Document 3Q0Jpa4gvkvVjvbjJazg3X2x6

DownloadSearchable PDFRandom document
`I OF T H E N A T I 0iVA L C A N C E R I N S T I T U T E In This Issue : 987 EDLTORMLS Promise and Problems of Bcl-2 Antisense Therapy, J. C. Reed Ductal Carcinoma In Situ of the Brean: a New Model, D. F.HaFes 988 991 NEWS Are Changes in End-of-Life Cancer Care Coming? IOM,ASCO Tackle Issues, C. McNeil Virtual Reality Researchen Seek to Minimize Invasive Diapostic Procedures, H. Heimer Keying Educational Messages to High-Risk Populations, L Doley Star Bite: US. Cigar Consumption, 193-1996, D. R. Shopland Kiwi Conundrum: Screening for Prosrate Cancer, B. K u s h Common Ground Found: "Overall h c e r Mortality Rates Are TLllling," K. J. Smeirz Awards, Appointments, Announcements 994 997 998 999 loo0 1001 1002 COI\1IMENTARY REVIEW A Science for the A r t of Consensus. D. L Sockerr __ Tocopherols and the Etiology of Colon Cancer, W. L Stone, A. M 'upas 1003 .. . 1006 SPECIAL ARTICLE National Institutes of Health Consenms Development Conference Statement: Br&t Cancer IO15 ,. - . . .. Screening for Women Ages 4049,Jannary 21-23,1997, NarionaI .`,mitures of Health Consensus Development Panel * ARTICLES .. . Induction of ApoptosiS in S d - C e l l L m g Cancer CeUs by an Anti mse Oligodeoxynudeotide Targeting the Bcf-2 Coding Sequence. .LZiegier, G.H. Luedke, D.Fubbro, K.4.Almrann, R. A. Srahel. U.Zangemeister-W i n k 1027 Characterization of p53 and pZ1 Functional Intemtfions in Glioma Cells en Route to Apoptmis, 1036 C. Gome:-rMan;ano, 1. Fwyo. A. P.K!ritsis, T.J. McDonnell. P. .4. Steck, et al. Association of Tumor Angiogenesis With Bone ~ M m ~wMicrometaz:;sesin Breast Cancer Patients, S.B. Fox R. D.Leek J. Bliss, J. L Mansi, B. Gusrerson, K.C. Garrer, A. L Ham`s 1044 Antitumor Efficacy of Adenocarcinoau Cells Engineered to Prodxe Interleukin 12 (n-12) 1049 or Other Cytokines Compared With Exogenous IL-12,F. Caallo, P. SignoreNi, M. Giovarelli. P. Musiani. A. Modesti, :M. J. BnuzC; M. P.Colombo, G.Forni REPORTS BRIEF COMbIUNICA'I'ION BOOK REVIEW Assessment of Hormone Dependence of Comedo Ductal Carcinornn In Situ of the Breast P. A. Holland W. F. Knox. C.S. Pozm,A. Howell, E. Anderson, .A. D. Baildam, N. J. Bundred Benzene and the Dose-Related Incidence of Hematologic Neoplas;:~in~ China, R. B. Hayes, S.-A? Yin, IM. Doserneci, G.-LLi. S.xucholder, L B. Travis, C.- k'. Li,IV. Rothman, er al. _- Aspartame Consumption in Relation to Childhood Bmin Tumor ?isk: Results From a Case-Control Study, J. G. Gurney, 1..ti. Pogoda, E. A. Holly, S.S. fil,chr. 5. Preston-Manin ~-~ ~~~~ The Cure of Childhood Leukemia: Into the Age of Sliracles, R. C Ribeiro Books Received 1059 1065 1072 1075 1075 (Conrenit continurd on back cover) rn Visit the Journal's World Wide Web sue at http://www.oup.co.uL-,;t1ci/ rn . i .- -.i in cYcinama in situ [if rhc brcast. Cancer 1992:69:1174-81. the HR for all hematologic neoplasms combined was 22 (95% confidence in- terval [CII = 1.1421, and, for the com- -tor in ducui wcinoma in kitu or {he brait. Incidence of Hematologic Surgery 1993113:134-37. 117) Hellman 1. Harris JR. Caiicllos GP.Cancer of Neoplasms in China bination of acute nonlvmphocytic leukemia and related rnyelodysplastic syndromes, the RR was 3.2 (95% CI = 1.0-10.1). For individuals who were oc- cupationally esposed to benzene at con- stant levels of 25 ppm or more, the RR for the combination of acute nonlym- ,Votes Supponcd by a grant from *he i\ssociation for htemationll Cancer Research. C.S. Potten and A. Howell arc funded by the C u . u Rerearth CamTaign. Tratqis, Chin- Yurig Li,iVatitatiie1 Rorhman, Robert :V. Hooi.er, Martlia S. Liizet * For the Chinese Acndem?. of phocytic leukemia and related myelo- dysplastic syndromes was 7.1 (95% CI = 1.1-23.7). 1Yorker-s with 1 0 o r more years of benzene exposure had an RR of developing non-Hodgkin's lym- phoma of 4.2 1 9 5 5 CI = 1.1-15.9), and We thank MI%.Marjorie Ga~ieIlfor typing the manuscrip Miss Lesley Shaw for technical assis- ~ C C .and Dr. E. 9. Faragher for statistical assis- -1IIC.S. h n d p t received October IO. 1996; revised .+MI 24, 1997; accepted May 3, 1997. Preventive J4ediciize-iVarional Cancer Institute Beii,-ene Study Group the development of this neoplasm was linked most strongly to exposure that had occurred at least 10 years before diagnosis (Le., distant exposure) (Pfor trend = .005. two-sided). In c a n t r a t , ... Background: Benzene is a widely distributed environmental contaminant known to cause leukemia, particularly acute nonlymphocFtic leukemia, and the risk for the combination of acute nonlymphocytic leukemia and related myelodysplastic s?ndromes was significantly increased among those with more recent benzene exposure (P for perhaps other hematologic neoplasms 'trend = .003, two-sided), but it was not and disorders. Few epidemiologic stud- linked to distant exposure (Pfor trend ies, however, have been able to address = 3 1 , two-sided). Conclusions: The re- relationships between the extent of ben- sults of this stud?