcinoma (NSCLC), and all patients were treated by surgery without any adjuvant treatment except postoperative radiotherapy for N2 or T3 disease with chest involvement. Chemotherapy was only proposed in a palliative purpose when a metastatic disease was diagnosed. We reported the pathological findings and the clinical outcome of these 18 patients in order to better specify the morphological and molecular characteristics of this distinct subset of tumour and to evaluate the result of applying to LCNEC the NSCLC standard treatment.
2. Patients and methods Registries of the Department of Thoracic Surgery and of the Department of Pathology from our institution were reviewed to cull all cases of LCNEC resected between 1996 and 2000. During this period, 1118 patients underwent surgery for primary lung cancer in our institution including 46 (4%) NE tumours (6 SCLC, 9 TC, 5 AC and 26 LCNEC). The 46 NE tumours were reviewed by two pathologists. Hematoxylin /eosin stained slides, alcian blue sections and immunohistochemical stains for chromogranin A and synaptophysin were available in all cases for review. Only 18 diagnosis of LCNEC were confirmed by both pathologists according to the WHO 1999 criteria [4]: (1) a tumour with a NE morphology (organoid nesting, palisading, rosettes, trabeculae); (2) high mitotic rate: 11 or greater per 2 mm2; (3) necrosis (often large zones); (4) cytological features of a non-SCNEC: large cell size, low nuclear to cytoplasmic ratio, vesicular chromatin and frequent proeminent nucleoli, some tumours have fine nuclear chromatin and lack nucleoli, but qualify as non-SCNEC because of large cell size and abundant cytoplasm; (5) positive immunohistochemical staining for one or more NE markers including chromogranin A, synaptophysin and NCAM. Of the 26 initially so-called ‘LCNEC’, eight tumours were reclassified into combined LCNEC (squamous cell carcinoma or adenocarcinoma with a NE component, (n /3), in AC (n /1), in SCLC (n /1), in LC (n /2), in
basaloid carcinoma (n / 1). No NE tumour from other subtype was reclassified in LCNEC. We thus retained 18 confirmed LCNEC (1.6% of resected tumours during this period). For each case, we specified the mitotic activity and morphological features such as nuclear details, frequency of rosettes structures and abundance of necrosis. Mitoses were counted at a high power field (HPF) of 40 / with a field-of-view number of 20 (2 mm2) in the areas of highest mitotic activity. Necrosis consisted of either punctate foci or large infarctlike zones. Rosettes were counted at a HPF of 40 / with an average on 10 fields. Immunohistochemical stains were performed using paraffin embedded tissue and the streptavidinbiotin-peroxidase method in a Techmate 500-220 automated immunostainer. We tested the following antibodies directed against NE markers: Chromogranin A (DAK A3, 1:300, DAKO), Synaptophysin (SY38, 1:10, DAKO), NCAM (123C3, 1:50, DAKO) and against P53 (D0-7, 1:50, DAKO) and Bcl-2 (124, 1:50, DAKO). At least 20% of positive cells were recorded to quote P53 or Bcl-2 positivity. The percentage of positive nuclei to quote P53 positivity was of 20%. We thus retained 18 LCNEC that fulfilled totally the WHO criteria and exhibited both morphologic and immunohistochemical features in order to constitute a homogenous series. The pre-operative assessment included for each patient, computed tomography scanner of the lungs, adrenal glands and brain, liver ultrasonography, bone scan and bronchial endoscopy. The pathological status of the disease was assessed according to the International Staging System [7]. We reviewed the therapeutic management including surgical procedure and postoperative treatment and the clinical outcome including global survival and time to progression for each patient. Survival from the time of surgery was calculated according to the Kaplan/Meier method [8]. Survival curves according to nodal status (pN0 vs. pN1/pN2) were also obtained.
3. Results 3.1. Pathological characteristics (Table 1) By definition, all selected tumours had a NE appearance with organoid, trabecular or rosette-like patterns. The frequency of rosette-structures was variable (from less than 1 to 5 for 10 HPF). The mitotic index was high from 35 to 71 and a median of 52. In all patients, histologic NE differentiation was confirmed by the presence of positive immunohistochemical staining for chromogranine A (18 cases) or synaptophysin (15 cases). NCAM was expressed in 11 cases. Bcl-2 was positive in nine cases and p53 was over-expressed in 15 cases, often with intensity.
J. Mazie`res et al. / Lung Cancer 37 (2002) 287 /292
289
Table 1 Histological and immunohistochemical characteristics N
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Diagnosis
Histology
Immunohistochemistry
Biopsy
Surgery
M.I.
