Ultrastructural heterogeneity in undifferentiated bronchial carcinoma

JOURNAL OF PATHOLOGY, VOL. 171: 53-57 (1993)

ULTRASTRUCTURAL HETEROGENEITY IN UNDIFFERENTIATED BRONCHIAL

CARCINOMA NAOMI CARTER, FIONA NELSON AND JOHN R. GOSNEY

Department of Pathology, University of Liverpool, Liverpool, I/. K.

Received 22 April I993 Accepted 19 May I993

SUMMARY Electron microscopy is often suggested as a useful aid to the classification of light microscopically undifferentiated

bronchial malignancies, features such as dense-core vesicles, desmosomes or tonofilaments, and microacini, allowing their designation as endocrine, squamous, or adenocarcinomas respectively. However, there is no reason to suppose that the heterogeneity of malignant bronchial tumours so often apparent by light microscopy or on immunolabelling might not occur at the ultrastructural level too. Extensive sampling of all deposits from eight subjects coming to necropsy with undifferentiated bronchial carcinoma revealed ultrastructural features of glandular and squamous differentiation to be widespread and often to occur together, although dense-core vesicles were not seen in any of the tumours studied. Heterogeneity was present within individual tumour deposits and particularly between different deposits of those tumours which had disseminated, such that any ultrastructural diagnosis would have been significantly influenced by sampling. Such variation should be borne in mind when ultrastructural features are used to classify bronchial malignancies.

KEY WORDS-Bronchial carcinoma, ultrastructure, differentiation, heterogeneity.

INTRODUCTION

The World Health Organization classification of malignant epithelial tumours of the lung' is based solely on their light microscopical appearance and takes no account of features of a particular line of differentiation which might be apparent ultrastructurally. Since electron microscopical examination of histologically undifferentiated tumours reveals many to possess features of squamous (epider- moid), glandular, or endocrine differentiati~n,~-~ it has been suggested that their routine ultrastructural examination might permit more accurate typing and allow more appropriate treatment. However, the morphological variability of bronchial carcinomas sometimes seen at the level of the light microscope would, if manifest also at the ultrastructural level, where the sampling error is even greater, cast doubt on the utility of electron

Addressee for correspondence: Dr J. R. Gosney, University Department of Pathology, Duncan Building, Royal Liverpool University Hospital, Liverpool L69 3BX, U.K.

01993 by John Wiley & Sons, Ltd. 0022-34171931090053-05 $07.50

microscopy as an aid to their classification. The aim of this study was to assess the extent of ultrastructural heterogeneity within and between individual deposits of a series of disseminated bronchial tumours, undifferentiated at the level of the light microscope, and to see how it might influence their classification.

MATERIALS AND METHODS Electron microscopical examination of tissue

from all primary and metastatic deposits of light microscopically undifferentiated bronchial carcinomas from eight subjects coming to necropsy with disseminated disease was undertaken. At the time of necropsy, all apparent deposits were sampled for histological and ultrastructural study. All necropsies were performed within 24 h of death.

For light microscopy, ten blocks of tissue were taken from the primary tumour and an appropriate number from its secondary deposits according to their size. These were fixed in neutral buffered

54 N. CARTER ET AL.

Table I-Ultrastructural features of eight ‘undifferentiated’ bronchial carcinomas*

