Erythropoietin-independent colonies of red blood cells and leukocytosis in a worker exposed to low levels of benzene.

Erythropoietin-independent colonies of red blood cells and leukocytosis in a worker exposed to low levels of benzene. Scand J Work Environ Health 1994;20:306--8. BACKGROUND - Exposure to high levels of benzene commonly results in the suppression of hemo poiesis, although cases of leukocytosis and leukocytosis with thrombocytosis have been reported. No hematologic abnormalities have generally been found with exposure to low levels of benzene. METHODS - A pipe fitter exposed to low levels of benzene (time-weighted average 0.9 ppm) devel oped leukocytosis. His blood counts and growth of erythroid burst forming units (BFU-E) was fol lowed with and without the addition of erythropoietin. RESULTS - Erythropoietin-independent BFU-E colonies were increased to 40 per 4 x 10 4 cells (nor mal < 3 per 4 x 10 4 cells). Both the leukocyte count and the number of erythropoietin-independent BFU-E colonies decreased when exposure to benzene was terminated. On reexposure the white blood count again increased. After the work was terminated, the white blood count returned to normal, as did the number of erythropoietin-independent BFU-E colonies. over a period of 12 months. CONCLUSIONS- Our findings suggest that even low levels of benzene can result in perturbations of the hemopoietic system. Further studies are warranted to determine whether these findings arc idio syncratic, coincidental, or a more general phenomenon.

Although exposure to benzene ge ne rally results in the suppressio n of hemopoiesis , ca se s of leukocytosis and leukocytosis with thromboc yt osis ha ve been reported (I). Ak so y et al in vesti gated 2 17 apparently healthy workers exposed to concentrations of benzen e between 30 and 210 ppm for 3 months to 17 years . Five workers were found to have isolated leukocytosis, and two had leu kocytosis with concomitant thrombocytosis ( I). On the other hand, Townsend & Fishbeck (2) studied 282 men, most having been exposed to less than 10 ppm for less than I year to over 20 years, and none of them were found to have either leukocytosi s or thrombocytosis. Still tho se exposed showed a trend towards a higher white blood cell count than the re ference group did. Fishbe ck et al (3) studied 10 workers ex pos ed to varying concentrations of benzene and found one worker with intermittent leukoc ytosis . Collins et al (4) studied 200 persons exposed to 0 .01-1.4 ppm; yet all but two workers were ex pos ed to less than 0. 2 ppm. They found no difference in the mean white blood cell values of those exposed to benzene and 268 unexposed workers at the sam e plant.
It ha s been demonstrated in vitro that the prolifer ation of er ythroid colonies from peripheral blood 306 and bone marrow of patients with myel oproli ferat ive disord er s occ urs without the additi on of exogen ou s erythro poietin (5-10) , ev en in th ose with normal total red blood cell mass. Th is phenom en on has been rel at ed to hypersen sitivity of the er ythro id ste m ce ll to erythropoi tin (8) and interleukin-3 (9). Mon ocytes ma y promote thi s independence ( 10) .
In the pre sent case study we investigated the ability of the per ipheral blood of a 43-y ear-old fitter with exp osure to benzene and leukocytosis to produce in vitro erythropoietin-independent burst-forming unit (BFU-E) colonies, as has been found in myeloprolifer ative di sorders.

