Placenta(1991), 12,653-661
Transfer of Acipimox Across the Isolated Perfused Human Placenta HAN-Y GHABRIAL, MARK A. CZUBA, CHERYL K. STEAD, RICHARD A. SMALLWOOD & DENIS J. MORGAN” University of Melbourne, Department of Medicine, Repatriation General Hospital, Melbourne, Victoria Victorian College of Pharmacy, Melbourne, Victoria, Australia at: Victorian College of Pharmacy, 381 Royal Parade, Parkville, Melbourne, Victoria, 3052 Australia n To tahom correspondence should be dressed
Paperaccepted 13.5.1991
SUMh%ARY The placental transfer of the new lipid-lowering agent, acipimox was investigated in the isolated perfused human placenta. Placentas obtained at caesarean section were perfused for 120 min, with both maternal and fetal circuits in closed recycling mode. Acipimox was added to either the maternal circuit alone eve experiments) or to both maternal and fetal circuits simultaneously @ve experiments) to achieve initial concentrations ofSug/ml. Antipyrine (IOpg/ml) and l-(14C)-leucine (250pM) were added in like fashion as reference compounds. Two hours after addition to the maternal circuit alone antipyrine was close to equilibrium across the placenta, but equilibration of acipimox was incomplete fetal/maternal ratio = 0.58 + 0.11). Maternal to fetal placental clearance of actpimox (0.80 + 0.18 ml/min) was 25per cent of antipyrine clearance. AJter simultaneous administration to both maternal and fetal n’rcuits the 1-(‘4C)-leucinefetal/maternal ratio was 1.44 It 0.13 at 120 min, whereas maternal and fetal concentrations of acipimox and antipyrine were at equilibrium for the duration of the experiment (fetal/maternal ratio of acipimox at 120 min = 1.10 f 0.06). This study shows that acapimox is transfered across the human placenta by da&ion at a slow rate. The low permeabilityof the placenta may afford some protection to the fetus fromacipimox administered to the mother in vivo.
Acipimox (5-methylpyrazine carboxylic acid-4-oxide) is a new lipid-lowering agent (Fuccella et al, 1980) with a molecular structure similar to nicotinic acid. Acipimox appears to modify lipoprotein metabolism (Sirtori et al, 1981; Stuyt et al, 1985; Taskinen & Nikkila, 1988) and may be useful in patient type II and IV hyperlipoproteinaemia and hypertriglyceridaemia (Taskinen & Nikkila 1988). Such patient groups may include women of child-bearing potential, but there is no information concerning the likely exposure of the fetus to acipimox administered to the mother. 0143-4004/91/060653
+ 09 $05.00/O
0
1991 Baillike Tindall Ltd
654
Placenta(I 991), Vol. 12
This drug presents several interesting characteristics for the study of placental transfer. It is ionized at physiological pH (carboxylic acid group pKa = 3.25) and is relatively polar. These factors would be expected to lessen placental permeability for this drug (Mihaly & Morgan, 1984). However, acipimox resembles the vitamin nicotinic acid, and as active maternal to fetal transport occurs with other vitamins, such as thiamine (Dancis et al, 1988a) and riboflavin (Dancis et al, 1988b), acipimox may also be actively transported to the fetus. We have studied the placental transfer of acipimox in the isolated perfused human placenta and compared its rate of transfer with those of a freely diffusible molecule, antipyrine, and an actively transported amino acid, l-[‘4C]-leucine.
