Document 9Y43rarg9rwQxv2j3o74b5o7

B2HZENE TCXICIT?; review or nacsrr irrsiATQRE / Prepared Zor the American Petroleum Institute sy 3emard 0. Goldstein, tt.2. Deparsient aZ Savironmental and Ccaonunity Medicine CMDNJ-Rutgers Medical School Piacataway, SJ 00854 Pebcuary 2, 1983 MCD 000008339 i>rreoppcT--k This review of Che literature concerning the toxicity of benzene focuses al most entirely on work published from 1978 to the present. It in intended as a fol lowup of the information in "Benzene Toxicity: X Critical Evaluation* edited by Lankin and Goldstein (1977). That was a multi-authored volume containing over 1,000 references which described in depth the known benzene literature to that point. This present single-authored review will not attempt to be as comprehensive In its coverage. Bather, the focus will be on reviewing and interpreting the recent literature particularly pertinent to scientific questions underlying the regula tion of benzene exposure. The following brief overview of the Literature until 1977 and cf the hematologic basis cf benzene toxicity is intended as background for the present review. For those interested, a sore detailed description is provided in La a kin and Goldstein (1977) and other scholarly reviews (Haley, 1977; White at al., 1980; Gold stein and Snyder 1962; Snyder et al., 1982). Benzene has been known to produce hematological toxicity in man since the nineteenth century. The earliest reports of benzene coxacity focussed on its ability to profoundly decrease all formed elements in tne blood (red blood cells, white blood cells, and platelets). This disease, known as aplastic anemia, is characterized by loss of bone marrow cells which are the precursors cf mature cir culating blood cells. Investigation of workplaces with significant benzene ex posure revealed a wide spectrum of response ranging from severe aplastic anemia to mild decreases in red blood cells (anemia), white blood cells (leukopenia), and platelets (thrombocytopenia); the milder form of this disorder often being known as pancytopenia. In association of benzene exposure with acute myelogenous leukemia was first noted over five decades ago. Tha evidence confirming a causal relation ship between benzene exposure this disease comes from a variety of different ap-- MCD 000008340 diverse disorders as acute aad chronic lymphatic leukemia* chronic syelogecoua leu* kesla* multiple myeloma and various types of lymphoma including Hodgkin's Disease, say also be related to benzene exposure- Reports of such disorders linked to benzene exposure are tabulated by Goldstein (1977). Information available in 1977 about the mechanism of benzene toxicity pointed strongly to the eotlon of a metabolite of benzene rather than benzene Itself in producing bone marrow toxicity. Other studies in both animals and man suggested that host factors* such as age* sex* obesity* and genetic factors* sight modulate the extent of benzene toxicity in terms of its aplastic effects. Although dose-res ponsive pancytopenia was readily produced evidence of leukemia in laboratory animals exposed to benzene was not observed. Also of note was the consistent failure to observe mutagenesis due to benzene in bacterial systems such as the Ames test des pite ample evidence that chromosomal damage was produced in animals and man. vtuntvv tv yaw Among the more important studies published in the pest five years has been that of Sinslcy et al (1981) which extended and elaborated on the prior study by Infante et al (1977). These investigators from JfXQSS originally reported seven deaths due bo leukemia in a cohort of 738 white male rubber workers employed at any time during the period of 1940-19*9 v* died between January 1, 1950 and June 30*' 1975 (Infante et al* 1977) Of the seven cases* 4 were reported as acute myelogenous leukemia* 2 as acnocytlc leukemia* and i as chronic ayelogecoua leukemia. The expected cumber of deaths in the cohort was 1.38 based on the general population of TS white sales and 1.38 based upon a control group of white males employed-in the same state and at the sane time la manufacture of a fibrous glass product. These findings* and the asser tion that the workers had been exposed to benzene levels within allowable workplace standards* led to an attempt by OSSA to promulgate a sore rlgcrcus standard of 1 ppm TVA as compared to the existing '0 pern TVA. Although the standard remains ac 10 pFm MOD 000008341 -3- recently to exceedingly high levels of benzene comes from industrial hygiene sur reys conducted by the University of Mortb Carolina in 1973-74 in which levels of 193255 ppm benzene were noted inside a dryer and it was reported that respirators were often not worn. It is further seated that men could stay up to 30 minutes at this location but it is not clear whether that occurred without use of a respirator. Similarly, during a 1972 shutdown operation* it was noted by the safety manager that during a short-term check an employee was not wearing respiratory protection in an area in which bes2aae levels were estimated to be over 200 pps. Using this monitoring data, 31nsky et al (1981) have attempted to estimate the tfeozene exposure levels for the individuals who developed leukemia* They present short came reports including the type of leukemia, interval between exposure and oaset of disease, and time period of employment* For the 7 individuals originally reported by Infante et al (1977) TYA exposure levels while at work are estimated to be 35 ppm (one worker), 40 ppm (three workers), "less than 100 ppm" (one worker), usable to estimate due to exposure occurring between 1940-42 (one worker), and, in a worker employed for a very short period of