Case No. Site LM

Blocks forEM DCV A M TJ MG D T K

1 Primary Pulmonary Pulmonary

Nodal 2 Primary

Cerebral 3 Primary

Hepatic 4 Primary 5 Primary 6 Primary

Hepatic 7 Primary

Hepatic Nodal

Adrenal Uterine Renal

8 Primary Cerebral

Renal

U u u u U

u ?a u ?a

u ?a u ?a u ?sq

u ?sc/lc u ?sc/lc u ?sq

?scllc ?sc/lc ?sc/lc

U

U

U u U

10 3 6 2

10 4

10 6

10 10 10 5

10 5 2 4 4 4

10 5 5

0 4 5 8 10 6 8 1 0 2 2 3 2 3 3 0 0 6 6 6 6 6 6 0 0 0 0 2 0 0 0 0 0 5 7 4 6 4 1 0 1 0 4 4 4 3 3 3 0 0 1 2 1 0 2 6 8 0 0 0 0 . 5 0 4 6 0 0 0 0 3 0 8 1 0 0 0 5 6 1 0 8 3 8 1 0 9 4 4 1 0 0 5 0 0 0 0 0 0 5 3 2 0 0 0 0 0 5 6 6 0 0 0 0 0 3 3 2 0 0 0 0 0 1 1 1 0 0 0 0 0 3 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 2 0 0 0 8 6 5 0 0 0 0 0 0 0 0 0 2 0 2 0 1 1 0

*Figures refer to the number of blocks from each deposit showing each feature. LM=light microscopy; u=no recognizable differentiation; ?sq=possible squamous carci-

noma; ?a=possible adenocarcinoma; ?sc/lc= possible small celYlarge cell carcinoma; DCV= dense-core vesicles; A= acini; M =microvilli; TJ = tight junctions; MG= mucous granules; D = desmosomes; T = tonofilaments; K = keratohyaline granules.

formalin and embedded in paraffin wax. Sections were cut at 4pm and stained with haematoxylin and eosin and by the periodic acid Schiff technique combined with alcian blue for general histology and identification of mucins, respectively.

For electron microscopy, between 40 and 50 1-2 mm cubes were cut from the primary tumour and each of its metastases using a razor blade, care being taken to sample, as far as possible, only tissue which appeared viable to the naked eye. These cubes were fixed in 2 per cent glutaralde- hyde, post-fixed in osmium tetroxide, dehydrated through alcohols, and embedded in epoxy resin. Semi-thin sections 1 pm thick were cut and stained with toluidine blue and examined by light microscopy. Up to ten blocks of viable tumour tissue from each deposit were selected for electron microscopy, ultrathin sections being cut and stained with uranyl acetate and lead citrate and

Ultrastructural indicators of squamous, glandular, and endocrine differentiation were then sought. The first was considered to be indicated by desmosomes, bundles of tonofilaments, and keratohyaline granules. The second was considered to be present if microvilli, intracellular microacini, tight junctions, and mucous granules were seen. The third was deemed present if dense-core (neurosecretory) granules were observed. For each tumour deposit, the number of tissue blocks in which these structures were identifiable was noted.

RESULTS

The light microscopical and ultrastructural features of the eight primary tumours and their metastases are summarized in Table I.

Light Microscopy examined with a Corinth 500 transmission electron Light microscopical examination of the ten microscope. blocks of tissue from each of the eight primary

ULTRASTRUCTURE OF BRONCHIAL CARCINOMA 55

tumours revealed a modest intralesional variation in morphology with occasional features suggesting possible differentiation along squamous, glandular, or endocrine lines in five (Table I). In none, however, were these sufficiently impressive to allow eventual classification of these primary tumours as anything other than ‘undifferentiated’ carcinoma. A similar histological heterogeneity was also seen within and between metastases, but here too no features were sufficiently marked to allow any of the tumours to be classified more precisely. Thus, in six tumours, the possible presence of occasional intercellular bridges and mucin droplets in either the primary deposit or its metastases suggested differentiation along squamous and glandular lines. In five tumours, a second population of cells smaller than the non-descript population of large cells which predominated could be identified, suggesting the possibility that these might be examples of the so-called ‘small celYlarge cell’ variant of small cell carcinoma, but the pleomorphism of the large cell component made this doubtful.

Electron microscopy Because of extensive sampling, we were able

ultrastructurally to examine ten blocks of viable tissue from all eight primary growths and between two and six from their metastases (Table I). Despite their being obtained post-mortem, preser- vation was generally excellent and ultrastructural features were clearly visible (Fig. 1). Microacini and desmosomes were particularly well preserved and easy to identify. Most of the former were intercellular, although some deposits contained significant numbers of intracellular microacini too. Desmosomes were usually well formed, with tonofilaments present as discrete cytoplasmic bundles, which sometimes could be seen inserting into them. Dense-core vesicles were not seen in any of the tissues studied.

