Meeting of investigators on the histological definition of precancerous lesions: Geneva, 27 November-1 December 1972: report to the Director-General

WORLD HEALTH ORGANIZATION CAN/73.1 ORGANISATION MONDIALE DE LA SANTE ENGLISH 0NLY MEETING OF INVESTIGATURS ON THE HISTULDGICAL DEFINITION OF ratcmcsaous LESIONS ,snseu a , fi;‘rdq,a? 0* H ‘- OoNtGeneva, 27 November - 1 December 1972 §EEHP ”r 11:? ha} 0 \rt' : Lt: ‘14::- 'v : 2 3 3.5. 5..- U3 E REPORT TO THE DIRECTOR-GENERAL a Jrhi , _:i f List of Participants Members Dr F.K. Mostofi, Armed Forces Institute of Pathology, Washington D.C. 20305, United States of America Dr K. Dota, Department of Pathology, University of Tokyo, Faculty of Medicine, Hongo, TDRYD.Japan Dr J.J. Pindborg, Department of Oral Pathology. The Royal Dental College, Universitetsparken 4, Copenhagen fl, Denmark (Chairman) Dr H.E. Poulsen, Institute of Pathological Anatomy, Kommunehospitalet, 1399 Copenhagen K, Denmark Dr G. Riotton, Centre de Cytologie et de Dépistage du Cancer, 34 Boulevard de la Cluse, Geneva 1205, Switzerland Dr R.E.J. ten Seldam, Department of Pathology; University of Western Australia, Perth, Australia Dr L.M. Shabad, Department of Experimental Oncology, Institute of Experimental and Clinical Oncology of the USSR Academy of Medical Sciences, Karshirskoye Shosse 5, Moscow m-ara, USSR Dr P. Shubik, Eppley Institute for Research in Cancer, University of Nebraska, College of Medicine, 42nd and Dewey Avenue, Omaha, Nebraska 63105, United States of America Dr A.C. Thackray, Bland-Sutton Institute of Pathology, Middlesex Hospital Medical School, London W1? ?PP, England (Rapporteur) Dr L.E. Zimmerman, Armed Forces Institute of Pathology, Washington D.C. 20305, United States of America Secretariat Dr N.F. Napalkov, Chief, Cancer Unit, WHO, Geneva, Switzerland Dr L.H. Sobin, (Secretary), Pathologist, Cancer Unit, 0H0. Geneva, Switzerland Mrs R. Lunt, Scientist, Cancer Unit, WHD, Geneva, Switzerland Dr H. Tulinius, International Agency for Research on Cancer, Lyon, France Dr V. Turusov, International Agency for Research on Cancer, Lyon, France The issue of this document does not constitute Ce document ne conatitue one one publication. formal publication. it should not be reviewed, Ii ne doit faire I‘objet d’aucun compte rendu ou abstracted or quoted without the agreement of resume ni d‘aucune citatlon aana l‘autorlaation do the World Health Organization. Authors alone FOrganisation Mondiale de la Santa. Les opinions are reapenaible for views expressed m sinned exprirneea clans lea articlea aignea n'engagent articles. oue Ieura auteurs. CAN/?3.1 page 2 Agenda 1. Organization of work 2. Current information on the WHO IRCs for Histological Classification of Tumours 3. Precancerous lesions and pre~invasive cancers 3.1 Currant concepts 3.2 Histological classification. nomenclature and definitions 3.3 Comparisons between different sites 3.4 Evidence for prOgression and regression 4. ' Future directions for the work of the IRCs on precancerous and prerinvasive lesions 5. Other business OPENING OF THE MEETING On behalf of the Director-General Dr M.A. Akhmeteli. Director of the Division of Non- Communicable Diseases,opened the meeting. He noted that the Organisation is particularly concerned with integrating the work-done by various IRCs for histological classification and one of these subjects is precancerous lesions. This problem is important not only to the histopathologist and the experimental pathologist but also to the programme of cancer control because of the possibility of identifying patients with high risks of developing cancer. Dr Pindborg was elected Chairman and Dr Thackray served as Rapporteur. The meeting had at its disposal 15 documents and working papers. BACKGROUND FOR THE MEETING Dr Sobin gave a review of the current status of the programme on the International Histological Classification of Tumours, noting that at the present time seven publications had appeared and that the work of six additional IRCs had been completed. Dr Sobin, introducing the work to be carried out, started by recalling a report IrOm 1954 of a WHO Expert Committee (THE 275 , in which appeared the statement, ”A systematic re-evaluation and classification of prercancerous lesions in manr in pathological terms, would be of great value”. He emphasized that the object of the meeting would be to attempt to define. name and classify precancerous lesions at the major clinically accessible epithelial sites. The emphasis in this report has been placed on those lesions common to the fields of interest of the IRCs represented at this meeting. DEFINITIONS The participants discussed a number of tentative definitions related to the problem of ”precancer" and suggest the following for further consideration. Precancerous lesions Precancerous lesion: 3 morphologically altered tissue in which cancer is more likely to occur than in its apparently normal counterpart. Some lesions have a very high probability of developing cancer, i.e. 20% and over, others much lower: and the magnitude of the potentiality for many others is unknown. CAN/73 . 