Protective Alterations in Phase 1 and 2 Metabolism of Aflatoxin B 1 by Oltipraz in Residents of Qidong, People s Republic of China

Transcription

1 Protective Alterations in Phase 1 and 2 Metabolism of Aflatoxin B 1 by Oltipraz in Residents of Qidong, People s Republic of China Jia-Sheng Wang, Xinnan Shen, Xia He, Yuan-Rong Zhu, Bao-Chu Zhang, Jin-Bing Wang, Geng-Sun Qian, Shuang-Yuan Kuang, Audrey Zarba, Patricia A. Egner, Lisa P. Jacobson, Alvaro Muñoz, Kathy J. Helzlsouer, John D. Groopman, Thomas W. Kensler Background: Residents of Qidong, People s Republic of China, are at high risk for development of hepatocellular carcinoma, in part due to consumption of foods contaminated with aflatoxins, which require metabolic activation to become carcinogenic. In a randomized, placebo-controlled, double-blind phase IIa chemoprevention trial, we tested oltipraz, an antischistosomal drug that has been shown to be a potent and effective inhibitor of aflatoxin-induced hepatocarcinogenesis in animal models. Methods: In 1995, 234 adults from Qidong were enrolled. Healthy eligible individuals were randomly assigned to receive by mouth 125 mg oltipraz daily, 500 mg oltipraz weekly, or a placebo. Sequential immunoaffinity chromatography and liquid chromatography coupled to mass spectrometry or to fluorescence detection were used to identify and quantify phase 1 and phase 2 metabolites of aflatoxin B 1 in the urine of study participants. Reported P values are two-sided. Results: One month of weekly administration of 500 mg oltipraz led to a 51% decrease in median levels of the phase 1 metabolite aflatoxin M 1 excreted in urine compared with administration of a placebo (P =.030), but it had no effect on levels of a phase 2 metabolite, aflatoxin mercapturic acid (P =.871). By contrast, daily intervention with 125 mg oltipraz led to a 2.6-fold increase in median aflatoxin mercapturic acid excretion (P =.017) but had no effect on excreted aflatoxin M 1 levels (P =.682). Conclusions: Intermittent, high-dose oltipraz inhibited phase 1 activation of aflatoxins, and sustained low-dose oltipraz increased phase 2 conjugation of aflatoxin, yielding higher levels of aflatoxin mercapturic acid. While both mechanisms can contribute to protection, this study highlights the feasibility of inducing phase 2 enzymes as a chemopreventive strategy in humans. [J Natl Cancer Inst 1999; 91:347 54] Hepatocellular carcinoma (HCC) is one of the most common cancers worldwide and results in more than deaths annually in the People s Republic of China. HCC is the leading cause of cancer death in Qidong, a city in eastern Jiangsu Province, People s Republic of China, and accounts for up to 10% of all adult deaths in some of the rural townships (1,2). It has been postulated that chronic infection with hepatitis B virus and exposure to aflatoxins in the diet contribute to the extraordinarily high risk of HCC in Qidong (1,2). Aflatoxins are potent hepatocarcinogens produced by some strains of Aspergillus fungi and are consistent contaminants of the food supply in this area, particularly in corn, peanuts, soya sauce, and fermented soy beans. Two nested case control studies in an ongoing cohort from nearby Shanghai have demonstrated a strong interaction between hepatitis B virus and aflatoxins for risk of HCC (3,4). A similar chemical viral interaction has been observed in Taiwan (5). From a public health perspective, these findings suggest that hepatitis virus vaccination programs and efforts to reduce aflatoxin exposures could have a major impact on the incidence of this disease. Indeed, a universal vaccination program against hepatitis B virus that started a decade ago in Taiwan is now resulting in lower rates of HCC in children (6). The extent of aflatoxin contamination in foods is a function of ecology and is not completely preventable. Secondary prevention programs, such as chemoprevention, may be useful in this setting. Experimentally, aflatoxin-induced hepatocarcinogenesis can be inhibited by more than a score of different chemopreventive agents (7,8). One of the most potent and effective agents in these animal models is the antischistosomal drug oltipraz (9). Dietary administration of oltipraz to rats afforded complete protection against aflatoxin-induced hepatocarcinogenesis when administered before and during the period of carcinogen exposure. Chemoprevention in these animals is reflected in lowered levels of several aflatoxin biomarkers (9 11). The protective actions of oltipraz in this model are thought to result primarily from an altered balance between the activation and detoxification of aflatoxin B 1 (AFB 1 )in the hepatocyte. As shown in Fig. 1, anticarcinogenic concentrations of oltipraz in the diet can markedly induce activities of detoxifying phase 2 enzymes such as glutathione S-transferases (GSTs) in rat tissues. This induction facilitates conjugation of glutathione to the ultimate carcinogenic species, aflatoxin-8,9-oxide, thereby enhancing its elimination as a mercapturic acid (AFB NAC) and coordinately diminishing DNA adduct formation (12,13). Formation of these DNA adducts is an essential but insufficient component of aflatoxin-induced hepatocarcinogenesis. Induction of GSTs by oltipraz in primary cultures of human hepatocytes has been observed (14), although the catalytic activity of human GSTs toward the 8,9-epoxide appears to be lower than that of rodent GSTs (15). Oltipraz can also influence phase 1 enzymes, particularly cyto- Affiliations of authors: J.-S. Wang, X. Shen, X. He, A. Zarba, P. A. Egner, J. D. Groopman, T. W. Kensler (Department of Environmental Health Sciences), L. P. Jacobson, A. Muñoz, K. J. Helzlsouer (Department of Epidemiology), The Johns Hopkins University, Baltimore, MD; Y.-R. Zhu, B.-C. Zhang, J.-B. Wang, Qidong Liver Cancer Institute, Qidong, Jiangsu Province, People s Republic of China; G.-S. Qian, S.-Y. Kuang, Shanghai Cancer Institute, Shanghai, People s Republic of China. Correspondence to: Thomas W. Kensler, Ph.D., Department of Environmental Health Sciences, The Johns Hopkins School of Hygiene and Public Health, 615 N. Wolfe St., Baltimore, MD ( tkensler@jhsph.edu). See Notes following References. Oxford University Press Journal of the National Cancer Institute, Vol. 91, No. 4, February 17, 1999 ARTICLES 347

