Document BR2LvkMdMRROB2KVOKL8Ozb4k

R&S 114633 m BIO-MEDICAL RESEARCH DOCUMENT DESCRIPTION FORM Duplicate.in all cards: 63 68 69 7I 76 ^ year as-1961- File number [Right justify rNuwerie onlvl 77 78 II Sub-Index Code Author (s), as Last Name FS (No Punctuation) and coden for journal as JAMA preceeded by one blank space 1 ^3 20 21 _________ __________ 40 41_________._______________ .y 60 61 62 11 12 13 Title of Report; end with space-hyphen-hyphen-space. Follow with Index Terras * separated from each other with comma-space. Avoid other punctuation; do not abbreviate. ,, , l . -~W49.t ii.9 /&P y.-y-; //K'*'-'. ^s ' /)O X7~-- *-- - '2'', - .?. / 7 JF/?* *-**.4 A 3Cr~ 21 22 23 m_________ ________ Source (Journal, Vol., Number, Pages, Date ) jr /y's f? ^7'' 1 jrT' K~:`jei?s~~r '/y.r ' y y" 7 ' /-'ri* ' .. *1 62 31 32 Brief Summary 12 10 SUMMARY: 61 62 61 62 63 64 `po Box No 6 Bessemer Road Welwyn Garden City Hertfordshire AL7 1HD Telephone Welwyn Garden 23400 (STD Code 07073) Telex 264251 Dr T R Torkelson Health & Environmental Science 1803 Building Dow Chemical Midland Michigan 48640 USA Your ref. Our ref JS/SEW/DS0-16 Tel ext 3162 icn Chemical Industries Limited Petrochemicals & Plastics Division Date 18 August 1981 R&S 114634 Dear Ted PVC DUST You may already have a copy of this paper. However, I remember that the last time we got any information about this work it came from Trent Lewis to our Government, to us and eventually back to American industry. X know we all heard Groth give this paper in March 1980 but just in case I am sending you a copy for your file. The paper did remind me that it is now a Tong time ago since we had that famous Symposium in Bethesda. We knew at the time that it was a bit of a non-event but I had expected to find the papers published by this time. Obviously NIOSH and the other agencies have had other priorities imposed on them since that time! With best wishes. Yours sincerely J Stafford Health & Environment Protection POLYVINYL CHLORIDE DUST by David H. Groth Dennis W. Lynch William J. Moorman Lloyd E. Stettler Irene R. Lewis , William D. Wagner Choudarl Kommineni Division of Biomedical and Behavioral Science Department of Health and Human Services National Institute for Occupational Safety and Health Robert A. Taft Laboratories Cincinnati, Ohio 45226 R&S 114635 ABSTRACT Rats, guinea pigs and monkeys were exposed by inhalation (6 hours/day, 5 days/week) for up to 22 months to a 10 mg/mJ concentration of PVC dust. Autopsies on rats and guinea pigs were performed after 12 months of exposure and on monkeys after 22 months of exposure. Lung function tests were performed on monkeys after 9, 14 and 22 months of exposure. Aggregates of alveolar macrophages containing PVC particles were found in the lungs of all animals. These aggregates were more numerous in the monkey lungs. No fibrosis or significant cellular infiltrates were present in or near these cellular aggregates. No significant effects on pulmonary function could be demonstrated in the monkeys exposed to PVC. Under the conditions of this experiment, inhaled PVC produced a benign pneumoconiosis. R&S 114636 R&S 114637 INTRODUCTION Approximately seven billion pounds of polyvinyl chloride (FVC) were produced in the United States in 1979. There are two major types of PVC production processes, suspension polymerization and emulsion polymerization. The latter process produces particles which are respirable and much smaller than those in the former process. These polymers of vinyl chloride are usually compounded with a variety of ingredients (e.g., plasticizers, light and heat stabilizers, pigments and fillers) and processed in several different ways to produce thousands of end products in common use in our society. About 40Z of PVC is used to make sewage pipes, water pipes and conduits. It is also used to make construction siding, window sashes, electrical wire and cable insulation, packaging films (for meat, etc.), vinyl floor tile, wall coverings, phonograph records, shower curtains, bottles, fabrics for clothes, furniture, automotive parts and many other products. Chest x-ray abnormalities, respiratory dysfunction and pulmonary granulomas have been reported in workers exposed to PVC dust during PVC manufacturing and fabrication (Tribukh, et al., 1949; Szende, et al., 1970; Vertkin and Mamoutov, 1978; Mastrangelo, et al., 1979). Consequently, experimental studies were initiated in 1975 to study the effects of inhaled PVC dust in animals. R&S 114638 MATERIALS AND METHODS Test Material Ten pounds of PVC (trade name Geon 121) were obtained from Goodrich Chemical Company and used in this experiment. The manufacturer's specifications stated a particle size range of 0.5-1.5 urn. Electron microscopic examinations in this laboratory confirmed that at least 90% of the particles were less than 1.5 um in diameter. The PVC was stored in its covered, fiber board shipping container throughout the duration of the study. Dust Generation and Measurements PVC powder was packed daily or every other day into a Wright dust feeder at a pressure of 3,000 lbs./in. . The generated dust was fed into the mainstream of the inhalation exposure chamber intake air supply and through a static eliminator before entering the 160 ft.^ chamber." ""Airflow through the chamber was maintained at 40 ft.^/minute with a negative chamber pressure of 0.2 inches of water. Animals were exposed to PVC dust 6 hours per day, 5 days per week for up to 22*5 months. Four chamber dust samples were collected each day by drawing chamber environment air, at a rate of 10 liters/minute for 20 minutes, through DM 5 ym pore size mecricel filters with a vacuum pump. In addition, simultaneous samples consisting of particles with an aerodynamic diameter =7 ym (respirable fraction) were col lected with a 10-plate, horizontal elutriator. All samples were weighed immed iately after each sampling period and the chamber dust concentrations calculated. Adjustments in the dust generating system after each sampling period were made when necessary to maintain a concentration approximating 10 rag of respirable 2 / dust/m3 of air. Daily and grand means (derived from averaging the daily means) for the total and respirable dust fractions were calculated. Midway through the experiment two, six-hour chamber air samples were collected on charcoal tubes while PVC dust was being generated. Those two samples and the bulk PVC were analyzed for vinyl chloride by gas chromatography. Samples of dust that were allowed to settle on formvar-coated copper grids placed in the chamber during the PVC dust generation were examined by transmission electron microscopy. Animals Animals used in this study were male. Cesarean-derived, Sprague-Dawley rats (Laboratory Supply Company, Inc., Indianapolis, Indiana); male. Hartley guinea pigs (Sweetwater Farms, Hillsboro, Ohio); and imported, adult, mala Cynomoigus monkeys (Primate Imports Corp., Long Island, New York). The treatment and control groups each contained 80 rats, 40 guinea pigs and 10 monkeys. Because of difficulties in obtaining imported monkeys from the supplier, the control group of monkeys was received several months before the exposed group of monkeys. They were not randomly assigned, because the control group was also used for the control group in another ongoing experiment. Consequently, at time of first exposure, the mean weight of the control monkeys was 4219 grams and that of the exposed group 3361 grams. In the ensuing 224 months, the controls gained 552 grams in weight, whereas, the exposed group gained 2116 grams. 3 R&S 114639 The rats and guinea pigs were quarantined for two weeks and the monkeys were quarantined for one month prior to the initiation of the inhalation exposures. Stainless and galvanized steel open wire mesh cages were used as exposure caging to provide adequate distribution of the dust aerosols within the exposure chambers. All of the animals in the study were individually marked by toe-clipping or tatoo. All three species were individually housed during the 6-hour exposures, whereas the rats and guinea pigs were housed two to four animals per cage at all other times. Control animals were housed in similar cages in separate animal quarters and exposed to filtered air 24 hours/day. The exposed animals were also housed in the animal quarters except during the 6-hour inhalation exposure. Rats, guinea pigs, and monkeys were fed standard laboratory pellet diets (Rodent Labora tory Chow, Guinea Pig Chow, and Monkey Chow-Jumbo from Ralston Purina, St. Louis, Missouri). Monkeys were given fresh fruit (oranges, bananas, or apples) twice a week. Tap water was available libitum except during the exposure period. Food was available to the rodents at all timer except during exposure. The monkeys were fed once daily at the cessation of the exposure period. Pathology Within one day after the last exposure day, all the surviving rats and guinea pigs were autopsied. Their lungs were inflated with 10% buffered formalin and sections of liver, spleen, heart, kidney, pancreas, adrenals, thyroid, testis, and urinary bladder from each animal were fixed in 10% buffered formalin. Hematoxylin and eosin stained sections of each lobe of each lung and of each of the above tissues were prepared for and examined by light microscopy. R&S 114640 4 R&S 114641 Within 48 hours of the last exposure day, all surviving monkeys were autopsied and sections of the same aforementioned tissues plus tracheobronchial and mesenteric lymph nodes, prostate, stomach, duodenum and skin were fixed and processed in the same manner as those for the rodents with the exceptions that two sections from each lobe of the lungs were examined, and some of the lung tissue was processed for examination by transmission electron microscopy. Pulmonary Function Before onset of exposures and after 9, 14 and 22^ months of exposure to PVC dust, the following pulmonary function tests were performed on fasted, anesthe tized (pentobarbital anesthetic) exposed and control monkeys: total lung capacity (TLC), vital capacity (VC), inspiratory capacity (IC), residual volume (RV), forced expiratory volume in 0.5 seconds (FeV 0.5), peak expiratory flow ( maximum expiratory flow volume (FEF) at 50%, 252 and 10% of vital capacity, resistance (RL) , compliance (CL), closing volume (CV), N2 washout (A^/lOO- ml) and volume of isoflow (VisoV). These tests were performed with a variable pressure, whole-body plethysmograph. Means and standard deviations for all parameters in each group were calculated at each interval and comparisons made between groups using parametric and non-parametric statistical analyses. -- Results The mean total dust concentration in the chambers for the rodents was 13.0 mg/m^ I (S.D. - 2.34) and for the monkeys 12.9 mg/m^ (S.D. - 4.69). The mean respirable dust concentration was 10.4 mg/m^ for both the rodents and the monkeys. The monkeys were exposed on 464 days for a total of 2818 hours while the rodents were exposed on 245 days for a total of 1426 hours. 