Large cell carcinoma of the lung with neuroendocrine

Transcription

1 ORIGINAL ARTICLES: GENERAL THORACIC Clinical-Pathologic Analysis of 40 Patients With Large Cell Neuroendocrine Carcinoma of the Lung Carolyn M. Dresler, MD, Jon H. Ritter, MD, G. Alexander Patterson, MD, Eric Ross, PhD, Marci S. Bailey, RN, and Mark R. Wick, MD Washington University School of Medicine, St. Louis, Missouri Background. The identification, classification, and appropriate treatment of patients with pulmonary carcinomas demonstrating neuroendocrine differentiation remains controversial. Methods. Patients undergoing resection of lung cancer at Washington University since 1986 were reviewed to identify all large cell neuroendocrine carcinomas. Cases were segregated into large, small, or mixed cell categories, and graded as moderate ( atypical carcinoid ) or poorly differentiated (all higher grade lesions). All patients charts were reviewed and referring physicians contacted to ascertain cancer treatment after resection and follow-up status. Results. Forty patients were identified with large cell neuroendocrine carcinoma: 8 moderate and 32 highgrade. Average follow-up was 19.8 months. Stage distribution was as follows: I, 25; II, 6; III, 6; and VI, 3. Fifteen patients have no evidence of disease, 15 are dead of disease, and 6 are alive with disease. Five-year survival of the stage I patients is 18%; all-stage 5-year survival is 13%. Of the 15 patients who died of their disease, 80% had stage I or II disease. Postoperative chemotherapy, radiation therapy, or both were given to 9 of 26 patient in stage I, with six deaths (67%). Six of 17 patients (35%) with stage I disease died after no postoperative intervention. Conclusions. Large cell neuroendocrine carcinomas identified by histologic examination have a remarkably poor prognosis even in very early stage disease. Adjuvant therapy did not improve survival. (Ann Thorac Surg 1997;63:180 5) 1997 by The Society of Thoracic Surgeons Presented at the Poster Session of the Thirty-second Annual Meeting of The Society of Thoracic Surgeons, Orlando, FL, Jan 29 31, Address reprint requests to Dr Dresler, Department of Surgery, Fox Chase Cancer Center, 7701 Burholme Ave, Philadelphia, PA Large cell carcinoma of the lung with neuroendocrine differentiation has been increasingly well studied. Most studies to date have explored the increasing variety of immunologic stains available for neuroendocrine markers. Despite many studies, one or even a specific battery of markers is not sufficiently specific to identify lung carcinomas with neuroendocrine differentiation [1]. In addition, the number of patients with this histologic type is still relatively small in comparison with squamous cell carcinoma or adenocarcinoma. Thus, the appropriate treatment and prognosis of these patients remains unclear. The pathologic clarification of specific characteristics of these tumors is critical to correctly categorize them. Significant controversy exists over the classification of these carcinomas as atypical carcinoids, malignant carcinoids, or neuroendocrine carcinomas. Often these tumors have a mixed phenotype of neuroendocrine aspects combined with other non small cell components. Potentially, these pathologic characteristics of large cell neuroendocrine carcinomas may be pivotal in determining appropriate therapy. Several reports have suggested that these tumors are more chemosensitive, similar to small cell carcinomas [2, 3]. More recently, however, identification of specific neuroendocrine markers has not predicted response to chemotherapy or survival. Therefore, this group of non small cell lung cancer remains a puzzle. It is critical that we establish pathologic guidelines for classification of these tumors to assist in determining the accurate prognosis and to help guide the treatment of these not so uncommon patients. Material and Methods The tumor registry at Washington University was reviewed to cull all cases of large cell neuroendocrine carcinoma of histologic grades II and III/III seen between the years of 1986 and All slides of non small cell carcinoma encountered during that time were reviewed by two pathologists (J.H.R. and M.R.W.), who segregated those tumors that fulfilled predefined criteria for the diagnosis of large cell neuroendocrine carcinoma [4]. These requirements included the following histologic features: (1) an organoid growth pattern, with cells arranged in relatively uniformly sized clusters, cords, ribbons, or trabecula; (2) a monotonous cytologic appearance; (3) size of the neoplastic cells ranging between four and six times that of normal mature lymphocytes (40 to 60 m in diameter); (4) dispersed or evenly granular nuclear chromatin, with or without small nucleoli or molding of nuclear membranes; (5) nuclear volume at least two thirds of total cellular volume; (6) foci of geographic necrosis; and (7) mitotic activity in the range of five or more division figures per 10 microscopic fields 1997 by The Society of Thoracic Surgeons /97/$17.00 Published by Elsevier Science Inc PII S (96)

