Document xzqEYoXJegmB3nbo3xG5p1OLg

American Journal of Industrial Medicine 2:187-196 (1981) R&S 116767 Classical Syndromes in Occupational Medicine ^Pathological Lessons From Vinyl Chloride Angiosarcoma Hans(Popper) MOrPhDrand Irving J.^elikofpMfl^ Key words: vinyl chloride, angiosarcoma, pathognomonic syndromes, occupational medicine me* Hyperplastic or neoplastic nodules with potential transition to frank hepatic carci noma may be produced in experimental HggaoEi particularly rodents [Ncwberne and Butler, 1978], by an abundance of industrial and other chemicals. On the other hand, firm epidemiologic evidence of causation of tumors in the itmmn kyer by the vast majority of the industrial agents is not available today [Popper and Seiikoff, 198 la], Heperiri tumors, however, in experimental animals are conventionally considered as proof of carcinogenic potential, possibly expressing itself in other organ sites in man, rather than evidence for likely hepatic carcinogenicity per se. Thus, they are viewed in this way by regulatory agencies [Nicholson, 1981]. The important exception to this uncomfortable uncertainty in regulatory action is the experience with vinyl chloride, where the relation between experi mental production of hepatic lesions and the recognition of a specific tumor in man de serves recounting as an interesting example of a reliable application of observations in experimental animals to man. Viola et al [1971] in 1970 described tumors of lung, bones, and particularly in the zymbal glands near ihe ear in rats which had been exposed to inhalaiion of gaseous vinyl chloride. In 1974 Maltoni of Bologna [Maltoni and Lefamine, 1974a] recorded in an Stratton Laboratory for the Study of Liver Diseases. The Mount Sinai School of Medicine of the City University of New York (H.P.), Environmental Sciences Laboratory. The Mount Sinai School of Medicine of the City University of NcwYork i R&S 116768 'tfatfcaisse. '7'- ir?vs ,t i.\"^,i,>; s',..*;' .......... ' 18S Popper ud ScfflcofT Italian journal, not widely read in the United States, angiosarcomas in the liver and other sites in tats as the result of inhalation studies. Significant numbers of hepatic angiosarco mas developed under doses of from 30,000 to 250 ppm, given five times a week for 135 weeks. This observation had not been widely appreciated when Creech and Johnson [1974], also in 1974, described angiosarcomas Of the liver in four workers in Louisville who had been engaged in the polymerization of vinyl chloride monomer to the plastic polyvinylchloride. Since this is a rare tumor in man, its observation in a relatively small work force raised the suspicion of a causal relation to vinyl chloride exposure. This called attention to Maltoni's observations. Nonspecific lesions in the liver had been previously reported, particularly in the Eastern European literature, and it also had been brown that vinyl chloride workers frequently suffered from a characteristic scleroderma-like skin and bone abnormality, acioosteolysis [Jflhe and Lange, 1972]. In Germany, therefore, workers in vinyl chloride plants were screened by dermatologists, who detected enlarge ment of the spleen. That led to thorough investigation of such workers by internists, who recognized by combined clinical, laparoscopic, and histologic investigations a peculiar hepatic fibrosis which produced in some of these workers the picture of idiopathic portal hypertension, which sometimes led to bleeding of esophageal varices [Maisteller et al, 1973]. Peculiarly enough, angiosarcoma was not found in this group at the time, although it was subsequently also noted [Lange et al, 1974], Stimulated by the coincidence of these observations, a conference on the toxicity of vinyl chloride/polyvinylchloride was speedily convened by Doctors I. J. Selikoff and E. C. Hammond at die New York Academy of Sciences