5 R&S 114642 The vinyl chloride concentrations in the chambers at the two sampling periods were 0.01 and 0.02 ppm. A small non-quantifled amount of vinyl chloride wasliQli^ detected in the bulk PVC. Electron micrographs of the settled chamber dust showed individual particles ranging from 0.13 to 1.68 pm in diameter and agglomerates measuring up to 12.8 ym in diameter. Although the agglomerates measuring greater than 7 ym in diameter accounted for only 2.3% (5/222) of all the particles counted, they were estimated to account for more than 20% of the mass. Pulmonary Function A summary of the extensive pulmonary function evaluations indicated some signs of loss of lung recoil pressure, probably a result of the animals' aging process. In most cases, differences were noted during the second and third testing periods (exposure months 9 and 14) and were indicative of some small airway obstruction. At this time, however, these differences'Trere not statistically significant (see Table 1). At the last evaluation (month 22) compared with baseline (preexposure) data, there were no significant differences for any parameter tested. Because of the differences in animal sizes between the control and exposed groups for all except the 14 month interval, meaningful comparisons could be made only at that time period. At that time, the mean weight of the controls was 4470 grams and that of the exposed was 4190 grams. The results appear in Table 1. No statistically significant differences existed. Impairment of respiratory function does not appear to be indicated under the conditions of this study from exposure to respirable PVC dust. 6 Pathology Ten exposed and 10 control monkeys were autopsied 22}s months after initiating exposures. No gross abnormalities were seen in the lungs or other organs that could be related to the exposures. Light microscopic examination of the lung sections of exposed animals revealed macrophage aggregates (Figures 1 and 2) in the alveolar walls, alveoli, alveolar ducts, respiratory bronchioles and around veins. The majority of the aggregates were 100-200 urn in diameter and some were as large as 475 um in diameter. Their concentration varied from 3-7 per microscopic field at 100X magnification. The tightly packed, large, spherical macrophages contained colorless cytoplasm which appeared light blue under phase DO contrast microscopy. Only a rare aggregate contained any other type of infil C/> trating cells, which usually consisted of a few polymorphonuclear leukocytes. O) No other lesions that could be related to experimental treatment were present, U CO and no metaplasia, interstitial fibrosis, nodular fibrosis or pneumonitis were 4 present. The tracheobronchial lymph nodes contained aggregates of the same type of macrophages described above in the lungs. Transmission electron micro scopic examination of the macrophage aggregates revealed numerous round particles that do not normally occur in macrophages and were of the same size aiid shape as PVC particles (Figure 3). Analysis of these particles with the microprobe re vealed high concentrations of chlorine (Figures 4 and 5), thus, further estab lishing their identity as PVC particles. A few aggregates of birefringent particles appeared in the lungs and high concentrations in the lymph node of both the exposed and control monkeys. These particles were identified as mica, keolin and quartz by the use of microprobe, electron and x-ray diffraction methods. The bulk PVC did not contain these minerals. No macrophage aggregate^ were present in the control monkey lungs. No treatment related lesions were seen in sections of the other organs examined. 7 R&S 114644 ( closely packed macrophages in the lesions shown by Ar^naud, et al., the morphology of the lesion is similar to that seen in our exposed monkey lungs. Part of the compactness of the cells in their lesion could be explained on the basis of artifactual compression caused during the taking of the lung biopsy. No excess collagen was apparent in the light microscopic pictures of the granulomas they showed in their article, although they did refer to their case as a "case of discrete pulmonary fibrosis." Their electron microscopic pictures of the PVC particles in in vitro cell systems were very similar to what we saw in the macrophages in the monkey lungs. The PVC particles they showed in a macrophage from the lung biopsy appear to be clustered in lysosomes and the individual particles are not as dense as the ones we found in the monkey lungs. That difference might be related to the different exposure durations and differing intervals since last exposure. Their patient was exposed to PVC for 23 years, (our monkeys were exposed for 22*s months) and had no PVC exposure for 6 years (our tissue was sampled within 48 hours .of last exposure). It is possible that some of the ingredients in the PVC particles in the human lung were altered and leached out over a period of time.~""~ Szende, et al. (1970), reported that the lung biopsy specimen from a worker who shoveled PVC for one year showed moderate diffuse fibrosis and contained a granuloma with concentrically arranged connective tissue fibers. We did not see this in the monkey lungs. The apparent difference in response of the lung tissue-from two workers exposed to PVC compared to the monkeys exposed to PVC might be related to a variety of factors, including differences in the PVC dusts. Since various types of 10 R&S 114645 emulsifiers and initiators are used in making PVC, and since these are not completely removed from the powdered PVC, they might explain the differences observed. It would be prudent in the future to obtain more information on PVC manufacturing processes and content of PVC dusts when studying their health effects. Mastrangelo, et al. (1979), published health findings on employees who were currently working in a PVC production factory. Twenty cases of pneumoconiosis were found in a population of 1216 workers. Although the data were not given, the authors stated that slight restrictive respiratory function impairments were associated with chest x-ray changes in a small percentage of cases. In our study, no differences in pulmonary function tests could be detected between exposed and control monkeys after 1A months of exposure to 13 mg PVC/m . Whether any pulmonary function deficits would have occurred upon further exposure cannot be ascertained from this study. The lack of any significant pulmonary function abnormality in these monkeys, however, does not mean that animals or humans exposed to other types of PVC dust or mixtures of PVC dust with vinyl chloride (or a multitude of other ingredients that might be used in fabricating PVC products) might not exhibit pulmonary function changes. For the first time, we have been able to show that a PVC dust generated under controlled experimental conditions is respirable and is deposited in cell aggregates in the lungs of monkeys, that PVC particles within macrophages are fairly unique in appearance, and that their identity can be confirmed with microprobe analyses. No pulmonary function deficits were found after 1A months of exposure to PVC concentrations of 13 mg/m , thus, confirming the pathologi diagnosis of a benign pneumoconiosis. 11 R&S 114646 ACKNOWLEDGMENTS The authors wish to acknowledge the following people who assisted in this project: George Madden, Brandon Barton, John Clark, David Brewer, Margrit Stoll, Myra Springs, Hazel Patterson, Lea Kalejs, Ardith Groce, John Holtz, Charles Gorski, Richard Niemeier, Richard Hornung, and Patricia Combs. 12 REFERENCES 1. Aranaud, A., Pommier de Santi, P., Garbe, L., Payan, H. and Charpin, J. Polyvinyl chloride pneumoconiosis. Thorax 33^:19-25, 1978. 2. Frongia, N., Spinazzola, A. and Bucarelli, A. Experimental.lung damage from prolonged inhalation of PVC dust in a work environment. La Medicine del Lavoro 65(9/10):321-342, 1974. 3. Lilis, R., Anderson, H., Nicholson, W.J., Daum, S., Fischbein, A.S. and Selikoff, I.J. Prevalence of disease among vinyl chloride and polyvinyl chloride workers. Ann. N.Y. Acad. Sci. 246:22-41. 1975. 4. Mastrangelo, G., Manno, M., Marcer, G., et al. Polyvinyl chloride pneumoconiosis: epidemiological study of exposed workers. J. Occup. Med. 21(8):540-542, 1979. 5. Popow, J. Influence of polyvinyl chloride (PVC) dust on the respiratory system in the rat. Roczniki Akademii Med. im. Juliana Marchlewskiego w Bialymstoku 24:5-48, 1969. 6. Szende, B., Lapis, K., Nernes, A. and Pinter, A. Pneumoconiosis caused by the inhalation of polyvinylchloride dust. La Medlclna del Lavoro 61:433-436, 1970. 13 R&S 114647 R&S 114648 V 7. Tribukh, S.L., et al. Work conditions and industrial hygiene measures in the manufacture and utilization of vinyl chloride plastics, Hygiena and Sanitarya (USSR) 10^38-44, 1949, 8. Vertkin, Yu. X. and Mamontov, Yu. R. On the state of the bronchopulmonary system in workers engaged in the manufacture of articles made of polyvinyl chloride. Gigiyena Truda 14(10):29-32, 1970. 14 R&S 114649 FIGURES Figure 1 - Macrophage aggregates in lung of monkey exposed to PVC dust for 22 months. Aggregates are primarily In respiratory bronchioles and alveolar ducts. (Original magnification on 35 mm film was 40x.) Figure 2 - Macrophage aggregates in lung of monkey exposed to PVC dust for 22 months. (Original magnification on 35 mm film was 250x.) Figure 3 - Transmission electron micrograph of a macrophage from the lung of a monkey exposed to PVC dust for 22 months. Note the numerous round and oval, electron dense particles comprising most of the cytoplasmic space. These are the inhaled PVC particles. Figure 4 - Scanning electron micrograph of a 5 um section of lung of monkey exposed to PVC dust for 22 months. This photo shows the numerous PVC particles which obscure the cellular outlines of the macrophages in which they are residing. Figure 5 - Chlorine x-ray map of the same area and magnification depicted in Figure 4. The white dots indicate the presence of the element chlorine. Note that the dots are clustered in the same area as the PVC particles, thus confirming the identification of the PVC particles. Figure 6 - Macrophage aggregates in the lung of a rat exposed to FVC dust for 12 months. Note vacuolated appearance of cells. (Original magnification on 35 mm film was 250x.) R&S 114650 TABLE 1. Results of Pulmonary Function Tests in Monkeys _____________ After 14- Months Exposure to PVC Dust* Pulmonary Function Test RL, cmH^O/L/sec CL, ml/cmH-0 TLC, ml VC, ml IC, ml RV, ml RV/TLC, % FeV 0.5/FVC, 7. PF, ml/sec FEF 50%, ml/s FEF 25%, ml/s FEF 10%, ml/s CV, ml AN2/100 ml Viso V, ml Groups____________ Control Exposed 12.1+4.0 22.9+10 357+39 334+36 185+19 40+10 11,3+2 ' 87.7+5.3 943+91 920+91 648+173 262+114 21+9 1.05+0.3 28.7+19 13.8+3.9 17.5+7.5 360+75 339+73 210+58 36+11 9.9+2 86.5+8.1 960+139 905+138 655+235 269+134 21+11 0.67*3.4 19.9+14.3 *Data are presented as means + cne standard deviation for each of the parameters. R&S -u CJJ cn i _' A iNalcoCfiemu cide^ and Nemacur, a pesticide,. for.Chemagro klbrp^a -division of Mobay Chemical: Corp., and falsely reported the1 studies ran IS months, whenVthey really /rah - only: 14 months. .The indictment also , said they un derreported the number of animals tested. - ^ A spokesman tor Mobay, a Pittsburgh-; -- . ^ - i *^yVT, based liiiit of Bayer'AG of West.Germany,' ti SptdalUi Tuk Wuj. Strkkt Journal- . . said.the concern has authorization from the vCHICAGO-Afederal grand juryindicted U.S. Environmental Protection Agency to four tormerV:officials of .Nalco Chemical sell Sencor and Nemacur, and is doing so. -= Co.'S todusfrialBio-Test Laboratories Inc; The former Nalco unit officials also were unit on cbajjas^f felsifying testTesults on charged with-falsifying results of tests-on- chemicals'and drugs between 1969 and!976. Trichlorocarbanilide, <TCC) an antibacterial . An eight-count, indictment alleged .that agent-used in deodorant soaps, and on Na defendants ialtered testresultsofstud-; proxen, an anti-inflammatory arthritis drug. conducted on mice and rats.iThe .studies Government prosecutors said the scientists were.lorthree -companies seeking federal: concealed evidencethatTCC produced atro-; pKiedjtesticles to .mice; that swere fed;the, substance. The.accused officials .also are charged with falsifying data <cm blood, and: ihine. tests oniNaproxen,^accordingto the. indiclment:-,'->~^^ Hie -Food and Drug -Administration said. Naproxen--and TCC have;_beeittested and' foundAafe, and-both, drugs are -currently on ' the marketV-- -v'- r. jApokesmanforNalco said IBTLabora-- tones .