Annals of Oncology 2: 355-360, 1991. O 1991 Kluwer Academic Publishers. Printed in the Netherlands.
Original article Prognostic impact of histologic demonstration of Chromogranin A and Neuron Specific Enolase in pulmonary adenocarcinoma B. G. Skov,1 J. B. Serensen,2 F. R. Hirsch,3 L. I. Larsson4 & H. H. Hansen2 'Dept. of Pathology, Rigshospitalet, University of Copenhagen, Copenhagen; 2Dept. of Oncology, Rigshospitalet, Copenhagen; 3Dept. of Oncology, Rigshospitalet, and Dept. of Internal Medicine C, Bispebjerg Hospital, Copenhagen; 4Dept. of Molecular Cell Biology, Statens Seruminstitut, Copenhagen, Denmark
Summary. One hundred-fourteen patients with inoperable adenocarcinoma of the lung (ACL) were evaluated by immunohistochemistry with monoclonal antibodies against Neuron Specific Enolase (NSE) and Chromogranin A (Chr A) in order to determine the frequency and prognostic impact of such antigen expression. All patients were previously untreated and received chemotherapy according to a prospective randomized trial. The tumors of 18 patients (16%) had more than 10% positive cells stained with anti-NSE, 59 (52%) had 1-10% positive cells and those of 37 patients (32%) contained no NSE-positive cells. The corresponding figures for Chr A were: 22 patients (19%), 51 patients (45%) and 41 patients (36%), respectively. Forty-four per cent of the patients with more than 10% positive NSE cells responded to chemotherapy (either complete or partial remissions) compared to 17% of the patients with fewer than 10% positive cells (p < 0.025). The corresponding values for Chr A were 30% responders versus 19% responders (not statistically significant). The median survival for patients with more than 10%, 1-10% or no NSE-positive cells was 262 days, 231 days and 159 days, while, for Chr A it was 245 days, 200 days and 238 days, respectively. The survival curves for both NSE and Chr A according to the various levels of positivity were not significantly different. The presence of neuroendocrine marker in pulmonary adenocarcinoma seems to be associated with increased sensitivity to chemotherapy. Key words: pulmonary adenocarcinoma, diagnosis, histopathology, Neuron Specific Enolase, Chromogranin A
Introduction The histologjcal classification of lung cancer has great influence on the choice of therapy. Small-cell carcinoma of the lung (SCLC), frequently exhibiting neuroendocrine markers [1, 2], is more sensitive to chemotherapy than adeno-, large-cell- or squamous cell carcinoma. When chemotherapy is given to patients with inoperable adenocarcinoma of the lung (ACL), the response rates are usually reported to be around 25% [3, 4], as compared to 80-85% for SCLC. Histological diagnosis of ACL is traditionally carried out by using Haematoxylin-Eosin (HE) and mucin stains, according to WHO criteria [5]. However, the interest in application of immunohistochemistry to the study and categorization of lung cancers has increased during recent years [6, 7, 8,9]. Enolases are glycolytic enzymes necessary for the anaerobic catabolism of glucose. The enzymes are dimers, variously made up of alpha, beta or gamma subunits. The alpha dimer is widespread, while the beta dimer is mainly found in muscles, and the gamma dimer (NSE) mainly in neurons and neuroendocrine cells [10, 11]. The latter has accordingly been used as a
marker for tumours with neuroendocrine features, including SCLC [12, 13]. Several studies have also demonstrated NSE immunoreactivity in non-SCLC tumours [14, 15], and in tumours not included in the APUD system, e.g., breast carcinomas [16, 17], and the specificity of NSE for the diagnosis of SCLC is therefore questionable. Chromogranin A (Chr A) is an acidic glycoprotein of high molecular weight originally isolated from the adrenal medulla [18]. It has been reported to occur in neuroendocrine cells and neoplasms, including SCLC, and Chr A and dense core granules are expressed concordantly [19, 20]. It is released together with neuroendocrine peptides and amines through exocytosis from granules, but its physiological relevance is still unknown [19]. High plasma concentrations have been demonstrated in patients with SCLC, especially with extensive disease, whereas low plasma concentrations or concentrations below detection limits have been found in patients with other lung tumours, including ACL [21, 22]. Furthermore, Chr A has been demonstrated in cell lines of classic SCLC, whereas cell lines of non-SCLC do not contain Chr A [23]. Accordingly, it has been hypothesized that Chr A might be used as a diagnostic marker
356 for neuroendocrine rumours and as a quantitative indicator of disease activity in SCLC [21, 22]. In the present study, the frequency of immunohistochemical demonstration of NSE and Chr A was evaluated in tumours from a series of patients with inoperable ACL treated with cytostatic agents in order to learn whether such expression may presage an increased response rate to chemotherapy. Materials and methods
details (see ref. 27). A human fetal lung was used as positive control for NSE staining, and a human fetal jejunum was used for the Chr A staining. As negative controls the primary antibodies were omitted, but otherwise the same procedure was used. Previous studies [28] employing the present antibody to NSE and type-matched IgG on similarly fixed SCLC tissue, have demonstrated an absence of unspecific staining reaction and high grade of sensitivity with this monoclonal antibody. No specimens were available for serum analysis of NSE and Chr A, or for electron microscopy.
Patients Evaluation The study included patients up to 70 years old with ACL, stages Ilia, IHb or IV. The tumours were classified according to the criteria recommended by WHO [24]. All of the patients received cytostatic treatment in a prospective trial and were randomized to either vindesine as a single agent or lomustine plus cyclophosphamide plus methotrexate, or all four drugs [3]. Patients with measurable (bidimensionally measurable) as well as patients with evaluable (unidimensionally measurable) disease were included. Response assessment was performed according to the WHO criteria [25] and was verified by two observers. No differences were observed with respect to response rate or survival between the three treatment regimens [3]. Histological material The histologic specimens consisted of tissue from primary tumours obtained by bronchoscopy or thoracotomy. All specimens were taken prior to chemotherapy. The tumour tissues were fixed in 10% formaldehyde, embedded in paraffin, sectioned and stained with HE and mucin stains (PAS and Alcian blue) for conventional histologic examination, as described previously [26]. From each tumour 1-2 blocks were available and from each block 1—2 slides were selected containing sufficient tumour tissue. The slides were then stained according to the protocol. For immunohistochemistry, 5 micrometer sections were blocked for endogenous peroxidase activity by methanol H 2 O 2 treatment for 30 minutes and were then stained by the indirect immunoperoxidase method, using a monoclonal antibody to NSE (Sanbio, Holland), and a monoclonal antibody to Chr A (Hybritech Inc., USA) as primary antibodies, both incubated at a 1:100 dilution in BSA/TBS (0,25% Bovine Serum Albumin and 0.01 Molar Tris Buffer, pH = 7.4 containing 0.15 Molar Sodium Chloride) for 1 h. The site of antigen-antibody reaction was revealed by the indirect peroxidase method using peroxidase-labelled anti-mouse IgG (Dakopatt A/S Copenhagen, Denmark) as secondary antibody (1:20 for 1 hour). The peroxidase activity was detected by incubation in diaminobenzidine-hydrogen peroxidase medium for 10 minutes. For further methodological
The immunostaining was assessed by one observer (BGS). The number of stained cells in each section was scored and classification was as follows: neg: no staining of tumour cells, +: 75% of the tumour cells stained. The tumours containing more than 10% staining cells were so few in number that they were combined into one group to make statistical analysis feasible. Differences in staining intensities were not evaluated. The immunohistochemical evaluation was blinded for the clinical results. Comparison of response rates was based on the Chi-square test. The analysis of survival times was performed by the Kaplan-Meier method and the log rank test [29, 30]. Statistical significance was assumed when p was less than 0.05.
