Pharmacology & Toxicology 1991, 69, 132-139.
Acute Hepatotoxicity after High-Dose Methotrexate Administration to Rats R. M. Bremnes', E. Smeland', N.-E. Huseby*, T. J. Eide', and J. Aarbakke' 'Department of Pharmacology, 'Clinical Chemistry, and 'Morphology, Institute of Medical Biology, P.O. Box 977, University of Tromse, N-9001 Tromsa, Norway (Received February 14, 1991; Accepted April 9, 1991) Abstract: Acute hepatotoxicity after administration of 10-1000 mg/kg methotrexate (MTX) to rats was studied by monitoring serum transaminases, liver morphology, and disposition kinetics of MTX and 7-hydroxy-methotrexate (7OH-MTX). Half the control rats and rats administered 1000 mg/kg MTX, had their bile duct cannulated. One to 2 hr after administration of 1000 mg/kg MTX, 50% of MTX treated bile-drained rats (Ek)developed cholestasis despite similar or larger initial bile flow rates than those which did not develop cholestasis (Ebn,controls). In E, animals, peak serum ASAT and ALAT levels were 6- and 4-fold higher than that of the control rats, and morphologically, prominent hepatocytic changes and grossly dilated bile canaliculi were found. Immediately prior to cholestasis, the E, animals reached biliary 7-OH-MTX levels (8.3 k 1.3 mM, meanf S.E.M.) which were equivalent to the threshold level for precipitation of 7-OH-MTX in rat bile in vitro, and 3-fold higher than the corresponding levels of 7-OH-MTX in the bile of E, rats. Ninety-five YOof the drug in the precipitated material was 7-OH-MTX. Hence, 7-OH-MTX may play a role in acute MTX hepatotoxicity, a dose-limiting toxicity that may not be counteracted by leukovorin rescue.
The folate antagonist methotrexate (MTX) is widely used in the treatment of human malignancies (Jolivet et al. 1983), recalcitrant psoriasis (McDonalds 1985) and rheumatoid arthritis (Wilke & Mackenzie 1986). High-dose MTX (HDMTX) infusions (1-34 g/m2) are currently used in several antineoplastic therapy regimens (Jolivet et al. 1983), yielding up to millimolar serum concentrations of 7-hydroxymethotrexate (7-OH-MTX), MTX's major extracellular metabolite (Slerrdal et al. 1986; Borsi et al. 1990; Wolfrom et al. 1990). The low aqueous solubility of 7-OH-MTX, when compared to the parent drug, has rendered it a possible mediator of renal toxicity during HD-MTX therapy (Jacobs et al. 1977; Crom et al. 1987). MTX administration is associated with both acute and chronic hepatotoxicity. Acute MTX-associated hepatotoxicity has been reported after HD-MTX infusions (Jaffe & Traggis 1975; Perez et al. 1979; Breithaupt & Kuenzlen 1983; Banerjee et al. 1988), whereas chronic hepatotoxicity may occur during long-term lowdose MTX treatment for acute leukaemia and autoimmune diseases (McIntosh e f al. 1977; Ashton et al. 1982; Shergy et al. 1988). It has been postulated that the hepatotoxicity of MTX is caused by a liver derived catabolite of MTX (Sherlock 1986). However, hitherto there are no conclusive data on hepatotoxic mechanisms, e.g., whether the liver damage is mediated by metabolites or by the parent compound. It is also unknown whether the acute and chronic toxicity share a common mechanism. The rat is so far the only animal model of human MTX hepatotoxicity, and has hitherto been employed only in studies of hepatotoxicity due to chronic MTX exposure (Custer et al. 1977). In the present study, we have examined the acute hepatotoxicity of MTX doses of 10-1000 mg/kg by assessing derangements in rat liver function and morphology. Further, we have investigated the significance of biliary 7-OH-MTX and bile-
acid levels in cholestasis development in bile-drained animals administered 1000 mg/kg MTX. Materials and Methods Drugs and solutions. L-Gl~tamyl3,4-[~H]-MTX (specific activity 37.1 Ci/mmol, purity 99.2% by HPLC) was purchased from New England Nuclear, Boston, U.S.A. Formulated MTX (purity 99% by HPLC) was a gift from Nycomed, Norway. 7-OH-MTX was a gift from Dr. F. M. Sirotnak, Memorial Sloan-Kettering Cancer Center, NY, U.S.A. The assay reagents for serum ASAT, ALAT, ALP, and creatinine determination were purchased from Boehringer Mannheim, FRG. The assay reagents for bile acid determination were purchased from Nycomed, Norway. All other reagents were of analytical grade. Experimental and control groups. Male Wistar rats weighing 250-340 g (obtained from Charles River, WIGA Gmbh, 8741 Sulzseld 1, FRG) were used for the experiments. Thirty eight rats were randomly allocated to six groups [controls (CTR",,, CTRBD), A, B, C, D and El, of which group CTRBD and E consisted of 4 and 8 animals, respectively, and the other of 6. The common bile duct of each CTRB, and E animal was cannulated as previously described (Bremnes et al. 1989a). [3H]MTX doses of 10, 50, and 250 mg/kg were administered to group A, B, and C animals, respectively. Group D and E animals received 1000 mg/kg [3H]MTX, while the controls were administered equivalent volumes of isotonic saline. Four E rats ceased to secrete bile 1-2 hr after termination of drug infusion, whereas the other 4 did not. These subgroups were designated Ek and E,, respectively. Experimentalprocedure. Methotrexate solutions were prepared by dissolving in drug in isotonic saline to concen-
133
ACUTE HEPATOTOXICITY AFTER HIGH-DOSE METHOTREXATE IN RAT
