Document wgbXJK5rJY8JQMdGBJED27GRV

*** ,,r aw a.lS|'?,3 im flr/f/jfr ofIndustrial Medicine 1980;37:213-221 a ^'Angiosarcoma of the liver: annual occurrence and aetiology in Great Britain P J BAXTER.1 P P ANTHONY,1 R N M MACSWEEN.* AND P J SCHEUER4 From the Health and Safety Executive, Employment Medical Advisory Service,1 London NWl SDT, the Area Department of Pathology,* Exeter EX2 5DY, Western Infirmary? Glasgow Gil 6NT, and Royal Free Hospital? London iVH'J 2QG, UK abstract The annual occurrence of angiosarcoma of the liver (ASL) In Britain from 1963 to 1977 was studied, including clinical and occupational details for those cases agreed as ASL by a panel of hlstopathologists. Thirty-five cases (28 men, six women, and one infant girl) were agreed as ASL. The increase in themeidenefof A$>L observed in recent years was attributable toThorotrast(thori um dioxide) usage (eight cases) and exposure to vinyl chloride (two cases) in the past. In its clinical i presentation and prognosis ASL resembled primary liver carcinoma, except that extrahepatic metastases were found in only eight (23 %) cases, and haemoperitoneum was more common in those cases due to Thorotrast. The results suggested a possible increased risk of ASL in the electrical and plastics fabrication industries, but information on exposure was inadequate to implicate specific chemicals. The clinical features ofone case were indicative of arsenical intoxication, but medications in the other patients did not appear to be of aetiological importance. VVe present our findings on the annual occurrence of Office for Population Censuses and Surveys (England angiosarcoma of the liver (ASL) in Britain during a and Wales) and the General Register Office IS-ycar period, 1963-77, and the clinical details Of (Scotland), identities recorded cases by the regular those cases agreed as ASL by a panel of pathologists. review of death certificates. The underlying cause of ai The role of known or suspected aetiological factors,1 death is routinely coded by the Registrar General's including vinyl chloride, Thorotrast {thorium staff under the International Classification of ? dioxide), inorganic arsenic and copper, and haemo- Diseases, 8th revision, as primary, secondary, or chromatosis. is also discussed. Many other chemicals unspecified neoplasms of the liver and intrahepatic 34 -1 occurring in the environment naturally or from bile ducts (ICD numbers 155.0, 133.1, 197.7. 197.3. industrial sources have been linked with liver 211.5); neoplasms of unspecified nature of the liver tumours in animal studies,1 but the role of these and-biliary passages (230.5); and haemangioma or substances in man remains to be resolved. The lymphangioma (227). Additional sources of notifica advantage of monitoring a tumour as rare in the tions are hospital pathologists, publications, and general population as ASL is that it may be used as cancer registries (England and Wales only). Recorded a marker tumour to identify human carcinogens, and cases are identified if any of the following diagnoses we have adopted this approach. are mentioned: haemangio orangiosarcoma, Ktipffer cell sarcoma, haemangioendotheUoma, haemangio- Method blastoma (adults only), or malignant angioma. For all the recorded cases the hospital pathologists In 1974 an ASL register was set up in Britain by the concerned are requested to send histological materia! Employment Medical Advisory Service (EMAS) of from the tumour for review by a panel of three the Health and Safety Executive, and a preliminary hlstopathologists (PPA, RNMMS, and PJS). report on the incidence of ASL in 1963-73, with Material for control purposes in the study of the details of the method of working of the panel, has J 974-7 cases of ASL was obtained from cases been published.1 EMAS. in collaboration with the recorded as dying from other liver tumours (selected i randomly from death certificates) or because the Received 4 Jurv* 1979 Attested 3 July 1979 diagnosis was of special interest--for example, angioma, haemoperitoneum--or mentioned Thoro- 213 \ V a.;-' V-`- VL a v. ss 'a. K.. iv. if ir w< !W !