- suggest that benzene zene exposure and the level of risk. Pur- exposure is associated with a spectrum pose: A large cohort study was c a m e d of hematologic neoplasms and related out in China to evaluate the risks of disorders in humans. Risks for these developing specific hematologic neo- conditions are ele\ ated at average ben- plasms and selected related disorders zene-exposure levels of less than 10 in relationship to quantitative estimates ppm and show a tendency. although not of occupational benzene exposure. a strong one, to rise with increasing lev- Methods: A cohort of 74828 benzene- els of exposure. The temporal pattern exposed and 35805 unexposed workers of benzene exposure appears to be im- employed from 1971 through 1987 in portant in determining the risk of de- 12 cities in China was identified and veloping specific diseases. (J Natl Can- followed to determine the incidence of cer Inst 1997;89: 1063-711 hematologic neoplasms and related dis- orders. Estimates of benzene exposure were derived from work histories and available historic benzene measure- ments. Existing pathologic material and supporting medical records were reviewed to establish diagnoses of disease. Relative risks (RRs) (i.e.. ratios of incidence rates for specific hematologic neoplasms and related disorders in the *.-lft7/iuti~irrsof arcfiirc. R. B. Hayes. M. Dosemeci. S. Wxholder. L. B. TrJvis. N.Rothman. R. N. Hoover. Sf.S. Liner. D:..:.cion of Cancer Epidemiology 2nd Generic<. S x i n d Cmcer Insriturc. Bzrhesda. MD: S.-S. Y::. G.-L. Li. Chiiie>e ACJ& Journal of the National Cancer ~ns~iruteV.ol. 89. No. 14. July 16. 1997 REPORTS IOhS Benzene is a widespread contaminant in the air and groundwater. originating from industrial sources, cigarette smoke. gmoline. and automobile emissions ( I 1. There is scientific agreement that benzene causes leukemia. particularly acute nonlymphocytic leukemia (ANLL). but controversy still exists with regard to the level of risk at low exposures (2-4) as well as the potential association with other hematologic neoplasms (5). The epidemiologic evidence for a doseresponse relationship between benzene and leukemia derives primarily from three small cohort studies (6-8)and their recent updates (9-1I ) . These investigxions have generally reported on workers exposed to benzene at relatively high levels, or they have identified such small numbers of leukemia cases as to preclude stable exposure-response estimates. In 1981, the Chinese Academy of Preventive Medicine (CAPM) cmied out a national occupational survey and identified more than 500000 benzex-exposed workers, characterized by a Sroad range of benzene exposure levels (12). A subsequent follow-up survey of IS 460 benzene-exposed and 28 257 unexposed workers from 1972 through 1981 found an increased risk of mortality due to leukemia (standardized mortality ntio [SMR] = 5.7). but the exposure-responserelationship was not quantified (13). Since 1987. the U.S. National Cancer Institute has collaborated with the CM.M and its team in 12 Chmese cities to identify all incident cases of hematolo,o' ic neoplasms and related disorders in an expanded study population with 6 additional years of follow-up. A major focus of the collaborative effort was to evaluate quantitative data with regard to benzene exposure. Here. we assess the exposureresponse relationship between the extent of benzene exposure and the risk for hematologic neoplasms and selccxd related disorders. Materials and Methods Study Population and Exposure Assessment The benzene-exporcd group con\i>:cd of 74 323 worker.; emph,yed from 1972 through 1987 in 673 factories in I2 citieb in China. A varie::. of indubtries and occupations uhin;: benrenr: was :i;:Jied. including painting. printing. and the m a n u f s u r e of fwtw a r . point. and other chemicals. A:: unexposed ci~mparisongroup was usembltd from .workersernploycd during the same period in w ) r i units where bci:rciic w;ih : T I ~ I ~ i w din 69 of these factories and in 40 ;1JditiiiiiaI ".i<iorichIcxated in the 12 study cities. n i c iiiicxpo.~.! $roup consisted of 35 805 workers. niix inve>tig.::#linu a s approved by the CAPkl Insiitutioiial Rc,.~i'uBoard in accordance with an &-smiice tiled L i i l i the U.S. Depanment of Health and Iluiiian Scr\ iccq. Subjects were identified thriiugli l'ictof: ndinililhtmtive records. from which dr:iitiigr:iphic .:d job hirtory h t a Cjob title snd dates ol'ciiiploynic:::I ucre abstncted ( /-I./S). A w n g e wcuparicinal :\powre (time-weighted avenge) to bcwcne Iin r;:::+ o i < I p m per million [ppml. I to 4. 5 to <IO. !O to <25. 25 to 40.and 350 ppm) was estimatcJ by local industrial hygienisrs and other ixcup;it!$.nalhealth personnel. using available anihient benrL::e exposure measurements and detailed pmcluc::on and related process information. for \even cx!:ndar periods (1949 through 1959, 1960 througii lY64. 1963 through 1969. 1970 through 197-1. 1975 through 1979. 1980 through IYU.and 1%:-I for study-specific job titles in each factory. Work 5istories were then linked to the benzene-exposur: data to provide individual timespccitic benz;.N*c.-exposure estimates. Subject Fo!!o\v-up With the UX of occupational personnel and health records and. needed. contacts with physicians. colleagues. ne-.[ of h n . or others. subjects were followed through December 31. 1987. for the occurrence of hem;.:ologic neoplasms and other selected hematologic Ji-ordcrs as well as for vital status and cause of death For cases of hematologic neoplasms and other s&-!ed hematologic disorders. available patholosic m.:::rial and medicd records were reviewed by exp;.