METHODS Perfusion techniques The perfusion technique was based on the method of Penfold et al (1981), as modified by Ching et al (1987, 1988). Briefly, placentas were obtained within 5 min of delivery by caesarean section and the fetal chorionic artery and vein supplying a peripheral cotyledon were cannulated with vinyl tubing and cleared of blood by flushing with warm, heparinized, Compound Sodium Lactate Injection B.P. at 3 ml/mm for 5 min. The placenta was placed on a platform with the perfused cotyledon exposed and supported by plastic mesh (3 mm grid) with the fetal side uppermost. Two blunt-ended cannulas (made from 19 gauge needles) were inserted 0.5 cm into the maternal decidual basal plate and the maternal perfusate flow rate was increased slowly to 14-15 ml/mm. Fetal flow rate was 7.5-8 ml/mm. The pet&ate consisted of Krebs-Ringer phosphate-bicarbonate electrolyte solution with added 30 g/l w/v dextran (MW 70 000), 1 g/I w/ v b ovine serum albumin and 1 g/l w/v glucose. The placenta and maternal (600 ml) and fetal (150 ml) per&sate reservoirs were housed in a thermostatically controlled cabinet at 37°C. Each circuit was pumped separately using an LKB peristaltic pump in a closed recycling mode. Perfusate passed from the reservoir, through a coarse filter (pore size approximately 1 mm), silastic membrane oxygenator and a bubble trap to the arterial inflow cannula. Indices of placental viability were maternal oxygen consumption > 1.5 ~mol/min, stable perfusate backpressure ~60 mmHg, with the absence (i.e. ~10 ml/h) of fetal pet&sate leakage, and active I-[‘4C]-leucine transfer from mother to fetus. Experimental design Maternal Administration. In five placental preparations the transfer of acipimox, 1-[‘4C]-
leucine and antipyrine was examined following administration to the maternal reservoir only. The maternal reservoir contained an initial concentration of 5 &ml acipimox (Farmitalia Carlo Erba, Melbourne, Australia), 20 ,~g/rnl antipyrine and 0.25 mu l-[‘4C]-leucine (specific activity 67,~Ci/nunol) (Amersham International, Buckinghamshire, UK). Per&sate (2 ml) was collected from the fetal and maternal circuits before and every 10 min after dosing for 120 min, and an equal amount of blank perfusate was added to replace that lost due to sampling. Maternal and Fe&Administration. This experiment was similar in design to that for maternal
dosing, except that in the five placental preparations used, both maternal and fetal reservoirs
Ghabrial et al: Transjir ofAcipimoxAcross Human Placenta
655
were spiked to the above concentrations. Perfusate samples were taken from both maternal and fetal circuits before and every 15 min after dosing, for 120 min. In one study sampling was extended to 150 min. Protein Binding of Acipimox Binding of acipimox to bovine serum albumin in perfusate solution was assessed by equilibrium dialysis. Solutions of 0, 1,5 and 10 g/l (w/v) of albumin in perfusate were spiked to give a concentration of 5 pg/ml of acipimox. Two millilitres of this solution were dialyzed against protein-free perfusate (2 ml, pH 7.4) in perspex chambers separated by cellulose dialysis membrance (type 20, Union Carbide Corp, NY). Dialysis cells were incubated at 37°C for 12 h in an oscillating water bath. Unbound fraction was calculated as the ratio of acipimox concentration in buffer/acipimox concentration in dialyzed perfusate at equilibrium. Assays Aczpimox: To a screw-topped culture tube was added 1 ml of sample, 2 pg of sulphanilamide (in 100~1) and 1 ml of 1Mphosphoric acid. To this, 10 ml of ethylacetate/isopropanol(90: 10) was added, the tube was capped and vortexed for 3 min, and centrifuged. The organic layer was transferred to a tapered glass tube and evaporated to dryness. The residue was reconstituted in 150 ~1 of mobile phase of which 20 ~1 were injected into the HPLC. A standard curve was constructed by plotting the peak area ratio of acipimox to sulphanilamide versus known concentrations. The standard concentrations were 0,0.1,0.5, 1,2, 5 and 10 &ml. Quality control samples of known concentrations were analysed with each assay run. The HPLC system comprised an M6000A pump, a Novapak phenyl 4 pm radial compression module housed in a Z-module, a Lambda max 481 variable wavelength detector set at 264 nm and a Maxima 820 data station to collect and process the signal from the detector (all from Milliporen;Vaters, Milford, Massachusetts). The mobile phase was 25 mu di-potassium hydrogen phosphate buffer containing 40 ml of methanol, 5 mmol of tetrabutyl ammonium and 5 mmol triethylaminefl with pH adjusted to 6.8 using concentrated phosphoric acid. This was delivered at 3 ml/min. The coefficient of variation (precision) was 4.6 per cent (n = 6) at 1 pug/ml and 2.9 per cent (n = 6) at 5 pug/ml and accuracy was 7 per cent at 1 &ml and 4.2 per cent at 5 pug/ml. Antipyrine: Antipyrine was assayed by HPLC according to the method of Shargel et al (1979). Precision and accuracy were 6 per cent and 9.5 per cent, respectively, at a concentration of 6.25 pug/ml. l-[‘4C/-leucine: Aquasol (DuPont) (10 ml) was added to 200 ~1 of sample and counted on a Packard Scintillation counter, Model 3255. Quench and efficiency correction was carried out using external standardization. Calculations In maternal dosing experiments the transplacental clearance was calculated by fitting the following equation which describes equilibration between two compartments (Simon, 1972) to the fetal perfusate concentration data.