Results Of the 259 patients included in the study, the primary diagnosis for 39 was based on cytology alone, and these patients were excluded from the study. Five patients were considered not to have adenocarcinoma when the HE and mucin-stained sections were reviewed, and there was insufficient tissue available from 101 patients for immunohistochemistry examination. The remaining 114 patients with histologically verified ACL were included in the study. There was no significant difference in either rates of response to chemotherapy (0.3 < p < 0.4) or survival times (0.3 < p < 0.4) for the 114 patients included compared to the 145 patients not included in the study, nor was there any difference in the response rates or survival times between the three treatment regimens in the study [3]. Table 1 summarizes the staining results of the tumours with the two monoclonal antibodies. NSE was expressed in 77 tumours (69%), with more than 10% of the cells being positive in 18 tumours (16%). Corresponding values for Chr A were 73 tumours (64%) and 22 tumours (19%), respectively. The staining was mainly cytoplasmic, and the staining intensity varied within and between sections.
357 Table 1. NSE and Chr A in pulmonary adenocairinoma correlated with age, sex, performance status, extent of disease, and response rate. Characteristic
Chr A
NSE
All patients
Neg
+
++
Neg
+
++
N (%)
N(%)
N(%)
N(%)
N(%)
N(%)
N(%)
No of patients
37 (32)
59 (52)
18(16)
41 (36)
51 (45)
22 (19)
259(100)
Age in years Median Range
53 37-69
56 32-69
60 31-70
53 37-69
57 32-69
57 31-70
57 31-70
Sex Male Female
19(17) 18(16)
32 (28) 27 (24)
12(10) 6 ( 5)
23 (20) 18(16)
27 (24) 24(21)
14(12) 8 ( 7)
139 (54) 120 (46)
Kamowsky's performance 90-100% 70-80% 50-60%
14 12 11
25 25 9
8 8 2
18 12 11
21 21 9
9 11 2
88 115 56
8 10 19
19 15 25
2 6 10
10 14 17
15 10 26
4 7 11
65 71 123
Negative and +
++
Negative and +
++
17%
44%
19%
30%
Stage
ma mb IV
Response rate Neg. - all tumour cells negative,
+ — 0-10% positive tumour cells,
Forty-four per cent of the patients with tumours containing more than 10% NSE-positive tumour cells responded to therapy (partial or complete response), whereas only 17% of those with less than 10% positive cells responded (p < 0.025). Corresponding values for Chr A were 30% versus 19% responders (not statistically significant).
24%
- - > 10% positive tumour cells.
The median survival time for patients with NSEnegative tumours was 159 days versus 231 days for patients with 0-10% NSE-positive tumour cells and 262 days for those with more than 10% NSE-positive cells. The median survival time for patients with any NSE positivity was 236 days (not statistically significant, Fig. 1). Corresponding figures for Chr A were 238, 100
- -•- 0% NSE positive ceBs (n -37) — 1-10% NSE positive cels(n-59) >10% NSE positive cells (n-18)
\ L 80-
'Ul
...
0% Chr-A positive cels(n-41) 1-10% Chr-A positive cells(n-51) >10% Chr-A positive eels (n-22)
|60V)
-c
£40
htrs
20-
•-"In —i
30 Months
12
\
1
1
18
24
1—•
30 Months
Fig. I. Survival curves for NSE according to the 3 levels of Fig. 2 Survival curves for Chr A according to the 3 levels of posipositivity of the tumours. tivity of the tumours.
358
200, 245 and 210 days (not statistically significant, Fig. 2). There were no statistically significant differences in either response rates (13 versus 31%) or survival times (202 versus 261 days) for patients with tumours negative for both NSE and Chr A (31 patients) compared to patients who had tumours with more than 10% positive cells for both antibodies (16 patients). None of the patients had tumours with more than 10% NSE-positive cells and simultaneously no Chr A positive cells, and only one patient had a tumour negative for NSE but with more than 10% Chr A-positive cells. Discussion Studies of NSE in patients with ACL report varying numbers of positive tumour cells [8, 9]. Said et al., using 3 different antibodies to NSE, have demonstrated diffuse cytoplasmic staining in 2 of 9 adenocarcinomas, while Addis et al., using a polyclonal antibody, demonstrated NSE in 3 of 39 ACL [8, 9]. Other studies have demonstrated no positive tumour cells [12, 14]. Several factors may explain these inconsistent results. Firstly, different fixatives, antibodies and methods have been used. Paraffin-embedded material fixed in formalin and Zamboni's fixative have been used in several studies with inconsistent results [8, 9, 12, 14, 15]. Because many of the studies were based on small numbers of tumours, or different primary antibodies were used, it has been difficult to compare their results. Furthermore, the sources of the histological material used varied considerably. Biopsies from broncho- and mediastinoscopies and specimens from thoracoscopies and lung resections differ in size and quality. A study by Bergh et al. showed that 35% of small and/or squeezed or crushed biopsies with SCLC were positive for NSE compared to 83% of larger and better preserved biopsies [15]. Only a few previous studies have evaluated Chr A in lung carcinomas. In a study of 10 non-SCLC cell lines Chr A could not be demonstrated, whereas Chr A was immunocytochemically demonstrated in 11 of 18 cell lines of classical SCLC and with radioimmunoassay in 15 of 18 cell lines [23]. Wilson et al. [31] studied 10 formalin-fixed SCLC with monoclonal anti-Chr A and found 4 tumours with positive cells, whereas neither of the two squamous cell carcinomas were positive. Graziano et al. [32] demonstrated Chr A in 3 of 45 non-SCLC and Mooi et al. [33] found 3 of 11 nonSCLC lung tumours positive for Chr A. Again the differences in results may be due to the use of different antibodies. The antibodies recognize only part of the antigen, and different antibodies recognizing different epitopes may behave in different ways with respect to the fixation. However, lung tumours are often heterogeneous and may contain several distinct tumour cell types [34]. If