in May, 1974 [Selikoff and Ham mond, 1975]. It surveyed older studies as well as recent investigations dealing with epide miology, metabolism of the agent, clinical changes, and morphologic findings in animals and man. Ever since, a larger number of studies have been carried out all over the world. Moreover; Maltoni provided additional evidence about the lowest vinyl chloride levels which induce angiosarcoma in rodents [Maltoni and Lefamine, 1974b]. Evaluation of these data resulted in the reduction of the maximum permissible concentration in the ambient air in U.S. factories to 1 ppm with excursions not exceeding 5 ppm for short intervals. Similar reductions were legislated in other countries [Lelbach and Marsteller, 1981]. Here, the similarity in the evolution of the various stages of vinyl chloride-induced hepatic lesions in man and experimental animals [Popper etal, 1978; Popper etal, 1981b] will be detailed to provide lessons for environmental pathology. The initial lesion in both man and rodents (Figs. 1,2) is focal proliferation ofhepatocytes which show variations in size and appearance of nuclei, subsequently associated with an increase of sinusoidal cells of various types (Fig. 3). This in turn is accompanied by an increase of perisinusoidal collagen fibrils, recognized by electron microscopy, which also reveals an increase of fat-storing interstitial, or Ito, cells, which have been considered precursors of fibroblasts [Kent et al, 1976]. Silver impregnation visualizes the increase of reticulin fibers in cir cumscribed zones. These histologic features have served in coded routine microscopic examination of liver biopsy specimens to identify workers who have been exposed to vinyl chloride and other chemical agents. They were absent in controls [Tambuno et al, 1979]. They thus may represent a morphologic diagnostic criterion. sincc-biochemicaP hepatic tests have been las useful in thcJdentification of injury in such-woricers-Jlowever, the precursorlesion in the biopsy specimens does not exclude the simultaneous presence of fully developed angiosarcoma in other parts of the liver. It is therefore an indication for diagnostic follow-up by radionuclide scanning and angiography and subse quent confirmation by open liver biopsy [Dannaher et al, 1981a]. ver and other c angiosareo* week for 135 and Johnson in Louisville co the plastic atively small :. This called ;n previously a known that like skin and /, therefore, ded enLargeemists, who is a peculiar aathic portal steller et al, ne, although e toxicity of ikoff and H. f and, with! s in animals r the world, oride levels valuation of ition in the n for short Marsteller, de-induced cal, 1981b] >ion in both ariations in soidal cells isinusoidal increase of fibroblasts 5ers in cir* licroscopic exposed to urro et al, ochemical ers. How* lultaneous erefore^an ind si Vinyl Chloride Angiosarcoma 189 Rg. 1. Rat exposed to 500/ppm vinyl chloride by-inhalation for 126 weeks (material of Dr. C. Maltoni, Bologna). In center note sharply defined area in which hepatocytes are markedly enlarged and sinusoidal cells somewhat increased (H&E, X 100). Rg. 2. Mouse exposed to 2500/ppm vinyl chloride by inhalation for 6 months. Note irregularly limited area center where hepatocytes are markedly enlarged and sinusoidal cells are proliferated. (H&E, X 200) 190 Popper and Seiikoff Fig. 3. Biopey specimen of vinyl chloride worker. Scattered hepatocytes are enlarged and sinusoid*! cells of different type etc conspicuously increased (H&E, X 340). In both man and animals the proliferation of sinusoidal cells that is described above may become more conspicuous, and cells with hyperehromatic spindle-shaped nuclei gradually exceed the other sinusoidal cells in number (Fig. 4). Eventually the features of intralobular angiosarcoma develop which, in rodents, is always multicentric. Obstruction