Basaloid Carcinoma of the Lung: A Really Dismal Histologic Variant?

Transcription

1 Carcinoma of the Lung: A Really Dismal Histologic Variant? Dae Joon Kim, MD, Kil Dong Kim, MD, Dong Hwan Shin, MD, Jae Y Ro, MD, and Kyung Young Chung, MD Departments of Thoracic and Cardiovascular Surgery, and Pathology, Yonsei University College of Medicine, and Department of Pathology, Ulsan University Medical College, Asan Medical Center, Seoul, South Korea Background. carcinoma of the lung has been reported as an uncommon and highly aggressive form of nonsmall cell lung cancers. Even in stage I and II of basaloid carcinoma, a 5-year survival rate of only 15% has been reported and it has been suggested that different treatment modalities for basaloid carcinoma should be considered. The aim of this study was to determine the prognostic implications of a basaloid carcinoma of the lung. Methods. This study included a series of 291 surgically resected lung tumors, which were originally diagnosed as a poorly or undifferentiated carcinoma, a small cell carcinoma, or an atypical carcinoid. Of these, 35 basaloid carcinoma patients were identified and compared with 167 poorly differentiated squamous cell carcinoma () patients in terms of the preoperative clinical data, the procedure performed, and the survival outcome. Results. The overall incidence of basaloid carcinoma was 4.8%. The actuarial 5-year survival rate was 40.6% in patients with and 36.5% in those with basaloid carcinoma (p 0.86). In stage I and II patients, the actuarial 5-year survival rate was 53.9% in the group and 57.2% in the basaloid group (p 0.97). There were no differences in the recurrence rate and the relapse pattern (p 0.584). Cox s proportional hazards model revealed that an age equal to 60 years old (hazard ratio 2.179, p 0.000) and an advanced stage (hazard ratio 2.264, p 0.000) were the risk factors for postoperative survival in both groups. Conclusions. carcinoma of the lung does not have a worse prognosis than the other nonsmall cell lung cancers. Although it is obvious that a basaloid carcinoma is a unique histologic entity, it does not require a different treatment modality due to the similar clinical behavior with other nonsmall cell lung cancers. (Ann Thorac Surg 2003;76:1833 7) 2003 by The Society of Thoracic Surgeons carcinoma of the lung was initially described by Brambilla and colleagues [1] in 1992 and is now defined according to the 1999 revised WHO/IASLC classification of lung tumors as a variant of squamous cell carcinoma or large cell carcinoma depending on the presence or absence of squamous carcinoma component [2, 3]. In a recent collective study [4], although the difference in the survival rate between a basaloid carcinoma and a poorly differentiated squamous cell carcinoma () in all stages was not statistically significant, the 5-year survival rate for stage I and II resected tumors was different (15% in basaloid carcinoma patients, compared with 47% in patients). They concluded that because this subset of nonsmall cell lung cancers (NSCLC) had a poorer prognosis than other NSCLC, a new treatment modality should be considered. The aims of this study were to evaluate the prognostic implications of basaloid carcinoma and to determine the clinical behavior (particularly in stage I and II) of this uncommon malignancy. Accepted for publication June 13, Address reprint requests to Dr Kyung Young Chung, Department of Thoracic and Cardiovascular Surgery, Yonsei University College of Medicine, CPO Box 8044, Seoul, South Korea ; kychu@yumc.yonsei.ac.kr. Material and Methods From 1991 through 2000 a total of 731 patients underwent a thoracotomy for a curative resection of lung cancer at our institute. Of these, 291 patients were diagnosed with a poorly or undifferentiated carcinoma, a small cell carcinoma, or an atypical carcinoid. The hematoxylin-eosin stained sections in all patients were initially examined by one of the authors (D.H.S.) in a blinded manner, and another pathologist (J.Y.R.) independently reviewed them. All discrepancies were solved by additional reviews until reached a mutual agreement. The patients with no consensus were excluded for further analysis. The pathologic diagnosis of basaloid carcinoma was based on the following criteria already described by Brambilla and coworkers in 1992 [1]: 1. Solid lobular or anastomotic trabecular pattern growing invasively in a fingerlike fashion from the bronchial and/or glandular duct lining; 2. Small cuboidal to fusiform cells of mean diameter 12 to 15 m, with moderately hyperchromatic nuclei and without prominent nucleoli. There was a scant but visible cytoplasm, and no nuclear molding; 2003 by The Society of Thoracic Surgeons /03/$30.00 Published by Elsevier Inc doi: /s (03)

