Seminars in Surgical Oncology 653-63 (1990)
Clinical Applications of the Pathological Properties of Small Cell Carcinoma, Large Cell Carcinoma, and Adenoid Cystic Carcinoma of the Lung TERUYOSHI ISHIDA,MD, TOKUJIROYANO, MD, and KEIZO SUGIMACHI,MD From the Department of Surgery II, Faculty of Medicine, Kyushu University, Fukuoka, Japan
Lung cancer has considerable treatment problems, with a poor 5-yr survival rate after surgery. Application of histopathological and immunohistopathological subtyping have proven to be powerful tools for the assessment of prognosis. Results in 33 patients with small cell carcinoma, 44 with large cell carcinoma, and five with adenoid cystic carcinoma of the lung are discussed. The 5-yr survival rates in patients with small cell carcinoma of the oat cell type and intermediate type were 24 and 44%, respectively. Argyrophilic granules and neuron-specific enolase, neuroendocrine markers, were detected more frequently in the oat cell type rather than in the intermediate type. In contrast, keratin, epithelial membrane antigen, and carcinoembryonic antigen, epithelial origin markers, were present more frequently in the intermediae type than in the oat cell type. The 5-yr survival rates with large cell carcinoma of the compact growth type and the loose structure type were 46 and 28%, respectively, based on evidence of morphologically intercellular cohesion. The epithelial origin markers were detectable and more frequent in the compact growth type than in the loose structure type. The growth patterns of adenoid cystic carcinoma are classified as tubular, cribriform, and solid. The solid pattern was the most aggressive with extensive perineural invasion. We propose that the pathological properties of lung cancer should be examined as a prognostic implication of subtyping. KEYWORDS:lung cancer, histological subtyping, immunohistochemical markers
INTRODUCTION Primary lung cancer has been categorized into four main groups, according to the World Health Organization Histological Typing of Lung Tumors [ 11. These subgroups are squamous cell carcinoma, small cell carcinoma, adenocarcinoma, and large cell carcinoma. There are differences in both the biologic behavior and clinicopathologic manifestations among the four subgroups [2]. Although specific criteria have been given to the major cell types, criteria for the subgroups frequently have been either minimal or misleading. Electron microscopy [3-6], immunohistochemistry [7-1 l], and basic experimental research techniques [ 121 have profoundly altered and immensely improved our under-
0 1990 Wiley-Liss, Inc.
standing of the embryogenesis, histogenesis, and morphogenesis of all lung cancers. The 5-yr survival rate of patients with lung cancer is now reported to reach about 30%, in cases of surgical treatment [ 131. Both small cell carcinoma and large cell carcinoma show a fairly rapid growth and dissemination. However, large cell carcinoma has a prognosis similar to that of adenocarcinoma, which is midway between squamow cell carcinoma and small cell carcinoma. On the other hand, adenoid cystic carcinomas of the
Address reprint requests to Teruyoshi Ishida, MD, Department of Surgery 11, Faculty of Medicine, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812, Japan
54
Ishida et al. TABLE I. Survival Rates of 33 Patients With Small Cell Carcinoma of the Lung
Group Stage
I I1 III Subtype Oat cell type Intermediate type Total
No. cases
Survival rates (%) 1 Yr
21 3 9
85
11 21
55 85
33
75
67 56
2 Yr
1*
4 Y'
5 Y'
1
79
1
65
57
1
57
33
J
22
22
1
11 J
46
24
62
55
24 55
24 44
58
46
42
38
o *
J
3 Y'
*Difference comparing stage I with stage I11 (P < .05).
MATERIALS AND METHODS
lung are relatively rare and develop principally in the trachea and large bronchi [ 141. The tumor seems to grow slowly and to have a high rate of local invasiveness and a low potentiality for distant metastasis [ 151. Although this tumor can be treated surgically, the invasiveness presents surgical limitations [ 161. We report here on the prognostic implications of clinical staging, histological subtyping, and histochemical and immunohistochemical tumor markers for patients with small cell carcinoma, large cell carcinoma, and adenoid cystic carcinoma of the lung. The 5-yr follow-up data on patients who underwent surgical treatment are also described.
From October 1974 through March 31, 1986, 92 Japanese patients with small cell carcinoma, large cell carcinoma, and adenoid cystic carcinoma of the lung were surgically treated in the Department of Surgery 11, Faculty of Medicine, Kyushu University, and the Department of Chest Surgery, National Kyushu Cancer Center Hospital. The stage of the disease was classified according to the TNM classification of the UICC [17], including the review of surgical and pathologic reports on the resected specimens. Thirty-three patients had small cell carcinoma, 54 large cell carcinoma, and five adenoid
Fig. 1. A: Oat cell type (HE, x420). The cells are small, round, hyperchromatic, loosely growing without any characteristic architec-
ture. B: Intermediate type (HE, X420). The cells are larger, spindled, and with a sparse cytoplasm.
Pathological Studies of Lung Cancer
55
TABLE 11. Immunohistochemistryof Small Cell Carcinoma of the Lung Subtype
Grimelius
Oat cell type
411 1 (36%)
Intermediate type
Total
NSE
Keratin
EMA
CEA
1*
911 1 (82%)
311 1 (27%)
3/11 (27%)
1/10 (lo%)]
1/20 J (5%) 513 1 (16%)
12/20 (60%) 21/31 (68%)
12/20 (60%) 15/31 (48%)
1 1/20 (55%) 1413 1 (45%)
13/20 (65%) 14/30 (47%)
*
NSE: neuron specific enolase, EMA: epithelial membrane antigen, CEA: carcinoembryonic antigen. *Difference between two groups of subtype (P < .05).
cystic carcinoma. The patients included 73 men and 19 women. The age at operation varied from 31 to 83 years with a mean of 62 yr. The extent of surgical excision was lobectomy in 77, pneumonectomy in eight, partial resection in three, and exploratory thoracotomy in three. The resected specimens were fixed in 10% formalin prior to embedding in paraffin. For histological studies, the sections were stained with hematoxylin and eosin (HE), Grimelius for argyrophilia, alcian-blue, and periodic acid-Schiff (PAS). All tumors were reviewed and subdivided by the criteria of the current WHO classification [ 11. For the immunohistochemical studies, tumors were stained with rabbit anti-neuron-specific enolase
(NSE) serum, mouse anti-keratin serum, mouse antiepithelial membrane antigen (EMA) serum, rabbit anticarcinoembryonic antigen (CEA) serum, rabbit antisecretory component (SC) serum, rabbit anti-lactofemn (LF) serum, rabbit anti-desmin serum, and rabbit antis-100 protein (S-100) serum, using an immunoperoxidase technique of the avidin-biotin-complex method [ 181. The mouse anti-keratin antibody was purchased from Imrnunotech (Marseille, France), and all other antibodies were obtained from Dakopatts (Glostrup, Denmark). The process of immunohistostaining was as follows: the deparaffinized sections were incubated in 0.3% H202 in methanol for 30 min to eliminate the en-
Fig. 2. A: Grimelius staining for argyrophilia (X560). Black granules are present in the cytoplasm of many cells. B: Immunohistochem-
istry for NSE (x450). Note the diffuse cytoplasmic staining of the tumor cells.