expression of NSE reflects neuroendocrine differentiation, as has been repeatedly proposed [10, 12], our investigation has detected such differentiation by immunocytochemistry in ACL. Demonstration of NSE in adenocarcinomas, which show no morphologic resemblance to SCLC, supports the concept that all lung tumours might originate from a stem cell type which is able to differentiate in several ways, including the development of neuroendocrine features - the Unitarian theory of development of lung carcinomas [35]. Irrespective of the biological significance of an immunohistochemical demonstration of NSE in ACL, the present study has demonstrated that the method can be used to detect a subgroup of ACL with high NSE reactivity and significantly better response to chemotherapy than tumours with less NSE reactivity. The initial number of patients in the present study was 259, but in about half of these the diagnosis had been based on cytologic material or biopsies from histologic material which was inadequate for immunocytochemical analysis. Therefore, very small and crushed biopsies were excluded from our material, and this may well have influenced the frequencies observed. To assess whether the 145 patients with material not eligible for immunocytochemical study constituted a selected group, we calculated their response rates and survival times but found no difference compared with the 114 patients included in the study. We therefore hypothesize that the finding in 114 patients is also true for the entire cohort. The clinical value of immunocytochemical demonstration of neuroendocrine features in non-SCLC has been the subject of three recent publications. Berendsen et al. [36] have prospectively studied the demonstration and prognostic impact on non-SCLC tumours of a series of MOC-antibodies reactive with SCLC-associated antigens, previously characterized as neuroendocrine-related differentiation antigen. In their study of 141 patients with non-SCLC tumours, they noted the presence of markers indicative of neuroendocrine differentiation in about 30% of the patients, most pronounced in adenocarcinomas. One of the antibodies, MOC-1, recognized a non-SCLC subgroup with worse prognosis in surgically treated patients. There was an equal distribution in neuoendocrine markers between metastatic lesions and primary tumours, indicating that neuroendocrine features of non-SCLC do not correlate with a propensity for early metastasis per se. Whether patients with neuroendocrine tumours have a different response to chemotherapy was not answered in the study by Berendsen. Linnoila et al. [37] described a group of non-SCLC patients who were treated with chemotherapy; they noted more responses in the patients in whom a neuroendocrine differentiation was recognized. However, the survival time was equal in both groups with or without neuroendocrine differentiation. In our study there was a tendency towards prolonged survival in the NSE-
359
positive group of patients with ACL. Finally, Graziano et al. [32] retrospectively reviewed 52 patients with non-SCLC tumours, including 26 who responded to chemotherapy and 26 without tumour response. They related the clinical outcome to the expression of neuroendocrine markers (NSE, Chr A and Leu 7). In the latter study, two markers were positive in 10/26 responders (38%) and in 0 of 26 patients without chemotherapy response. Responders with two or more positive markers had longer survivals than did responders with less than two positive markers. The authors emphasized, however, that the findings should be interpreted with caution due to the relatively small numbers of patients in the study and its retrospective nature. It seems justified to conclude from the present study and the few other reports in the literature that the presence of neuroendocrine markers in non-SCLC tumours might be associated with increased senstivity to chemotherapy. However, future larger prospective studies are needed to confirm the benefit of chemotherapy for this subgroup of non-SCLC patients. This study was presented at the 25th Annual Meeting of the American Society of Clinical Oncology, 1989 (abstract 857). Acknowledgement This work was supported by the Danish Cancer Society. References 1. Richardson RL, Greco FA, Oldham RK et al. Tumor products and potential tumour markers in small cell lung cancer. Semin Oncol 1978; 5: 253-62. 2. Carney DN, De Leij L. Lung cancer biology. Semin Oncol 1988; 15: 199-214. 3. Serensen JB, Hansen HH, Dombernowsky P et al. Chemotherapy for adenocarcinoma of the lung (WHO III): A randomized study of Vindesine versus Lomustine, Cyclophosphamide, and Methotrexate versus all four drugs. J Clin Oncol 1987; 5: 1169-77. 4. Osterlind K, Horbov S, Dombernowsky P et al. Vindesine in the treatment of squamous cell carcinoma, adenocarcinoma and large cell carcinoma of the lung. Cancer Treat Rep 1982; 66: 305-9. 5. World Health Organisation. Histological Typing of the Lung tumours. International Histological Classification of Tumours, No. 1, 2nd ed. Geneva: WHO. 1981. 6. Moss F, Bobrow L, Sheppard MN et al. Expression of epithelial and neural antigens in small-cell and non-small-cell lung carcinoma. Pathol 1986; 149: 103-11. 7. Mooi WJ, Wagenaar SS, Schol D et al. Monoclonal antibody 123C3 in lung tumour classification. Immunohistology of 356 resected tumours. Molecular and Cellular Probes 1988; 2: 31-7. 8. Said JV, Vimadalal S, Nash G. Immunoreactive Neuronspecific Enolase, Bombesin and Chromogranin as markers for neuroendocrine lung tumours. Hum Pathol 1985; 16: 236-40. 9. Addis BJ, Hamid Q, Ibrahim BN et al. Immunohistochemical markers of small cell carcinoma and related neuroendocrine tumours of the lung. J Pathol 1987; 153: 137-50.