1 page 3 Precancerous conditions Precancerous condition: a generalized state associated with a significantly increased risk of cancEr. Examples of precancerous conditions are sideropenic dysphagia and aeroderma pigmentosum. Apart from precancerous conditions and lesions the possibility also exists that ill-defined large areas have become conditioned to the development of precancerous and cancerous lesions, e.g. irradiated skin and oral submucous fibrosis. Carcinoma in situ Carcinoma in situ: a lesion in which all or most of the thickness of the epithelium shows malignant cellular features, loss of polarity and loss of stratification, but there is no stromal invasion. {Synonyms and related terms: preinvasive carcinoma; intraepithelial carcinoma; intraepidermal carcinoma; noninvasive carcinoma.) The term carcinoma in situ should not imply that all such lesions will eventually become invasive although it is recognised that a substantial proportion will do so. It must be clearly recognized that in some situations carcinoma in situ is thought of as a precancerous lesion and in other sites as a cancer and prognosis and treatment vary accordingly. The present report is concerned only with an effort to establish a basis for the use of terms to describe similar morphological lesions in different tissues. In the skin, there are a number of distinct clinico-pathological conditions (e.g. Bowen's disease, erythroplasia of Queyrat) which all could be called carcinoma in situ although in practice this term is rarely if ever applied to such lesions. These lesions have certain distinguishing histological (and/or clinical) features but they also have certain features in common. These are loss of polarity of cells, pleomorphism, usually polychromasia of both nuclei and cytoplasm and the common presence of a variable number of mitotic figures which may or may not be abnormal. Invasion into the underlying dermis is absent. In the cervix, carcinoma in situ is confined to the epithelium of the surface and/pr the glands. Carcinoma in situ may be recognised in three histological forms. In the classical type the normal stratification of the epithelium has been lost from the basement membrane to the surface. The cells on the surface may he parakeratotic. The nuclei are tightly packed and are usually hyperchromatic. They often tend to be arranged in a vertical or diagonal rather than a horizontal manner. Cytoplasm is sparse and there is little, if any, differentiation. Mitoses are frequent, extending to the more superficial layers. Atypical mitotic forms are often found. Lesions with pronounced cellular abnormalities may show a degree of cellular maturation, usually more marked towards the surface. Such cases are included in the group classified as carcinoma in situ. There is a variant in all respects similar to the classical type except for smaller, more tightly packed cells similar to those found in small cell, non-keratinizing epidermoid carcinoma. In addition, a keratinizing form of carcinoma in situ occasionally occurs. The lesion is more apt to appear in the portio vaginalis. may": 3 . 1 page 4 In the oral mucosa. the term carcinoma_in sitn is used For that condition in which the uncle thickness of the epithelium shows malignant cellular features, characterized by oleomorphism with loss of polarity and surface stratification. Nuclei are hyperchromatic and show variations in size and shape. The nuclearrcytoplasmic ratio is altered. Nucleoli are enlarged and often multiple. Mitoses usually occur in all parts of the epithelium. Invasion is absent. In the conlunctiva and cornea. the whole thickness of the epithelium is replaced my cells showing malignant features (i.e. large hyperchromatic nuclei, altered nuclear— oytoplasmic ratio, increased mitotic activity, usually at all levels in the epithelium, abnormal mitotic figures, and pleomorphism} with a loss of polarity and surface stratification. Stromal invasion is absent and keratiniaation is minimal or absent. The affected epithelium is typically thickened and abruptly demarcated from the adjacent normal epithelium. In the stomach, carcinoma in gitu has been described but is extremely difficult to identify. Intramucosal carcinoma, i.e. invasion limited to the lamina propria, has been recognised but must not be considered as carcinoma in situ. In the urinary tract, carcinoma_iflnjitu is used to denote the lesion of a flat mucosal surface in which the normal rladder epithelium is replaced in all or most of its thickness by epithelial cells showing malignant features (‘.e- large hyperchromatic nuclei, altered nuclear-cytoplasmic ratio, increased mitotic activity in superficial layers, abnormal mitotic activity at any level, pleomorphlsm of cells, giant cells, loss of polarity and loss of stratification). The cells may or may not retain their transitional appearance. The cells may be less cohesive and may be desquamated leaving a denuded surface. The changes are usually seen on the surface but may also be seen in Bruno's nests and cystitis cystica. However, there is no invasion. Epithelial dysplasia The participants agreed that the term dysplasia is preferred to atypia, which is usually applied to purely cellular features. Dysplasia: a lesion in which part of the thickness of the epithelium is replaced by cells showing varying degrees of atypia, loss of polarity andXor loss of stratification less pronounced than in carcinoma in situ. There is no stromal invasion. Three grades of dysplasia - mild. moderate and severe, are commonly described. In dysplasia of the conjunctiva and cornea, some, but not all of the features of carcinoma in site are present to a variable degree. In some instances only the absence of full thickness involvement or the absence of mitotic figures beyond the basal cell layer serve to differentiate dysplasia from carcinoma in situ. In others the alterations are mild to moderate and the histologic picture does not closely resemble carcinoma in situ. In the oral musoca, the term epithelial dysplasia is used when some or all of the following changes are present: irregular epithelial stratification; hyperplasia of the basal layer; drop-shaped rete pegs; increased number of mitotic figures {a few abnormal mitoses may be present}; increased nuclearscytoplasmic ratio; loss of polarity of the basal cells: nuclear polymorphism: nuclear hyperchromatism; enlarged nucleoli; keratinisation of single cells or groups in the crickle cell layer; and loss of inter- cellular adherence. CAN/73, page 5 Dysplasia of the cervix tends to occur distally to the external os, but may at times involve the endocervical glands. There are varying degrees of atypia of the Epithelium featuring crowding of cells, loss of polarity. nuclear enlargement, hyperchromatism and a degree of impaired cytoplasmic maturation. In a severe stage, the lesion may be difficult, if not impossible, to distinguish from carcinoma in situ. In the urinary tract, the term dysplasia is generally not employed. PRECANCEROUS LESIONS AND CDNDITIDNS BEING CONSIDERED BY IRCs IhC on Uterine and Placental Tumours _ Most of the work of the group was devoted to changes in the squamous epithelium of the cervix and to changes in the endometrium. Cervix, squamous epithelium. There has been general agreement in the group to the use of the widely accepted terms dysplasia and carcinoma in situ. It was also agreed that dysplasia be divided into mild, moderate and severe. Dysplasia appears to occur from four to six years earlier than carcinoma in situ. Present evidence indicates that some of these lesions revert to normal, but some progress to carcinoma in situ, or even invasive carcinoma. Although a high percentage of untreated carcinoma in situ will eventually progress to invasive cancer, relatively few dysplastic lesions will do so in a similar period of time. Dysplasia commonly co-exists with carcinoma in situ and/or invasive cervical cancer. Cervix, columnar epithelium. Epithelial changes in glandular cells may be pronounced to the extent that they are indistinguishable from adenocarcinoma of the cervix. When these changes are confined to the epithelium they may be regarded as "adenocarcinoma in situ". Such lesions are not commonly encountered in the cervix. Endometrium. Various types and degrees of glandular hyperplasia including polypoid adenomas are encountered in the endouetrium. Glandular hyperplasia may have either a proliferative adenomatous pattern or a cystic ”Swiss cheese" pattern. Glandular hyperplaaia offers no particular diagnostic difficulties in its usual form. In more florid forms (atypical hyperplasia) stratification, disturbance in polarity, epithelial atypia and the crowding of cells, can present a difficult diagnostic problem. The known clinical association of hyperplasia with endometrial carcinoma also adds to the difficulty of interpretation. * Work is continuing on the conditions mentioned above and also on some other lesions such as hydatidiform mole and cervical papilloma. IRC on Oral Precanceroua Lesions The 1RD has concentrated its work on the precancerous lesions, leukoplakia and erythroplakia and the possible precancerous conditions lichen planus and submucous fibrosis. The work has been carried out on the basis of a set of definitions accepted at two meetings of the IRS. The evaluation of the histological features observed in the above-mentioned conditions has been done by two different methods: one codifying the pathologists' total interpretation without listing the histological components and the other, termed the Epithelial Atypia Index, grading by photographic standards the components of atypia, basically without interpretation. At the next meeting of the IRC (in March 1973) these methods will be evaluated. 1'. flahlrla.l Dame '3 it is now a well-estahiishei fact that oral leuhonlaaias only become malignant in 3‘51. whereas erythroplahias aooea r to involve a much h.ighe r risk {or cancer development. It 15 “fit FBt flflfisible eaactly to r welict which of the histological features are the best prognostic indicators. During the work of the [RC it has become clear that about 20% of all the leukoplakias are associated with a Candida infection. These cases are characterised by a higher frequency of epithelial dysolasia. The possible role of Candida albicans in epithelial dysplasia remains to he eluci ated. Furthermore, it apoears that some of the oral epithelial dysplasias are reversible. The IRC has also included in its work the epithelial dyselasias found in the oral mucosa in individuals with particular smohing hahits from the different parts of the world. IRC on Tumours of Eye and Ocular sdnesa This recently as t.ac lisher IRE has not vet hegun its work, so the presentation represented only the views of the teed of the centre- The pre cerous l_esion.s of the cornea and conjunctiva were considered under two categcries? those predisposing to the occurrence of squamous celi carcinoma, —-I -aid ose predisposing to the occurrence of malignant melanoma. In the first category the most trequent and mo : important is actinir keratosis (in the precancerous condition, seroderma oigmentosum. the conjunctiva along with the skin exhibits a greatLly increased tendency to develop actinic heratosis and squamous cell carcinoma]. Actinic keratosis, generally involving the most exposed parts of the conjunctiva, is characterised by histologic features similar to those observed in actinie {senile} keratosis f the skin. The affected epithelium typicallv is located over a oingueoula or pterygium. It is greatlv thickened and eahihits marked as raheratosis and orthokeratosis, varying degrees of cytologic abnormalities, e.g. enlarged hyperchromatic nuclei with prominent nucleoli, multi nucleated cells_ increased nuclearrcytoplasmic ratio, increased mitotic activity, abnormal mitotic figures, dyskeratosis, premature and irregular keratiniaation and a tongue~1ihe1ownward. proliferation ol‘ large hyperohromatic hasaloi d cells but without actual invasion. While the frequency with which actinio keratosis gives rise to squamous carcinoma is not known, there are reasons he believing that most carcinomas of the hulbar conjunctiva arise in areas of aLt.i nic heratcsis. These are usually well differentiated he! atini.ting sou nous cell carcinomas hat grow out from tiie surface as a large papillary mass {verrucous ca; cinema}. de:o invasion occurs late and metastasis is rare. Also in the first category there is the broad spectrum of lesions falling in the d ’splasia - carcinoma in situ growns, w—u—q—m—u—n—u— Predisnosing to the development of malignant melanoma are: congenital melanosis, naeti and acquired melanosis. Congenital melan r:sis, while frequently giving rise to melanomas of the uvea, only rarely gives rise to melanomas in the conjunctiva or lids Naevi and a-cnuired melanosis are the important r:rerur.st fl 1 ul' Hf If P {Jniinctivaal melanomas each accounting for about one—third of the cases {c-ne—tniro alsc coherently arising 'Hie novn‘ from seemingly normal conjunctivasl. THC on Shin Tumours In the skin a large number of precancerous oondi.