Document dYzExawL1ZK1z2m6Gjgyno5VR

FILE NAME: Asbestos Textile Product Use (ATPU) DATE: 1984 Sept DOC#: ATPU042 DOCUMENT DESCRIPTION: Journal Article - Malignant Mesothelioma Clinical and Epidemiological Features Reprinted from the " S i4 MedicalJournal", Vol 6 6 ,1 5 September 1984, p p 407 - 412 Malignant mesothelioma -- clinical and epidemiological features A report of 80 cases K. SOLOMONS D epartm ent of Occupational Medicine, National Centre far Occupational Health, Johannesburg K. SOLOMONS, m 8 c h b South African mines produced 249187 metric tons of asbestos in 1979, of which 53% was crocidolite, 31% chrysotile and 16% amosite. Exports accounted for 95% of the total production, the remaining 5%being used locally. While the 12 050 miners (Table I) who mined this tonnage are at risk of developing asbestosinduced diseases, it is not possible at present to estimate the numbers o f people in the rest of the population who are at a similar risk from either occupational or environmental exposure. Almost 100 new cases of mesothelioma are confirmed nation wide by the South African Asbestos Tumour Reference Panel (referred to as the Panel) each year. This figure, which excludes asbestos tumours other than pleural and peritoneal meso theliomas, underestimates the true incidence of mesothelioma in the country. 407 408 SA MEDIESE TVDSKRIF DEEL66 15 SEPTEMBER 1984 TABLE I. ASBESTOS MINING DATA Miners*______ '______ _________________ Asbestos production (metric tonnes) Year 1977 1978 1979. 1980 1981 1982 Total 23367 13311 12050 12626 11316 10724 White 1096 718 678 706 617 649 Coloured and Black 22271 12 593 11372 11920 10 699 10 075 Totalf 380164 257 325 249187 276 734 235 943 -- Amositet -- -- 39058 -- 56834 -- Chrysotile -- -- 91 828 -- 76772 -- Crocidohte -- -- 118301 -- 102337 -- * Bulletin of Statistics Pretoria Central Statistical Services, 1979,1982,1983 \South African Statistics Pretoria Central Statistical Services, 1982 11 7 XSouth African Mining and Engineering Yearbook Johannesburg Jim Emery. 1981 16,1982 21 Valuet (R1 000) 137779 121682 107 048 102148 117335 -- M ost cases diagnosed in the Johannesburg/W itwatersrand area are referred for an asbestos-exposure history to the clinic attached to the National Centre for Occupational Health. Between 1972 and June 1983 150 documented cases of meso thelioma were seen at the clinic. A repon on the first 70 cases was published by Cochrane and W ebster' in 1978. This article reviews the findings in the subsequent 80 cases. The objectives of the review are to examine the epidemiology of mesothelioma in South Africa and the value of clinical and epidemiological features as prognostic indices for survival. Subjects and methods All cases of malignant mesothelioma proven histologically from biopsy specimens and seen by the clinic between January 1977 and June 1983 were included in the study group. Data oh each subject were collected from the clinic's files. Where the records were incomplete, additional data were obtained from general practitioners, thoracic surgeons, hospitals, the Panel, the Workmen's Compensation Commissioner, the South African Transport Services, the Medical Bureau of Occupational Diseases, employers and family members. All cases of meso thelioma confirmed by the Panel over the same period, excluding those in the study group, were also evaluated. Only basic data on age, sex, race and exposure history were available for this group and they were not followed up for survival data. The function of this reference group was to act as a body to which clinical findings in the study group could be generalized, if it resembled the study group in significant features, and to provide a broader population from which epidemiological data could be gathered. Diagnosis was either made by the Centre's pathologists or subsequently corroborated by them. The tumour type and the presence or absence of parenchymal asbestosis were also deter mined on histological examination of biopsy material which was usually obtained at open thoracotomy. Asbestos-exposure histories were taken from all patients or their families at the request of the attendant thoracic surgeon or the Centre's pathologists. Exposure history was considered positive if the patient recalled working with or being environmentally exposed to asbestos. Environmental exposure included childhood, domestic, neighbourhood or any other definite exposure which was not occupational. Possible exposure was recorded if the patient had no recall of specific asbestos exposure, b u t when he/she had either worked in an industry or spent time' in an environment