2 Fig. 1. Steps where oltipraz might affect the metabolism of aflatoxin B 1 (AFB 1 ). AFM 1 the metabolite aflatoxin M 1 ; GSH glutathione. chrome P450 activities. Enzyme kinetic studies on heterologously expressed human CYP1A2 indicate that oltipraz is a competitive inhibitor, with an apparent inhibition constant (K i ) of 10 M (16), a pharmacologically achievable concentration in rats and humans (17). CYP3A4 can also be inhibited but with an eightfold higher K i value (16). Inhibition of CYP1A2 by oltipraz results in diminished metabolism of aflatoxin to the genotoxic 8,9-epoxide and the hydroxylated metabolite aflatoxin M 1 (AFM 1 ) in primary cultures of rat and human hepatocytes (16). Urinary excretion of AFM 1 also drops dramatically immediately after oltipraz administration to aflatoxin-treated rats (18). Thus, both inhibition of cytochrome P450s and induction of electrophile detoxification enzymes are likely to contribute to chemoprevention by oltipraz. Phase I clinical trials with oltipraz have indicated that plasma concentrations of the drug equivalent to those observed in the rodent chemoprevention studies are easily obtained in humans without substantial adverse effects (17,19). Moreover, pharmacodynamic action, as evidenced by elevations in the levels of messenger RNA transcripts and/or activities for detoxification enzymes, has been reported in these human trials (20). As a result of these collective clinical and experimental findings, a phase IIa chemoprevention trial with oltipraz was conducted in Qidong in 1995 (21). Individuals randomly assigned to a group receiving 500 mg oltipraz once a week exhibited a decline in levels of a serum biomarker, aflatoxin albumin adducts, during the intervention stage of the clinical trial (22). This article confirms the multiple mechanisms of action of oltipraz by detailing the inhibitory effects of this high-dose intermittent oltipraz intervention on levels of AFM 1 (the major phase 1 oxidative metabolite of aflatoxin) in the urine of these study participants during this period. Moreover, low-dose sustained treatment with oltipraz leads to increased excretion of a major phase 2 metabolite, AFB NAC. SUBJECTS AND METHODS Overall Design and Structure The phase IIa chemoprevention trial with oltipraz was a randomized, placebocontrolled, double-blind study. Signed informed consent was obtained from all participants in accordance with institutional and federal guidelines of the People s Republic of China and of the United States. Two hundred forty adults in good general health without any history of major chronic illnesses and with detectable serum aflatoxin albumin adduct levels at baseline were randomly assigned into one of three intervention arms: A) placebo, B) 125 mg oltipraz administered daily, or C) 500 mg oltipraz administered weekly. The trial included men and women and did not exclude individuals positive for hepatitis B virus surface antigen who had evidence of normal liver function. The rationale, methods, participant characteristics, compliance, adverse events, and initial results on modulation of biomarkers from this trial have been reported (21 23). Study participants were recruited from Daxin Township, Qidong. Daxin is a rural farming community of approximately residents and is located at the mouth of the Yangtze River, 15 km southeast of Qidong. After an initial screening of 1006 residents, 233 eligible participants actually reported to the Daxin Medical Clinic on the first day of the study (July 9, 1995), where they completed another physical examination and provided blood and urine samples. One additional participant, included in the randomization scheme, missed this initial visit but was allowed to participate starting at week 3. All study participants remained eligible as determined on-site and were given their randomized identification number and their first dose of study drug at the clinic. Thereafter, two identical capsules containing either placebo (arm A) or active drug (125 [arm B] or 250 [arm C] mg oltipraz) were administered daily for 8 weeks. In practice, each daily administration in intervention arm B contained one capsule with 125 mg oltipraz and one placebo capsule. In intervention arm C, individuals received 500 mg oltipraz (two 250- mg capsules of oltipraz) on the first day of each weekly cycle, followed by two placebo capsules per day for the next 6 days. Blood and urine samples, collected throughout the intervention and follow-up periods, provided the basis for measuring aflatoxin biomarkers. Overnight urine samples were collected on the second, third, and fourth mornings of weekly cycles 1, 3, 5, 9, 13, 15, and 17. Urine samples were collected at the participants homes by village doctors and delivered by motorbike courier to the Qidong Liver Cancer Institute by midmorning of each collection day at which time they were logged in, distributed into several tubes, and frozen at 20 C. At the conclusion of the clinical trial, samples were shipped by air to Baltimore, MD, and stored at 80 C before assay. Analysis of Urinary Levels of AFM 1 and AFB NAC Five milliliters of urine was adjusted to an acidic ph with 0.5 ml of 1 M ammonium formate (ph 4.5), and the volume was increased to 10 ml with water. The sample was then applied to a 1-mL preparative monoclonal antibody column at a flow rate of 0.3 ml/minute as described previously (24,25). The affinity column was then washed twice with 5 ml of phosphate-buffered saline (ph 7.4) and