-is ""still .in"existence,^ working with the "government: and- some ^oompanies on- past studies. .But,he -said the wholly owned' .......... subsidiary is -'^winding-down" to.prepare to: three counts of mailfraud, one count of-wire* cease operations.;.,. yes'tvVfyit&t--? baud and four counts -of submitting false! / Nalco has been named in four lawsuits in documents to the government The mail and connection with, the IBT Laboratories stud wire fraud chargestcany jnaixmum penai^ ies, accordingto the^company's 1980 annual ties of five years', imprisonment and iSLfXir report. Mobay istoitag. Nalco for breach of fine;-the iaisedocuments charges; canyiba contract And misrepresentation. The-com maximum penalty<of five -years in jail .and pany deniesliability for.Its unit's conduct," $10,000 In the report ; says. It. also -says Nalco made- The- indictment,-charged .that the'ctour provision-to Its 1978financial statements fori ~men--c-o--n-d--u-c-t-e--d-.-s-t-udies on S--e--n-cor, a~.h--er.bi dis:-c--o-n--tin'uinwg --IBT ^Laibao/jriawtmoriiicebs uoMpecriatuiovnius.. t 9P (n Brief Report M? ft?f R&S 114653 Hepatic Angiosarcoma' Possible Relationship to Long-term Oral Contraceptive Ingestion Paul S! Monroe, MO; Robert H. Riddell, MD; Mark Siegler, MD; Alfred L.Bsker, MD " A 32-year-old woman with an eight-year history of oral contraceptive use' was found at laparotomy to have a hepatic angiosarcoma. The prolonged exposure to contraceptive hormones may have been a factor in the development of this rare tumor. The estrogens and progestins contained in ral contraceptives are biochemically similar to the anabolic-androgenic steroids that have recently been reported to cause hepatic angiosarcoma. : IJAMA 19815246:64-65)::^ i. , ,_ i. . THE USE* of oral contraceptives'has ... defects in the right lobe and a large cold been associated with the development area in the region of the porta hepatis, of a variety of hepatic tumors, includ ing hepatic adenoma, focal nodular hyperplasia, and hepatocellular carci noma.' We describe a patient in whom' hepatic angiosarcoma developed after taking oral contraceptives for eight years.. The increasedincidence _jof with Blightly increased - activity in this area during the arterial phase of the flow -study-Hepatic ultrasound showed inhomogeneity of the posterior right lobe but no definite mass. Celiac angiography dis closed a large avascular lesion of the right lobe of the liver with a small central area ~ s-that filled during the venous phase, sug- hepatic angiosarcoma among users of - gesting a-hemangioma or adenoma. The androgenic-anabolic steroids3 and .the '--patient was gravida 2, para 2,. and had resemblance of vascular and hepatic used an oral contraceptive for eight years. parenchymal abnormalities in users ..She had not used oral contraceptives after of oral contraceptives to`the precnr- ' ftabdl*Kation three years before admis- . , , .. - ---sion. There was no history of exposure to sor lesions of hepatic angiosarcoma thoro- . suggest that this may be more thaiua^ ^ 0T^ic .^.-chance association.^^,laparotomy, many firm, white nod- " ules involved both lobes of the liver. -i^`. -- - 11 ----- tBtiiroinpbsvy csTpwemciTmTeittnnsx wwaevre* tt.aken, and a dlag- ' A 32-year-old woman was admitted in ' mosis of hemangioendothelial sarcoma (he May 1979 with a history -of several epi patic angiosarcoma) was made (Fig 1 and sodes of severe right-sided abdominal pain 2). While some areas of liver contained occurring during two years. On admission - normal endothelial cells, other sections she appeared welL An abdominal exami showed atypical Kupffer cells that were nation demonstrated an exquisitely tender enlarged, hyperchromatic, and pleomor liver but no bruits or rubs. Laboratory phic (Fig 3). tests disclosed the following values: hema- In the year after surgery, the pain had tocrit, 41%; WBCs, 7,300/cu mm; differen- not'recurred and the patient has felt tial cell count, normal; total bilirubin, 0.6 completely well. A liver Bean showed no - mg/dL; alkaline phosphatase, 53 IU (nor progression of the disease. Because of the mal,.10 to'85 IU);`aspartate aminotrans- atypical course for an hepatic angiosarco ferase, 18 IU; alanine aminotransferase, 8 ma, the histology slides and electron IU; albumin, 4.0 g/dL; and globulin, 3.2 micrographs were reviewed by specialists - g/dL. The prothrombin time was not pro at other institutions. All concurred in the longed. The caranoembryonic antigen and .diagnosis of angiosarcoma. In addition, a-fetoprotein values were -within normal immunoperoxidase staining of the tissue limits. An oral cholecystogram was nor using a factor VIII antiserum was per mal, but a technetium Tc 99m sulfur formed. This was strongly positive, con colloid liver .scan .showed multiple filling firming the endothelial origin of the From the Liver Study Unit (Dr> Monroe and Baker) and the Section of General Internal Medi cine (Dr Siegler), Department of Medicine and the Department of Pathology (Dr Riddell), University of Chicago Hospitals and Clinics. Raprint requaata to Department of Medicine, Univeraity of Chicago, Box 400. 950 E 60th St, Chicago.-(L 60837 (Dr Baker). tumor. Comment More than 300 cases of liver adeno mas and focal nodular hyperplasia have been reported in women taking oral contraceptives.1 Regression of Fig 1.--Tumor nodule showing spectrum of endothelial cell changes ranging from spin- dle-ahaped to ovoid, containing central nucleoli with incomplete vasoformation (he- matoxyiin-eoain. original magnification X300). - ' -"u . Fig 2.--More cellular part of second nodule in which tumor cells form capillary channels (arrow) containing. RBCa. Electron micros copy confirmed that these were endothelial calls (hematoxylin-eosin, original magnifi cation X300). .. __ Fig 3.--Biopay specimen from groaaly nor- I mal liver, demonstrating sinusoids lined with dyplastic Kupffer calls (arrow) (hamstoxylin-eosin, original magnification X500). 64 JAMA, July 3, 1981--Vol'246, No. 1 Angiosarcoma--Monro# etsl Hi *; V - ' . liver adenomas after cessation of hor monal therapy and recurrence of these tumors after surgical resection in patients who continue to use oral contraceptives suggest.^ that these drugs are linked to the pathogenesis of this tumor.1 The precise role of oral contraceptives in production of focal nodular hyperplasia is uncertain.' Oral contraceptives may also be a factor in causing malignant tumors of the liver, particularly since foci of hepatocellular carcinoma have been found within several Oliver Adeno mas.' *- Hepatic angiosarcoma'developed in the patient, after using oral contra ceptives for a total of eightVears. In the year after diagnosis, she has had no progression of disease, in contrast to the short survival for most patients with hepatic angiosarcoma. The slow progression resembles the apparently indolent course reported for hepatocellular carcinoma associ ated with oral contraceptivelise,5 sug gesting that the drug may have been a factor in causing and promoting growth of the tumor. ^ " Additional factors support a possi ble etiologic relationship between oral contraceptives and hepatic an giosarcoma. First, hepatic angiosar coma developed in a patient .after using diethylstilbesterol for 13 years for the treatment of prostatic carci noma.* Second, users of oral con traceptive pills have hepatic paren chymal and vascular changes that resemble precursor lesions of hepatic angiosarcoma, including hyperplasia of hepatocytes and sinusoidal cells, sinusoidal dilation, and peliosis hepatis.li7 These changes are similar to those produced in humans by vinyl chloride, arsenic, and thorotrast,1? previously shown to cause hepatic angiosarcoma. Also, liver adenomas from women who use oral contracep tives demonstrate more extensive pe liosis hepatis and arterial, intimal, and medial hypertrophy with occlu sive phlebitis and thrombi than do tumors from nonusers of these medi cations.1 - Third, users of androgenic-anabol ic steroids, C-17-nr-alkyl-substituted steroids that are closely related bio chemically to the synthetic estrogens and progestins used in oral contracep tives, have been associated with the production of vascular and parenchy mal changes resembling those seen in patients with oral contraceptive pill exposure, as well as probable precur sor lesions of hepatic angiosarcoma and liver adenomas and hepatocellu lar carcinoma. In 168 cases of hepatic angiosarcomas, four patients were found to have had long-term andro genic-anabolic steroid exposure.1 Fur ther systematic study of patients with and without oral contraceptive pill exposure will be necessary to establish these drugs as an unequivo cal cause of this rare tumor. Hans Popper, MD, PhD, Peter Scheuer, MD, and Juan Rosai, MD, reviewed the histology slides and electron micrographs. Dr Rosai also performed immunoperoxidase staining of the .tumor tissue. References 1. Klatskin G. Hepatic tumors: Possible rela tionship to use of oral contraceptives. Gastroen terology 1977;73:386-394. 2. Falk H, Thomas LB, Popper H, et ah Hepatic angiosarcoma associated with andro genic-anabolic steroids. Lancet 1979;2:1120-1123. 3. Ross D, Pina J, Mirxa M, et al: Regression of focal nodular hyperplasia after discontinuation of oral contraceptives. Ann Intern Med 1976: 85:203-204. 4. Knowles DM; Casarella WJ, Johnson PM, et al: The clinical, radiologic, and pathologic char acterization of benign hepatic neoplasm. Medi cine 1978;57:223-237. 5. Neuberger J, Portmann B, Hunnerley HB, et al: Oral-contraceptive-assodated liver tumors: Occurrence of malignancy and difficulties in diagnosis. Lancet 1980;1:273-276. 6. Hoch-Ligeti C: Angiosarcoma of the liver associated with diethylstilbesterol. JAMA 1978; 240:1510-1511. 7. Cole FM, Sweeney GD: Change in rat hepatocyte plasma membranes caused by syn thetic estrogens. Lob Invest 1980;42:225-232. 8. Popper H, Thomas LB, Telles NC, et ah Development of hepatic angiosarcoma in man induced by vinyl chloride, thorotrast, and arsen ic. Am J Pathol 1978:92:349-369. ..... R&S 114654 Computed Tomography for the Evaluation of Thoracic Masses in Children Susan B. Shurin, MD; John R. Haagja, MD; Robert E. Wood, MD; Frank P, Ittleman. MD * Large chest masses in two young children were evaluated using thoracic computed tomography (CT). Air was identified within the mass four days before it was visible on the plain roentgenograms in one patient. Rapid injection of contrast material with CT scanning demonstrated that the pulmonary blood flow passed through the mass, instead of being displaced by it, in the other child. Thus, both lesions were identified as originating in the pulmonary parenchyma rather than in the mediastinum. (JAMA 1981;246:65-67) From th* Department of Pediatrics. Rainbow Babies & Childrens Hospital (Drs Shurin and Wood); the Departments of Radiology (Dr Haags) and Surgery (Dr Ittleman). University Hoapitsls Ot Cleveland; and Case Western Reserve University School ot Medicine (Dr* Shurin, Haags, and Wood). Cleveland. Dr Ittleman is now with the Department of Surgery, University ot Vermont Medical Center. Burlington. Reprint requests to Division ot Hematology, Rainbow Babies & Childrens Hospital. 