*T V AP000I0627 i i. Angiosarcoma of the liver: annual occurrence and aetiology in Great Britain 215 show the increase expected if national awareness had led to more recording of ASL. Also, in the pros pective studies of Thorotrast patients3* and British workers exposed to vinyl chloride7 deaths from ASL have occurred only In recent years. Taking the ten-year period 1968-77 as being most representative, the annual number of cases in whleh no causal agents had been positively identified was fairly constant, both for the recorded cases and those agreed by the panel (figs lb and 2b). On average, seven cases a year were recorded from all sources, and among those for whom histological sections were obtainable, about two cases a year were agreed as ASL. For death certificate cases I Fig 1(a) Annual Incidence ofaU recorded cates (including three controls), and numbers ofcases agreed only, the average number ofcasts reported was five a year, with about one a year agreed by the panel. as ASL by panel, 1963-77. Fig 1(b) At infig 1(a), but excluding cases ofknown aetiology, AGI DISTRIBUTION The distribution of the ages at death of the agreed cases ofASL is shown in fig 3; there was no apparent until 1975 and 1977, when it rose to !2 in both years. The numbers of cases agreed by the panel remained at two to three a year until 1975 and 1977, when they rose to seven and nine respectively. If the cases of difference between cases of known and unknown aetiology, though the numbers of both were small. The mean age (and range) for men was 58-3 (30-78) years, and for women 65-2 (44-84) yean; the infant girt died aged 8 months. known aetiology (vinyl chloride and Thorotnst, to be described below) were removed from considera tion an increasing incidence was not so apparent (fig lb), suggesting that the rise had been due to a real increase in the number of cases rather than an improvement in case finding after publicity of the vinyl chloride problem in 1974. This interpretation is supported by the annual numbers of death certiflcatM for ASL (figs 2a and b>. These did not SOCIAL CLASS * For the male cases agreed as ASL, the occupations recorded on the death certificates were assigned to the appropriate social class according to the scheme of the Registrar General* by a statistician who was not provided with any other information about the cases. The social class distribution was then com pared with that obtained by a proportional analysis f male deaths aged 15-64 in England and Wales, 1970-2," and also of the male population of Britain ( H I- Fig 2(a) Annual incidence ofall eases Identified by death certificate search, and numbers ofcases agreed as ASL by panel, 1963-77. 0-U -IS Age (years) - $5 - W-64 Fig 2(b) As in fig 2(s), but excluding eases ofknowst Fig 3 Agedis/ribution ofmale andfemale eases i aetiology. agreed by panel as ASL, 1963-77, AP00010628 i i ) .1 te .n '.d I On Angiosarcoma of the liver: annual occurrence and aetiology in Great Britain 2t7 Tabic 4 Industries employing at least two of the agreed adult eases of ASL according to aetiology. {No of eases in parentheses) AetMeej Industry A'< ofpatients Unksovtn. or doubtful (24) Mm in*) Women (J> VJnyt afclerida (2) Men in TDorouastW Men {Ti S!*ecneal flaulaa fibrienic Truipocl Comtfuetton Coal iriintne Fannins Cokc-otmi (|ss end uact) Ctblt ffloklflf Cisrrfn* DomMtc sorrioo PVC ffluwnetuit Transport Ceoiuuetlon Coalmining * 3 3 3 3 2 2 2 2 2 2 2 third as a haulage worker for most of his working life. In the remainder of the industries shown in table 4, farming is noteworthy because of the known use of arsenical pesticides. Both men who had held farming jobs had been labourers at the beginning of their working lives, but details of pesticide usage were unobtainable. Information was provided for one of the two men who had worked in cablemaking; he had been an inspector of brass-plated wire in an entirely non-rubber part of a tyre factory, and was not directly exposed to chemicals or PVC. The other man had been employed in this industry from 194L to 1946, before PVC was generally used in cable manufacture. The two men employed in coke ovens had held this occupation for five years and over 30 years. Of the five women, four had either worked in catering or domestic service for at least five years. Vinyl chloride cases death certificates, as mentioned In the results of the The two men in this category had been process social