where asbestos contamination was likely. Exposure was considered negative if neither positive nor possible exposure were recorded. Negative exposure means that no exposure history was elicited, not that exposure had never occurred; exposure could have occurred in infancy or childhood, or been negligible or obscure. Patients were staged retrospectively from data obtained from surgeons' operative findings and notes at the time of diagnosis, according to the system advocated by M attson2 (stage I -- ipsilateral pleura and lung only; stage I I -- chest wall invasion, mediastinal or pericardial involvement, contralateral lung or pleural involvement; stage III -- extrathoracic extension: (a) nodes outside chest, (b) diaphragm atic penetration t perito neum; stage IV -- distant metastases). It must be emphasized that staging was not an active process and that no standard staging system was used by the different surgeons. Treatment was broadly divided into the major modality used rather than specific agents and regimens since these varied widely, largely because treatm ent was not carried out at a single centre with standardized regimens. The data were analysed statistically by a variety of techniques including analysis ofvariance and product-limit survival analysis. Percentages have been rounded off to the nearest point. Results The study group comprised 80 subjects. The mean age at the tim e of diagnosis was 57,2 years and the range 24 - 87 years. T he male :female ratio was 10:1, the group consisting of 73 men (91%) and 7 women (9%). T here were 64 Whites (80%), 13 Blacks (16%) and 3 Coloureds (4%). One subject had a primary peritoneal mesothelioma, but the pleura was the primary site in the rest. These mesotheliomas arose on the left side in 37 cases (46%) and on the right side in 40 (50%). No data were available in the remaining 2 cases. In 72% of cases (58/80) the Panel had confirmed the diagnosis of malignant mesothelioma; 3 patients had never been subm itted and 19 patients had been submitted bu t not yet examined (since submission o f'th is article the diagnosis of mesothelioma was confirmed by the Panel in these latter 19 patients). A positive history of previous asbestos exposure was obtained from 71 subjects (89%); 4 (5%) gave a history of possible exposure and in 5 (6%) no history of prior exposure was elicited. Similar studies in the past have found between 18% and 97% positive asbestos-exposure histories.3 In 59 cases (74%) there was a history of only occupational exposure, 7 (9%) had a history of only environmental exposure and 5 (6%) had a history Of both occupational and environmental exposure. The remaining 9 (11%) had either no exposure or only possible exposure. The 4 subjects with possible exposure had possible occupational exposure, one as a boilermaker, another as a fitter and turner on naval ships, one handling construction material and the fourth as a m arket agent handling hessian sacks which had previously been used to transport asbestos. Among those 68 subjects with occupational exposure, 15 (22%) were exposed in the mining industry, 15 (22%) on the SA MEDICAL JOURNAL VOLUME 66 15 SEPTEMBER 1984 . 409 railways, 8 (12%) in the construction industry, 6 (9%) in the from first exposure to diagnosis was 34,3 years. In one case the marine-engineering industry, 2 each in the electrical, engineering lag period was as short as 4 years and in another as long as 68 and battery industries, 7(1096) in the primary asbestos manufac years. There is a possibility that the person with the 4-year lag turing industry and 11 (16%) in other assorted industries (Table had environmental exposure to asbestos as a child, but this could II). The activities among those subjects occupationally exposed not be confirmed by a carefully taken history. included lagging of boilers and pipes in 23 cases (34%), At the time of diagnosis 29 (43%) of the 68 subjects whose underground mining in 8 (1296), surface mining jobs in 5 (7%) smoking status was known were still smoking, 8 (12%) had and transport o f raw asbestos in a further 6 cases (9%). Drilling, \ stopped smoking within 10 years of their presentation, 9 (13%) , cutting and sawing of asbestos products (sheets, boards, cloth) / had stopped more than 10 years before their presentation and 22 ' accounted'for exposure in 11cases (.16%) and the use of asbestos (32%) were lifelong non-smokers. The smoking status of the i gloves, blankets and aprons for another 3 cases (5%). One of this remaining 12 subjects was not established. \ latter group of 3 had further exposure from the asbestos present , Dyspnoea