time, a possible slhgle exposure episode to 1,000 ppm benzene* The study of Sinsky et al. (1981} has bees interpreted as indicating that Densene levels at or near the present standard are leukemogenlo (Sosenstock, 1982; White et al., 1981). This of course is a possible interpretation of the relation of benzene to acute myelogenous leukemia in man. However, in my Judgment the 3iask7 et al. data, and the original findings of Infante et al., do cot provide significant direct support for leukesogenesis occurring at levels of benzene at or near the present 0SHA standard. This interpretation Is due both to the aaericail7 low in cidence of leukemia and to the marked heterogeneity of benzene exposure in the work force studied. Ives when benzene exposure leads to an increase in leukemia in cidence sufficient to be statistically appreciable, as in the Infante et al study, the total incidence of leukemia has fortunately always been a very small fraction of ** *978). Aksoy continued to review and summarize hta experience wits shoe workers and ethers occupationally exposed to benzene in Turkey ( Askoy 1 980 a, b). During the 1980s shoe workers in Istanbul replaced gasoline with benzene as their solvent of choice for working with adhesives. Cases of aplastic anemia followed 'by cases of acute leukemia were observed in this work population. Exposure levels were well over ICO ppm benzene. The recent review articles contain further substantiation of a decrease in the incidence of leukemia among shoe workers in recent years which could be attributed to the gradual prohibition and discontinuation of the use of benzene in Istanbul beginning 1969. The substitutes now in place contain little or to benzene (AksoyandErdam* 1978). After reaching a high of seven cases in the year 1973* leukemia incidence in shoe workers in 197* was four cases* in 1975 three cases* and is 1976*78 oo cases. Aksoy does report four cases of leukemia in benzene exposed individuals in 1979 however only one was a shoe worker. Also of interest is that 11 of the *4 individuals with leukemia had a documented period of pancytopenia preceding the onset of leukemia. Ho evidence is presented which would allow one to conclusively exclude the possibility of a pancytepenlc period preceding the onset of leukemia in the-other 33 patients. In studying those patients who originally presented with pancytopenia* Aksoy noeed complete remission in 23* death due to aplastic anemia or its complications in 14* and leukemia in six of a total of 44 patients. In a tabulation of *2 individuals with leukemia and benzene exposure Aksoy noted 16 with acute myelogenous leukemia* eight with erythroleukeaia which can be considered a variant of acute myelogenous leukemia; one with acute prcayelocytlc leukemia* also a variant of acute myelogenous leukemia* and seven in dividuals with preleukemia which usually also evolves into acute myeloblastlc leu kemia. In addition he reports four individuals with acute lymphoblastic leukemia* three with acute monocytic leukemia, two with chronic myelogenous leukemia* and one with acuta undlfferentlaeed leukemia. Ho Individuals with cnrooic lymphocytic leu- MCD 0000083^3 -7- forming their risk estimate, and as the mean and median age of his patients with myelogenous leukemia is veil below tbat at which peak incidence occurs, tnis tends to underestimate the relative risk. Aksoy's observation tbat many of tbe in dividuals with benzene induced pancytopenia progress through preleukemla into acute myelogenous leukemia Is In keeping wltb tbat reported by many authors in in dividual case reports following exposure to benzene around tbe world. These obser vations bad led hematologists to accept tbe causal relationship cf benzene to acute myelogenous leukemia decades ago. Finally, if sore proof were needed, the virtual cessation of sew oases of both pancytopenia and leukemia among these workers fol lowing a decrease In benzene usage fits a temporal pattern that Is conclusive. There thus remains no question that an epidemic of aplastic anemia and acuta myelogenous leukemia occurred due to exposure to high levels of benzene In the Turkish shoe in dustry. -- Pagsetto et al. (1979} have presented an Interesting follow-up on 38 workers previously exposed to benzene levels generally between 5-50 ppm until 1964. This study is notable for the relatively extensive industrial hygiene and hesatologleal studies that were performed between I960 to 1984 which resulted in reasonably detailed characterization of the workforce exposed to beszene. The observed am bient levels of benzene were confirmed by measurement of urinary phenol excretion. Mild hematological abnormalities were present in a number of the workers when studied in 1961 and 1963. Since cessation of benzene exposure in 1964, retesting performed in 1973 and 1977 revealed what appeared to be complete recovery. Mo hematological sallgandes were observed during this 13 year follcv-gp period. Al though this lack of leukemia In these well-studied workers Is reassuring, the cohort is far too small to be meaningful. The Occupational Health Studies Group of the University of ttorth Carolina School of Public Health has conducted a number of studies of zortality of vcrkers in the GS rubber industry. They noted an excess of_deaths attributed to neoplasms of -9MCD 000008344 ciioaen to separate this from the acute myelogenous leukemia group. Also rather la the finding that of the 3? cases of leukemia in Company A only oca in dividual had chronic myelogenous leukemia, k more expected incidence for this age group would he the five of 27 cases of leukemia reported in Company B. The use of death certificates to sub-divide the type of