of the sinusoidal lumen by multiple layers of these cells leads to enlarging areas of ne crosis. While exposure of adult rodents to vinyl chloride results in angiosarcoma, similar exposure of newborn rats has produced mainly hepatocellular carcinoma [Maltoni and Lefemine, 1974b] (Fig. 5), which might be explained by altered hepatocellular biotransformation in the newborn state. Another frequent feature of the precursor stage is dilata tion of the sinusoids with an increased number of lining cells. This dilatation is not pre dominant around the tributaries of the hepatic veins and therefore can not be explained by passive congestion. Rather it is a specific effect of the agents on the sinusoidal wall. Alterations of the lining cells have been seen here by electron microscopy, which also reveals accumulation of platelets [Popper et a], 1977]. The progressive sinusoidal dilata tion leads to the more frequent form of angiosarcoma, the trabecular angiosarcoma (Fig. 6) , which is characterized by large bloody cysts traversed by cords consisting either of initially hyperplastic hepatocytes or of portal and central canals. All these structures are lined by increased connective tissue and angiosarcoma cells which invade the cords (Fig. 7) . Advancing fibrosis and neoplasia lead, as a rule, to atrophy and disappearance of the hepatocytes (Fig. 8), although in some instances they appear, even in later stages, rather hyperplastic and resemble hepatocellular carcinoma. Moreover, solid nodules of angiosar comas often form. They might consist of spindle shaped cells, but also of large polyhedral tumor ceils with abundant eosinophilic cytoplasm (Fig. 9). These have also been called Vinyl Chloride Angiosarcoma 191 R&S 116771 described above :-shaped nuclei >' the features of ric. Obstruction ng areas of ne uroma, similar a [Maltoni and dlular biotrans' stage is dilata tion is not prehe explained by inusoidal wall, py, which also ausoidai dilataosarcoma (Fig. tsting either of ' structures are the cords (Fig. warance of the r stages, rather es of angiosar- Rg. 4. Autopsy specimen of vinyl chloride worker. The sinusoidal cells, of different type, ate enlarged, sometimes in layers, their nuclei are frequently hyperehromatic or multiple. Some sinusoids are widened (H&E, Rg. 5. Very young rat exposed to 500/ppm vinyl chloride fay inhalation five times a week for 76 weeks (material of Dr. C. Maltoni, Bologna). Note neoplastic nodules (between curved arrows) and hepatocellular carcinoma (between straight arrows) with dilated vascular spaces (H&E, X 100), r>-5 192 Popper end SeUkolf (Z5*fc afessS ssa Rj. 6. Autopsy specimen ofvinyl chloride workcrJTrabcculir ingjosircoma. The hepsiocytes, arranged in cotds, arc hypoplastic (right lower aspect). The dilated*spices are lined by angiosarcoma cells (H&E, X 40). |5L. & 'Stl Pig. 7. Mouse exposed si 2500/ppm vinyl chloride five times s week ford months by inhalation. Polyhedral angiosarcoma cells line dilated blood space which is traversed by bepaiocytic cords (straight snows) and rem nants of portal tract (curved arrow). Both ate surrounded by angiosarcoma cells (HAE..X 150). m R&S 116772 Vinyl Chloride Angiosarcoma 193 rf'- 5- Fig. 9. Autopsy specimen of vinyl chloride worker. Clusters end cords consisting of solid angiosarcoma (arrows) in blood spaces (H&E, X 100). 194 Popper and Selikoff histiocytoid [Rosai et al, 1979], The derivation of the tumor ceils is not hilly established, even on fine-structured analysis. However, demonstration of factor VIII in the tumor cells strongly supports an endothelial cell origin [Fortwengler et al, 1981] and speaks against a. Kupffer cell tumor. Still, at this time, the noncommittal term `angiosarcoma' is preferred. Usually the picture of trabecular angiosarcoma is so characteristic that it is readily diag nosed in biopsy specimens. At human autopsies, however, all forms of angiosarcoma are almost always multicentric. This observation discourages surgical intervention even in eariy stages. Chemotherapy offers limited promise [Dannaher et al, 1981b], Thus vinyl chloride produces a well defined histologic sequence virtually identical in and man. Moreover, there is evidence on the basis of limited human material that prolonged exposure to inorganic aisenicals, for instance Fowler's solution in the treatment of asthma or psoriasis, has led to the same sequence. This however, has not been reproduced in experimental animals [see Popper et al, 1978]. Several decades ago the use ofinorganic arsenical pesticides in vineyards, now prohibited, was associated with angiosarcoma, cirrhosis, and hepatocellular carcinoma [Roth, 1957]. In India, increased arsenic in drinking water has been associated with idiopathic portal hypertension [Dana, 1976], supporting the assumption that this rare disorder in the Western world may be explained by cither arsenic or possibly other agents, so far unidentified. Study of more extensive material or oriental idiopathic portal hypertension suggests, however, a differ ent pathogenesis, namely, a disorderof intrahepatic portal vessels (Futugawa et al, 1980]. Thoiutrasr, used years ago in radiologic visualization, mainly of cerebral vessels, has been followed by both angiosarcomas and carcinomas of either hepatocellular or cholangiocellular type; in a few specimens the precursor lesion has been identified, and animal experimental reproduction has also been reported [see Popper et al, 1978]. A survey of hepatic angiosarcomas observed during a ten-year period in the United States [Popper et al, 1978; Falk et al, 1981] has regularly shown the presence of the same precursor lesion in almost all instances. This suggests that af this time in idfcpaihic eases there are some so far unidentified-giu/imnmCTral-^artrtfc Armnie is a candidate. Moreover,-this survey has discovered patients with angiosarcoma in whom a history of intake oflarge doses of androgenic/anabolic steroids was recorded [Falk et al, 1979]. This suggests that these steroids may join vinyl chloride, thorotrast, and arsenic as etiologic factors in angiosarcoma. Recently such tumors have also been observed following inges tion of oral contraceptives [Monroe et al, 1981] or long-term estrogen therapy for carci noma of the prostate [Ham et al, 1980]. Since all these steroids produce, besides cholesta sis, sinusoidal dilatation and adenomas, it appears now that there may be a group of agents responsible for a previously unrecognized sequence that is associated neither with cir rhosis nor with elevated serum alpha-fetoprotein. It starts with an identical precursor stage, namely, mixed sinusoidal and hepatocellular hypexplasia, not necessarily recog nized by common hepatic tests. The sequence may terminate in idiopathic portal hyper tension or hepatic tumors like adenoma, carcinoma, or angiosarcoma. Thus, a group of environmentally induced hepatic lesions is now being recognized which may possibly also be induced by other agents, so far not appreciated. Substances related to vinyl chloride, such as vinyl bromide [Bolt et al, 1979] or vinylidine chloride [Reitz et al, 1980] have produced a similar sequence in experimental animals. This calls for increased vigilance in the use of chemically related substances produced in industry. In contrast it should be stressed that for the hepatocellular nodules without significant participation of sinusoidal ceils, produced in rodents by many industrial and other environmental agents, a human counterpart has so far not been convincingly demonstrated, with the exception of the Vinyl Chloride Angiosarcoma 195 R&S 116775 mycotoxin aflatoxin. Even for the latter, the human evidence is so far still circumstantial [Popper 1979]. This lack of proven examples of human hepatocellular carcinoma caused by industrial agents which produce lesions in experimental animals including, in the case of nitrosamines, primates, does not exclude their carcinogenic potential in man. Almost all are metabolized in the human liver and a promoting effect cannot be excluded. In recent years, the promoting action in hepatocarcinogenesis has been given greater importance over the initial effect of DNA alteration, recognized by mutagenesis. More over, since most human hepatocellular carcinomas are associated with two principal fac tors, namely alcoholism and hepatitis B infection [Popper, 1979] the causal relation may be more readily overlooked than in the case of agents producing the more unusual an giosarcoma. Delineation of the promoting effect of the other industrial agents, perhaps by alteration of microsomal biotransformation or by stimulation of proliferation of hepatocytes, is a task of the future. So far the number of recorded human hepatic angiosarcomas defined as associated with vinyl chloride exposure seems not to exceed 95 worldwide [Falk et al, 1981; Baxter et al, 1980; Pialat et al, 1980; Wagoner et al, 1980], although listings are incomplete. Most recent cases have been reported in Germany and France but no new cases of an giosarcoma were detected in the last seven years from those plants in the U.S. where they were originally described [Dannaher et al, 1981a], Hepatocellular carcinoma has been very rare [Koischwitz et al, 1981]. Other manifestations of "vinyl chloride disease" [Lelbach et al, 1981; Veltman, 1980] ar^also not common. Recognition of the sequences induced by vinyl chloride and the other agents men tioned and their reproduction in experimental animals are an important lesson in environ mental medicine. It has not only facilitated clinical recognition but, even more important, has led to control of exposure and avoidance of additional disease. Too, it has spurred the search for similar sequences resulting from chemically related agents. Finally, human pathology has been enriched by description of entities produced by exposure to environ mental, including industrial, agents. ACKNOWLEDGMENTS Supported in part by National Institute of Environmental Health Sciences grant no. ES00928. REFERENCES Baxter PI, Anthony PP, MacSwecn RNM, Scheuer PJ (1980): Angiosarcoma of the liver annual occurrence and aetiology in Great Britain. BrJ Indust Med 37:213-221. Bolt HM. Laib RJ. StSckle G (1979): Formation of pre-neopiastic hepatocellular foci by vinyl bromide in newborn tats. Toxicology 43:83-84. Creech JL Jr, Johnson MN (1974): Angiosarcoma of liver in the manufacture of polyvinyl chloride. J Occup Med 16:130-131. Dannaher CL, Tamburro CH. Yam LT (1981a): Occupational carcinogenesis: The Louisville experience with vinyl chloride-associated hepatic angiosarcoma. Am J Med 70:279-287. Dannaher CL, Tamburro CH, Yam LT (1981b): Chemotherapy of vinyl chloride-associated hepatic angiosar coma. Cancer 47:466-469. Datta DV (1976): Aisenic and non-cirrhotic portal hypertension. Lancet 1:433-435. Falk H. Popper J. Thomas LB, Ishak KG (1979): Hepatic angiosatcoms associated with androgenic-anabolic steroids. Lancet 2:1120-1123. Falk H. Herbert J, CrowleyS. Ishak KG, ThomasLB, Popper H. Caldwell GG (1981): Epidemiology of hepatic angiosarcoma in the United States 1964-1974. Environ Health Fctspcct (in press).' Fortwengier HP Jr, Jones D. Espinosa E. Tamburro CH (1981): Evidence for endothelial cell origin of vinyl chloride-induced hepatic angiosarcoma. Gastroenterology 80:1413-1419. R&S 116776 rvj** 196 Popper and SeUkoff FuUgawm S, Fukazawa M. Horisawa M, Musha H, Ito T, Sugiun M, Kameda H, Olcuda K (1980): Porto- giaphic liver changer in idiopathic none .irhodc portal hypertension. Am J Roentgenol 134:917-923. Ham JM. Pirol* RC, Crouch RL (1980): Hem&ngioendothelial saicoroa of the liver