~L'.S. and Chinese hematopathologists (16). Bccaube myelodysplastic syndromes (MDS) (17) m y be precursors to AVLL (18) and were nor syste::i.itically distinguished from ;LULL in past epidemiclopic studies. ANLL and .MDS were combined for .cveral analyses. On averag:. benzene-exposed and unexposed subjects werc rolloued for 10.5 years and 1 I.? yeus. respecri..c'ly. Most person-yen of follow-up occurred duricg active employment (SSQ among benzene-expoc-.d and 9 I 9 among unexposed workers): only a si-sll percentage of person-yean were atuibutable to irorkers who had moved from the study cities ( 2 . 5 5 among benzene-exposed and 2.49 among .inexposed workers). Approximately 2% of the st!:d) subjects died during follow-up (1369 benzer.: :\posed and 598 unexpoxdl. Only I47 ( 0 . 2 9 1e x ; v w d and YO (0.35)unexposed work- ers were lost : 3 fnilow-up. Statistical .?nalysis Subjects en::!(T!ed less than 6 months and those hired before :I..:exposure assessment period (1949 through 1987, ' I .fct of e x p o d md O.Y% of unex- posed) were :\<luded from the analysis. Personyears at risk *.\?:: dccutnulated from 6 months ifier study entrl, iJ :lux! I . 1972. or. if hired later. the date of hire) : :hc csrliest of the following events: date or'diagnl - I \ 111' uxnccr or ;1 related Jihcise. due of death. datc . I ( !in( employment in a job ol' un- knou n cxpuri.-: ( <1'i 01'perstin-yemi. Jur: i)f 105s to follow-up [.II '; of whjccts). or end of follow-up (December 3 I . 1OS7t. Subjects were c a t e g o r i d by occupation.; iii Ltutingr applications (iiiostly painters). chemic!. pnlduciion. shoe priiduction. and .- other or mixcd wcupatitins (i.c., mow tlrall oiie 0 1 the above categories). To provide stable ri3k estimates. hirly hrti;id p u p s were dctinrtf for domion of e x p t i ~ u(~<j. 5-9. and 310 YeXiJ. average (<IO. 10-24.and 32.5 ppm). and cuinulative expcxurc (CUI.UI-ltM). and ,100 ppiii-ycmi to k n x n e . Each measun' ol knzene e x p c ~ s u rf~or each individual wiu: aI111~cdi t ~ change with time: penon-years and dixea.w even[\ were ;Issigned to benzene+\porure levels with .I I.p y e u la;: lie.. according to the level 1.5 y e m prcviously). A lag period of 1.5 years was used in inaking these calculations because more recent CAPOsures are unlikely to be biologically linked to the development of diagnosed cancers (19,201. To evaluate the rernponl component of disease development funher. we pmitioned the cumulative CSFUsure at 3 given time into recent (1.5 to IO !ex. earlier, and distant (10 or more years earlier) e s p ~ i sure. Because 3 substantial number of individu.d. were historically exposed to a relatively ctable a\ ~ e n g e m o u n t of benzene while working. we also developed a measure of constant exposure. where follow-up was censored 1.5 y e n after the individual's exposure level changed (<IO. 10-24.and 215 ppm) for the first time. Ratios of incidence rxes for hematologic neoplasms and related disorders in [he benzene-exposed group compared with ntes in the unexposed group (RRs = relative risks) were determined using Poisson regression analysis (2/). \ k i t h smtification by age and ses. Tests for l i n e v trend (two-sided) of increasing risk with increasing ester11 of benzene exposure were carried out. on the basis or the mean of years in durationexposure cartigories. the log of the mean of parts per million in averageexposure categories. and the log of the m e a of pans per million-yean in cumulative exposure categories. Results The RR for all hematologic neoplasms combined among workers who were occupationally exposed to benzene was 1.6 (95% confidence interval [CI]= 1.14.7) (Table 1). Statisticdly significant excesses were found for XNLL (RR = 3.0: 95% CI = 1.0-8.9) and for the combined group of ANLUMDS (RR = 4.1: 9 5 5 CI = 1.4-1 1.6) (Table 1). Excesses were also noted for other leukemias (RR = 2.0; 95% CI = 0.7-3.4) and non- Hodgkin's lymphoma (N.HL) (RR = 3.0: 95% CI = 0.9-10.5), but the results were not statistically significant. Increased risk was found for hematolosic neoplasms among workers hired before 1971 (the beginning of the cancer risk assessment period) (RR = 2.9: 9 5 4 CI = 1.5-5.4) and thereafter (RR = 2.5: 95% CI = I . 1-5.4). Fifteen of the 16 ben- zene-exposed YHL cases occurred among workers hired before 1977, (RR = 4.1: 95% CI = 1.2-14.4). For AYLL. signifi- cantly increased risk was found for work- ers hired after 1972 (RR = 5.1; 95% CI = l.S-l7.Zj,while the risk for ANLU 1066 REPORTS Journal of the National Cancer Institute. Vol. 89. No. 14, July 16. 199: ,.+.. . - . . . .. . . . ...-.-.. - i . . > , , .. ,_ ,, - -. .. . ---&- --...-.- ..- _- Table 1. Rclative risk for hematologic neoplasms and related cnnditions. on rhe basis of selected occupational :.h;iractcristics, for workers exposed to benrrne Study gmup All exposed subjects Pcnonyeus. x 10' 698 Mc~ exposure. Mean ppm*.t exposure. y* 22.5 9.3 ExposcJ subjects: y of hire e1972 3 1972 Exposed subjects: occupation Coaiingsil Rubber Chemical Shoe Otherhixed All unexposed subjects (referentfl 404 194 350 34 88 68 I58 405 24.9 19.2 21.5 53.5 24.8 21.8 17.4 0 13.2 4.0 9.4 11.1 8.5 7.9 9.8 0 All hemtologic neoplasm$ 2.6 158) 959C1J = I.U.7 Relative r1.L (No. of cscs) NHL 3.0(16) 95QCl = 0.9- 10.5 Leukrnim 2.5(3) 958CI = 1.2-5. I ANLL 3.0(21) 958C1 = I .0-8.9 ANLLMDS ~~ 4.1 ( 2 8 ) 958CI = 1.4-1 1.6 Other 1euliemta.s 2.0(17) 95ciCI = 0.7-5.4 2.9 (ai 2.5 114) 4.1 (15) 0.5 ( I ) 2.4(3) 3.4 ( I ? ) 2.6(12) 5.1 (9) 4.0 ( 19) 5.1 (9) 3.2 (13) 2.0 ( 4 ) 2.1 1 2 , 1.8 13 5.0 I 141 1.9 15, 2.7 (121 l.O(III 1.6 (4) 4.0( I 7.3 ( 5 ) 1.6(1) 4.1 ( 5 ) 1.0 (3) 2.2 (17) 1.3(l) 3.6 (7) 2.5 (4) 2.5 (9) 1.o (9) 2.9 (IO) f.0(1) 4.5 (I) 1.5 ( I ) 3.1 ( 5 ) 1.0 (4) 4.2 (14) 6.1 ( 2 ) 4.5 (4) 1.3 ( I ) 4.4 (7) 1.o (4) -1.7 (7) (0) 2.8 (3) 3.2 (3) 2.0 (4) I .o ( 5 ) 'Time-weighted avenge. lag I .5 y. 