Ultrastructural features of both squamous and glandular differentiation were identified in all but one of the primary tumours examined, the single exception being case 7, in which features of only squamous differentiation were seen. The five metastases of this particular tumour also displayed features of squamous differentiation only. Of the eight metastases of the remaining seven primary tumours, five showed features of both squamous and glandular differentiation; one (case 6 ) of just squamous; one (case 1) of just glandular; and one (case 8) of neither. These ultrastructural features

varied from block to block within any single tumour deposit and particularly between individual deposits of the same tumour, as well as being often at odds with the impression gained by light microscopy.

DISCUSSION

Extensive sampling of the eight tumours examined in this study reduced problems of post- mortem autolysis and permitted a reasonable representation of the range of ultrastructural appearances of each deposit to be obtained, We were readily able to identify features of glandular and squamous differentiation, especially microacini and desmosomes, although dense-core vesicles were not seen. This was despite the impression, on light microscopical examination, that five tumours might show endocrine differentiation. Some previous studies of the ultrastructure of ‘undifferentiated’ or ‘large cell’ bronchial carcinomas have described dense-core vesicles in a significant proportion?6 and we cannot exclude the possibility that these structures were being lost as a result of autolysis and, therefore, that endocrine differentiation was under-represented in our series. However, the number of cases that we studied was small, and many other investigations of the ultrastructure of such tumours have reported many, sometimes all, to show features of only glandular and/or squamous differentiati~n~.~,~ when examined by electron microscopy.

The fact that features of glandular and squamous differentiation were present in so many of the tumours studied is not, therefore, surprising. Nor is their co-existence, which was the rule rather than the exception in the neoplasms that we studied. Such co-existence is well described in previous ultrastructural and Carlile and Edwards” describe ultrastructural evidence of glandular differentiation even in tumours which could be classified as squamous by light microscopy.

The important point to emerge from the present investigation is that ultrastructural heterogeneity in histologically undifferentiated bronchial malignancies, within but particularly between deposits, can be considerable. This is not unexpected, since the morphological heterogeneity of bronchial carcinoma is well recognized at the level of the light microscope and when studied by immunolabelling for intermediate filament proteins and

56 N. CARTER ET AL.

Fig. I-Ultrastructural evidence of glandular and squamous differentiation in a histologically undifferentiated bronchial carcinoma with metastases to the lung and hilar lymph nodes (case 1). (a) shows intracellular acini from an intrapulmonary metastasis which also contained features of squamous differentiation in the form of intercellular bridges (b). Both structures were present in all six blocks from this particular deposit, but were less prevalent in a second intrapulmonary metastasis and in the primary growth, and absent from a metastasis in a hilar lymph node. (c) shows a well-formed desmosome from the primary growth. Similar desmosomes were widespread in the pulmonary metastases, but again were absent from that in a hilar lymph node. (d) shows bundles of tonofilaments in the same tumour deposit as the microacinus shown in (a). Uranyl acetate and lead citrate; original magnifications (a) x 31 000; (b) x 40 000; (c) x 46 900; (d) x 6700

ULTRASTRUCTURE OF BRONCHIAL CARCINOMA 57

those antigens considered to indicate endocrine differentiation. However, the prevalence of heterogeneity at the ultrastructural level has been little explored. Mooi et a1.” recently compared super- ficial and deep samples from a series of resected pulmonary tumours by electron microscopy. They found that in only two of 44 cases did the ultrastructural features differ significantly between the two samples, although a further nine showed a lesser discrepancy, with one of the pair showing evidence of a second line of differentiation. They did not, however, compare primary tumours with their metastases or different metastases with each other, and it was between such deposits that we found ultrastructural heterogeneity to be particularly marked.