associated with long-term estrogen therapy in a man. Dig Dis Sd 23:879-883. JQhe S. Lange C-E (1972): Sklerodennieanige Hautvetaendetungen. Raynaud-Syndrom und Akro-osteolyaen bet Atbeitern der PVC-betstelienden Industrie. Dtsch Med Wochenschr 97:1922-1923. Kent G. Gay S. Inouye T, Bahu R, Minick OT. Popper H (1976): Vitamin A-containing lipocytes and formation of type 01 collagen in liver injury. Proc Natl Acad Sd USA 73:3719-3722. Koischwitz D. Lelbach WK. Laclater K. Hennanutz D (1981): Das vinyldriorid-induziene Lebetangiosaricom und hepatozeilulaere Kaninoro. Fortschr Rdntgenstr 134:283-290. Lange C-E. iOhe S. Veltman G (1974): Ucber das Auftreten von Augiosarkomen der Leber bd zwd Atbeitern der PVC-hentellenden Industrie. Dtsch Med Wochenschr 99:1398-1399. Lelbach WK. MantellerHJ (1981): Vinyl chloride associated disease. Eigeb Inner Med u Kinderheil 47:1-110. Mahoni C. Lefcmine G (1974a): Le potenzialita dd saggi sperimentali nella ptedirione dd riidti oncogen! ambientali: un esempio: il dororo di virile. Aecad Nazionaie dd Lined 36:1-11. Mahoni C. Lefctnine G (1974b): Carcinogenicity bioassays ofvinyl chloride. 1. Research plan and early results. Environ Re* 7:387-403. MamellerHI. Ldbadi WK. MilllerR, JflheS. Lange CE, RohnerHG, Veltman O(1973): Chronisch-toxische Leberscfaaedcn bd Arbeitern in der pyc-Produfction. Dtsch Med Wochenschr 98:2311-2314. Monroe PS, Riddell RH, Siegler M. Baker AL (1981): Hepatic angiosarcoma. Possible relationship to long term oral contraceptive ingestion. JAMA 246:64-63. Newbcme PM. Butler WH (eds) (1978): "Rat Hepatic Neoplasia." Cambridge: MIT Press. Nicholson WJ (ed) (1981): "Management of Assessed Risk for Carcinogenesis." New York: New York Acad emy of Sciences. ffdat J, Pasquier B, Prim M. Kopp N (1980): Hepatic lesions caused by vinyl chloride monomer (CVM). Son Hop Paris 56:1188-1202. __ Popper H (1979): Hepatic cancels in man: quantitative-perspectives. Environ Res 19:482-494. Popper H. SelikofT U (1981a): What is environmental pathology? Am J Med 70:218-220. Popper H, Mahoni C, SelikofT U (1981b): Vinyl chloride-induced hepatic lesions in man and rodents. A com parison. Liver 1:7-20. tapper H. Thomas LB, Schaffoer F, Mahoni C. SelikofT U (1977): Interaction between sinusoidal ceils and hepatocytes in human aid experimental angiosarcoma induced by environmental factors. In Wisie , Knook DL (eds): "Knpffor Cells and Other Liver Sinusoidal Celts." Amneidam: Elsevier/North Hol land Biomedical Proas, pp 173-181. Popper H, Thomas LB, Tellet NC, Falk H, SelikofT U (1978): Development of hepatic angiosarcoma in man induced by vinyl chloride, thorotmst and arsenic. Am J Pathol 92:349-376. Rehz RH, Wtunabc FG. McKenna MJ, Quasi JF. Gehring PJ (1980): Effects of vinylidine chloride on DNA synthesis and DNA repair in the tat and mouse: A comparative study with dimethylnitrosamine. Toxicol Appl Pharmacol 52:357-370. Rosa) J, Gold J, Landy R (1979): The htstiocytoid hemangiomas. A unifying concept embracing several previ ously described entities of skin, soft tissue, large vessels, bone, and heart. Hum Pathol 10:707-730. Roth F (1937): Anen-Leber-Tbmoren (Haetnangioendothellom). Zschr Kiebsfbisch 61:468-303. SdikofT U, Hammond C (eds) (1973): "Tbxkity of Vinyl Chloride-Polyvinyl Chloride." Ann N Y Acad Sri. p 246. Thmbuno CH. Makk L. Popper H (1979): Early hepatic histological alterations among chemical (vinyl mono mer) workers. Gastroenterology 77:A43. Veltman G (1980): Klinische Befunde und aibeitsmedizinische Aspekte der Vmylchloridkiankheil. Theta- piewoche 30-.5331-3338. Viola PL, Blgotti A. Caputo A (1971): Oncogenic response of rat skin, bungs, and bones to vinyl chloride. Cancer Res 31:316-522. WagonerJK. Infame PF. ApfeldorfRB (1980): Toxicity of vinyl chloride and polyvinyl chloride as seen through epidemiologic observations. J Toxicol Environ Health 6:1101-1107. m