and chest pain were the presenting symptoms in in asbestos-coated welding rods. The work activities of th e ' 75% of all cases, dyspnoea alone in 23 cases (29%), chest pain / remaining 12 (18%) were not specified. alone in 14 cases (18%) and both in 22 cases (28%). In 4 cases the ` The types of asbestos involved were predominantly mixed, tumour presented incidentally, 3 at routine chest radiography although Cape crocidolite as the sole type was documented in 17 and 1 at unrelated abdominal surgery. The remaining cases cases. Amosite asbestos from the Penge mine was implicated as, presented in various ways including recurrent chest infections, the sole type in 4 cases. A fifth patient with only amosite H orner's syndrome, chest swelling, abdominal mass and ascites. exposure was excluded from the study group because the Panel Associated symptoms included weight loss, cough, tiredness and diagnosed squamous carcinoma of the bronchus and not meso haemoptysis. Pleural effusions, detected clinically and radiolo- thelioma. Similarly a patient with squamous carcinoma of the bronchus (an amosite miner) was excluded from the reference gically, were present in 42 cases (56%) and absent in 33 (44%). Parenchymal asbestosis, established on histological examina group. The reference group included 8 cases of mesothelioma tion oflung tissue specimens was present in 18 cases, absent in 42 with amosite exposure. N o further data on asbestos type were and unknown in the remaining 20. The histological type of available. tum our was established in 36 cases (45%). Of these, 29 (80%) The mean duration of asbestos exposure was 13,6 years, but were epithelial, 2 (5%) were sarcomatous and 5 (14%) were ranged from 1 week to 47 years. The mean lag period of time mixed. The stage of disease at diagnosis was ascertained in 50 cases (63%). O f these, 9 (18%) were in stage 1 ,25 (50%) in stage II, 14 (28%) in stage III and 2 (4%) in stage IV. TABLE II. SOURCE OF AS8EST0S EXPOSURE No. of patients A number of cases had evidence of haematogenous and distant tumour spread, which confirms reports that malignant meso thelioma invades the bloodstream and metastasizes (Table III).4 Mining In 5 of the 10 cases on which full autopsies were performed, Mining 8 internal abdominal organs were affected by haematogenous Milling and sorting 4 spread. In the 5 remaining cases spread was confined to Transport 2 intrathoracic organs. The superior vena cava syndrome and Shaft-sinking 1 Horner's syndrome were observed in 4 and 2 cases respectively. Railways Workshops 11 Transport Construction Carpentry 4 TABLE ill. SITES OF SPREAD IN 52 CASES 3 No. % Construction site 3 Ipsilateral lung 32 62 Plumbing 1 Chest wail 26 50 Asbestos insulation spray 1 Thoracotomy scar 11 21 Marine engineering Pericardium 12 23 Naval shipyards 3 Opposite pleura 3 6 Non-naval engine-rooms 3 Opposite lung 12 23 Engineering 2 Mediastinal structures (lymph nodes, ' Electricians 2 great vessels, myocardium) 16 31 Battery industry Diaphragm 14 27 Manufacture 1 Peritoneum 6 12 Destruction 1 Intestine 5 - 10 Asbestos board and sheet manufacture S Liver B 15 Asbestos manufacturers' agents 2 Kidney 2 4 Fitting and turning 4 Adrenals 1 2 Brake-lining repairs 1 Spleen 1 2 Stove manufacture 1 Mesentery 2 4 Cem ent m anufacture 1 Abdominal wall 1 2 Mattress manufacture 1 Scalp 1 2 Bottle-making (asbestos belts and pad repair on conveyor belt) 1 Foundryman 1 Market agent (asbestos-contaminated hessian bags) Environmental Nit known Total Information about treatm ent was known in 69 of the 80 cases. 1 A total of 5 different therapeutic modalities were employed, 7 either alone or in a variety of combinations. These included 5 radiotherapy in 22 cases, systemic chemotherapy in 20 cases, 80 surgery in 23 cases, immunotherapy with BCG in 11 cases and intrapleural mustine in 26 cases. In 29 cases only one of these 410 SA MEDIESE TYDSKRIF DEEL 56 15 SEPTEMBER 1984 methods of treatment was used. In 19 cases a combination oftwo methods was used, in 12 cases a combination o f three methods and in 2 cases a combination of four different m ethods. In 5 cases no .anticancer therapy was given. Radiotherapy and systemic chemotherapy were used in combination, on their own, and with other methods in a total of 18 cases. Survival data were available in 80% of cases, and recorded in months from diagnosis to death and from onset of symptoms to death. The mean survival time from diagnosis to death was 8,6 months and ranged from 1 week to 64 months; 85% of subjects died within 12 months of diagnosis. The mean survival time from the onset of symptoms to death