trations of 1, 5, 25, and 100 mg/ml MTX, thus allowing administration of equal volumes to all rats. To each drug solution was added tritiated MTX ([3H]MTX)to an activity of 9.2 pCi/ml. The tracer amount of isotope to which each animal was exposed (0.09 pCi/g) is considered harmless (Short & Woodnott 1969). The pH of the isotonic saline and MTX solutions was adjusted to 8.9 prior to administration. All animals were anaesthetized (fentanyl 0.3 mg/kg intraperitoneally, maintenance 0.08 mg/kg/hr intramuscularly) during the experiments and had their right external jugular vein cannulated (Bremnes et al. 1989a). The drug and diluent were administered as infusions (10 min.) through the central venous catheter, which was flushed with heparinized (10 IU/ml) saline immediately after infusions and after each subsequent blood sampling. Blood and bile samples for MTX and 7-OH-MTX determination were obtained according to schedules in fig. 3 and fig. 6 (Bremnes et al. 1989a). Voided urine was collected during procedures. After sacrifice, the urine bladders were aspirated to assure complete collection. pH was measured in all bile and urine samples. Blood samples for enzyme and creatinine analyses were obtained immediately after cannulation of the jugular vein, and at 3, 6, and 10 hr after infusions. After separation of serum, all samples were stored protected from light at -20" for a maximum of 4 weeks. The rats were hydrated with 0.06 M NaHCO, in isotonic saline, 10 mg/kg/hr and 6 ml/kg/hr to bile-drained and undrained rats, respectively. Venous samples for blood gas and haematocrit analyses were drawn at the end of the experiments. Immediately before sacrifice, one specified liver lobe was biopsied in each rat. Analyticalprocedures. Analysis of MTX and 7-OH-MTX concentrations in serum, bile, and urine were performed by reverse phase high-pressure liquid chromatography (HPLC), fraction sampling and determination of radioactivity as reported previously (Bremnes et al. 1989a) with one exception: All urine samples from animals treated with MTX doses exceeding 50 mg/kg were alkalinized (0.1 M NaOH) prior to analysis. The assay detects both MTX and its major extracellular metabolites 7-OH-MTX and 2,4diamino-N 10-methylpteroic-acid (DAMPA), with no interference from polyglutamates 1-3 of MTX. Determination of serum aspartate aminotransferase (ASAT), alanine aminotransferase (ALAT), and alkaline phosphatase (ALP) was performed on a BM/Hitachi 737 at 37" (Committee on Enzymes 1974). Serum creatinine was determined with a kinetic Jaffe method using the same instrumentation. Determination of total bile acids in bile was performed using the photometric Sterognost-3a Pho method. Although bile acids (pK, 5) are essentially dissociated in bile at pH > 8 excreted as bile salts (Hofman 1988), the designation "bile acids" will be employed in the following. For light microscopy (LM) the rat liver tissue was immersion fixed in 4% phosphate buffered formaldehyde, dehydrated, embedded, cut at 5 pm, and stained with haematoxylin & eosin and van Gieson. Following fixation in
McDowells fixative overnight, tissue for electron microscopy (EM) was washed in phosphate buffer prior to postfixation in 1% aqueous OsO, for 2 hr. The tissue was then block-stained in Uranyl-acetate before dehydration in a series of graded ethanol, and subsequently embedded in Epon/Araldite. Ultrathin sections were cut on a Reichert ultracut and contrasted for 12 min. in 5 % aqueous uranylacetate and 10 min. in Reynold's lead citrate. The sections
I
OJ
I
0
2
s
4
a
1
Ib
TIME(h)
Fig. 1. Serum concentrations of aspartate aminotransferase (top panel) and alanine aminotransferase (bottom panel) versus time prior to and after short-time infusions of 1000 mg/kg ['H]methotrcxate ([3H]MTX) or diluent in anaesthetized rats. (Open circle), undrained controls, n =6; (open triangle), bile-drained controls n =4; (open square), undrained rats administered MTX, n = 6; (open inverse triangle), bile-drained rats administered MTX, n = 4; (open diamond), bile-drained rats administered MTX which developed cholestasis, n=4. *, significantly higher than the other groups; **, significantly higher than controls and undrained rats given MTX; ***, significantly higher than undrained rats administered MTX. Data are given as mean+S.E.M.
134
R. M. BREMNES ET ,415.
were examined in a Jeol 1200 EX electron microscope. LM was done on liver biopsies from all animals, whereas EM was performed on biopsies from all animals of group E and three randomly selected animals of group CTR,, CTR,, and D. Calculations. The MTX serum concentrations were analyzed according to a three-compartment open model. Pharmacokinetic parameters were obtained by means of linear regression analysis in a semilogarithmic data set, and refer to the triexponential equation: C =Ae-"'+ Be-P'+ Ge-Y' Total clearance, C I , was calculated by the equation
+ + +
C1,= Dose/(AUCo A/a B/P G/y)
where AUCo is the area under curve during drug infusion (10 min.), calculated by a triangular area, A, B, and G are the zero-time intercepts of the extrapolated lines of the a-, P-, and y-phases, respectively. The central volume of distribution, V,, was obtained by dividing the dose by (A B + G), and the apparent volume of distribution in the postdistributional phase, V,, was calculated by dividing total clearance by y. Statistical analyses were performed by one-way analysis of variance and estimation of least significant distance (StatgraphicsR; STSC, Rockville, MA, U.S.A.), with the exception of bile acid parameters which were statistically analyzed by the non-parametric Mann-Whitney U-test (MicrostatR; Ecosoft, Inc., Indianapolis, IN, U.S.A.). Statistical significance was defined as P < 0.05. All results are expressed as meanfS.E.M.