2 1834 KIM ET AL Ann Thorac Surg BASALOID CARCINOMA OF THE LUNG 2003;76: Peripheral palisading with radially arranged cells at the periphery of lobules; 4. A high rate of mitosis, between 15 and 44 per 10 high-power fields. The prime differential diagnosis was a large cell neuroendocrine carcinoma (LCNEC). Neuroendocrine differentiation was determined by immunohistochemical staining with CD56 (Boehringer Mannheim, Mannheim, Germany), chromogranin A (DAKO Corporation, N.S., Glostrup, Denmark), and synaptophysin (DAKO). A tumor that stained focally to least one of the three neuroendocrine markers was judged positive for neuroendocrine differentiation. Immunohistochemical stains using 34 E12 (DAKO) and thyroid transcription factor-1 (TTF-1; Neomarkers, Fremont, CA) were performed to distinguish basaloid carcinoma from LCNEC. The diagnosis of a basaloid carcinoma was made by consensus after an optical microscopic review and immunohistochemical staining. The clinical information was collected from the medical records including the patient s sex, age, smoking status, functional class, history of other disease, surgical data, and survival outcome. The location of the tumor was determined by computed tomography and bronchoscopy. The stage of disease was based on the TNM classification of the International Union Against Cancer [5]. The actuarial survival rate was calculated from the time of surgical treatment until August 20, 2002 using the Kaplan-Meier method, and the curves obtained were compared with a log-rank test [6]. The risk factor analysis for postoperative survival was determined by Cox s proportional hazards multivariable regression model [7]. A p value less than 0.05 was considered statistically significant. Results Thirty-five of 291 patients who underwent a pathology review had the histologic and morphologic characteristics of a basaloid carcinoma. Their initial diagnosis were a poorly differentiated squamous cell carcinoma in 27 patients, a large cell carcinoma in 2 patients, an atypical carcinoid in 1 patient, small cell carcinoma in 1 patient, and a basaloid carcinoma in 4 patients, as illustrated in Table 1. The overall incidence of a basaloid carcinoma was 4.8%. Among these 35 basaloid carcinoma patients, 24 patients were classified as a pure form (basaloid pattern in at least 80% of the tumor bulk). The remaining 11 patients were classified as having the mixed form (basaloid pattern in at least 60% of tumor bulk), and all were associated with a squamous cell carcinoma. Immunohistochemical staining with CD56, chromogranin A, and synaptophysin revealed that 14 of 35 patients tested positive, but none tested positive for all three neuroendocrine markers, as revealed in Table 2. Immunoreactivity for 34 E12 was detected in 28 patients (80%) of basaloid carcinoma, and immunostains for TTF-1 were all negative. The 167 patients remaining were classified with, Table 1. Result of the Optical Microscopic Review Initial Diagnosis Given New Diagnosis of Carcinoma (13.9) PDAC 44 0 (0.0) Large cell carcinoma 27 2 (7.4) Atypical carcinoid 8 1 (12.5) Small cell carcinoma 13 1 (7.7) carcinoma 5 4 (80.0) Total (12.0) poorly dif- PDAC poorly differentiated adenocarcinoma; ferentiated squamous cell carcinoma. and the clinical features and survival outcomes of the two groups were compared: 167 s and 35 basaloid carcinomas. The preoperative clinical features of the two groups are listed in Table 3. There was no history of a previous cancer with a basaloid histology, which suggested that this lung cancer could be metastatic. The age, sex, smoking history, functional class, predicted FEV1 (% of normal), and tumor location in the two groups were similar. There was also no statistical differences in the clinical data in the stage I and II tumors between two groups, as presented in Table 4. The surgical procedure, TNM stage, and postoperative treatments are summarized in Table 5. The complete resection rate, surgical mortality, TNM classification, and postoperative treatment between the two groups were similar. A pneumonectomy was performed in 99 patients (59.3%) in the group and in 13 patients (37.1%) of the basaloid carcinoma group (p 0.017). The surgical data in the stage I and II patients were similar in terms of the proportion of the pneumonectomy as well as the other variables, as illustrated in Table 6. A complete follow-up was obtained in all survivors (159 s and 34 basaloids) and the median follow-up time was 26.5 months. Ninety-six patients in the group and 22 patients in the basaloid carcinoma group died during the follow-up. The survival data were available in all patients and the cumulative survival of each group of patients is illus- Table 2. Pattern of Immunohistochemical Staining in Carcinomas CD56/CGA/Syn With 34 E12( ) With TTF-1 ( ) / / / / / / / / / / CGA chromogranin A; Syn synaptophysin; TTF-1 thyroid transcrition factor 1.