was 13,6 months and ranged from 3 to 64 months. The mean duration of time from onset of symptoms to diagnosis was 5,5 months and ranged from 2 weeks to 21 months. Only 23 o f the subjects are on record as having been compensated for mesothelioma. In terms of the legislation, 25 of the subjects (31%) were not eligible for compensation, but-the remaining 55 (69%) were all eligible. Only 27 of these subjects are known to have claimed compensation; 4 were turned down and the remaining 23 were compensated. Compensation was awarded to 3 people who had both mesothelioma and asbestosis at a time when mesothelioma had not yet become a disease for which compensation was paid, b u t when asbestosis was. Compensation was paid to 6 miners (5 W hite, 1 Black), 11 railway workers (all W hite) and 6 industrial workers (5 W hite, 1 Black). I t seems that White railway workers and miners receive compensation in a high percentage of cases, b u t Black miners and railway workers and most industrial workers are less frequently compensated. At the end of the survey period 15% o f the subjects in the group were still alive, 19% had been lost to follow-up and 66% had died. Reference group T he reference group comprised 546 cases, in 505 of which the diagnosis of mesothelioma had already been confirmed by the Panel; 41 were yet to be examined by the Panel. The Panel had confirmed 481 of the 505 cases as definite mesothelioma and 24 as probable mesothelioma. T he mean age o f the control group was 55,1 years and ranged between 20 and 90 years. The sex o f 14 subjects in the reference group was not known, but of the remaining 532 the male:female ratio was 2,9:1 -- 395 males (74%) and 137 females (26%). The racial classification of 14 cases was not known, but of the remaining 532, Whites comprised 43% (228 cases) of the group, Coloureds 20% (107 cases) and Blacks 37% (197 cases). Sketchy positive exposure histories were available in 307 cases (56%), and possible exposure histories in a further 71 cases (13%). No history was available for the remaining 168 cases (31%). Occupational exposure occurred in 213 subjects (69%) and environmental exposure in 94 (31%) of the 307 with a positive history of exposure. Mixed environmental and occupational exposure was noted in 13 of these 307 cases (4%). Mining accounted for exposure in 143 (67%) of the 213 occupational exposure cases and industrial exposure for the remaining 70 cases (33%). The histological type of tum our was known in 89 cases (16%); 58 (65%) were epithelial, 16 (18%) were sarcomatous and 15 (17%) were mixed. Parenchymal asbestosis was documented in 41 cases (8%). N o further data on asbestosis were available since no parenchymal biopsy material was submitted for histological examination in the remaining cases. It is not known how prevalent asbestosis was among the other 505 subjects. Histories of cigarette smoking were available in 88 cases. Sixty subjects had a positive history of smoking and the remaining 28 were non-smokers. Results of statistical analysis The results of product-limit survival analysis indicated that the outcome or duration of survival was unaffected by the stage of the disease, treatm ent, presenting symptoms, smoking status, side on which the lesion occurred, presence or absence of asbestosis and pleural effusion, lag period from first exposure to diagnosis, histological type of tum our or source of asbestos exposure (Table IV). T he num ber of cases may have been too small for significance to be detected in the latter two categories. Significant differences in outcome were detected with regard to age at diagnosis, duration of exposure to asbestos and the number of therapeutic modalities used in treatment. The mean duration of survival from diagnosis in patients over 50 years of age was 14,9 months, while that for patients under 50 years was 7,4 months (P = 0,03). This difference disappeared, however, when survival was measured from the onset of symptoms (P = 0,13). Similarly, when survival from diagnosis to death was correlated with the duration of asbestos exposure, patients with TABLE IV. FACTORS PREDICTING OUTCOME Factor P value -- diagnosis to death* P value -- onset of symptoms to death f Histological tumour typej: Source ot exposure Smoking Side ot tumour Asbestosis Lag period Presenting symptoms Effusion Stage Therapy Age at diagnosis! Duration of exposure! No. of therapeutic modalities! 