+
Results Changes in serum transaminases. Serum transaminase levels 3, 6 , and 10 hr after infusion of 1000 mg/kg MTX to D animals, were comparable to controls (fig. 1). Rats of control groups (CTR,, CTRBD) showed similar although slowly increasing serum transaminase levels during the 10 hr experimental period. While serum ASAT concentrations of the E,, rats were equal to that of the CTR,, group and ALAT levels were higher at only 3 hr after MTX infusions (93 k28 IU/l, meanf S.E.M.), the Ebcrats displayed greatly
3.0
1
2.5
-
I
n c
1 ' . 2.0 E v
Y
1.5
-
E
1.0-
Fig. 2. Electron microscopic details in liver of rats administered 1000 mg/kg methotrexate (MTX). The top panel shows the junction between two hepatocytes of a rat which were not drained for bile (D). The bile canaliculi and the cell organelles are here inconspicuous. The bottom panel displays a 2- to 3-fold distension of the bile canaliculi (c) with deformation of the microvilli and vesicular alteration of membranous material within an otherwise empty channel in a bile-drained rat which developed cholestasis of the common bile duct due to precipitation of 7-OH-MTX (Ek). Numerous dense bodies Cphagolysosomes, p) are present at the biliary pole within the hepatocytes. Note the increase of vesicular smooth endoplasmic reticulum (s) and the corresponding decrease in rough endoplasmic reticulum (r) relative to that observed in the top figure. The nuclei (0) and mitochondria (m) are also labelled. Original magnification x 10,000.
0.5
WIX i.v. r
1
1
-30 -20 -10
I
0
10
20
I
I
I
30 40 50
I
I
60
70
TIME (min)
Fig. 3. Biliary flow in anesthetized and bile-drained rats prior to, during (10 min.), and following the administration of 1000 mg/kg [3H]methotrexate (['HIMTX). Group E, (closed circles) developed cholestasis 1-2 hr after MTX infusion whereas group E,, (open circles) did not. * denotes bile sample intervals at which the bile production rates were significantly different between the groups. Data are given as mean+S.E.M., n = 4 in each group.
135
ACUTE HEPATOTOXICITY AFTER HIGH-DOSE METHOTREXATE IN RAT
elevated ASAT and ALAT levels with peak concentrations of 836f 189 (mean+ S.E.M.) and 242+34 U/l, respectively. Animals of group A-C displayed serum transaminase levels equal to control values. Morphological hepatic changes. Light microscopy (LM) revealed mild to moderate steatosis in all animals of group A-D, and most pronounced in those administered the highest MTX dose. Steatotic vesicles were also observed in the control groups, but here the changes were rather scarce. Judged by LM and electron microscopy (EM) examination, there were no morphologic differences between the unstained and bile-drained control groups. EM examination of D animals showed lipid accumulation corresponding to the light microscopical findings and in addition a slight proliferation of phagolysosomes not found in A-C animals. The bile-drained Ebn animals demonstrated only slightly more marked hepatic changes compared to rats of group D. LM examinations of liver biopsies from E, animals showed, however, an increase of cell mitosis, double nuclei formation, and single cell necrosis as evidence of severe toxic cell injury. No signs of bile accumulation within the bile ducts and acini were demonstrated by the van Gieson staining. EM examination of these animals revealed a striking dilation of the bile canaliculi with distortion of the microvilli, and vesicular arrangement of membranous material within an otherwise empty channel (fig. 2). Numerous
100000
1
10-1 I
I
I
I
I
I
0
1
2
3
4
5
I 6
TIME(h)
Fig. 4. Bile-concentration of methotrexate (open symbols) and 7hydroxymethotrexate (closed symbols) versus time following a short-time infusion of 1000 mg/kg [3H]methotrexate to 8 anaesthetized rats. Four animals (Ek,solid line) ceased to secrete bile 1-2 hr after termination of drug infusion, whereas the other rats (Ebnr dashed line) did not.
Table I. Methotrexate and 7-hydroxymethotrexate in precipitated bile samples. Four of 8 bile-drained rats developed cholestasis after administration of 1000 mg/kg ['HIMTX, in which precipitations were observed in the bile ducts and sampled bile. Concentrations of methotrexate and 7-hydroxy-methotrexate in the total bile samples, supernatants, and precipitates are listed. The supernatants and precipitates were prepared in the following way: Supernatants were siphoned off after centrifugation of bile samples at 10,000x g for 10 min. Pellets were then washed once in NaCl prior to resuspension of supernatants and precipitates in the initial volumes in 0.5 M NaOH. Data are given as mean+S.E.M., n = 12. Total bile sample Supernatant Precipitate MTX 7-OH-MTX
mg/ml YO(mean) mg/ml % (mean)
6.4 0.3 68.0 3.0f0.3 32.0
7.1 k 0.4 80.7 1.7k0.2 19.3
+
0.06 0.02 5.1 1.2k0.2 94.9
phagolysosomes as well as loss of rough endoplasmic reticulum with an increase of vesicular smooth endoplasrnic reticulum were also demonstrated in the hepatocytes of these rats. Biliaryflow, 7-OH-MTX, and bile acids. Prior to and during MTX infusion, the bile flow was comparable between Ebn and E, rats (fig. 3). Only 30 min. or more after drug infusion, the bile flow was significantly smaller in animals of group E,c as compared to Ebn.The mean bile flow during the first hour after cessation of MTX administration was 1.83, 2.47, and 1.78 ml/hr (means) for the E,, E,,, and CTR,, group, respectively. Prior to cholestasis in E, animals, we measured biliary 7-OH-MTX levels 3-fold higher than Ebn animals, peak concentrations 8.3+ 1.3 mM and 2.8f0.2 mM, respectively (fig. 4). Meanwhile, peak MTX levels of MTX in bile of E, rats were somewhat smaller (18.6 versus 20.9 mM, means) (fig. 4). While the cumulated biliary recovery as MTX was considerably smaller in the E, (6.6 f0.8%, mean fS.E.M.) than in the E,, animals (18.3 f 1.4%) due to development of cholestasis in the former group, the biliary recovery as 7-OH-MTX was slightly larger (2.28% versus 2.05%, means). In the E, rats, macroscopic precipitations were observed in bile samples as well as in the bile duct at laparotomy. Analysis of the biliary precipitates found 7OH-MTX to constitute 94.9% and MTX 5.1 YOof the drug content in the precipitated material (table 1). Biliary pH levels remained constant at 8.07 f0.03 throughout the experiments, with no discrepancy between CTR,, E,, and E, animals. Bile samples from E,, and E, animals demonstrated a rapid decline in bile acid concentrations, partly due to the bile drainageper se (fig. 5). Additionally, the Ek rats demonstrated somewhat lower biliary concentrations and smaller biliary secretion of bile acids as compared to the Ebnanimals (statistical significance by the non-parametric Mann-Whitney U-test, not by one-way analysis of variance) (fig. 5 & 6).