3 Ann Thorac Surg KIM ET AL 2003;76: BASALOID CARCINOMA OF THE LUNG Table 3. Clinical Data of With Carcinoma (n 167) and Carcinoma (n 35) Table 5. Surgical Data of With Carcinoma (n 167) and Carcinoma (n 35) 1835 Age, median (years) Male 158 (94.0) 31 (91.2) Cardiovascular disease a 18 (10.7) 6 (17.6) Functional class b 2 46 (27.4) 11 (32.4) Smoking history 126 (75.4) 22 (62.9) Packs-year FEV1 (% of normal) Cancer history 7 (4.2) 2 (5.9) Stomach 2 1 Colorectal 3 0 Breast 1 1 Uterine cervix 1 0 Tumor location c Central 92 (55.1) 21 (60.0) Peripheral/intermediate 75 (44.9) 14 (40.0) a History of documented hypertension, coronary arterial occlusive disease, or cerebrovascular accidents. b According to the New York Heart Association functional class. c According to the computed tomographic and bronchoscopic findings. poorly differentiated squamous cell carcinoma. trated in Figure 1. The actuarial 5-year survival rate was 40.6% in the group and 36.5% in the basaloid carcinoma group (p 0.86). The median survival was 34.0 months in the group and 34.4 months in the basaloid carcinoma group. The survival of the patients with stage I and II disease are depicted in Figure 2. The actuarial 5-year survival rate was 53.9% in the group and 57.2% in the basaloid carcinoma group (p Table 4. Clinical Data of With a Stage I, II Carcinoma (n 85) and Carcinoma (n 19) Age, median (years) Male 81 (94.2) 18 (94.7) Cardiovascular disease a 14 (16.3) 4 (21.1) Functional class b 2 18 (20.9) 3 (15.8) Smoking history 64 (75.3) 13 (68.4) Packs-yr FEV1 (% of normal) Cancer history 2 (2.3) 1 (5.3) Stomach 1 1 Colorectal 0 0 Breast 0 0 Uterine cervix 1 0 Tumor location c Central 41 (48.2) 10 (52.6) Peripheral/intermediate 44 (51.8) 9 (47.4) a History of documented hypertension, coronary arterial occlusive disease, or cerebrovascular accident. b According to the New York Heart Association functional class. c According to the computed tomographic and bronchoscopic findings. poorly differentiated squamous cell carcinoma. Operation performed Pneumonectomy 99 (59.3) 13 (37.1) Bi-, lobectomy 68 (40.7) 22 (62.9) Complete resection 145 (86.8) 31 (88.6) Operative mortality 8 (4.8) 1 (2.9) Pathologic staging a Stage I 54 (32.4) 14 (40.0) Stage II 31 (18.6) 5 (14.3) Stage IIIA 53 (31.7) 13 (37.1) Stage IIIB 21 (12.6) 2 (5.7) Stage IV 8 (4.8) 1 (2.9) Postoperative treatment Chemotherapy 9 (5.7) 4 (11.8) Radiation therapy 40 (25.1) 9 (26.4) Combined 41 (25.8) 7 (20.6) None 69 (43.3) 14 (41.2) a According to the AJCC TNM cassification (1997). poorly differentiated squamous cell carcinoma. 0.97). There was no significant difference in the survival between the two groups in stage I and II disease. The survival was further analyzed in the patients with stage I disease. There were 52 patients in the group and 14 patients in the basaloid carcinoma group. In these patients with no nodal metastasis, the actuarial 5-year survival rate and median survival were 62.1% and months in the group, and 71.8% and 79.6 months in the basaloid carcinoma group. No statistical difference was found between two groups in the stage I disease (p 0.79). A tumor recurrence occurred in 53 patients (33.3%) from the group and 13 patients (38.2%) from the basaloid carcinoma group. The sites of the firstdocumented recurrence are indicated in Table 7. The majority experienced a distant relapse and the tumor recurrence was noted quite soon after the operation. The Table 6. Surgical Data of With Stage I, II Carcinoma (n 85) and Carcinoma (n 19) Operation performed Pneumonectomy 41 (48.2) 7 (36.8) Bi-, lobectomy 44 (51.8) 12 (63.2) Complete resection 84 (98.8) 18 (94.7) Operative mortality 5 (5.9) 0 (0.0) Postoperative treatment Chemotherapy 3 (3.8) 2 (10.5) Radiation therapy 15 (18.7) 5 (26.3) Combined 7 (8.7) 1 (5.3) None 55 (68.8) 11 (57.9) poorly differentiated squamous cell carcinoma.