hematological malignancy is clearly a perilous undertaking, particularly In the differentiation of leukemias. This probably accounts for the variable findings between companies, although the pos sibility that soma highly unusual process was occurring in Company A cannot be ex cluded* Even if this were the case, it seems most unfortunate that the investigators undertook this major effort without attempting to confirm the hematological diagtests of these individuals through careful review of the clinical findings. In view of these problems it is net surprising that no association was found between myelogenous leukemia and solvent use* The authors conclude that these findings sup port their previously reported association of lymphatic leukemia with work history of possible solvent exposure but that the association was weaker than had been noted. However, it is also stated that a recently detailed environmental evaluation reported as an abstract is said to support tbe initial findings and that further studies are therefore required to resolve the issue. Perhaps of note is that of the jsb categories evaluated lymphatic leukemia and monocytic leukemia were par ticularly observed in geaeral service workers. Monscn and his colleagues at the Harvard-School of Public Health have also con tinued their series of studies evaluating the health of rubber workers. Secant studies inelude an assessment of cancer morbidity and mortality among 131570 white sale rubber workers who had worked for at least five years at an Akron, Ohio ruber plant (Monscn and Fine, 1978). Occurrence of cancer was measured by death cer tificates and by a survey of local hospital tumor registries. Comparison of mor tality rates with these of US white males was performed as was analysis of cancer morbidity rates among different subgroups of workers. Excesses in a cumber cf dif- MCD 000008345 -i i. SMH reported is that for central nervous aystea neoplasms with an SiH of 69. This in aaid to be the only statistically significant effect. Aiaong sale tire workers SKRs for leukemia were 129 and stomach cancer 1U5. gilpikari (1932) baa also published a study evaluating mortality among sale rubber workers in Finland shoving a decrease in risk of cardiovascular disease and an apparent Increase In cancer incidence. However, the cohort size is too 1 for further evaluation. Another paper compiling a series of studies Is that by Burrows (1930) coo* seating on health care of workers In research laboratories. In three studies in dif ferent countries, that of the American Chemical Society membership, sale graduates of the School of Chemical Engineering, Stockholm, and graduates and fallows of the fioyal Institute of Chemistry In Great Britain, there is an increase In tha number of neoplasms of the hematological system. Two studies from Sweden suggest an association between lymphoma and solvent exposure. Hardelletal. (1930) reported on a case control study of 169 cases of all types of lymphoma and for each ease there are two controls. A statistically sig nificant relative risk of 2.8 was observed for these with a history of substantial exposure to solvents, while those with a lesser exposure bad a relative risk of 1.2. Increased relative risks were also observed In Individuals exposed to pheccxy acids and to cilorophenols and there seems to be an enhanced risk with combined exposure to the various chemicals. It should be noted that only one individual gave a history of substantial exposure to benzene. In another study, Olsson and Brandt (1981) used a standardized interview in 61 consecutive sale patients with a diagnosis of nonHodgJOn's lymphoma to evaluate a possible exposure to organic solvents. They report that in those who had a history of organic solvent exposure there was a such higher incidence of supradiaphragmatic presentation of disease as compared to those without such history of exposure. Again Is this study an overt history of benzene exposure was found in only one patient. Benzene was apparently banned for oc cupational use in Sweden in 1972. Sowever, the extent to which other solvents con- MCD 000008346 risk 0.70 and 0.9* respectively) A relative risk of 3 *3^ was observed for exposure to benzene, but due to the small numbers thin wan not statistically significant. Of interest in that 3 of 4 patients with a history of benzene exposure were diagnosed an having chronic lymphocytic leukemia. Timocen and Ilvonen 0978) suggested that contact with hospitals is an tiolcgic factor in leukemia (defined as acute leukemia and chronic myeloid leu kemia) This was based on a study of 45 adults and a control group of patients from the same hospital in Finland* Ho significant difference in the use of chemicals between the two groups was observed, although there was apparently no history of the use of benzene by any of the study population. The study of 71anna and Polan led Smith and Licklaa (I960) to evaluate the ex tent to which Hodgkin's Disease and ncn-Sodgkin's lymphoma is associated with ben zene exposure in Tasmania* Their study includes n new cases diagnosed between 1972 and 197?* Using the same criteria for occupational benzene exposure used by Tlanna and Folan, they were unable to confirm an association with lymphoma* Of note, however. Is that a statistically significant increase in deaths due to acute syelogenous leukemia was observed suggesting that benzene exposure bad in fact oc curred. Cytogenetic abnormalities have been demonstrated to result from heaatotoxlc levels of benzene in previous studies in animal s and aan (see below). Another re cent, and somewhat controversial, study reporting cytogenetic abnormalities in workers exposed to relatively low levels of benzene Is that of Picciaao ('979) The author reported an increase in chromosome aberration rates in the peripheral lym phocytes of 52 workers in a chemical company as opposed to a 4A person control group observed for pre-employment examination* The age of the exposed group was sig nificantly greater than that of the ccatrol group but was said by the author not to affect the results. The workers were reported to Save been espcsed to less than 10 MCD 000008347 -15- cohort studies of entire Industries. Tie finding of fewer than expected leukemia deaths in the entire cohort of this industry in which benzene exposure is known to occur apparently conceals a true benzene effect which can only be observed with more careful evaluation of the workforce assessing the degree and duration of exposure. It should be noted that in this study, as in many others, the observed leukemia In* alienee, although less than that expected based upon a control group derived from the general, population. Is actually more than that observed for all neoplasms or all causes of death. This 'healthy worker effect' is well known and is almost uniformly demonstrated in studies of this type. Such results argue in favor of the use of proportional mortality ratios as opposed to standard mortality ratios. Another approach to evaluate mortality among workers at petroleum refineries and petrochemical plants was taken by Thomas et al C1990}. They studied the nor* tall17 experience of 3,105 members of the Oil, Chemical and Atomic Workers Inter* national Onion COCAV) reported by locals la Texas to have died between 1947*77 and for whom death certificates were available. The proportional &ortalit7 ratio was calculated based on expected mortality for the general population adjusted for age, race, and cause of desth. The workers were divided into four categories depending upon the type of work performed. A statistically significant increase in the SKR for all cancers in white sales (1.17) was observed. This was particularly notable for skin (1.76), brain and CSS (1.46) and respiratory system (1.25)i all of which were statistically significant. 63 hematologic neoplasms were observed as compared to 54.4 expected resulting in a PMR of 1.16 which was not statistically significant. The subgroup with the highest PMfl for hematologic neoplasms was that of 'petroleum refining and production of petrochemicals" (44 observed, 56.3 expected; PMH 1.21). In this work* category, there was a tendency to as increase In the FMB for leukemia with increasing length of membership is the Galon. This was also observed when the expected tumor frequency was calculated based upon data for Texas sales. However, in zo case were the findings for hematologic neoplasms statistically significant. Mod 0000834q -17- below 1s that benzene say be capable of producing non- hematological neoplasms is man. Furthermore, the lack of ayeiogenous leukemia la .Maitool's studies could even be interpreted as indicating that benzene causes these other tumors at a higher la* cldence than It does leukemia. Obviously, one can aot readily extrapolate relative cancer incidence from animals to man; there being oany instances of species dif ferences in organ selectivity of a carcinogen. The findings by Kaltoni et al., If one accepts their validity, do however serve to raise the hypothesis that benzene vlU cause non-hematologlcal neoplasms In man. This has also been suggested by Iksoy (1580a, D) who coted five cases of lung cancer associated with chronic benzene exposure (see below) Of note is that It Is at least possible to begin to evaluate this hypothesis. There are numerous cohort studies of work groups who have been ex posed to benzene and for whom cancer mortality has been evaluated, is discussed above, an increased incidence of hematological neoplasms has been observed for many of these groups. Inasmuch as the causal link between benzene exposure and at least certain hematological neoplasia is now beyond reasonable doubt, one can as a first approximation utilize an observed increase la hematological cancers la these work groups as a Barker for significant benzene exposure. Therefore, la cohorts selected In this way it should be possible to evaluate the relative Increase in risk of non- hsmatcloglcal cancers from this same benzene exposure. There are of course a number of problems with as approach of this type. These include the fact that all hematological neoplasia have cot been proven to be associated with benzene; that ex posure in none of the cohorts is to benzene alone but to a aixture of workplace . and factors; asd that for oany of these cohorts Che increases in hematological neoplasia are too small to reach statistical significance. Two other considerations which are Inherent in an approach of this type sust be kept in Bind . in actual causative role for bensene in a specific son-hematalcglc cancer say net be observed in a cohort with benzene-induced hematological neoplasia sinpiy because the relative risk for these hematological cancers is greater than that for the other neoplasm. Stated another way, one say have a cohort of a sufficient sine to observe a MCD 0000083^9 -19- these caveats, it should be pointed out that a staple sign test or other acn~ parametric statistical approach which evaluated simply whether or set each finding was above or below the expected incidence, night veil lead to a statistically sig nlficant association wltn hematologic cancers, depending upon how the studies were chosen. To sumnarise, epidemiological studies since 1973 have provided further con firmatory evidence that occupational exposure to benxane results is hematological neoplasms, particularly acuta myelogenous leukemia