10. Schmechel D, Marangos PJ, Brightman M. Neuron Specific Enolase is a molecular marker for peripheral and central neuroendocrine cells. Nature 1987; 276: 834-6. 11. Tapia FJ, Barbosa AJA, Marangos PJ. Neuron-specific enolase is produced by neuroendocrine tumours. Lancet 1981; i: 808-11. 12. Sheppard MN, Corrin B, Bennett MH et al. Immunohistochemical localization of neuron specific enolase in small-cell carcinoma and carcinoid tumours of the lung. Histopathology 198; 48:171-81. 13. Carney DN, Ihde DC, Cohen MH et al. Serum neuron specific enolase: a marker for disease extent and response to therapy of small-cell lung cancer. Lancet 1982; i: 583-5. 14. Dhillon AP, Rode J, Dhillon DP et al. Neural markers in carcinoma of the lung. Br. J. Cancer 1985; 51:645-52. 15. Bergh J, Esscher T, Steinholtz LM et al. Immunohistochemical demonstration of neuron-specific enolase (NSE) in human lung cancers. Am J Clin Pathol 1985; 84:1-7. 16. Nesland JM, Holm R. Neuron specific enolase immunostaining in the diagnosis of breast carcinomas with neuroendocrine differentiation. Its usefulness and limitations. J Pathol 1986; 148: 35-43. 17. Virones SA, Bonnin JM, Rubinstein LJ et al. Immunohistochemical demonstration of neuron-specific enolase in neoplasms of the CNS and other tissues. Arch Pathol Lab Med 1984; 108:536-40. 18. O'Conner DT, Frigon P, Sokoloff RL. Human chromogranin A: purification and characterization from catecholamine storage vesicles of human pheochromocytoma. Hypertension 1984; 6 2-4. 19. O'Conner DT. Widespread immunoreactivity in polypeptide hormone producing tissues and in serum. Regul Pept 1983; 6: 263-80. 20. Gazdar AF, Helman LJ, Israel MA et al. Expression of neuroendocrine cell markers L-dopa decarboxylase, chromogranin A, and dense core granules in human tumors of endocrine and nonendocrine origin. Cancer Res 1988; 48: 4078-82. 21. Sobel RE, O'Conner DT, Addison J et al. Elevated serum Chromogranin A in small-cell lung carcinoma. Ann Intern Med 1986; 105:698-700. 22. O'Conner DT, Deftos LJ. Secretion of Chromogranin A by peptide-producing endocrine neoplasma. N Engl J Med 1986; 314: 1145-51. 23. Deftos LJ, Linnoila RI, Carney DN et al. Demonstration of Chromogranin A in human neuroendocrine cell lines by immunohistology and immunoassay. Cancer 1988; 62: 92-7. 24. Mountain CF. A new international staging system for lung cancer. Chest 1986; 89 suppl: 225s-33s. 25. World Health Organisation. Handbook for Reporting results of Cancer Treatment. Geneva, World Health Organization. 1979. 26. Serensen JB, Hirsch FR. Olsen J. The prognostic implication of histopathologic subtyping of pulmonary adenocarcinoma according to the classification of the World Health Organization. Cancer 1988; 62: 361-7. 27. Larsson LI. Immunohistochemistry. Theory and Practice. CRC Press, Boca Raton, FL. 1988. 28. Jargensen LGM, Hirsch FR, Skov BG et al. Occurrence of Neuron Specific Enolase in tumour tissue and serum in smallcell lung cancer. Br J Cancer 1991; 63: 151-3. 29. Kaplan EL, Meier P. New parametric estimation from incomplete observation. J Am Statist Assoc 1958; 53:457-81. 30. Peto R, Pike MC, Armitage P et al. Design and analysis of randomized clinical trials requiring prolonged observation of each patient Br J Cancer 1977; 35:1-39. 31. Wilson BS, Lloyd RV. Detection of Chromogranin in neuroendocrine cells with a monoclonal antibody. Am J Pathol 1984; 115:458-68. 32. Graziano SL, Mazid R, Newman N et al. The use of neuroendocrine immunoperoxidase markers to predict chemo-
360
33.
33.
34. 35.
therapy response in patients with non-small-cell lung cancer. J Qin Oncol 1989; 7: 1389-1406. O'Conner DT, Burton D, Deftos LJ. Chromogranin A: immunohistology reveals its universal occurrence in normal polypeptide hormone producing endocrine glands. Life Sci 1983; 33: 1657-63. Mooi WJ, Dewar A, Springall D et al. Non-small-cell lung carcinomas with neuroendocrine features. A light microscopic, immunohistochemical and ultrastructural study of 11 cases. Histopathology 1988; 13: 329-37. Roggli JG, Vollmer RT, Greenberg SD et al. Lung Cancer heterogeneity. A blinded and randomized study of 100 consecutive cases. Hum Pathol 1985; 16: 569-79. Gazdar AF, Carney DN, Guccion JC et al. Small cell carcinoma of the lung: Cellular origin and relationships to other primary tumors. In Greco FA, Oldham RK and Bunn PA (eds.): Small cell lung cancer, Grune & Stratton, New York, 145-75, 1981.
Book review Surgical oncology, current concepts and practice.
C. S. McArdle (ed). Butterworths, London, Boston, Singapore, Sydney, Toronto, Wellington, 1990. 342 pp, £55.00 This is an excellent, easy-to-read, comprehensive review of the current therapeutic concepts and practice with respect to solid tumors. Our colleagues from the UX. are to be congratulated on this series of chapters, written by well-known experts in the field. The subjects reviewed include carcinoma of the oesophagus, gastric cancer, carcinoma of the pancreas and biliary system, colorectal cancer, endocrine tumors, breast cancer, kidney tumors, bladder cancer, prostate and testis cancer, head and neck tumors, malignant melanoma, bone and soft tissue tumors and malignant lymphoma. The two final chapters are dedicated to principles of cancer chemotherapy and to recent advances in cancer research. Despite the fact that each of the chapters is written by a different expert, all of them are outlined according to common guidelines, with many subtitles including epidemiology, etiology, pathogenesis, diagnosis, clinical features, surgical therapy and combined modality approaches. An extensive, up-to-date review
36. Berendsen H, Leij L de, Poppema S et al. Clinical characterization of non-small-cell lung cancer tumors showing neuroendocrine differentiation features. J Clin Oncol 1989; 7: 1614-20. 37. Linnoila RI, Jensen S, Steinberg S et al. Neuroendocrine differentiation correlates with favourable response to chemotherapy in patients with non-small-cell lung cancer. Fifth World Conference on Lung Cancer (abstr) 1988; 3.08. Received 19 November 1990; accepted 9 January 1991. Correspondence to: Fred R. Hirsch, MD Department of Oncology Rigshospitalet Blegdamsvej 9 2100 Copenhagen, Denmark
Annals of Oncology 2: 360, 1991.