analysis. Unfortunately, no further workers in the manufacture of PVC for three and a information was obtainable for two of these cases; halfyears* and 20 years1*; their ages at death were 37 according to the death certificates they had been an and 71. Both had worked in autoclaves, where electronic testing engineer and power station exposure to VCM would have exceeded 200 ppm. engineer. The two others, the first a construction and Neither had been employed in occupations where design engineer at a power plant and the second a heavy exposure to PVC dust could have occurred. Post Office engineer supervising telephone and postal power plant, had been in the industry for most of Thorium dioxide cases their lives. One of the two remaining cases had been Transport and construction Industries also feature a self-employed electrical fitter all his life, and the in this category. The two men in coalmining had other a boilerman and odd job man for the electricity been faceworkers, and an injury at work to one had board. It wu not possible from the histories obtained resulted in his receiving Thorotrast during further to derive a specific exposure which these six men may medical investigations. have had in common. At least three of them, however, probably handled or repaired electrical Other cases components for some period in their careers, but Occupations mentioned only once did not show any more precise information was not available. actiological pattern. In the infant ease, occupational Of special interest in relation to vinyl chloride is histories were obtained from both parents to study theiindtn* three cases among plastics fabricators* the possible role of transplacental carcinogens. The According to the death certificates the occupation of father had worked for three years in market two of these men was "machine, filter." The three gardening, followed by three years as a mechanic and men had been employed in factories that manu painter of old vehicles. He then worked as a refuse factured goods from raw plastic, including, at some collector a year before and after the child was bom. time, PVC; two factories had also produced rubber The mother had worked as a hairdresser since articles (the use of cbloropreoe. a possible cause of leaving school until she was seven months' pregnant. ASL,U was excluded in one of these firms). The She admitted having received heavy exposures to records were inadequate, however, for confirming hair lacquer from hand-pumped sprays, and had whether or notthesd iMft hid come into contact with rarely used canned aerosols at work. PVC 1 ---------------------------- " In the construction industry two men had been MEDICATIONS building labourers and one a clerk. All three men in Ofthe agreed cases, seven men and one woman had the transport industry had been.road vehicle drivers. been neurosurgical patients investigated with intra Three men had worked in collieries, one as a clerk, arterial Thorotrast 28-36 years before their deaths. another as a "colliery worker'* for one year, and the Another woman, who died aged 70, was suspected. AP00010629 Angiosarcoma of tkt liver; jtwntaf occurrence and aetiology in Great Britain 219 cases, the sites bei** *~!en in four cases, the cause; thus reliance on death certificates could lungs in tw.r. anc "J-r is': ;t*:un and porta hepatis in underestimate the trueincidcncc byat least half. In the each of the :ema:r ~z 'n two of the cases with Thorotrast study" six cases of ASL were eventually splenic hvorve--iT . -s Ji-rant metastases were Identified whereas onlythreewere recordedasASLon found in the her.,.*. and adrenal gland. the death certificates; this finding in a virtually Jn most putlcr.i ;:= :-cz enlarged by the complete follow-up of patients with an above- presence of vasculi.