Fig. 3. Immunohistochemistry in small cell carcinoma, showing diffuse cytoplasmic and membrane staining for keratin (A, x450), EMA (BY X450) and CEA (C, X450).
Pathological Studies of Lung Cancer
57
TABLE 111. Survival Rates of 54 Patients With Large Cell Carcinoma of the Lung Survival rates (%)
No. cases Stage I
I1 111 IV
Cohesiveness Compact growth type Loose structure type Subtype Differentiated type Undifferentiated type Giant cell type Total
3 Yr
4 Yr
17 7 24 6
94 86 1 * 39 J 17
33 17
28 26
70 50
66 42
62 38
38
64 1
56
53
10 6
30 83
30 83
30 61
46 15 67
54
60
54
50
42
**
5 Y' 51
8 26 1,1
69 27 51 28
46 28
I* 1
39 15 67 37
*Difference comparing stage I with stage I11 (P < .05). **Difference comparing differentiated type with undifferentiated type (P < .05).
dogenous peroxidase activity. When using antikeratin antibody, the deparaffinized sections were treated with 0.001% trypsin in phosphate-buffered saline (PBS) (pH 7.4) for 30 min at 37°C prior to blocking the endogenous peroxidase. After being washed in PBS and incubating with normal goat serum, each section was incubated with primary antibodies at a dilution of 1:lOO for 30 min at
37°C. The sections were then exposed to biotinylated secondary antibody and avidin with biotinylated horseradish peroxidase (Vector Laboratories, Burlingame, CA) for 30 and 60 min, respectively. The peroxidase reaction was performed by a 5-min incubation in 0.01% H202 and 0.05% diaminobenzidine. Counter-staining was done with methyl green. Staining reactions were
Fig. 4. A: Compact growth type (HE, X420). Large polygonal tumor cells have proliferated in a solid nest and thin fibrovascular septa are present. Note the intercellular cohesion. B: Loose structure type
(HE, x420). The tumor cells are diffusely proliferated with loss of intercellular cohesion.
58
Ishida et al.
TABLE IV. Immunohistochemistry of Large Cell Carcinoma of the Lung Subtype
Keratin
Compact growth type
22/26 (79%)
Loose structure type
9/23 (35%)
EMA
1* ’
25/28 (89%) 9/24 (35%)
31/49 (57%)
Total
1*
’
34/52 (63%)
sc
LF
12/28 (43%)
9/27 (32%)
2 1I28 (75%)
4126 (15%)
2/22 (7%)
5/24 (19%)
16/52 (30%)
11/49 (20%)
2615 2 (48%)
CEA
1*
EMA: epithelial membrane antigen, SC: secretory component, LF: lactofenin, CEA: carcinoembryonic antigen. *Difference between two groups of subtype ( P < .05).
evaluated as positive or negative, and a positive reaction was considered only when strong brown deposits were visible. The survival rate was calculated by the Kaplan-Meier life-table method [ 191. Survival comparisons were tested by the Z value and generalized Wilcoxon test [ZO]. The chi-square test was also used to test correlations between detectability of tumor markers and the subtype of the carcinomas. The differences were considered to be statistically significant when the P value was less than .05.
RESULTS Small Cell Carcinoma Of 33 patients with small cell carcinoma, there were 21 patients with stage I disease, three with stage I1 and nine with stage 111, and the survival rates of the patients are shown in Table I. The overall 5-yr survival rate was 38%, and the survival rate was more favorable in those in the early stage. Using the morphologic criteria of WHO, the patients were divided into two subtypes: 11 patients with oat cell type (Fig. 1A) and 21 with intermediate type (Fig. 1B). However, one patient was unclassified. The 5-yr survival rate in those with the intermediate type was 44%, and that with the oat cell type was 24%. Survival rates in those with the intermediate type tended to be bener than in those with the oat cell type, albeit with no statistical significance. Histochemical and immunohistochemical studies were done to detect argyrophilic granule, NSE, keratin, EMA, and CEA, and the number of positive cases is shown in
Table 11. The incidences of argyrophilic cells (Fig. 2A) and/or NSE-positive cells (Fig. 2B) were greater in the cases of oat cell type than that of the intermediate type, but the difference was significant only in conjunction with the detection of argyrophilic cells ( P < .05).In contrast, the incidences of keratin, EMA, and/or CEApositive cells (Fig. 3) were greater in the cases of intermediate than of the oat cell type, but the difference was significant only along with the detection of CEA-positive cells (P < .05).Patients with keratin and EMA usually survived longer than did those with argyrophilia, NSE, and CEA, but the statistical difference was not significant.
Large Cell Carcinoma Of 54 patients with large cell carcinoma, there were 17 patients with stage I, seven with stage 11, 24 with stage 111, and six with stage IV, and the survival rates of patients were shown in Table 111. Using the morphologic criteria of WHO, large cell carcinomas were composed of large-sized and polygonal tumor cells with no evidence of maturation toward squamous cell carcinoma and/or adenocarcinoma. The patients were divided into two subtypes according to the existence of intercellular cohesion: 28 patients with compact growth type (Fig. 4A) and 26 with loose structure type (Fig. 4B). The 5-yr survival rates of the patients with compact growth type and loose structure type were 46 and 28%, respectively (P < .05).Immunohistochemical studies were done to detect keratin, EMA, SC, LF, and CEA, as shown in Table IV. The incidences of ker-
TABLE V. Treatment and Prognosis of Adenoid Cystic Carcinoma Case 1 2 3 4 5
Location Distal trachea Lt. main bronchus Rt. tr. intermedius Lt. lower lobe br. Lt. segmental br.
br.: bronchus; lob.: lobectomy.
Size (mm) 15 X 10 10 x 10 9 X 8 65 X 72 29 X 32
20 X 20 x 17X 95 X 36 X
Treatments Resection + Radiation Lt. upper sleeve lob. Rt. middle and lower lob. Lt. pneumonectmy Lt. upper lobectomy
Results No recurrence (3 yr, No recurrence (4 yr, No recurrence (1 yr, No recurrence (2 yr, No recurrence (8 yr,
1 mo) 3 mo) 6 mo) 9 mo) 2 mo)
Pathological Studies of Lung Cancer
atin-, EMA-, and CEA-positive cells were significantly greater in cases of compact growth type than of the loose structure type (P < .05). The patients were further separated into another three subtypes according to differentiation: the differentiated type characterized by immunoreactivities of each of the markers of keratin, EMA, SC, LF, or CEA, the undifferentiated type characterized by non-immunoreactivity of those markers, and the giant cell type characterized by highly pleomorphic multinucleated cells, with HE stain. The differences were significant between the 39% survival rate in those with the differentiated type and 15% in those with the undifferentiated type ( P < .05). The survival rate was poor in those with the loose structure type as well as in those with the undifferentiated type.
59
S-100 protein were positive in all lesions with the solid pattern.
