Medical Hypotheses (1999) 53(3), 243245 1999 Harcourt Publishers Ltd Article No. mehy.1998.

0753

Differentiation, histogenesis and morphogenesis: their implications for tumor diagnosis

G. E. Quinonez
Department of Pathology, Health Sciences Centre, University of Manitoba, Winnipeg, Manitoba, Canada

Summary Differentiation, as employed in tumor classification for histopathology, refers to the resemblance of neoplastic cells to their presumed cell/tissue of origin. Since differentiation for diagnosis is obtained by analysing histological images, this histogenetic concept creates practical limitations. Morphogenesis, a variant of differentiation, is a wider concept since it takes into consideration the nature of the tumor cells, their organization and their synthetic products. Consequently, morphogenesis, rather than differentiation, seems to be a better predictor of histogenesis in tumor diagnosis. 1999 Harcourt Publishers Ltd

INTRODUCTION Differentiation and histogenesis are two intimately related concepts on which tumor classification and nomenclature are based. Since histogenesis is only a conclusion arrived at by implication from differentiation, an understanding of both terms has attracted attention and challenges in the pathology literature. In an article published in 1986 (1), Dr Victor E. Goulds analysis of these concepts clarified their distinctiveness in embryology and pathology and emphasized the impact of immunohistochemistry (IHC) in tumor classification. Since then, a shift toward the study of the genome has halted further thought on these concepts as applied to the practice of tumor diagnosis. The recently revived concept of morphogenesis, a variant of differentiation, is a study of form from a wider perspective than differentiation since it simultaneously takes into consideration the nature of the cells, their organization and their synthetic products (24). Consequently, morphogenesis may prove to be a better predictor of histogenesis than differentiation and it appears to give a more conceptually valid system of clas-

sification and nomenclature for tumors that that obtained from differentiation. TUMOR PHENOTYPE THE BASIS FOR MORPHOLOGICAL DIAGNOSES When making a diagnosis of a tumor by light microscopy, pathologists observe static images of tissue structural abnormalities that correspond to phenotypes (differentiation) and relate these phenotypes to their presumed tissue or cell of origin (histogenesis). The malignant phenotype of diagnostic relevance has a structural and functional nature. Structural abnormalities are observed as tissue architectural arrangements (tissue patterns) and cell morphological features (cell patterns) and are evaluated at three levels of biological organization: tissue, cellular and sub-cellular. At the tissue level, light microscopy and histochemistry (e.g. reticulin stain) reveal patterns of cell organization. At a cellular level, the same techniques demonstrate differences between neoplastic cells and normal cells. At the sub-cellular level, transmission electron microscopy (TEM) reveals cytoplasmic features and cell organization patterns impossible to uncover by other techniques and it may demonstrate a considerably greater degree of morphologic differentiation than that revealed by light microscopy alone. Functional abnormalities are expressed in tumor cells by distorted secretory activities and by their antigenic constitution. For diagnosis, the secretory activity is identified by histochemistry, IHC and TEM whereas the cells 243

Received 30 January 1998 Accepted 13 March 1998 Correspondence to: G. E. Quinonez MD, MSc, FRCPC, Department of Pathology, Health Sciences Centre, University of Manitoba, 820 Sherbrook Street, Winnipeg, Manitoba, R3A 1R9 Canada. Phone: +1 204 787 4639; Fax: +1 204 787 1790

