Document 3ZB8ddLODeNg61pLL7g3JayJ

\\g j ^ ZA**'/ ! 5* 4 4 - >*"? t\fa4^L) !*? 7 & rw | ! tecs nas. i )(fM ini Ths QTfi l Tht 1i i ff# bbiti i nvtet P*du xiuU, Angiographic and Radionuclide Characteristics of Hepatic Angiosarcoma Found in Vinyl Chloride . Workers1 Joseph G. Whelan, Jr., M.D., John L. Creech, M.D., and Carle H. Tamburro, M.D. Hepatic angiosarcoma, recently dricovored In a largo series ol vinyl chloride workers, demonstrates characteristic, angiographic and radionuclide changes. Tumors exhibiting central hypevaseularity with puddling are usually surrounded by a peripheral stain. A negative peripheral delect is demonstrated on hepatic scan. Healing hepatic Infarction secondary to wedged hepatic venography creates a false-positive lesion on angiography similar to angiosarcoma. Splenomegaly and systemic venous hypertension develop in a number ol these patients, bJOSX tepms: Liver, infarction Liver neoplasms, angiography Liver Neoplasms, ra dionuclide diagnosis a Sarcoma Spleen, hypertrophy Vinyl Chloride Dfegnosiic FU*Ckjyy The n Radieto"gy 118:549-557, March 1976 ..JuM * <f mecjNcs 1964, 8 cases of hepatic angiosarcoma have w* f been found in Louisville, Ky. among workers asso- *bra>ed with the polymerization of vinyl chloride, a male- 476. h essential to the production of plastics. In 1973, sch and Johnson (1) reported the first case of an- ercoma of the liver in vinyl chloride workers. Sub sequent investigation of records of employees from two vinyl chloride plants in Kentucky who died between 1947 and 1973 revealed an elevenfold high6< incidence of death from cancer of the liver.and biliary tract.than expected when compared to the same age group and sex (14) of other workers in the country. Other nonma- 1 Case 1. Hepatic arteriogram, 8 seconds. A peripheral stain surrounds a large tumor within the homo of tho right hopitfc liX.u (:tr`' t'.nl hypovnscukirity is present containing large areas of contrast puddling Change* of peirocit hop.iits are present throughout most <'! la!.'.}, Casi. I. MmTc.wlf\jr ccRoid liver-spleon scan, anterior view, 1 September 1972. A bilobulnr nass Is prrsnnt within Bio right kiho r ' '0*o ttio left lobo of tlio liver 7 Caj:l II. |,1JAu iivi'r scan, anterior *n'w, A large defect present within the entire left lobo and medial right lobe. ` i'k* Departments ol fUfiohxry IJ G. IV . CUmcni Instructor. Sundry (J !.. C . Clirik\|l A-r.isl.int Piri(,-.;,or>. nut M.'di-:in.> <C H T . I i'Mil Pimm l>>r i>f Vinyl Chlciuih* hroyrcJ). Umvi!l .ily nf l nm-ivill.- l'c(hjtof f.Vilii.iilo .md SI Anthony's ll,i .pii.,|, l u<u'iv>ll>v Ky (>iriiln'..vtir4i in r.i-i'i-<'" ty/l* I I'v tin' ti V ILv-ii ViOyl i'.),inn.t<>O,mt nnl M.'hoatl C.,u < li>-.t<li,i>- runtr u 1 fiOl (*.1/ Vi.''K' All .iirllws ;uf> nicmfii-r, of tl < O'lir.men T*.,rri. Vnlyl o)<1or hh' t'lnjrc!. I It:iw-)y of Lihk^viIIi' Si t*! >! Mo In liu>, l OiM'.vitlo. Ky *.li.ui AP00008354 . 550 r Joseph G. Whelan, jr. and Or --s t i i i M,Kn Id, 6 i Ii , 2 "n. r?- Pig. 4. case in. Negate arteriogram, A. Arterial phase; 6 seconds. The hecalic artery is of normal size. Circumferential hypervascuferity is present. The intrahepatfc arteries are small and tortuous suggesting cirrhosis, B. Arterial hepatogram phase; 10 seconds. Note circumferential tumor stein and puddling of contrast agent with a central area Of hypovascularity. The stain persisted for 34 seconds. [Reproduced with permission of jama (9)J i t" / A \* . I Pig. 5. Case IV, *fcT sulfur colloid liver-spieen scan. A. Posterior view. A 10-cm defect is seen in the lower- right lobe. The spleen appears lower due to the position of the diaphragm. B. Right lateral view. The peripheral delect is present along the posterior margin o! the right hepatic looe. ilgnnnt diseases of the liver and spleen have recently been described (4. 20) and most of these changes have been defected through a systematic screening program (2. 9). Ad t. ISO wotkers of a Louisville vinyl chloride poly merization plant have had liver scans; 20 of .these indi viduals were subsequently followed with hepatic end splenic angiography, trnnsjuguiar hepatic venography, venous pressure studies and liver biopsies. Four men were found to have hepatic angiosarcoma. We will describe the angiographic and radionuclide characteristics of hepatic angiosarcoma found in these workers. / l Fig 6 CaC*. IV, Hop.il.c .inqiofjr.irn. A. Arturial ptv.so: at 4 soooiuK. ilx-,e rvj evtdwn'V* M tWiv.i-.nil.Mity Mmlm.it rii*pl.icpn','' inl/Ahnp.Mic arterial branches ot IhO rich! k>b is soon. A siricturn is suory.trw in r<ikt nrrtot n tnnuwi. tfw arc., <>l n*rijr G. Venous phase. at 14 tfcorvls tho penplioral slam is idi'iihlitkl (arrows) through thu onu* phase Scullftred .non1; of | " ' oro also pruseitt. Nolo central hypovaseulanty. 