and its variants. There has been marginal improvement in the ability to make a dose response estimation of the levels of benzene responsible for adverse hematological effects, including Leu- kola. This is in part due to many additional thorough studies of benzene exposed cohorts, confirmatory evidence concerning the relation of benzene exposure ft aplastic anemia and acute myelogenous leukemia is Turkish leather workers, and th follow-up information obtained by Biasky et al. (T981}, concerning exposure is s workforce of rubber workers previously shown to have a high incidence of leukemia. However, the data do not appear to provide any evidence concerning the shape c: the benzene-leukemia dose response relationship, and particularly, the relation ship to leukemia of benzene levels at or below the current occupational standard. srrmm rs un rs 7TTM There been an explosive increase in the number of publications detailiz laboratory investigation into the metabolism and tcsicity of benzene. The fol lowing discussion is aimed at *1 gM4 g*** "g certain of the directions in th: research, Ho attempt will be made to provide a detailed description of the liter cure or to reference every publication. Perhaps the most notable finding in the past five years has been the deveic meet of arn-ai models of benzene carcinogenesis. The previous inability to deac strate that benzene produced tumors in animals had been intriguing; particularly -21- cidence of other pathological lesions, described as nodular hyperplasia and dys plasia, were reported in the benxene-exposed rata, with again a zero incidence in the control anInals. Subsequently, an increased incidence of aammary tumors has also been observed in this experimental group (unpublished data). In the third ex periment, 7 week old Sprague-Cawley rats were given 500 sg/kg bensane, toluene, xylene, or ethyl benzene by stcmaeh tube once dally, 4-5 times weekly, for 10b weeks. Observations at 84 weeks revealed an increased incidence'"of ryabal' gland tumors in the benzene-exposed rats* So hepatomas had bees observed although the seas latency period for this tumor based on their previous study had not yet been reached. The incidence of hematologic neoplasms also did sot differ from the csn~ trol group. Of note* however, was the observation of a statistically significant increase la tumors of the oral cavity. 1 few tumors of this type were observed la th ` first study, also performed by stomach tube. t is interesting to speculate that tfco different patterns of neoplasms observed la these studies sight reflect variation la the site of action of besxeso depending upon the mode of administration. Ob- vleusly, mere Informatics Is needed concerning the results of these ongoing experl* seats. The presence of hepstie pathology aad tumors is particularly intriguing, sac deserves further evaluation, restudy of the pathological findings in tht UTH Institute of Environmental Medicine studies which also exposed Spregue-Dawle: rats to similar Penmens concentrations. Is discussed above, these fishings should lead to a closer examination of the aon-bematologleal cancer mortality experleno of workers with knows exposure to benzene. Zshalatlon toxicology studies at the Sew fork adversity Institute of Savlrcs Medicine by C. Snyder and hi-9 colleagues have provided clear-cut evidence c benxese induced lymphoma and the first suggestion of syelogesous leukemia : laboratory animals. The inhalation regimen used in this laboratory consists of ei posures six hours a day, five days a week, for up to lifeline. In 057/BL nice expos< to 300 ppm besnene a higher incidence of an inducible thymic lysphcoa was obsem MCD 000008351 dueed myelogenous leukemia. Sued a model would allow evaluation of whether pancytopenia.is a necessary prerequisite for the development of acute myelogenous leukesla. Studies by this group exposing various mouse strains and Sprague-Dawley rats to 100 ppa or 300 ppa benzene up to llfetiae are suaaarized by Goldstein and Snyder (1982) The hesatologleal findings in these animals document strain differences, as well as the narked relative resistance of rats as ccapared to alee which Is reflected in benzene pharaaccJcLnetlcs (Snyder et al, 1981a). Lyaphecytopenia is the earliest finding and the sost sensitive indicator of benzene effects in these aalaals among standard heoatologic tests (Snyder et al, 1978;1980;1982; Green et al, 1981a) 1 proliferative effect on nyeloblasts and/or pronyelocytes was observed In CE-1 alee, the strain which appears to bo susceptible to benvene-1 nrtuced leukemia (Snyder et al, 1981) Other studies by this group evaluating the interaction of ben zene and ethanol ingestion, and assessment of bone marrow cells In culture are described below. the mechanism by which benzene night cause leukeaia or other cancers is still ' unknown. Benzene has not been found to be autageolc in standard bacterial tests. Segatlve studies include that of Florin at al. (I960) who tested benzene concentra tions of 0.03, 03 3, and-30 aicroooles/plate g Salmonella typftlmurium strains 7198 and Til 00 with and without an activated liver preparation from 3-methyl cholanthrene-induced rats. Despite these segatlve findings is vitro there is so question that benzene produces cytogenetic abnorsalitles la vivo la both animals and man. Hecent observations Include demonstration of a positive mlerosucleus test In animals administered benzene (Hite et al., 1980; Diaz et al.,1980; Tunek et al. ,1982). However, one does wonder about the validity of the mlerosucleus test as a marker for genetic toxicity due to hesatologleal poisons. Perhaps the sost In teresting of the short-ters in vivo tests for genetic