4 1836 KIM ET AL Ann Thorac Surg BASALOID CARCINOMA OF THE LUNG 2003;76: Table 7. Mode of Recurrence Recurrence 53 (33.3) 13 (38.2) Local 8 (15.1) 2 (15.4) Distant a 45 (84.9) 11 (84.6) Lung 17 (37.8) 3 (27.2) Bone 12 (26.7) 5 (45.5) Brain 10 (22.2) 1 (9.1) Liver 6 (13.3) 2 (18.2) a Site of first-documented recurrence. poorly differentiated squamous cell carcinoma. Fig 1. Actuarial survival of the patients with a and a basaloid carcinoma in stage I IV, p ( poorly differentiated squamous cell carcinoma.) ; basaloid carcinoma. disease-free time was 7.0 months in the group and 10.0 months in the basaloid carcinoma group (p 0.422). In terms of recurrence, the recurrence rate and the pattern of relapse in the two groups were similar. Multivariate analysis was performed to analyze the patient s survival with respect to age, history of cardiovascular disease, preoperative functional class, type of operation, incomplete resection, histology, and pathologic staging, as demonstrated in Table 8. The analyses indicated that an age equal to 60 years old (hazard ratio 2.179, p 0.000) and advanced stage of III and IV (hazard ratio 2.264, p 0.000) were significant prognostic factors adversely influencing the postoperative survival regardless of tumor histology, or basaloid carcinoma. Comment Brambilla and coworkers [1] initially described the characteristic histopathologic features of a basaloid carcinoma of the lung. A basaloid carcinoma usually characterizes a solid, lobular pattern of growth, scant but visible cytoplasm, moderately hyperchromatic nuclei without prominent nucleoli, peripheral palisading, and a high rate of mitosis [1 3]. This uncommon subtype of NSCLC was introduced as a distinct entity in the recently revised WHO/IASLC classification of lung tumors in 1999 [2]. However there are some difficulties in diagnosing a basaloid carcinoma of the lung, as several studies have pointed out [8, 9]. First, a preoperative diagnosis of a basaloid carcinoma is difficult, particularly when a small number of cells are obtained from a bronchoscopic bi- Table 8. Cox s Proportional Hazards Model of Factors Influencing Postoperative Survival Variable Hazard Ratio 95% CI p Value Fig 2. Actuarial survival of the patients with a stage I and II and basaloid carcinoma, p ( poorly differentiated squamous cell carcinoma.) ; basaloid carcinoma. Age Cardiovascular disease Functional class histology Incomplete resection Pneumonectomy Advanced stage ( III) CI confidence interval.