2101 Adal bert Rd, Cleveland, OH 4410 (Dr Shurin). JAMA, July 3, 1981--Vol 246, No. 1 PULMONARY consolidation may ap pear radiologically as a solid mass. Without invasive procedures, it may be difficult to determine whether such a lesion originates in the medi astinum or in the pulmonary paren chyma. Distal infection may accom pany bronchial obstruction, so the presence of fever, cough, or leukocyto sis does not exclude'a neoplasm or congenital anomaly as a component of a chest mass. This is especially true in children, in whom pulmonary or mediastinal masses may reach mas- Thoracic Masses--Shurin et al 65 sPer> e 0uArcJ T&f&z/son r* *7 ^L1 ' Journal of Analytical roMcoiogy Vol 4. Jurlly/August 1980 Phosgene in the Thermal Decomposition Products of Poly (Vinyl Chloride): Generation, Detection and Measurement James E. Brown and Merritt M. Birky 30 Center for Fire Research, United States Department of Commerce, National Bureau of Standards. Washington. D.'C. 20234 (/> , 114655 Abstract analytical study was made to determine whether carbonyl chloride (phosgene) is formed during the thermal decomposition of polylvinyi chloride), PVC. Four methods of decomposition were studied: (1) ther mal degradation of PVC in a resistively heated furnace, (2) electrical overloading of a PVC clad wire, (3) electri cal arcing between electrodes partially covered with PVC, end (4) electric arc initiated flaming combustion in a cup furnace. Results are reported which show that significant quantities of phosgene can be generated from PVC by the electric arc method. Lesser amounts were found in the other scenarios. The measurements, identification and quantification of phosgene in the decomposition product, were obtained through the use of gas chromatography, inlrared spectroscopy and mass spectroscopy. While the study was not mecha nistic in nature, phosgene is postulated to result from secondary reactions of the PVC products. Reaction mechanisms are suggested.^ Introduction Vinyl chloride polymers arc frequently used in many applieaiions, perhaps lhe most familiar application is lot electrical insulation. These polymers are also widely used in construc tion, in appliances, in furnishings and in transportation vehicles. Total production Tor 1978 is estimated to be 2.3 million metric tons (I). Buildings, including home furnishings and appliances, provide the major market for the production. The role of poly(vinyl chloride). PVC. in leading to human file lai.ilnies has been the subject of a number of investiga tions (2.3.4). In addition, the thermal degradation ol I'VC has been studied extensively trom an analytical point ol view (5,(1) and various reviews of such studies have been prepared (7,8,9). Toxicological studies involving the exposure of animals to the thermal degradation products of PVC have also been reported (2,3). I 'ii;"i-<wrfcil |w 'hese various studies is whether or not catlmiivl vhlotidc* is a product of thermal degradation of PV( . A niviiului'kal study hy Barrow cl al (2) suggests that the tnsirit v ol l*V( decomposition products is due pimcipally to * I r I. wtdi little , ouii ihnuou from othci components, Ihe w.ork reported by VYoollev (ft) and 1 suchiya cl al (') seems to lend suppott to Batiow's coiklusion llowevct. Woolley's negative results regttuling phosgene ate limned In an instrumental detectability ol 50 ppm. a level that would be quite significant toxieologiv allv (10). On (he other hand, lialtaiis el al (II) tepoiicd the presence of phosgene in the combustion pioduets ol two PVC composite materials which wetc hut tied in a closed container. These workers topoticil that hunting <IA g ol I'VC (mm con veyor hells and cables gave respectively 0 (KH9 ami 0 0135 per cent of COCh in a 5 liiei cltamhci. 1 ailicr Coleman and Thomas (12) reported ( ( K'U eimccnn.tnoiis up to 3 ppm and to 1 ppm respect tv civ lot uitsiahili/ctl PVC and stabilized PVC composites binned al 900'C in a dosed env it oilmen!. In fatal fires involv mg PVC, vaiious claims have been made which attribute at least pan ol the toxicological hazard to phosgene (13,14). I he bases lot die claims, howevet, are*not well documented in those teports. Ncvcnhdcss, Or- reports do raise a fundamental question that the nnaly Inal measurements referred to above do not clearly tesolve, either av a conseqnenee of ittsti iimcinul Imuiniioii ot the limitation imposed by the type of decomposition scettat to chosen. Based on these uiuciiaiiihcs and the toxicological signifi cance of phosgene. a studv was cat tied out m whidt analyses lor COCli vvetc pet lot mod on decomposition products obtained by lout methods, each ol which tepiescnis a potential real situation The lour methods ate: (I) diicci thcimal degradation of PVC in a resistive' liimuee: (2) elcelneal overload of a copper wue wrapped with P\ C insulation: (3) an deeiiieal discharge between wires coveted hy P\ C insiihiiion; and (4) eleetrie arc inil trued flaming combustion in a cup fin mice. The first method is equivalent to the analytical type of studies abends iclcienced in the liicinmrc. The most signifkant dilleiencc between this study and other work is that more sensitive unalvtkal techniques were used in this study. Mote importantly, the second anti third methods were not previously lepiesented and wcie chosen based on expected modes of dcgiadation involving eleetrieal overload and arcing ctmtlnion.s (e g. lefctenee (7)). Carbonyl chlnruJf1 whoso ImimuiM is COC'^ has three Known synonyms: phosgene, chlnrpfor niy I ( himtdo and r*u Ihacid riirhioride In Ihe lollow- rng t<>*1 t ho synonym phns<|i.....ami ihi* turunda aio ustnl inintchangeabiy 10 donnie Ibis 4.(.*ni|KMind 166 11"pr pf|iH lnn* (photocopying H,f t*t |ih u t.il i nnt#M>t of this |iujl U,d IS | if nhihitu. 1 <*> itl'tml |mild'si n'r s ptM nttSMOU R&S 114656 Journal of Analytical Toxicology, Vol. 4, July/August 1980 Experimental Reference Gas Reference gas mixtures containing about 60, 80, 110 and 600 ppm of COCI2 were prepared in metal cylinders from the vapor of liquid COCI2 and pressurized with nitrogen to 500 psi. The exact concentration of the individual reference gases is not known. However, the method of preparation, purity of the reagent gases and the expected stability lead to an estima ted accuracy of about 5*170 (15). PVC Composite Samples The PVC composite materials used in this study were formulations commonly employed as electrical wire insula tion. Typical formulations included about 457o PVC resin, 32''o plasticizer, with the remaining fraction composed of fillers, pigments and stabilizers. The PVC resin used in this study was a commercial material whose reported values of number-averaged and weight-averaged molecular weights are 22,400and 57,700, respectively. HV IGNITION ARC }tt fonn supra Figure 1. Pyrolysis/Combustion Apparatus. Pyrolysis/Combustion Chambers A diagram of the thermal degradation system is shown in Figure 1. The system consists of a 3 liter resin reaction flask, which could be opened and closed by a ground glass top which was fined with standard tapered joints. For the first group of experiments, the PVC was thermally degraded in a resistive!) heated furnace. A quartz boat, used as the sample holder, was positioned within the furnace and located near the center of the chamber. A rubber septum was fitted to one of the standard tapered joints through which gas samples were taken from the chamber by a gas-tight hypodermic syringe. To accommodate the pressure increase in the chamber during thermal decomposition, a rubber balloon was attached to another joint in the flask cover as show n. For the electrical overload heating of PVC samples, the Ifurnace resistive elements were replaced with a PVC insulated coppct wire. A power supply was used to heat the wire by a current overload. Dual strands of number 12 copper wire, about 10 cm in length, connected the power supply to a 10 cm coil ol number 12 copper wire. The coil was coated with PVC insulation. Figure 2. Diagram of Flash-Fire Call used for Electric Arc Decom position of PVC. A cell was used to degrade PVC samples in the presence of an electric arc. The cell, shown in Figure 2, was designed to study the Hash-fire potential of materials. This system was constructed of a Pyrex cylinder 50 cm in length and 5,cm in diameter having a volume of about one liter. The ends of the cylinder were fitted with spring-loaded covers sealed by O-rings. Parallel platinum electrodes about one centimeter in length were located near one end of the cell to which a 10 kv AC power supply was connected for the generation of an electric arc for the decomposition of PVC. The cell also con tained sampling ports, each with a rubber septum, which xvere located near the ends and middle of the cell. Analytical Instrumentation In this study, the analytical instrumentation consisted of gas chromatography (GC), infrared spectroscopy (1R), and a gas chromatography-mass spectroscopy-minicomputer (GC - MS MC) system. The primary GC measurements of phosgene were accomplished with an instrument equipped with a linearized electron capture detector (ECD). The GC was fitted with a glass column about 2 meters in length and 2 mm i.d. which was packed with an acid treated silica gel (Chromosil 310). lisposito el al (16) had previously shown that this packing proxides good recoveries and resolution in GC determinations of COCK in air. The column was maintained at 70*C. The cat tier gas was 5"<i methane in argon at 30 ml 'minute. A second column packed with 30ro diisodecyl phthalate on a solid support (Chromosorb \Y) was used at about fO'C for continuation of the measurements using the Chromosil 310 packed column. The chromatograph was interfaced with a 167 Journal of Analytical Toxicology. Vol. 4, July/August 1980 reference gas coci? PVC DECOMPOSITION PRODUCTS t I w? u DMt. MINUTES Figure 3. Chromatograms Obtained from Silica Gf Packed Column showing COCI2 Peak in Reference Gas and PVC Decomposition Product. sample. I or the 1R measurements the decomposition products were expanded or introduced into an evacuated one-meter folded-path length gas cell which was equipped with KBr win dows and gold front surface mirrors. The decomposition of the PVC by an electrical overload was accomplished by applying power to the PVC clad coil in the sample chamber for a period of about 3 to 5 minutes, or until decomposition was complete. Sampling of the products and analysis of these products were accomplished by the tech nique described for the rcsistively heated furnace experiments. For the electric arc- study, a piece of PVC insulation ap proximately '/; cm long was placed on an electrode in the "flash-fire" cell. A 10 kv ac potential w^s applied between the electrodes until the sample was degraded. The gaseous degra dation products were analyzed as described previously. The GC method of analysis for COCJ2 in PVC degradation products was also applied to gases taken from a 200 1 animal-exposure chamber in which a PVC composite (elec trical insulation) was degraded by flaming combustion. The details of the animal exposure study are the subject of another study 10 be reported at a later date. recording integrator (printer/ploiter). In the GC/MS/MC system, the quadrupole spectrometer W'as equipped w'ith an electron impact ionization source which, along with the MS electrometer, served as the system's GC detector. This GC was interfaced to the MS by a membrane separator and w-as fitted with the silica gel packed column maintained at 70C. Helium was used as the carrier gas at a flow rate of 30 mL/min. The infrared spectrophotometer employed in obtaining spectra of PVC decomposition products and phosgene was a high resolution grating instru ment having a wide range of capabilities for scan speed and scale expansion. Typically, the scan speed and slit program or resolution were set at values deemed appropriate for quan titative determinations. The spectra were recorded with the ordinate scale expansion set at 10.x in the linear absorbance mode. Methods of PVC Decomposition and Sampling for COCI2 Approximately 0.6 g of PVC were placed in the quartz boat and ignited with the resistance heater. To achieve ignition, the go furnace was heated to a bright red color or until ignition of the CO PVC occurred, which was approximately 1 to 2 minutes after .j, applying the voltage to the heater. After ignition, the power to the heater was terminated and the sample was allowed to burn until it self-extinguished or the sample was consumed. The deCO gree of combustion depended on the formulation but was consistent within a given formulation. Alter extinguishment. 