of the most relevant literature concludes each chapter. The ones on cancer of the oesophagus, colorectal cancer and breast tumors are of outstanding value, whereas little information is given on the current muJtidisciplinary approach to head and neck tumors and bone/ soft tissue tumors. Unfortunately, there are no chapters on lung cancer, the most common solid rumor in Western countries, or anal carcinoma, a paradigm of multimodality therapy. In addition to the well-written final chapters on medical therapy the Editor should have included a similar one on the principles and concepts of radiotherapy (including radiobiology, radiation beam sources, field planning, dose fractionation). Some of these aspects are important for the operating surgeon. Despite these minor flaws, I greatly enjoyed reading this book. It has great potential value for the general surgeon and for the non-surgical oncologist interested in this field, and should be placed in all surgical libraries. U. Metzger Zurich
Original article Prognostic impact of histologic demonstration of Chromogranin A and Neuron Specific Enolase in pulmonary adenocarcinoma B. G. Skov,1 J. B. Serensen,2 F. R. Hirsch,3 L. I. Larsson4 & H. H. Hansen2 'Dept. of Pathology, Rigshospitalet, University of Copenhagen, Copenhagen; 2Dept. of Oncology, Rigshospitalet, Copenhagen; 3Dept. of Oncology, Rigshospitalet, and Dept. of Internal Medicine C, Bispebjerg Hospital, Copenhagen; 4Dept. of Molecular Cell Biology, Statens Seruminstitut, Copenhagen, Denmark
Summary. One hundred-fourteen patients with inoperable adenocarcinoma of the lung (ACL) were evaluated by immunohistochemistry with monoclonal antibodies against Neuron Specific Enolase (NSE) and Chromogranin A (Chr A) in order to determine the frequency and prognostic impact of such antigen expression. All patients were previously untreated and received chemotherapy according to a prospective randomized trial. The tumors of 18 patients (16%) had more than 10% positive cells stained with anti-NSE, 59 (52%) had 1-10% positive cells and those of 37 patients (32%) contained no NSE-positive cells. The corresponding figures for Chr A were: 22 patients (19%), 51 patients (45%) and 41 patients (36%), respectively. Forty-four per cent of the patients with more than 10% positive NSE cells responded to chemotherapy (either complete or partial remissions) compared to 17% of the patients with fewer than 10% positive cells (p < 0.025). The corresponding values for Chr A were 30% responders versus 19% responders (not statistically significant). The median survival for patients with more than 10%, 1-10% or no NSE-positive cells was 262 days, 231 days and 159 days, while, for Chr A it was 245 days, 200 days and 238 days, respectively. The survival curves for both NSE and Chr A according to the various levels of positivity were not significantly different. The presence of neuroendocrine marker in pulmonary adenocarcinoma seems to be associated with increased sensitivity to chemotherapy. Key words: pulmonary adenocarcinoma, diagnosis, histopathology, Neuron Specific Enolase, Chromogranin A
Introduction The histologjcal classification of lung cancer has great influence on the choice of therapy. Small-cell carcinoma of the lung (SCLC), frequently exhibiting neuroendocrine markers [1, 2], is more sensitive to chemotherapy than adeno-, large-cell- or squamous cell carcinoma. When chemotherapy is given to patients with inoperable adenocarcinoma of the lung (ACL), the response rates are usually reported to be around 25% [3, 4], as compared to 80-85% for SCLC. Histological diagnosis of ACL is traditionally carried out by using Haematoxylin-Eosin (HE) and mucin stains, according to WHO criteria [5]. However, the interest in application of immunohistochemistry to the study and categorization of lung cancers has increased during recent years [6, 7, 8,9]. Enolases are glycolytic enzymes necessary for the anaerobic catabolism of glucose. The enzymes are dimers, variously made up of alpha, beta or gamma subunits. The alpha dimer is widespread, while the beta dimer is mainly found in muscles, and the gamma dimer (NSE) mainly in neurons and neuroendocrine cells [10, 11]. The latter has accordingly been used as a
marker for tumours with neuroendocrine features, including SCLC [12, 13]. Several studies have also demonstrated NSE immunoreactivity in non-SCLC tumours [14, 15], and in tumours not included in the APUD system, e.g., breast carcinomas [16, 17], and the specificity of NSE for the diagnosis of SCLC is therefore questionable. Chromogranin A (Chr A) is an acidic glycoprotein of high molecular weight originally isolated from the adrenal medulla [18]. It has been reported to occur in neuroendocrine cells and neoplasms, including SCLC, and Chr A and dense core granules are expressed concordantly [19, 20]. It is released together with neuroendocrine peptides and amines through exocytosis from granules, but its physiological relevance is still unknown [19]. High plasma concentrations have been demonstrated in patients with SCLC, especially with extensive disease, whereas low plasma concentrations or concentrations below detection limits have been found in patients with other lung tumours, including ACL [21, 22]. Furthermore, Chr A has been demonstrated in cell lines of classic SCLC, whereas cell lines of non-SCLC do not contain Chr A [23]. Accordingly, it has been hypothesized that Chr A might be used as a diagnostic marker
356 for neuroendocrine rumours and as a quantitative indicator of disease activity in SCLC [21, 22]. In the present study, the frequency of immunohistochemical demonstration of NSE and Chr A was evaluated in tumours from a series of patients with inoperable ACL treated with cytostatic agents in order to learn whether such expression may presage an increased response rate to chemotherapy. Materials and methods
details (see ref. 27). A human fetal lung was used as positive control for NSE staining, and a human fetal jejunum was used for the Chr A staining. As negative controls the primary antibodies were omitted, but otherwise the same procedure was used. Previous studies [28] employing the present antibody to NSE and type-matched IgG on similarly fixed SCLC tissue, have demonstrated an absence of unspecific staining reaction and high grade of sensitivity with this monoclonal antibody. No specimens were available for serum analysis of NSE and Chr A, or for electron microscopy.
Patients Evaluation The study included patients up to 70 years old with ACL, stages Ilia, IHb or IV. The tumours were classified according to the criteria recommended by WHO [24]. All of the patients received cytostatic treatment in a prospective trial and were randomized to either vindesine as a single agent or lomustine plus cyclophosphamide plus methotrexate, or all four drugs [3]. Patients with measurable (bidimensionally measurable) as well as patients with evaluable (unidimensionally measurable) disease were included. Response assessment was performed according to the WHO criteria [25] and was verified by two observers. No differences were observed with respect to response rate or survival between the three treatment regimens [3]. Histological material The histologic specimens consisted of tissue from primary tumours obtained by bronchoscopy or thoracotomy. All specimens were taken prior to chemotherapy. The tumour tissues were fixed in 10% formaldehyde, embedded in paraffin, sectioned and stained with HE and mucin stains (PAS and Alcian blue) for conventional histologic examination, as described previously [26]. From each tumour 1-2 blocks were available and from each block 1—2 slides were selected containing sufficient tumour tissue. The slides were then stained according to the protocol. For immunohistochemistry, 5 micrometer sections were blocked for endogenous peroxidase activity by methanol H 2 O 2 treatment for 30 minutes and were then stained by the indirect immunoperoxidase method, using a monoclonal antibody to NSE (Sanbio, Holland), and a monoclonal antibody to Chr A (Hybritech Inc., USA) as primary antibodies, both incubated at a 1:100 dilution in BSA/TBS (0,25% Bovine Serum Albumin and 0.01 Molar Tris Buffer, pH = 7.4 containing 0.15 Molar Sodium Chloride) for 1 h. The site of antigen-antibody reaction was revealed by the indirect peroxidase method using peroxidase-labelled anti-mouse IgG (Dakopatt A/S Copenhagen, Denmark) as secondary antibody (1:20 for 1 hour). The peroxidase activity was detected by incubation in diaminobenzidine-hydrogen peroxidase medium for 10 minutes. For further methodological
The immunostaining was assessed by one observer (BGS). The number of stained cells in each section was scored and classification was as follows: neg: no staining of tumour cells, +: 75% of the tumour cells stained. The tumours containing more than 10% staining cells were so few in number that they were combined into one group to make statistical analysis feasible. Differences in staining intensities were not evaluated. The immunohistochemical evaluation was blinded for the clinical results. Comparison of response rates was based on the Chi-square test. The analysis of survival times was performed by the Kaplan-Meier method and the log rank test [29, 30]. Statistical significance was assumed when p was less than 0.05.