136
R. M. BREMNES ET AL. 35
F &3
T
30
I E
Y
30
25-
E
L d z
0 0
20ffJ
E
$!
Y E 10-
5
4
0
15 10
5-
0
20
Y
15-
.
, 1
. 2
,
.
,
3
. 4
,
. 5
,
.
,
#
o
! 0
6
.
, 1
TIME(h)
Fig. 5 . Biliary concentrations of bile acids versus time post bile duct cannulation in 2 untreated rats (open triangles) and 8 rats administered 1000 mg/kg ['Hlmethotrexate (open circles). All animals were anesthetized and bile-drained during experiments. #denotes the time of cessation of bile secretion in 4 rats which developed cholestasis. Data are given as meanfS.E.M., n=8. Insert shows biliary concentrations of bile acids in the 4 rats which did (closed squares) and 4 rats which did not (open squares) develop cholestasis after 1000 mg/kg ['Hlmethotrexate. *, bile fractions with significantly different biliary bile acid concentrations, n = 4 in each group.
Renal excretion of M T X and 7-OH-MTX. The urinary MTX recovery increased significantly with dose increments from 250 to 1000 mg/kg (table 2). Though the urinary recovery of MTX remained unaffected by the ablation of the enterohepatic MTX recirculation and the termination of biliary excretion, the urinary excretion of 7-OH-MTX increased considerably in the latter case. Urine pH increased slowly as a result of NaHCO, administration, and group CTR,, CTRB, A, D, E,, and E, reached pH 7 during the experiments. Group Ebnand CTRB, animals voided significantly less urine (means 6.6 and 5.7 ml, respectively) during the experiments (10 hr) than groups CTR,, A-D, and E, animals, 9.5, 9.5, and 10.9 ml, respectively. Pharmacokinetics of M T X and 7-OH-MTX. Serum MTX concentrations declined triphasically in all groups of animals except E,. The pharmacokinetic variables are given in table 3. A dose-dependent decline in total body clearance (Cl,) appeared at dose increments from 250 to 1000 mg/kg, where E, rats showed the smallest C1, and the longest second phase half-lives (tliza).Ebnrats demonstrated significantly more rapid initial and terminal phase half-lives and smaller V, and V, compared to animals of group A-D. Peak serum 7-OH-MTX levels prior to cholestasis in E, animals (61 It 23pM) surpassed peak concentrations of group Ebn (21 k3.3 pM) 3-fold. During cholestasis in E, rats, peak serum 7-OH-MTX levels increased 5-fold to 320 pM (fig. 7). Subsequently, serum 7-OH-MTX concentrations declined
.
, 2
.
,
.
,
. 4
3
,
. 5
,
. 6
TIME(h)
Fig. 6. Biliary secretion rate of bile acids versus time after bile duct cannulation in 4 rats which did (closed circles) and 4 rats which did not (open circles) develop cholestasis. All rats were anaesthetized, bile-drained, and administered short-time infusions of 1000 mg/kg ['Hlmethotrexate. Data are given as mean+S.E.M., n = 4 in each group.
monophasically, with significantly longer half-lives (342 f97 min) in group E, than in group E,, (27.6f 2.1 min.) and A-D (2942 min., means). At termination of experiments, all groups had comparable venous pH levels (7.39-7.43, means), whereas haematocrit values were slightly higher in bile-drained rats (CTRBn Ebn,E,) (0.224.24, means) versus animals of group CTRU, and A-D (0.194.21, means). Discussion
Acute hepatotoxicity, as evidenced by pathological increases in serum ASAT levels, occurs in 81-100% of patients adTable 2. Percentage of cumulative urinary recovery of administered ['Hlmethotrexate as methotrexate and 7-hydroxy-methotrexate in anesthetized rats during 10 hr after administration of 10, 50, 250 and 1000 mg/kg. Group E animals were in addition bile-drained during the experiments. Four E animals (E,) developed cholestasis within 1 to 2 hr after methotrexate administration. Data are given as mean+ S.E.M.
Group
l-hydroxyDose No. of Methotrexate (YO) methotrexate (%) mg/kg animals 0-10 hr 0-10 hr
6 25k1.8 0.19k0.03 6 30+ 1.8 0.09k0.02 6 36k3.4 0. I3 k0.06 6 54+ 1.8" 0.30 +O. 1 1 bo 1000 4 51 k6.9 0.18 k0.04 1000 4 48 6.2 4.55 k 1.22' E, ' P
Acute Hepatotoxicity after High-Dose Methotrexate Administration to Rats R. M. Bremnes', E. Smeland', N.-E. Huseby*, T. J. Eide', and J. Aarbakke' 'Department of Pharmacology, 'Clinical Chemistry, and 'Morphology, Institute of Medical Biology, P.O. Box 977, University of Tromse, N-9001 Tromsa, Norway (Received February 14, 1991; Accepted April 9, 1991) Abstract: Acute hepatotoxicity after administration of 10-1000 mg/kg methotrexate (MTX) to rats was studied by monitoring serum transaminases, liver morphology, and disposition kinetics of MTX and 7-hydroxy-methotrexate (7OH-MTX). Half the control rats and rats administered 1000 mg/kg MTX, had their bile duct cannulated. One to 2 hr after administration of 1000 mg/kg MTX, 50% of MTX treated bile-drained rats (Ek)developed cholestasis despite similar or larger initial bile flow rates than those which did not develop cholestasis (Ebn,controls). In E, animals, peak serum ASAT and ALAT levels were 6- and 4-fold higher than that of the control rats, and morphologically, prominent hepatocytic changes and grossly dilated bile canaliculi were found. Immediately prior to cholestasis, the E, animals reached biliary 7-OH-MTX levels (8.3 k 1.3 mM, meanf S.E.M.) which were equivalent to the threshold level for precipitation of 7-OH-MTX in rat bile in vitro, and 3-fold higher than the corresponding levels of 7-OH-MTX in the bile of E, rats. Ninety-five YOof the drug in the precipitated material was 7-OH-MTX. Hence, 7-OH-MTX may play a role in acute MTX hepatotoxicity, a dose-limiting toxicity that may not be counteracted by leukovorin rescue.
The folate antagonist methotrexate (MTX) is widely used in the treatment of human malignancies (Jolivet et al. 1983), recalcitrant psoriasis (McDonalds 1985) and rheumatoid arthritis (Wilke & Mackenzie 1986). High-dose MTX (HDMTX) infusions (1-34 g/m2) are currently used in several antineoplastic therapy regimens (Jolivet et al. 1983), yielding up to millimolar serum concentrations of 7-hydroxymethotrexate (7-OH-MTX), MTX's major extracellular metabolite (Slerrdal et al. 1986; Borsi et al. 1990; Wolfrom et al. 1990). The low aqueous solubility of 7-OH-MTX, when compared to the parent drug, has rendered it a possible mediator of renal toxicity during HD-MTX therapy (Jacobs et al. 1977; Crom et al. 1987). MTX administration is associated with both acute and chronic hepatotoxicity. Acute MTX-associated hepatotoxicity has been reported after HD-MTX infusions (Jaffe & Traggis 1975; Perez et al. 1979; Breithaupt & Kuenzlen 1983; Banerjee et al. 1988), whereas chronic hepatotoxicity may occur during long-term lowdose MTX treatment for acute leukaemia and autoimmune diseases (McIntosh e f al. 1977; Ashton et al. 1982; Shergy et al. 1988). It has been postulated that the hepatotoxicity of MTX is caused by a liver derived catabolite of MTX (Sherlock 1986). However, hitherto there are no conclusive data on hepatotoxic mechanisms, e.g., whether the liver damage is mediated by metabolites or by the parent compound. It is also unknown whether the acute and chronic toxicity share a common mechanism. The rat is so far the only animal model of human MTX hepatotoxicity, and has hitherto been employed only in studies of hepatotoxicity due to chronic MTX exposure (Custer et al. 1977). In the present study, we have examined the acute hepatotoxicity of MTX doses of 10-1000 mg/kg by assessing derangements in rat liver function and morphology. Further, we have investigated the significance of biliary 7-OH-MTX and bile-
acid levels in cholestasis development in bile-drained animals administered 1000 mg/kg MTX. Materials and Methods Drugs and solutions. L-Gl~tamyl3,4-[~H]-MTX (specific activity 37.1 Ci/mmol, purity 99.2% by HPLC) was purchased from New England Nuclear, Boston, U.S.A. Formulated MTX (purity 99% by HPLC) was a gift from Nycomed, Norway. 7-OH-MTX was a gift from Dr. F. M. Sirotnak, Memorial Sloan-Kettering Cancer Center, NY, U.S.A. The assay reagents for serum ASAT, ALAT, ALP, and creatinine determination were purchased from Boehringer Mannheim, FRG. The assay reagents for bile acid determination were purchased from Nycomed, Norway. All other reagents were of analytical grade. Experimental and control groups. Male Wistar rats weighing 250-340 g (obtained from Charles River, WIGA Gmbh, 8741 Sulzseld 1, FRG) were used for the experiments. Thirty eight rats were randomly allocated to six groups [controls (CTR",,, CTRBD), A, B, C, D and El, of which group CTRBD and E consisted of 4 and 8 animals, respectively, and the other of 6. The common bile duct of each CTRB, and E animal was cannulated as previously described (Bremnes et al. 1989a). [3H]MTX doses of 10, 50, and 250 mg/kg were administered to group A, B, and C animals, respectively. Group D and E animals received 1000 mg/kg [3H]MTX, while the controls were administered equivalent volumes of isotonic saline. Four E rats ceased to secrete bile 1-2 hr after termination of drug infusion, whereas the other 4 did not. These subgroups were designated Ek and E,, respectively. Experimentalprocedure. Methotrexate solutions were prepared by dissolving in drug in isotonic saline to concen-
133
ACUTE HEPATOTOXICITY AFTER HIGH-DOSE METHOTREXATE IN RAT
trations of 1, 5, 25, and 100 mg/ml MTX, thus allowing administration of equal volumes to all rats. To each drug solution was added tritiated MTX ([3H]MTX)to an activity of 9.2 pCi/ml. The tracer amount of isotope to which each animal was exposed (0.09 pCi/g) is considered harmless (Short & Woodnott 1969). The pH of the isotonic saline and MTX solutions was adjusted to 8.9 prior to administration. All animals were anaesthetized (fentanyl 0.3 mg/kg intraperitoneally, maintenance 0.08 mg/kg/hr intramuscularly) during the experiments and had their right external jugular vein cannulated (Bremnes et al. 1989a). The drug and diluent were administered as infusions (10 min.) through the central venous catheter, which was flushed with heparinized (10 IU/ml) saline immediately after infusions and after each subsequent blood sampling. Blood and bile samples for MTX and 7-OH-MTX determination were obtained according to schedules in fig. 3 and fig. 6 (Bremnes et al. 1989a). Voided urine was collected during procedures. After sacrifice, the urine bladders were aspirated to assure complete collection. pH was measured in all bile and urine samples. Blood samples for enzyme and creatinine analyses were obtained immediately after cannulation of the jugular vein, and at 3, 6, and 10 hr after infusions. After separation of serum, all samples were stored protected from light at -20" for a maximum of 4 weeks. The rats were hydrated with 0.06 M NaHCO, in isotonic saline, 10 mg/kg/hr and 6 ml/kg/hr to bile-drained and undrained rats, respectively. Venous samples for blood gas and haematocrit analyses were drawn at the end of the experiments. Immediately before sacrifice, one specified liver lobe was biopsied in each rat. Analyticalprocedures. Analysis of MTX and 7-OH-MTX concentrations in serum, bile, and urine were performed by reverse phase high-pressure liquid chromatography (HPLC), fraction sampling and determination of radioactivity as reported previously (Bremnes et al. 1989a) with one exception: All urine samples from animals treated with MTX doses exceeding 50 mg/kg were alkalinized (0.1 M NaOH) prior to analysis. The assay detects both MTX and its major extracellular metabolites 7-OH-MTX and 2,4diamino-N 10-methylpteroic-acid (DAMPA), with no interference from polyglutamates 1-3 of MTX. Determination of serum aspartate aminotransferase (ASAT), alanine aminotransferase (ALAT), and alkaline phosphatase (ALP) was performed on a BM/Hitachi 737 at 37" (Committee on Enzymes 1974). Serum creatinine was determined with a kinetic Jaffe method using the same instrumentation. Determination of total bile acids in bile was performed using the photometric Sterognost-3a Pho method. Although bile acids (pK, 5) are essentially dissociated in bile at pH > 8 excreted as bile salts (Hofman 1988), the designation "bile acids" will be employed in the following. For light microscopy (LM) the rat liver tissue was immersion fixed in 4% phosphate buffered formaldehyde, dehydrated, embedded, cut at 5 pm, and stained with haematoxylin & eosin and van Gieson. Following fixation in
McDowells fixative overnight, tissue for electron microscopy (EM) was washed in phosphate buffer prior to postfixation in 1% aqueous OsO, for 2 hr. The tissue was then block-stained in Uranyl-acetate before dehydration in a series of graded ethanol, and subsequently embedded in Epon/Araldite. Ultrathin sections were cut on a Reichert ultracut and contrasted for 12 min. in 5 % aqueous uranylacetate and 10 min. in Reynold's lead citrate. The sections
I
OJ
I
0
2
s
4
a
1
Ib
TIME(h)
Fig. 1. Serum concentrations of aspartate aminotransferase (top panel) and alanine aminotransferase (bottom panel) versus time prior to and after short-time infusions of 1000 mg/kg ['H]methotrcxate ([3H]MTX) or diluent in anaesthetized rats. (Open circle), undrained controls, n =6; (open triangle), bile-drained controls n =4; (open square), undrained rats administered MTX, n = 6; (open inverse triangle), bile-drained rats administered MTX, n = 4; (open diamond), bile-drained rats administered MTX which developed cholestasis, n=4. *, significantly higher than the other groups; **, significantly higher than controls and undrained rats given MTX; ***, significantly higher than undrained rats administered MTX. Data are given as mean+S.E.M.
134
R. M. BREMNES ET ,415.
were examined in a Jeol 1200 EX electron microscope. LM was done on liver biopsies from all animals, whereas EM was performed on biopsies from all animals of group E and three randomly selected animals of group CTR,, CTR,, and D. Calculations. The MTX serum concentrations were analyzed according to a three-compartment open model. Pharmacokinetic parameters were obtained by means of linear regression analysis in a semilogarithmic data set, and refer to the triexponential equation: C =Ae-"'+ Be-P'+ Ge-Y' Total clearance, C I , was calculated by the equation
+ + +
C1,= Dose/(AUCo A/a B/P G/y)
where AUCo is the area under curve during drug infusion (10 min.), calculated by a triangular area, A, B, and G are the zero-time intercepts of the extrapolated lines of the a-, P-, and y-phases, respectively. The central volume of distribution, V,, was obtained by dividing the dose by (A B + G), and the apparent volume of distribution in the postdistributional phase, V,, was calculated by dividing total clearance by y. Statistical analyses were performed by one-way analysis of variance and estimation of least significant distance (StatgraphicsR; STSC, Rockville, MA, U.S.A.), with the exception of bile acid parameters which were statistically analyzed by the non-parametric Mann-Whitney U-test (MicrostatR; Ecosoft, Inc., Indianapolis, IN, U.S.A.). Statistical significance was defined as P < 0.05. All results are expressed as meanfS.E.M.