Our findings carry two implications: one theoretical, one practical. In the first instance, our study provides further support for the now widely accepted view that the major types of bronchial carcinoma are variations on a theme rather than distinct entities. Its findings are consistent with the idea that all such tumours originate from a com- mon stem cell in the pulmonary epithelium with the potential for subsequent differentiation along one or more pathways. ‘Multidirectional’ differentiation within the same tumour may be the rule rather than the exception.” In the second instance, our study carries obvious practical implications for the classification of the disease in general and the possible contribution of electron microscopy in particular. Any of the tumours included in the present study could have been classified on ultrastructural grounds as squamous or adenocarcinoma depending on the amount of tissue examined and, in particular, from where it was taken. Even widespread sampling of a metastasis for ultrastructural examination might give a fundamentally different picture from that obtained by studying another deposit or the primary growth.

Whether the prevalence of the various ultrastructural features of differentiation of a disseminated bronchial tumour is influenced by its location or related to the properties which deter- mine the site to which it metastasizes in the first

place is unknown. Such questions are of considerable theoretical interest but, from a practical view- point, such variation as is described here should be borne in mind whenever ultrastructural features of bronchial tumours are used in an attempt to increase the accuracy of their classification.

ACKNOWLEDGEMENTS

We are grateful to Gordon Mellor for his invalu- able technical assistance. This work was part of a larger study supported by the Mersey Regional Health Authority.

REFERENCES 1. The World Health Organization histological typing of lung

tumours. Am J CIin Pathol 1982; 77: 123-136. 2. Churg A. The fine structure of large cell undifferentiated carcinoma

of the lung. Hum Pafhol 1978; 19 143-156. 3. Azar HA, Espinoza CG, Richman AV, Saba SR, Wang T-Y.

‘Undifferentiated’ large cell malignancies: an ultrastructural and immnnocytochemical study. Hum Path01 1982; 13: 323-333.

4. Warren WH, Memoli VA, Kittle CF, Hensik RJ, Faber LP, Gould VE. The biological implications of bronchial tumors. J Thorac Cardiovasc Surg 1984; 87: 274-282.

5. Albain KS, True LD, Golomb HM, Hoffman PC, Little AG. Large cell carcinoma of the lung. Ultrastructural differentiation and clinicopathologic correlations. Cancer 1985; 56: 1618-1623.

6. Kodama T, Shimosato Y, Koide T, Watanabe S , Teshima S. Large cell carcinoma of the lung-ultrastructural and immunohistochemical studies. Jpn J Clin Oncol 1985; 15 4 3 1 4 l l .

7. Mooi WJ, Zdndwijk NV, Diiigemans KP, Koolen MGJ, Wagenvoort CA. The ‘grey area’ between small cell and non-small cell lung carcinomas. Light and electron microscopy versus clinical Java in 14 cases. J Path01 1986; 1 4 9 49-54.

8. Mooi WJ, Dewar A, Springall D, Polak JM, Addis BJ. Non-small cell lung carcinomas with neuroendocrine features. A light microscopic, immunohistochemical and ultrastructural study of 11 cases. Histopathology 1988; 1 3 329-337.

9. Wick MR, Berg LC, Hertz MJ. Large cell carcinoma of the lung with neuroendocrine differentiation: a comparison with large cell ‘undifferentiated pulmonary tumors. Am J Clin Pathol 1992; 97: 796-805.

10. Carlile A, Edwards C. Poorly differentiated squamous carcinoma of the bronchus: a light and electron microscopic study. J Clin Pafhol 1986; 3 9 284-292.

11. Gosney JR. Pulmonary Endocrine Pathology: Endocrine Cells and Endocrine Tumours of the Lung. Oxford: Butterworth-Heinemann, 1992: 83-100.

12. Mooi WJ, Dingemdns KP, Wagenaar SS, Hart AAM, Wagenvoort CA. Ultrastructural heterogeneity of lung carcinomas: representa- tivity of samples for electron microscopy in tumor classification. Hum Path01 1990; 21: 1227-1234.

Ultrastructural heterogeneity in undifferentiated bronchial carcinoma

Documents

Transcript of Ultrastructural heterogeneity in undifferentiated bronchial carcinoma