DISCUSSION The operative management of small cell carcinoma has remained controversial. Following the demonstration by the British Medical Research Council in 1966 that radiotherapy was superior to surgical therapy, surgical treatment for small cell carcinoma has been out of favor and generally the establishment of the diagnosis of small cell carcinoma indicated that the patient is inoperable, on that basis alone [21]. On the other hand, clinical investigation by the Veterans Administration Surgical Oncology Group in 1982 showed the overall 5 yr survival rate to be 23% for patients given surgical treatment of small cell carcinoma and that the survival time is related to the Adenoid Cystic Carcinoma TNM staging system [22,23]. Other researchers have The clinical data of five patients with adenoid cystic reported that there were differences in the natural behavcarcinoma are shown in Table V. The locations of the ior or response to therapy of the various subtypes. Histumors were in the trachea, main bronchi, lobar bronchi, tologic subtyping of small cell carcinoma showed progor segmental bronchi. None was presented as a periph- nostic significance, oat cell type behaving more favoreral pulmonary lesion. All patients have been closely ably than the intermediate type in patients with extensive followed for 1.5-8 yr. disease and on chemotherapy [24]. The presence of ulHistologic features of the five cases of adenoid cystic trastructural features of epithelial differentiation correcarcinoma are summarized in Table VI. The cribriform lated in a highly significant manner with localized dispattern was evident in all five cases (Fig. 5A); however, ease at presentation, operability, and long-term survival the tubular pattern was identified in only three lesions after resection [3]. Our present study showed a 5 yr survival rate of 38% (Fig. 5B), and the solid pattern, in four lesions (Fig. 5C). Lesions with the tubular pattern were located in the more in patients with resected small cell carcinoma of the proximal area of the tracheobronchial tree, whereas le- lung. The survival rate in patients with stage I disease sions with the solid pattern were located in the distal was 57% at 5 yr. The survival rate of the intermediate tracheobronchial tree. Perineural invasion was evident in type seems to be better than that of the oat cell type. On two lesions with the solid pattern. Immunohistochemical the other hand, the additional diagnostics of histochemstudies were done to detect SC, LF, EMA, desmin, and istry and immunohistochemistry were used in the hope of S-100 protein and the findings are shown in Table VII. obtaining a more consistent subtype of small cell carciOf the five lesions with the cribriform pattern, SC and LF noma. Argyrophilic granule [25] and NSE [ 7 ] , as histowere identified in cells lining the gland-like lumen of all logical markers of neurosecretory differentiation, are delesions (Fig. 6A, B). Desmin and S-100 protein were tectable in small cell carcinomas of both oat cell and positive in the non-lining cells of four lesions (Fig. 6C, intermediate types. EMA [26] and keratin and CEA [27], D). Of the three lesions with the tubular pattern, SC and as histological markers of epithelial differentiation, have LF were positive, and of the three lesions with the solid also been seen in tumors of both subtypes. The neurosepattern, SC, LF, and EMA were negative. Desmin and cretory markers are frequently demonstrated in the oat
TABLE VI. Histological Features of Adenoid Cystic Carcinoma Growth patterns Case 1
2 3 4 5
Tubular
Cribiform
Solid
Perineural invasion
+ ++ +
++
-
-
-
+ + ++ ++
+
++ + +
-
+
+
-
+ + : more than 50% of the resected specimen; + : focal areas of the resected specimen.
Fig. 5 . A: Cribriform pattern of adenoid cystic carcinoma, which is composed of the two-cell pattern, showing epithelial cells in the inner layer and myoepithelial cells in the outer layer (HE, X250). B: Tubular pattern of adenoid cystic carcinoma, showing the most differentiated form. The tubular nests tend to be smaller with a single lumen and fibrous stroma (HE, x250). C: Solid pattern of adenoid cystic carcinoma, showing the most poorly differentiated form. The solid nests are filled with small cells and a few lumen. Perineural invasion is prominent (HE, X250).
Pathological Studies of Lung Cancer
cell type and epithelial markers in the intermediate type. Thus, the presence of these markers does not reflect the survival time in those with a small cell carcinoma, but small cell carcinoma is suggested to derive from a primitive multipotential stem cell differentiating into epithelial or neuroendocrine origin carcinoma. Large cell carcinomas are microscopically defined as solid anaplastic carcinomas without specifically differentiating features of squamous cell, small cell, or adenocarcinoma [2]. However, there are difficulties in accurately diagnosing these diseases, particularly when only hematoxylin and eosin stains are used [28]. Some large cell carcinomas show ultrastructural evidence of squamous cell carcinomas and/or adenocarcinoma [4]. Patients with ultrastructural evidence of squamous differentiation had a better prognosis than did those with adenocarcinomatous differentiation [5]. However, other investigators found no correlation between the ultrastructural evidence of differentiation and the prognosis [2]. The objective of our study was to analyze the prognostic implications of histological and immunohistochemical subtyping of large cell carcinoma of the lung. The survival of patients with the compact growth type was better than that of the loose structure type, as based on histological findings. Immunohistochemically, the epithelial markers were more frequently demonstrated in the compact growth type than in the loose structure type. The prognosis of patients with the differentiated type, as determined using such epithelial markers, was more favorable than for those with the undifferentiated type. Giant cell carcinoma, a variant of large cell carcinoma, tends to occur in younger patients and the clinical course is poor and often fatal [29]. In our series, a good survival time may be related to the small number of cases and early stages of the diseases. Adenoid cystic carcinoma is the most common of the salivary gland-like tumors arising from the trachea and bronchi [ 141. The infiltrative nature of the tumor requires a more radical resection. If the tumor shows markedly infiltrative growth at the surgical margin, radiation therapy will be palliative for the residual tumor [30]. Attempts have been made to correlate the different clinicopathologic aspects of adenoid cystic carcinoma with the biologic behavior [311. Growth patterns and vascular
and/or perineural involvement have been implicated with the prognosis of adenoid cystic carcinoma. All three patterns of growth, i.e., tubular, cribriform, and solid, showed an infiltrating growth to the adjacent structures. As compared with the cribriform and the tubular pattern, the solid pattern was the most aggressive. In fact, our series showed that the histologic evidence of perineural invasions was the most frequent in areas of the solid pattern of growth. The solid pattern may carry a worse prognosis in terms of development of distant metastasis and overall survival. Many investigators consider that adenoid cystic carcinoma is primarily a proliferation of both myoepithelial cells and secretory epithelial cells [ 6 ] ,and the acini and collecting ducts of bronchial gland may be the origin of the adenoid cystic carcinoma [32]. On the other hand, adenoid cystic carcinoma with myoepithelial features has been reported to be more resistant against irradiation than that having secretory features [32]. These findings suggest that the recurrence resulted from a proliferation of the remaining malignant cells with myoepithelial features. Immunohistochemically, SC, FL, and EMA were common in cells lining the cystic spaces and contained tissues of the tubular and cribriform patterns, but were rare in the solid pattern. On the other hand, desmin, an intermediate-sized filament of the muscle type [33], is used as a marker of muscle differentiation. S- 100 protein is not only distributed in Schwann cell tumors but also in myoepithelial cells of breast tumors and mixed tumors of the salivary gland [34]. In our current series, immunoreactivities of desmin and S-100 were found in the nonlining cells in cases of the tubular and cribriform patterns. The solid pattern of adenoid cystic carcinoma was positive for desmin and S-100 protein but negative for SC, LF, and EMA. These findings suggest that the solid pattern of adenoid cystic carcinoma, the most aggressive histologically, might originate from the myoepithelial cells, and, at the early stage of development, myoepithelial cells predominantly proliferate. We propose that some small cell carcinomas, large cell carcinomas, and adenoid cystic carcinomas may have some differentiating features if examined histologically and immunohistochemically . The subtyping of
TABLE VII. Immunohistochemistry of Adenoid Cystic Carcinoma Growth Dattern Tubular (n = 3 ) Cribriform (n = 5) Solid (n = 4)
61
sc
LF
EMA
Desmin
s-100
3 5 0
3
5
2 3
3 5
0
0
4
3 5 4
SC: secretory component; LF: lactofemn; EMA: epithelial membrane antigen; S-100: S-100 protein.