244

Quinonez

antigenic constitution is determined by IHC and immunocytochemistry. DIFFERENTIATION Differentiation, as used in pathology, can have more than one connotation. It might refer to the evolutionary cytodifferentiation of embryonic cells (5) or to changes from a lower to a higher state of specialization (6). The first definition is used in developmental pathology and the second one in experimental carcinogenesis. In genetic terms, differentiation is the process of regulated gene expression that gives rise to different phenotypes from a common genotype (7). The definition most often used by pathologists to classify tumors is stated as the extent to which parenchymal tumor cells resemble comparable normal cells, both morphologically and functionally (8). From morphology and when applied to tumor diagnosis, differentiation is really the phenotypic expression of neoplastic tissues observed at the time of the histopathological diagnosis and at a specified microscopic magnification. It results from the activation of (a) portion(s) of the cellular genome in stem cells or clones by microenvironmental factors (9,10). HISTOGENESIS Pathologists refer to tumor histogenesis as a hypothesis as to the cell of origin, based on an assessment of differentiation and a conclusion arrived at by inference, and not by direct observation of the original cell or cell clone (1). These definitions refer to a cell for the tumors origin (1) and infer that histogenesis is an assumption based on the analysis of an image (10). Unfortunately, the process of inference from differentation to histogenesis is limited in diagnosis. A tumor cell looking morphologically similar to another cell from the presumed tissue of origin does not mean that the former has arisen from the latter. The pathologists assumption relating a histological image to the presumed tumor cell of origin, nevertheless, has some validity because there is a dynamic aberrant process of differentiation in neoplastic cells caused by an altered genome and distinct from normal differentiation. Tumor and normal cell genomes are similar and microenvironmental conditions are responsible for distinct expressions in neoplastic cells as compared to normal cells (7). Morphological differentiation, therefore, is phenotypically associated with histogenesis. Phenotypically, differentiation imperfectly reproduces the tissue of origin in two possible ways: the tumor resembles the tissue from which the tumor arises (e.g. a well-differentiated adenocarcinoma of colon) or the tumor corresponds to metaMedical Hypotheses (1999) 53(3), 243245

plasias observed in the presumptive tissue of origin (e.g. squamous metaplasia and squamous cell carcinomas of cervix and lung, tubal metaplasia and serous carcinoma of endometrium (7). Alternatively, the tumor may not at all resemble the likely tissue of origin (e.g. alveolar soft part sarcoma, extraskeletal myxoid chondrosarcoma) (7). MORPHOGENESIS Differentiation, as applied to morphology and for the stated reasons, is limited for infering histogenesis when making tumor diagnoses. A revived concept that, if properly understood, can improve this inference is morphogenesis, a variant of differentiation. In the last 15 years, morphogenesis, both structurally and functionally, has been extensively studied experimentally in salivary gland tumors with the intention of making a positive impact on the classification and nomenclature of these neoplasms (24). Those studies allow for a reassessment of morphogenesis as applied to the cognitive process of tumor diagnosis. Morphogenesis is defined as the evolution and development of form, as the development of the shape of a particular organ or part of the body, or the development undergone by individuals who attained the type to which the majority of the individuals of the species approximate (12). Taking into consideration that morphogenesis so defined is the evolution and development of form, that a tumor is also a system that attains the type to which the majority of systems of its class approximates, and that a morphological diagnosis results from studying the phenotype of the form, it appears that the definitional aspects of morphogenesis as applied to tumors can be also accommodated to the diagnosticians needs. This can improve the way in which we use differentiation in diagnosis. Morphogenesis, as a morphological concept, is really the interpretation of the tumor phenotype as put in evidence by the neoplasm architecture and its is based on the nature of the cells, their organization and their synthetic products. Histomorphological variations are explained by the degree of resemblance of a group of characteristics if tumor cells to their normal counterparts (well, moderately and poorly differentiated), cell organization (sheaths, cords, nest, trabeculae, etc.) and substance production (mucus, collagen, etc) (3). Morphogenesis is, for this reason, a wider and more useful concept to infer histogenesis than the extrapolation of isolated feactures of differentiation (e.g. bundles of tonofilaments by TEM: squamous cell carcinoma; keratin positivity by IHC: carcinoma). These examples illustrate the risk of taking ancillary techniques in isolation and upon which no final diagnosis should be based. An aberrant immunohistochemical result,
1999 Harcourt Publishers Ltd