0,0951 0,5023 0,6097 0,0522 0,3934 0,4702 0,2908 0,8232 0,9587 0,9423 0,0310 0,0340 0,0000 - 0,0003 -- 0,2509 0,7747 0,1186 0,1440 0,1503 0,7563 0,2856 0,8160 0,8063 0,1341 2,2912 0,9845 Correlation between factor and survival time from diagnosis to death expressed as a P value fCorrelation between factor and survival time from onset of symptoms to death expressed as a P value (Numbers too small to exclude significance Tendency towards significance, i e subjects with right-sided pleural mesotheliomas had a tendency to survive longer than subjects with left-sided tumours -- 15.4 months as opposed to 8,63 months respectively ^Significant (P<0.05) SA MEDICAL JOURNAL VOLUME 66 15 SEPTEMBER 1384 411 exposure of between 30 and 40 years survived for a mean of 22,1 months whereas patients with less than 1 year's exposure survived for a mean duration of 5,6 months (P = 0,03). This difference disappeared when survival was measured from the onset of symptoms (P = 0,29). Patients treated with one or more anticancer modalities survived significantly longer from diagnosis than untreated patients (F< 0,0001). While patients treated with four modalities survived significantly longer than patients treated with one or three modalities (P = 0,01 and 0,03 respectively), they had no advantage, as regards survival over patients treated with two modalities (P --0,08). All these differences, however, disappeared when survival was measured from the onset of symptoms. x The number of cases was too small to permit testing of the effects of combinations of factors such as stage and treatment or stage, treatm ent and histological type on outcome. This drawback restricts the interpretation of both negative (stage and therapy) and positive (age and therapeutic modalities) associations. The study group resembled the reference group with regard to> age (57,2 years and 55,1 years, P = 0,16) and histological type of tum our'(P = 0,21), but had significantly more Whites (80% compared with 43%, P < 0,0001) and fewer females (9% compared with 26%, P = 0,0006) than the reference group. Furthermore, fewer subjects in the study group were exposed environmentally (9% compared with 30%) and through mining (22% compared with 46%), but a higher proportion were exposed in industry (60% compared with 23%). These differences are explained by the fact that the study group came mostly from the industrial sector of the Witwatersrand, whereas the reference group was drawn from a nationwide selection and particularly included the populations at risk because of asbestos mining. Mining places large numbers of people at risk, especially women and children through environ mental contamination. Thus 52 of the 71 women (73%) in the reference group with known exposure sources were exposed environmentally, and more than half of those exposed environ mentally were women (57% as opposed to 43% of the men). It is not possible to comment on the clinical implications of these epidemiological features since the numbers in the study group were too small to indicate the effect, if any, that the source of asbestos exposure might have on survival. Discussion Mesothelioma is a rare tumour. Its incidence has been reported as 1 and 2,2/million per annum in Canada and the USA respectively.5,6 The reporting of cases in South Africa ii unfortunately unreliable. Mesothelioma was only coded sepa rately as a specific cause ofdeath by the Department of Statistics in 1977. In 1978 a total of 5 deaths were officially recorded as due to mesothelioma; in the study group alone 4 deaths occurred in that year. The most recent official data available on causes of death are dated 1978 for Whites, Coloureds and Asians (with 4 deaths from mesothelioma, all in Whites) and 1979 for Blacks (in which year 18 deaths from mesothelioma were recorded).7' 6 In these circumstances the most reliable mesothelioma inci dence data at present come from the Panel records. Unfortu nately not even the Panel records all diagnosed mesothelioma cases since there is no statutory obligation for cases to be reported to the Panel or any other centre. Over the 3-year period 1980 -1982 only 57 of the 78 mesothelioma cases diagnosed (and compensated) by the Medical Bureau for Occupational Diseases were reported to the Panel (unpublished records of the MBOD, 1980- 1983). In addition to these 21 unreported cases, there are 3 from the study group which were also not reported to the Panel. There is also cause to believe that significant numbers of cases occur which are never diagnosed, particularly among Blacks.1 These unrecognized cases contribute further to the extent whereby the best-available figures underestimate the true inci dence. The total number of miners in the study and reference groups, including the 21 miners from the Medical Bureau for Occupational