62
Ishida et al.
Fig. 6. Immunoperoxidase micrograph of adenoid cystic carcinoma. Many tumor cells in the cells lining the cystic spaces of the nests show a positive reaction of secretory component (A, X450) and lactofemn
(B, X450). Small tumor cells arranged in the nests show positive reaction of desrnin (C, x450) and S-100 protein (D, x450).
such carcinomas may reveal a different biologic behavior; hence, the prognosis and treatment can be better assessed.
Kyushu Cancer Center Hospital, Fukuoka, for providing valuable information and surgical specimens.
ACKNOWLEDGMENTS The authors are grateful to Drs. Nobuyuki Hara and Mitsuo Ohta, Department of Chest Surgery, National
REFERENCES 1. The World Health Organization Histological Typing of Lung Tumors (WHO), 2nd edition. Am J Clin Pathol 77:12>136, 1982. 2. Albain KS, True LD, Golomb HM, Hoffman PC, Little AG: Large cell carcinoma-Ultrastructural differentiation and clinicopathologic correlations. Cancer 56: 1618-1623, 1985.
Pathological Studies of Lung Cancer 3. Iglehart JD, Wolfe WG, Vernon WB, et al.: Electron microscopy in selection of patients with small cell carcinoma of the lung for medical versus surgical therapy. J Thorac. Cardiovasc. Surg. 90:35 1-360, 1985. 4. Churg A: The fine structure of large cell undifferentiated carcinoma of the lung. Hum Pathol 9:143-156, 1978. 5. Horie A, Ohta M: Ultrastructural features of large cell carcinoma of the lung with reference to the prognosis of patients. Hum Pathol 12:423-432, 1981. 6. Chaudhry AP, Leifer C, Cutler LS, et al.: Histogenesis of adenoid cystic carcinoma of the salivary glands, light and electromicroscopic study. Cancer 58:72-82, 1986. 7. Said JW, Vimadalal S, Nash G, et al.: Immunoreactive neuronspecific enolase, bombesin, and chromogranin as markers for neuroendocrine lung tumors. Hum Pathol 16:236-240, 1985. 8. Delmomte VC, Alberti 0, Saldiva PHN: Large cell carcinoma of the lung-Ultrastructural and immunohistochemical features. Chest 90:52&527, 1986. 9. Ishida T, Nishino T, Oka T, et al.: Surgical treatment of patients with small cell carcinoma of the lung. A histochemical and immunohistochemical study. J Surg Oncol40: 188-193, 1989. 10. Ishida T, Kaneko S, Tateishi M, et al.: Large cell carcinoma of the lung. Prognostic implications of histopathological and immunohistochemical subtyping. Am J Clin Path01 (in press). 11. Ishida T, Nishino T, Oka T, et al.: Adenoid cystic carcinoma of the tracheobronchial tree. Clinicopathology and immunohistochemistry. J Surg Oncol41:52-59, 1989. 12. Carney DN, Gazdar AF, Nan M, Minna JD: Recent advance in the biology of small cell lung cancer. Part 1., Establishment, classification, growth characteristic and oncogene analysis of small cell lung cancer cell line. Second Workshop on Small Cell Lung Cancer, Gleneagles, September 1984. 13. Mountain CF: Assessment of the role of surgery for control of lung cancer. Ann Thorac Surg 24:365-373, 1977. 14. Markel SF, Abell MR, Haight C, French AJ: Neoplasms of bronchus commonly designated as adenomas. Cancer 17:590-608, 1964. 15. Payne WS, Ellis FH Jr, Woolner LB, Moersch HJ: The surgical treatment of cylindroma (adenoid cystic carcinoma) and mucoepidermoid tumors of the bronchus. J Thorac Cardiovasc Surg 38:709-726, 1959. 16. Boyd AD, Soencer FC, Lind AL: Why has bronchial resection and anastomosis been reported infrequently for treatment of bronchial adenoma? J Thorac Cardiovasc Surg 59:359-365, 1970. 17. TNM Classification of Malignant Tumor, International Union Against Cancer (UICC), 3rd edition, Geneva, 1978. Enlarged and Revised, 1982, pp 4 1 4 5 .
63
18. Hsu SM, Raine L, Fanger H: The use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: A comparison between abc and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29:577-580, 1981. 19. Kaplan EL, Meier P: Nonparametric estimation from incomplete observation. J Am Stat Assoc 53:457481, 1958. 20. Gehan E: A generalized Wilcoxon test for comparing arbitrarily single-censored samples. Biometrika 52203-224, 1965. 21. Fox W, Scadding JG: Medical Research Council comparative trial of surgery and radiotherapy for primary treatment of small-celled or oat-celled carcinoma of bronchus. Lancet 2:979-986, 1966. 22. Shields TW, Higgins GA, Matthews MJ, Keehn, RJ: Surgical resection in the management of small cell carcinoma of the lung. J Thorac Cardiovasc Surg 84:481488, 1982. 23. Meyer JA: Indication for surgical treatment in small cell carcinoma of the lung. Surg Clin North Am 69:1103-1115, 1987. 24. Davis S, Stanley KE, Yesner R, et al.: Small-cell carcinoma of the lung-According to histologic subtype. Cancer 47: 18631866, 1981. 25. Grimelius L: A silver stain for a2 cells in human pancreatic islets. Acta SOCMed Uppsala 73:243-270, 1968. 26. Sloane JP, Ormerod MG: Distribution of epithelial membrane antigen in normal and neoplastic tissues and its value in diagnostic tumor pathology. Cancer 47:1786-1795, 1981. 27. Said JW, Nach G, Tepper G, Banks-Schlegel S: Keratin proteins and carcinoembryonic antigen in lung carcinoma. Hum Pathol 147C76, 1983. 28. Yesner R: Large cell carcinoma of the lung. Semin Diagn Pathol 2:255-269, 1985. 29. Flanagan CP, Roeckel IE: Giant cell carcinoma of the lung. Anatomical and clinical correlation. Am J Med 36:214-224, 1964. 30. Vieta JO, Maier HC: The treatment of adenoid cystic carcinoma (cylindroma) of the respiratory tract by surgery and irradiation therapy. Dis CheTt 31:493-511, 1957. 31. Nascimento AG, Amaral ALP, Prado LAF, et al.: Adenoid cystic carcinoma of salivary glands, a study of 61 cases with clinicopathologic correlation. Cancer 57:312-319, 1986. 32. Hoshino M, Yamamoto I: Ultrastructure of adenoid cystic carcinoma. Cancer 25:186198, 1970. 33. Gabbiani G, Kapanci Y, Barazzone P, Franke WW: Immunochemical identification of intermediate-sized filaments in human neoplastic cells, a diagnostic aid for the surgical pathologist. Am J Path01 104:206-216, 1981. 34. Nakajima T, Kameya T, Watanabe S, et al.: An immunoperoxidase study of S-100 protein distribution in normal and neoplastic tissues. Am J Surg Path01 6:715-727, 1982.