2 Ann Thorac Surg DRESLER ET AL 1997;63:180 5 LARGE CELL NEUROENDOCRINE CARCINOMA 181 Table 1. Histologic Grading Criteria for Neuroendocrine Carcinoma of the Lung and Related Tumors Tumor NHT Cell Size ( m) Geo Nec Mitoses NM Nucleoli Grade 1 NEC Classic carcinoid /10 HPF 0 Usually 0 Grade 2 NEC Atypical carcinoid to 5 and 10 per 0to 10 HPF Grade 3 NEC, large-cell type Large cell anaplastic to 10/10 HPF to to carcinoma Grade 3 NEC, small-cell type Oat cell carcinoma to 10/10 HPF 0to Large cell carcinoma with occult neuroendocrine differentiation a Large cell anaplastic carcinoma to Variable 0 to to a Large cell carcinoma with neuroendocrine features often lacks an organoid growth pattern on scanning microscopy as well as cellular monotony, and does not have the generic cytoarchitectural features of NEC. It is instead indistinguishable from truly undifferentiated (anaplastic) large cell carcinoma of the lung. Geo Nec foci of spontaneous geographic tumor necrosis; HPF high-power ( 400) microscopic fields; NEC neuroendocrine carcinoma; NHT nosologically homologous term; NM nuclear molding; 0 absent; present but inconspicuous; scattered but obvious; prominent. at a magnification of 400. Aside from the last three of these criteria, all of the other features would be readily recognized by pathologists as common to the entire spectrum of neuroendocrine neoplasms, including those in the low-grade category. As such, it is acknowledged that these standards doubtlessly resulted in inclusion of some lesions that would have been considered in the formerly employed diagnostic category of atypical carcinoid, which we do not use because of its current diagnostic adulteration [5], but examples of classic carcinoid were excluded because of the distinctive and generally indolent behavioral attributes of those tumors. Criteria for grading of neuroendocrine carcinomas of the lung are provided in Table 1, along with an attempt at correlating the resulting categories with probable nosologic homologues that can be found in prior works. In an attempt to provide a straightforward working definition of large cell neuroendocrine carcinoma, we believe that its overall cytoarchitectural features are remarkably similar to those of classic small cell neuroendocrine ( oatcell ) carcinoma, but the individual size of constituent cells is larger by a factor of 1.5 to 2 times. This statement applied to all cases in the current series. Representative cases are illustrated in Figures 1 through 3. The type of operation, radiation therapy and chemotherapy, and stage at presentation were documented by chart review. The patients were followed up until the time of submission of this article. As this review of these patients is retrospective, it entailed a wide variety of treatment approaches. Analysis of data necessarily was descriptive. Statistical analysis was performed using log-rank tests [6] to determine whether clinical characteristics (stage I, II, III, or IV), cell size (mixed versus large), differentiation (moderate versus poor), and treatment information were associated with significant differences in the time to failure. All pairwise comparisons of the different stages were performed. For these comparisons the type I error was adjusted using the Bonferroni procedure to control the experimentwise type I error. Failure time was defined as either time to recurrence or time to death where the cause of death was determined to be the tumor. Individuals who were disease free at the end of the observation Fig 1. Grade 2 neuroendocrine carcinoma ( atypical carcinoid ) demonstrates moderate to large cells, scattered mitoses (less than 5 per 10 high power fields), rare individual cell necrosis or small foci of necrosis in the center of nests, and modest pleomorphism. (A) Low-power magnification shows organoid nests of cells, some peripheral palisading, and focal necrosis in the center of one of the nests. (B) Higher power magnification shows midsized to large cells with some mild pleomorphism. The cells have granular chromatin and small chromocenters, without nucleoli. Mitoses are not seen in this field. Focal necrosis is seen.