Disease records, is 171. Of these 44 (26%) were White, 40 (24%) Coloured and 87 (51%) Black. The White to Coloured and Black ratio is therefore almost 1:3. The race distribution of asbestos miners for the period 1977 - 1982 averaged 1 White to nearly 18 Coloureds and Blacks (Table I). The number of Coloureds and Blacks at risk from asbestos mining exposure is therefore nearly 18 times the number of Whites. Since the average duration of employment of Black miners is shorter and their rate of turnover higher than those of the White miners, it is likely that greater numbers and an even higher proportion of Black miners are at risk. While the evidence for a dose-response relationship for mesothelioma is not as strong as it is for lung cancer in relation to asbestos exposure, it is reasonable to assume that people with shorter durations of exposure have a lower relative risk of contracting mesothelioma than those with longer exposures.10 However, most South African asbestos miners work on the mines for longer than 2 - 3 months and would therefore have an exposure greater than 2 - 3 fibre-years/ml, a level which Nicholson et al. 10regard as sufficient to impart a significant risk for the development o f asbestos disease. Itis beyond the scope of this article to estimate the size of the at-risk population in South Africa or to put a figure to the true incidence of mesothelioma, but it is clear from the above discussion that in spite of the relatively large numbers of Black miners found to have meso thelioma, significant numbers of cases must occur which are never recognized or diagnosed. The extent of underdiagnosis is probably more marked in the non-mining industries. Of all the mesotheliomas arising from non-mining industry, 87% occurred among Whites and only 13% among Coloureds and Blacks. The ratio of Blacks to Whites is not as high in non-mining industry as in the mining industry, but is still high enough to expect that greater numbers of mesotheliomas would occur among Coloured and Black workers. The Black: White ratio in the two industries where most non-mining mesotheliomas arose in this study, the construction industry and the railways, is 6,5:1 and 1,3:1 respectively.11A similar situation applies as regards the environ mentally exposed cases where 45% of all cases detected occurred among Whites and 55% among Coloureds and Blacks. From the mean figure of 96,3 new cases of mesothelioma per year found in this study, an annual incidence figure, age- adjusted to the 1980 US population, of 7,2 per million population was derived (Table V). As has been shown, even this figure underestimates the true incidence -- further research is needed to arrive at more reliable estimates. The most plausible explana tion for the South African incidence being more than three times higher than that found in the US is that asbestos is not mined or milled there in any major quantity, whereas these activities are central to the South African asbestos industry The Canadian incidence rate is much lower than the South African rate, presumably because although much mining and milling is carried out in Canada, only chrysotile is worked- in Canada whereas South Africa mines crocidolite and amosite as well, both of which are considered to be more carcinogenic in man than chrysotile.3 Another area of interest in the context of mesothelioma in South Africa is compensation. Mesothelioma was made an occupational disease, sufferers of which were eligible for compen sation in the non-mining sector, in 1979, although it has been compensated in the mining industry under a provision for compensation of pneumoconiosis since 1962. In this series of .48 industrially exposed non-mining subjects, fewer than half (38%) of all those eligible have been compensated. The situation in the mining industry is much better and 88 subjects were compensated by the mines over the study period. Even so only 1of the 8 Black miners from the study is known to have been compensated to 412 SA MEDIESE TYDSKRIF DEEL 66 15 SEPTEMBER 1984 TABLE V. AGE-ADJUSTMENT TO US POPULATION (1980) FOR ANNUAL BEST-AVAILABLE SOUTH AFRICAN MESOTHELIOMA INCIDENCE RATE Age group (yrs) No. of cases,1977- 1983 Adjusted No. of cases, 1977- 1983* Annual age-specific incidence rate/million population! Annual age-adjusted agespecific incidence rate/ million population! 