Results Of the 259 patients included in the study, the primary diagnosis for 39 was based on cytology alone, and these patients were excluded from the study. Five patients were considered not to have adenocarcinoma when the HE and mucin-stained sections were reviewed, and there was insufficient tissue available from 101 patients for immunohistochemistry examination. The remaining 114 patients with histologically verified ACL were included in the study. There was no significant difference in either rates of response to chemotherapy (0.3 < p < 0.4) or survival times (0.3 < p < 0.4) for the 114 patients included compared to the 145 patients not included in the study, nor was there any difference in the response rates or survival times between the three treatment regimens in the study [3]. Table 1 summarizes the staining results of the tumours with the two monoclonal antibodies. NSE was expressed in 77 tumours (69%), with more than 10% of the cells being positive in 18 tumours (16%). Corresponding values for Chr A were 73 tumours (64%) and 22 tumours (19%), respectively. The staining was mainly cytoplasmic, and the staining intensity varied within and between sections.
357 Table 1. NSE and Chr A in pulmonary adenocairinoma correlated with age, sex, performance status, extent of disease, and response rate. Characteristic
Chr A
NSE
All patients
Neg
+
++
Neg
+
++
N (%)
N(%)
N(%)
N(%)
N(%)
N(%)
N(%)
No of patients
37 (32)
59 (52)
18(16)
41 (36)
51 (45)
22 (19)
259(100)
Age in years Median Range
53 37-69
56 32-69
60 31-70
53 37-69
57 32-69
57 31-70
57 31-70
Sex Male Female
19(17) 18(16)
32 (28) 27 (24)
12(10) 6 ( 5)
23 (20) 18(16)
27 (24) 24(21)
14(12) 8 ( 7)
139 (54) 120 (46)
Kamowsky's performance 90-100% 70-80% 50-60%
14 12 11
25 25 9
8 8 2
18 12 11
21 21 9
9 11 2
88 115 56
8 10 19
19 15 25
2 6 10
10 14 17
15 10 26
4 7 11
65 71 123
Negative and +
++
Negative and +
++
17%
44%
19%
30%
Stage
ma mb IV
Response rate Neg. - all tumour cells negative,
+ — 0-10% positive tumour cells,
Forty-four per cent of the patients with tumours containing more than 10% NSE-positive tumour cells responded to therapy (partial or complete response), whereas only 17% of those with less than 10% positive cells responded (p < 0.025). Corresponding values for Chr A were 30% versus 19% responders (not statistically significant).
24%
- - > 10% positive tumour cells.
The median survival time for patients with NSEnegative tumours was 159 days versus 231 days for patients with 0-10% NSE-positive tumour cells and 262 days for those with more than 10% NSE-positive cells. The median survival time for patients with any NSE positivity was 236 days (not statistically significant, Fig. 1). Corresponding figures for Chr A were 238, 100
- -•- 0% NSE positive ceBs (n -37) — 1-10% NSE positive cels(n-59) >10% NSE positive cells (n-18)
\ L 80-
'Ul
...
0% Chr-A positive cels(n-41) 1-10% Chr-A positive cells(n-51) >10% Chr-A positive eels (n-22)
|60V)
-c
£40
htrs
20-
•-"In —i
30 Months
12
\
1
1
18
24
1—•
30 Months
Fig. I. Survival curves for NSE according to the 3 levels of Fig. 2 Survival curves for Chr A according to the 3 levels of posipositivity of the tumours. tivity of the tumours.
358
200, 245 and 210 days (not statistically significant, Fig. 2). There were no statistically significant differences in either response rates (13 versus 31%) or survival times (202 versus 261 days) for patients with tumours negative for both NSE and Chr A (31 patients) compared to patients who had tumours with more than 10% positive cells for both antibodies (16 patients). None of the patients had tumours with more than 10% NSE-positive cells and simultaneously no Chr A positive cells, and only one patient had a tumour negative for NSE but with more than 10% Chr A-positive cells. Discussion Studies of NSE in patients with ACL report varying numbers of positive tumour cells [8, 9]. Said et al., using 3 different antibodies to NSE, have demonstrated diffuse cytoplasmic staining in 2 of 9 adenocarcinomas, while Addis et al., using a polyclonal antibody, demonstrated NSE in 3 of 39 ACL [8, 9]. Other studies have demonstrated no positive tumour cells [12, 14]. Several factors may explain these inconsistent results. Firstly, different fixatives, antibodies and methods have been used. Paraffin-embedded material fixed in formalin and Zamboni's fixative have been used in several studies with inconsistent results [8, 9, 12, 14, 15]. Because many of the studies were based on small numbers of tumours, or different primary antibodies were used, it has been difficult to compare their results. Furthermore, the sources of the histological material used varied considerably. Biopsies from broncho- and mediastinoscopies and specimens from thoracoscopies and lung resections differ in size and quality. A study by Bergh et al. showed that 35% of small and/or squeezed or crushed biopsies with SCLC were positive for NSE compared to 83% of larger and better preserved biopsies [15]. Only a few previous studies have evaluated Chr A in lung carcinomas. In a study of 10 non-SCLC cell lines Chr A could not be demonstrated, whereas Chr A was immunocytochemically demonstrated in 11 of 18 cell lines of classical SCLC and with radioimmunoassay in 15 of 18 cell lines [23]. Wilson et al. [31] studied 10 formalin-fixed SCLC with monoclonal anti-Chr A and found 4 tumours with positive cells, whereas neither of the two squamous cell carcinomas were positive. Graziano et al. [32] demonstrated Chr A in 3 of 45 non-SCLC and Mooi et al. [33] found 3 of 11 nonSCLC lung tumours positive for Chr A. Again the differences in results may be due to the use of different antibodies. The antibodies recognize only part of the antigen, and different antibodies recognizing different epitopes may behave in different ways with respect to the fixation. However, lung tumours are often heterogeneous and may contain several distinct tumour cell types [34]. If