0.1 to 0.2 mL of the gaseous products were drawn from the chamber with a gas tight syringe and in jected into the gas chromatograph for analysis. From the re sulting chromatogram, the COCI2 peak areas and retention times were calculated by the integrator interfaced with the chromatograph. The concentration of the COCI2 in the de composition products was then computed from calibration data obu >ncd from the LCD response to known quantities of COCI2 reference gases. In some ca..ex tr.lrr.reJ spectra w^ere made oT the decom position products loliowing extinguishment of the PVC 168 coct2 REFERENCE GAS T PVC PRODUCTS i$]\----- TIME MINUTES Figura 4. Chromatograms Obtained from Phthalata Ester/Soltd Support Column showing COCIj Peak in Relarenca Gas and PVC Dacompoaition Product. Results and Discussions Identification of Phosgene Identification of phosgene in PVC decomposition products was accomplished primarily by employing gas chromatog raphy. Additionally confirmation evidence of this product was obtained through infrared spectroscopy and mass spectro scopy. Figure 3 represents a typical chromatogram of the phosgene reference gas, and a chromatogram of a product gas mixture generated from PVC in air by the electric arc method. It is apparent that the retention time of a peak near 1.8 minutes in the PVC decomposition products corresponds to the retention time of COCls in the reference gas chromatogram. To confirm that the peak ar 1.8 minutes is COCI2, a second column material with different properties was used. Com ponent characterization by retention limes in two columns of different polarities is generally considered positive identi- Journal of Analytical Toxicology. Vol. 4, July/August 1980 cation of an unknown compound (e.g. reference 17). For this confirmation, the GC was equipped with the diisodecyl phthalate/Chromosorb W packed column. Figure 4 shows chromatograms obtained using the column packed with the phthalate ester/solid support for the reference COCIj gas and the products of another PVC decomposition. As noted in the figure, the peak with the retention time of 0.77 minutes in the decomposition products is essentially the same as the retention lime ol COCK in the reference gas. For the mass spectral identification of phosgene in the PVC products, mass spectra were obtained of components eluting from the GC equipped with the silica gel packed column at the retention time of COC^. Figure 5 shows a total ion current chromatogram which represents the total ions generated by the mass spectrometer in the range m/e 33 to 150 as the COCIj reference eluted from the chromatograph. The peak at 2 minutes is COCIj and the peak at 0.4 minutes was found to be 1,1 dichloroethane, an impurity, whose parent ion m/e 98 and base peak m/e 63 are the same as COCK. In Figure 6 chro matograms of the total ion current and a single ion, m/e 63, are shown for PVC decomposition products generated by an electric arc. The peaks at 2 minutes correspond to the retention time of the COCI2 reference, li was also feasible to monitor all of the major m/e ions of COClj by the use of the computer interfaced with the GC/MS. The major ion Figure 5. Mass Spectral Total Ion Current (or COCIj Reference Eluting from GC Column. R&S 114658 TIME, MINUTES Figure 6. Mats Spectral Total Ion Current and Fragment m/e 63 Ion Current of Component ol PVC Product Eluting from GC at Retention TlmaolCOCIj. Flgur* 7. Major ton* ot COCI2 Monltorad In Chromatograma ol COCIj and a PVC Dacompoiltlon Product. 169 30 fio CO 1.2 4* O) cn co 0.9 x Uoi GQ 0.6 CO GO CELL BACKGROUND 0.3 Journal of Analytical Toxicology. Vol. 4, July/August 1980 C0C12 R EFERENCE PVC PRODUCTS i !1111I 950 850 750 III________ III 950 850 750 Figure 6. Infrared Absorbance Band of COCIj near 84S cm'1. fragments of COCH are m/e 35 (Cl"1"), 47 (CCI+), 63 (COCI+), and 98 (COCIj"*")- This result is demonstrated in Figure 7 for COCH reference gas and the component in the PVC products which has the retention time of COCh. Here the COCIj retention time (about 1,7 minutes) is slightly less than the previous case (about 2 minutes) as a result of a small increase in the helium flow rate. This body of evidence corroborates the chromatographic identification of phosgene as a component in the decomposition products of PVC. I he infrared evidence ot COCH as a decomposition product of PVC is shown in Figure 8 by ait absorbance band of COCIs and the same band in the decomposition products spectra. This band arises from the Cl-C-Ct stretching vibration (< 4) of phosgene and is centered near 845 cm'*. Gcncrnllv the samples were scanned from about lJ50 cm'* to 750 cm"' which was suflicient to obtain the absorbance due to phosgene in this region. The presence of phosgene in the decomposition products of PVC has been shown from different perspectives hv three independent analytical techniques. The GC results alone provide convincing evidence for the presence of phosgene, ilowc'ct. the mas-, spinal and infrared data unequivocally corroborate this conclusion. 170 Quantitative Measurements of COCI2 Figure 9 is given in order to show the response of the ECD equipped gas chromatograph to various quantities of COCH reference gas. Here the resulting peak areas are plotted against the calculated volume of COCH in Nj which was injected into the chromatograph from a 100 uL syringe. These data were fitted to a line bv linear least squares. The result of this fit gave the ECD response as 1.53 n I0*6 nanoliters (nL) per area unit. Flic con elution coefficient of 1.05 wav obtained from the fit of these data. The response curve demonstrates that the ECD response is essentially linear at least to 12 nL of COCI2. Although the response of the ECD to COCI2 may be viewed as a calibration, in practice the sensitivity of the ECD was determined for each decomposition experiment. . The results shown in Table 1 are the results obtained from the combustion/pyrolysis of various PVC samples in the 3 1 chamber furnace arrangement. The concentrations of COCI2 shown in column 2 are derived from GC analyses of 0.1 to 0.2 ml gas samples of the decomposition products taken from the chamber at various times after the initial sample heating. In column 3, the weight ratios (mg/g) of COCI2 to PVC were calculated as a means of normalizing the various quantities and PVC formulations to the estimated total amount of COCI2 formed in the decomposition process. The weight Journal of AnalyticalToxicclogy, Vol. 4, July/August 1980 analyses of PVC combustion products from a 200 1 animalexposure chamber. Column 3 refers to the times that evac uated sampling bulbs svere equilibrated to the pressure in the animal-exposure chamber. The ratios of COCI2/PVC are similar to those reported in Table I. Here the PVC combustion was accomplished using a "Potts furnace" (18) heated to SOO^C and equipped with an electric arc ignition source to insure flaming combustion. Details of this experi ment arc the subject of another study which is to be reported. R&S 114660 Figure 9. Response ot ECD to COCIj. Table II. Chromatographic Estimation of COCI? Levels of 200 Liter Animal-Exposure Chamber After Flaming Combustion of PVC Composite C. Experiment No. (COCI2) ppm 1. 4.0g Sample II. 7.0g Sample 0.55 0.50 0.55 1.28 1.26 1.26 COCI2/PVC mg/g Remarks 0.11 0.10 0.11 0.15 0.14 0.14 Bulb filled at 3 mm. Bulb filled at 7 mm. Bulb filled at 13 min Bulb filled at 19 mm Bulb filled at 23 mm Bulb filled at 30 min ratios of COCIj to PVC loading is seen to be of the same order of magnitude for the two composite samples. When the actual amount (about 45/b) of PVC resin in the composite is considered, the COCN/PVC ratio is near that calculated for the homopolymer. PVC resin. Table II giscs the replicate results obtained from GC Table I. Chromatographic Estimation of COCI2 Levels in Products of PVC Thermally Decomposed in a 3 Liter Chamber. Sample/ Weight | COCI2], COCI2/PVC. ppm (v/v) mg/g Remarks Composite A: Run 1 (0.78g) Run II (0.67g) Run III (0.59g) Composite B. Run 1 (0.63g) Run II (0.72g) Run III (0.63g) PVC Resin: Run 1 (0.26g) Run II (0.38g) 6 7 5 - 0.5 3 0.09 0.10 0.09 - 0 01 0.06 Sampled at-3 min. Sampled at 21 mm Sampled at 3 min. Sampled at 4 min. Sampled at 17 min 6 2 - 0.6 6 7 2 6 2 0.12 0.04 - 0.01 0.10 0.12 0.03 0.12 0.04 Sampled at 4 min Sampled at 12 min Sampled at 26 min Sampled at 3 mm. Sampled at 7 min. Reheated 3 min.. sampled at 30 min. Sampled at 4 min. Sampled at 12 mm - 0.3 2 3 - 0.2 2 - 0.01 0.09 0.14 < 0.01 0.06 Sampled at 4 min. Reheated 6min,, sampled at 30 min Sampled at 10 min Sampled at 5 min. Sampled at 10 mm The results shown in Table III were obtained on products evolved from PVC wire insulation which was pyrolyzed solely by the heat produced by the electrical overload on the wire. It is seen that the ratios of COCH/PVC are about the same magnitude as those in the combustion/pyrolysis degradation in this chamber (Table I). A marked increase in the levels of COCH was found in the products when PVC was degraded by an electric arc. The results of these experiments are listed in Table IV. Run I of sample B showed the highest levels of phosgene. The reason for the higher concentration in this particular run is not known at this time. Table III. Gas Chromatographic Estimation of COCI2 Levels in Pyrolyzate of PVC on Electrically Over Loaded Wires in a 3 Liter Chamber. Sample \COCI2\. COCI2 ppm, {v/v) PVC. mg/g Remarks Composite D. Run 1 (0.45g) Run II (0.45g) Composite E: Run 1 (0 53g) <0.1 1.3 3 <0.1 4 6 <0.1 4 2 -- 0 04 0.08 _ 0.11 0.16 Heated 3 mm . sampled at 4 min. Sampled at 60 min Reheated 3 min.. sampled at 66 min Sampled at 2 mm. Reheated 3 min.. sampled at 15 min Sampled at 4 mm. -- 0.09 0.05 Heated 3 min., sampled at 4 mm. Reheated 2 mm.. sampled at 21 mm Sampled at 30 min 171 Journal of Analytical Toxicology, Vol. 4, July/August 1980 R&S 114661 Table IV. Chromatographic Estimation of Level of COCI2 in Products of PVC Decomposed by an Electric Arc in 1 Liter Chamber. Sample/ Weight (COC/2), COCI2/PV-C. ppm(v/v) mg/g Remarks Composite E; Run 1 (O.llg) Run II (0.24g) Composite A; Run 1 (0.21 g) Run II (0.21g) Run III (0.21 g) Composite B: Run 1 (0.l8g) Run II (0.19g) Run III (0.21g) 4 0.15 Sampled at 4 min. 39 1.5 Sampled at 24 min. 29 1.1 Sampled at 84 min. 4 0.07 Sampled at 4 min. 32 0.56 Sampled at 6 min. 23 0.40 Arc on 3 min., Sampled at 24 min. 15 0.3 Sampled at 3 min. 11 0.22 Sampled at 25 min. 17 0.34 Sampled at 70min. (IR Cell) 35 0.70 Sampled at 15 mm. 70 1.6 Sampled at 5 min 56 1.3 Sampled at 15 min. 14 0.28 Sampled at 12 min. 11 0.22 Sampled at 20 min. (1R Cell) Table V. Extinction Coefficients E855cm~^ of COCI2 IR Absorption from GC Determined Concentrations. lCOC/2], ppm Absorbance x left at 855cmE, {ppm-m~1) x 1(fi 10 1.5 1.72 18 2.9 1.61 3.0 1.58 33 4.4 1.33 29 4.2 1.48 32 4.2 1.39 35 5.7 1.64 Avg. = 1.53 Std. Dev. = 0.14 The R branch absorbance (maximum at about 855 cm'*) of the COCI2 band was used for the quantitative measure ments of this compound in the PVC decomposition products. The extinction coefficient for the absorbance at 855 cm'1 was derived from the absorbance of known volumes (50 ml. aliquots) of 110 ppm COCIj in Nj which were injected into the onc-mctcr folded-path length sample cell. Table V sum marizes these results. The trend toward lower coefficients at higher COCI2 concent tat ions suggests that COClj was undergoing decomposition in the 1R cell. Table VI compares the estimations of COCI2 levels found in additional combustion/pyrolysis and