The Effects of Benzene on Human Health

17. Evidence of the harmfulness of benzene comes both from studies of human populations exposed to the chemical in their workplace and also from observations made in studies in laboratory rats and mice. The Panel have put greatest weight on the studies of human populations, and it is these that we discuss in most detail here, since it is difficult to extrapolate from results in animals to effects on people. Having said this, there are also problems in extrapolating from the available studies in man to possible effects on the general population, since these studies have been carried out on workers exposed to benzene at concentrations considerably higher than those found in ambient air. As an illustration, a worker exposed to 1000 ppb (one fifth of the United Kingdom workplace exposure limit) may take in as much as 10,000 µg of benzene per day, compared to the 650 µg that someone living in an inner city environment might take in. It is also true that the estimates of past exposure of workers in published studies are only approximations to the actual exposures which occurred over that period; in many cases the exposures are likely to have been underestimated, thus probably overestimating the likely risks to workers of exposure to a given concentration.
18. In spite of these reservations, it has proved possible to draw some broad conclusions from the published literature. Benzene is readily absorbed into the body when breathed into the lungs, about half of it being retained. As it is more soluble in fat than in water, it is distributed in the body to fatty tissues including the brain and the bone marrow where blood cells are made. In the absence of further exposure, the benzene is eliminated by chemical breakdown in the body or by metabolite excretion in the urine, 80% being eliminated within about two days.
19. The distribution to fatty tissue is reflected in benzene's toxic effects. Acute exposure to extremely high concentrations may cause narcotic, or anaesthetic, effects and deaths of workers have been recorded after exposures to concentrations of several thousand parts per million (or several million ppb) in confined spaces. Very high levels of exposure (well over 5000 ppb) on repeated occasions have led to the development of severe and sometimes fatal damage to the blood-forming elements of the bone marrow, causing it to be unable to manufacture essential blood cells. Such serious consequences are not, of course, a risk attendant upon exposure to the concentrations of benzene observed in ambient air and will not occur in workers except as a result of unforeseen and accidental exposure to very high concentrations.
20. The effect of long-term exposure to benzene which is of most concern, however, is leukaemia and in particular several types of this disease known collectively as the non-lymphocytic leukaemias³. Leukaemia was first described among workers, exposed to very high concentrations such as those that may cause other bone marrow effects, employed, for example, in the shoe manufacturing industry in Italy and Turkey. Thereafter, a series of studies of groups of workers in the synthetic rubber and petroleum industries, in which generally much lower exposures to benzene had occurred, has been carried out, and these have confirmed the increase in risk of non-lymphocytic leukaemias in some of these workers.
21. Moreover, exposures of rats and mice to inhaled concentrations of 10,000 ppb or more for most of their short lifetimes have shown that benzene induces similar effects in these animals, as well as causing other types of malignant disease. These types of experiments, and others, on cells, designed to determine the mechanisms of benzene's action have shown that it acts on the genetic material of the cells. Such a genotoxic action of a chemical is generally taken to indicate that the possibility of it causing malignant disease exists even with very small exposures. While this could be strictly interpreted as meaning that there is no safe concentration to which people can be exposed, a realistic view is that the risks become increasingly small as the cumulative exposure of an individual is reduced and that, for all practical purposes, there is a concentration at which the risks are exceedingly small and unlikely to be detectable by any practicable method. This is the view that the Panel take.
22. Further problems in extrapolating from risks in groups of industrial workers to the general population relate to the size and make-up of the industrial group studied. In the studies reported in the medical literature, the industrial groups (often called 'cohorts', that is defined groups of workers followed forward over time to estimate their risks of death from various diseases) have varied in size from about 750 to 4600. Since in normal circumstances leukaemia is a relatively uncommon condition, occurring in about 1 person in 6000, it is often difficult to exclude the possibility that small excesses of cases of the disease are simply chance occurrences. Thus, the smaller the cohort, the stronger a leukaemia-causing effect has to be in order for it to be detected.
23. A final point the Panel have considered in recommending a Standard for benzene concerns the differences between past industrial cohorts and the general population. The former generally comprised fit young and middle aged males, whereas the general population of course includes also children, pregnant women, the elderly and the sick, some of whom may be unduly sensitive to toxic chemicals. The industrial cohort is potentially exposed to high concentrations for perhaps 8 hours per day, five days weekly for 40 years, whereas the general population is exposed to much lower concentrations, but throughout their lifetimes.