'ppm = panls) per million. $Hematologic neoplasms (International Classification of Diseases [ICD]9: 30-208): NHL = non-Hodgkin', lymphoma (ICD9: 200. 102): leukemia (ICD9: 1(Y-208); ANLL = acute nonlymphocytic leukemia (ICD9:205.0. 206.0. 207.0): MDS = myelodysplvdc s! .!dromes (ICD-02: 9980-9989); other leukemias: includes leukemias other than ANtL and leukemias not otherwise specified i,ICXk 204, 205.1-205.9. 206.1-206 9. 207.1-207.9. 208). See (34) for ICD9 and (35) for ICD-01. 8CI = confidence interval. !!Coatings: painters and other coating application worken. meferent relative risk = 1.0 for unexposed workers. with all risks adjusted for age and sex. MDS was significantly increased among both earlier (RR = 4.0; 95% CI = 1.311.9) and later (RR = 5.1; 95% C1 = I .5- 17.2) hires. Although occupationspecific results were limited by small numbers. risks for A N L U l D S were el- evated among coatings workers (RR = 4.1;95% CI = 1.4-12.6). rubber workers (RR = 6.1; 95% CI = 1.1-33.3). chemical workers (RR = 4.5; 95% CI = 1.1- 18.1). and among those with other or mixed occupations (RR = 4.4; 95% CI = 1.3-15.2). Risk for NHL was also el- evated among several occupational groups. but only chemical workers showed a statistically significant excess (RR = 7.8; 95% CI = 1.9-32.5). Benzene-exposed subjects experienced significantly incresed risks for hematologic neoplasms at avenge benzene exposures of less than 10 ppm (RR = 2.2; 95% CI = 1.13.2) and cumulative expo- sures of less than 40 ppm-years (RR = 2.2: 95% CI = 1.14.5) (Table 2). Trends of increasing risk for hematolo,O' IC neoplasms were seen with respect to increasing average exposure (including the subcohort of subjects with a constant level of occupational exposure) and cumulative exposure to benzene. Risk for NHL in- creased with increasing duration of expo- sure. from 0.7 ( 9 5 9 CI = 0.1-7.2) among those exposed for fewer than 5 years to 4.2 (95% CI = 1.1-15.9) associated with exposures of 10 or more years, but the exposure-response patterns were less clear for the other measures of exposure. ANLL and AiLIAIDS both showed patterns of increasing risk with increasing average exposure to benzene, with more consistent exposure-response patterns for A N L W D S than for ANLL alone. The link of ANLUMDS with avenge exposure was strongest when restricted to subjects with constant levels of occupational exposure: R R s rose from 3.2 (95% CI = 1.0-10.3) for those with constant lowlevel exposures (<IO ppm) to 7.1 (95% CI = 2.1-23.7) for those with constant high- level exposures ( 3 2 5 ppm). Risk for ANLUMDS did not increase with duration of exposure to benzene (Table 2). The RRs associated with fewer than 5 years. 5-9 years, and 10 cr more years of exposure were 11.7 (95% CI = 2.947.3: based on four cases). 5.2 (95% Ci = 1.419.3; based on five cases), and 2.5 (95% CI = 0.8-8.9; based on 10 cases). respec- tively. among benzene-exposed workers hired before 1972 and 5.7 ( 9 5 9 CI = I .5-2 I .5; based on six cases). 3.5 (95% CI = 0.6- 19.8;based on two cases), and 7.2 ( 9 j % CI = 0.8-65.6; based on one case) a w n g workers hired in 1972 or later. In nrtitlier time period was longer duntion of bciizrne exposure clearly linked to incr:.ssing risk of ANLUErlDS. Risk for ANLWMDS increased with increasing cumlative exposure to benzene. but the hishest risks were not seen at the highest exposure level (cumulative ppm-years: 40-99, RR = 6.0 [95% CI = 1.8-20.61; cu!iiulative ppm-yean: 3100, RR = 4.4 195% CI = 1.4-13.51). Although there W:I< some evidence of increased risk for leckemias other than iLvLL among benrex-exposed workers (Table 1). clear p2::erns of increasing risk with increasing exi?osurewere not observed (Table 2). We partitioned cumulative benzene ex- posure into recent (<IO years prior to diagiiosis) and distant (210 years prior to di::yosis) exposure. NHL was strongly IiilAed with distant exposure to benzene (P for trend = .005;based on 13 exposed ~ 3 . 2 ~ b) u. t the association with exposure in :he most recent 10 years was weak ( P fcr trend = .15; based on 16 exposed c a w ) . NHL showed no clear link with Work that involved only recent exposure ( R R = 1.5 [95% CI = 0.3-7.61; based on thl-cc exposed cases) (Table 3). In cont r s t . risk of ANLUMDS was signifi- Journal of the National Cmcer Institute. Vol. 89,No. IJ. July 16, 1997 REPORTS 1067 Table 2. Relative risk (RR). for hemvolagic neoplasms and related conditions Jccoding Io extent of exposure to benzene . _. All I ~ : : I I : I I O ~ I ~ I C nLnpl;r\ms C.i.C\ 24 16 18 11 RR -7.7- 3.1 2.a . w 3 ?.I 95'i ClX 1.14.2 1.5-6.5 1.15.7 I .M.I 8 II 3.2 2.9 ,002 1.3-7.6 I J6.6 16 3.1 1.5-6.6 17 25 3.3 2.0 1.6-6.9 1.0-3.9 .:r 18 2.2 1.14.5 II 2.9 1.3-6.5 29 27 1.J-5.2 .w I .: NHL c2.L.s RR 95': CI 727 70 3 I 4 II 6I9 -2.7 1.7 4.7 .04 3.0 - 3.5 .I5 0.7 3.3 4.1 .01 3.3 1 . 1 3.5 n: 0.7-10.6 0.3-10.2 I.I-I a. I 0.8-1 1.6 0.7-17.3 0.1-7.2 0.7-14.7 1.1-15.9 0.8-13.1 0.1-1 1.1 0.9-13.2 . 3 LeuAcrnm CAC> RR '95'7 CI I5 2.0 0.94.5 13 IO 3.7 2.3 1.6-6.7 0.9-5.7 .02 12 7 7 I4 I 1 13 1.7 4.0 2.S .W 4.0 3.1 I .5 0 . 7 4 1 1.5-10.7 1.0-7.4 1.7-9.6 1.3-7.5 0.6-3.6 I1 a .YS 1.9 3.1 0.84.7 I.ZS.0 19 2.7 . l ) l 1.2-6.0 9 ANLL C .\es RR 95'; CI 79 5 64 5 966 5 5 II 2.0 5.8 2 6 .04 1.9 4.9 4.4 ,008 5.6 3.6 1.6 .SI 1.9 4.3 3.6 ,116 0.6-7.0 1.5-18.9 0.7-9.9 0.5-6.8 1.2- 19.8 1.2- 16.4 1.7-18.7 1.0-12.9 O.IJ.