Case
A 43 -ye ar-old titter wa s employed in maintenance wo rk at a terminal for the shipping of che mica ls in the Haifa harbor ov er a 10-year peri od . Hi s job required joining, cleaning, and rep airing pip elines. H is che mical exposure included benzene (3 % by vo lume) , toluene, xy lene, phenol, meth yl chloride, white spirit, acetone, eth ylacetate, amy l alco hol, and in or gan ic acid s. His personal le vel of ben zene exposure wa s found to be 0.9 ppm (time-weig hted avera ge ), and that of phenol was 0.1 ppm. Monitorin g test s are done in the harbor every three months and th e ti me-weighted averages for benzene in the are as where he worked ranged from 0. 3 to 1.2 ppm ov er the last year. During a routine annual examinat ion , he wa s found to have a white bl ood ce ll growth facto r for the growth of mouse keratinocyte s (13). This finding may be analogous to our finding of the erythropoietin-independent growth of red cell progenitor colonies. A rapid decrease in the incidence of leukemia when exposure to benzene is terminated has been reported and is consistent with the finding of a reversible step in the carcinogenic process ( 14). Finally , two cases of chroni c myelogen ous leukemi a have been reported among workers exposed to high levels of benzene (14). Our results should be interpreted with caution. The fitter was exposed to many other chemicals, and our findin gs could have been a result of exposure to one of these chemicals or merely coincid ental. Still perturbati ons in the hemopoietic system have not been reported with the other chemicals. There are many causes of leukocytosis, such as mild chronic infections, inflammatory diseases, and allergic reac tions. Our patient however was asymptomatic. The work challenge test was positi ve, and no leukocytosis was found durin g the follow-up after the exposure was termin ated. There fore we believe that benzen e was likely to be responsible for the effects noted. Further studies of workers with and without leukocytosis and exposed to low levels of benzen e are warranted in order to substantiate our finding s. Animal studies and epidemiologic studies have suggested that the "no effect" exposure level for the hematologic effects of benzene in humans is 25 ppm, the timeweighted average suggested by the American Conference of Governmental Industrial Hygienists (4). In a study of workers exposed chronically to levels under I ppm with occasional levels over 100 ppm, no effects were found on hematologic or biochemical parameter s, proliferati ve rate indices, sister chromatid exchanges, and urine mutagenicit y, except for a small but significant increa se in mean corpuscular volume (15). Such tests may, however, lack sensitivity.
lt is uncertain whether our findin gs were idiosyncratic or a general phenomenon related to the devel-count of IS 100 . mrrr", with a norm al differential count.
Repeated testing revealed a consistentl y elevated count (figure I ). The leukocyte alkaline phosphatase score was ele vated on two separate occa sions (230 and 300 with a control of 58, N < 100). In a urine collection the phenol level was found to be 36 mg .1-1 •

Methods
Peripheral blood was collected with preservative-free heparin. Low-density mononuclear cells were isolated by fractionation on Ficoll-Hyp aque density gradient , washed with phosphate-buffered saline, and resuspended in Iscove medium supplemented with 10% fetal calf serum. BFU-E colonies were grown from peripheral blood according to a microwell modification of the methylcellulose technique (10). Lowdensity mononuclear cells (4 x 10 4 cells per ml) were suspended in Iscove's modified Dulbe cco' s medium (Biological Industries, Kibbutz Beth Haemek, Israel) with 0.8% methylcellulose, 10% fetal calf serum , 0.01% 5-thi oglycerol, and a penicillin-streptomycin solution. BFU-E colonies were scored after 14 d of incubation in a fully humidified atmosphere with 5% carbon dioxide at 37°C. The colonies with at least three cluster s of 50 cell s were defined as BFU-E . Selected colonies were stained with benzidine for the ident ification of their erythroid lineage; otherwise they were identified by their reddi sh color.

Discussion
Leukem ogenesis is thought to be a multistep process. Whether the findings from our patient repre sent one reversible step in the carcinogenic process is uncertain, however. Benzene has been shown to activate protein kina se C both in vitro and in intact platelets ( I I) . Protein kinase C plays a central role in signal transduct ion and is thought to mediate seve ral tumor promoters ( 12). Introducing oncoge nicall y activated receptor kinases has been shown to alleviate completely the requirements for epidermal

Results
The number of erythropoietin-independent BFU-E colonies was increa sed to 40 per 4 x 10 4 cells (normal < 3 per 4 x 10 4 ) . When the exposure to benzene was terminated, the BFU-E (erythropoietin-independent) decreased to 20 and then to 7 and the leukoc yte counts returned to normal. After reexposure, the patient' s leukocyte count again increased to over 18 000. He has subsequently stopped working and stopped smoking, and his leuko cyte count has remained normal over a year of follow-up. The erythropoieti n-independent BFU-E colonies decreased to 2 per 4 x 10 4 • The leukocyte alkalin e phosphatase scores also returned to normal. opment of malignancy. We are planning a study of progenitor cells in a cohort of workers exposed to low levels of benzene. Since the abnormalities found in our patient were reversible, it is unlikely that he is at increased risk for developing leukemia. Still he is possibly more sensitive to the effects of benzene than other workers, and we have recommended that he not be returned to work in areas where there are even low exposure levels of benzene.