5 Ann Thorac Surg KIM ET AL 2003;76: BASALOID CARCINOMA OF THE LUNG opsy and fine needle aspiration. Second, the requirement of a content of at least 60% basaloid cells is arbitrary. This criteria might cause interobserver difference because basal cells are noted in all squamous cell carcinomas and in other carcinomas in situ. Finally, a differential diagnosis of the pure form of basaloid carcinoma from LCNEC and an intermediate type of small cell lung cancer (SCLC) is required because these tumors have some similarities in their morphology. Sturm and colleagues [10] reported that TTF-1 expression excluded a diagnosis of basaloid carcinoma, and positive 34 E12 staining excluded a pure LCNEC. In this study positive immunohistochemical staining for 34 E12 was noted in 80% of basaloid carcinomas, and TTF-1 immunoreactivity was detected in none of them. Therefore, TTF-1 and 34 E12, in association with the specific neuroendocrine markers, are believed to be a useful panel of antibodies for differentiating basaloid carcinomas from LCNECs. Moro and coworkers [4] reported that in the early stages of the disease (stage I and II), the probability of a 5-year survival in basaloid carcinomas was 15% compared with 47% in s. Therefore, they concluded that a basaloid carcinoma has an intermediate prognosis between a SCLC and a resectable NSCLC. However, their report revealed a difference only in stage I and II patients. In this study, there was no statistical difference with respect to the preoperative clinical data, the type of operation, and the postoperative treatment between patients with basaloid carcinoma and in stage I and II diseases. The 5-year survival rate was similar in the two groups with stage I and II diseases (57.2% in basaloid group and 53.9% in group). During the same time period, the 5-year survival rate was 55.8% in patients with stage I and II NSCLCs at our institute. As for recurrence, there was no difference in the recurrence rate as well as the pattern of relapse between two groups. In our series, Cox s proportional hazards model indicated that an age equal to 60 years old and an advanced stage (equal to III), not the basaloid histology, were adverse prognostic factors for the postoperative survival. Brambilla and coworkers [1] suggested that a high rate of proliferation markers and the small size of the tumor cells were responsible for the poor prognosis of basaloid carcinoma. However, it is questionable as to whether the above histologic findings have more impact on survival than variables such as the TNM staging, age, or others. A basaloid carcinoma needs to be compared with other NSCLCs in stage T1N0 or T2N0 in order to clarify the relationship between the histologic findings and the survival outcome. In our series there was no statistical difference in survival between the two groups in stage I disease. The actuarial 5-year survival rate was 62.1% in stage I s and 71.8% in stage I basaloid carcinomas (p 0.79). Although the number of stage I basaloid carcinomas was small, it is thought that the basaloid histology itself has a similar outcome with other NSCLCs. More patients and a thorough histologic workup are needed to demonstrate this hypothesis. In conclusion, basaloid carcinoma of the lung does not have a worse prognosis than other nonsmall cell lung cancers. Because the postoperative survival and recurrence were not different from those of other NSCLCs, newer treatment protocols specific to basaloid carcinoma are not warranted. This work was supported by Yonsei University Faculty Research Fund (2002). References Brambilla E, Moro D, Veale D, Brichon PY, Paramelle B, Brambilla C. Basal cell (basaloid) carcinoma of the lung: a new morphologic and phenotypic entity with separate prognostic significance. Hum Pathol 1992;23: Travis WD, Colby TV, Corrin B, Shimosato Y, Brambilla E. (In collaboration with Sobin LH and pathologists from 14 countries). Histological typing of lung and pleural tumors. World Health Organization. International Histological Classification of tumors. 3rd edition. Berlin: Springer, Brambilla E, Travis WD, Colby TV, Corrin B, Shimosato Y. The new World Health Organization classification of lung tumors. Eur Resp J 2001;18: Moro D, Brichon P, Brambilla E, Veale D, Labat F, Brambilla C. bronchial carcinoma: a histologic group with a poor prognosis. Cancer 1994;73: Hernabek P, Hutter RVP, Sobin LH, Wager G, Wittekind C, editors. TNM atlas. 4th edition. Berlin: Springer, Kaplan E, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958;53: Cox DW. Regression models and life tables. J R Stat Soc 1972;34: Dugam JM. Cytologic diagnosis of basal cell (basaloid) carcinoma of the lung. A report of two cases. Acta Cytol 1995;39: Foroulis CN, Iliadis KH, Mauroudis PM, Kosmidis PA. carcinoma, a rare primary lung neoplasm: report of a case and review of the literature. Lung Cancer 2002;35: Sturm N, Lantuejoul S, Laverriere MH, et al. Thyroid transcription factor 1, and cytokeratin 1,5,10,14 (34 E12) expression in basaloid, and large-cell neuroendocrine carcinomas of the lung. Hum Pathol 2001;32: