Document baR8XLL0qboMQZn8Xkv3agxpk

Ivmph m and -flanges tization Arch. pfwtics, -ondon, t 1 ( 1 PRIMARY MALIGNANT MESOTHELIAL TUMORS IN CANADA, 1960-1968 A Pathologic Review by the Mesothelioma Panel of the Canadian Tumor Reference Centre Alison D. McDonald, MD,* Desmond Magner, MD,' and Gail Eyssen, MSc* Six pathologists reviewed 119 of 165 cases reported as primary malignant mesothelial tumors, between 1960 and 1968, in a Canadian national survey. Based on histologic observations only, the panel was in favor of the diagnosis in 50%, uncertain in 14%, and against in 36%. This distribution was similar for pleural and peritoneal tumors. In males, the diagnosis was favored more often than in females--61% compared to 26%. Clinical and additional pathologic information changed the pathologists' opinion little. A member of the UK Mesothelioma Panel who reviewed 44 of the 119 cases recorded essentially simi lar opinions. Subsequently, a further 71 cases were registered between 1968 and 1970, and reviewed pathologically. The 99 cases in all, in which the panel favored the diagnosis of malignant mesothelioma, had been exposed to as bestos slightly more frequently than the rest. Twelve of 26 (46%) tumors classified histologically as mixed were associated with asbestos exposure com pared with 8 of 73 (11%) classified as epithelial or mesenchymal. SURVEY WAS CARRIED OUT, DURING 1968 A and 1969, of all fatal cases of primary malignant mesothelial tumors in Canada diag nosed between January 1960 and June 1968.4 Ascertainment was based on replies to an in- The research was assisted by grants from the Insti tute for Occupational and Environmental Research of the Quebec Asbestos Mining Association and from the National Cancer Institute of Canada. Associate Professor. Department of Epidemiology and Health. McGill University, Montreal, Canada. t Registrar. Canadian Tumor Reference Centre; Pro fessor and Head, Department of Pathology, University of Ottawa. Ottawa, Canada. t Lecturer, Department of Epidemiology and Health, McGill University, Montreal, Canada. Members of the Mesothelioma Panel of the Canadian Tumor Reference Centre who carried out the review: G. M. Fraser, MD, Dalhousie University, Halifax; R. Guy, MD, Hopitai du Sacre-Coeur, Montreal; M. Kannerstein, MD, Barnert Memorial Hospital Center. Paterson, N.J., and Mount Sinai Hospital, New York; R. Lefebvre, MD, Hotel-Dieu, Montreal; D. Magner, MD, Canadian Tumor Reference Centre and University of Ottawa; J. P. Tremblay, MD, Hopitai St-Vincent de Paul. Sherbrooke. The review was organized by A. D. McDonald, MD and S. Ducic, MD of the Department of Epidemiology and Health, McGill University, who also studied the certified causes of death. P. Bolduc, MD, helped in se lecting controls and G. Eyssen, MSc, in analyzing the results. Later, W. T. E. McCaughey, MD, of the United Kingdom Mesothelioma Panel reviewed 44 cases. Address for reprints; A. D. McDonald, MD. Depart ment of Epidemiology and Health, 1110 Pine Ave. West, Montreal 112, Canada. The authors wish to thank all pathologists in Can ada for their collaboration. Received for publication September 22. 1972, quiry about these tumors received from every pathologist in Canada who was a member of the Canadian Association of Pathologists or of the Quebec Association of Laboratory Physicians. The 165 cases studied were those which the reporting pathologists thought on balance were primary malignant mesothelial tumors. The methods used to assemble the cases and two matched control series and to study them epidemiologicaliy have been de scribed.4 In brief, detailed occupational, resi dential, and smoking histories were sought from relatives and friends of the deceased. The field workers did not know which were cases and which controls, and the information collected was coded "blind." An association between primary mesothelial tumors and an occupational history of asbestos exposure was found in 20% of the 101 male cases for whom a questionnaire was completed compared with 3% of the controls and in 2% of the 51 female cases compared with none in the controls. The survey subsequently has been continued in a similar manner and, between July 1968 and December 1970, 71 cases were registered --39 male and 32 female.5 Occupational or do mestic exposure was found in 32% of males compared with 11% of controls and 13% of females compared with 3% of controls. Thus, although all rates were higher, the excess asbestos exposure was of a similar order. No 869 HWBUI0006949 870 Cancer April 1973 Vol. 31 differences have been found in occupational exposure of lesser degree and history of resi dence in an asbestos mining area. Almost all the excess of occupational exposure was in the manufacture and industrial application of asbestos rather than in mining and milling. Pathologic studies have been carried out in three phases. Phase 1: A pathologic review of the original 165 cases was planned to find out: 1. the joint opinion of members of the Canadian mesothe lioma panel as to whether the cases studied were indeed primary malignant mesothelial tumors; 2. how much weight was given by the panel to histologic appearance alone, and to what extent did clinical, operative, surgical pathology, and autopsy reports modify the his tologic diagnosis; 3. how much observer varia bility there was among members of the panel, and 4. whether a history of asbestos exposure was related to the diagnostic opinion of the panel or to characteristics of the tumor. Phase 2: The additional 71 cases were re viewed pathologically. Phase 3: Cases accepted by the panel as me sotheliomas were classified histologically. Materials and Methods Phase 1: The pathologists who had reported the 165 cases were asked to send in paraffin blocks or tissue in fluid fixative. Material was obtained for 125 cases--in 3 it was known to have been lost and in 37 none could be found. In nearly all cases the specimens received were in the form of paraffin blocks of formalinfixed tissue, taken at autopsy in 89 cases and at operation in 36 cases. The pleural surgical material varied from small portions of pleura, with or without subjacent tissue, to pneumo nectomy specimens. In only one peritoneal tumor had the diagnosis been made on the basis of a surgical specimen. This consisted of a tumor mass measuring 7.0 x 3.5 cm from the right iliac fossa. Six sets of slides of each of the 125 cases were prepared at the Canadian Tumor Refer ence Centre. Slides were stained with hema toxylin and eosin. In each case, one to three additional slides showing tumor were stained with mucicarmine and by the PAS method after diastase digestion. The contributing pathologists were asked to furnish a clinical summary including com ments on duration of illness, presenting symp toms, presence or absence of pleural or perito neal effusions, other physical signs, x-ray findings, and therapy. Although the informa tion provided varied in amount and was not extensive, a synopsis of the main clinical events was given in most cases. The pathologists were also asked to describe the gross pathology of surgical specimens and the gross findings at autopsy. This request was usually met by cop ies of surgical pathology reports or, for autop sies, by a description of the mesothelial tumor and a list of the main autopsy findings, in cluding sites of metastases. Six members of the Canadian Tumor Refer ence Centre Mesothelioma Panel, which in cluded an American pathologist who was also a member of the U.S. panel, took part in the review. The study was carried out during a period of 4 days in the Department of Epide miology and Health, McGill University. The pathologists worked independently in separate rooms and were helped by clerks who checked the case numbers and recorded observations. Each pathologist reviewed cases in the same order, first recording his opinion on histology alone and then, after reading the related clini cal information and description of the gross pathology (but not any occupational history), his final judgment. The additional informa tion usually allowed pathologists to make some estimate of how closely the macroscopic appearance of the tumor resembled their con cept of a typical mesothelioma and to what extent the possibility of a primary tumor else where could be discounted. At each stage the pathologists were asked to classify cases into one of the five categories given below. In order to introduce an element of doubt into each diagnosis, comparable specimens from 17 cases with pleural or peritoneal metastases from known extra-thoracic primary tumors were introduced into the series. It was decided not to search for asbestos bodies. Classification by pathologists 0. Not a primary mesothelial tumor 1. Unlikely to be a primary mesothelial tumor {--tending towards unlikely unable to decide + tending towards probably 3. Probably a primary mesothelial tumor 4. Definitely a primary mesothelial tumor Simplified version against uncertain in favor ed od to miptoray nalot nts ere of at )paplor in- 'erinIso :he ;a de ns. me >gy nioss 7). naike pic aniat A.se:he j no In t ito 17 ses ors led tt i ( If they used category 2 (possible) they were asked to indicate by a -- or a + whether they leaned towards unlikely or probable. A simpli fied version of the classification was derived as indicated above and the six opinions on each case examined. If a majority was against or in favor of the diagnosis the case was classified as such and if opinions were evenly balanced, as uncertain. A panel opinion was obtained for 119 of the 125 cases reviewed. In three of the remaining six cases, no tumor was seen by all six pathologists. In the other three cases, the number of pathologists who found no tumor was five, four, and two, and the cases accepted as a mesothelioma by one, none, and three of the pathologists, respectively. Clinical reports were missing in four other cases so a final di agnosis was made in only 115. A few months after the panel had finished its work, there was opportunity for a patholo gist who was a member of the United King dom Mesothelioma Panel to review in the same manner about a third of the series. For this, 44 cases were selected by a random proce dure from the three groups in favor, uncer tain, against, into which the cases were classi fied by the panel. Phase 2: Essentially the same procedure was followed in the classification of the 1968-1970 cases, but material was circulated and no con trols used. Phase 3: All cases accepted by the panel as mesotheliomas for whom a full occupational and environmental history had been ob tained--59 in the original series, and 40 in the subsequent series--were classified histologi cally by one member of the panel (DM) with out knowledge of the occupational history. Findings Phase 1: The results of the review of 119 cases are shown in Table 3. The panel was in favor of the diagnosis on histologic grounds alone in 61.3% of males and 25.6% of females and at the final stage they were in favor of the diagnosis in 62.8% of males and 35.1% of fe males. The panel was against the diagnosis of primary mesothelial tumor in 16 of the 17 cases of secondary pleural or peritoneal tu mors with an identified primary elsewhere. In Table 2, a classification is presented of all 690 individual opinions of the panel mem bers on the 115 cases reviewed both histologi cally and with clinical and gross pathologic in formation. When, on histologic appearance, pathologists tended to be in favor of the diag nosis, the additional information resulted in an increased sureness. In cases where they were against the diagnosis, the additional in formation resulted in some changes equally for and against. Using the simplified classifica tion, in only 21 cases was a change made in the panel opinion after consideration of the clinical and gross pathologic data, 13 being moved up a group and 8 down a group, but the overall proportions remained much the same. In the 78 male cases with complete clinical and other information, the panel was in favor of the diagnosis in 62% of 63 pleural and 64% of 14 peritoneal tumors. The correspond ing figures for females were 37% of 19 pleural and 29% of 14 peritoneal tumors. They also accepted one male, one of two female pelural and peritoneal tumors, and one of two female pericardial tumors. The availability of clinical Panel opinion of diagnosis In favor No Per cent Table 1. Results of Pathologic Review* Males 49 (46) 61.3 Histology only Females 10(10) 25 6 based on Histology and clinicopathologic information All 59(56) 49.6 Males 49(46) 62.8 Females 13(13) 35.1 All 62 (59) 53.9 Uncertain No Per cent 12(11) 15.0 5(3) 12.8 17(14) 14.3 8(7) 10.3 4(3) 10.8 12(10) 10.4 Against No Per cent 19(17) 23.8 24(21) 61.5 43 (38) 36. 1 21 (19) 26.9 20(16) 54. 1 41 (35) 35.7 All 80 ( 74) 39 (34) 119(108) * Figures in parentheses indicate cases studied epidemiological!)'. 78(72) 37 (32) 115(104) HWBUI0006951 872 Cancer April 1973 vl- 31 Table 2. Classification by the Six Pathologists on Histologic Appearances (115 Cases--690 Observations) before and after Study of Additional Clinical and Pathologic Information 121 105 64 12 112 156 120 Total fNot Against j Unlikely Primary classifiestion on histologic Uncertain appearance [Possible -- Possible (Possible -f In favor j Probable (Definite 96 9 1 15 67 16 3 19 31 i 14 4 7 11 211 1 1 107 1 93 111 1 17 7 78 7 7 4 1 25 1 72 44 12 151 2 7 97 59 169 1 47 49 Revised classification after review of clinical and pathologic information Z C rr *2<. rs v> ft "3 g cr I "3 is fstT V 5! 4- O 690 ft e>ra C3 So. dea or res' (15 to th< lit* tiv 19 ce 7C ] P O' b c 2 v VCT information and an account of the gross pa In the 44 cases selected for the British pathol thology did not raise the proportion of favora ogist, there was rather less agreement between ble opinions which amounted in males to individual members and the panel, but the 61% of 59 autopsy cases and 68% of 19 cases overall picture was similar. The opinions of in which there was only a biopsy and in fe the pathologist from the British panel resem males 33% of 24 autopsy and 31% of 13 biopsy bled those of one of the two pathologists less cases. in favor of the diagnosis than the rest of the ti Variation in opinion between pathologists panel, but his opinion fell within the range is shown in Table 3. Using the simplified clas observed among the members of the Canadian i\ sification, they each agreed with the panel panel. opinion in about 70% of cases. In the remain Death certificates were inspected for 107 of ing 30%, two members of the panel tended to the 115 cases classified by the panel on histo be more in favor of the diagnosis and two less. logic and clinical grounds. Of these, 49 (46%) / Table 3. Percentage of Cases in Which Individual Pathologists Agreed with the Panel Opinion Disagreement 1. In 115 cases with clinical information 2. In 44 selected cases -more in favor Pathologist Agreement of diagnosis A B C D E F A B C D E F W. McC. 70 4 7' l 61 0 7.' 2 .7 7/.1 (' .4 6-.6 51.6 S'). 1 .1.6 63.6 59.1 22.6 21.7 14.8 11.3 9.6 0.9 27.3 29.5 15.9 18.2 13.6 0.0 13.6 -less in favor of diagnosis 7.0 6.1 18.3 16.5 21.7 27.0 11.4 6.8 29.5 22.7 22.7 36.4 27.3 HWBUI0006952 deaths were certified as due to mesothelioma even within the three categories listed, but de or connective tissue tumor, 32 (30%) to other tailed descriptions will not be given here. Sev i! respiratory or gastrointestinal tumors, 16 eral microscopic patterns are illustrated (Figs. (15%) to tumors of other organs, and 10 (9%) 1-8). to non-malignant disease. The 60 cases that In epithelial tumors, the cells were gener the panel accepted as mesotheliomas were a ally oval, cuboidal, or polygonal with vesicu little more often certified as such or as connec lar nuclei and acidophilic or amphophilic cy tive tissue tumors (48%) than the 36 cases that toplasm, occasionally vacuolated. These the panel rejected (33%). formed tubules, sometimes lined by flattened Phase 2: Of 71 cases registered between July cells, and wider, irregular spaces into which 1968 and December 1970, 42 (59%) were ac papillary processes projected. Frequently, the cepted by the panel; they were uncertain in cells grew in solid masses with minimal evi 7% and rejected 34%. dence of tubules, larger spaces, or papillary Phase 3: The tumors in which the panel processes. Evidence of neutral mucopolysac opinion was in favor of mesothelioma dis charide (epithelial mucin) production was not played patterns which have been described by established. The stroma varied in amount and others.12-5 Of the 70 pleural tumors accepted was fibrous. In mesenchymal tumors, spindle- by the panel and investigated epidemiologi- shaped or fusiform cells with elongated, oval, cally, 22 were epithelial, 27 mesenchymal, and vesicular, or hyperchromatic nuclei were ar 21 mixed; of 23 peritoneal tumors, 16 were ep ranged in parallel fashion in sheets, fascicles, ithelial, 2 mesenchymal, and 5 mixed. Eight and whorls. Some tumors were highly cellular; cases showing approximately equal mesothe- others contained moderate or considerable liomatous involvement of pleura and perito amounts of collagen, laid down in sheets or neum were epithelial and a pericardial tumor, enlacing bands, often with hyaline, relatively mesenchymal. There were wide variations in acellular areas. In mixed tumors, epithelial the microarchitecture of individual tumors, and mesenchymal elements were usually inter- f V Fic. 1 (left). CTRC 14037. Malignant pleural mesothelioma, tubular epithelial pattern. The tubules are lined by cuboidal and flattened cells (x400). Fic. 2 (right). CTRC 14065. Malignant pleural mesothelioma, tubulopapillary epithelial pat tern. A group of small tubules are present at the top of the photograph. Below, papillary proc esses project into a space lined by tumor cells (X400). HWBUI0006953 874 Cancer ^pri7 1973 Vol. 31 Si *i 3* . Fic. 3 (Uft)- CTRC 13056. Malignant pleural mesothelioma, solid epithelial pattern. The tumor cells vary in size, some being very large with abundant cytoplasm. Nucleoli are prominent in many nuclei (x40O). Fig. 4 (right). CTRC 14051. Malignant pleural mesothelioma, mesenchymal pattern. The elongated tumor cells are arranged with their long axes parallel. Nuclei are oval or fusiform. Collagen fibers are inconspicuous (x40Q). HWBUI0006954 No. 4 Malignant Mesotheliomas in Canada McDonald ci al. 875 Fig. 7 (left). CTRC 14066. Malignant pleural mesothelioma, mixed pattern. The top half of the photograph shows the tubules formed by tumor cells of epithelial type. The bottom half, from an area in the same tumor, shows spindle-shaped tumor cells of mesenchymal type (x400). Fic. 8 (right). CTRC 14031. Malignant pleural mesothelioma, mixed pattern. The top third of the photograph shows a solid epithelial area with one raultinudeated tumor cell. In the middle third, from an area in the same tumor, are spindle-shaped tumor cells of mesenchymal type. In the bottom third, from a hilar lymph node in the same case, metastatic mesothelioma forms a tubular epithelial pattern (X400). mingled throughout the tumor but sometimes occupied different but adjacent territories. In several tumors, particularly those of pleural origin, undifferentiated areas were en countered in which the cells were pleo morphic and varied considerably in size and nuclear characteristics. However, even in these cases there were zones of sufficient differentia- Table 4. History of Definite or Probable Occupational or Domestic Exposure to Asbestos Related to Panel Opinion of Diagnosis Males Females Panel opinion of diagnosis Total Asbestos Exposure No. Percent Total Asbestos Exposure No. Percent In favor 67 17 25.4 32 3 9.4 Uncertain 10 4 40,0 5 1 20.0 Against 32 6 18.8 27 1 3.7 All 99 27 27.3 63 5 7.9 ------------------------------------------ ----------------------------------------------------------------------------------- ------- -------------------m Fig. 5 (left). CTRC 12092. Malignant pleural mesothelioma, mesenchymal pattern. Elongated tumor cells form enlacing strands with moderate collagenization (x400). Fig. 6 (right). CTRC 13503. Malignant pleural mesothelioma, mixed pattern. The top half of the photograph shows spaces lined by cuboidal tumor cells of epithelial type. The bottom half, from an area in the same tumor, shows spindle-shaped tumor cells of mesenchvma) type (X400). HWBUI0006955 876 Cancer April 1973 v0i. 31 Table 5. Histologic Classification and History of Asbestos Exposure in the 99 Cases Accepted bv the Panel (1960-1970) Definite or probable exposure to asbestos Occupational Domestic Epithelial 3 -- Males Mesenchymal 3 -- Mixed 11* 1* Epithelial -- 1 Females Mesenchymal -- 1 Mixed _.~ 1 N umber Percent 3 11.5 3 17.6 11 45.8 5.2 9.1 50.0 Total cases 26 17 24 19 n 2 * One man who worked as a geophysical explorer for an asbestos company had also been exposed domestically. i I tion to allow classification into one of three histologic types with fair certainty. Exposure to Asbestos A history of definite or probable asbestos exposure was slightly more frequent in cases in which the panel, after consideration of all available information, was in favor of the di agnosis than in cases in which they were against it or uncertain (Table 4). Table 5 shows the histologic classification and history of asbestos exposure in all 99 ac cepted cases. Mixed tumors were more often associated with asbestos exposure than the other two groups, and this type of tumor was infrequently found in women. Six of the 11 occupationally exposed men with mixed tu mors had worked in insulation or asbestos products manufacture compared with one of the 6 men with epithelial or mesenchymal tu mors. \ \ ADDENDUM Since this paper was submitted for publication, some of these data were included in a letter to the editor of the New England Journal of Medicine, 287:570-571, 1972. { REFERENCES 1. Churg, J., Rosen. S. H., and Mootten, S.: Histo logic characteristics of mesothelioma associated with as bestos, Ann. IV. y. Acad. Sri. 132:614-622, 1965. 2. Hourihane, D. O'B.: The pathology of mesothe lioma and an analysis of their association with asbestos exposure. Thorax 19:134-143, 1964. 3. McCaughey, W, T. E.: Primary tumors of the pleura. J. Path. Bacteriol. 76:517-529, 1958. 4. McDonald, A. D.. Harper, A.. Attar. O. A., and McDonald, J. C: Epidemiology of primary malignant mesothelial tumors in Canada. Cancer 26:914-919, 1970. 5. McDonald, A. D., and McDonald, J. C.: Epidemiclogic surveillance of mesothelioma in Canada. In preparation. ( t j I I \ 1 I HWBUI0006956