expression of NSE reflects neuroendocrine differentiation, as has been repeatedly proposed [10, 12], our investigation has detected such differentiation by immunocytochemistry in ACL. Demonstration of NSE in adenocarcinomas, which show no morphologic resemblance to SCLC, supports the concept that all lung tumours might originate from a stem cell type which is able to differentiate in several ways, including the development of neuroendocrine features - the Unitarian theory of development of lung carcinomas [35]. Irrespective of the biological significance of an immunohistochemical demonstration of NSE in ACL, the present study has demonstrated that the method can be used to detect a subgroup of ACL with high NSE reactivity and significantly better response to chemotherapy than tumours with less NSE reactivity. The initial number of patients in the present study was 259, but in about half of these the diagnosis had been based on cytologic material or biopsies from histologic material which was inadequate for immunocytochemical analysis. Therefore, very small and crushed biopsies were excluded from our material, and this may well have influenced the frequencies observed. To assess whether the 145 patients with material not eligible for immunocytochemical study constituted a selected group, we calculated their response rates and survival times but found no difference compared with the 114 patients included in the study. We therefore hypothesize that the finding in 114 patients is also true for the entire cohort. The clinical value of immunocytochemical demonstration of neuroendocrine features in non-SCLC has been the subject of three recent publications. Berendsen et al. [36] have prospectively studied the demonstration and prognostic impact on non-SCLC tumours of a series of MOC-antibodies reactive with SCLC-associated antigens, previously characterized as neuroendocrine-related differentiation antigen. In their study of 141 patients with non-SCLC tumours, they noted the presence of markers indicative of neuroendocrine differentiation in about 30% of the patients, most pronounced in adenocarcinomas. One of the antibodies, MOC-1, recognized a non-SCLC subgroup with worse prognosis in surgically treated patients. There was an equal distribution in neuoendocrine markers between metastatic lesions and primary tumours, indicating that neuroendocrine features of non-SCLC do not correlate with a propensity for early metastasis per se. Whether patients with neuroendocrine tumours have a different response to chemotherapy was not answered in the study by Berendsen. Linnoila et al. [37] described a group of non-SCLC patients who were treated with chemotherapy; they noted more responses in the patients in whom a neuroendocrine differentiation was recognized. However, the survival time was equal in both groups with or without neuroendocrine differentiation. In our study there was a tendency towards prolonged survival in the NSE-
359
positive group of patients with ACL. Finally, Graziano et al. [32] retrospectively reviewed 52 patients with non-SCLC tumours, including 26 who responded to chemotherapy and 26 without tumour response. They related the clinical outcome to the expression of neuroendocrine markers (NSE, Chr A and Leu 7). In the latter study, two markers were positive in 10/26 responders (38%) and in 0 of 26 patients without chemotherapy response. Responders with two or more positive markers had longer survivals than did responders with less than two positive markers. The authors emphasized, however, that the findings should be interpreted with caution due to the relatively small numbers of patients in the study and its retrospective nature. It seems justified to conclude from the present study and the few other reports in the literature that the presence of neuroendocrine markers in non-SCLC tumours might be associated with increased senstivity to chemotherapy. However, future larger prospective studies are needed to confirm the benefit of chemotherapy for this subgroup of non-SCLC patients. This study was presented at the 25th Annual Meeting of the American Society of Clinical Oncology, 1989 (abstract 857). Acknowledgement This work was supported by the Danish Cancer Society. References 1. Richardson RL, Greco FA, Oldham RK et al. Tumor products and potential tumour markers in small cell lung cancer. Semin Oncol 1978; 5: 253-62. 2. Carney DN, De Leij L. Lung cancer biology. Semin Oncol 1988; 15: 199-214. 3. Serensen JB, Hansen HH, Dombernowsky P et al. Chemotherapy for adenocarcinoma of the lung (WHO III): A randomized study of Vindesine versus Lomustine, Cyclophosphamide, and Methotrexate versus all four drugs. J Clin Oncol 1987; 5: 1169-77. 4. Osterlind K, Horbov S, Dombernowsky P et al. Vindesine in the treatment of squamous cell carcinoma, adenocarcinoma and large cell carcinoma of the lung. Cancer Treat Rep 1982; 66: 305-9. 5. World Health Organisation. Histological Typing of the Lung tumours. International Histological Classification of Tumours, No. 1, 2nd ed. Geneva: WHO. 1981. 6. Moss F, Bobrow L, Sheppard MN et al. Expression of epithelial and neural antigens in small-cell and non-small-cell lung carcinoma. Pathol 1986; 149: 103-11. 7. Mooi WJ, Wagenaar SS, Schol D et al. Monoclonal antibody 123C3 in lung tumour classification. Immunohistology of 356 resected tumours. Molecular and Cellular Probes 1988; 2: 31-7. 8. Said JV, Vimadalal S, Nash G. Immunoreactive Neuronspecific Enolase, Bombesin and Chromogranin as markers for neuroendocrine lung tumours. Hum Pathol 1985; 16: 236-40. 9. Addis BJ, Hamid Q, Ibrahim BN et al. Immunohistochemical markers of small cell carcinoma and related neuroendocrine tumours of the lung. J Pathol 1987; 153: 137-50.