Differentiation, histogenesis and morphogenesis

245

for instance, is useless and confusing histogenetically (13). Morphogenesis, as a mental construct of the tumor organization (form), is based on the perception of cell and tissue patterns as detected by light and electron microscopy, and complemented by other techniques such as immunohistochemistry. Using this concept, the perception of a static phenotype and its connection to a genomically related normal tissue may give even a false inference of structural evolution. Morphogenesis in this context, like differentiation, does not confirm histogenesis, which has a genetic basis, but it is better suited than differentiation to explain the cognitive association that occurs in diagnosis. The experimental studies carried out in salivary gland neoplasms have demonstrated that efforts for identifying the basic anatomical salivary gland unit (i.e. the ductoacinus) and its cell components (luminal and abluminal) are the basis for understanding morphogenesis and for improving tumor nomenclature and classification of complex neoplasms (24). This is exactly what is expected from a diagnosis. The morphogenetic unit is not a genome fragment but the organized phenotype as observed by light and electron microscopy and complemented by IHC (3). The extension of studies like those to other tumor systems, including the incorporation of molecular biology techniques, is expected to produce similar results. This could be the case, for instance, with tumors from breast (e.g. ductal carcinomas, phylodes tumors and metaplastic carcinomas), pancreas (e.g. carcinomas and islet cell tumors), ovary (e.g. surface epithelial tumors and stromal tumors), thyroid (e.g. follicular carcinomas, papillary carcinomas and medullary carcinomas) and lung (e.g. bronchogenic carcinomas and neuroendocrine carcinomas).

REFERENCES
1. Gould V. E: Histogenesis and differentiation: a re-evaluation of these concepts as criteria for the classification of tumors. Hum Pathol 1986; 17: 212215. 2. Dardick I., van Nostrand A. W. P. Morphogenesis of salivary gland tumors. A prerequisite to improving classification. Pathol Annu 1987; 22 (pt1): 153. 3. Dardick I. Histogenesis and morphogenesis of salivary gland neoplasms. In: Ellis G. L., Auclair P. L., Gnepp D. R., ed. Surgical Pathology of the Salivary Glands. Philadelphia: WB Saunders, 1991: 108128. 4. Dardick I. Salivary Gland Tumor Pathology. New York: Iga KuShain, 1996: 1729. 5. Bolande R. P. Developmental Pathology. Illinois: Am Assoc Pathol, 1979: 1. 6. Berenblum I. The nature of tumor growth. In: Florey H., ed. General Pathology. Philadelphia: WB Saunders, 1970: 645667. 7. Hall PA. Differentiation, stem cells and tumour histogenesis. In: Anthony P. P., MacSween R. N. M., eds. Recent Advances in Histopathology. Edinburgh: Churchill Livingstone, 1992: 115. 8. Cotran R. S., Kumar V., Robbins S. L. Pathologic Basis of Disease, 5th edn. Philadelphia: WB Saunders, 1994: 245252. 9. Farber E., Sarma D. S. R. Hepatocarcinogenesis: A dynamic cellular perspective. Lab Invest 1987; 56: 422. 10. Mills S. E. Sometimes we dont look like our parents (editorial). Modern Pathol 1995; 8: 347. 11. Henderson D. W., Papadimitriou J. M., Coleman M. Ultrastructural Appearances of Tumors, 2nd edn. Edinburgh: Churchill Livingstone, 1986: 2. 12. Histogenesis, morphogenesis. Dorlands illustrated Medical Dictionary, 28th edn. Philadelphia: WB Saunders, 1994: 769, 1058. 13. Mierau G. W., Berry P. J., Malott R. L., Weeks D. A. Appraisal of the comparative utility of immunohistochemistry and electron microscopy in the diagnosis of childhood round cell tumors. Ultrastruct Pathol 1996; 20: 507517.

1999 Harcourt Publishers Ltd

Medical Hypotheses (1999) 53(3), 243245