Table 3 Leukaemia deaths in workers occupationally exposed to benzene

Authors	Study	Number of Subjects	No of Deaths	Exposure Estimates (ppb-years unless otherwise indicated)	SMR†	(95% CI)‡
Rinsky et al 1987	Goodyear Pliofilm Plant workers employed in a department where benzene was used for at least one day during 1940-1965. Deaths occurring after 1950 counted.	1165	2	1-39,990	109	(12-394)
			2	40,000-199,990	322	(36-1165)
			2	200,000-399,990	1186	(133-4285)
			3	>400,000	6637	(1334-19,393)
Wong 1987	Chemical Workers exposed to benzene for at least 6 months between 1946 and 1975 (Only the cumulative exposure group is included here)	3536	5*	500 Estimates for 30 year equivalent levels of exposure	91*	(29-213)
			5*	2000 Estimates for 30 year equivalent levels of exposure	147*	(48-343)
			5*	>2000 Estimates for 30 year equivalent levels of exposure	175*	(57-409)
Bond et al 1986	Dow Chemical Workers exposed in organic and resin synthesis (excludes confounding exposures)	888	3	1500 Estimates for individual cases	162	(33-461)
				25,400 Estimates for individual cases
				351,000 Estimates for individual cases
Hurley et al 1991	Coke Oven and other Coal Product Workers (National Smokeless Fuels) (British Steel Corporation)	3812	3	Estimated exposure concentrations 310-1320 ppb**	42	(9-123)
		2708	2		41	(5-147)
Yin et al 1987	Chinese Factory Workers exposed to benzene in painting, shoe-making, rubber synthesis, leather, and adhesive and organic synthesis factories	15,643	17	Concentrations where leukaemias occurred: (2000-345,000 ppb)**	501	(292-802)
		12,187	8		830	(358-1635)
* All lymphopoietic and haematopoietic cancers.
** No data on duration of exposure given.
† SMR: Standardised Mortality Ratio, the ratio of the observed number of deaths in the occupationally exposed cohort to the expected number of deaths in a control group of this size of unexposed persons, multiplied by 100. For control groups the studies have used either the national populations or workers in other industries.
‡ CI: Confidence Interval, the 95% CI is the range in which, allowing for variability in study populations, there is a 95% chance of the true result falling. If the value 100 is not included in the 95% CI shown in the Table then there is only a 1 in 40 chance that the increase in leukaemia deaths in the benzene-exposed group is an erroneous result (p<0.025).

24. The Panel examined published reports of studies of leukaemia in a number of cohorts of workers exposed to benzene. While none provided wholly reliable information on the exposures of the workers, several included estimates that were felt to be reasonable. The most important of these are summarised in Table 3. The two studies considered most useful were those of a cohort of workers employed in an American rubber manufacturing plant between 1940 and 1965, and of a larger cohort of American chemical workers exposed to benzene between 1946 and 1975. Both studies showed an increased risk of non-lymphocytic leukaemias in workers with the highest exposures, estimated to have been greater than 200 parts per million (ppm) years (that is, equivalent to 10 ppm or 10,000 ppb for 20 years) in the former study. However, excess numbers of cases of leukaemia occurred at exposure to lower concentrations among workers in both cohorts (although the possibility that these were due to chance could not be ruled out) and in both cohorts there was evidence of a statistical relationship between exposure to benzene and likelihood of developing leukaemia. In examining these data, and taking account of the uncertainties discussed above, the Panel concluded that the risk of leukaemia in workers was not detectable when average exposures over a working lifetime were around 500 ppb.
25. The Panel are aware that the risk of non-lymphocytic leukaemia, in the form of acute myeloid leukaemia, is increased substantially in cigarette smokers, to the extent that the risk of this disease is almost doubled in those who smoke 20 cigarettes daily. If this effect were due to benzene, the studies of industrial cohorts would have been expected to have shown far more cases amongst the exposed workers. The Panel therefore conclude that this effect of smoking is not related to benzene alone, but to the mixture of carcinogens found in cigarette smoke.