Clinical Applications of the Pathological Properties of Small Cell Carcinoma, Large Cell Carcinoma, and Adenoid Cystic Carcinoma of the Lung TERUYOSHI ISHIDA,MD, TOKUJIROYANO, MD, and KEIZO SUGIMACHI,MD From the Department of Surgery II, Faculty of Medicine, Kyushu University, Fukuoka, Japan
Lung cancer has considerable treatment problems, with a poor 5-yr survival rate after surgery. Application of histopathological and immunohistopathological subtyping have proven to be powerful tools for the assessment of prognosis. Results in 33 patients with small cell carcinoma, 44 with large cell carcinoma, and five with adenoid cystic carcinoma of the lung are discussed. The 5-yr survival rates in patients with small cell carcinoma of the oat cell type and intermediate type were 24 and 44%, respectively. Argyrophilic granules and neuron-specific enolase, neuroendocrine markers, were detected more frequently in the oat cell type rather than in the intermediate type. In contrast, keratin, epithelial membrane antigen, and carcinoembryonic antigen, epithelial origin markers, were present more frequently in the intermediae type than in the oat cell type. The 5-yr survival rates with large cell carcinoma of the compact growth type and the loose structure type were 46 and 28%, respectively, based on evidence of morphologically intercellular cohesion. The epithelial origin markers were detectable and more frequent in the compact growth type than in the loose structure type. The growth patterns of adenoid cystic carcinoma are classified as tubular, cribriform, and solid. The solid pattern was the most aggressive with extensive perineural invasion. We propose that the pathological properties of lung cancer should be examined as a prognostic implication of subtyping. KEYWORDS:lung cancer, histological subtyping, immunohistochemical markers
INTRODUCTION Primary lung cancer has been categorized into four main groups, according to the World Health Organization Histological Typing of Lung Tumors [ 11. These subgroups are squamous cell carcinoma, small cell carcinoma, adenocarcinoma, and large cell carcinoma. There are differences in both the biologic behavior and clinicopathologic manifestations among the four subgroups [2]. Although specific criteria have been given to the major cell types, criteria for the subgroups frequently have been either minimal or misleading. Electron microscopy [3-6], immunohistochemistry [7-1 l], and basic experimental research techniques [ 121 have profoundly altered and immensely improved our under-
0 1990 Wiley-Liss, Inc.
standing of the embryogenesis, histogenesis, and morphogenesis of all lung cancers. The 5-yr survival rate of patients with lung cancer is now reported to reach about 30%, in cases of surgical treatment [ 131. Both small cell carcinoma and large cell carcinoma show a fairly rapid growth and dissemination. However, large cell carcinoma has a prognosis similar to that of adenocarcinoma, which is midway between squamow cell carcinoma and small cell carcinoma. On the other hand, adenoid cystic carcinomas of the
Address reprint requests to Teruyoshi Ishida, MD, Department of Surgery 11, Faculty of Medicine, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812, Japan
54
Ishida et al. TABLE I. Survival Rates of 33 Patients With Small Cell Carcinoma of the Lung
Group Stage
I I1 III Subtype Oat cell type Intermediate type Total
No. cases
Survival rates (%) 1 Yr
21 3 9
85
11 21
55 85
33
75
67 56
2 Yr
1*
4 Y'
5 Y'
1
79
1
65
57
1
57
33
J
22
22
1
11 J
46
24
62
55
24 55
24 44
58
46
42
38
o *
J
3 Y'
*Difference comparing stage I with stage I11 (P < .05).
MATERIALS AND METHODS
lung are relatively rare and develop principally in the trachea and large bronchi [ 141. The tumor seems to grow slowly and to have a high rate of local invasiveness and a low potentiality for distant metastasis [ 151. Although this tumor can be treated surgically, the invasiveness presents surgical limitations [ 161. We report here on the prognostic implications of clinical staging, histological subtyping, and histochemical and immunohistochemical tumor markers for patients with small cell carcinoma, large cell carcinoma, and adenoid cystic carcinoma of the lung. The 5-yr follow-up data on patients who underwent surgical treatment are also described.
From October 1974 through March 31, 1986, 92 Japanese patients with small cell carcinoma, large cell carcinoma, and adenoid cystic carcinoma of the lung were surgically treated in the Department of Surgery 11, Faculty of Medicine, Kyushu University, and the Department of Chest Surgery, National Kyushu Cancer Center Hospital. The stage of the disease was classified according to the TNM classification of the UICC [17], including the review of surgical and pathologic reports on the resected specimens. Thirty-three patients had small cell carcinoma, 54 large cell carcinoma, and five adenoid
Fig. 1. A: Oat cell type (HE, x420). The cells are small, round, hyperchromatic, loosely growing without any characteristic architec-
ture. B: Intermediate type (HE, X420). The cells are larger, spindled, and with a sparse cytoplasm.
Pathological Studies of Lung Cancer
55
TABLE 11. Immunohistochemistryof Small Cell Carcinoma of the Lung Subtype
Grimelius
Oat cell type
411 1 (36%)
Intermediate type
Total
NSE
Keratin
EMA
CEA
1*
911 1 (82%)
311 1 (27%)
3/11 (27%)
1/10 (lo%)]
1/20 J (5%) 513 1 (16%)
12/20 (60%) 21/31 (68%)
12/20 (60%) 15/31 (48%)
1 1/20 (55%) 1413 1 (45%)
13/20 (65%) 14/30 (47%)
*
NSE: neuron specific enolase, EMA: epithelial membrane antigen, CEA: carcinoembryonic antigen. *Difference between two groups of subtype (P < .05).
cystic carcinoma. The patients included 73 men and 19 women. The age at operation varied from 31 to 83 years with a mean of 62 yr. The extent of surgical excision was lobectomy in 77, pneumonectomy in eight, partial resection in three, and exploratory thoracotomy in three. The resected specimens were fixed in 10% formalin prior to embedding in paraffin. For histological studies, the sections were stained with hematoxylin and eosin (HE), Grimelius for argyrophilia, alcian-blue, and periodic acid-Schiff (PAS). All tumors were reviewed and subdivided by the criteria of the current WHO classification [ 11. For the immunohistochemical studies, tumors were stained with rabbit anti-neuron-specific enolase
(NSE) serum, mouse anti-keratin serum, mouse antiepithelial membrane antigen (EMA) serum, rabbit anticarcinoembryonic antigen (CEA) serum, rabbit antisecretory component (SC) serum, rabbit anti-lactofemn (LF) serum, rabbit anti-desmin serum, and rabbit antis-100 protein (S-100) serum, using an immunoperoxidase technique of the avidin-biotin-complex method [ 181. The mouse anti-keratin antibody was purchased from Imrnunotech (Marseille, France), and all other antibodies were obtained from Dakopatts (Glostrup, Denmark). The process of immunohistostaining was as follows: the deparaffinized sections were incubated in 0.3% H202 in methanol for 30 min to eliminate the en-
Fig. 2. A: Grimelius staining for argyrophilia (X560). Black granules are present in the cytoplasm of many cells. B: Immunohistochem-
istry for NSE (x450). Note the diffuse cytoplasmic staining of the tumor cells.
Fig. 3. Immunohistochemistry in small cell carcinoma, showing diffuse cytoplasmic and membrane staining for keratin (A, x450), EMA (BY X450) and CEA (C, X450).