Placenta(1991), Vol. I2 cF,t
=
Where C,,t is fetal perfusate concentration at time t, VM and V, are maternal and fetal perfusate volumes respectively and Kis the transfer rate constant. Placental clearance (CL), which represents the ease of transfer of the substance across the placenta, is given by (Simon, 1972): cL
= W?M*~F) VM +
(2)
b
Kwas calculated by fitting equation (1) to.CF,t versus time data obtained from the maternal dosing experiments. Fitting was by least squares non-linear regression using the GaussNewton-Marquardt method as performed on the computer program Minim 1.6 (Purves, 1989). Recovery of the dose was calculated as follows: (3)
RESULTS Maternal administration Figure, 1 shows the increasing concentrations of antipyrine and acipimox in the fetal perfusate for the first 120 min. At 120 min the fetal concentration of antipyrine had almost reached a
0
20
40
60
60
loo
I2C
Time (min) Figure 1. Typical fetal perfusate concentration prohles for antipyrine (0) and acipimox (+) in a maternal administration study. The fit (solid line) acuw&tg to equation (1) is shown and yiekkd the f&wing parameters: antipyrine-A = 13.5pg/ml, K = 0.01% min-‘, CL = 2.94 ml&n, ? = 0.9%; and acipimox-A = 3.9O@nl, K = 0.0066 min-‘, CL = 0.99 ml/m& 2 = 0.979.
Ghabrial et al: Transfer ofAcipimoxAcross Human Placenta
657
Time (min)
Figure 2. Mean z!z s.e.m. fetal/maternal ratios of perfusate I-[14C]-leucine (W) in the maternal administration studies.
concentrations
for antipyrine
(O), acipimox (+) and
plateau, whereas the fetal concentration of acipimox was increasing steadily. The ratios of fetal/maternal perfusate drug concentrations (F/M ratio) over this 120 min period are shown in Figure 2. At 120 min the F/M ratio for antipyrine was 0.94 + 0.11 (means + s.d.), indicating that fetal and maternal concentrations of this drug were very close to equilibrium. In contrast, at 120 min the F/M ratio for acipimox was only 0.58 f 0.11. Fetal perfusate concentrations of l-[14C]-leucine were still rising at 120 min, with the F/M ratio at this point being 1.02 k 0.12. Equation (1) fitted the fetal perfusate drug concentration profiles well and typical fits are shown in Figure 1. Parameter values and placental clearance values, calculated from equation (2), are shown in Table 1. The placental clearance of acipimox was 0.80 + 0.18 ml/ min, which was approximately one-quarter that of antipyrine (3.61 + 1.20 ml/min). The
Drug
Table 1. Kinetics of antipyrine,
acipimox and l-[‘4C]-leucine
120 min F/M ratio
K” (min-‘)
CLb (mLhin)
in maternal
dosing study Recovery W)
Clearance ratio’
Antipyrine
0.94 (0.11)
12.8 (0.79)
0.030 (0.010)
3.61 (1.20)
79.7 (4.92)
Acipimox
0.58 (0.11)
3.59 (0.47)
0.007 (0.002)
0.80 (0.18)
89.9 (11.8)
0.24 (0.08)
I-[14C]-leucine
1.02 (0.12)
(Z;
0.023 (0.004)
2.70 (0.45)
73.1 (18.5)
0.79 (0.18)
Means of five placental preparations with standard deviations in parenthesis. u From equation (l), A = dose/(VM + VF). b Placental clearance, from equation (2). ’ Clearance ratio of compound to that of antipyrine. d d/min/ml.