Necrosis (%)
Rosettes
Chrom
Synap
NCAM
p53 (%)
60 44 71 32 44 45 48 64 35 44 100 90 56 72 80 34 20 77
30 60 50 20 70 50 10 30 50 70 10 80 50 50 30 50 20 40
1 /3 B1 B1 B1 B1 B1 B1 B1 B1 1 /3 B1 B1 3 /5 B1 3 /5 B1 B1 3 /5
(95) (5) (80) (5) (90) (35) (50) (90) (70) (35) (15) (90) (80) (95) (90) (75) (50) (80)
Bcl-2 (%)
(70) (0) (60) (30) (5) (40) (0) (30) (0) (5) (0) (0) (50) (5) (5) (40) (20) (100)
M.I., mitotic index; Chrom, chromogranin A; Synap, synaptophysine; NCAM, neural cell adhesion molecule. Histological and immunohistological procedures and counting are detailed in Section 2.
3.2. Clinical and therapeutic characteristics (Table 2) Of the 18 patients studied, all were men. Median age was 63 with a range from 49 to 78 years. All but one had a smoking history. Three patients (17%) had a central tumour with positive bronchoscopy in two cases. The 15 other patients (83%) had a peripheral mass with negative bronchoscopy. Overall, the diagnosis of LCNEC was
established pre-operatively in only three cases (two bronchoscopy and one fine needle biopsy). In four cases, the pre-operative biopsy concluded in favour of unclassified carcinoma and in six cases, the biopsy was negative. In all cases, a complete surgical proceeding including a mediastinal lymphadenectomy was performed either by lobectomy (n /14) or by pneumonectomy (n / 4). Post-operative radiotherapy (50 /54 grays
Table 2 Clinical characteristics and outcome Age Stage (TNM)
Surgery (months)
TTP (months)
Survival or last follow-up (months)
Site of relapse
CHT (courses)
Response
69 59 54 43 64 69 51 45 70 66 63 52 68 49 78 54 67 72
L (LLL) L (RUL) L (LLL) P (R) L (RLL) L (RUL) L (RUL) P (R) L (RUL) L (RLL) L (RLL) P (R) L (ML) L (RUL) L (LLL) P (R) L (LUL) L (LUL)
5 11 / 3 / 1 3 6 4 1 2 / 3 1 3 5 8
Alive 11 20 Alive 10 Alive 4 10 9 8 11 7 Alive 8 1 10 10 12
/ Brain Liver / Bone / Brain, bone Brain, bone Bone Liver Bone Brain, bone / Liver / Brain Brain Brain
/ C/V C/V / C/V / C/V C/V C/V C/V C/V C/V / C/V / C/V C/V C/V
/ PD PR / NE / PD PD PD NE NE PD / PD / PR PD PD
T2N1M0 T3N0M0 T2N0M0 T2N0M0 T2N0M0 T2N1M0 T2N1M0 T3N2M0 T2N0M0 T2N0M0 T3N2M0 T2N1M0 T2N0M0 T2N0M0 T2N0M0 T3N0M0 T2N1M0 T2N1M0
(42)
(30) (22)
(12)
(4) (6) (4) (2) (4) (2) (3) (3) (4) (3) (4) (2) (1)
L, lobectomy; LLL, left lower lobe; LUL, left upper lobe; RLL, right lower lobe; ML, middle lobe; RUL, right upper lobe; P, pneumonectomy, R, right; FU, follow-up; TTP, time to progression; S, survival; CHT, chemotherapy; C, cis or carboplatine; V, vepeside; NE, non-evaluable; PR, partial response; PD, progressive disease.
J. Mazie`res et al. / Lung Cancer 37 (2002) 287 /292
290
in two grays daily fractions) was delivered using a 10 MeV linear accelerator on the mediastinum and/or the chest wall in four cases of tumours classified T3 and/or N2. None received adjuvant post-operative chemotherapy. Clinicopathologic stages corresponded to localised neoplasm in most of the cases (stage I [T1N0, T2N0], n /8; stage II [T1N1, T2N1, T3N0], n /8; stage IIIA [T1N2, T2N2, T3N2, T3N1], n /2). 3.3. Clinical outcome (Table 2) Follow-up was an average of 24 months. No patient was lost to follow-up. The evolution was pejorative for 14 patients (78%). A patient died of acute respiratory failure 45 days after surgery (5% perioperative mortality) and 13 patients relapsed with distant metastases (brain /7, bone /6, liver /4, adrenal glands /2). The metastatic spread was detected in 10 cases within 6 months after surgery. No loco-regional recurrence was noted. One-year survival was 27% and at the end of the follow-up, only four patients were alive without evidence of disease (22%). Survival was not significantly influenced by nodal status (P /0.9) (Figs. 1 and 2). Palliative chemotherapy was proposed for all patients in whom metastatic relapse was diagnosed. Chemotherapy regimen was either cisplatin (80 /100 mg/m2 per day, day 1) or carboplatin (Area Under the Curve from 4 to 5, day 1) every 4 weeks associated with etoposide (100/m2 per day, day 1 /3). They received an average of 3.2 courses (from 1 to 6). Among these 13 patients, three were not evaluable, eight did not show any objective tumour response and only two did show partial response (20%).