5^ 550 * .u i >t AP00008355 voi. ua IU:P/MlC /VAjiUvnnwi-x. MATERIALS AND. METHODS The radiologic screening procedure for vinyl chloride workers consists of a routine 6 ft. posterior-anterior and left lateral chest x ray, a 93mTc-suifur colloid liverspleen scan, panangiography of the liver and spleen as described by Viamonte et at. (30), transjuguiar hepatic biopsy as described by Hanafee (6) and Rosch (27). and portal pressures and venograms. If the liver-spleen scan is abnormal, or If there are persistent abnormal biochemical liver studies, panangiography is dono, including a selective hepatic and celi ac angiogram. The spleen is included In the celiac in jection. A superior mesenteric angiogram is obtained if there is an accessory or replaced right hepatic artery or if the portal vein is not visualized. In selective hepatic angiography, films are obtained at 2/second for 3 sec onds, 1/second for 4 seconds, and 1 film every 4 sec onds up to 35 seconds. Initially, the hepatic pressure studies and venography were performed via the femoral route. Later, the transjuguiar route was used in order to Include a liver biopsy. Three to five biopsies were ob tained from each patient. Venous pressures are obtained with use of the Mennen-Greathatch2 monitor. An average of three pres sures are recorded within the inferior vena cava, the right atrium, and the free and wedged hepatic vein posi tions. The Cook3 transjuguiar biopsy needle and catheter have been modified; both were lengthened 1.5 cm be cause the catheter could not oe properly wedged for the necessary pressure data in several patients, and the Teflon catheter was enlarged to 10 Fr to accommodate a 14-gauge biopsy needle to obtain better specimens. Also, the bevel of the needle was reversed to the lesser curve side to allow entrance into more sharply angled hepatic veins and better penetration of the superolater al wall of the right hopatic vein. CASE REPORTS Cass I. This 49-year-old man had worked as a chemical opera, tor in vinyl chloride polymerization for 16 years. Si/tco January 1964, he has had recurrent hematsmesis end metona. Initial upper gastrointestinal (0.1.) studios wots normal. In May 196S, hcpMosptoncmegaiy. esophageal varices, thrombocytopenia, and internal and external hemorrhoids developed. A portacaval shunt was performed and the splenoportogram was normal except for a possible 3-cm defect within the right hepatic lobe. Portal vein pressure measured 25 mm Hg. In October 1970. he underwent surgery for cholelithiasis and the liver biopsy revealod tile duct proliferation, portal fibrosis 'ad Kuplfer cell hyperplasia. In May 1971, bocause o( continued imptomotoiogy, a hepatic angiogram was obtained which revealed malignant neoplasm (Fig. t and Table i>. Biopsy suggested ana:i.ic,tic hepatoma. In September 1972 following ctmmdthcrapy. a Tc-sulfur colloid (Ivor scan demonstrated a bilobular defect * Mcnnon.C'GAtbateh Electronics. Ine., 10123 Main Street. Ctor- nen, N, Y. 10431. * Cook. Inc.. P. O. 80* 469. Bloomington. Indiana 47401. ( f Fig. 7. Case IV. Spicnie arteriogram. 10 seconds. There are two pie-shaptd delects on the lateral and inleroiateral aspect of tho spleen. Tlie spleen is not enlarged. within the liver (Fig. 2). The patient died in March 1973. Postmortem finding was hepatic angiosarcoma. Earlier liver biopsies also dem onstrated subcapsuiar end periportal fibrosis and Kupffer cell hyper plasia. Case II. This 37-year-oW man had worked as a chemical opera tor In vinyl chloride polymerization for 12 years. In September 1966, he presented with internal and external hemorrhoids which were then excised, in January 1970, painless melons occurred for the first time. Several upper CU. studies ware normal. In May 1970, molena recurred and he was found to hava hepalosplenomogaly, esophageal varices and thrombocytopenia. An ,MAu liver scan re vealed ) Urge lesion replacing tne left lobe with extension into the right lobe iFig. 3). Exploratory laparotomy revealed a markedly thickened liver capsule and hepatic angiosarcoma. In spite ol Co radiation and chemotherapy, massive ascites and liver failure devel oped and lie died in September 1971. Case III. This 42-year-ofd man had worked as a chemical opera tor In vinyl chloride polymerization for approximately 20 ye&rs. In August 19()7. he had a sympathectomy for treatment of Suorger's disease of the left tower extremity. A femoral angiogram revealed corrugation of the anierior and peroneal arteries. He had also boon hospitalized twice for alcolielism. in February 1974. through the vinyl chloride screening program (9), a *SmTe-sulfur colloid liver spleen scan detected a superficial 4-em lesion on the anterior Inferior surface of the right hepatic kibe. The spleen was normal An ,3,l rose bengal liver scan damonstrated the some defect. Hepatic angiography confirmed a lesion corre sponding to the liver scan dofect (Fig 4). The tumor margin was hyporvazeuiar and contained a hypovnscufnr center. Puddling and penphnral stain persisted from the early arterial to the late venous pluvte (TaolE Ik Splenic and portal veins were normal. Venous pres sures obtained were: wedged hepatic vein (WHV) ts min Hg and tree hepatic vein <FHV) 10 mm Hg. The free hepatic venogram was normal. The wedged hepatic venogram demonstrated portal vein re flux wiih hepntooctal flow. On exploratory laparotomy, the tumor could be seen on the surface ol the liver. Biopsy revealed angiosar coma with extensive necrosis, portal fibrosis and moderate cvrhosis In spite of chnmothnrapy, there was progressive enlargement of the tumor, massive ascites and liver failure. The patient died in March 1975. Autopsy revealed angiosarcoma. mulUlocnl Itvoughout the liver, as well as direct extension into the ihnphmgm and gnllbUddur. Mctaviases were found In regional lymph nodus, lungs, scalp anj skull. AP0d008J56 SLIM Hepatic Angiosarcoma in Vinyl Chloride Workers Diagnostic 551 Radoiggy MATERIALS AND METHODS The radiologic screening procedure for vinyl chloride `orfcers consists of a routine 6 ft. posterior-anterior and ill lateral chest x ray, a MrnTe-suHur colloid Hver- Dteen scan, panangiography of tho liver and spleen as . ascribed by Viamonte el at, (30), Iransjuguiar hepatic iopsy as described by Hanafee (6) and Rosch (27), and prtal pressures end venograms. .. if the liver-spleen scan is abnormal, or if there are srsistent abnormal