* :r.c >: tissue. Tumour nodules, average necropsy rate also suggests that the true sometimes rJseu i~:\s ths liver surface and incidence could be underestimated by half. On the containing cautrr.o or honeycombed areas were other hand the panel agreed the diagnosis in under often described, as v-rs -.crying amounts of fibrotic half of all the cases seen, so that incidence rates based (issue, and macro- or micro-nodular cirrhosis of the on death certificates might not be too far removed liver. In two cases tit -thvh the liver was cirrhotic, from the true incidence, at least among hospital the presence of a rumour was not confirmed until deaths. In Britain about half of all deaths occur in histological examination. In five Thorotrusl cases hospital, and on average 20-30% of these undergo the spleen was small and fibrotic. The liver of the necropsy.24 Hepatic cirrhosis and primary liver infant was enormously enlarged and contained cancer were the commonest misleading diagnoses, multiple umbilicated nodules of tumour tissue on its but we have no evidence' to suppose that many more surface; there were areas of small, thick-wailed cases of ASL were being overlooked in these channels filled with blood surrounded by normal categories. liver tissue, but no extrihcpatlc merastases, and the The age distribution of the adult cases resembled immediate cause of death was congestive cardiac that of primary liver carcinoma In Britain.1* ** There failure. In none of these cases washaemochromntosis was no conspicuous social class gradient, but surpri found. singly all four cases in social classes I and II were engineers in the electrical industry. The maletfemsle Discussion ratio for adults was 5:1 for all cases, and 4:1 for "idiopathic" cases only. This male preponderance Excluding the cases of confirmed aetiology, the has been shown for primary liver carcinoma3* as well incidence of ASL during a representative ten-year as ASL,4 suggesting that occupational factors may period was seven a year for all recorded cases, and be responsible. Male susceptibility has been observed two a year for cases agreed by the panel, in a British in animal studies of primary liver carcinoma,2* population of about 50 million. Surveys from else however, so the importance of occupational agents where have also shown the rarity of ASL. but they in these sex differences awaits further elucidation. have been undertaken in industrialised countries The commonest presenting symptom was upper where the incidence of primary cancer is low, though abdominal pain, and in this respect, and also in the in Britain it has probably increased slightly over time between onset of symptoms and death, ASL recent decades.1*19 resembled primary hepatic carcinoma.1* The median Necropsy surveys indicate that ASL accounts for survival time of three woeks after admission to less than 3% of primary liver tumours and less than hospital emphasises the difficulties in obtaining 15 per 100 000 of ail necropsies.20 21 Epidemiological occupational histories for survey purposes and the data in the USA-2 and the Netherlands23 show much -necessity of relying on the relatives for this infor the same rate of recorded cases as found in the mation. present study. Infant cases are extremely mre. Out Thus in its incidence and presentation ASL of four recorded infant deaths, tumour tissue was resembled primacy hepatic carcinoma, though the obtained for two, and one of these was agreed as number of cases of ASL on which to base these ASL. comparisons was small. Extrahep&tic raetostoses, The use of data soleiy from death certificates however, were found in only 23% of cases; this seemed to provide an insensitive measure orinddence figure should bo compared with 47 % obtained by compared with that from all sources, including MucSween for primary livercarcinoma in Scotland,1* hospital pathologists, published reports, and cancer and 69% reported by Popper tt el* in their cases of registries. The rising trend for all recorded cases for ASL, which were mostly drawn from the USA. The the last yean of the study period was not seen for the reason for our low figure is not clear, but may relate death certificate cases. Evidently data from death to differences in diagnostic criteria or aetiology. certificates alone cannot be relied on to detect those Rupture of the tumour resulting in intra- moderate increases in incidence that may be a clue to abdominal haemorrhage caused or contributed new environmental agents, in 17 cases agreed as ASL towards death in 26% of adult cases, and was the death certificate ive a different underlying significantly more common in those patients who fi % i \i .