4. Discussion Multiplication of several classifications and of alternative terms has led to considerable confusion in the diagnosis and treatment of NE tumours. LCNEC must
Fig. 1. Kaplan /Meier survival curves for all patients.
Fig. 2. Kaplan /Meier survival curves related to node status (N0: 10 patients vs. N1 and N2: 8 patients).
be strictly separated from other NE tumours such as SCLC and AC and also from NSCLC with NE differentiation. The distinction between AC and LCNEC is based on Travis’s works [3] that had identified a better correlation with the clinical prognosis by considering the threshold mitotic rate at 10 per 2 mm2. In fact, LCNEC and SCLC have very high mitotic rates and also more extensive necrosis than AC [4]. Consistent with this, we report in our series a median mitotic index of 52 per 2 mm2 and the presence of necrosis in all slides. Furthermore, LCNEC cells more closely resemble those of a large cell carcinoma than a carcinoid tumour. In the NE tumours, SCLC and LCNEC have been kept separate because they are both high-grade tumours. We confirm these data with on one hand the high mitotic rate and on the other hand the high incidence of p53 and Bcl-2 abnormalities (respectively in 83 and 50% of the tumours). A high frequency of p53 mutations has already been described [9]. According to our series and to the literature, it appears that more than a variant of LC, LCNEC should be considered as a major histological type with specific features. LCNEC should be distinguished not only from AC or SCLC but also from common NSCLC, which do not show NE morphology by light microscopy but demonstrate immunohistochemical and/or ultrastructural evidence of NE differentiation [10]. These tumours should be classified according to the conventional typing herein, with NE differentiation noted [4]. Furthermore, the term ‘combined large cell neuroendocrine carcinoma’ should be used for LCNEC with components of adenocarcinoma, squamous cell carcinoma or giant cell carcinoma. Of the 26 initially called LCNEC, three tumours were reclassified in this subtype. These patients were eliminated from our study because we do not know yet if there is any difference in their biological and clinical behaviour. It will be of high interest to determine whether there are differences in survival or response to therapy among these various subtypes.
J. Mazie`res et al. / Lung Cancer 37 (2002) 287 /292
Most of the studies attempting to make a recommendation about the therapeutic option for LCNEC have used several classifications and have probably included some SCLC, AC and undifferentiated large cell carcinomas. In Dresler’s series a mitotic threshold at five is used to distinguish AC from LCNEC. This means that this study includes tumours with between 5 and 10 mitoses per 2 mm2 and fall below the lower limit proposed by the WHO classification [5]. The series of Jiang included as well four combined tumours of 22 with a squamous or a glandular component [10]. In the series of Carretta, AC and LCNEC were considered together as one group and in the Japanese series of Ishida, undifferentiated carcinomas were also retained [6,11]. These inclusions do not allow the authors to make meaningful conclusions about the clinical outcome and the therapeutic management. In order to avoid any confusion, we make a special effort with the pathologic analysis and immunohistochemical staining. Thus, all tumours fulfilled rigorously the predefined pathological criteria. Furthermore, as adjuvant chemotherapy was not considered as a standard treatment for LCNEC in our institution, no patient received post-operative chemotherapy and only four patients received adjuvant radiotherapy for N2 and/or chest wall involvement. Our study confirms that LCNEC have a dramatically poor prognosis even in the early stages of the disease. One-year survival rate of 27% is less than expected for stage-comparable NSCLC [7]. According to previous studies, the fatal outcome rate in LCNEC is between 57 and 100% [12 /15]. Using the new modified criteria, the 5 and 10-year survival rate for this sub-type is of 27 and 9%, respectively [3]. For Jiang and colleagues, the clinical outcome for LCNEC was significantly worse than for stage-comparable non-small-cell carcinomas [10]. For Wick and co-workers and, more recently, for Iyoda et al., LC with NE features appear to be more clinically agressive than classic LC [16,17]. For Dresler and associates, despite the tumour’s propensity to be present at low stage, the mortality was impressive after resection (survey at 5 years was of 13%) [5]. The problem of a surgical procedure in LCNEC is that even if its usefulness is unproven in the