once with 10 ml of water to remove nonspecifically bound materials. Aflatoxin derivatives were eluted from the immunoaffinity column with 2 ml of 80% methanol in water. The eluate was reduced to about 100 L with an argon stream and mixed with an equal amount of 5 mm triethylammonium formate (ph 3.0) before analysis by high-performance liquid chromatography (HPLC). AFM 1 and AFB NAC were analyzed by reversed-phase HPLC on a Hewlett- Packard model 1040A diode-array detector connected in series with a Dynmax FL-2 fluorescence detector (366-nm excitation wavelength and 436-nm emission wavelength) to quantify aflatoxin metabolites. The HPLC column used was a C m (4 250 mm) Microsorb analytical column (Rainin Inst. Co., Woburn, MA), and chromatographic separation was obtained by a 5% 25% ethanol linear gradient in water generated over a 25-minute period followed by isocratic elution with 25% ethanol in water, all at a flow rate of 1 ml/minute. The mobile phase was buffered with 5-mM triethylammonium formate (ph 3.0), and the column temperature was maintained at 45 C. The eluted peaks were integrated and AFM 1 and AFB NAC were quantitated with the regression formulae obtained from standard curves for each metabolite. Authentic AFB NAC (25) was eluted at 27.1 minutes and AFM 1 was eluted at 34.4 minutes. The limit of detection of this fluorescence HPLC method was 0.5 pg for AFM 1 and 2 pg for AFB NAC. The experimental conditions for metabolite analyses were optimized during method development for several parameters, including volume of urine sample, size of immunoaffinity column, and constitutents of the immunoaffinity resin. An equal admixture of two monoclonal antibodies (2B11 and 2F5) was used. The capacity of the column to bind aflatoxin derivatives was assessed with 348 ARTICLES Journal of the National Cancer Institute, Vol. 91, No. 4, February 17, 1999

3 analysis by liquid chromatography mass spectrometry. A Thermal Systems Products HPLC was used to provide a constant flow of 200 L/minute to an ODS J-sphere M-80 column (2 250 mm) (YMC, Inc., Wilmington, NC). A gradient starting at 4% acetonitrile/2% methanol and finishing at 13% acetonitrile/12% methanol in 25 minutes was used for separating aflatoxins. A buffer containing 0.1% formic acid (ph 2.5) was used throughout the run. The HPLC column was maintained at 55 C and the column effluent was directed through a UV detector (365 nm) and into the electrospray ionization interface on the mass spectrometer. The sample was scanned from 200 to 600 atomic mass units at 1 second per scan. A collision energy of 22% was used for collision reaction monitoring. Statistical Analyses Fig. 2. A) Representative chromatogram of aflatoxin M 1 (AFM 1 ) immunopurified from 5 ml urine collected from a study participant and detected with a high-performance liquid chromatography (HPLC) fluorescence system. B) Full positive-ion mass spectrum of immunoaffinitypurified and HPLC-separated AFM 1. C) Positive-ion collision reaction monitoring of the m/z 329 ion shown in B. D) Full positive-ion mass spectrum of immunoaffinity-purified and HPLCseparated aflatoxin B 1 mercapturic acid (AFB NAC). E) Positive-ion collision reaction monitoring of the m/z 492 ion shown in D. AFM 1 and [ 3 H]AFB 1. Up to 500 ng of AFM 1 applied to a 1-mL affinity column could be recovered at rates of 90% 98%. To ensure quality control of the analysis, all measurements were conducted on blinded samples and 5% of the samples were analyzed pairwise with spiked (0.1 ng of AFM 1 added) and unspiked sample for the same individuals. Cross-analysis with a separate HPLC fluorescence system was also carried out on all outlier samples. The variances between the pairwise and cross-analyses were less than 5%. Characterization of Urinary AFM 1 and AFB NAC A Finnigan LCQ liquid chromatography mass spectrometry system was used to perform electrospray ionization mass spectrometry in positive-ion mode to confirm the identity of AFM 1 and AFB NAC. The elution fractions of either AFM 1 or AFB NAC from the HPLC fluorescence system were collected and loaded onto a Waters Oasis column (3 ml) to remove salts. The column eluate was then reduced to about 100 L with an ultra-high-purity argon stream before The distributions of the levels of both urinary metabolites were highly skewed. Therefore, nonparametric Wilcoxon rank sum tests were used to compare urinary metabolite levels in each treatment arm to the results in the placebo arm. For this analysis, values of L/2 (where L is the limit of detection) for AFM 1 and AFB NAC were inserted for all nondetect values, as described by Hornung and Reed (26). To test whether lower levels of AFM 1 excretion at week 5 predicted the overall change in aflatoxin albumin adducts from baseline through week 13, we used the slope of adducts in each individual as the outcome and, by linear regression methods, determined whether the individuals with the lowest levels of urinary AFM 1 at week 5 had the fastest declining adducts. Because a decline in aflatoxin albumin adducts had only been observed in those individuals randomly assigned to receive the weekly dose of 500 mg oltipraz (22), this latter analysis was restricted to that group. All reported P values are two-sided. RESULTS Validation of the Analytic Method Fig. 2, A, shows a representative chromatogram of AFM 1 and AFB NAC immunopurified from 5 ml of urine collected from a study participant and detected by the HPLC fluorescence system. Peaks corresponding to AFM 1 and AFB NAC by retention time and fluorescence characteristics were then isolated from multiple urine samples and subjected to liquid chromatography mass spectrometry analysis. Fig. 2, B, shows the mass spectrum of 200 pg of AFM 1 isolated from these urine samples. The positive molecular ion (MH + )atm/z 329 is identical to that observed with authentic standard (18). Fig. 2, C, shows the data obtained by collision reaction monitoring of the m/z 329 ion. The major fragmentation ions for the m/z 329 ion are m/z 301, 273, and 259, and they reflect the loss of CO +, loss of another CO + from the m/z 301 ion, and loss of CH 2 from the m/z 287 ion, respectively. Fig. 2, D, shows the mass spectrum of AFB-NAC isolated from urine. The molecular ion m/z 492 is identical to that observed with authentic standard and the mercapturic acid characterized in the urine of rats receiving AFB 1 (25). A monosodium adduct at m/z 514 is also observed. Collision reaction monitoring of the m/z 492 ion is shown in Fig. 2, E. The major fragmentation ion is m/z 329, which reflects the loss of the N-acetylcysteine group. Characteristic aflatoxin fragments at m/z 311 and 271 are also seen. Diminution of Urinary AFM 1 (Phase 1) Excretion Analyses were conducted on 189 urine samples collected during week 5, the midpoint of the active intervention phase of the Journal of the National Cancer Institute, Vol. 91, No. 4, February 17, 1999 ARTICLES 349

4 study. Because of the short half-life of urinary aflatoxin metabolites, this period was judged a priori as most likely to reveal possible treatment-related effects on biomarker levels. Samples collected at other time points have not been analyzed for urinary aflatoxin biomarkers. As shown in Table 1, 197 individuals remained active participants in the intervention trial at this time. One hundred ninety-five of these participants provided an overnight urine sample on the morning of the second day of weekly cycle 5. Six of these samples had abnormally low levels of creatinine (<12 mg/dl) and were excluded from further analysis. Review of study compliance, as judged by pill counts, indicated that all but four of these 189 individuals took their assigned capsules during the 48 hours before this urine collection. Thus, in practice, urine was collected hours after administration of the weekly dose of 500 mg oltipraz and hours after administration of the most recent of the daily doses of 125 mg of oltipraz. AFM 1 could be detected in 154 (81.5%) of the 189 urine samples. Samples in which AFM 1 could not be detected were distributed evenly between arms, accounting for 19.4% (14 of 72 samples), 17.5% (10 of 57 samples), and 18.3% (11 of 60 samples) of the placebo arm and the arms receiving 125 mg oltipraz daily and 500 mg oltipraz weekly, respectively. Fig. 3 (left) shows the distributions of the levels of AFM 1 in the three intervention arms. The median level of AFM 1 in the urine of participants receiving placebo was 9.3 pg/mg of creatinine, with a range from nondetectable to pg/mg. Administration of 125 mg oltipraz daily for 4 weeks had no statistically significant (P =.682) effect on urinary excretion of AFM 1 (median, 7.1 pg/mg; range, nondetectable to 70.3 pg/mg). However, administration of 500 mg oltipraz once a week for 4 weeks before collection led to a statistically significant (P.030) 51% reduction in the amount of AFM 1 excreted (median, 4.6 pg/mg; range, nondetectable to 25.3 pg/mg). In a previous report (22), we have shown that there was a statistically significant (P.008) weekly decline of serum aflatoxin albumin adduct levels in the group receiving weekly oltipraz. Comparisons of these slopes of curves showing changes in the levels of aflatoxin albumin adducts as a function of time in individuals randomly assigned to the arm receiving 500 mg oltipraz with the levels of AFM 1 excretion presented in this Participants Table 1. Study follow-up by intervention group Placebo arm Oltipraz arms 125 mg daily 500 mg weekly Total Randomly assigned, No Returned to clinic for beginning * of trial, No. Active participants at week 5, No Provided urine sample at week , No. Sufficient creatinine for biomarker analyses ( 12 mg/dl), No *Six eligible individuals who were randomized into the study subsequently declined to participate at the beginning of the intervention. Twenty-six of the withdrawals during the first month of the intervention were determined in the field to be related to drug treatment and were distributed as two, 15, and nine in the placebo arm, the 125-mg arm, and the 500-mg arm, respectively. The primary reason for withdrawal by the other 11 participants was that they were tired of participating (21). study did show a moderate association between these two biomarkers. Specifically, for the lowest (<2.96 pg/mg), middle ( pg/mg), and highest (>7.36 pg/mg) tertile groups of AFM 1 excretion at week 5, the mean values of the slopes of the adducts were , , and pmol/week. Individuals ranked in the two lowest tertiles of AFM 1 levels showed similar rates of decline in aflatoxin albumin adducts, with those in the lowest tertile showing a borderline nonstatistically significant decline (P.078) when compared with those in the highest tertile. Elevation of AFB NAC (Phase 2) Excretion AFB NAC could be detected in 124 (65.6%) of the 189 urine samples, with the samples in which AFB NAC was not detected distributed as 43.1% (31 of 72 samples), 21.1% (12 of 57 samples), and 35.0% (21 of 60 samples) of the placebo arm and arms receiving 125 mg oltipraz daily and 500 mg oltipraz weekly, respectively. Fig. 3 (right) shows the distributions of the levels of AFB NAC in the three intervention arms. The median level of AFB NAC in the urine of participants receiving placebo was 7.1 pg/mg of creatinine, with a range of nondetectable to pg/mg. Administration of 125 mg oltipraz daily for 4 weeks before collection led to a statistically significant (P.017) elevation in the amount of AFB NAC excreted (median, 18.6 pg/mg; range, nondetectable to pg/mg). This increase was primarily driven by the diminished number of nondetectable values within this treatment group. By contrast, administration of 500 mg oltipraz once a week for 4 weeks had no statistically significant (P.871) effect on urinary excretion of