Pathological Studies of Lung Cancer
57
TABLE 111. Survival Rates of 54 Patients With Large Cell Carcinoma of the Lung Survival rates (%)
No. cases Stage I
I1 111 IV
Cohesiveness Compact growth type Loose structure type Subtype Differentiated type Undifferentiated type Giant cell type Total
3 Yr
4 Yr
17 7 24 6
94 86 1 * 39 J 17
33 17
28 26
70 50
66 42
62 38
38
64 1
56
53
10 6
30 83
30 83
30 61
46 15 67
54
60
54
50
42
**
5 Y' 51
8 26 1,1
69 27 51 28
46 28
I* 1
39 15 67 37
*Difference comparing stage I with stage I11 (P < .05). **Difference comparing differentiated type with undifferentiated type (P < .05).
dogenous peroxidase activity. When using antikeratin antibody, the deparaffinized sections were treated with 0.001% trypsin in phosphate-buffered saline (PBS) (pH 7.4) for 30 min at 37°C prior to blocking the endogenous peroxidase. After being washed in PBS and incubating with normal goat serum, each section was incubated with primary antibodies at a dilution of 1:lOO for 30 min at
37°C. The sections were then exposed to biotinylated secondary antibody and avidin with biotinylated horseradish peroxidase (Vector Laboratories, Burlingame, CA) for 30 and 60 min, respectively. The peroxidase reaction was performed by a 5-min incubation in 0.01% H202 and 0.05% diaminobenzidine. Counter-staining was done with methyl green. Staining reactions were
Fig. 4. A: Compact growth type (HE, X420). Large polygonal tumor cells have proliferated in a solid nest and thin fibrovascular septa are present. Note the intercellular cohesion. B: Loose structure type
(HE, x420). The tumor cells are diffusely proliferated with loss of intercellular cohesion.
58
Ishida et al.
TABLE IV. Immunohistochemistry of Large Cell Carcinoma of the Lung Subtype
Keratin
Compact growth type
22/26 (79%)
Loose structure type
9/23 (35%)
EMA
1* ’
25/28 (89%) 9/24 (35%)
31/49 (57%)
Total
1*
’
34/52 (63%)
sc
LF
12/28 (43%)
9/27 (32%)
2 1I28 (75%)
4126 (15%)
2/22 (7%)
5/24 (19%)
16/52 (30%)
11/49 (20%)
2615 2 (48%)
CEA
1*
EMA: epithelial membrane antigen, SC: secretory component, LF: lactofenin, CEA: carcinoembryonic antigen. *Difference between two groups of subtype ( P < .05).
evaluated as positive or negative, and a positive reaction was considered only when strong brown deposits were visible. The survival rate was calculated by the Kaplan-Meier life-table method [ 191. Survival comparisons were tested by the Z value and generalized Wilcoxon test [ZO]. The chi-square test was also used to test correlations between detectability of tumor markers and the subtype of the carcinomas. The differences were considered to be statistically significant when the P value was less than .05.
RESULTS Small Cell Carcinoma Of 33 patients with small cell carcinoma, there were 21 patients with stage I disease, three with stage I1 and nine with stage 111, and the survival rates of the patients are shown in Table I. The overall 5-yr survival rate was 38%, and the survival rate was more favorable in those in the early stage. Using the morphologic criteria of WHO, the patients were divided into two subtypes: 11 patients with oat cell type (Fig. 1A) and 21 with intermediate type (Fig. 1B). However, one patient was unclassified. The 5-yr survival rate in those with the intermediate type was 44%, and that with the oat cell type was 24%. Survival rates in those with the intermediate type tended to be bener than in those with the oat cell type, albeit with no statistical significance. Histochemical and immunohistochemical studies were done to detect argyrophilic granule, NSE, keratin, EMA, and CEA, and the number of positive cases is shown in
Table 11. The incidences of argyrophilic cells (Fig. 2A) and/or NSE-positive cells (Fig. 2B) were greater in the cases of oat cell type than that of the intermediate type, but the difference was significant only in conjunction with the detection of argyrophilic cells ( P < .05).In contrast, the incidences of keratin, EMA, and/or CEApositive cells (Fig. 3) were greater in the cases of intermediate than of the oat cell type, but the difference was significant only along with the detection of CEA-positive cells (P < .05).Patients with keratin and EMA usually survived longer than did those with argyrophilia, NSE, and CEA, but the statistical difference was not significant.
Large Cell Carcinoma Of 54 patients with large cell carcinoma, there were 17 patients with stage I, seven with stage 11, 24 with stage 111, and six with stage IV, and the survival rates of patients were shown in Table 111. Using the morphologic criteria of WHO, large cell carcinomas were composed of large-sized and polygonal tumor cells with no evidence of maturation toward squamous cell carcinoma and/or adenocarcinoma. The patients were divided into two subtypes according to the existence of intercellular cohesion: 28 patients with compact growth type (Fig. 4A) and 26 with loose structure type (Fig. 4B). The 5-yr survival rates of the patients with compact growth type and loose structure type were 46 and 28%, respectively (P < .05).Immunohistochemical studies were done to detect keratin, EMA, SC, LF, and CEA, as shown in Table IV. The incidences of ker-
TABLE V. Treatment and Prognosis of Adenoid Cystic Carcinoma Case 1 2 3 4 5
Location Distal trachea Lt. main bronchus Rt. tr. intermedius Lt. lower lobe br. Lt. segmental br.
br.: bronchus; lob.: lobectomy.
Size (mm) 15 X 10 10 x 10 9 X 8 65 X 72 29 X 32
20 X 20 x 17X 95 X 36 X
Treatments Resection + Radiation Lt. upper sleeve lob. Rt. middle and lower lob. Lt. pneumonectmy Lt. upper lobectomy
Results No recurrence (3 yr, No recurrence (4 yr, No recurrence (1 yr, No recurrence (2 yr, No recurrence (8 yr,
1 mo) 3 mo) 6 mo) 9 mo) 2 mo)
Pathological Studies of Lung Cancer
atin-, EMA-, and CEA-positive cells were significantly greater in cases of compact growth type than of the loose structure type (P < .05). The patients were further separated into another three subtypes according to differentiation: the differentiated type characterized by immunoreactivities of each of the markers of keratin, EMA, SC, LF, or CEA, the undifferentiated type characterized by non-immunoreactivity of those markers, and the giant cell type characterized by highly pleomorphic multinucleated cells, with HE stain. The differences were significant between the 39% survival rate in those with the differentiated type and 15% in those with the undifferentiated type ( P < .05). The survival rate was poor in those with the loose structure type as well as in those with the undifferentiated type.
59
S-100 protein were positive in all lesions with the solid pattern.