therapeutic management, it is often difficult to assess a rigorous diagnosis on biopsy specimen, therefore, the whole operative piece is needed to confirm the diagnosis. In our series, diagnosis of LCNEC was determined pre-operatively by bronchial endoscopy or transparietal fine needle biopsy in only three patients. This can be explained in part by the peripheral situation of most of the tumours (15 of 18 in our series) as already described [6,18], and by the difficulty in assessing a definitive diagnosis with immunhistochemical study on a small sample of tumour. Given our experience, we do think that cytology with immunohistochemistry on fine needle biopsy is not sufficient to accurately diagnose these tumours and
291
most of the time, surgery remains the only way to assess a rigorous diagnosis. Most of the patients had a metastatic relapse and none had a local recurrence. The fact that metastatic spread is detected a few months after surgery (77% within 6 months) proves that LCNEC is a very aggressive tumour which has a clinical behaviour closer to SCLC than to AC. In another series including 42 LCNEC or mixed carcinomas, the metastatic spread rate was of 57% after surgery [19]. This way of relapse, similar to SCLC, underlies the necessity of an effective chemotherapy. In our series, a combination of cisplatin and etoposide resulted in only 20% partial response. We can nevertheless speculate on a potential efficacy as neoadjuvant or adjuvant therapy. Even if this response rate is probably superior in an earlier stage, metastatic LCNEC seems to be less chemosensitive than SCLC for whom an objective response from 29 to 58% is usually reported [20]. Some authors have suggested that LCNEC are chemosensitive, similar to SCLC. At the opposite, Lai et al. suggest an insensitivity to chemotherapy due to the high frequency of expression of the MDR 1 gene in the sub-group of lung cancer with NE markers [21]. Very few studies discuss the usefulness of adjuvant chemotherapy. For Dresler, adjuvant chemotherapy does not improve survival but because it was a retrospective study, criteria for post-operative treatment were not reported and treatment protocols were not standardised [5]. Carretta et al. report only one patient who received chemotherapy and died as a result of disease recurrence 2 months after surgery [6]. For Moertel et al. a response rate of 67% can be obtained with the platin-etoposide regimen in 18 patients with anaplastic NE carcinoma but the primary site was lung in only three patients [22]. There is no reliable prognostic factor that can help us evaluate the risk of recurrence after surgery. Even the lymph node involvement does not seem not to be a prognostic factor in our series as it had been reported for AC and to the contrary of non-small cell carcinoma. Some genetic and molecular markers have been described as prognosis indicators and may help to identify high-risk cases in which adjuvant chemotherapy could be recommended [9]. Furthermore, mitotic rate or molecular abnormalities such as p53 or Bcl-2 do not seem to be correlated with clinical outcome as the four patients who were still alive at the end of the follow-up did not show different histological or molecular characteristics from other patients. Because of such a small population, a statistical analysis cannot be applied to show correlation. Even if LCNEC is a distinct subset of lung tumours, we do not know yet if we have to consider it as a SCLC with a two-stage classification (localised or diffuse) and a chemotherapy-based treatment or as a non-small cell carcinoma with a surgical treatment. The staging system
292
J. Mazie`res et al. / Lung Cancer 37 (2002) 287 /292
used in NSCLC seems to be inappropriate for LCNEC, as we could not distinguish the survival rates between the different stages. Our population is too small to allow definitive assertions but we can conclude that applying to LCNEC the NSCLC standard treatment leads to a poor prognosis. LCNEC is a particular subset of tumour that should not be treated with protocols designed for non-small cell carcinomas. In fact, surgery and adjuvant radiotherapy seems to be efficient in the loco-regional control of the disease as no loco-regional relapse was noted. The very high mortality rate is entirely due to metastatic relapse. This finding underlines the necessity of an adjuvant effective chemotherapy. The low response rate of the platin-etoposide combination suggests that new chemotherapy regimens should be evaluated in large-scale trials to assess the optimal treatment of LCNEC.
Acknowledgements The authors thank Bill Mazie`res and Maite´ Turmon for their help in reviewing the manuscript.