electric arc experiments by infrared and chmmaiographic techniques. Generally, there is good agterment between the concentrations determined by GC and IK methods, although in some cases, the IR method gave lower \alue. oi phosgene. It should be noted, however, that the primary purpose of the IR analysis was qualitative 172 Table VI. COCI2 in PVC Pyrolyzates Estimated by IR and GC Techniques. Decomposition Sample/ |COC/2| .ppm (v/v) Method Weight IR analysis GC analysis Electric Arc Furnace Composite A; Run 1 (0.l7g) Run II (0.14g) Composite B: Run 1 (0.20g) Run II (0.18g) Run III (0.17g) Composite A: Run 1 (0.61 g) Composite B; Run 1 (0.59g) Run II (0.63g) 27 10 54 21 13 1 5 1 25 11 91 39 19 3 7 3 confirmation of the GC findings. A significant part of this difference may be attributed to the time required to obtain the respective chromatograms and spectrograms; the GC samples generally were taken from the IR cell before the scans of the 855 cm"1 band were made. The use of infrared spectroscopy for qualitative and quantitative determinations of COCI2 in this study was not the method of choice with respect to ease of sample handling and analysis time required. However, spec troscopy did provide an independent means for confirming the presence of COClj in PVC decomposition products and estimating its concentration. The results of the above measurements show various quan tities of COCI2 are produced during the combustion/pyrolysis and electric arc decomposition of PVC in air when PVC is degraded under the experimental conditions described. The yield of COCI2 from the combustion type experiments was low (0.06 to 0.15 mg/g). The concentration of COCI2 was of the order of 0.5 to 1.3 ppm (Table II) when a sufficient mass of PVC was degraded (4 to 7 g decomposed in 200 1 chamber) to produce significant adverse effects on animals. This concentration of COCI2 is low and by itself is not considered to be the major cause of the observed toxicity. However, as noted in Table IV, the electric arc experiments produced 10 to 20 times more, COCI2 than the combustion/pyrolysis experiments. This concentration (30-50 ppm) has obvious toxicological implications. The toxicological significance of the above analytical data can be estimated from the reported toxicity of phosgene in air (10). Concentrations of the order of 3 to 5 ppm cause effects to the eyes and throat, 25 ppm is dangerous for 30 to 60 minute exposure, and 50 ppm is rapidly fatal even after a short exposure. This information suggests that electric arc experi ments can produce toxicologically significant concentrations of COCI2. However it should be noted that exposure of animals to PVC decomposition products generated by an electric arc have not been carried out to establish the toxicological significance of this finding nor have experiments been performed to establish whether other products of signifi cance may also be produced that are unique to an electric arc. Additionally, the results of the electric arc decomposition of PVC insulation have not been investigated by full scale studies. Journal of Analytical Toxicology, Vol. 4, July/August 1980 R&S 114662 Precursors for COCI2 Formation The reaction mechanisms for the. formation of phosgene during the decomposition of PVC were not explored during this investigation. Stull (19) showed by thermodynamically calculated equilibrium for phosgene, chlorine, and carbon monoxide that the mole fraction of phosgene rapidly decreases as the temperature is raised significantly above 200C. On the other hand, the environment around or near a PVC material undergoing pyrolysis and combustion is seen to consist of a complex mixture and is not expected to approach equilibrium conditions. Nevertheless, phosgene may be formed by secon dary reactions involving the decomposition products of PVC at various proximities to the "decomposition plasma". David and Staley (20) have summarized methods by which phosgene may be prepared. Included is the reaction of metal halides (MCI2) with carbon monoxide. r (I) CuCl2 + CO ^--"~Cu + COCl2 At temperatures of 450C to 750C, copper chloride and carbon dioxide are also reported to yield phosgene in quan tities as great as those with carbon monoxide. Additionally, Gmelins Handbuch (21) shows that chlorine may be generated from HC1 according to the following equation: (2) 4hci+o2 yp'-yqrc 2h2o+2 ci2. Chlorine in the presence of carbon (soot) reacts with carbon monoxide to form phosgene (22). (3) CO+ Cl2* . COCl2 The above sequence of reactions, although unexplored in this investigation, suggests possible routes by which phosgene may be formed as a result of PVC decomposition, especially in electrical fires. It can be postulated that a high intensity electrical discharge could produce vapors of copper and copper chloride in presence of PVC as precursors to carbonyl chloride. Similarily, the question is raised whether antimony chloride, generated from antimony oxide added to PVC as a fire retardant, would also contribute to COCI2 production. It appears that the established processes for the formation of phosgene cannot occur directly by fragmentation or depolymerization of the polymer (-CH2CHCl-)n and subsequent oxidation of the fragment to COCl2, or by concerted reactions of oxidation and depolymerization. This indicates that phosgene is formed by a secondary reaction involving the pyrolyzates and oxygen. Since the greater amount of phosgene is found by the electric arc decomposition, the process of formation through a free radical mechanism or highly activated components, is suggested. The parameters influencing the quantities of phosgene formed may include current density, experimental scale and PVC formulation. Conclusions The use of gas chromatography, infrared and mass spectro scopy, and gas chromatography in concert with the latter two has confirmed that phosgene can be generated by thermal degradation of PVC, especially in electrical fires. The process by which this substance is generated remains an unanswered question. The levels of COCI2 found in the products of PVC generated by an electric discharge show that this substance may be produced in significant quantities in electrical fires involving PVC. This study suggests that additional investiga tions need to be undertaken to determine the factors or parameters that influence the amount of phosgene produced during thermal decomposition of PVC and other chlor ine-containing polymers. Acknowledgments The authors gratefully acknowledge the efforts of Mr. William D. Dorko, research chemist in the Center for Analytical Chemistry, who prepared the phosgene reference gases, and of Mrs. Maya Paabo, research chemist in the Center for Fire Research, who obtained the mass spectra. The authors are also grateful to Dr. Roland E. Florin, research chemist in the Center for Material Science, for valuable discussions and suggestions. References 1. "Materials 79: What's coming up in resins? What's the supply? What's the price situation?". Mod. Plast. 45-69 (Jan. 1979) 2. C.S. Barrow, H. Lucia, and Y.C. Alarie. A comparison of the acute inhalation toxicity of hydrogen chloride versus the thermal decomposition products of poly( vinyl chloride). J. Comb. Tox., 6: 3-12 (1979) 3. C.S. Barrow, Y. Alarie and F.F. Stods. "Sensory irritation evoked by the thermal decomposition products of plas ticized poly(vinyl chloride)." Paper presented at an Inter national Symposium on Toxicity and Physiology of Combustion Products. Univ. of Utah, Salt Lake City, (March 1976) 4. R.F. Dyer and V.H. Esch. Polyfvinyl chloride) toxicity in fires, hydrogen chloride toxicity in fire fighters. J. Am. Med. Assoc. 235: 393-7 (1976) 5 Y. Tsuchiya and K. Sumi. Thermal decomposition products ol poly(vinyl chloride). J. Appi. Chem. 17: 364-6 (1967) 6. W.D. Woolley. Decomposition products of PVC for studies of fires. Br. Polym. J. 3: 186-93 (1971) 7. A. Tewarson, "The effect of fire-exposed electrical wiring systems on escape potential from buildings -- Part 1. A literature review of pyrolysis/combustion products and toxicity -- poly(vinyl chloride)." Factory Mutual Research Corp Tech. Report No. 22491. Rc75-T-47 (Dec. 1975) 8. P.L. Palokoff, N,L. Lapp and R. Reger. Poly(vinyl chloride) pyrolysis products. Arch. Environ. Health. 30: 269-71 (1975) 9. D.J. Rasbash Smoke and toxic products produced at tires. Trans. J. Plastics Inst. 55-61 (Jan. 1967) 10. N.l. Sax. Dangerous Properties ol Industrial Materials. 3rd Edition. Rcinhold Book Corp.. N.Y., N.Y. 1968. p. 1019. 11 V. Baltalis, Yu. M. Markovich. I.F. Yanembash. and A.S Grebtsova. Toxicity of combustion products of some synthetic materials during a .subsurtace lire. Razrab Mestorozhd. Polezn. Iskop. (Kiev) 39: 158-61 (1975). CA 84: 85715(1976). 12. E.H, Coleman and C.H. Fish. The products of com bustion of chlorinated plastics. J. Appl. Chem. 4: 379-83 (1954) 173 Journal of Analytical Toxicology, Vol. 4, July/August 1980 13. L. Donovan, "Fighting the secret killer in fires". RN, 41: (No. 2), 58-63(1978). 14. R.B. Durant. Lessons to be learned. Navy Lifeline, Occup. Safety and Health J. 10-11 (Jan/Feb. 1978). 15. B.C. Cadoff, P.A. Pella, E.E. Hughes, E-P. Scheide. A.A. Angotti, and J.K. Taylor, "Development of Analytical Reference Materials and Contaminant Generation Sys tems--Work Completed During Fiscal Year 1973". Na tional Bureau of Standards (U.S.), Interagency Report 73*411 (January 1974). 16. G.G. Esposito, D. Lillian, G.E. Podolakand R.M.Truggle, Determination of Phosgene in Air by Gas Chromatogra phy and Infrared Spectrophotometry. Anal. Chem. 49, 1774-78(1977). 17. S.D. Nogare and R.S. Juvet, Jr. Gas-Liouid Chromatog raphy. Chap. 11. Interscience Publishers, N.Y., N.Y, 1962. 18. J.W. Polls and T.S. Lederer. A method lor comparative testing of smoke toxicity. J. Comb, Tox. 4:114-62 (1977) 19. D.R. Stull. Chemical thermodynamics and fire problems. Fire Research Abstr, Rev. 13:161-86(1971) 20. D.J. David and H.B. Staley. Analytical Chemistry of the Polyurethanes. Chap. 2, Volume 16, Part 3, Wiley-lnterscience. N.Y., N.Y. 1969. 21. Gmelins handbuch der anorganischen chemie. 8th Edition. Sec. 6A "Chlor". Verlag Chemie, Weinheim 1968, pp. 14-17. 