~ 0.5-7.0 1.1-16.0 1.1-11.6 - ANI.LS1DS C;l.es RR 9 5 7 CI I1 9 a 3.2 5.8 4.1 .01 1.0-10.1 1.8-18.8 1.2-13.2 IO 4 a IO 7 I 1 3.2 5.1 7.1 .0003 6.6 4.3 2.8 1.0-10.3 1.3-20.6 2.1-23.7 2.0-21.6 1.3-14.5 0.0-9.1 7 7 14 .-r8 2.7 6.0 4.4 01 0.8-9.5 1.8-20.6 1 . ~ i 3 . s 4 W c r 'xkemiu CSeS RR 95'; CI 8 1.9 0.6-5.9 4 2.1 0.673 5 2.1 0.6-7. I 27 63 2 1.6 3.2 1.4 .31 0.5-5.1 0.5-13.4 0.3-7.4 5 z7 0.7-9.7 5 2.7 0.9-9.3 7 I .4 OSJ.~ .a: 6 2.0 0.6-6.6 3 2.1 0.5-9.0 a 1.9 0.66.0 .36 5 . k ? X J n-y, -x IO-' 352 I57 1aa 324 aa 121 249 n o 278 300 119 279 'R-ferent: RR = 1.0 for unexposed workers, with all risks adjusted for age and KX. t S H L = non-Hodgkin's lymphoma; ANLL = acute nonlymphccytic leukemia: MDS = myelodysplastic syndroma: other leukemia: includes 1eu;:rnias other than . X W Land leukemias not otherwise specified (see Table 1 for ICD codes). $FPm = pan(s) per million. ST!IIS category encompasses a p u p of person-yem for workers who were always exposed 31 the indicated levek ]If\dues are two-sided. ?Croups for trcnd test: unexposed m d exposure of <IO. 10-39, 4-99,100-399,and ppm-yeus. WI = confidence interval. cantly increased among those who had ries (<1960, 1960-1971. and 31972) was on]) recent benzene exposure (RR = 6.0 included in the statistical re,gression mod- [ 9 j 5 CI = 1.9-18.43; based on 14 ex- els. the findings were unchanged (data not posed cases). Risk of ANLUMDS was shown). also strongly associated with increasing amciints of recent (P for trend = .003) Discussion but not distant ( Pfor trend = .51) expo- In earlier reports (22.23). we showed sure. that employment in benzene-exposed jobs SHL was linked to distant cumulative in China. regardless of the exposure level. exFosure to benzene among coatings was linked to hematologic disease mortal- workers (the largest occupational group, ity and the incidence of leukemia, NHL. mok:ly painters) (Pfor trend = .IO:based MDS, and aplastic anemia. This investi- on ' h e exposed cases) and among those gation provides evidence that benzene wit!? other benzene-exposed occupations may cause hematologic neoplasms and re- (PFh,rtrend = .005; based on 10 exposed lated disorders at average exposures of ca>:\). ANLUMDS was linked to recent less than IO pprn and cumulative expo- cui**lilativrexposure among coatings sures of less than 40 ppm-years. Increas- w o r k s (P for trend = .OOl;based on 14 ing risks for ANLL were linked with in- e x p w d cases) and among workers in creasing levels of benzene exposure, o t h r occupations (P for trend = .06; particularly for exposures that had oc- basc'd on I4 exposed cases). When the curred within the 10 years prior to diag- ye:!. o f initial employment in study facto- nosis. and showed a stronger link for the combined categories of ANLL? IDS than for A . i L alone. Risk for NHL increased with duration of exposure and N 2s strongest among workers who had di>::int ( 2I O years) exposure. No excess was noted for NHL among workers hired siiice 1972. but foilow-up for this group \vas relatively short. In rodents, benzene causes .! range of neoplasms. including lymphor .:s and tumors of epithelial origin. but. fi-r reasons not understood. increased risk fdr granulocytic leukemias has not bee:i reponed (24-26). In earlier epidemiologic studies. benzene has been linked mo.: strongly with ;LULL. while results for 0::xr hematologic neoplasms have been ir:onsistent (5).However, the epidemioIog;<tlaubasc for qwtitativr assessment of i.incer risk associated with benzene ex;'osure is sparse. No increases in risk were 5wnd in ;I IO(.> REPORTS Journal of the Notional Cancer Institute. Vol. 89,No. 14, July 16. 1997 Table 3. Relvrrc nsk (RR). for ulectcd hern.:t*lopc conditions among hcnacne-exp.wd workcn according IO recency of exposure Di\tant c x y w r c . ppm-yng - None c;L\cs P-Y (XlU')Il RR 955 crq 40 CLUS P-Y ( X I 0 7 RR 959 CI 340 Cses P-Y (XI07 RR YS% CI Total cases P-Y (x103) RR 95% CI SHLt Recent exposure. ,ym-yrh: None <?O 240 Reftrcnt 0 -3 0 6 0 -9 3 218 2.5 0.5-13.2 3 97 3.4 0.7-17.4 1 31 6.4 1.04 I . 1 8 347 3.4 0.9-13.0 0 136 I 46 2.6 0.3-25.3 7 I59 4.7 1.2- 18.9 8 342 1.9 0.5-10.8 Total 3 354 I .5 0.3-7.6 4 I47 3.I 0.7-14.3 9 I97 4.9 1.3-18.9 ANLUMDSt Recent exposurc. ppm-yrsS None c40 r J O Referent 0 -3 0 6 0 9 5 218 3.7 I .O-13.9 3 97 2.4 0.5-10.7 I 31 1.0 0.2-I 8.5 9 347 2.9 0.9-9.3 9 I36 9.1 1.8-19.7 4 46 7.4 1.8-19.7 6 I59 2.7 0.8-9.8 19 342 5.3 1.8-15.6 Tod I4 354 6.0 I .9-18.4 7 147 3.8 1.1-13.0 7 I97 2.5 0.7-8.7 Linear vend tests: (never cqused [referent]. A to <IO, 10-.:9,10-99, 100-399.and 3400 ppm-yn, in time period). NHL Recent exposux. P = .15. Distant exposure. P = .005. ANLlMDS Recent exposure. P = ,003 Disrvlt cxposurr. P = SI *Referent: RR = 1.0 for aexposed workers. wirh dl r i A adjusted for age and KX. TNHL non-Hodgkin's ::rmphoma; ANtL = acuu noni) mphocyuc leukemia:MDS = rnyelcdysplvtic syndromes (Jcc Table 1 for ICD codes). %Recentexposure: cumuhve benzene exposure (pans pe: million [ppm]-yeus, lag 1.5 yeus) in period e10 ye= prior to diagnosis. $Distant exposure: c u m u h v e benzene exposure (ppm-yes. lag 1.5 yeus) in a10 yevs prior ro diagnosis. I1P-Y(XI@`) = person-ycn per 1ooO. CI = confidence inrent; _- case-conuol study (3, a i leukemia (based on 14 exposed cases). nultiple m y e l o m (based on seven exposxi cases), or NHL (based on nine exposd asses) among Canadian petroleum disa5ution workers exposed