toxicology recently* used with Mphangw fgcsi. Tice et al (1980) demonstrated an increase In SCS in bone marrow cells of DBA/2 alee inhaling 31 PP of benzene for MCE 000qq52 -25 tur techniques which allow the growth ana identification of different stages of naturing hesaecpoietlc precursor cells. Exposure to benzene has been demonstrated to inhibit growth of bone marrow cells in culture by a number of investigators using different techniques and different species* Findings include a decrease in bone marrow colony forming cells fallowing ezpoaure of BDF1 dee to 4680 ppm benzene for 8 hours (Oyekl et al, 1977) i a decrease in stem cells measured as spleen colony forming cells in C77B1/6J female mice exposed to 400 ppm benzene 6 hours dally for 9- 11 days (Sarigaya et al, 1981); a decrease in stem cells observed during exposure of C57S1/6 male mice to 4,000 ppm benzene 6 hours daily, five days weekly* for up to six weeks (Gill et al* 1980), and the prevention of the development of spleen colonies in C57B1/6 mice transplanted with normal marrow cells prior to exposure to 100 ppm benzene for 8 days (Gill et al*1980). In this latter study, the authors report a dose-dependent decrease in white blood cell count in which there is an approximately 80S decrease is mice continuously exposed to 100 ppm benzene for only 4 days* Ssak and Speck (1982) have recently reported that rabbits treated subcutaneously with 0.2 ml/kg benzene, 5-7 days weekly, for 2 months develop not only an inhibition of in vitro colony formation of their own bone marrow, but also of normal bone marrow ooeultivated together with the marrow from the benzene treated rabbits. This in teresting and surprising finding, vhlcn suggests chat benzene is not only directly syelotoxic but also results in bone marrow cells with growth inhibitory potential, requires replication and extension. Green et al (1981b), at the Hew fork University Institute of Environmental Medicine, performed as extensive dose response study evaluating splenic colony forming cells and grsnuloeyte/macrophage progenitor cells in CD-I mice. Effects an splenic granulocyte/ macrophage progenitor cell levels and os femoral bone marrow and splenic colony forming units were observed following exposure to 103 ppm benzene or higher levels 6 hours daily for five days. In a second study, alee were exposed to 9.6 ppm benzene, 6 hours daily, five days weekly, for a total of.50 exposure days (e-- mcd 0008353 -27- catechol, hydroquincne, and benzoqulnone la rat tissue following inhalation of ban* 2ace. Fraa catechol and hydroquinoca vara found to persist in the bona narrow longer than benzene or phenol (Richert at al, 1979). Greenlee at al OS81) have suggested that the mechanism by which polyphenol aatabolltes produce heaatotozlelty Is through the autozldatlon of hydroqulnone and 1 ,2,4-benzenetriol resulting in the formation of aemlquinone and qulnone oxidation produets as well as superoxide anion radical. Irons and Heptun 0980) have advanced the interesting hypothesis that the polyhydroxy netaholltes or their autozldatlon produces interfere with olcrotuhule function perhaps resulting in eytoskeletal abnormalities central to benzane tox *icity. The findings in these and other studies provide relatively convincing evidence of a role for polyhydroxy metabolites in benzene toxicity. There is however still some question concerning why short and long-term studies of the tox icity of.phenol or other non-benzene precursors of these polyphenol metabolites do aot demonstrate pancytopenia. It is possible that the polyphenol metabolites par ticipate in benzene toxicity, particularly the lymphctcxlc effects, but do not ac count for the totality of the observed effects. Other studies by Irons and colleagues include an assessment of bone marrow cell cycle kinetics in rats given benzene by subcutaneous infection in which the findings suggest that benzene toxicity Is relatively specific for the cell dif ferentiation and cell cycle phase of bone marrow precursor cells (Irons at al, 1979) 1 series of studies by this group have included evidence that Eba lTmphoeytopenlc effect of benzene in rabbits primarily affected B lymphocytes, although a decrease in T lymphocytes was also observed at higher doses (Irons and Moore, 1980). The affects of benzene on T cell and S cell function have also been studied by Karaulov and Frasch (1978). Vlerda et al (1981) found a variety of al terations in lymphocyte responses, including inniblticn of 3 and T cell sitogenesls, in C5731/6 mice injected intraperitoneally with *4-660 ag/kg benzene for three days. Of interest is that pretreatment with Aroclor 1154 altered acme but MOD 000008354 -29- b present in other benseae metabolising systems (Andrews et al, 1979; Snyder et al, 1982), Of particular interest is the recent observation by Ealf et al (1982) that benzene inhibits SNA synthesis in mitochondria from liver and bone narrow. Gill and Afaned (1981) noted a significant extent of binding of label derived from radioactive benzene to bone narrow Mitochondria in a study which also observed that 172 of the label in the narrow was bound to nucleic acids. This evidence of a Mitochondrial ef fect opens a new line of investigation into the biochemical basis of benzene tox icity. Goldstein et al (1981) have suggested that a ring-opened fora of benzene may be responsible in part for its heaatotoxicity. Evaluation of transtrans-eueonal-- debyde, which is formed in the reaction of hydroxyl radicals with benzene, has found this compound to be heaatotesLc but there is as yet no direct evidence that it is formed fros benzene in vivo. Studies of interactive effects have included the observation that 52 or 152 ethanol