10. Schmechel D, Marangos PJ, Brightman M. Neuron Specific Enolase is a molecular marker for peripheral and central neuroendocrine cells. Nature 1987; 276: 834-6. 11. Tapia FJ, Barbosa AJA, Marangos PJ. Neuron-specific enolase is produced by neuroendocrine tumours. Lancet 1981; i: 808-11. 12. Sheppard MN, Corrin B, Bennett MH et al. Immunohistochemical localization of neuron specific enolase in small-cell carcinoma and carcinoid tumours of the lung. Histopathology 198; 48:171-81. 13. Carney DN, Ihde DC, Cohen MH et al. Serum neuron specific enolase: a marker for disease extent and response to therapy of small-cell lung cancer. Lancet 1982; i: 583-5. 14. Dhillon AP, Rode J, Dhillon DP et al. Neural markers in carcinoma of the lung. Br. J. Cancer 1985; 51:645-52. 15. Bergh J, Esscher T, Steinholtz LM et al. Immunohistochemical demonstration of neuron-specific enolase (NSE) in human lung cancers. Am J Clin Pathol 1985; 84:1-7. 16. Nesland JM, Holm R. Neuron specific enolase immunostaining in the diagnosis of breast carcinomas with neuroendocrine differentiation. Its usefulness and limitations. J Pathol 1986; 148: 35-43. 17. Virones SA, Bonnin JM, Rubinstein LJ et al. Immunohistochemical demonstration of neuron-specific enolase in neoplasms of the CNS and other tissues. Arch Pathol Lab Med 1984; 108:536-40. 18. O'Conner DT, Frigon P, Sokoloff RL. Human chromogranin A: purification and characterization from catecholamine storage vesicles of human pheochromocytoma. Hypertension 1984; 6 2-4. 19. O'Conner DT. Widespread immunoreactivity in polypeptide hormone producing tissues and in serum. Regul Pept 1983; 6: 263-80. 20. Gazdar AF, Helman LJ, Israel MA et al. Expression of neuroendocrine cell markers L-dopa decarboxylase, chromogranin A, and dense core granules in human tumors of endocrine and nonendocrine origin. Cancer Res 1988; 48: 4078-82. 21. Sobel RE, O'Conner DT, Addison J et al. Elevated serum Chromogranin A in small-cell lung carcinoma. Ann Intern Med 1986; 105:698-700. 22. O'Conner DT, Deftos LJ. Secretion of Chromogranin A by peptide-producing endocrine neoplasma. N Engl J Med 1986; 314: 1145-51. 23. Deftos LJ, Linnoila RI, Carney DN et al. Demonstration of Chromogranin A in human neuroendocrine cell lines by immunohistology and immunoassay. Cancer 1988; 62: 92-7. 24. Mountain CF. A new international staging system for lung cancer. Chest 1986; 89 suppl: 225s-33s. 25. World Health Organisation. Handbook for Reporting results of Cancer Treatment. Geneva, World Health Organization. 1979. 26. Serensen JB, Hirsch FR. Olsen J. The prognostic implication of histopathologic subtyping of pulmonary adenocarcinoma according to the classification of the World Health Organization. Cancer 1988; 62: 361-7. 27. Larsson LI. Immunohistochemistry. Theory and Practice. CRC Press, Boca Raton, FL. 1988. 28. Jargensen LGM, Hirsch FR, Skov BG et al. Occurrence of Neuron Specific Enolase in tumour tissue and serum in smallcell lung cancer. Br J Cancer 1991; 63: 151-3. 29. Kaplan EL, Meier P. New parametric estimation from incomplete observation. J Am Statist Assoc 1958; 53:457-81. 30. Peto R, Pike MC, Armitage P et al. Design and analysis of randomized clinical trials requiring prolonged observation of each patient Br J Cancer 1977; 35:1-39. 31. Wilson BS, Lloyd RV. Detection of Chromogranin in neuroendocrine cells with a monoclonal antibody. Am J Pathol 1984; 115:458-68. 32. Graziano SL, Mazid R, Newman N et al. The use of neuroendocrine immunoperoxidase markers to predict chemo-
360
33.
33.
34. 35.
therapy response in patients with non-small-cell lung cancer. J Qin Oncol 1989; 7: 1389-1406. O'Conner DT, Burton D, Deftos LJ. Chromogranin A: immunohistology reveals its universal occurrence in normal polypeptide hormone producing endocrine glands. Life Sci 1983; 33: 1657-63. Mooi WJ, Dewar A, Springall D et al. Non-small-cell lung carcinomas with neuroendocrine features. A light microscopic, immunohistochemical and ultrastructural study of 11 cases. Histopathology 1988; 13: 329-37. Roggli JG, Vollmer RT, Greenberg SD et al. Lung Cancer heterogeneity. A blinded and randomized study of 100 consecutive cases. Hum Pathol 1985; 16: 569-79. Gazdar AF, Carney DN, Guccion JC et al. Small cell carcinoma of the lung: Cellular origin and relationships to other primary tumors. In Greco FA, Oldham RK and Bunn PA (eds.): Small cell lung cancer, Grune & Stratton, New York, 145-75, 1981.
Book review Surgical oncology, current concepts and practice.
C. S. McArdle (ed). Butterworths, London, Boston, Singapore, Sydney, Toronto, Wellington, 1990. 342 pp, £55.00 This is an excellent, easy-to-read, comprehensive review of the current therapeutic concepts and practice with respect to solid tumors. Our colleagues from the UX. are to be congratulated on this series of chapters, written by well-known experts in the field. The subjects reviewed include carcinoma of the oesophagus, gastric cancer, carcinoma of the pancreas and biliary system, colorectal cancer, endocrine tumors, breast cancer, kidney tumors, bladder cancer, prostate and testis cancer, head and neck tumors, malignant melanoma, bone and soft tissue tumors and malignant lymphoma. The two final chapters are dedicated to principles of cancer chemotherapy and to recent advances in cancer research. Despite the fact that each of the chapters is written by a different expert, all of them are outlined according to common guidelines, with many subtitles including epidemiology, etiology, pathogenesis, diagnosis, clinical features, surgical therapy and combined modality approaches. An extensive, up-to-date review
36. Berendsen H, Leij L de, Poppema S et al. Clinical characterization of non-small-cell lung cancer tumors showing neuroendocrine differentiation features. J Clin Oncol 1989; 7: 1614-20. 37. Linnoila RI, Jensen S, Steinberg S et al. Neuroendocrine differentiation correlates with favourable response to chemotherapy in patients with non-small-cell lung cancer. Fifth World Conference on Lung Cancer (abstr) 1988; 3.08. Received 19 November 1990; accepted 9 January 1991. Correspondence to: Fred R. Hirsch, MD Department of Oncology Rigshospitalet Blegdamsvej 9 2100 Copenhagen, Denmark
Annals of Oncology 2: 360, 1991.
of the most relevant literature concludes each chapter. The ones on cancer of the oesophagus, colorectal cancer and breast tumors are of outstanding value, whereas little information is given on the current muJtidisciplinary approach to head and neck tumors and bone/ soft tissue tumors. Unfortunately, there are no chapters on lung cancer, the most common solid rumor in Western countries, or anal carcinoma, a paradigm of multimodality therapy. In addition to the well-written final chapters on medical therapy the Editor should have included a similar one on the principles and concepts of radiotherapy (including radiobiology, radiation beam sources, field planning, dose fractionation). Some of these aspects are important for the operating surgeon. Despite these minor flaws, I greatly enjoyed reading this book. It has great potential value for the general surgeon and for the non-surgical oncologist interested in this field, and should be placed in all surgical libraries. U. Metzger Zurich