AFB NAC (median, 8.3 pg/ml; range, nondetectable to pg/ml). DISCUSSION As summarized in Fig. 4, oltipraz exerts multiple effects on the metabolism of AFB 1 in people exposed to this potent hepatocarcinogen. Dose- and schedule-dependent inhibition of cytochrome P450-dependent activation and induction of GSTmediated detoxification of AFB 1 have been observed in this clinical trial. Induction of phase 2 enzymes, particularly GSTs, has long been suggested as an important mechanism for achieving chemoprevention (27,28). Indeed, several novel chemopreventive agents, including oltipraz (29) and sulforaphane (30), the active principle in broccoli sprouts (31), were initially identified as candidate chemopreventive agents based on their phase 2 enzyme-inducing properties. Also, overexpression of GSTs renders V79 cells resistant to nucleic acid alkylation by AFB 1 (32). Thus, the present finding that daily treatment with 125 mg oltipraz elevated median urinary levels of AFB NAC 2.6-fold compared with a placebo confirms the functional ability of oltipraz to induce phase 2 enzymes and thereby alter carcinogen disposition in humans, as predicted from studies in rodents in vivo and in primary cultures of rodent and human hepatocytes in vitro (9,12 14,16). Administration of similar doses of oltipraz to people during phase I chemoprevention trials has lead to elevations in the levels of messenger RNA transcripts and the specific activities of several phase 2 enzymes in surrogate tissues, such as lymphocytes, and possible target tissues, such as colonic mucosa (17,20). Overall, these findings establish a proof of principle and provide strong justification for mechanism-based ap- 350 ARTICLES Journal of the National Cancer Institute, Vol. 91, No. 4, February 17, 1999

5 Fig. 3. Box plots (with horizontal lines at 25th, 50th, and 75th percentiles and bars at the 10th and 90th percentiles) of the distribution of excreted amounts of aflatoxin M 1 (AFM 1 )(left) and aflatoxin B 1 -mercapturic acid (AFB NAC) (right) after 4 weeks of the intervention. Urine samples from 72, 57, and 60 participants in the placebo arm and the arms receiving 125 mg oltipraz daily and 500 mg oltipraz weekly, respectively, were collected. Aflatoxins in the urine were separated by sequential immunoaffinity chromatography and liquid chromatography, and material in the eluate was detected with the fluorescence detection system. AFM 1 was not detected in 19.4%, 17.5%, and 18.3% of samples and AFB NAC was not detectable in 43.1%, 21.1%, and 35.0% of samples in the three arms, respectively. Fig. 4. Summary of the effects the different dose schedule regimens of oltipraz on the urinary excretion of phase 1 (aflatoxin M 1 [AFM 1 ]) and phase 2 (aflatoxin B 1 -mercapturic acid [AFB NAC]) metabolites of aflatoxin B 1 (AFB 1 ). The median levels of AFM 1 and AFB NAC in the treatment arms are shown relative to levels in the placebo arm. AFBO aflatoxin-8,9-epoxide; AFB SG aflatoxin glutathione conjugate. All P values are two-sided. proaches for the identification and development of chemopreventive agents. Somewhat surprising is the apparent inability of the highdose weekly regimen of 500 mg oltipraz to elevate urinary levels of AFB NAC. However, this seemingly anomalous outcome may simply reflect a masking effect exerted by a second mechanism of action for oltipraz, namely, inhibition of cytochrome P450 activities. Weekly administration of 500 mg oltipraz led to substantial reductions in the excretion of AFM 1, the major oxidative metabolite of aflatoxin. Oltipraz is largely a competitive inhibitor of CYP1A2, the principal enzyme involved in the metabolism of AFM 1 from AFB 1 with an apparent K i of 10 M (16). Pharmacokinetic studies conducted during phase I trials indicated that peak plasma concentrations of approximately 1 M oltipraz occurred with a dose of 125 mg oltipraz, whereas administration of 500 mg produced peak plasma concentrations of 20 M (17). Thus, if peak plasma concentrations reflect intracellular concentrations of the drug, then diminution of urinary AFM 1 levels with the 500-mg dose, but not the 125-mg dose, is consistent with the enzyme kinetics and pharmacokinetics data. As shown in Fig. 4, aflatoxin-8,9-epoxide, which is principally formed by CYP1A2 at ambient environmental exposures to AFB 1 (33), serves as a substrate for GSTs (15). Further metabolism of the aflatoxin glutathione conjugate leads to AFB NAC, the ultimate excretion product. Inhibition of initial substrate formation (i.e., aflatoxin-8,9- epoxide), which is only seen in the group receiving 500 mg oltipraz weekly, would likely mask any elevation of GST activities that may have occurred. Indeed, Langouët et al. (16) reported twofold to fourfold elevations in the levels of and classes of GST proteins in primary cultures of human hepatocytes containing a concentration of oltipraz comparable to that in plasma of humans treated with a single dose of 500 mg oltipraz. However, this inductive effect was not associated with an increased formation of aflatoxin glutathione conjugate because it was overridden by an inhibitory effect of oltipraz on AFB 1 activation by CYP1A2. Thus, elevations in glutathione or mercapturic acid conjugates may only be observed under circumstances in which rates of formation of aflatoxin-8,9-epoxide are unchanged, as appears to be the case in the group receiving 125 mg oltipraz daily. More detailed kinetic analyses of the entire pathway, including determination of the rate-limiting step, will be required to confirm this hypothesis. The persistence or transience of the effects of 500 mg oltipraz administered weekly on excretion of AFM 1 and possible masking of GST induction will be assessed in a follow-up phase IIb intervention trial by examining urinary biomarker levels at different times relative to drug administration. Because AFM 1 is formed by the same cytochrome P450 that yields the 8,9-epoxide, AFM 1 may serve as a reasonable surrogate for the genotoxic potential of aflatoxin exposures in individuals. Such a possibility has been examined in residents of Fusui County, Guangxi Autonomous Region, People s Republic of China, where a high incidence of HCC has also been reported. Zhu et al. (34) analyzed AFM 1 concentration in urine samples by enzyme-linked immunoabsorbent assay and noted correlations between levels of AFM 1 excretion and levels of AFB 1 in corn and peanut oil samples collected from different households. Using immunoaffinity and HPLC methods, Groopman et al. (24) have observed that measurements of urinary excretion of AFM 1 Journal of the National Cancer Institute, Vol. 91, No. 4, February 17, 1999 ARTICLES 351