‘r one soon . In some only clinical experience is evidence of the precancerous nature of ti {:J I ‘“ a condition ;e.v. raciationE . f .dermatitis, ohagedenic ulcers, tat andfor oil dermatoses. hurn scars, atc.}. casfts.1 page 7 In addition there are a number of precancerous lesions, the most important of which is actinic keratosis of which it is believed that 25% will show invasive squamous cell carcinoma in due time if left untreated. Even so extensive local destruction andfor metastases are extremely rare in this condition in contradistinction to the same histological type of carcinoma arising in scars or "de novo”. The squamous cell carcinomata arising in phagedenic ulcers form an intermediate group. Most behave like verrucous carcinoma but some will form metastases and kill the patient. In Bowen's disease of the skin the lesions are sharply demarcated and the whole thickness of the epidermis is involved with loss of polarity and stratification, hyperchromatism and clumping of noelei and a varying degree of clear cell vacuolation of epidermal cells. The lesions associated with arsenic poisoning are indistinguishable from classical Bowen's disease. Very similar histological features are sometimes observed in actinic keratoais. ' In about 5% of cases of Bowen's disease invasion of the dorm ultimately occurs but in only about 2% do metastases occur. Associated with Bowen's disease is a cancer proneness expressing itself in the later develcpment of other skin tumours and/or internal cancers. In erythroplasia of Queyrat the histological features of the lesion may be strikingly similar to those of Bowen‘s disease. Invasion takes place in about 14% and metastases are formed in about 4%; but cancer proneness has not been established. Intra-epidermal carcinoma of Jadassohn is a controversial entity as many unrelated intrarepidermal carcinomata and even irritated aeborrhoeic keratoses have been described under this diagnosis. However, the classical lesion occurs for 40% on the lower limbs and less than 1% will in due time show invasion. Paget's disease, in particular the extra—mammary form, shows a tendency to lateral extension intra-epidermally of either the continuous layer of Paget cells or of isolated cells or cell groups which are large, partly vacuolated and contain moon-polysaccharides. Both Junctional and compound naevi can he precursors to malignant melanoma but the incidence is completely unknown. In precancerous melanosis two forms are distinguished, the classical Hutchinson's melanotic freckle on the face and similar lesions but located elsewhere on the body. The histological features overlap to a great extent but the clinical prognosis varies greatly. While malignant melanoma develops later and usually_slowly in Hutchinson's melanotic freckle, the chance of malignant melanoma developing in a melanotic freckle elsewhere on the body is much greater and then bears a considerably worse prognosis. The possibility of Junctional naevi giving rise to a melanotic freckle which is followed by malignant melanoma certainly exists but the frequency of such occurrence is totally unknown. IRC on Male Urogenital Tract Tomours Relevant precancerous mucosal changes are individually identified (lrfi below) or collectively referred to as "unstable mucosa" (maladie de la muqueuse). These lesions may he focal, or diffuse and multifocal. The lesions listed are those that mar be considered precancerous. However. -arcinoma.. may arise ”de novo” without showing any of these changes. 