0 -1 9 -- -- -- 20-29 12 15 0,5 3 0 -3 9 42 52 2,5 -- 0,090131 0,347905 40-49 106 130 9,0 0,9042165 50-59 144 177 18,3 1,8843363 60-69 129 158 26,8 2,2322845 70-79 66 84 30,0 80-89 7 9 11,2 1,534977 90+ 1 1 5.2 0,2238565 Total 509 626 7,2177078 "Adjusted for cases with unknown age on proportional basis tSouth African 1980 population figures source Mid-Year Estimates, Republic of South Africa 1980 (Statistical News Release) Pretoria Central Statistical Services. December 1982 JUS 1980 population figures source 1980 Census o f Population US Department of Commerce Washington, OC Bureau of the Census. 1981 3 date. Attending medical practitioners shoulder a large portion of the responsibility for ensuring that details of all patients eligible for compensation are subm itted to the appropriate centres. U nder the Occupational Diseases in Mines and Works Act, there is a legal obligation on medical practitioners to notify all cases of mesothelioma in people employed at mines or controlled works to the Director ofthe Medical Bureau for Occupational Diseases. No such provision exists with regard to industrial workers. The obligation on medical practitioners to apply for compensation from the W orkmen's Compensation Commissioner in these industrial cases is even more crucial since the responsible employer, on whom the onus for reporting accidents and diseases and initiating compensation claims normally rests, is often no longer in business. Significant numbers o f subjects in both the study and the reference groups were exposed to asbestos environmentally (9% and 30% respectively). A t present these persons have no recourse to compensation since existing compensation laws only cover persons exposed occupationally. No other provision is made to cover them or their families against losses suffered as a result of the disease. Malignant mesothelioma affects a relatively young population in South Africa. It is noteworthy that 10% of all the subjects in the study contracted the disease before reaching the age o f 40 years, and 31%before reaching 50 years of age. O f the 54 subjects under 40 years of age, 16 (30%) had been exposed environ mentally, presumably from an early age. H alf of the women (12/24) under 40 years of age were environmentally exposed. Large, well-controlled, random prospective cohort studies, as advocated by Hillerdal,12are necessary to advance our knowledge of treatm ent and other factors which may ultimately lead to a better outlook for. patients with this disease which has such a dismal prognosis. W ith a mean lag period of 34 years, as found in this study, and with the continuing mining, manufacturing and consumer utilization of asbestos and asbestos products in South Africa even today, medical practitioners should-be seeing cases of mesothelioma well into the next century, and should have ample opportunity to study the disease more comprehensively. I thank D r G. Reinach of the Institute for Biostatistics for statistical help and Dr J. C. Cochrane of the National Centre for Occupational Health for his support and advice. Thanks are also due to Professor I. Webster of the South African Asbestos Tumour Reference Panel and D r H. Faure of the Medical Bureau for Occupational Diseases for making their records freely available. REFERENCES 1. Cochrane JC, Webster I. Mesothelioma in relationto asbestos fibre exposure; a review of 70 serial cases. 5 A frM e d J 1978; 54:279-281. 2. Mattson K. Natural history and clinical staging of malignant mesothelioma. E u rJ Resptr Dis 1982; 63: suppl 124,87. 3. Beckiake MR. Asbestos-related diseases of the lung and other organs; their epidemiology and implications for clinical practice Am Rev Respir Dis 1976; 114: 187-227. 4. Parkes WR. Asbestos-related disorders. B r J Dis Chest 1973; 67: 261-300. 5. McDonald AD, Harper A, El Attar OA, McDonald JC. Epidemiology of primary mesothelioma tumors m Canada. Cancer 1970; 26: 914-919. 6. Brenner J, Sordillo PP, Magill GB, Golbey RB. Malignant mesothelioma of the pleura: review of 123 patients. Cancer 1982; 49: 2431-2435. 7. South African Department of Statistics Deaths of Whites, Coloureds andA starts (Report No. 07-03-17). Pretoria: Government Printer, 1978. 131. 8. South African Department of Statistics. Deaths o f Blacks (Report No. 07-05- 01) . Pretoria: Government Printer, 1978:85. 9. South African Department of Statistics. Deaths o f Blacks (Report No. 07-05- 02) . Pretoria: Government Printer, 1979: 91. 10. Nicholson WJ, Perkel G , Selikoff IJ. Occupational exposure to asbestos population at risk and projected mortality'-- 1980 - 2030. A m JIndM ed 1982; 3:259-311. 11. Central Statistical Services. South African Statistics. Pretoria: Government Prmtcr, 1982: 7.4. 12. Hillerdal G. Prognostic factors and treatment results of 4 225 patients with diffuse malignant mesothelioma. E u rJ Respir Dis 1982; 63: suppl 124, 89.