6 and the labile DNA adduct aflatoxin-n 7 -guanine showed strong and highly statistically significant correlations with aflatoxin intake in this region. Moreover, direct associations between excretion of AFM 1 and levels of aflatoxin albumin adducts in individuals were noted in this ecologic survey (35). In a prospective, nested, case control study, Qian et al. (4) reported that the relative risk of HCC for individuals whose urine contained AFM 1 was 4.4 (95% confidence interval ) compared with those not excreting this biomarker. Similarly, Yu et al. (36) reported a statistically significant dose response relationship between urinary AFM 1 levels and HCC. The odds ratio encompassing the highest with the lowest tertile of AFM 1 levels was 6.0 (95% confidence interval ). Thus, urinary levels of AFM 1 may provide some index of altered risk for use in chemopreventive interventions. Indeed, Scholl et al. (18) have reported that excretion levels of AFM 1 fell precipitously when oltipraz was administered to rats continuously exposed to AFB 1. Diminished AFM 1 excretion persisted for the duration of the intervention but rebounded rapidly when the intervention was discontinued. This transient nature of the inhibition in vivo reflects the competitive inhibition of CYP1A2 by oltipraz and the relatively short half-life ( 24 hours) of the target enzyme (16,37). There is a moderate association between downward modulation of the slopes of curves for serum aflatoxin albumin adducts as a function of time and diminished excretion of AFM 1 in the urine of study participants randomly assigned to the arm receiving 500 mg oltipraz weekly. This association suggests that inhibition of cytochrome P450 activity leads to the observed decline in serum aflatoxin albumin adducts. By contrast, enhanced glutathione conjugation of the 8,9-epoxide to yield AFB NAC does not appear to be associated with altered aflatoxin albumin adduct levels. A comprehensive analysis of the predictive value of the aflatoxin albumin biomarker has been conducted in rats. This analysis indicated a reasonable association between the level of the aflatoxin albumin biomarker and risk of HCC at the population level but no association between biomarker levels and individual risks of HCC (11). Moreover, measurements of adduct levels, in both DNA and protein, consistently underestimate the chemopreventive efficacy of oltipraz in rodent aflatoxin hepatocarcinogenesis models (9 11). Thus, adduct biomarkers, particularly the aflatoxin albumin adduct, do not provide the full picture. Reduction in AFM 1 formation may have direct consequences in addition to those reflected in its role as a surrogate marker for diminished aflatoxin genotoxicity. Although AFM 1 is considerably less active as a hepatocarcinogen in the rat (38), it is equipotent to AFB 1 as a hepatotoxin (39). Perhaps reduction in AFM 1 production by oltipraz and the attenuation of the contribution of AFM 1 to the cytotoxic autopromoting component of aflatoxin hepatocarcinogenesis are important elements of the overall chemopreventive outcome. Other oxidative metabolites of aflatoxin (e.g., aflatoxin P 1 and aflatoxin Q 1 ) are much less toxic than AFB 1 or AFM 1 ; however, it is not known whether inhibition of CYP1A2 by oltipraz appreciably shunts aflatoxin metabolism to other cytochrome P450 enzymes. Finally, additional chemopreventive mechanisms unrelated to effects on the metabolism of aflatoxin have also been identified for oltipraz (40 42). Measurement of urinary levels of AFM 1 can serve as a biomarker for aflatoxin exposure. Cheng et al. (43) using similar analytic methodologies have recently reported on excretion rates of urinary AFM 1 in 69 counties throughout rural China. They reported the mean and highest levels of AFM 1 excretion to be 3.2 and 108 ng per 12 hours, respectively. If an excretion rate of 0.7 g of creatinine per 12 hours (normal range, g per 12 hours) is assumed, then the mean level of excretion in the placebo arm of this trial would be 12 ng per 12 hours. The nearly fourfold higher excretion rate in the present study reflects the selection of Qidong as a study site because of the known high prevalence of aflatoxin-contaminated foods in the diet and the elevated risk of HCC in this region. In the study by Cheng et al. (43), investigators intentionally dosed themselves with 1.0 gof pure AFB 1 and determined that 5% 6% of the administered dose was excreted over the subsequent week as AFM 1. Other reports from field studies have estimated that 1% 2% of ingested AFB 1 is eliminated as AFM 1 in humans (34). By extrapolation, it can be inferred that the AFB 1 exposure of the residents of Daxin Township in Qidong in the summer of 1995 was approximately 1 2 g/day. However, because of the heterogeneity of aflatoxin contamination of foodstuffs, it is likely that there is large interand intra-individual variation in this estimate. Nonetheless, this estimate is less than half that reported for the region 15 years earlier (44) and may reflect a switch from corn to rice as the primary dietary staple within the past decade. After rising steadily during the 1970s and 1980s, age-adjusted incidence rates for HCC in Qidong have plateaued and perhaps begun a modest decline over the past few years. Whether such trends reflect reduced aflatoxin exposures, the implementation of hepatitis B virus vaccination programs, or additional combinations of factors remains unclear. Nonetheless, incidence rates for HCC remain untenably high. If sustainable over the long term, statistically significant alterations in the formation and fate of the ultimate carcinogenic metabolite aflatoxin-8,9-epoxide (as brought about through multiple mechanisms with oltipraz at one point in this intervention) could provide substantive protective effects against the adverse actions of aflatoxin in this population. A follow-up 12-month phase IIb intervention with oltipraz will be conducted in this region from 1999 through The primary goal will be to assess the full extent and persistence of the initial modulation of aflatoxin biomarkers seen in the phase IIa trial. Participants will be randomly assigned to receive placebo or 250 or 500 mg oltipraz once a week. The exclusive selection of a weekly schedule is driven in part by the effects of 500 mg oltipraz weekly on the urinary and serum aflatoxin biomarkers, as well as some practical considerations. The weekly schedule is likely to improve compliance, both by attenuating the intense monitoring required with daily administration of study drug and by reducing the occurrence of adverse events. In this phase IIa trial, fewer adverse events were reported by individuals in the weekly arm than in the daily arm (21), perhaps reflecting the lower cumulative dose. Drug costs are also substantially reduced in a weekly intervention, rendering the widespread use of oltipraz in high-risk populations more feasible. The upcoming phase IIb trial is also an intermediate step in the development of chemopreventive strategies in that phase III studies with a duration of 5 years or more will ultimately be required to establish the extent to which a decrease in the concentration of aflatoxin biomarkers translates into a reduction or delay in the development of HCC. REFERENCES (1) Zhu YR, Chen JG, Huang XY. Hepatocellular carcinoma in Qidong County. In: Tang ZY, editor. Primary liver cancer. Beijing (China): Springer-Verlag; p ARTICLES Journal of the National Cancer Institute, Vol. 91, No. 4, February 17, 1999

7 (2) Yu SZ. Primary prevention of hepatocellular carcinoma. J Gastroenterol Hepatol 1995;10: (3) Ross RK, Yuan JM, Yu MC, Wogan GN, Qian GS, Tu JT, et al. Urinary aflatoxin biomarkers and risk of hepatocellular carcinoma. Lancet 1992; 339: (4) Qian GS, Ross RK, Yu MC, Yuan JM, Gao YT, Henderson BE, et al. A follow-up study of urinary markers of aflatoxin exposure and liver cancer risk in Shanghai, People s Republic of China. Cancer Epidemiol Biomarkers Prev 1994;3:3 10. (5) Wang LY, Hatch M, Chen CJ, Levin B, You SL, Lu SN, et al. Aflatoxin exposure and risk of hepatocellular carcinoma in Taiwan. Int J Cancer 1996;67: (6) Chang MH, Chen CJ, Lai MS, Hsu HM, Wu TC, Kong MS, et al. Universal hepatitis B vaccination in Taiwan and the incidence of hepatocellular carcinoma in children. Taiwan Childhood Hepatoma Study Group. N Engl J Med 1997;336: (7) Kensler TW, Davis EF, Bolton MG. Strategies for chemoprotection against aflatoxin-induced liver cancer. In: Eaton DL, Groopman JD, editors. The toxicology of aflatoxins: human health, veterinary, and agricultural significance. San Diego (CA): Academic Press; p (8) Hayes JD, McLeod R, Ellis EM, Pulford DJ, Ireland LS, McLellan LI, et al. Regulation of glutathione S-transferases and aldehyde reductase by chemoprotectors: studies of mechanisms responsible for inducible resistance to aflatoxin B 1. IARC Sci Publ 1996: (9) Roebuck BD, Liu YL, Rogers AE, Groopman JD, Kensler TW. Protection against aflatoxin B 1 -induced hepatocarcinogenesis in F344 rats by 5-(2- pyrazinyl)-4-methyl-1,2-dithiole-3-thione (oltipraz): predictive role for short-term molecular dosimetry. Cancer Res 1991;51: (10) Groopman JD, DeMatos P, Egner PA, Love-Hunt A, Kensler TW. Molecular dosimetry of urinary aflatoxin-n 7 -guanine and serum aflatoxin albumin adducts predicts chemoprotection by 1,2-dithiole-3-thione in rats. Carcinogenesis 1992;13: (11) Kensler TW, Gange SJ, Egner PA, Dolan PM, Munoz A, Groopman JD, et al. Predictive value of molecular dosimetry: individual versus group effects of oltipraz on aflatoxin albumin adducts and risk of liver cancer. Cancer Epidemiol Biomarkers Prev 1997;6: (12) Kensler TW, Egner PA, Dolan PM, Groopman JD, Roebuck BD. Mechanism of protection against aflatoxin tumorigenicity in rats fed 5-(2-pyrazinyl)-4-methyl-1,2-dithiol-3-thione (oltipraz) and related 1,2-dithiol-3-thiones and 1,2-dithiol-3-ones. Cancer Res 1987;47: (13) Davidson NE, Egner PA, Kensler TW. Transcriptional control of glutathione S-transferase gene expression by the chemoprotective agent 5-(2- pyrazinyl)-4-methyl-1,2-dithiole-3-thione (oltipraz) in rat liver. Cancer Res 1990;50: (14) Morel F, Fardel O, Meyer DJ, Langouet S, Gilmore KS, Meunier B, et al. Preferential increase of glutathione S-transferase class alpha transcripts in cultured human hepatocytes by phenobarbital, 3-methylcholanthrene, and dithiolethiones. Cancer Res 1993;53: (15) Raney KD, Meyer DJ, Ketterer B, Harris TM, Guengerich FP. Glutathione conjugation of aflatoxin B 1 exo- and endo-epoxides by rat and human glutathione S-transferases. Chem Res Toxicol 1992;5: (16) Langouet S, Coles B, Morel F, Becquemont L, Beaune P, Guengerich FP, et al. Inhibition of CYP1A2 and CYP3A4 by oltipraz results in reduction of aflatoxin B 1 metabolism in human hepatocytes in primary culture. Cancer Res 1995;55: (17) Gupta E, Olopade OI, Ratain MJ, Mick R, Baker TM, Berezin FK, et al. Pharmacokinetics and pharmacodynamics of oltipraz as a chemopreventive agent. Clin Cancer Res 1995;1: (18) Scholl