22. I.D. Morton and E. Hoggarth. "Carbonic acid and its derivatives." Chapter 14 in Chemistry of Carbon Com pounds. E.H. Rodd (Ed.) Volume 1, Part B. Elsevier Publishing Co., New York. 1952. Manuscript received January 28,1980 Revision received March 10,1980 9 9 fctt SVU MATERIALS AND METHODS present conflicting data on PVC toxicity. This may be due to the different formulations of PVC examined and/or differences of experimental designs. In the only previ ously reported inhalation study,1 the authors suggest that continuous (24 hr) .exposure of guinea pigs for 2-7 months in a PVC bagging plant resulted in the formation of granulomatous foci in the lung. The nature of the PVC used in that study-possibly a mixture of homopolymer and paste polymer-is not described, and the level of exposure is poorly defined. Animals. Eighty 10-wk-old female albino rats of the Sprague Dawley C.D. strain (Charles River, U.K. Ltd., Margate, Kent) were used in this study. These were ran domly allocated to either the control or PVC-exposed group so that each group consisted of 40 rats. The rats were housed in groups of four, in polypropylene cages with stainless steel mesh floors. Food and water were^Bj plied ad libitum while the rats were in the cages. The room temperature was maintained at 20 2C, and light Using a simple in vitro technique of hemolysis it has ing was controlled to provide a 12-hr light period each day. been shown that different formulations of PVC exhibit Aerosol exposure. A paste polymer, PVC-7, which con differential biological potential, as assessed by their ability tains sodium lauryl (dodecyl) sulphate surfactant was used to react with red blood cell (RBC) membranes.1,3 Thus, throughout the study. The PVC-treated animals were homopoiymers are not hemolytic, whereas paste polymers exposed to a mean aerosol concentration of 10 mg/m3, containing surfactants (detergents) on their particle sur i.e., "nuisance" dust level (Guidance Note EH 15/77, face produce extensive RBC damage in vitro. The chemi Health and Safety Executive), on 5 days for 6 hr/day for cal formulation of the detergent used in the preparation 5 days/wk up to 15 wk. Groups of 10 exposed rats were of the PVC polymer was, therefore, considered important examined with control animals at 3, 9, and 15 wk after in its hemolytic reaction.3 Similar findings were reported the experiment was initiated. One group of animals was by other investigators4,5 who found that certain emulsion maintained without further contact with PVC for 15 wk (paste) polymers were cytotoxic to rat peritoneal macro after being previously exposed to the dust for 15 wk. A phages maintained in vitro, which is a test system consid nonexposed control group was also maintained during ered by some investigators to be indicative of particle this time period. fibrogenicity in vivo. These investigators,5 however, subse Exposures were carried out in a perspex, cubic, 240-L quently concluded that some paste polymers containing chamber which was operated under dynamic conditions detergents produce a "false-positive" result in the in vitro at a slight positive pressure. The aerosol of PVC particles system because the intratracheal 2-mg injection of the was generated using a Wright dust feed mechanism identical polymers into rat lungs produces no progressive (L. Adams Ltd.; Minerva Road, Chase Estate, London fibrogenic response (unlike a quartz). All the polymers NW10), Following preliminary experiments to determine tested in experimental animals produce only a mild inflam optimum operating conditions, a dust packing force of matory response, and it was therefore concluded that 0.2 tons was used. The mechanism was operated with PVC per se is an inert material.5 standard canister liner at a gear ratio of 4 : 1 reduced. Total chamber air was supplied through the dust feed Intratracheal instillations of high doses (25 mg) of PVC which was run at 1.05 kg/cm2 (25 L/min). The aeroso^^ have been shown to induce biochemical and histopatho- was introduced into the exposure chamber at the base^B 23 & CD logical changes in rat lung tissue;6 other studies7 have indi cated that the cellular and biochemical effects at the alveolar surface and in lung tissue of instilled PVC particles are dose-dependent. However, aggregates of PVC dust center and then ascended through an area clear of holding cages to descend through the cages in which the animals were individually housed, and left the chamber through a series of holes at the base. The chamber was held in a oo>> given intratracheally, like many other materials given in this way, can produce a different, usually more vigorous, response from that produced by inhaled dust Additionally, it is not possible to convert a single intratracheal instilla fume cupboard which was at negative pressure with respect to the laboratory, and the exhausted dust was removed by -an absolute filtration uniL Control animals were placed in tion of a known volume of dust to a period of exposure at an identical chamber, but were exposed only to the same a multiple or fraction of the dust level, which means that for assessing safe occupational exposure levels, the results from intratracheal instillation studies are open to critical interpretation. The aim of the present study was to determine whether the inhalation of a paste polymer, PVC-7, containing the detergent sodium dodecyl sulphate,3'7 at "nuisance" dust level (10 mg/m3) would produce any alteration in the compressed air as that used to operate the dust feed mechanism. Aerosol concentration was determined twice daily by gravimetric analysis of samples collected with an openface filter holder operated in the vertical position. Due to the extremely low aerosol concentration, a sampling rate of 10 L/min was required to ensure accurate assess ment of total chamber concentration. The particle size distribution in the chamber was measured using an Ander lungs of experimental animals. The parameters chosen for son mini-sampler. investigation included a study of the biochemical integrity Treatment of animals prior to biochemical analyses. In of the alveolar surface by estimation of free cell number the biochemical studies six exposed and six control rats and enzyme activities, and determination of pulmonary were examined at each chosen time interval and treated surfactant levels and soluble protein. In addition to histo- identically. pathological examination of the lung, DNA, and protein Each animal was given an intravenous tail vein injection 'synthesis' and hydrolytic enzyme activity of alveolar tis of 75 pG of (methyl-3H]-thymidine (52 Ci/mmol) and sue were monitored. 2.5 pCi of l-[U-14C]-proline (282.5 mCi/mmol) and left January/February 1981 (Vol. 36, No. 1) -L~ .Jj*- REFERENCES 1. Taylor, j, s. 1979. Environmental Chloracnc: Update and Overview. Ann NY Acad Sci 320: 295-307. 2. Morse, D. L.; Baker, Jr., E. L.; Kimbrough, R. 0.; and Wisseman, III.C. L. 1979. Propanil-Chloracne and methomyl toxicity in workers of a pesticide manufacturing plant. Clin Toxicol 15: 13-21. 3. Sundstom, G.; Jansson, B.;and Renberg, L. 1978. Determina tion of the toxic impurities 3,3 ,4,4-tetrachloroazoxybenzene in commercial diuron, linuron and 3,4-dichloroaniline samples. Chemosphere 12: 973-79; Bunce, N. J.; Corke, C. T.; Merrick, R. L.; and Bright, J. H. 1979. 3,3',4,4-teuachloroazobenzene as a contaminant in commercial propanii. Chemosphert 5: 283-84. 4. Bartha, R.; Linke, H. A. B.; and Pramer, D. 1968. Pesticide transformations: production of chloroazobenzenes from chloroanilines. Science 161: 582-83. 5. Poland, A.; Glover, E.: Kende, A. S.;f DeCamp, M.;and Giandomenico, G. M. 1976. 3,4,3 ',4-tetrachloroazoxybenzene and azobenzene: potent inducers of aryl hydrocarbon hydroxylase. Science 194: 627-30. 6. Hsia, M. T. 5.; Bairston, F. V. Z.; Shih, L. C T.; Pounds, J. G.; and Allen, J. R. 1977. 3,4,3 ,4-tetrachloroazobenzene: A potential environmental toxicant. Res Commun Chem Pathol Pharmacol 17: 225-36. 7. Hsia, M.,T. S., and jSurant, G. F. 1979. Preparation and spectral analysis of 3,3 ,4,4-tetrachloroazobenzene and the correspond ing azaxy and hydrazo analogs. / Assoc Off Anal Chem 62: 746-50. 8. Taylor, J. S.; Wuthrich, R. C.; Lloyd, K. M.; and Poland, A. 1977. Chloracne from manufacture of a new herbicide. Arch Dermatol 113:616-19. 9. Kimbrough, R. D.; Carter, C D,; Liddle, J. A,; Cline, R. E.; and Phillips, P. E. 1977. Epidemiology and pathology of a tetrachlorodibenzodioxin poisoning episode. Arch Environ Health 77-86. 10. Adams, E. M,; Irish, D. D.; Spencer, H. C.; and Rowe, V. K. 1941. The response of rabbit skin to compounds reported to have caused acneform dermatitis. tnd Med 10: (Ind Hyg Sec) 2: 1-4. 11. Schwetz, B. A.; Norris, J. M.; Sparschu, G. L.; Rowe, V. K.; Yehring, P. J.; Emerson, J. L.; and Gerbig, C 1973. Toxi cology of chlorinated dibenzo-p-dioxins. Environ Health Perspect 5: 87-99. R&S 114665 ^ Effects in the Rat of Inhaling PVC Dust at the Nuisance Dust Level (10 mg/m ) R. J. RICHARDS, Ph.D. F. A. ROSE, Ph.D. T. D. TETLEY, Ph.D. Department of Biochemistry University College Cardiff, Wales United Kingdom L. M. COBB, Ph.D. MRC Radiobiology Unit Harwell, Oxon United Kingdom C. J. HARDY Department of Inhalation Toxicology Huntingdon Research Centre Cambs., United Kingdom ABSTRACT. Rats inhaled a paste polymer polyvinyl chloride dust at an aerosol concentration of 10 mg/m3 for 6 hr/day, 5 days/wk for a 15-wk period. A small num ber of randomly scattered lung lesions were detected at 15 wk; these lesions were also present 15 wk after exposure to polyvinyl chloride had ceased. Few if any biochemical changes were detected at the alveolar surface in lung tissues or other organs of polyvinyl chloride-exposed rats. It is therefore concluded that at "nuisance" dust level this form of polyvinyl chloride polymer exhibits a weak biological reactivity. DURING ONE STAGE in polyvinyl chloride (PVC) manufacture the material exists as a fine, easily respirable dust. Consequently, investigators have examined the effects of this dust on mammalian lung. R&S 114666 Table 1.-Cellular and Biochemical Studies on Rats Exposed to 10 mj'm3 PVC tXist foe Different Time Periods Exposure Period/ Animal Group Lung Weight/ Body Weight Ratio x 1Q2 Number Free Celis/Animal X.10'6 Pulmonarvt Surfactant mg; Animal Pulmonary! Lavage Protein mg/Animal Acid RNAase units/106 Ceils Acid RNAase units/g Lung DNA TM mg/g Lung 3 Week! Control PVC-Ex posed 0.53 (.086) 0.55 (.082) 6.71 (0.93) 8.86 (1.52) 0.50 0.52 2.25 0.68 (0.12) 74 (10) 2.62 0.69 (0.29) 75 (19) 3 Weeks Control PVC-Exposed 15 Weeks Control PVC-Ex posed 0.47 (.036) 0.48 (.037) 0.43 (.023) 0.40 (.045) 19.35 (4.64) 13.79 (4.63) 11.00 ( 3.61) 11.00 (1.08) 0.60 1.50* 0.66 0.62 3.19 3.96 2.86 3.11 1.17 (0.27) 1.39 (0.31) 108 (12) 104 (12) 0.90 (0.31) 1.10 (0.21) 83 ( 5) 104 (16)* 15+ Wk 15 Wk Clearance Control PVC-Ex posed 0.39 (.028) 0.41 (.063) 9.26 ( 2.93) 11.59 (1.97) 1.28 1.20 3.50 4.13 0.96 (0.22) 0.72 ( 0.23) 45 ( 8) 50 ( 9) NOTE: Numbers within parentheses refer to standard deviation. Significantly different from control. + Values represent mean of pooled samples from six rats. 4.54 (0.76) 4.06 (0.43) 4.93 (1.07) 5.30 (0.87) 4.01 (0.87) 4.78 (0.32) 4.23 (0.25) 4.27 (0.29) for exactly 1 hr. They were then sacrificed by intraperi tonea! pentobarbitone injection, a blood sample was then removed from the hepatic portal vein, and the lungs per fused via the heart with 0.15 M NaCI to remove blood from the vascular bed. The lungs were removed from the pleural cavity and lavaged six times with 0.15 M NaCI, after which they were dried between paper towels and weighed. The lavaged lung tissue and pooled washes were kept on ice prior to further treatment (see below). Other tissues taken for examination were the liver, spleen, kid ney and sections of the gastrointestinal tract which were stored frozen until required. Fractionation of lavage fluid and biochemical analysis. The free cell population from each animal lavage was obtained by centrifugation at 300? for 20 min at 4aC. The supernatant fraction from six rats in each group was pooled and centrifuged at 1000? for 1 hr at 4C. The supernatant fraction from this centrifugation contained the majority of soluble alveolar lavage protein. The pellet was resuspended in 4 M NaCI, mixed well, and centrifuged at 1500 ? for 25 min at 4C, and the resulting separation gave a pellicle of lipoprotein-rich material, designated pul monary surfactant, which floated to the top of the tube. This was collected, dialyzed against distilled water, freezedried, and weighed.9 The free cell population was counted and the levels of acid RNAase and acid protease were 16 determined as described previously.9 The incorporation of [3H]-thymidine into tissue DNA and [14C)-proline into tissue protein was determined as follows. Body tissues were homogenized in 0.15 M NaCI and samples suspended in a final concentration of 0.2 M perchloric acid (PCA) for 1 hr. The mixture was centrifuged at 1000? for 20 min. and the supernatant, containing free label, discarded. This procedure was repeated once more and the resulting pellet was then resuspended in 0.5 M PCA at 70C for 20 min. to solubilize the DNA. The supernatant derived from centrifuging this mixture at 1000? for 20 min. was stored and the process repeated. The supernatant fractions were pooled and samples taken for chemical analysis of DNA and direct counting of [3H] label. The remaining pellet was suspended in distilled water and samples digested in 1 M NaOH to assay for protein content or taken up in Soluene for determination of (14C) radiolabel. The efficiency of counting was determined by using internal standards and results were expressed as disintegra tion per minute of incorporated radiolabel per mg DNA or protein for each tissue examined. Histopathoiogy. The trachea and lungs from four con trol and four PVC-exposed rats were removed after the rats had been killed by exsanguination under deep pento barbitone anaesthesia. The lungs were infused to constant pressure (10 cm water) with buffered 10% formalin, and Archives of Environmental Health - 3(H1Thymidine in Lung DNA dpm x 10~3/mg DNA Protein mg/g Lung 7.5 ( 4.5) 10.4 ( 3.6) 40.0 ( 48.0) 25.6 (14.2) 14.3 ( 6.9) 10.3 ( 2.3) 50.6 ( 8.52) 52.0 ( 7.46) 77.3 (19.72) 87.7 (10.68) 70.9 (10.50) 77.5 ( 4.14) 14(clProline in Lung Protein dpm/mg Protein 164 (55) 165 (47) 160 (23) 120 (23)* 156 (20) 165 (21) 7.7 ( 2.4) 7J ( 1.9) 80.5 ( 6.84) 91.2 (11.58) 94 (33) 74 ( 6) Fig. 1. Lung from rat exposed to PVC dust for 3 wk. Transmission electron microscopy of alveolar macrophage containing numerous smooth-surfaced particles {arrows) identical to the inhaled PVC (scale unit = 2 jum). transverse sections prepared from embedded material. Sections were stained with hematoxylin/eosin, Masson's trichrome stain for collagen,10 silver stain for reticular 2Q fibers,11 and with modified Sudan IV to identify PVC pa particles.13 Some material was also processed for electron C/) microscopy. 