biochemical liver studies, panangl- jraphy Is done. Including a selective hepatic and celf- ; angiogram. The spleen Is included in the celiac In- ction. A superior mesenteric angiogram is obtained if re is an accessory or replaced right hepatic artery or the portal vein is not visualized. In selective hepatic igiography, films are obtained at 2/second for 3 sec- ids. 1 /second for 4 seconds, and 1 film every 4 sec- ids up to 35 seconds. Initially, the hepatic pressure udles and venography were performed via the femoral ufe. Later, the transjugular route was used in order to elude a liver biopsy. Three to five biopsies were op ined from each patient. Venous pressures are obtained with use of the Mon- n-Greatbatch* monitor. An average of three pres- res are recorded within the inferior vena cava, the tht atrium, and the free and wedged hepatic vein posi es. The Cook3 Iransjuguiar biopsy needle and catheter ve been modified: both were lengthened 1.5 cm be- US6 the catheter could not be properly wedged for j necessary pressure data In several patients, and the flon catheter was enlarged to 10 Fr to accommodate 14-gauge biopsy needle to obtain better specimens. ;o, the bevel of the needle was reversed to the lesser -ve side to allow entrance into more sharply angled patic veins and better penetration of the superolater- vall of the right hepatic vein. CASE REPORTS 'asc l. This 49-yerr-oW man had worked as chemical operain vinyl chloride polymerization for 16 years. Since January 4 ho has had recurrent hemotemesis end tnelena. initial upper irointestinal (Q.l ) studies wero normal, fn May 1965, hcpatosplengnly, esopltogeal varices, ttoroniboeyloponta. and lniorn.it and vn.ll hemorrhoid* developed. A portacaval shunt was per formed the splenoportogram was normat except for a possible 3-em tet within the right hepatic too*. PortAl vein pressure measured nm Hg. in October 1970. he underwent sirgery for cholelithiasis the liver biopsy revealed bile duct proliferation, portal tibrosls KupHor cell hyperplasia. In May 1971. because Ol continued Ptumatotogy, a hopatic angiogram was obtained whirn rovait-vi >>mani neoplasm (Fig, 1 and Taku 4, Oippsy sugt-sii*.1 anasc hepatoma. In September 197? following chentothi.rncy. a '.-iullur colloid liver scan demonstrated a OtlOOutar defect '*,nufi|.(Vnntb.itch Electronics. Inc., TOtMan Stc.*ii. Ci.ir* M V. 10431. HiW. Int , P O tins -inn. ptoonuw|ti*i. Irat'.iiri 4/401 f r Fig. 7. C*se iv. Splenic aneriogrjm, 10 seconds. There ere two pie-shaped defects on the lateral and infarolaical aspect of the spleen. The spleen is not enlarged. wltitfn the liver (Fig. 2). The patient died fn March 197$. Postmortem finding was hepatic angiosarcoma. Earlier liver biopsies also domonstrated tuticapstfar and periportal fiorosis and Kupffar ceil hyper plasia. Case II. This 37-yesr-otd man had wortcod as a chemical opera tor in vinyl chloride polymerization for 12 years. In September 19CG. ha presented with internal and external hemorrhoids which wore then excised. In January 1970, painless melena occurred for me first time. Several uppur G t. studies were normal. In May 1970. metftna recurred and he was found to have hepatosplonomegAly. esophageal varices end thrombocytopenia. An ,MAu liver scan re vealed a large lesion replacing the left lobe with extension into the right lobe (Fig. 3). Exploratory laparotomy revealed a markedly thickened liver capsule and hopatic angiosarcoma. In spite of `sCo radiation and chemotherapy, massive ascites and liver failure devel oped and he died in September 1971. Case ill. This 42-year-old man had worked as a chemical opera tor In vinyl chloride polymerisation for approximately SO years, m August 1387. he had a sympathectomy for treatment of Buerger's disease of the left tower extremity. A femoral angiogram revealed corrugation of the anterior and peroneal arteries. He had also been hospitalised twice for alcoholism. In February 1974, through the vinyl chlortoo screening program (9|, a ,,Tc-uMur conoid liver spleen scan detected a superficial 4-cm lesion on the anterior inferior surfpes of the right hepatic lobe. The spleen was normal. An "`I rose bengal liver scan rtrmunjtrstcd the same defect. Hepatic angiography confirmed a lesion corre sponding to the livur scan defect (Fig. 4). The tumor margin was hv- porvusculrif and contained a hypovascuUr center. Puddling and pe ripheral stain persisted from the early ntlorml lo me late venous phase (Table t>. Splenic and portal vein* were normal Venous pres sures obtained were: wedged hepatic vein (WHVl IS mm Hg ,i,vi free hepatic vein (FHV] to mm Hg Tho hoo ticjwtic venogram wnr. normal The wedged hop.itic venogram demonstrated pnrt.il vein r.- flux with l*,p.ilupl.il Now On o>p`oKi;oiy laparotomy, v itan.it could be veen on tho emt-KV of the liver Diopty rew.Ji'if tirtplu .!!- coma wan extensive n,'t*ro ;;i. pi mat (itKpn.v ,uM cirihn- sis. fn spite Of ChemottWnpy, t*i-*ro wav [imoii'v.ivi' ,<i it tfi'i*rrwvtf nf thn lumix. m.e.itivo ,i-:i~t,.vul bwi t.ntiue tn,- {>.!,<-ivi ilu-.l .n M,vch ly.'ft Amni^y fCva.il-Mi ,in,|hvMri:irm.i nuillifui .it tfirt)ii.jJi.>i1| It hi (ivit, J-, well .e, iir t r.l.-n .mn miu Uv>- ifi.i(il,}m ;uu1 i(.lt bUiitili-' McD.i'.t.isr,at *. fixm.l hi n-cjxut.il lymph < li-n. tuiuj-.. .. ; .... . .! AP00008357 SS2 c Jose ph G. Whelan, Jr jo Others rr Mareh * Fig. 8. Case fV. Fi* hapatie vonogram. Th* hepatic vein is markedly displaced eephaisd by tumor with amputation of tributary branches (arrow). This is the "thorn sign'* es described by Doehner (Radiology 90:1124-1131. Jun 1968). Fig. 9. Cass IV. Operative portal venogram. The exposure is made at the completion of a hand injection of 49 ml of Renografin76. The fntrahepatic portal veins are marked/ displaced by tumor, almost completely compressing the left branch. The ares of tumor it hypovascular with no evidence of tumor stain or puddling. Fig. 10. '"Tc-sulftir colloid liver scan, ante rior view. A peripncral 2.5-cm delect is present on the dome of the right hopetic lobe (single arrow). A defect from the eonhuonee el hepatic veins appears more media! (double arrows!. Case IV. A 45-yeor-old man had worked as a chemical operator in vinyl chloride polymericalion lor 12 years. During a screening lest in February 1974. he was found to have a oerslstent'y elevated lac tic cid dehydrogenase. A **mTc-suffur colloid liver spleen scan re vealed a large defect on the posterior periphery of the right lobe of the liver (Fig. 5]. Hepatic angiography demonstrated a multilobular tumor corresponding to the liver scan defect (Fig. 6). The angio graphic featwes are described in Table t. Two pfe*shnpod defects were present within Ihj spleen (Fig. 7). Venous pressures obtained were: right atrium <RA) 14 mm Hg; interior vnta cava (IVC) 18 mm Hg; VVHV 20 mm Hg, and FHV 14 mm Hg. One hepatic vein branch was partially compressed by tumor (Fig. 8). On wedged venography, there was portal vein reflux with some portal stagnation. Exploratory Isparoiomy and biopsy revealed hepalie angiosarcoma, capsular fi brosis, sinusoidal dilatation and portal fibrosis. The spleen was nor mal to palpation. A portal venogram riumariclratcd marked displace ment by tumor (Fig. 9). Portal vein pressu'o measured 20 mm Hg. In spite of *aCo radiation and chemotherapy, the tumor pro gressed. Massive ascites and liver failure developed and the patient died tn July 197S. At postmortem, tho angiosarcoma was multifocal throughout the liver and extended into the diaphragm. Them were extensive metnstases to regtonal, i:trajv:ritoitoal and mudiar.tin.il lymph nodes, lungs, adrenal ginrtrJs and c.-rulmlUurv Th>* spleen was normal and froe of motastjsis in spite of thu angiographic twvlmgs. Case V: Mao-posiflvo jnptosnrrwna. Tim 50-y< nr-oW man . hail beon a chemic.il upr>i.it<w hi vinyl rjikvidn (Kdymcii/.ituMt for tfl yu.ns. It* Mjy tin innjfi Itto vinyl lihlondu M'leeuintj jmxji.iin, a *tmTc-suifur colloid liver spleen scan was done which suggest defect on the dome of the right hepatic lobs (Fig. 10), that was In the anterior and right lateral views, initial hepatic angiogr, was unsuccessful. The following day, transjugular hepatic ve. studies and biopsy were performed, A wedged hepatic venor demonstrated portal vein reflux with hepatopetai (low (Fig, ll). day later, e repeat hepatic angiogram vis the transaxflUry rout vested a 3 5 X 4,5-cm oval lesion within the superior right hej lobe consisting of a peripheral stain with a central hypovasi area (Fig. 12). Seven days toter. a right hepatectomy was performed bee. the angiographic pattern resembled trot of angiosarcoma On examination, the lesion appeared as a 2-cm pyramidal lesion the base on the liver capsule. This was found to be a boating ini which corresponded to tno exact configuration'and area of wedged venogram. In addition, there was some degree of port, brosis. PATHOLOGY Hepatic angiosarcoma Is a tumor believed to ar from the endothelial colls lining tha sinusoids. 7 tumor is also referred to as malignant hemangio dothelioma. or Kupffer cell sarcoma. The histolo characteristics have recently been well described ( 22, 29), and the spread of this turner is mainly by dir invasion and by the coeiomic route (22). Metast; tumor has been found within portaf veins, bone marr and peritoneal cavity, lung, spleen and the thyroid 10, 22, 29). Edmundson reported that up to 50% those tumors have pulmonary mctaslases (3). Seven of the 8 vinyl chloride workers had turn* outside the liver. Thrco of these directly invaded the nphragm. one extended into tho duodenum and anoti had readied the posterior abdominal wall on followexploration. There is no histologic evidence of splu involvement in nny of those cases. Subcapuular fibrosis, portal fibrosis, sinusoidal dito (ion and atypical sinusoidal lining colls have been ioi to coexist with fhis tumor (2, 4, 10, 22, 29). il is prt willy full Hint (ho lesions ol poitnl fibrosis, sinusoidal l.ii.iiion. and atypical lining cells mny represent a pr AP00008358 i. ir* Hepatic Angiosarcoma in Vinyl Chloride Workers Diagnostic 553 Radiology I > v t Fig. 11. Tranijugufar wedged hepatic venogram: 6 ml of Renografin-76 injected at 2 ml par second. Thera is portal vain reflux with hepatopstal flow. Fig. 12. Hepatic arteriogram 2 days following the wedged hepatic venogram. A 3.5 X 4.5-cm oval lesion containing a peripheral stain and hypovoscular center persisted into the late venous phase. This corresponds to the area of the wedged * hepatic venogram which was histologically proved to be an Infarct. rsor stage in the development of angiosarcoma (4). had varying degrees of portal and subcapsular fibrosis. I groups of vinyl chloride workers with tumors have Up to 40% of these tumors have been reported to have s I i I ig IJ A ra-yu-w-oM vinyl ctvtoi*l>` worker with nnnorm.il liv.v iuii.nuvi >.!niis ,m.-l m>niril hvir serin A lli*|MtiC nngtOijr.ini. iv fviiciv.im pfshin, 24 Ae IH7*. ni'.inlmi'Oii of cw>Im -t wni k (vnm il II Hop.ll'C now r`ikx]f,iin, 2U Aii>j ItlT-t A auvil.iim :'.<tt><M|i .ul.u '.t.iin c; will) jmii (.>1 vui tolUi* .vhl n..p.|o- It hi'iMttC AtlilKjij'rim. Iip'jmliiiy.im ptviv*, 30 Ain) 13/4 A in-tipluM.il t >.>< ' ***-. .il i> ,il pi'fM'.fs up t(l 31 ^CondS. A loltful-w >ii t n( < t o i -ri M* CtTrti.it ,in> i 4 iyp*>v.it.cnl.it\y. niiv-.pnvtnv)tul'w> AP00008359 Arterial Hcpjttc artery size MO tumor) Encasement Displacement Hypcrvascufority Central vascularity A. V. shunts Peliosis henatiS Tvmor stain Central Peripheral Time Maximum stain Puddling time Venous Portal Last film in series Table Is Angiographic Features of Hepatic Angiosarcoma Case l Case 111 Normal No Yes Yes (2 sec) Hypovascular No Yes (left lobe) No Yes 4-15 sec* 8 sec 2-15 sec* Not demonstrated (portacaval shunt); splenic vein; 11 sec Normal No No Yes (2.S see) Hypovascular No No No Yes 5-34 sec* 10 sec 5-14 sec 10 see Caso IV Normal No Yes No Avascular No No No Yes 6-30 sec 6 see 6-40 sec* 13 sec coexisting cirrhosis (22). The relationship of cirrhosis to this tumor is not determined. Only 2 of the 8 vinyl chlo ride workers with tumor had cirrhosis. DISCUSSION Uver and Spleen Scans Angiosarcoma appears as a negative defect on liver scan with use of either 1MAu, ,3,l-rose bertgal, or mTc-sulfur colloid. Sufficient #7Ga studies have not been performed to be conclusive. Although most le sions appear Initially as a single focus, one patient pre sented with two separate lesions. At serial scons and autopsy, the tumor Is seen to progress to multifocal le sions throughout tha liver. Even though most lesions, when detected, were rather large, they .appeared pe ripheral in location. This was also found to be true at surgery. In the tangential view, there is no radioactivity surrounding the defect and no activity within the defect itself. We used the camera technique to view the stud ies of 2 other patients with this tumor. These also ap peared as peripheral lesions. Without the history of vinyl chloride exposure, it would be difficult to differen tiate this lesion from an extrahopatic contiguous lesion (24). it is also difficult to differentiate from normal ana tomical defects such as portahepatis, gallbladder fossa, confluence of hepatic veins, or rib and xiphoid. Periph eral development of this tumor is possibly related to the subcapsular fibrosis found in these workers (7. 29). Their peripheral location is also the most likely reason that these tumors show early direct extension through the liver capsule. Four of the 8 vinyl chloride workers with tumor had splenic enlargement, demonstrated in one patient after the tumor was delected. Three of these patients pre sented with Bantl's syndrome. Watson and Bnttnxe (31) altribulod splenic enlargement to massive replacement of liver parenchyma, or invasion of tho portal vein by tumor. Splenic enlargement occurred in one of our pa tients (Case I) prior to massive tumor involvement. In 3 others, portal venograms did not demonstrate occlusion or invasion. On postmortem examination in Cases III and (V, no tumor was demonstrated within the portal veins. We have found, in fact, approximately 10% of the 1,160 employees screened to have splenic enlarge ment without evidence of tumor. Splenic enlargement may be due to stimulation of splenic cells by chronic exposure to vinyl chloride, thus producing an increase in splenic blood flow. Further hemodynamic studies are currently being conducted. Angiographic Characteristics Table f contains a summary of the angiographic fea tures in Cases I, HI and IV. The hepatic arteries are of normal size with some displacement by tumor in Cases I and IV, and of increased size of the left hepatic artery not supplying the tumor in CASE i. Hypervascufar changes were present in Cases I and HI around tho pe riphery of the tumor but not in Case IV in which the number of terminal vessels decreased. No true tumor vessels were demonstrated. Changes of peliosis hepatis as, described by Pliskin (19), were present within the left lobe in Case I. The central portion of tumors in these patients has some degree of hypovascularity. most likely secondary to necrosis as was found in the biopsy of Case ill. A peripheral tumor stain was present in all 3 cases, appearing in the mid&rteriai phase and extending up to approximately 34 seconds. The stain in Case I was faint at 15 seconds (the last film obtained In that series). The hemodynamics of the hepatic circula tion may have been adored in this case due to the pre vious portacaval shunt. Puddling within Iho hypovascuiar portion of the tumor appeared for the same period of time. The amount of puddling seemed to increase wit*1 the increased size of the tumor, again, probably roloied to the amount of necrosis. Maximum stain and puddlind occurred prior to tho portal vein phase: these change4 appeared closer to tho portal vein phaso in the small'1' AP00008360 s 2 * 2 Z " I " Z 2 22 O 0 O CD SsSSals* fiSSSa-J< 0 o 0 no o o o * f o! ? S 5 *? Sf ?s I IX XXX X 3 21 *7 2 X 2 X 3 722 7 2 22 2 * 7" I* f? 2 Sai;o o 00 0 ill Hf *- r- _ s- - s- -- W^ I * . S * i ? S 1 I i. -i -- O ~ -is-- X~ $? S ' B. 3 2? ?2 IX 111 *x||2 II 7 73 2 iL 9 3~ IX IIIIn2 2X21 o 7iS- ov xf I? 22 v 2 2 g^l22<? -fncSpjr > 3 =.- 2 u* |S|S?I ?* -- ka C>b AP00008361 tumor In Cash til. No stain or puddling could ho demon strated in the portal venogram in' Case IV during sur gery. The tumor, therefore, appears to receive its major blood supply, from the hepatic arlary. Even though the peripheral slain persists wotl into the venous phase, lit tle ol this appears (o bo due to the porta! circulation. The peripheral stain in these casos probably represents sinusoidal dilatation (2. 29). The portal and hepatic veins In these cases are displaced but have not been obliterated. The angiographic characteristics of hepatic angiosar coma, therefore, appear to be: (a) a normal-size he patic artery supplying tlm tumor which becomes dis placed as the tumor enlarges; (b) a peripheral tumor slain and puddling appearing durjng the midarleriai phase and persisting as late as 34 seconds; and (c) some degree of hypovascularity of the central portion Of the tumor including the smaller tumors. Angiographic Differentia! Diagnosis Multiple vascular lesions of the liver in adults are to be considered with hepatic angiosarcoma (Table tl). Al though some variations exist, this review includes the major description of these .lesions. Well-differentiated hepatomas differ from angiosarcoma due lo enlarged hepatic arteries with true neovascularlty and a rather diffuse stain which lasts only through the early venous phase (8, 16, 20). Multicentric avascular hepatomas (26), anaplastic hepatomas (8), and cholangiocarsinomas (7, 6) are excluded because of their relative avascularity with smalt scattered areas of stain and lack of puddling. Liver cell adenomas are excluded due to the presence of neovaseularity and lack of a peripheral stain (18). This lesion could simulate angiosarcoma if central necrosis develops as In the case described by Sacket et at. (28). A large cavernous hemangioma has peripheral puddling of contrast material surrounding a central avascular area of fibrosis (11. 17, 21). However, there is no evidence of peripheral stain. Metastatic le sions (15, 20) and hamartomas are excluded since a stain appears only during the arterial phase. Recently, bile duct hamartomas of the liver have been visualized as clusters of ring-like stains lasting from the arterial to the tute venous phnso. diffusely involving the liver (12). Angiosarcoma has not presented with this pattern in any of our cases. Macronodular regeneration of cirrho sis can be excluded since the stain appears rather uni form in the venous phase with no evidence of puddling (13. 25) and the scan may also be normal (25). Focal nodular hyperplasia mny appear os a diffuse stain in the arterial and venous phases (13, 25); however, there is no evidence bf central hypovascularity.or puddling of contrast substance. Here, also, the scan may be normal (25). Angiographic changes of poliosis hnpntis have re cently boon described (10). This may also represent tho changes as seen in hepatic vein occlusion (5) and those described by Vinmonte et ol. (30) in longstanding portal vfein obstruction. These lesions, however, tend to be numerous, involve the entire fiver, and have d diffuse stain throughout tho nodules. The stain does seem to persist into tho late venous phase. In 4 of the 50 vinyl chloride workers evaluated, varying degrees of peliosis hepatis have been demonstrated angiographicaily. This pattern Is also suggested within the loft lobo in Case I. Some of these patients are being followed closely since this histologic change may be one of the prelimi nary findings before the occurrence of angiosarcoma in vinyf chloride workers (2. 4). These nonmnlignont changes will be the subject of a future publication. Hepatic infarction secondary to wedged hepatic ven ography was found to ciosefy resemble angiographic changes of angiosarcoma and, in fact, led to a right he patic loboctomy in Case V. At the time of this discov ery, another worker had just been studied by hepatic angiography, followed by wedged venography (Fig.-13. A and 8). Repeat angiography confirmed an area of in farction (Fig. 13, C). Serial liver scans failed to detect this lesion. In 3 cases in which the wedged venogram was obtained first, the angiographic pattern was similar. Hepatic Venography and Venous Pressures Venography and pressure studies ore carried out to evaluate the overall condition of the liver. The free he patic venogram was normal except when enlarging tumor has caused displacement. No occlusive changes have been demonstrated. Because of iatrogenic lesions created by wedged venography in 3 patients, this study has been discontin ued as part of the routine screening procedure. All vinyl chloride workers were asymptomatic following wedged venography. Serial'hepatic scans from 1 day to 7 months postinjection have been negative. The small he patic infarct is probably insignificant due to the marked regenerative power of tho liver. Wedged hepatic ven ography is still used in cases of nonopadification of the portal vein during angiography to differentiate reverse flow from obstruction. This study, if performed, should follow angiography. Hepatic venous pressures were elevated In Cases III and IV. In a number of these workers, the free hepatic vein pressure and the inforior vena caval pressure were found to be elevated from 7 to 25 mm Hg. The majority of the corrected sinusoidal pressures Iwvo been within the normal range* (0-5 mm Hg). Poital pressure, how ever, may not correlate with the corrected sinusoidal pressure since periportal fibrosis is one of the predomi nant lesions found among tlvose workers. A more de tailed investigation is being conducted to further evalu ate the systemic and splanchnic venous beds. SUMMARY Isotopic and angiographic studies of hepatic angio sarcoma tn vinyl chloride* workers me presented. Of the screening procedures included in our project (9). the AP00008362 .. .1 :t:. ,I , 1/ irJ- in i-n* he tlV :ic 13. in- er.t un :-r. to *d n >yi d 7 d- d |1- e 0 id Ut t? y n t liver scan has be'-., the most useful in detecting the tumor. No singlo biochemical fiver test has. no yet. con sistently been found to ho abnormal in these patients. The peripheral defect on the fiver scan may be difficult to differentiate from a normal anatomical defect or an BXtrahepatie lesion contiguous with the liver. The angiogrophic features which are characteristic, if not diag nostic, of this lesion are peripheral tumor stain, pud dling Of contrast agent extending from the midartcriat phase up to 34 seconds, and some degree of central hypovascularily within the tumor. Transjugulnr hepatic pressures and biopsies are essential procedures since systemic venous hypertension seems to develop In some workers. Subcapsutar fibrosis, portal fibrosis, si nusoidal dilatation and endotheiiai lining cell hyperpla sia are associated with the tumors. Follow-up studies will be helpful to determine whether these changes are reversible or can be arrested after removal of these men from vinyl chloride exposure, and the effects of therapeutic and other treatment regimens. AcKMOWLtoQMeNrs: Our gratitude ) extended to Ors. John Watte, William Joule end John M?Ctana. Associate Radiologists, and to Ora- Will Ward. James Kurfees, Robert Arnold, and Stanley Low. nbraun tor their assistance and cooperation. A special thanks to Mr. Jack Oltty, senior medical student, for his extensive review of the case histories. Department of Radiology St. Anthony Hospital 13 )3 St. Anthony Place Louisville, Ky. 40204 REFERENCES 1. Creech JL Jr. Johnson AIN; Angiosarcoma of liver in the manufacture of polyvinyl chloride. J Oecup Med 18: t5C-151, Mar 1874 2. Creech JL Jr. Makk l_ Wtiotan JG Jr, ei al: Hepatotoxicity among polyvinyl chloride (RVC) production workers during first year of surveillance program (abst), Gastroenterology $7:728, Oct 1974 3. Edmondson HA: Haemangioendo.theHal sarcoma, fin} Tu mors of the Liver end Intrahepaiic Site Ducts. Atlas of Tumor Pathol ogy, Sec VII. Fasc 25, AFIP, Washington D. C-, 1958 4. Falk H, Creech JL Jr, Heath CW Jr. et al: Hepatic disease among workers at a vinyl chloride polymerization plant. JAMA 230: 59-63, 7 Oc.t 1974 5. Galloway $, Casarelta WJ, Price JB: Unllobuiar vcno-occtusiva disease of the fiver. Am J Roentgenol 119:99-94, Sep 1973 6. Harulee WN. Weiner M: Tronsjuguiar percutaneous chol angiography. Radiology 98:35-39. Jan 1967 7. Kaudo J, Rian R: Cholengiocarcinoma. Radiology 100: S73-S80. Sep 1971 8. Kido C. Sasaki T, Kaneko M: Angiography of primary fiver cencer.AmjRoemgcnol 113:70-91.Sep t9?t 9. Makfc L. Creeeh JL, Whelan JG Jr. at at: Liver rlnm/u;. angiosarcoma in vinyl chloride workers. A systematic doloetioi > gram. JAMA 230:64 -08. 7 Oct 1974 10. Makk L. Delorme F. Creech JL. et ai: Clinical and mnteg-cai lentures of hepatic angiosarcoma^ in vinyl chlocwlo woi Cancer (to bo published) 11. McLoughlin MJ: Angiography in cavernous homnngior the liver. Am J Roontgcnot 113:50-55. Sep 1971 12. McLoughlin MJ. Phillips MJ: Angiographic findings in r pi* bile-duct hamartomas of the livor. Radiology 1i6:4t-43 1975 13. McMullen CT, Montgomory JL: Arterioyaphlc (iivtirvffocat nodular hyperplasia of the (aver and review of the litor.r Am J Roentgenol 117:380-387. Feb 1973 14. Monson. P.R. Poier* JM. Johnson MN: Proportional mo> 1y among vinyl-chloride workers. Lancet 2:397-398. 17Auij ID,* T5. Nehosur RA, Pollard JJ. Edmunds LH, ei at: Indication seiectivo eotiae and superior mesenteric angiography. Am J Rr gonol92:il00-ti09.Nov 1964 16. Nebesar RA. Pollard JJ, Stone DL: Angiographic diaqn ol malignant disease of the liver. Radiology 66:284-292. Feb 10' 17. Neiman HL, Goldstein HM: Angiography of benign and : lignant hepatic masses. Semin Roentgen to-.197-205, jut to ?s 18. Paiucmskas aj, Baldwin J, McCormack Kft: Livcr-ccH c. noma. Radiology 89:444-447, Sep 1957 19. Pliskin M; Peliosis hepatis. Radiology 1t4;29-30, , 1975 20. Pollard JJ, Fleischli OJ, Nabesar RA: Angiography of patic neoplasms. Radiol Clin N Am >;3t-41, Apr 1970 2t. Pollard JJ. Neoesar RA. Manoso LF: Angiographic ding sis of benign diseases of the liver. Radiology 88:276-263, Feb 1C- . 22. Pollard 5M, Miilward-Sadler CH: Malignant haemangiojJothclioma Involving tl>* liver. J Clin Path 27:214-221, Mar 1974 23. Port RB, Petasroek JP. Ranniger K; Angiographic Oem stralion of hepatoma in association v;ith Fanconi's anemia. An Roentgenol 113:82-83, Sep 1971 24. Potchen J. Addsteifl SJ, Hoffer P8. et el: Nuclear Rftd ogy Syiiabus, Set 7, American College Radiology. Chicago, t 1974.pp 182-164 25. Rabinowliz JG. Kinkabwala M, (Jlreich S: Macrowegencr. ing nodule of the cirrhotic liver. Am J Roentgenol 121:401-411. J. 1974 26. Reuter SR. Redman HC. Slders 09: The spectrum of a Biographical findings in hepatoma. Radiology 94:89-94, Jan 1970 27. Rosch j. Lakin PC, Antonovic R. el al: Transjugular a proach to liver biopsv and transhtpatic cholangiography. N Engi Med 239:227-231. 2 Aug 1973 28. Sackett jF, Moscnihei WT, House RK, et ai: Scintillate scanning of liver cell adenoma. Am J Roentgenol 113:58-60, S< 1971 29. Thomas LS, Popper H, Berk PO. et at: Vinyl-chloride li duced liver disease. From idiopathic portal hypertension (Q.antt syndrome) to angiosarcomas. N Engl j Med 292:17-22. 2 Jan I97i 30. Vi&monte M Jr. Warren VVD, Fomon JJ: Liver pnnang ogrepby in tho assessment ol portal hypertension In Nvcr cirrhosi Radiol Clin N Am 8:147-167. Apr 1970 31. Watson RC, Baltaxe HA: The angiographic appaarnnee primary and secondary tumors of the liver, Radiology 101:539-54? Oec 1971 32. Wilson RH. McCormick WE. Tatum CF, et at; Occupations eeroosteolysiS. JAMA 231:577-581, 21 Aug 1967 .) 