1. Papillary hyperplasia {papillary cystitis) 2. Proliferative cystitis {von Bruno's nests) 3. Cystitis cystica and glandularis 4. Glandular change (metaplasia) 5. Squamous change (metaplasia) 6. "Nephrogenic adenoma” Transitional cell papilloma. This is defined as a tumour in which the epitheliu is indistinguishable from normal bladder in all respects; It constitutes about 3~fi% of epithelial tumours of bladder. It is regarded as the most common precancerous lesion of the bladder. "Inverted transitional pacilloma". he natural history of the somcalled inverted transitional papilloma has not yet been established, but this lesion mav also he precance The bases for regarding 1 to 6 as precancerous are that they: (a) precede the development of carcinoma in experimental animals; {h} precede chem cally induced cancers of bladder in man; to} they are sometimes followed by fully developed cancer in man and animals; (d) they occur at the advancing margins of fully developed cancers. Information is fragmentary regarding the percentage of these that become actually cancerous} IRE for Nomenclature in Cytologv The_IRC has completed its work on nomenclature of female genital tract cytology and this will be published soon. It includes definitions and descriptions of precancerous lesions and carcinoma in situ of the cervix and endometrium. Exfoliative cytology allows follow+up of a lesion with minimal or no disturbance of its biological behaviour. The IRE has now held its first meeting on the cytology of the extra genital sites. IHC for Gastro-Desophageal Tumours oesophagus The following may have significance as precancerous lesions and conditions: 1. Genetic conditions: e.g. tylosis (keratosis calmaris et plantaris) in which there is a high incidence of oesophageal carcinoma CAN/73 .1 page 9 2. Congenital and acquired heterotopic epithelium: gastric gland heterotopia Barrett oesophagus 3. Chronic irritation; submucous fibrosis leukoplakia (e.g. megaoesophagus) epithelium overlying mesenchymal tumours scar from corrosive oesophagitis 4. Sideropenic dysphagia (Plummer-Vinson oesophagitis) Stomach Precancerous lesions: 1. Villous adenoma - high potential for deyeloping carcinoma 2. Tubular adenoma (especially flatly elevated type) - moderate potential for developing carcinoma 3. Intestinal metaplasia - low potential for developing carcinoma 4. Chronic ulcer - questionable 5. Peuts—Jeghers' (and others} polyposis - low potential for developing carcinoma It must he pointed out that a majority of carcinomas apparently arise "de novo". Precancerous condition: 1. Pernicious anaemia Experimental aspects Experimental procedures in chemical carcinogenesis now permit many studies to be undertaken in large, controlled animal populations that may further clarify problems facing the pathologist diagnosing precancerous lesions and carcinc in situ. Although most of the early studies in this field were concerned with the investigation of early changes in the skin induced mainly by polycyclic aromatic hydrocarbons, more recent studies have involved many other organ systems including stomach, oesophagus, lung and bronchi, kidney, oral cavity. New carcinogens and the use of various methods of administration to various species now provide systems in which many specific problems can be studied. Studies now on record have already made many basic contributions that still require application. However, additional studies still remain to be done and there is unquestionably a need for joint planning of such investigations by pathologists in clinical practice and experimental carcinogenesis workers. In particular certain tissues with specific features such as the mucosa of the oral cavity, the eye, the cervix uteri and the urinary bladder require additional study. Finally, animal systems have only been used in a limited manner to study the progression andfate of tumours and their precursors. Such studies should be extended but are expensive and painstaking. canfla - 1 page 10 Follow-up studies It is important that a special effort be made to carry out follow-up studies to obtain more information on the natural history of precancerous lesions and carcinoma in situ in man. RECOMMENDATIONS 1. That WHU continue its programme on the histological definition of precancerous lesions. 2. That close collaboration he maintained between individual IRCs involved in this work. 3. That the definitions adopted in this report be broughtto the attention of the other IRCs dealing with histological classification of tumours. 4. That long term clinical follow—up studies be promoted to further elucidate the nature of precancerous lesions including carcinoma in situ, particularly in regard to their progression and regression. 5. That experimental investigations be conducted to obtain relevant data on the natural history of precancerous lesions including carcinoma in situ. 6. That more interaction occurs between experimentalists and other pathologists and that experimental findings and methods now available be evaluated for current use and additional applied experiments recommended to elucidate specific practical problems.