Classification of Neoplasms: A Critical Appraisal

Publisher Website
Google Scholar
Abstract
CLASSIFICATION OF NEOPLASMS: A CRITICAL APPRAISAL ROBERT W. HUNTINGTON* and ROBERT W. HUNTINGTON IHf In scientific or practical study, schemes of classification and nomenclature have to be recognized as at once indispensable and fallible. They are devices of human students, hence subject to all human fallibilities. A scheme improvised for opening a new field of study will be based on conjecture, and one used for some time in a field under continued exploration cannot be expected to incorporate the newest data and insights . Yet in medicine, as in other scientific and practical endeavors, workers must communicate through a mutually understandable terminology . While it is no longer hoped that dreaded things can be made harmless by proper placatory pronunciation of their names [1], the attempt to cope scientifically with diseases demands an effort to give them right names, indicative of causes and results, and to group them properly in order to facilitate comparison and contrast, which are indispensable to the accumulation of knowledge. Thus, while traditional classifications are not to be taken too seriously, taxonomy is not to be regarded as a hopelessly frivolous operation. The proper attitude toward schemes of classification is one of critical scrutiny in the hope of improvement. Ceteris paribus, schemes based on discernment of alternatives, or of clusters and gaps, will be more satisfactory than those imposed on even and continuous distributions. Thus humans, and even fishes [2, 3], can usually be classed as male or female, and human blood groups can be listed as O, A, B, or AB despite such variants as A2, Ax, or Bombay. However, in tabulations of human weight, height, or age, groups must be demarcated arbitrarily. Dialectically speaking, classification into groups presupposes discontinuity, and application of classificatory grouping to a continuous series is like applying a polygon to a circle. Just *Director of medical services and consulting pathologist emeritus, Kern Medical Center, Bakersfield, California. Address reprint requests to 470 Wellington, Cambria, California 93428. !Assistant professor of pathology, University of Wisconsin Medical Center, Madison, Wisconsin 53706. Perspectives in Biology and Medicine ¦ Winter 1977 | 215 as a polygon approaches a circle by acquiring an infinite number of sides, a descriptive nosology of a continuously variable set would require infinite variations for a complete fit. All too often the classifier sets up more categories than he or his colleagues can use, with the fit still far short of perfection. Problems Involved A philosopher glancing at tumor biology might anticipate that classification of neoplasms would be difficult under any conceivable scheme, and might recall Cassirer's comment [1] that human capacity to make language is closely related to capacity to make myths. Morphologic study over the past century has demonstrated a huge and bewildering array of patterns, and more recent biochemical study has revealed another large range of discouragingly independent variables. As Foulds [4] has suggested, each individual neoplasm may be unique, and grouping , however necessary for study and management, inevitably arbitrary. Uncertainty in application of the traditional scheme of classification is vividly recorded in the assembled monthly minutes [5] of the regional study groups of the California Tumor Tissue Registry. Unanimity is unusual, while near-tie votes are not. Split is inter- rather than intragroup , with some groups all for alternative A, others for B, and intragroup consensus clearly representing persuadability rather than firm individual judgment. Yet the participants are outstanding both in concern with tumor pathology and in opportunity for self-education. Further indication of the unsatisfactory state of tumor classification is afforded by the vigor of attempts to improve it. In the endeavor to get tumor pathologists to "speak the same language," there have been elaborate definitions of categories from scholarly international committees. "Tumor slide conferences" are regular features of meetings of pathology societies. A moderator selected for expertise in the tumor group in question tries to teach colleagues to see the sections as he sees them and call them as he calls them. A rebel might suggest that at times the process comes perilously close to brainwashing. Yet the divergencies of interpretation remain. Current etiquette demands that when a patient with a neoplasm is transferred to another clinic the record must include not only the histopathologic diagnosis but also the...

This publication has 2 references indexed in Scilit: