| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Article |
Departments of Obstetrics and Gynecology (A.W., Y.Ok., T.F.), Radiology (Y.Og.), and Medicine (T.S.), Kochi Medical School, Kochi 783-8505, Japan
Address all correspondence and requests for reprints to: Akihiko Wakatsuki, M.D., Department of Obstetrics and Gynecology, Kochi Medical School, Oko cho, Nankoku, Kochi 783-8505, Japan. E-mail: . wakatuki{at}kochi-ms.ac.jpm
Abstract
The purpose of the present study was to investigate the effects of tamoxifen on the size and oxidative susceptibility of low-density lipoprotein (LDL) particles in breast cancer patients with tamoxifen-induced fatty liver. We investigated the following breast cancer patients: 13 receiving no tamoxifen (group A), 13 receiving tamoxifen 40 mg daily but without fatty liver (group B), and 13 receiving tamoxifen 40 mg daily with fatty liver (group C). Plasma lipids and diameter of LDL particles were measured. Susceptibility of LDL to oxidation was analyzed by incubation with CuSO4 while monitoring conjugated diene formation and assaying thiobarbituric acid reactive substances (TBARS). Plasma total and LDL cholesterol concentrations in groups B and C were significantly lower than those in group A. In group C, concentrations of plasma triglyceride (TG) and TBARS were significantly greater, but LDL particle diameter and lag time for LDL oxidation were significantly smaller than those in groups A and B. Plasma TG concentrations correlated negatively with computed tomography ratio of liver to spleen (r = -0.76; P < 0.001). LDL particle diameter correlated negatively with plasma TG (r = -0.62; P < 0.001) and TBARS (r = -0.44; P < 0.01), but positively with LDL lag time (r = 0.47; P < 0.01). Tamoxifen-induced fatty liver in breast cancer patients may be atherogenic, via increased TG and consequent small, easily oxidized LDL particles.
TAMOXIFEN, A POTENT ANTAGONIST of estrogen, is widely used as adjuvant therapy for patients with breast cancer (1). Estrogen replacement therapy (ERT) has beneficial effects on plasma lipids (2), low-density lipoprotein (LDL) oxidation (2), hemostatic factors (3), and endothelial function (4). Epidemiological studies indicate that postmenopausal ERT significantly reduces mortality from coronary heart disease (CHD) and other cardiovascular disease (5, 6, 7). Accordingly, concern arises that tamoxifen, an antiestrogenic agent, might increase the risk of CHD. In fact, however, tamoxifen has been reported to decrease overall incidence of fatal myocardial infarction (8) through favorable effects on lipid metabolism via reduced plasma concentrations of total cholesterol (TC) and LDL cholesterol (LDL-C) (9). Nonetheless, in a study by Ogawa et al. (10), 24 of 66 patients (36%) with breast cancer who received tamoxifen for 3–5 yr developed fatty change of the liver. Because fatty change of the liver may be associated with hypertriglyceridemia, this subgroup of tamoxifen-treated breast cancer patients might have increased risk of CHD. Indeed, in contrast with favorable effects of tamoxifen on plasma TC and LDL-C concentrations, tamoxifen also has been reported to increase plasma triglyceride (TG) concentration (11). Thus, plasma TG concentration is likely to be elevated in tamoxifen-treated patients with fatty liver. However, lipid metabolism in breast cancer patients with tamoxifen-induced fatty liver has not been evaluated.
Hypertriglyceridemia has been considered a risk factor for CHD. McNamara et al. have suggested that plasma TG concentration is the single most important factor affecting size of LDL particles (12). We similarly have found that an increase in plasma TG concentration may reduce the size of LDL particles (13, 14). LDL are heterogeneous in size and density (15), and not all LDL subfractions are equally atherogenic. Small LDL particles are more susceptible than larger particles to oxidative modification (16), an initial step in atherogenesis, and small particles are associated with increased risk of CHD. Therefore, tamoxifen may be atherogenic in the patients who develop fatty liver from the drug.
In the present study, we investigated lipid metabolism, LDL particle diameter, and susceptibility of LDL to oxidative modification in breast cancer patients with tamoxifen-induced fatty liver, and compared these parameters with those in subjects without fatty liver who received adjuvant tamoxifen as well as those in patients not receiving tamoxifen.
Subjects and Methods
Subjects
We studied 39 postmenopausal breast cancer patients who satisfied the following conditions during this period. None of them smoked, used caffeine or alcohol, or had a history of hypertension, thyroid disease, liver disease, diabetes mellitus, or cardiovascular disease; and none were currently taking any medication known to influence lipoprotein metabolism. No subjects underwent exercise or dietary therapy during the study period. Written informed consent was obtained from each subject before admission to the study. The study design was approved by the ethics committee of Kochi Medical School.
Study design and determination of fatty liver
Abdominal computed tomography (CT) examination was carried out annually for breast cancer patients who received adjuvant tamoxifen, and we evaluated the CT attenuation number (Housefield Unit) ratio of liver to spleen. The criteria used for diagnosing fatty liver was the CT attenuation number ratio less than 0.9 (10). No subjects showed fatty changes of the liver before tamoxifen treatment.
We compared hormones, lipid metabolism, glucose, insulin, LDL particle diameter, and susceptibility of LDL to oxidation between three groups of breast cancer patients: group A consisted of 13 patients without adjuvant tamoxifen; group B included 13 patients not developing fatty liver but receiving adjuvant tamoxifen (40 mg daily) for 12–54 months; group C included 13 patients with fatty liver associated with adjuvant tamoxifen (40 mg daily) administered for 11–46 months. Venous blood samples were drawn into tubes containing 1 mg/ml EDTA between 0800 and 1000 h following a 12-h fast. Samples were centrifuged immediately at 1500 x g for 20 min at 4 C to obtain plasma.
Measurement of lipids, hormones, glucose, and insulin, and isolation of LDL
Plasma concentrations of TC and TG were measured by enzymatic methods as previously described (17). The concentration of HDL cholesterol (HDL-C) was determined by similar methods after apolipoprotein B-containing lipoproteins had been precipitated with sodium phosphotungstate in the presence of magnesium chloride (17). LDLs (density, 1.019–1.063) were fractionated from freshly drawn (<24 h) plasma samples by ultracentrifugation according to the method of Havel et al. (18). Concentrations of LDL-C were assayed enzymatically (17). Plasma concentrations of apolipoprotein AI, AII, and B were measured by a turbidimetric immunoassay (19). Plasma glucose and insulin concentrations were determined by hexokinase technique and RIA, respectively. Insulin resistance was estimated from the fasting glucose and insulin (20). Plasma concentrations of estradiol, LH, and FSH were measured by RIA.
LDL particle diameter
LDL was subjected to gradient gel electrophoresis using 2% to 15% nondenaturing polyacrylamide-agarose gels. The apparatus was filled with a buffer consisting of 0.025 mol/liter Tris, 0.192 mol/liter glycine, and 0.1% SDS at pH 8.4. After a 2-h pre-electrophoresis at 200 V, 5 µl aliquots of the LDL fraction (5–10 µg protein) were applied to the gel. A reference protein mixture (Molecular Weight Marker, Daiichi, Tokyo, Japan) and carboxylated latex beads (Duke Science, Palo Alto, CA) were used as standards for molecule size and mass. Electrophoresis was performed at 4 C for 26 h as follows: 2 h at 30 V, 12 h at 50 V, and 12 h at 150 V (21). The gels were then fixed in 40% acetic acid for 1 h, stained for 45 min in 0.1% Coomassie G-250 (Nacalai, Kyoto, Japan) prepared in 10% acetic acid and 30% ethanol, and then destained in a mixture of 7.5% acetic acid and 10% ethanol. The gels were subjected to a gentle horizontal rotation during the fixing, staining, and destaining. The distribution profile of the LDL subfractions was determined by densitometric scanning of gels at 633 nm (Shimadzu, Kyoto, Japan). The apparent diameter of the major LDL subfractions was determined by comparing the results with a calibration curve constructed using of ferritin (12.20 nm), thyroglobulin (17.00 nm), and latex beads (38.00 nm) as reference samples (21).
Susceptibility of LDL to oxidation
To remove EDTA, the isolated LDL fraction was dialyzed in darkness at 4 C for 48 h against 30 mM sodium phosphate buffer containing 150 mM NaCl, which was made oxygen-free by vacuum degassing followed by purging with nitrogen. The buffer was changed after 24 h of dialysis.
Analysis of conjugated dienes formation.
The EDTA-free dialyzed LDL subfraction was diluted with dialysis buffer to a final concentration of 200 µg/ml. The protein concentration of the LDL subfraction was determined using the method of Lowry et al. (22). Oxidation was initiated by the addition of 2.0 µM CuSO4. The kinetics of the formation of conjugated dienes was determined by monitoring the change in the absorbance at 234 nm on a Beckman Model DU 640 spectrophotometer equipped with a 12-position automatic sample changer. The absorbance at 234 nm was recorded at 37 C every 3 min for 4 h. The lag phase, propagation phase, and decomposition phase were determined as previously described (23). The lag time was defined as the interval between addition of CuSO4 and the intercept of the tangent of the slope of the absorbance curve with the time-scale axis during the propagation phase. The maximal oxidation rate was calculated from the slope of the tangent, using a molar extinction coefficient for conjugated dienes of 
234 = 29,500/mol/cm, and expressed as nanomoles of diene formed per minute per milligram of LDL protein. The maximal increase in absorbance was determined from the absorbance curve as the absorbance at the beginning of the decomposition phase minus the absorbance at the start of the lag phase. The corresponding amount of dienes was calculated as described for the oxidation rate (23).
Measurement of thiobarbituric acid reactive substances (TBARS) level. The EDTA-free dialyzed LDL subfraction (200 µg/ml) was oxidized by the addition of 5 µM CuSO4 and incubated at 37 C for 3 h. The concentrations of TBARS in the LDL subfraction, with and without addition of CuSO4, were determined according to the method of Ohkawa et al. (24). In brief, 1.5 ml 20% acetic acid (pH 3.5) and 1.5 ml of an 0.8% TBA solution were added to the LDL solution, and the volume was brought to 4.0 ml with distilled water. The mixture was shaken thoroughly and heated in an oil bath at 95 C for 60 min. After the mixture was cooled with tap water, 1.0 ml distilled water and 5.0 ml butyl alcohol and pyridine (15:1, vol/vol) were added, and the sample was shaken gently for 5 min. After centrifugation at 1500 x g for 10 min, the butyl alcohol-pyridine phase containing the TBARS was separated, and its absorbance was measured at 532 nm. The results were expressed as mole equivalent malondialdehyde per milligram of protein, using malondialdehyde from tetramethoxypropane as a standard and double-distilled water as a control.
Statistical analysis
Data are expressed as the mean ± SD. Differences among the three groups were analyzed by one-way ANOVA. If the ANOVA indicated a significant difference, Scheffé’s multiple comparison procedure was used to determine which groups were different. Regression lines were determined by the least squares method. A level of P less than 0.05 was accepted as statistically significant.
Results
General physiological characteristics
No significant differences were found in age, body mass index, or plasma concentrations of estradiol, LH, or FSH between groups A, B, and C. The mean CT attenuation number ratio of liver to spleen was significantly smaller in group C (range, -0.40 to 0.84) than that in group B (range, 1.09–1.59) (Table 1
).
|
Plasma TC and LDL-C concentrations in groups B and C were significantly lower than in group A. In group C, the plasma TG concentration was significantly higher, whereas the HDL-C concentration was significantly lower, than in groups A or B (Table 2). Plasma concentrations of TG correlated negatively with CT ratio of liver to spleen (r = -0.76; P < 0.001; Fig. 1). No significant differences were found in plasma concentrations of apolipoprotein AI, AII, and B, and lipoprotein (a), fasting glucose and insulin, and insulin resistance among three groups (Table 2).
|
|
The mean diameter of LDL particles in group C was significantly smaller than in groups A or B. LDL lag time in group C was significantly shorter than in groups A or B. The concentration of LDL-derived TBARS in group C was significantly higher than in groups A or B. All three of these markers of atherosclerosis fell outside of the normal range in group C, but not in group B (Table 3).
|
). LDL particle diameter correlated positively with lag time for LDL-related conjugated diene formation (r = 0.47; P < 0.01; Fig. 3A) and negatively with LDL-derived TBARS concentration (r = -0.44; P < 0.01; Fig. 3B).
|
|
Tamoxifen has been reported to exert an antiatherosclerotic effect by reducing plasma concentrations of TC and LDL-C (9). These favorable effects on lipid metabolism may reduce the incidence of CHD. In the present study, degrees of reduction in plasma TC and LDL-C concentrations were similar in tamoxifen-treated subjects with or without fatty change of the liver. Although we did not determine the pretreatment levels of TC and LDL-C, tamoxifen may decrease the plasma concentration of LDL-C in breast cancer patients irrespective of the presence of fatty liver. Tamoxifen has been reported to have both estrogenic and antiestrogenic effects (25). Because estrogen reduces plasma concentrations of LDL particles by stimulating hepatic LDL receptors (26), tamoxifen-induced decrease in LDL-C concentration can be explained by its partial estrogenicity. Gylling et al. (27), however, maintain that the decrease in serum cholesterol from tamoxifen results from interference with cholesterol synthesis, specially inhibition of conversion of 
8-cholesterol to lathosterol.
Although adjuvant tamoxifen has favorable effects on TC and LDL-C, it induces severe fatty changes of liver (10), which may be associated with dyslipidemia. In our subjects with tamoxifen-induced fatty liver, plasma TG concentration was increased, whereas HDL-C concentration was decreased. CT examination carried out before tamoxifen treatment did not show fatty change of the liver in all subjects. CT ratio of liver to spleen correlated negatively with plasma TG. Although we did not determine the pretreatment levels of TG, these data indicate that in subjects with tamoxifen-induced fatty liver, a decrease in CT ratio may be accompanied by a concomitant elevation of plasma TG during the 28 months of tamoxifen treatment. Thus, tamoxifen-induced fatty change of the liver may be associated with increased plasma TG concentrations.
LDL particle diameter was significantly reduced in subjects with fatty liver. We previously reported that increased plasma TG appeared to reduce the size of LDL particles (28). In the present study, plasma TG concentration correlated negatively with LDL particle diameter, suggesting that in tamoxifen-treated breast cancer patients who develop fatty liver, an increase in plasma TG concentration may reduce the size of LDL particles. In previous reports, we considered the mechanisms underlying production of small LDL particles; hypertriglyceridemia may initially enhance cholesteryl ester transfer protein-induced lipid transfer reactions, resulting in TG-rich and cholesteryl ester-poor LDL particles (29). Hydrolysis of the enriched TG by lipolytic enzymes such as lipoprotein lipase or hepatic TG lipase may then induce formation of LDL particles that are abnormally small (30). Fasting insulin (31) or insulin sensitivity (32) may be associated with the size of LDL particles. In the present study, fasting glucose and insulin concentrations and insulin resistance estimated from fasting glucose and insulin did not differ significantly between three groups. Therefore, these factors may not affect LDL particle diameter in breast cancer subjects receiving tamoxifen.
Smaller, denser LDL particles are associated with increased risk of CHD, because these particles are more susceptible than larger particles to oxidative modification (16), an initial step in the atherosclerotic process. In breast cancer subjects treated with tamoxifen, the lag time indicating intrinsic antioxidant activity of LDL particles was shortened in subjects with fatty liver. In the same group, the concentration of LDL-derived TBARS resulting from lipid peroxidation of LDL also was increased. These findings indicate that LDL particles in breast cancer subjects with tamoxifen-induced fatty liver may be particularly susceptible to oxidative modification. LDL particle diameter correlated positively with oxidation lag time and negatively with concentrations of LDL-derived TBARS, indicating that decreased size may render LDL particles more susceptible to oxidation, which is consistent with our previous findings (33). Thus, tamoxifen-induced hypertriglyceridemia may enhance oxidative susceptibility of LDL particles by reducing their diameter.
In contrast with our findings, in vitro studies have shown that tamoxifen and its metabolite 4-hydroxytamoxifen protected LDL against lipid peroxidation (34). Guetta et al. (35) have demonstrated that short-term tamoxifen treatment for 2 months in postmenopausal women did not change plasma TG concentration, and actually significantly prolonged LDL lag time. Like tamoxifen, ERT inhibits LDL peroxidation in postmenopausal women by its antioxidant effects (2). However, ERT increases plasma TG concentrations, which may decrease LDL particle diameter, as we previously demonstrated (13, 14). Although estrogen has an antioxidant effect that opposes oxidation of LDL particles (2), this benefit can be offset by hypertriglyceridemia-associated small LDL particles that are highly susceptible to oxidation (33). As with ERT in patients with elevated TG, tamoxifen-induced fatty liver with an associated increase in plasma TG concentration could counteract the favorable effects of tamoxifen, resulting in enhanced peroxidation of the small LDL particles. According to Ogawa et al. (10), onset of fatty liver occurred an average of 18 months after initiation of tamoxifen therapy. Because long-term administration of tamoxifen is recommended in patients with breast cancer, short-term tamoxifen treatment may be insufficient for evaluating lipid concentrations, including LDL size and oxidation. In addition, our results demonstrated that in subjects without fatty liver whose plasma TG concentrations were unchanged, tamoxifen treatment neither reduced LDL particle size nor increased LDL oxidative susceptibility. These findings suggest that adjuvant tamoxifen may have little protecting effect against LDL peroxidation. Further studies are needed to definitively investigate possible antioxidant effects of tamoxifen.
Adjuvant tamoxifen treatment in breast cancer patients exerts beneficial effects on lipids by reducing TC and LDL-C concentrations. However, tamoxifen-induced increase in plasma TG in subjects with fatty liver can be atherogenic via a decrease in LDL particle size that is associated with highly oxidation susceptibility. Studies are required to investigate whether increased plasma TG concentrations during adjuvant tamoxifen may increase the risk of CHD. In addition, tamoxifen-induced fatty change of the liver may progress to liver cirrhosis with steatohepatitis (36). Regular abdominal CT and/or plasma lipid examination may be necessary during tamoxifen therapy.
Acknowledgments
Footnotes
This study was conducted without any financial support.
Abbreviations: CHD, Coronary heart disease; CT, computed tomography; ERT, estrogen replacement therapy; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; TBARS, thiobarbituric acid reactive substances; TC, total cholesterol; TG, triglyceride.
Received December 13, 2001.
Accepted April 16, 2002.
References
ethinyl estradiol. Proc Natl Acad Sci USA 83:792–796
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