predominantly :a less than 1 pprn benzene. Also, among Scandinavian service station workers exposed on avenge to less than 1 ppm bezene, no excesses were found of Ieukemia (based on 28 exposed cases) or acute nyeloid leukemia (based on 13 exposed cases) (I)..4mOng 4607 benzene-expose2 and 3074 unexposed U.S. workers 16hrisk rose in an txposure-response faskion for lymphatic and hematopoietic c a ~ ~ e (rbsased on 18 cases) and for leukeaia and aleukemia (based on six cases) uith respect to benzene exposure categcezed as none. less than 15 ppm-yews, 15-59 ppm-years. and 60 ppm-years or more. In another study (7) among 956 benzece-exposed workers in Michigan, seven dexhs caused by lymphatic or hematopoieli; cancer were identified. including four aused by leukemia (all myelogenous leukemias). Average expcsures to benzene were 1-5 ppm for b e t of the individuals who died of leukeri3 and about 18 ppm for the fourth. Cor.pared with general population rates, total deaths caused by myelogenous leukemia, were significantly elevated: the R R b were 1.7 among workers with less than 12 ppm-yem' cumulative benzene e x p w r e (based on two cases) and 2.5 arncng workers with 83 ppm-years or moi? of benzene exposure (based on one C351). I n a third cohort study (8), 1165 rubber hydrachloride manufacturing workers in the TJnited States were followed through 19s:. and I5 deaths caused by lymphatic or I:trnatopoietic cancer, including nine cac>cd by leukemia (six by ANLL) and focr caused by multiple myeloma, were idt..:iified. The leukemia deaths occurred am,.ng workers who were generally exp o ~ dto benzene at levels of more than 20 ;'prn, and the risk increased with incre:..\ing exposure, showing significant exLx e s with cumulative exposures of grc::icr than 200 ppm-years, consistent with an exponential in risk with increasing cumulative dose (8). Updated reports (11,27) on this cohort, with follow-up through 1987, show similar doseresponse results for acute myeloid leukemia (AML) (based on six cases) (71) but indicate that risk for the non-AML leukemias (based on eight cases) was also in excess (27). Other investigatots (10.28) suggest that exposures in this cohort may have been higher, but this suggestion has also been disputed (29). In contrast to the findings among rubber hydrochloride workers (8.1f), our study shows excess risk at relatively low levels of exposure (<IO pprn avenge and 4 0 ppm-yean cumulative), but it shows a relatively modest dose-response effect, with proportionally smaller increases in risk at increasing levels of exposure. While the precise shape of the benzene exposure-response curve can only be estimated with caution from either of these studies, our larger study is in agreement with human (30) and rodent (31) meta- bolic studies showing a shift at higher Icv- Journal of the Notional Cmcer Institute. Vol. Y9. Ng1. 14. July 16. 1997 REPORTS 1069 . els of benzene exposure in the proportion calibntion of exposure levclh was not observation suggests that the effects arc of the likely toxic precursors. such as hy- possible. particularly for disti.iguishing more likely due to the common exposure droquinone and muconaldehyde. small differences in low-levcl c: \.posures. to benzene than due to other exposures. MDS is a known precursor of acute Hematologic disease diapnosc. for this While this study is larger than previous myeloid leukemia after treatment with al- investigation were based on c:it,:l'ul noso- investigations and includes workers with kylating agents (18). Earlier, we showed logic review of available path(>Itrgicma- a wide nnge of exposures to benzene. the (22) that MDS was linked to employment terial and related documentatiorl by expert estimates of risk. as measured by statisti- in benzene-relatedjobs (RR = x : 95% CI hematopathologists (16). an inipi ovcment cal CIS. are still fairly broad. and would = 1.7 to x ; based on seven cases). and over previous death certific.:tc.-based benefit from the larger numbers that could here we relate A N L U I D S to the extent studies ( 6 . i , l l ) .This effort %.is of par- be provided by continued follow-up of of benzene exposure. Earlier mortality ticulv importance in China, bt'.c..'iuse re- this population. Nevertheless. the obser- studies of benzene-exposed populations cent nationwide vital statistics ::nd cancer vation of doubled risk for ANLUMDS at may have failed to identify this associa- incidence validation studies ar: unavail- an avenge exposure of less than IO ppm tion because .LIDS has only begun to be able. Because benzene-exposec! workers benzene. the proportionally smaller in- more widely recognized in the past 10-15 Ye known to be at excess risk ''or leuke- creases in risk with increasing exposure. years and patients may die of complica- mia. it is possible that addititml early the elevated risk for .ANLU;C.fDSwith re- tions relating to the cytopenias of MDS cases were identified in the exposed cent exposure, and the possible links with before its evolution to ANLL. In our group compared with the ulicxposed NHL are all provocative new observa- study, A N L m l D S was linked to recent group due to more thorough sur?cillance. tions that should enhance our efforts [O benzene exposure, and additional distant Such a bias. however. could not have understand benzene carcinogenesis in hu- exposure did not appear to funher in- been substantial, because overall excesses man populations. crease risk. This finding suggests