3 182 DRESLER ET AL Ann Thorac Surg LARGE CELL NEUROENDOCRINE CARCINOMA 1997;63: months. Table 2 describes the stage at presentation, histology, treatment, operative procedure, and disease status. The collection of data concerning the site of first recurrence was incomplete because of the retrospective nature of the study. However, in the patients for whom data were available, the most common site of recurrence was in the brain (n 6), followed by local recurrence (n 3), breast (n 2), and bone, adrenal gland, and liver (n 1 each). To simplify analysis, all varieties of differentiation labeled as poor, whether poor/neuroendocrine features, poor/mixed, or poor, were grouped together. Similarly, the single patient with cell size reported as large and small was amalgamated in the large group. No correlation could be identified between the histopathologic findings and the stage or disease status. Figure 4 demonstrates the survival curves. Statistically significant differences were identified between stage IV and stages I, II, and III patients. There were no statistically significant differences among stages I, II, or III. There was no prolongation of survival or disease-free status identified in patients who received adjuvant therapy compared with those who did not (Fig 5). Fig 2. Grade 3 neuroendocrine carcinoma, large cell type ( large cell neuroendocrine carcinoma ) contains large cells, poorly differentiated. There is extensive necrosis and moderate to marked pleomorphism, the nuclei may be vesicular or stippled, they may have nucleoli, and there is a very high mitotic rate. (A) At low-power magnification, most of the field is geographic necrosis, surrounded by nests of viable tumor. (B) High-power magnification shows large cells and visible cytoplasm, with vesicular to granular chromatin, some cellular molding, some nucleoli, and four mitoses in this field alone. Comment The nosology of neuroendocrine lung tumors (indeed, or neuroendocrine neoplasms anywhere in the body) can only be described as confusing. Many practitioners still regard grade 1 neuroendocrine carcinoma ( classic carcinoid ) as a benign neoplasm, despite the fact that up to 15% of such lesions demonstrate metastasis and thereby prove that all such lesions are low-grade malignancies. Atypical carcinoid has often been used as a wastebasket category in which to place lesions that cannot be classified anywhere else by the pathologist, and, at best, the criteria used for the recognition of this neoplasm are considered perplexing. Moreover, there has been no diagnostic provision for those tumors considered in this period, who died of an unrelated cause, or who were lost to follow-up were considered censored. In the cases where very small samples were being compared, bootstrap methods were used to verify the log-rank tests, which assume large samples [7]. In all of the cases examined, the bootstrap results confirmed the significance (or lack of significance) of the log-rank tests. Results Forty patients were identified as having large cell neuroendocrine carcinoma of the lung. The average age was 65.9 years (range, 29 to 83 years). There were 18 women (45%) and 22 men (55%). Follow-up was an average of Fig 3. Small cell carcinoma contains small hyperchromatic cells, little cytoplasm, and molding. This is a grade 3 neuroendocrine carcinoma, small cell type.

4 Ann Thorac Surg DRESLER ET AL 1997;63:180 5 LARGE CELL NEUROENDOCRINE CARCINOMA 183 Table 2. Treatment, Histology, and Outcome (n 40 patients) Cell Size Differentiation Procedure Stage Treatment Status Large Mixed Lobe; RLL; med T1 N0 None NED Large Moderate Lobe; LUL T2 N1 RT DOD Large Moderate Lobe; RLL; med T2 N0 None NED Large Moderate Lobe; RML, RLL T2 N2 Cp/VP 4/RT AWD Large Moderate Lobe; RUL T3 N0 None NED Large Moderate Wedge; LLL T1 N0 None NED Large Moderate Wedge; LLL T1 Nx a None NED Large Moderate Wedge; RLL T1 Nx CAV 2 DW/OD Large Moderate Wedge; RUL T1 M1 None AWD Large Poor Lobe; LUL T2 N0 CAV/Carbo VP DW/OD Large Poor Lobe; LUL T2 N2 Cp/VP 4 NED Large Poor Lobe; RUL T1 N0 CAV 3 DOD Large Poor Lobe; RUL T1 N2 Cp/VB/Mito DOD Large Poor Lobe; RUL T2 N1 None DOD Large Poor Lobe; RUL; med T2 N2 Cp 7/RT AWD Large Poor Lobe; LLL T1 N0 None NED Large Poor Lobe; LLL; med T1 N0 None NED Large Poor Lobe; LUL T1 N0 None DOD Large Poor Lobe; LUL T1 N0 None DW/OD Large Poor Lobe; LUL T1 N0 None NED Large Poor Lobe; RLL T1 N0 CAV 4 DOD Large Poor Lobe; RLL T1 N0 None DOD Large Poor Lobe; RLL; med T1 N0 Cp/VP 5 NED Large Poor Lobe; RUL; med T1 N0 None DOD Large Poor Lobe; RUL; med T2 N1 Carbo/VP 6 AWD Large Poor Lobe; RUL; med T2 N1 M1 RT AWD Large Poor Wedge; RML T1 N0 Cp/VP 4 DW/OD Large Poor Wedge; RUL T1 Nx Cp/VP 1 DOD Large Poor Wedge; LLL T1 N0 None DOD Large Poor Wedge; LUL T1 N2 RT DOD Large Poor Wedge; LUL T1 Nx None NED Large Poor Wedge; RUL T1 Nx Cp/VP 4/RT NED Large Poor/mixed Lobe; LUL T1 N1 CAV 3 DOD Large Poor/mixed Lobe; RUL T1 N0 None DOD Mixed Poor Endobronchial bx; RUL T3 M1 CAV 5 DOD Mixed Poor Lobe; LUL T1 N0 None NED Mixed Poor Lobe; LUL; med T2 N1 Carbo/VP 4/RT NED Mixed Poor Lobe; RUL T1 N0 None AWD Mixed Poor Lobe; RUL; med T1 N0 Cp/VP 6 NED Mixed Poor Pneumonectomy T1 N1 CAV/Cp/VP 4/RT DOD a T1Nx were stage 1 for analysis purposes as resection was limited to a wedge. AWD alive with disease; Carbo carboplatin; CAV cyclophosphamide, doxorubicin, vincristine; Cp cisplatin; DOD died of disease; DW/OD died without disease; LLL left lower lobe; lobe lobectomy; LUL left upper lobe; med mediastinoscopy; Mito mitomycin; NED no evidence of disease; RLL right lower lobe; RML right middle lobe; RT radiation therapy; RUL right upper lobe; VB vinblastine; VP etoposide. series, which strongly resemble small cell carcinomas but are composed of much larger cells; as a consequence, they have been included with the large cell anaplastic carcinomas, not further specified. In response to these problems, we recommend our scheme of classification. It has the advantages of simplicity and reproducibility, giving the practitioner two important pieces of information as integral parts of the diagnostic label. First, it is stated unequivocally that a neoplasm is malignant and epithelial in general (ie, a carcinoma) and neuroendocrine in particular. Second, a relative grade of malignancy is provided of necessity. These points are, in our opinion, a logical and supportable response to the inevitable question of why another classification scheme?. It should also be emphasized that large cell neuroendocrine carcinoma and other forms of unqualified neuroendocrine carcinoma differ from those tumors that are termed large cell carcinoma with neuroendocrine features [4]. The latter lesions are recognized as having endocrine characteristics only after specialized immunohistologic

5 184 DRESLER ET AL Ann Thorac Surg LARGE CELL NEUROENDOCRINE CARCINOMA 1997;63:180 5 or ultrastructural studies have been done, as opposed to the identification of a neuroendocrine lineage by routine microscopy, as is possible with straightforward neuroendocrine carcinoma. Hence, we would suggest that the lesions previously categorized as large cell carcinoma with neuroendocrine features should be called large cell carcinoma with occult neuroendocrine differentiation. We have accumulated over the past 9 years a relatively large group of patients that have been carefully characterized pathologically and treated in a variety of methods. Because of the concern of the more aggressive nature of these tumors, we treated many patients with adjuvant chemotherapy and/or radiation therapy, despite their Stage I classification. Although the treatment protocols were not standardized, they were generally cis-platinum based and usually for 4-6 courses. Even in Stage I, this relatively aggressive management did not alter survival. This insensitivity to chemotherapy may have been suggested in 1989 by Lai and colleagues [8]. They examined the multiple drug resistance gene (MDR1) expression in a variety of lung cancer and lung cancer cell lines. This MDR1 gene is frequently expressed in tumors that are chemoresistant or have developed chemoresistance. In their study, only in the subgroup of lung cancer with neuroendocrine markers did a majority of tumors demonstrate the MDR1 gene, perhaps predicting the findings in our study. The staging system used in lung cancer is supposed to differentiate