References [1] Gould VE, Linnoila RI, Memoli VA, Warren WH. Neuroendocrine cells and neuroendocrine neoplasms of the lung. Pathol Annu 1983;18:287 /330. [2] Travis WD, Linnoila RI, Tsokos MG, et al. Neuroendocrine tumours of the lung with proposed criteria for large-cell neuroendocrine carcinoma. An ultrastructural immunohistochemical, and flow cytometric, study of 35 cases. Am J Surg Pathol 1991;15:529 /53. [3] Travis WD, Rush W, Flieder DB, et al. Survival analysis of 200 pulmonary neuroendocrine tumours with clarification of criteria for atypical carcinoid and its separation from typical carcinoid. Am J Surg Pathol 1998;22:934 /44. [4] Travis WD, Colby TV, Corrin B, Shimosato Y, Brambilla E. Histological Typing of Lung and Pleural Tumors. In: Travis WD, Colby TV, Corrin B, Shimosato Y, Brambilla E, editors. World Health Organisation. International histological classification of tumours, 3rd ed.. Berlin, Heidelberg, New York: Springer, 1999. [5] Dresler CM, Ritter JH, Patterson GA, Ross E, Bailey MS, Wick MR. Clinical-pathologic analysis of 40 patients with large cell neuroendocrine carcinoma of the lung. Ann Thorac Surg 1997;63:180 /5.
[6] Carretta A, Ceresoli GL, Arrigoni G, et al. Diagnostic and therapeutic management of neuroendocrine lung tumours. A clinical study of 44 cases. Lung Cancer 2000;29:217 /25. [7] Mountain CF. Revisions in the International System for Staging Lung Cancer. Chest 1997;111:1710 /7. [8] Kaplan EM, Meier P. Non parametric estimations from incomplete observations. J Am Stat Assoc 1958;53:457 /81. [9] Brambilla E, Negoescu A, Gazzeri S, et al. Apoptosis-related factors p53, Bcl2, and Bax in neuroendocrine lung tumours. Am J Pathol 1996;149:1941 /52. [10] Jiang SX, Kameya T, Shoji M, Dobashi Y, Shinada J, Yoshimura H. Large cell neuroendocrine carcinoma of the lung: a histologic and immunohistochemical study of 22 cases. Am J Surg Pathol 1998;22:526 /37. [11] Ishida T, Kaneko S, Tateishi M, et al. Large cell carcinoma of the lung. Prognostic implications of histopathologic and immunohistochemical subtyping. Am J Clin Pathol 1990;93:176 /82. [12] Hammond ME, Sause WT. Large cell neuroendocrine tumours of the lung. Clinical significance and histopathologic definition. Cancer 1985;56:1624 /9. [13] Neal MH, Kosinski R, Cohen P, Orenstein JM. A typical endocrine tumours of the lung: a histologic, ultrastructural, and clinical study of 19 cases. Hum Pathol 1986;17:1264 /77. [14] Mooi WJ, Van Zandwijk N, Dingemans KP, Koolen MG, Wagenvoort CA. The ‘grey area’ between small cell and nonsmall cell lung carcinomas. Light and electron microscopy versus clinical data in 14 cases. J Pathol 1986;149:49 /54. [15] Piehl MR, Gould VE, Warren WH, et al. Immunohistochemical identification of exocrine and neuroendocrine subsets of large cell lung carcinomas. Pathol Res Pract 1988;183:675 /82. [16] Wick MR, Berg LC, Hertz MI. Large cell carcinoma of the lung with neuroendocrine differentiation. A comparison with large cell pulmonary tumours. Am J Clin Pathol 1992;97:796 /805. [17] Iyoda A, Hiroshima K, Toyozaki T, Haga Y, Fujisawa T, Ohwada H. Clinical characterization of pulmonary large cell neuroendocrine carcinoma and large cell carcinoma with neuroendocrine morphology. Cancer 2001;91:1992 /2000. [18] Jung KJ, Lee KS, Han J, et al. Large cell neuroendocrine carcinoma of the lung: clinical, CT, and pathologic findings in 11 patients. J Thorac Imaging 2001;16:156 /62. [19] Rusch VW, Klimstra DS, Venkatraman ES. Molecular markers help characterize neuroendocrine lung tumours. Ann Thorac Surg 1996;62:798 /809 (discussion 809-10). [20] Greco FA, Hainsworth JD. Practical approaches to the treatment of patients with extensive stage small cell lung cancer. Semin Oncol 1994;21:3 /6. [21] Lai SL, Goldstein LJ, Gottesman MM, et al. MDR1 gene expression in lung cancer. J Natl Cancer Inst 1989;81:1144 /50. [22] Moertel CG, Kvols LK, O’Connell MJ, Rubin J. Treatment of neuroendocrine carcinomas with combined etoposide and cisplatin. Evidence of major therapeutic activity in the anaplastic variants of these neoplasms. Cancer 1991;68:227 /32.