DISCUSSION The operative management of small cell carcinoma has remained controversial. Following the demonstration by the British Medical Research Council in 1966 that radiotherapy was superior to surgical therapy, surgical treatment for small cell carcinoma has been out of favor and generally the establishment of the diagnosis of small cell carcinoma indicated that the patient is inoperable, on that basis alone [21]. On the other hand, clinical investigation by the Veterans Administration Surgical Oncology Group in 1982 showed the overall 5 yr survival rate to be 23% for patients given surgical treatment of small cell carcinoma and that the survival time is related to the Adenoid Cystic Carcinoma TNM staging system [22,23]. Other researchers have The clinical data of five patients with adenoid cystic reported that there were differences in the natural behavcarcinoma are shown in Table V. The locations of the ior or response to therapy of the various subtypes. Histumors were in the trachea, main bronchi, lobar bronchi, tologic subtyping of small cell carcinoma showed progor segmental bronchi. None was presented as a periph- nostic significance, oat cell type behaving more favoreral pulmonary lesion. All patients have been closely ably than the intermediate type in patients with extensive followed for 1.5-8 yr. disease and on chemotherapy [24]. The presence of ulHistologic features of the five cases of adenoid cystic trastructural features of epithelial differentiation correcarcinoma are summarized in Table VI. The cribriform lated in a highly significant manner with localized dispattern was evident in all five cases (Fig. 5A); however, ease at presentation, operability, and long-term survival the tubular pattern was identified in only three lesions after resection [3]. Our present study showed a 5 yr survival rate of 38% (Fig. 5B), and the solid pattern, in four lesions (Fig. 5C). Lesions with the tubular pattern were located in the more in patients with resected small cell carcinoma of the proximal area of the tracheobronchial tree, whereas le- lung. The survival rate in patients with stage I disease sions with the solid pattern were located in the distal was 57% at 5 yr. The survival rate of the intermediate tracheobronchial tree. Perineural invasion was evident in type seems to be better than that of the oat cell type. On two lesions with the solid pattern. Immunohistochemical the other hand, the additional diagnostics of histochemstudies were done to detect SC, LF, EMA, desmin, and istry and immunohistochemistry were used in the hope of S-100 protein and the findings are shown in Table VII. obtaining a more consistent subtype of small cell carciOf the five lesions with the cribriform pattern, SC and LF noma. Argyrophilic granule [25] and NSE [ 7 ] , as histowere identified in cells lining the gland-like lumen of all logical markers of neurosecretory differentiation, are delesions (Fig. 6A, B). Desmin and S-100 protein were tectable in small cell carcinomas of both oat cell and positive in the non-lining cells of four lesions (Fig. 6C, intermediate types. EMA [26] and keratin and CEA [27], D). Of the three lesions with the tubular pattern, SC and as histological markers of epithelial differentiation, have LF were positive, and of the three lesions with the solid also been seen in tumors of both subtypes. The neurosepattern, SC, LF, and EMA were negative. Desmin and cretory markers are frequently demonstrated in the oat
TABLE VI. Histological Features of Adenoid Cystic Carcinoma Growth patterns Case 1
2 3 4 5
Tubular
Cribiform
Solid
Perineural invasion
+ ++ +
++
-
-
-
+ + ++ ++
+
++ + +
-
+
+
-
+ + : more than 50% of the resected specimen; + : focal areas of the resected specimen.
Fig. 5 . A: Cribriform pattern of adenoid cystic carcinoma, which is composed of the two-cell pattern, showing epithelial cells in the inner layer and myoepithelial cells in the outer layer (HE, X250). B: Tubular pattern of adenoid cystic carcinoma, showing the most differentiated form. The tubular nests tend to be smaller with a single lumen and fibrous stroma (HE, x250). C: Solid pattern of adenoid cystic carcinoma, showing the most poorly differentiated form. The solid nests are filled with small cells and a few lumen. Perineural invasion is prominent (HE, X250).
Pathological Studies of Lung Cancer
cell type and epithelial markers in the intermediate type. Thus, the presence of these markers does not reflect the survival time in those with a small cell carcinoma, but small cell carcinoma is suggested to derive from a primitive multipotential stem cell differentiating into epithelial or neuroendocrine origin carcinoma. Large cell carcinomas are microscopically defined as solid anaplastic carcinomas without specifically differentiating features of squamous cell, small cell, or adenocarcinoma [2]. However, there are difficulties in accurately diagnosing these diseases, particularly when only hematoxylin and eosin stains are used [28]. Some large cell carcinomas show ultrastructural evidence of squamous cell carcinomas and/or adenocarcinoma [4]. Patients with ultrastructural evidence of squamous differentiation had a better prognosis than did those with adenocarcinomatous differentiation [5]. However, other investigators found no correlation between the ultrastructural evidence of differentiation and the prognosis [2]. The objective of our study was to analyze the prognostic implications of histological and immunohistochemical subtyping of large cell carcinoma of the lung. The survival of patients with the compact growth type was better than that of the loose structure type, as based on histological findings. Immunohistochemically, the epithelial markers were more frequently demonstrated in the compact growth type than in the loose structure type. The prognosis of patients with the differentiated type, as determined using such epithelial markers, was more favorable than for those with the undifferentiated type. Giant cell carcinoma, a variant of large cell carcinoma, tends to occur in younger patients and the clinical course is poor and often fatal [29]. In our series, a good survival time may be related to the small number of cases and early stages of the diseases. Adenoid cystic carcinoma is the most common of the salivary gland-like tumors arising from the trachea and bronchi [ 141. The infiltrative nature of the tumor requires a more radical resection. If the tumor shows markedly infiltrative growth at the surgical margin, radiation therapy will be palliative for the residual tumor [30]. Attempts have been made to correlate the different clinicopathologic aspects of adenoid cystic carcinoma with the biologic behavior [311. Growth patterns and vascular
and/or perineural involvement have been implicated with the prognosis of adenoid cystic carcinoma. All three patterns of growth, i.e., tubular, cribriform, and solid, showed an infiltrating growth to the adjacent structures. As compared with the cribriform and the tubular pattern, the solid pattern was the most aggressive. In fact, our series showed that the histologic evidence of perineural invasions was the most frequent in areas of the solid pattern of growth. The solid pattern may carry a worse prognosis in terms of development of distant metastasis and overall survival. Many investigators consider that adenoid cystic carcinoma is primarily a proliferation of both myoepithelial cells and secretory epithelial cells [ 6 ] ,and the acini and collecting ducts of bronchial gland may be the origin of the adenoid cystic carcinoma [32]. On the other hand, adenoid cystic carcinoma with myoepithelial features has been reported to be more resistant against irradiation than that having secretory features [32]. These findings suggest that the recurrence resulted from a proliferation of the remaining malignant cells with myoepithelial features. Immunohistochemically, SC, FL, and EMA were common in cells lining the cystic spaces and contained tissues of the tubular and cribriform patterns, but were rare in the solid pattern. On the other hand, desmin, an intermediate-sized filament of the muscle type [33], is used as a marker of muscle differentiation. S- 100 protein is not only distributed in Schwann cell tumors but also in myoepithelial cells of breast tumors and mixed tumors of the salivary gland [34]. In our current series, immunoreactivities of desmin and S-100 were found in the nonlining cells in cases of the tubular and cribriform patterns. The solid pattern of adenoid cystic carcinoma was positive for desmin and S-100 protein but negative for SC, LF, and EMA. These findings suggest that the solid pattern of adenoid cystic carcinoma, the most aggressive histologically, might originate from the myoepithelial cells, and, at the early stage of development, myoepithelial cells predominantly proliferate. We propose that some small cell carcinomas, large cell carcinomas, and adenoid cystic carcinomas may have some differentiating features if examined histologically and immunohistochemically . The subtyping of
TABLE VII. Immunohistochemistry of Adenoid Cystic Carcinoma Growth Dattern Tubular (n = 3 ) Cribriform (n = 5) Solid (n = 4)
61
sc
LF
EMA
Desmin
s-100
3 5 0
3
5
2 3
3 5
0
0
4
3 5 4
SC: secretory component; LF: lactofemn; EMA: epithelial membrane antigen; S-100: S-100 protein.