1f AP00008363 sal v Hepatic Angiosarcoma in Vinyl Chloride workers diagnostic 557 ftidetogy sr scan has been the most useful in detecting the nor. No single biochemical liver test has, as yet. con tents been found to be abnormal in these patients, e peripheral defect pn the liver scan may be difficult differentiate from a normal anatomical defect or an trahepatic lesion contiguous with the liver. The angio- iphic features which are characteristic, if not diagStic, of this fesion aro peripheral tumor stain, pud- ng of contrast agent extending from the midarterial ase up to 34 seconds, and some dogree of central povascularity within the tumor. Transjugular hepatic assures and biopsies are essential procedures since stemic venous hypertension seems to develop in me workers. Subcapsular fibrosis, portal fibrosis, si- soidal dilatation and endothelial lining cell hyperpiai are associated with the tumors. Follow-up studies II be helpful to determine whether these changes are versible or can be arrested after removal of these sn from vinyl chloride exposure, and the affects of srapeutic and other treatment regimens. acknowisogmcnts: Our gratitude is extended to Drs. John tfls, William joule end John MpCtane, Associate Radiologists, and Ors. Will Ward. James Kurfees. Robert Arnold, and Stanley Low* xaun for their assistance end cooperation. A special thanks to . Jack Ditty, senior medical student, lor hit extensive review of case histories. Bailment of Radiology Anthony Hospital 13 Si. Anthony Place uijvlIle.Ky. 40204 REFERENCES 1. Creech JL Jr. Johnson MN: Angiosarcoma of liver In the '.nufaeture of polyvinyl chloride. J Occup Wed 16:150-151, War 74 .2 Creech JL Jr. Makk L. Whelan JG Jr, et al: Hepatotoxicity long polyvinyl chloride (PVC) production workers during first year survoilljnc* program tabs!). Gastroenterology 67:766. Oct 1974 3. Edmundson HA: Hnomangioondothotinl sarcoma, (in [ To ys of the Liver and Intrafrepstic Bite Ducts. Atlas of Tumor Pallioty. Sec VII, Fasc 75, AFlP, Washington 0. G., l9SS 4. Falk H, Creech JL Jr. Heath CW Jr. el al: Hepatic disease tong workers at a vinyl chloride polymerization ptant. JAMA 230: t-63. 7 Oct 1974 5. Galloway S, Casarelln WJ, Price J8: Unilobuljr veno-oeJsivo d'seose ol the liver. Am J Roon'ocnol 1,19:89 94. Sep 1973 6. Han.ifce WN, Weiner M: Tran^ugul.v percutaneous ehol9*1Faphy. ft.idiaiogy &6:3S-39. .Lm (967 7. ' Kaude J, Rian R; ClKrtangiocarcmomn. Radiology 100: '3-580, Sep 1971 6. Kido C. Snsaki T, Kaneko M: Angiography of primary liver wser, Am j Roentgenol 113:70-61, Sep 1971 9. Makk L, Creech JL. Whelan JG Jr. et al: Liver damage and angiosarcoma in vinyl chloride workers. A systematic detection-pro gram. JAMA 230:64-68, 7 Oct 1974 10. Makk L, Delorme F. Creech JL. et el: Clinical and morpho-' logical features of hepatic angiosarcoma in vinyl chloride workers. Cancer (la be published) tl. Mcloughlin MJ: Angiography in cavernous hemangioma of the fiver. Am J Roentgenol 113:50-55. Sep 1971 12. McLoughltn MJ. Phrlfips MJ: Angiographic findings In multi ple bile-duct hamartomas of die liver. Radiology 116:41-43, Jid 1975 13. McMullen CT. Montgomery JL: Artcriographie findings of focal nodular hyperplasia of IIh* liver and review of ttte literature. Am J Roentgonol 117:380-367. Feb 1973 14. Monson. RR. Peters JM, Johnson MN: Proportional mortali ty among vinyl-chloride workers. Lancet 2:397-398. 17 Aug 1974 1$. Nebesor RA, Pollard JJ. Edmunds LH. et at: Indications for selective celiac and superior mesenteric angiography. Am j Roenttfenol 92:1100-1109. Nov 1964 16. Nebesar PA, Pollard jj. Stone OL: Angiographic diagnosis of malignant disease of the liver. Radiology 66:284-292. Feb 1966 17. Neiman HL, Goldstein HM: Angiography ol benign and ma lignant hepatic messes. Semin Roontgen 10:197-205. Jul 1975 16. Paludinskas AJ. Baldwin J, McCormack KR: Liver-cell ade noma. Radiology 66:444-447, Sep 1967 19. Ptiskin M: Psliosis hepatla. Radiology 114:29-30, Jan 1975 20. Pollard JJ, Floischll DJ, Nebesar RA; Angiography of hepelic neoplasms. Radiol Cnn N Am 8:31-41. Apr 1970 21. Pollard JJ. Nebesar RA. Mattoso LF: Angiographic diagno sis of benign diseases of the liver. Radiology 86:276-283. Feb 1966 22. Pollard SM, M.ilward-Sadier GH: Malignant haomangioen- dothelioma involving the liver. J Clm Path 27;2t4-221. Mar 1974 23. Port RB. Pe'asniek JP, Ronnigar X: Anryographie demon- tralion of hopntoma In associi'tisn with Fanconi's anemia. Am J Roentgenol 113:82-83. Sep 197.1 24. Potchen Ej. Adelstein SJ. Hoffer PB. et alt Nuclear Radiol ogy Syllabus. Sel 7. American Cottage Rddlotopy. Chicago. III.. 1974, pp 162-164 25. Rablnowitz JG. Kfnkabwula M. IBrnich Si Macro-regenerat ing nodule of Ihe cirrhotic liver. Am j Roentgenol 121:401-411. Jun .1974 26. Reutor SR. Redman HC, Siders OB: The spectrum of angeographical findings in hepatoma. Radiology 94:89-94, Jan T&70 27. Rosch J. Lakin PC, Amcnovic R. et al: Transju-uier ap proach to liver biopsy and banshepanc cholangiography, N Engl j Med 289:227-231. 2 Aug 1973 . 28. Socket! Jr, Mosenthn) WT. House RK. el al: Scintillation scanning of livor coil adenoma. Am J Roentgenol 113:50-60, Sep 1971 29. Thomas LH. Popper H, Berk PO. et al: Vinyl-chloride in duced liver disease. From idiopathic portal hypertension (Banti't syndrome} to angiosarcomas. N Engl J Med 292:17-22. 2 Jan 1975 30. Vinmontc M jt, Warren wo. Femon JJ: Liver panangL OCjf.iphy n Iho .issiessmont of pofial hypnrtonr.ion m liver cirrhosis. Radiol Clm N An* 8:14 7- 167, Apr 1970 31. Watson RC. Oattnm HA: The angiographic appearance of primary onJ secon l.tiy uimors of tho kver. Radiology 101:537-546. Dec 1971 32. Wilson RH, McCormick WE. Tatum OF, et at: Occupational acroostcofysis. jama 201:577-581, 21 Aug 1967