that re- were similar when restricted to exposed cent benzene exposure is predictive of and unexposed deceased indiviLiuals (22). References subsequent risk for AiiL&IDS, analo- We compared disease rates ..;;long the gous to the short latency and wavelike increase then decrease in risk seen for several forms of radiation-induced leukemia (19) and for chemotherapy-related AML (18). In contrast. recent exposure benzene-exposed population - i i t h rates among a nonexposed industri:ii population. This comparison was an Advantage because population data on caiicer incidence and monality rates in China are ( I ) Wallace LA. Major sources of benzene exposure. Environ Health Penpecr 1989:Yz:165-9. (2) Hricko A. Rings of controversy xound benzene [newsj. Environ Healrh Pcrspect 1994: 102276-8I. (3) Schnatter AB. Armstrong T". Nicolich MJ. was only weakly linked to NHL, suggesting that benzene-related NHL may be associated with a longer induction period. Our study included workers who were hired before the period of risk analysis limited and because an earlier >tudy(33) found that comparisons of canc:r rates in benzene-exposed workers wi:h general population rates tended to underestimate leukemia risk among benzenc-exposed Thompson FS. K;lu XM, Huebner W.et 31. Lymphohaematopoietic malignancies and quantitative estimates of exposure to benzene in Canadian petroleum distribution worken. Occup Environ Sled 1996:53:773-8I. (4)Lynge E, Andersen A. Nilsson R. Barlow L. began (Januuy 1972). and we do not have information on the number of hematologic neoplasms that occurred in this earlier time period. Continued follow-up of the cohort. particularly those hired since workers. Our earlier investigation in this cohon ( 2 2 )showed that benzrr:-exposed and unexposed workers were closely comparable for total mortality (XR = 1.1; 95% CI = 1.0-1.2) and cmce: mortality Pukkala E. Nordlinger R. et 31. Risk of cancer and exposure to gasoline vapors. Am J Epide- miol 1997:145:449-58. (5)Slc.Michae1 .aC.Ycinogenicity of benzene. tolu- ene and xylene: e p i d e ~ o l o @ m d experimen- tal evidence. MRC Sci Pub1 1988:3-15. 1972. will provide improved insight into the temporal development of benzeneassociated diseases. While this investigation in China substantially increases the number of person- (RR = 1.1; 95% CI = 1.0-1.i)but that death rates in both groups we:> substantially lower than those expected by comparison with available vital sta:istics data for China. This latter finding u!!derscored (6)Wong 0.An industry wide monality study or' chemical workers occupationally exposed to benzene. 11. Dose response analyses [published emturn appeus in Br J Ind Sled 1987:U:7761. Br I Ind .Med 1987:44382-95. (7) Bond GG. 4lcLaren EA. Baldwin CL. Cook years of follow-up and. in particular. the number of hematologic neoplasm outcomes for determining quantitative relationships between benzene exposure and disease risk. certain design aspects should the need for an internal compxison with an unexposed population 01' Chinese workers from the same geogr:!$ic areas. A s in most industrial ser:ings. the workers in this investigation u ?re likely RR. An update of monaiity miong chemical workers exposed 10 benzene [published e m turn ~ p p e min Br J lnd Sled I9Y7;U:2l5]. Br J Ind >led 1986;43:685-91. (8)Rinsky R4. Smith AB. Hornung R. Filloon TG.Young RI. Okun AH. et a1. Benzene and be recognized. As in previous epidemiologic investigations, benzene-exposure estimates were based on historic data (15), charxrerizing only average exposure levels over broad time periods. We exposed to a number of chen-icals other than 'benzene and the observed :isks could be due to some other expo>~!-c.sH.owever. the subjects in this stud! :\ere emplo>ed in a variety o f OCWF :ions. and leukemia. An epidemiologic risk assessment. N Engl J bled 1987;316:1O-U-50. ( 9 ) P;L.tton M B . Chinchilli VSl. Brett SXI. Rodricks JV. Leukemia ri3k ~ssswiatedwirh benzene exposure in the pliotilm cohort. 11. R l h k estimates. Ri.\k .Anal 1994:14:1S5-61. have shown (32) that o u r benzeneexposure mnking and the occurrence of benzene poisoning are in good agreement. providing indirect support for the generd validity of our approach. but the precise exces>csof hematologic disea.2 were not restricxd to a piinicular subxr of benzene-related occupations. wit!. [he possible exception of the notably h;;hcr risks for S H L among chemical w c ~ ~ k x sT.his ( 10) Crump KS. Risk of benzene-induced leukemia: ;L wmiriviry Jna1yci.i ut' [he pliorilrn cohort with .dditimaI i ~ l l ~ w ~- i~l pdnew expciwrc earimarcs. J Tosic.ol Envirt~nH d t h IW4:42: 119-42. ( I I ) Wong 0.Risk of acute myeloid Icukaemia Jnd 1070 REPORTS Journal of :ICNalional Cancer Inhture, Vol. 89. No. 1-1. July 16. 199: !nu;:iplc mycloma in workcrs c x p o . 4 to ben- 7mc. &cup Environ Mcd IYYS:52:3XM. I/:) Yin SN. Li Q. Liu Y. Tian F. Du C. Jin C. Occupational expnxurc to hcnzcne in Chino. Br J Ind .Wed 1987:U:192-5. I/.?] Sin SN. Li G. Tain F. Fu 21. Jin C. Chcn YJ. et AI. Lcukaemia in knrunc workers: a revozpective cohon study. Br J Ind Mcd 1987:U: 124-8. 114) S i n SN. Linct MS. Hayes RB. Li G& D o x meci kl. Wang YZ. et 11. Cuhon study among workers exposed to benzcnc in Chinr 1. General methods md resources. Am J Ind Med I994:26:383-100. I 151 Do.scmeci .M. Li GL. Hayes RB. Yin SN. Linct 11. Chow WH. et al. A cohon study among workers exposed to knzcne in Chinr 11. Ex- powre iwcssment. Am J Ind Med 199426: 401-1 I. $ 1 6 )Travis LB. Li CY. Zhang ZN. Li DG, Yin SN. Chow WH. ei al. Hematopoietic malignancies and related disorders among benzcnetxposed workers in China. Leuk Lymphoma 199414: 91-102. I 17) Bennett JM. Catovsky D. Daniel ,MT. Flmdrin G. Galton DA. Gnlnick HR. et al. Proposals for the classification of the rnyelodysplastic syndromes. Brit J Haematol 19825 1:189-99. i 18) Park DJ. Koeffler HP. Therapy-related acute myelocytic leukemia. In: Wiernik PH. Canel- low GP. Dutcher JP, Kyle RA. editors. Neoplastic diseases of the blood. New Yo& Churchill Livingstone. 1996:381-107. i19) Committee on the Biological Effects of Ionizing Radiation. Health effects of exposure to low levels of ionizing radiation. BEIR V. Washington (DC): National .Academy Press. I990 I42I. 1201 CUKiS RE, Boice JD Jr. Stovall M.,Bcrnstcin L. Greenberg RS. Rannery JT.et al. Risk of leu- kemia after chemochenpy and radiation mat- ment for breast cancer. N Engl J Med 1992: 526: 1745-5 1. 121) Breslow NE. Day NE. Statistical methods in cancer research. Volume 11-The design and analysis of cohort studies. IARC Sci Pub1 I957:51: 1406. 122) Yin SN. Hayes RB. Linet MS, Li GI- Dosemeci M. Tnvis LB. et 31. A cohon study of canccr amnng bcnzcne-cxposed workers in China: over~llresults. Am J Ind Med lW:29 217-35. (23) Hayes RB. Yin SN. Dorcrncci M. Li GL. Wacholdar S. Chow WH. et aI. .Morulity among benzene-expod workers in China Envimn Health Pcrspcct 1996:I W: 1349-51. (24) Maltoni C. Cilibeni A. Cotti G. C m o B. Belp o g i F. Eenzcne. JII experimental muiuprcntid cucinogm: re.sulU of the long-rem bioassays prrformed ~t the Bologna Institute o i Oncology. Environ Health P e r s p n 198982:IWX ( 2 3 Huff JE. H u m a n JK. DeMarini DSL Eusiis S. .Maronpor RR. Peters AC. et al. .\lultiple-site carcinogenicity of benzene in F i r k r 344 mu and B6CjFI mice. Environ Heaich Perspect 1989:52:1 3 - 6 3 , (26) h m s G M Everitt JI. Irons RD. P ~ JXJ. Car- cinogenicity of inhaled benzene in CBA mice. Fundam .Appl Toxic01 I993:10jOtt. (27) Savitz DA. Andrews KW. Risk of rnyelogenous leukemia and multiple myelora in workers exposed to benzene [letter]. O c ~ Eqnviron Med 1996:53:357-8. (28) Paxton .MB. Chinchilli VM. B e n S.M. Rodncks JV. Leukemia risk associated with benzene exposure in the pliotilm cohon: I. Mortality update and exposure disuibuaon. Risk Anal 1994:14: 147-54. (29) Utterback DF. Rinsky RA. Benzex exposure assessment in rubber hydrochloride xorkers: a critical evaluation of previous estimes. A m J Ind &led 1995:17:661-76. (30) Rothman N. Li GL. Dosemeci M. Bahtold W, ,Mani GE. Wang YZ. et al. Herumtoxicity among workers heavily exposed :o benzene. Am J Ind Med 1 9 9 6 : 2 9 2 3 6 4 . (31) Henderson RF. Sabourin PJ. Bcchtold WE. Griffith WC. .Medinsky MA, B i r n h m LS. et al. The effect of dose. dose rate. mute of ad- ministration. and species on tissue m d blood levels of benzene metabolites. Envimn Hedth Penpect 1989:32:Y-17. (32) Dosemeci .M. Yin SN. Linet .M. Wxcholder S. Rothman N. Li GL.et d.Indirect \didation of benzene exposure assessment by =sociation with benzene poisoning. Environ Health Perspect 1996:lW:1343-7. (33) Wong 0. An industry wide m o d i r y study of chcmicil \\ 4 d.crs occupationally exposed in benzene. I. C;uncnl rcsults. Br J Ind Mcd 1987: JJ:365-8I . (34) World Hd:!! Organization. International Classsilication I)!. Dl\easc.s. Manual of the Intemational Stati.; ~ a Cllasxilication of Diwxus. Injuries. and C:iu%c\ of Death. Geneva: World Hrnlth Org.:i17at1cm. 1977:1-773. (3.5) World He:iI1!1Organization. Intcrnationa1CI&.silication o i D i ~ a w \for Oncology. Geneva: World Heii::: Organization. 1990:l-la. Notes ,Members of I!.,< Chinc3e Academy of Preventive iMedicine-Sation.:i Cancer Institute Bcnzqe Study Group are as fol! 'ivh: S.-Z. Wmg (Stations of Public Health and ?.-!mion of Infection. Shanghai. China). Z.-L. Ji::.; tlnstiiutes of Labor Health and Occupational Di.ase. Tianjin. China). T.-R. Dai (Stations of Pubi.; Health and Prevention of Infection. Chengdu. C'iinai. W.-U. Zhang (Institutes of Labor Health a n i Occupational Disease, Sichuan. China). X.-J. CL:o istations o i Public Health and Prevention of In!; .xion. Chonqing. China). P.-2. Ye (Institutes of LJ;..tir Health and Occupation4 Disease. Heilongjiar.;. China). Q.-R. Kou (Institutes of Labor Health 3nJ Occupational Disease. Shenyan. China). X.-C. Zh..:i_c (Institutes of Labor Health 3nd Occupational Di>:ise. Henan. China). X.-F. Lin (In- stitutes of Labor I-lealth and Occupational Dixase. Gumgzhou. China). J.-F. Meng (Institutes of Labor Hedth and 0cc::ptional Disease, Jiangxi. China), C.-Y. Ding (1nsti::ites of Labor Health and Occupational Disease, Jirilhou. China). J.3. Zho (Institutes of Labor Health .ind Occupational Disease. Nanchang, China). a d D.-G. Li, Z.4. Zhang (Peking Union ,Medical Hospital. Chinese Academy of LMedical Science.. Beijing. China). We thank Dr. \\'iiliam J. Blot. former chief of the Analytic Studies ?ranch. Sarional Cmcer Institute. for his i m p o m r :untributions to the earlier phases of this investigat...in. N'e also wish to acknowledge the suppon pro!dsd by Nancy Odaka of Westai. Inc.. Rockville. \lD. and Tom Helde, Information Management Scr. ices. Inc.. Rwkville. MD. in data management for his study. .Mmuscript rec:ived November 18, 1996: revised April 1.1997: ac xpted May 9. 1997. l~wriialof the National Cancer Institute. Vol. 89. NO. I 4 July 16. I997 REPORTS 1071