62
Ishida et al.
Fig. 6. Immunoperoxidase micrograph of adenoid cystic carcinoma. Many tumor cells in the cells lining the cystic spaces of the nests show a positive reaction of secretory component (A, X450) and lactofemn
(B, X450). Small tumor cells arranged in the nests show positive reaction of desrnin (C, x450) and S-100 protein (D, x450).
such carcinomas may reveal a different biologic behavior; hence, the prognosis and treatment can be better assessed.
Kyushu Cancer Center Hospital, Fukuoka, for providing valuable information and surgical specimens.
ACKNOWLEDGMENTS The authors are grateful to Drs. Nobuyuki Hara and Mitsuo Ohta, Department of Chest Surgery, National
REFERENCES 1. The World Health Organization Histological Typing of Lung Tumors (WHO), 2nd edition. Am J Clin Pathol 77:12>136, 1982. 2. Albain KS, True LD, Golomb HM, Hoffman PC, Little AG: Large cell carcinoma-Ultrastructural differentiation and clinicopathologic correlations. Cancer 56: 1618-1623, 1985.
Pathological Studies of Lung Cancer 3. Iglehart JD, Wolfe WG, Vernon WB, et al.: Electron microscopy in selection of patients with small cell carcinoma of the lung for medical versus surgical therapy. J Thorac. Cardiovasc. Surg. 90:35 1-360, 1985. 4. Churg A: The fine structure of large cell undifferentiated carcinoma of the lung. Hum Pathol 9:143-156, 1978. 5. Horie A, Ohta M: Ultrastructural features of large cell carcinoma of the lung with reference to the prognosis of patients. Hum Pathol 12:423-432, 1981. 6. Chaudhry AP, Leifer C, Cutler LS, et al.: Histogenesis of adenoid cystic carcinoma of the salivary glands, light and electromicroscopic study. Cancer 58:72-82, 1986. 7. Said JW, Vimadalal S, Nash G, et al.: Immunoreactive neuronspecific enolase, bombesin, and chromogranin as markers for neuroendocrine lung tumors. Hum Pathol 16:236-240, 1985. 8. Delmomte VC, Alberti 0, Saldiva PHN: Large cell carcinoma of the lung-Ultrastructural and immunohistochemical features. Chest 90:52&527, 1986. 9. Ishida T, Nishino T, Oka T, et al.: Surgical treatment of patients with small cell carcinoma of the lung. A histochemical and immunohistochemical study. J Surg Oncol40: 188-193, 1989. 10. Ishida T, Kaneko S, Tateishi M, et al.: Large cell carcinoma of the lung. Prognostic implications of histopathological and immunohistochemical subtyping. Am J Clin Path01 (in press). 11. Ishida T, Nishino T, Oka T, et al.: Adenoid cystic carcinoma of the tracheobronchial tree. Clinicopathology and immunohistochemistry. J Surg Oncol41:52-59, 1989. 12. Carney DN, Gazdar AF, Nan M, Minna JD: Recent advance in the biology of small cell lung cancer. Part 1., Establishment, classification, growth characteristic and oncogene analysis of small cell lung cancer cell line. Second Workshop on Small Cell Lung Cancer, Gleneagles, September 1984. 13. Mountain CF: Assessment of the role of surgery for control of lung cancer. Ann Thorac Surg 24:365-373, 1977. 14. Markel SF, Abell MR, Haight C, French AJ: Neoplasms of bronchus commonly designated as adenomas. Cancer 17:590-608, 1964. 15. Payne WS, Ellis FH Jr, Woolner LB, Moersch HJ: The surgical treatment of cylindroma (adenoid cystic carcinoma) and mucoepidermoid tumors of the bronchus. J Thorac Cardiovasc Surg 38:709-726, 1959. 16. Boyd AD, Soencer FC, Lind AL: Why has bronchial resection and anastomosis been reported infrequently for treatment of bronchial adenoma? J Thorac Cardiovasc Surg 59:359-365, 1970. 17. TNM Classification of Malignant Tumor, International Union Against Cancer (UICC), 3rd edition, Geneva, 1978. Enlarged and Revised, 1982, pp 4 1 4 5 .
63
18. Hsu SM, Raine L, Fanger H: The use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: A comparison between abc and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29:577-580, 1981. 19. Kaplan EL, Meier P: Nonparametric estimation from incomplete observation. J Am Stat Assoc 53:457481, 1958. 20. Gehan E: A generalized Wilcoxon test for comparing arbitrarily single-censored samples. Biometrika 52203-224, 1965. 21. Fox W, Scadding JG: Medical Research Council comparative trial of surgery and radiotherapy for primary treatment of small-celled or oat-celled carcinoma of bronchus. Lancet 2:979-986, 1966. 22. Shields TW, Higgins GA, Matthews MJ, Keehn, RJ: Surgical resection in the management of small cell carcinoma of the lung. J Thorac Cardiovasc Surg 84:481488, 1982. 23. Meyer JA: Indication for surgical treatment in small cell carcinoma of the lung. Surg Clin North Am 69:1103-1115, 1987. 24. Davis S, Stanley KE, Yesner R, et al.: Small-cell carcinoma of the lung-According to histologic subtype. Cancer 47: 18631866, 1981. 25. Grimelius L: A silver stain for a2 cells in human pancreatic islets. Acta SOCMed Uppsala 73:243-270, 1968. 26. Sloane JP, Ormerod MG: Distribution of epithelial membrane antigen in normal and neoplastic tissues and its value in diagnostic tumor pathology. Cancer 47:1786-1795, 1981. 27. Said JW, Nach G, Tepper G, Banks-Schlegel S: Keratin proteins and carcinoembryonic antigen in lung carcinoma. Hum Pathol 147C76, 1983. 28. Yesner R: Large cell carcinoma of the lung. Semin Diagn Pathol 2:255-269, 1985. 29. Flanagan CP, Roeckel IE: Giant cell carcinoma of the lung. Anatomical and clinical correlation. Am J Med 36:214-224, 1964. 30. Vieta JO, Maier HC: The treatment of adenoid cystic carcinoma (cylindroma) of the respiratory tract by surgery and irradiation therapy. Dis CheTt 31:493-511, 1957. 31. Nascimento AG, Amaral ALP, Prado LAF, et al.: Adenoid cystic carcinoma of salivary glands, a study of 61 cases with clinicopathologic correlation. Cancer 57:312-319, 1986. 32. Hoshino M, Yamamoto I: Ultrastructure of adenoid cystic carcinoma. Cancer 25:186198, 1970. 33. Gabbiani G, Kapanci Y, Barazzone P, Franke WW: Immunochemical identification of intermediate-sized filaments in human neoplastic cells, a diagnostic aid for the surgical pathologist. Am J Path01 104:206-216, 1981. 34. Nakajima T, Kameya T, Watanabe S, et al.: An immunoperoxidase study of S-100 protein distribution in normal and neoplastic tissues. Am J Surg Path01 6:715-727, 1982.