+
Results Changes in serum transaminases. Serum transaminase levels 3, 6 , and 10 hr after infusion of 1000 mg/kg MTX to D animals, were comparable to controls (fig. 1). Rats of control groups (CTR,, CTRBD) showed similar although slowly increasing serum transaminase levels during the 10 hr experimental period. While serum ASAT concentrations of the E,, rats were equal to that of the CTR,, group and ALAT levels were higher at only 3 hr after MTX infusions (93 k28 IU/l, meanf S.E.M.), the Ebcrats displayed greatly
3.0
1
2.5
-
I
n c
1 ' . 2.0 E v
Y
1.5
-
E
1.0-
Fig. 2. Electron microscopic details in liver of rats administered 1000 mg/kg methotrexate (MTX). The top panel shows the junction between two hepatocytes of a rat which were not drained for bile (D). The bile canaliculi and the cell organelles are here inconspicuous. The bottom panel displays a 2- to 3-fold distension of the bile canaliculi (c) with deformation of the microvilli and vesicular alteration of membranous material within an otherwise empty channel in a bile-drained rat which developed cholestasis of the common bile duct due to precipitation of 7-OH-MTX (Ek). Numerous dense bodies Cphagolysosomes, p) are present at the biliary pole within the hepatocytes. Note the increase of vesicular smooth endoplasmic reticulum (s) and the corresponding decrease in rough endoplasmic reticulum (r) relative to that observed in the top figure. The nuclei (0) and mitochondria (m) are also labelled. Original magnification x 10,000.
0.5
WIX i.v. r
1
1
-30 -20 -10
I
0
10
20
I
I
I
30 40 50
I
I
60
70
TIME (min)
Fig. 3. Biliary flow in anesthetized and bile-drained rats prior to, during (10 min.), and following the administration of 1000 mg/kg [3H]methotrexate (['HIMTX). Group E, (closed circles) developed cholestasis 1-2 hr after MTX infusion whereas group E,, (open circles) did not. * denotes bile sample intervals at which the bile production rates were significantly different between the groups. Data are given as mean+S.E.M., n = 4 in each group.
135
ACUTE HEPATOTOXICITY AFTER HIGH-DOSE METHOTREXATE IN RAT
elevated ASAT and ALAT levels with peak concentrations of 836f 189 (mean+ S.E.M.) and 242+34 U/l, respectively. Animals of group A-C displayed serum transaminase levels equal to control values. Morphological hepatic changes. Light microscopy (LM) revealed mild to moderate steatosis in all animals of group A-D, and most pronounced in those administered the highest MTX dose. Steatotic vesicles were also observed in the control groups, but here the changes were rather scarce. Judged by LM and electron microscopy (EM) examination, there were no morphologic differences between the unstained and bile-drained control groups. EM examination of D animals showed lipid accumulation corresponding to the light microscopical findings and in addition a slight proliferation of phagolysosomes not found in A-C animals. The bile-drained Ebn animals demonstrated only slightly more marked hepatic changes compared to rats of group D. LM examinations of liver biopsies from E, animals showed, however, an increase of cell mitosis, double nuclei formation, and single cell necrosis as evidence of severe toxic cell injury. No signs of bile accumulation within the bile ducts and acini were demonstrated by the van Gieson staining. EM examination of these animals revealed a striking dilation of the bile canaliculi with distortion of the microvilli, and vesicular arrangement of membranous material within an otherwise empty channel (fig. 2). Numerous
100000
1
10-1 I
I
I
I
I
I
0
1
2
3
4
5
I 6
TIME(h)
Fig. 4. Bile-concentration of methotrexate (open symbols) and 7hydroxymethotrexate (closed symbols) versus time following a short-time infusion of 1000 mg/kg [3H]methotrexate to 8 anaesthetized rats. Four animals (Ek,solid line) ceased to secrete bile 1-2 hr after termination of drug infusion, whereas the other rats (Ebnr dashed line) did not.
Table I. Methotrexate and 7-hydroxymethotrexate in precipitated bile samples. Four of 8 bile-drained rats developed cholestasis after administration of 1000 mg/kg ['HIMTX, in which precipitations were observed in the bile ducts and sampled bile. Concentrations of methotrexate and 7-hydroxy-methotrexate in the total bile samples, supernatants, and precipitates are listed. The supernatants and precipitates were prepared in the following way: Supernatants were siphoned off after centrifugation of bile samples at 10,000x g for 10 min. Pellets were then washed once in NaCl prior to resuspension of supernatants and precipitates in the initial volumes in 0.5 M NaOH. Data are given as mean+S.E.M., n = 12. Total bile sample Supernatant Precipitate MTX 7-OH-MTX
mg/ml YO(mean) mg/ml % (mean)
6.4 0.3 68.0 3.0f0.3 32.0
7.1 k 0.4 80.7 1.7k0.2 19.3
+
0.06 0.02 5.1 1.2k0.2 94.9
phagolysosomes as well as loss of rough endoplasmic reticulum with an increase of vesicular smooth endoplasrnic reticulum were also demonstrated in the hepatocytes of these rats. Biliaryflow, 7-OH-MTX, and bile acids. Prior to and during MTX infusion, the bile flow was comparable between Ebn and E, rats (fig. 3). Only 30 min. or more after drug infusion, the bile flow was significantly smaller in animals of group E,c as compared to Ebn.The mean bile flow during the first hour after cessation of MTX administration was 1.83, 2.47, and 1.78 ml/hr (means) for the E,, E,,, and CTR,, group, respectively. Prior to cholestasis in E, animals, we measured biliary 7-OH-MTX levels 3-fold higher than Ebn animals, peak concentrations 8.3+ 1.3 mM and 2.8f0.2 mM, respectively (fig. 4). Meanwhile, peak MTX levels of MTX in bile of E, rats were somewhat smaller (18.6 versus 20.9 mM, means) (fig. 4). While the cumulated biliary recovery as MTX was considerably smaller in the E, (6.6 f0.8%, mean fS.E.M.) than in the E,, animals (18.3 f 1.4%) due to development of cholestasis in the former group, the biliary recovery as 7-OH-MTX was slightly larger (2.28% versus 2.05%, means). In the E, rats, macroscopic precipitations were observed in bile samples as well as in the bile duct at laparotomy. Analysis of the biliary precipitates found 7OH-MTX to constitute 94.9% and MTX 5.1 YOof the drug content in the precipitated material (table 1). Biliary pH levels remained constant at 8.07 f0.03 throughout the experiments, with no discrepancy between CTR,, E,, and E, animals. Bile samples from E,, and E, animals demonstrated a rapid decline in bile acid concentrations, partly due to the bile drainageper se (fig. 5). Additionally, the Ek rats demonstrated somewhat lower biliary concentrations and smaller biliary secretion of bile acids as compared to the Ebnanimals (statistical significance by the non-parametric Mann-Whitney U-test, not by one-way analysis of variance) (fig. 5 & 6).