in drinicing water markedly potentiates the heaatotoxicity of 300 ppm ben zene inhaled by C5731/6 mice, six hours daily, five days weekly (Snyder et al, 1981; Saarson et al, 1982), Of interest is that soae facets of the heaatotoxicity of the two agents together differ from those observed with benzene alcca despite the fact that alcohol by Itself had oo overt hematological effect. The interaction of toluene and benzene continues to be evaluated. Combined exposure usually leads to a decrease in benzene-induced hematotoxlclty, apparently due to inhibition of ben zene aetabolism (Andrews et al, 1977). However, the importance of this finding to man has been questioned in a study by Sato and Hakajima (1979) who found no sig nificant interaction between benzene and toluene when disappearance rates in blood and in expired air were measured. Exposure was to 25 ppm benzene and TOO ppm toluene for two hours is three subjects. Tice et al (1982) also observed that the inhibition of SCS by toluene only occurred at higher benzene levels. An additional study sug gesting that lead, another common component of-gasoline, might interact with ben- 31 tiCD 00ooa355 cf dose-response affects in animals potentially extrapclatable to man. A fav of the sore important approaches Include the use of bone oarrev cell culture tech* nlques to obtain information about potential target cells of benzene toxicity. This is a major improvement over approaches measuring peripheral blood counts and bone sarrov cellularlty as an endpoint of benzene toxicity. Based on the data thus fart It is conceivable that careful dose-response studies evaluating laboratory animals, particularly sloe, exposed to benzene inhalation regimens approximating the current TL7 will demonstrate an effect on bone marrow hematopoietic precursor cells. The Implications of such findings would bear careful evaluation and would certainly be of concern to regulators and other decision maJcers. The ascertainment of a number of candidate besmtotoxic metabolites of benzene Is also potentially of importance. Evidence Is beginning to accumulate that in faot it may be sore than one metabolite that la responsible for the effects of benzene. It should be emphasized that studies of the tczlelty of benzene metabolites have thus far focussed predominately on endpoints related to aplasia rather than leukemia. Evidence that benzene does la faot cause cancer In laboratory animals has begun to accumulate. A model of benzene induced leukemia would be of particular value in answering questions concerning the shape and nature of the dose-response curve. The suggestion that benzene say cause non hematological neoplasia in animals, ether that of zymbal gland tumors, warrants a thorough reevaluatlon of the cancer sortallfiy experience of human cohorts in which benxene exposure is identifiable through the presence of a high incidence of hematological neoplasia. **Qb 0q Qq. q -33 be3l;cctappt Andrews, L.S., Eun,V.L., Witaer, C.M., Eocsis, J.J. and Snyder, 3. .Effects of toluene on the metabolism, disposition end hematopoietic toxicity of CJH]benzene. Biochea. Pharmacol. 26:293-300, 1977. Andrews, L.S., Sesame, 3.A. and Gillette, J.H. ^S-Benxene netabollsa in rabbit boss narrow. Life Sei. 25 : 567-572, 1979. Aksey, M. Mai 1 gnanclea due to oecupatlonal exposure to benzene. Saeoatologlca 65:370-373* 198aa. Aksoy, M. Different types of malignancies due to occupational exposure to benzene: A renew of recent observations la Turkey. Environ. Res. 23: 181-190, 1980b. Akaoy, M., Srdem, 5. Follow up study on the mortality and the development of leu kemia In 44 pancytopenia patients with chronic exposure to benzene. Blood 52: 285292, 1978. Aksoy, M., Srdem, S,,, Olncol, G., Leukemia in shoe-workers exposed chronically to benzene. Bleed 44:837-841, 1974. Alderson, M., Rushton, L. Mortality patterns in eight U.E. ell refineries. Aan. I. t. load. Sol. 381: 139*145, 1982. Baarsoa, S., Snyder, C.A., Green, J.D., Sellakumar, A., Goldstein, 3.0. and Alpert, R.2. The heaatotoxle effects of Inhaled benzene on peripheral blood, bone marrow, and spleen cells are increased by Ingested ethanol. Tox. Appl. Fharm. 64:393-404, 1982. Burrows, G.E. Health care of people at work. Screening of workers In research laboratories. J. Soe. Occup. Med. 30:164-168, 1980. Cronklte, E.P., Znoue, T., Carstea, A.L., Miller, M.D., Sullis, J.S. and Drew, R.T. Effects of benzene inhalation on murine plurlpetent stem cells. J. Tox. A Environ. 31th. 9:411-421, 1982. Dean, B.J. Genetic toxicology of benzene, toluene, xylenes and phenols. Mutation Ses. 47:75-97, 1978. BeCoufle, P., Cancer risks associated with employment in the leather and leather products industry. Arch* Environ. Health 34: 33-37, 1979. Deaton, J.E., Potter, G.D., Saatoludto, J.A. Effaces of lead on benzene secabollsa. J. Tox. A Environ. Hlth. 7:893-900, 1961. Dias, M., Reiser, A., Braler, L. and Dies, J. Studies on benzene sutagenesls. X. The mlcronucleus test. Experlsantia 36:297-299, I960* Delzell, 2., Moason, R.3., Mortality among rubber workers: 7. Processing workers. J. Occup. Med. 30:164-168, 1980. Esterline, P.E. Lymphomas and benzene. Lancet 2:1021, 1979* MCD 000008357 -34- Chem. Biol. Intar. 33:3*5-360, 1980. Greenlee, W.F., Sun, J.D. and Bun, J.S. A proposed mechanism of benzene toxicity: formation of reactive intermediates from polyphenol metabolites. Tox. Appl. Pharm. 59:187-195, 1981. Out, X., Satie, E.f Zizkova, L. Effect of phenobarbltal pretreatment on benzene biotransformation in the rat. Arch. Toxicol. *7:13-2*., 1981. Saak, H.L. and Specie, B. Inhibition of CFU- and BFO* by mononuclear peripheral blood cells during chronic benzene treatment u rabbiCa. Acta Baeoat. 