*t t i i r AP00010630 vt ' jft ", r ` ;..v\ 121 i f- Angiosarcoma ofthe liver: annual occurrence and aetiology in Great Britain 221 * Preussmen R. Hepatocarctnogens as potential risk for human liver cancer. In: Remmer H. Boll HM, Bannatlh endothelioma involving the liver. J Clin Pathol 1974:27: 214-21. I* V P. Popper H. cdk. Primary Liver Tumours. Lancaster: 14 MacSween RNM. A cQnicopathologlca) review of 100 MTP Press, 1974:11-29. eases of primary malignant tumours of the liver. J CHn . MacSween RNM. Vtt;rt JM. Rosa SK, Ferguson J, Pathol 1974(27:669-82. Johnstone JM. Sindison AT. Hacmangiocndothclisi Burnett RA, Patrick RS, Spill WGS, Buchanan WM. r sarcoma of the liver. J Paikol 1973:109:39^4. MacSween RNM. Hepatocellular carcinoma and hepatic 4 Popper H. Thomas LB, Telles NC. Falk H, SdikotT IJ. cirrhosis in the West of Scotland: A 25-year necropsy Development or hepatic angiosarcoma ht man induced review. J Clin Pathol I97S3I :I08-I0. by vinyl chloride, Thorotrut and arsenic. Am J Pathol " Alrenga DP. Primary angiosarcoma of the liver, int Surg I97t (92:349-70. 1975;60:198-203. * Smith PM, Wllliaim DMJ. Evans DMD. Hepatic angio 41 Rein FR, Hulh F. Sachs spoctane primlra maligne I sarcoma in a vinyl chloride worker. Bui! NY Acad Med geftssgeschwBbte der leber in einem koitektiv von 1976(92:447-52. 30,000 obduktlonen. Int Arch Occup Environ Health 4 Syed S, Jung RT. Cardiac tamponade caused by mftas- 1975;34:237-46. ttsing haemaagioendoiheiial sarcoma of (he liver. Brit ** Heath CW, Falk H, Creed: JL. Characteristics of cases of Heart J 1971;40:697-9. angiosarcoma of the liver among vinyl chloride workers T Fox AF, Collier PF. Mortality experience of workers in the United Stales. Ann NY Acad Sci 1975;246:231-6. exposed to vinyl chloride monomer in the manufacture u Dalderup LM. Angiosarcoma of the liver ie Great Britain. of polyvinyl ehloride in Great Britain. Br J fnd Med BrMedJ I9T7;1:1149.J0. 1977;34:1-10. ** Brlthh Medical Journal. Reviving the hospital necropsy. I Office of Population Censuses and Surveys. Classification Br MedJ 1977;l:l560u of Occupations, 1970. London: HMSO, 1970, H Johnson PJ, Krasner N, Portmann B, Eddleston ALWF, * Office of Population Censuses and Surveys. Registrar Williams R. Hepatocellular carcinoma in Great Britain: GtntraTs Decennial Supplement for England and Wales, Influence of age, sex, HBsAg status, and aetiology or \ 1970-3. London: HMSO. 1971. underlying cirrhosis. Gat 1978;19:1022-6. ** Office of Population Censuses and Surveys. Census 1971, ** Da Silva Horta J, Da Mona LC, Tavares MH. Thorium it t Great Britain, Economic Activity Part IV, London: dioxide effects In man: Epidemiological, clinical and HMSO, 1975. pathological studies (experience in Portugal). Environ 1I { u Tarring T. ed. NemFtnottt Mete! Worksof the World. 2nd ed. London: Metal Bulletin Books, 1974. * Aston SR, Thornton !, Webb JS, Milford BL. Purves JB. - Arsenic in stream sediments and waters of South West Ret 1974J:I31-S9. *' Ong GB, Taw JL. Spontaneous rupture of hepatocellular carcinoma. Br MedJ 1972:4:146-9. 44 Doll R. Vessey MP. Beasley RWR. et al. Mortality of !tr.f England. Sci Total Environ 1973;4:347-58. gasworkers--final report of a prospective study. Br j ** Colboum P, Altoway BJ. Thornton I. Arsenic and heavy Ind Med 1972 9:394-406. metals in soils associated with regional geochemical n Health and Safety Executive. Mortality In the British anomalies in South West England. Sci Total Environ Rubber Industries. London; HMSO On press). i; 1975;4:359-63. " Milby TH. ed. Vlnvl Chloride: an Information resource. infante PF. Mutagenic and carcinogenic risks associated with halogenated olefins. Environ Health Pcrspect 1977(21 231-4. United States: DHEW, 1978. 41 Ministry of Agriculture, Fisheries and Food. Survey of vinyl chloride content of pvlyvlnyl chloride for food f: ? II Lot FI, Harry DS. Angiosarcoma of the liver in a vinyl- contact and offoods. London: HMSO. 1978. t chloride worker. Lower I974;1:I31^4. " 8axtcr PJ. Longlands AO, Anthony PP, MacSween RNM, * 14 Donald D, Dawson AA. Microangiopathic haemolytic SeheuerPJ. Angiosarcoma ofthe liver: a marker tumour anaemia associated with malignant haemangio-endo- for the late effects of Thorotrest in Great Britain. Br theiioma. J Clin Pathol 1971(24^56-9. J Cancer (in press). J >' Pollard SM, Milhvard-Sadier GH. Malignant (Hemangio K I 4 i j i .a AP0001063I