136
R. M. BREMNES ET AL. 35
F &3
T
30
I E
Y
30
25-
E
L d z
0 0
20ffJ
E
$!
Y E 10-
5
4
0
15 10
5-
0
20
Y
15-
.
, 1
. 2
,
.
,
3
. 4
,
. 5
,
.
,
#
o
! 0
6
.
, 1
TIME(h)
Fig. 5 . Biliary concentrations of bile acids versus time post bile duct cannulation in 2 untreated rats (open triangles) and 8 rats administered 1000 mg/kg ['Hlmethotrexate (open circles). All animals were anesthetized and bile-drained during experiments. #denotes the time of cessation of bile secretion in 4 rats which developed cholestasis. Data are given as meanfS.E.M., n=8. Insert shows biliary concentrations of bile acids in the 4 rats which did (closed squares) and 4 rats which did not (open squares) develop cholestasis after 1000 mg/kg ['Hlmethotrexate. *, bile fractions with significantly different biliary bile acid concentrations, n = 4 in each group.
Renal excretion of M T X and 7-OH-MTX. The urinary MTX recovery increased significantly with dose increments from 250 to 1000 mg/kg (table 2). Though the urinary recovery of MTX remained unaffected by the ablation of the enterohepatic MTX recirculation and the termination of biliary excretion, the urinary excretion of 7-OH-MTX increased considerably in the latter case. Urine pH increased slowly as a result of NaHCO, administration, and group CTR,, CTRB, A, D, E,, and E, reached pH 7 during the experiments. Group Ebnand CTRB, animals voided significantly less urine (means 6.6 and 5.7 ml, respectively) during the experiments (10 hr) than groups CTR,, A-D, and E, animals, 9.5, 9.5, and 10.9 ml, respectively. Pharmacokinetics of M T X and 7-OH-MTX. Serum MTX concentrations declined triphasically in all groups of animals except E,. The pharmacokinetic variables are given in table 3. A dose-dependent decline in total body clearance (Cl,) appeared at dose increments from 250 to 1000 mg/kg, where E, rats showed the smallest C1, and the longest second phase half-lives (tliza).Ebnrats demonstrated significantly more rapid initial and terminal phase half-lives and smaller V, and V, compared to animals of group A-D. Peak serum 7-OH-MTX levels prior to cholestasis in E, animals (61 It 23pM) surpassed peak concentrations of group Ebn (21 k3.3 pM) 3-fold. During cholestasis in E, rats, peak serum 7-OH-MTX levels increased 5-fold to 320 pM (fig. 7). Subsequently, serum 7-OH-MTX concentrations declined
.
, 2
.
,
.
,
. 4
3
,
. 5
,
. 6
TIME(h)
Fig. 6. Biliary secretion rate of bile acids versus time after bile duct cannulation in 4 rats which did (closed circles) and 4 rats which did not (open circles) develop cholestasis. All rats were anaesthetized, bile-drained, and administered short-time infusions of 1000 mg/kg ['Hlmethotrexate. Data are given as mean+S.E.M., n = 4 in each group.
monophasically, with significantly longer half-lives (342 f97 min) in group E, than in group E,, (27.6f 2.1 min.) and A-D (2942 min., means). At termination of experiments, all groups had comparable venous pH levels (7.39-7.43, means), whereas haematocrit values were slightly higher in bile-drained rats (CTRBn Ebn,E,) (0.224.24, means) versus animals of group CTRU, and A-D (0.194.21, means). Discussion
Acute hepatotoxicity, as evidenced by pathological increases in serum ASAT levels, occurs in 81-100% of patients adTable 2. Percentage of cumulative urinary recovery of administered ['Hlmethotrexate as methotrexate and 7-hydroxy-methotrexate in anesthetized rats during 10 hr after administration of 10, 50, 250 and 1000 mg/kg. Group E animals were in addition bile-drained during the experiments. Four E animals (E,) developed cholestasis within 1 to 2 hr after methotrexate administration. Data are given as mean+ S.E.M.
Group
l-hydroxyDose No. of Methotrexate (YO) methotrexate (%) mg/kg animals 0-10 hr 0-10 hr
6 25k1.8 0.19k0.03 6 30+ 1.8 0.09k0.02 6 36k3.4 0. I3 k0.06 6 54+ 1.8" 0.30 +O. 1 1 bo 1000 4 51 k6.9 0.18 k0.04 1000 4 48 6.2 4.55 k 1.22' E, ' P