groups according to prognosis. In our group of patients we could not distinguish survival between Fig 5. Kaplan-Meier survival for patients with stage I neuroendocrine lung cancer: adjuvant (chemotherapy, radiation, or both) therapy versus no adjuvant therapy. Fig 4. Kaplan-Meier survival curves for all stages. groups I, II, or III. Although we did demonstrate a difference between stage IV and stages I, II and III, this is hardly surprising, in spite of the fact that only 2 patients were in the stage IV group. Even if we could apply the staging system of small cell lung cancer extensive versus limited the results would not be helpful. The majority of our patients would be in the limited stage and yet still have a remarkably poor prognosis. Something is worse within this category of lung cancer with neuroendocrine differentiation, and its etiology has not been identified. Recently we [9] carefully reviewed 136 patients with stage I non small cell lung cancer that was completely resected during the same time period as the present patient cohort was identified. The 5-year survival for this contemporary control group with no evidence of neuroendocrine differentiation was 57%. Although this survival rate is low according to recent reports for T1 N0 patients, it is considerably better than the 18% 5-year survival of patients with large cell carcinoma with neuroendocrine differentiation. In summary, our findings show that identification of neuroendocrine differentiation in lung carcinoma by itself portends a poor prognosis. The description of cell size and, more particularly, the degree of differentiation were not statistically predictive. More importantly, despite the tumor s propensity to present at a low stage, the mortality was impressive and did not correlate with

6 Ann Thorac Surg DRESLER ET AL 1997;63:180 5 LARGE CELL NEUROENDOCRINE CARCINOMA 185 stage. Therefore, despite the careful pathologic evaluation of these tumors, no specific characteristic was noted that could predict the prognosis of these patients. Clearly, more work, perhaps in the area of genetic markers, is necessary to elucidate this puzzling and impressively poor prognosis in patients with lung cancer with neuroendocrine differentiation. References 1. Addis BJ. Neuroendocrine differentiation in lung carcinoma. Thorax 1995;50: Graziano SL, Mazid R, Newman N, et al. The use of neuroendocrine immunoperoxidase markers to predict chemotherapy response in patients with non small cell lung cancer. J Clin Oncol 1989;17: Linnoila RI, Jensen S, Steinberg S, Minna J, Gazdar AF, Mulshine JL. Neuroendocrine differentiation correlates with favorable response to chemotherapy in patients with non small cell lung cancer. Lung Cancer 1988;4:A Travis WD, Linnoila RI, Tsokos MG, et al. Neuroendocrine tumors 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: Wick MR, Berg LC, Maj MC, Hertz MI. Large cell carcinoma of the lung with neuroendocrine differentiation. Am J Clin Pathol 1992;97: Cox DR, Oakes D. Analysis of survival data. New York: Chapman and Hall, Efron B, Tibshirani RJ. An introduction to the bootstrap. New York: Chapman and Hall, Lai S-L, Goldstein LJ, Gottesman MM, et al. MDR1 gene expression in lung cancer. J Natl Cancer Inst 1989;81: Dresler CM, Ritter JH, Wick MR, Roper CL, Patterson GA, Cooper JD. Immunostains for blood group antigens lack prognostic significance in T1 lung carcinoma. Ann Thorac Surg 1995;59: Notice From the American Board of Thoracic Surgery The part I (written) examination will be held at the Atlanta Airport Hilton and Towers, Atlanta Airport, Atlanta, GA, on February 1, The closing date for registration is August 1, To be admissible for the part II (oral) examination, a candidate must have successfully completed the part I (written) examination. A candidate applying for admission to the certifying examination must fulfill all the requirements of the board in force at the time the application is received. Please address all communications to the American Board of Thoracic Surgery, One Rotary Center, Suite 803, Evanston, IL by The Society of Thoracic Surgeons Ann Thorac Surg 1997;63: /97/$17.00 Published by Elsevier Science Inc PII S (96)