Document 51rD8k0kv5droRr0zn1BOJB5

INTER-0FFICE CORRESPONDENCE DATE February 28, 1995 -------------tl~D. Polle:t .-~'-L,', .L-- ]0/ s.:.JWECT: N.a..tJ.onal Car:ctor Ins t:i t;ut;e (NCl) Trip H. R. fricdlnndcr Report (2/13/95) Tom Armstrong, Michael Bird, Steven Phillips, Rob Schnatter and I ~et with members of the National Cancer Institute (NCl) on February 13th, 1995. The purpose of the meeting was to discuss with NC! i-nvestigators their studies of workers exposed to benzene in the Peoples Republic of China and for us to describe some of the work with which Exxon is involved in Canada, the UK and Australia. We .are monitoring the NCI studies because of their potential impact concerni-ng health risks at low benzene exposures and their interaction-with key activities currently underWay within and outside Exxon. !hsse activtties include the IP Epidemiology Study, the European Union Benzene Risk Assessment and the lOL Worker Study. The EBSI te~ summarized our work on benzene using ~aterials previously cleared for 4elease. The NCT epidemiology efforts consist of a large coho~t study-followed by a "nested case-control" assessment. The cohort study compares the health experience over time of' a large group of workers believed to have had substantial exposure to ben:zene to a set of control~; whose history and available work records.suggest an absence of -occupational benzene e~posure. The case-control study focuses on i~dividuals with or without the disorders of concern (e.g., certain leukemias, aplastic anemia, etc.) and assesses the relationship of having those disorders with specific expo~ure attributes (e.g., concentrations, durations, and frequency of benzene exposure). The cohort study is very large and complicated with over 75,000 benzene exposed workers and about 36,000 non- benzene cont~ols. l'he st:udy coven; 12 Chinese cities, nearly 700 factories, multiple job ca~egories and several industries. Petroleum industry workers are not significantly represented in the study. Painters, chemical manufacturing, and rubber industry ~orkers appeared to comprise the largest of the groups studied. The benzene exposure concentrations found were quite high. In the 1950~ and 1960s, jobs such as rubber gluers and drip painters were reportea to bave h-ourly exposures so111eti-mes exceeding 100 .ppm. Average daily. exposures exceeded 50 ppm for those workers and between 10 and 50 ppm for most others exposed, The exposure data" are derived fr_om a mix of c.ct:ual mea.surement data in about 40% of ca-tegories, and the residual 60l: are based on best: ~strial hygiene juligl:'ll!"ent:s. Lict:le information was .provided at the mee.ting as to the quality or validation of tha exp<>sur.e dat:a. While t:heir studies are scill in early stages, some initial results have been reported from t:he cohort study. In addition to incre-ases in acute myelogenous leukemia (a known target site), excesses were al.so not:ed for several other diseases of the blood-formin-g system such as non-Hodgkins lymphoma, lymphoid leukemia, and aplastic anemia. Also, there appears to be an increase in some solid tumors. The ease-control studies may provide further insight: into the.se results, scme of which (e.g., sGlid t:umors. lympho~as, etc.) appear co conflict with other published findings. BP-00010562 .._,,._, ,._._ l....r"'l- ~- - - 2. - P.03 Altho~..:gh motivated investigators from the NCI, the Chinese Academy of Preventive Medicine, and consulting organizations- are part:icipat:ing in t:he research, several important: mer:hodologtcsl issues could ~nfluence s t:udy qu.ality and conclusions: l. Accuracy of benzene exposur:~ data - Kuch of the exposure data ha::; bee:--. estiwa.t:ed based on "best judgements". Qual!..ty assurance of nnalycical measurements needs verification. We infer from our discussions with the NCI scientists that benzene exposures experienced by the ~orkers in the previous decades may have been even higher than those repor~ed. 2. Assessment of otha:r expoaures - Exposure to other agents, including carcinogens, needs more detailed consideration. Workers likely had significant exposures to other chemical and physical agents. The study impact of such exposures and life style factors such as smoking, diet, and dwellings (radiation, asbestos, and PAH issues) need further assessment to more fully define the role of benzene in the reported disorders. 3. Case ascertainment and verification- At least one incentive e~ists for more ..complete repoi'ting (or over-reporting) of leukemia in benzene workers than in other members of the population, If leukemia occurs in benzene-exposed workers, compensacion payments are made to the family. Thus, a potential bias exists for leukemia to be reported preferan- tially for benzene -workers compared to study "controls". Hematologic malignancies were verified by -c-linica~. laboratory and pathologic data review, thus making this study's results incomparable to most ether published studies. 4. Accuracy of diagnose-s of solid twnors - Many of the solid lung tumors reported were based orily on chest x-ray results. Without a confirmation from biopsy, autopsy or cytolo~y, tuberculosis (a common disorder in China) could often be misdiagnosed as lung cancer. The potential for this issue to create misleading results needs to be assessed. In summary. the NCI China program is the world's largest health study of heavily exposed benzene workers. There are many issues surro~nding the study, e.g., accuracy and completeness of expo~urs information, reliability of disease reporting, effect of lifestyle factors, etc. which may confound interpretation of results. However, a key point is that the reported health findings to date are in workers exposed to high- concentrations of ~enzene. The NCI studies do not: directly addr-ess the low dose exposure region of current regulatory and risk ~ssessment interest. However, because future interpretations or finaings could imply lower dose effects and impa~t policy. it is important that we continue to monitor these studies. NCI will present some of it:s results in the June Benzene Conference at Rutgers Univel;'s-!cy and anticipates publicat:ions in late 1995 or early 1996. Additional details are available ~pon requesc. BRF:jmk cc: T. ~-Armstrong, J. G. Bell. H. G. Bird, J. L. Cox, G. F. Egan, A.M. Goldstein, M. Harrison, D. R. Hudgins, A. H. Katz, T. J. McDonagh, R. D. Phillips, S. C. Phillips, J. T. Sanderson, A. R. Schnatter, F. R. Smith, G. V. ~ant, C. H. Yarborough BP-00010563