4* RESULTS o> ci Body weight and clinical signs. There was no effect on body weight of PVC-exposed rats nor any clinical signs that could be related to PVC exposure during the study. Chamber aerosol analyses. The chamber aerosol was maintained at an average concentration of 10.6 mg/m3 [standard deviation (SD) = 2.3, N = 146]. The highest and lowest concentration recorded was 15 mg/m3 and 7 mg/m3, respectively. The mass median aerodynamic diameter of the aerosol was 1.7 pm (<7g 4.64). Biochemical studies. Very few significant changes in lung weight/body weight ratio, pulmonary surfactant and alveolar surface protein, free cell number and enzyme activities or lung enzyme activities, protein and DNA `syn thesis' were detected in animals exposed to PVC dust throughout this study (Table 1). In addition, DNA and protein `synthesis' in the liver, spleen, and kidney, together with changes in acid protease levels in the free cells and lavaged lung tissue, were not detected in PVC-exposed rats (data not shown). Three significant changes in rats exposed to PVC were found: (1) elevated pulmonary sur factant, (2) depressed lung protein 'synthesis' at 9 wk, (3) elevated lung acid RNAase activity at 15 wk. Histopathology. At 3 wk the lungs of both control and PVC-exposed rats appeared normal, except that in the exposed group the alveolar macrophages were enlarged and contained numerous, 1-4 pm diameter, smooth surfaced particles similar in appearance to PVC dust (Fig. 1). At 9 wk there was evidence of a mild respiratory tract infection in both control and PVC-exposed groups characterized by increased numbers of mononuclear cells in the perivascular spaces, and in most animals, an increase in the volume and extent of bronchus-associated lymphatic tissue. The foamy macrophages seen at 3 wk in PVCexposed animals were less evident, possibly because of an increased turnover of cells in the 9-wk group. At 15 wk there was no sign of the respiratory tract infection seen in the rats at 9 wk. In the PVC-exposed animals there were aggregates of foamy macrophages occupying groups of up to 9 alveoli; also associated with these aggregates was hypercellularity due to an increase in mononuclear cells and fibroblasts, in the interstitium of the alveolar wall (Fig. This interstitial reaction was also associated with minim increase in collagen and reticular fibers. The smooth- January/February 1981 (Vol. 36, No. 1J 17 ^ r 899V VV Fig. 2. Lung from rat exposed to PVC dust for 15 wk. Note aggregation of foamy macrophages and hypercellularity of adjacent alveolar wall (arrows). Hematoxylin and eosin stain (X 250). surfaced particles that filled foamy macrophages stained red with modified Sudan IV, indicating that they were PVC dust.13 Fifteen weeks after the final exposure (30 wk from commencement of the experiment), the lungs of control animals appeared normal, while those previ ously exposed to PVC showed no significant change in the lesions detected at 15 wk (termination of exposure). DISCUSSION The results show that a paste polymer preparation of PVC dust inhaled by rats at "nuisance" dust level (10 mg/m3) for 6 hr/day, 5 days/wk, produces small, randomly scat tered, lung lesions after 15 wk of exposure. These lesions are characterized by hypercellularity of the interstitium of Phe alveolar walls in areas adjacent to macrophage aggre gates containing PVC. In addition, the lesions persist 15 wk after the cessation of animal exposure to the particulate. The PVC-induced lesions, however, show only minimal increase in collagen and reticular fiber formation, with no evidence of extensive fibrotic reaction. Similar conclusions are reported by other investigators5 who studied the effects of intratracheal instillations of different PVC formulations. The results of the current histopathological study there fore suggest that at "nuisance" dust level PVC has a rela tively weak biological reactivity, and this conclusion receives further support from the biochemical investiga tion. While PVC-exposed animals have elevated levels of pulmonary surfactant and show a decrease in lung protein `synthesis' 9 wk after exposure, the presence of a mild respiratory tract infection in all the animals during this time period prevents the establishment of any definite con clusions. Thus relatively few, if any, biochemical changes in PVC-exposed rats are detected at any exposure periodeither at the alveolar surface, the lung tissue, or other body 18 Archives of Environmental Health organs. Such results are in contrast to previous inhalation studies where rats exposed to 12 m^/m3 chrysotiie asbestos for 5 days/wk for 15 wk have elevated free cell numbers (2-3 X control animals) free cell and lung enzyme activity (3-10 and 1-8 X control, respectively), and pulmonary sur factant (12 X control).* Elevations in pulmonary surfactant and free cell numbers are also detected in rats inhaling amosite asbestos and fiber glass (12 mg/m3, 5 days/wk, for 8 wk).13 With dusts of "high" biological reactivity these biochemical changes at the lung surface persist with the progressive pathological development of lung disease. In summary, the present short-term study indicates that at "nuisance" dust level one form of paste polymer PVC has a weak biological reactivity, only detectable by histopathological examination, which reveals a small number of lung lesions in experimental animals. It is not possible at this time to interpret this finding in terms of the lesion likely to arise in man under similar conditions of exposure. Drs. Richards and Tetley would like to thank the Medical Research Council for financial support We are also grateful to the Huntingdon! Research Centre for the provision of inhalation expo sure facilities. Submitted for publication November 14, 1980; revised; accepted for publication December 8, 1980- Requests for reprints should be sent to: Dr. R.). Richards, Department of Biochemistry, University College Cardiff, P. O. Box 78, Cardiff CF11 XL, U.K. REFERENCES 1. Frongia, N.; Spinazoila, A.; and Burarelli, A. 1974. Lesioni polmonari sperimentali da inaiazione prolungata di polveri di PVC in ambiente di lavoro. Med Lav 65: 321*41. 2. Richards, R. J.; Desai, R.; Hext, P. M.; and Rose, F. A. 1975. Biological reactivity of PVC dust Nature 256: 664-65. 3. Richards, R. J.; Desai, R.; and Rose, F. A. 1976. A surface- active agent involved in PVC-induced haemolysis. Nature 260: 53-54. 4. Pigott, G. H. 1976. In vitro studies of a range of formulations of PVC Polymer Preprints 19: 29-34. 5. Pigott, G. H., and tshmael, J. 1979. A comparison between in vitro toxicity of PVC powders and their tissue reaction In vivo, Ann Occup Hyg 22: 111-26. 6. Agarwal, D. K.; Kaw, J. L.; Srivastava, S. P.,' and Seth, P. K. 1978. Some biochemical and histopathologlcal changes Induced by polyvinylchloride dust in rat lung. Environ Res 16: 333-341. 7. Tetley, T. D.; Rose, F. A.; and Richards, R. J. 1980. Biochemi cal and cellular reaction of PVC paste polymers and latex following intratracheal instillation into rats. Inflammation (in press). 8. George, G., and Richards, R. ]. 1979. Preliminary studies on the isolation, separation and identification of pulmonarylavage proteins from the rabbit. Biochem Soc Trans 7: 1285-87. % TetJey, T. D.; Hext, P. M.; Richards, R. J.; and McDermott, M. 1976. Chrysotile-induced asbestosis: Changes in the free cel] population, pulmonary surfactant and whole lung tissue of rats. Br j Exp Pathol 57: 505-14. 10. Bradbury, P., and Gordon, K. 1977. Connective tissue and stains. In Theory and Practice of Histological Techniques, I, D. Bancroft and A. Stevens, eds., p. 104. Edinburgh, Lon don, and New York: Church ill-Livingstone. 11. Gordon, H., and Sweets, H. H. 1936. A simple method for the silver impregnation of reticulin. Am J Pathol 12: 545-51. 12. Wilson, N. 1979. A method for staining polyvinyl chlorine In sections using Sudan IV. Stain Techno/ 54: 101-02. 13. Richards, R. j.; George, G.; Hunt, J,; and Tetley, T. D. 1980. Relationship between the hemolytic potential of certain par ticulates and their reactivity at the lung surface in vivo. In The In vitro Effects of Mineral Dusts, R. C. Brown, M. Chamberlain, R. Davies, and I. P. Gormley, eds., p. 323-332. London: Academic Press. INDUSTRIAL HYGIENE SAMPLING STRATEGIES (NIOSH 553), sponsored by the Midwest Center for Occu pational Health and Safety, will be held on April 22-24, 1981, at St. Louis, Missouri. The course content includes introduction to statistical sampling strategies, legal aspects of sampling strategies, fundamentals of statistics, estima tion and decision l and 11, exposure measurement sampling strategies, compliance vs. non-compliance, full period sam pling, grab sampling decisions, ceiling limit sampling, com pliance officer sampler strategies, and statistical workshops. Industrial hygienists and supervisors who are responsible for sampling industrial atmospheres, making decisions on such sample results and taking appropriate action in com pliance with OSHA regulations should plan to attend. Three points will be awarded toward maintenance of certification from ABIH. CEUs will also be awarded. There is a fee of $375.00. For further information, write or call Ruth K. McIntyre, Director, Continuing Education, Midwest Center for Occupational Health and Safety, 640 Jackson Street, St. Paul, Minnesota, (612) 221-3771. January/February 1981 (Vol. 36, No. 1) 33 </> fc. O) O CD 19 0i9 n ^ S18U -------- IMPERIAL CHEMICAL INDUSTRIES LIMITED PLASTICS DIVISION--WELWYN GAROEN CITY From." __ j stATTOKD Dinalan Mukgvr Health 4 brtraMttt Protection DSO-16 Sit 3162/3359/M96 3859 - Or Stafford'* Sec. 2096 - Telephone contact ia TS Mr Owen Jones Dr W G F Adams (6) Dr B Bennett Dr F W Best Dr D P Duffield Dr A P Wright Mr J C Edwards Dr D W Plester Dr D A Tester Dr J G Kamrauller Dr J P Tassignon Mr G Dupont Dr T B Torkelson Dr M N Johnson Mr S Nakamura PVC DUST Geoff Pigott of CTL has kept in touch with this work at Cardiff. The PVC materials were supplied by ICI and BP under CIA code numbers so anonymously that I have no idea whether PVC 7 came from ICI or BP. Perhaps Dr Tester could disentangle this for us? J Stafford js/sew/dso-i6 23 June 1981