P, Musser SM, Kensler TW, Groopman JD. Inhibition of aflatoxin M 1 excretion in rat urine during dietary intervention with oltipraz. Carcinogenesis 1996;17: (19) Benson AB 3d. Oltipraz: a laboratory and clinical review. J Cell Biochem Suppl 1993;17F: (20) O Dwyer PJ, Szarka CE, Yao KS, Halbherr TC, Pfeiffer GR, Green F, et al. Modulation of gene expression in subjects at risk for colorectal cancer by the chemopreventive dithiolethione oltipraz. J Clin Invest 1996;98: (21) Jacobson LP, Zhang BC, Zhu YR, Wang JB, Wu Y, Zhang QN, et al. Oltipraz chemoprevention trial in Qidong, People s Republic of China: study design and clinical outcomes. Cancer Epidemiol Biomarkers Prev 1997;6: (22) Kensler TW, He X, Otieno M, Egner PA, Jacobson LP, Chen B, et al. Oltipraz chemoprevention trial in Qidong, People s Republic of China: modulation of serum aflatoxin albumin adduct biomarkers. Cancer Epidemiol Biomarkers Prev 1998;7: (23) Zhang BC, Zhu YR, Wang JB, Wu Y, Zhang QN, Qian GS, et al. Oltipraz chemoprevention trial in Qidong, Jiangsu Province, People s Republic of China. J Cell Biochem Suppl 1997;28/29: (24) Groopman JD, Zhu JQ, Donahue PR, Pikul A, Zhang LS, Chen JS, Wogan GN. Molecular dosimetry of urinary aflatoxin DNA adducts in people living in Guangxi Autonomous Region, People s Republic of China. Cancer Res 1992;52: (25) Scholl PF, Musser SM, Groopman JD. Synthesis and characterization of aflatoxin B 1 mercapturic acids and their identification in rat urine. Chem Res Toxicol 1997;10: (26) Hornung RD, Reed LD. Estimation of average concentration in the presence of nondetectable values. Appl Occup Environ Hyg 1990:5; (27) Benson AM, Batzinger RP, Ou SY, Bueding E, Cha YN, Talalay P. Elevation of hepatic glutathione S-transferase activities and protection against mutagenic metabolites of benzo(a)pyrene by dietary antioxidants. Cancer Res 1978;38: (28) Wattenberg LW. Chemoprevention of cancer. Cancer Res 1985;45:1 8. (29) Ansher SS, Dolan P, Bueding E. Chemoprotective effects of two dithiolthiones and of butylhydroxyanisole against carbon tetrachloride and acetaminophen toxicity. Hepatology 1983;3: (30) Zhang Y, Talalay P, Cho CG, Posner GH. A major inducer of anticarcinogenic protective enzymes from broccoli: isolation and elucidation of structure. Proc Natl Acad Sci U S A 1992;89: (31) Fahey JW, Zhang Y, Talalay P. Broccoli sprouts: an exceptionally rich source of inducers of enzymes that protect against chemical carcinogens. Proc Natl Acad Sci U S A 1997;94: (32) Townsend AJ, Fields WR, Haynes RL, Doss AJ, Li Y, Doehmer J, Morrow CS. Chemoprotective functions of glutathione S-transferases in cell lines induced to express specific isozymes by stable transfection. Chem Biol Interact 1998; : (33) Gallagher EP, Kunze KL, Stapleton PL, Eaton DL. The kinetics of aflatoxin B 1 oxidation by human cdna-expressed and human liver microsomal cytochromes P450 1A2 and 3A4. Toxicol Appl Pharmacol 1996; 141: (34) Zhu JQ, Zhang LS, Hu X, Xiao Y, Chen JS, Xu YC, et al. Correlation of dietary aflatoxin B 1 levels with excretion of aflatoxin M 1 in human urine. Cancer Res 1987;47: (35) Gan LS, Skipper PL, Peng XC, Groopman JD, Chen JS, Wogan GN, Tannenbaum SR. Serum albumin adducts in the molecular epidemiology of aflatoxin carcinogenesis: correlation with aflatoxin B 1 intake and urinary excretion of aflatoxin M 1. Carcinogenesis 1988;9: (36) Yu MW, Lien JP, Chiu YH, Santella RM, Liaw YF, Chen CJ. Effect of aflatoxin metabolism and DNA adduct formation on hepatocellular carcinoma among chronic hepatitis B carriers in Taiwan. J Hepatol 1997;27: (37) Shiraki H, Guengerich FP. Turnover of membrane proteins: kinetics of induction and degradation of seven forms of rat liver microsomal cytochrome P-450, NADPH-cytochrome P-450 reductase, and epoxide hydrolase. Arch Biochem Biophys 1984:235: (38) Wogan GN, Paglialunga S. Carcinogenic effects of low dietary levels of aflatoxin M 1 in rats. Food Cosmet Toxicol 1974:12; (39) Pong RS, Wogan GN. Toxicity and biochemical fine structural effects of synthetic aflatoxins M 1 and B 1 in rat liver. J Natl Cancer Inst 1971: 47; (40) Primiano T, Kensler TW, Kuppusamy P, Zweier JL, Sutter TR. Induction of hepatic heme oxygenase-1 and ferritin in rats by cancer chemopreventive dithiolethiones. Carcinogenesis 1996;17: (41) Primiano T, Li Y, Kensler TW, Trush MA, Sutter TR. Identification of dithiolethione-inducible gene-1 as a leukotriene B 4 12-hydroxydehydrogenase: implications for chemoprevention. Carcinogenesis 1998: 19; (42) O Dwyer PJ, Johnson SW, Khater C, Krueger A, Matsumoto Y, Hamilton TC, et al. The chemopreventive agent oltipraz stimulates repair of damaged DNA. Cancer Res 1997;57: Journal of the National Cancer Institute, Vol. 91, No. 4, February 17, 1999 ARTICLES 353

8 (43) Cheng Z, Root M, Pan W, Chen JS, Campbell TC. Use of an improved method for analysis of urinary aflatoxin M 1 in a survey of mainland China and Taiwan. Cancer Epidemiol Biomarkers Prev 1997;6: (44) Sun TT, Chu YY. Carcinogenesis and prevention strategy of liver cancer in areas of prevalence. J Cell Physiol Suppl 1984;3: NOTES Supported by Public Health Service (PHS) grants ES03819 and P01ES06052 (National Institute of Environmental Health Sciences) and by PHS grant CA06973 and PHS contract N01CN25437 (National Cancer Institute), National Institutes of Health, Department of Health and Human Services. We thank Drs. Gary Kelloff, James Crowell, Earnest Hawk, Charles Boone, Kenneth Olden, and Gary Gordon for their helpful discussions and Drs. Nancy Davidson, Mary Gorman, and Hans Prochaska for their medical oversight in Qidong. We also thank the staff of the Daxin Medical Clinic, the village doctors, and the residents of Daxin Township for their participation. Manuscript received July 20, 1998; revised December 3, 1998; accepted December 21, ARTICLES Journal of the National Cancer Institute, Vol. 91, No. 4, February 17, 1999