67:27-33, 1982. Sakaaa, M., Eilplkarl, I. Canoer risk among rubber workers. J. toxic. Env. Health 6:1211-1218, 1980. Haley, T.J. Evaluation of the health effects of benzene inhalation. Clinical Toxichl. 11:531-5*8, 197T. Rani a, B.M., Talmege, M.H. Epidemiologic study of refinery and chemical plant workers. J. Oceup. Med. 24:203-212, 1982. Harden, 1. Eriksson, P., Leaner, P., Lusdgren, E. Malignant lympnoaa and exposure to chemicals especially organic solvents, ehlorophenels and phenoxy acids: a case control study. Sr. J. Cancer 43i 169-176, 1980. Herigaya, E., Miller,- E.M., Cronkite, E.P. and Drew, R.T. The deeeotion of In vivo bematotoxicity of benzene by in vitro liquid bone marrow cultures. Tox. Appl. Pharm. 60:348-353, 1981. Hite, M., Peobaro, M., Smith, I., Thornton, S. The effect of benzene in the mlcroaBdaua test. Mutation Res. 77:149-155, 1980. Infante, Sinsky, B.A., Wagoner, J.I., Toung, 3.J. Leukemia in benzene workers. Lancet 2:76-78, 1977. Irons, H.D., Heck, H.D., Moore, S.J. and Muirhead, E.A. Effects of short-term bentana adaiaistratioa on bone narrow cell cycle kinetics in the rat. Tox. Appl. Pharm. 51:399-409, 1979. Irons, H.D. and Moore, B.J. Effect of short term benzene administration on cir culating lymphocyte suhpopulatlons in the rabbit: evidence of a selective B-lym phocyte sensitivity. Sea. Come. Chem. Path. Pharm. 27:1*7-151, 1980. Irons, 3.0. and Meptun, O.A. Effects of the principal hydroxy-metabolites of ben zene on microtubule polymerization. Arch. Toxicol. *5:297-305, 1980. Half, G.7. r 3ushmore, T. and Snyder, R. Benzene inhibits RHA synthesis in mitochon dria from liver and bone marrow. Chem. Biol. Inter. *2:353-370, 1982. Eamlnska, 0., Jonek, J., Eamlnski, M., Eoehler, S. and Eonecki, J. The behavior of seme enzymes in the mouse liver due to chronic benzene intoxication. Acta. Sistocnem. Bd. 62:209-222, 1978. MCD 000008358 36-- dividuals occupationally exposed to benzene. Arch. Environ. Health33:3-10, 1978. Pagnotto, L.9., Elkins, a.3., Srugsch, H.C. Benzene exposure in the rubber coating industry - a follow-up. la. Ind. Hyg. Assoc. 1. UQ: 137-146, 1979. Pleclano, D. Cytogenic study of workers exposed to benzene. Environ. Has. 19:3338, 1979. Hao, G.S. and Pandy, S.P. Hepatic metabolism of heme in rats after exposure to ben zene, gasoline and kerosene. Arch. Toxicol. 44:313-317, 1980. Sickert, D.E., Baker, T.S., Bus, J.S., Barrow, C.S. and Irons, R.D. Benzene dis position la Che rat after exposure by inhalation. Tex. Appl. Pharmacol. 49:417* 423, 1979. Blnsky, B.A., Toung, 9.J., Smith, A.B.', Leukemia in benzene workers. Am. J. In dustr. Med. 2:217*245, 1981. 'Bosenstock, L. Leukemia and benzene. Annals Int. Med. 97:275*276, 1982. Bushtoa, L., Aldersen, M. The influence of occupation on health sose results from a study la the OX oil industry. Carcinogenesis 1:739*743, !980. Buahtos, L., Aldersen, M.8. A case*control study to investigate the association between exposure to benzene and deaths fpoa leukaemia in oil refinery workers. Br. J. Cancer 43:77-84, 1961. Sammetfi, 0.,Lee, S.V., Eoesls, J.J. and Snyder,.B. Partial bepatectoay reduces both metabolism and toxicity of benzene. J. Toxicol. Environ. Hlth. 5:785*, 1979* Sato, A. and Hakajlma, T. Dcae-depandant aetabolle interaction between benzene and toluene in vivo and in vitro. Toxicol. Appl. Pharmacol. 48:249-256, 1979. Salth, P.3., Llckiss, J.M. Benzene and lyaphooas. Lancet 1:719, 1980. Snyder, C.A., Goldstein, B.D., Sellakuoar, A., Bromberg, ., Laakin, S. and Albert, B.E. Toxicltyof chronic benzene inhalation: CD-I mica exposed to 300 ppm. Bull. Environs. Contaa. Toxicol. 29:385*391, 1982. Snyder, C.A., Green, J.9., Lofiue, J., Goldstein, B.9., 7alle, C.D. and Albert, R.E. Protracted benzene exposure causes a proliferation of myeloblasts and/or promyelocytes in CT-1 alee. Bull. Environm. Contaa. Toxicol. 27:17*22, 1981 (b). Snyder, C.A., Srlichnan, H.H., Laskin, S., Goldstein, 3.9. and Albert, 9.E. Pharzekoklnetica of repetitive benzene exposures at 300 ppm and 100 pps in AZS-mice and Sprague/Dawley rats. Tox. Appl. Pharm. 57:163*171, 1981 (al. Snyder, C.A., Goldstein, 3.9., Sellakuoar, A.B., Bromberg, I., Laskin, S. and Al bert, B.E. The inhalation toxicology of benzene: Incidence cf hematopoietic aeoplasas and heaaeotoxicity in AO/J and C57S1/6J mice. Toxicol. Appl. Phar isee!. 53:323-331, 1980. Snyder, C.A. t Goldstein, 3.D., Sellakuoar, A., Volaan, S.H., Bromberg, I., Erlichzan, M.B. and Laskin, S. Hesatotoxicity of inhaled benzene to sprague-Cawlsy rats 000008359 -38- 39:129-138, 1982. Oyeki, S.M., Aablear, A.L., Shoeaan, D.tf. and Sisal* T.U. Acute toxicity of benzene Inhalation to hemopoietic precursor cello. Toxicol. Appl. Pharmacol. 40:49-57, 197771 anna, H.J., Polan, A. Lymphomas and occupational benzene exposure. Lancet 1:1394-1395, 1979. White, M.C., Infante, P.F., Walker, B. Occupational expaore to benzene: 1 review of carcinogenic and related health effects following the O.S. Suprase Court Decision. An. J. Indust. Med. 1:233-243, 1980. Vlerda, 0., Irons, H.D. andQrenlee, W.F. Ivusotoxiclty in C57BL/6 nice exposed to benzene end arodor 1254. Toxicol. Appl. Pharmacol. 60:410-417* 1981. Vlerda, D. Greenlee, .7. and Irons, 8.0. Ismunotoxicity of benzans metabolites in G57BL/6 dee. Wildaan, J.M., Freedman, M.l., Boaman, J. and Goldstein, S.S. Benzene and lead in hibition of rabbit reticulocyte heme and protein syntheais: evidence for add!tire toxicity cf theae two components of ccsnereial gasoline. Bee. Tnamttn. Chen. Pathol. Pharmacol. 13:473-488, 1976. Volf, F.S., AndJelkoelch, 0., Smith A*, Tryroler, H. A case-control study of leu kemia is the O.S, rubber industry. J, Occup. Med. 23:103-108, 1981. MCD 000008360 -40