658
Plactwa (1991), Vol.12
0
15
30
45
60
75
90
105
120
Time (min)
Figure 3. Mean f s.e.m. fetal/maternal ratios of perfusate concentrations of antipyrine (Cl), acipimox (+) and I-[ 4C]-leucine (W) in maternal and fetal administration studies.
mean percentage of acipimox recovered in maternal and fetal perfusates at the end of the experiment [equation (3)] ranged from 73.7 per cent for l-[‘4C]-leucine to 89.9 per cent for acipimox (Table 1). Maternal and fetal administration Figure 3 depicts the mean F/M ratio in the experiments in which fetal and maternal drug concentrations were the same at the start. The F/M ratios of antipyrine and acipimox remained relatively constant over the 120 min whereas the F/M ratio of l-[‘4C]-leucine rose steadily throughout the experiment, reaching 1.44 f 0.13 at 120 min. At 120 min the F/M ratios of antipyrine (1.06 + 0.06) and acipimox (1.10 f 0.06) did not differ significantly, and their 95 per cent confidence intervals included the value of 1. At 120 min, the F/M ratio of l-[‘4C]-leucine was significantly greater than those of antipyrine and acipimox (P < 0.05). Placental viability Mean oxygen consumption by the perfused cotelydon was 5.08 f 0.56~moVmin in maternal administration experiments and 4.80 If: 1.67 pmol/min in dual administration experiments. Values were not corrected for placental weight as the whole placenta was not perfused. These values are similar to those obtained by us previously (Ching et al, 1987,1988). Protein binding No protein binding of acipimox was evident. After 12 h of dialysis the concentration of acipimox was similar in both chambers, regardless of the protein concentrations.
Ghabrial et al: Transfer
ofAcipimox Across Human Placenta
659
DISCUSSION Antipyrine diffuses rapidly across the placenta, so that the limiting factor in transfer is placental blood flow rate (Schneider et al, 1972). Antipyrine is therefore a useful reference compound for assessing the placental permeability and transfer rate of other compounds. In our study the mean placental clearance of antipyrine was 3.61 + 1.20 ml/min. This is somewhat greater than found previously [2.15 f 0.11 ml/min, Ching et al (1987); 2.71 + 0.66 ml/min, Ching et al (1988), probably representing a greater degree of overlap of the maternal and fetal circulations studied]. The placental transfer of acipimox was appreciably slower than that of antipyrine (Figure 1). Fitting the equation describing diffusion of drug from the maternal to the fetal compartment to the maternal dosage data yielded a placental clearance value that was only 2.5 per cent that of antipyrine. Protein binding of acipimox to bovine serum albumin (l-10 g/l) was negligible, therefore placental transfer rate of acipimox would not have been affected by protein binding in the perfusate, which contained 1 g/l albumin. In maternal administration experiments the mean recoveries of antipyrine, acipimox, and l-[14C]-leucine at the end of the experiment, calculated using equation (3), ranged from 75-90 per cent (Table 1). The drug not accounted for was probably taken up by placental tissue, with the relatively high recovery for acipimox indicating a low affinity of this drug for placental tissue. Most drugs studied in vitro and in vivo have shown a rapid placental transfer from mother to fetus (Mihaly and Morgan, 1984). However, the placental clearance of acipimox was relatively low, being only 25 per cent that of antipyrine. A number of other drugs have been found to have a similarly low placental clearance, including clorazepate (Guerre-Milo et al, 1979), cimetidine (Ching et al, 1987), cefoperazone, sulbactam (Fortunato et al, 1988) and pentamidine (Fortunato and Bawdon, 1989). This indicates that the permeability of the placenta is relatively low for these drugs, although not as low as for inulin, for which the placenta is highly impermeable, having a placental clearance of only 7 per cent relative to antipyrine (Nandakumaran et al, 1981; Contractor and Stannard, 1983). The physicochemical properties of a drug that determine its placental permeability are primarily lipid solubility, degree of ionization in blood and molecular weight (Mirkin and Singh, 1976). Antipyrine is highly lipid soluble, unionized at physiological pH and has a molecular weight of 188.2. Acipimox (MW = 154.1) is a relatively polar compound (freely soluble in water, insoluble in ether) and with a pK, of 3.25 is fully ionized at physiological pH. Therefore, these latter two properties probably both contribute to the low placental permeability of acipimox. In contrast, the low placental permeability of cimetidine appears to be due primarily to its low lipid solubility only, because it is predominantly unionized at physiological pH (Ching et al, 1987). Acipimox is a structural analogue of the vitamin nicotinic acid. The mechanism of placental transfer of nicotinic acid has not been investigated, but the mechanism of placental transfer of several other vitamins, including thiamine (Dancis et al, 1988a) and riboflavine (Dancis et al, 1988b) involves active maternal to fetal transport. If there were also an active placental transport mechanism for nicotinic acid, acipimox might be a substrate for this process and this could augment the exposure of the fetus to maternally administered drug. The maternal and fetal administration experiments were designed to detect the presence of active transport of acipimox. Figure 3 shows that when maternal and fetal perfusate concentrations were equal at the start of the experiment, they remained equal for the 120 min duration of the experiment. This finding rules out the possibility of active transport of acipimox in our preparation. To
660
Pkumta (1991), Vol. 12
demonstrate that the preparation was capable of active transport, we also included equal concentrations of l-[14C]-leucine in maternal and fetal perfusates at the start of the experiment. l-leucine is considered to be a useful indicator of active maternal to fetal placental transfer (Schneider et al, 1987). Figure 3 shows that in the maternal and fetal administration experiments, the F/M ratio of I-[‘4C]-leucine reached 1.44 by 120 min. Although some metabolism of the leucine during the 120 min period cannot be excluded (Schneider et al, 1987) the accumulation of l-[14C]-leucine in the fetal perfusate reflects active transport of this substance. This F/M ratio of 1.44 is similar to values obtained previously by us and others (Ching et al, 1987, 1988; Challier et al, 1985; Schneider et al, 1979). For drugs that rapidly diffuse across the placenta, equilibrium between maternal and fetal blood concentrations of drug will be rapid. Therefore, the degree of fetal exposure will depend on the equilibrium between maternal and fetal blood rather than the rate of placental transfer (Mihaly and Morgan, 1984). The relative maternal and fetal blood concentrations at equilibrium are determined primarily by maternal and fetal factors (e.g. clearance, blood pH, plasma protein binding) and the placenta plays an insignificant role for such drugs (Mihaly and Morgan, 1984). In contrast, the maternal administration studies with acipimox, which diffuses slowly across the placenta, showed that there is a substantial delay to equilibrium between maternal and fetal drug concentrations. The potential in vivo consequences of this slow equilibrium are that peak fetal plasma drug concentration will be lower than the maternal peak concentration during maternal treatment (Mihaly and Morgan, 1984), thereby attenuating the fetal effects of the drug. In conclusion, this study shows that the placental transfer of acipimox is slow and consistent with free diffusion between the maternal and fetal compartments. There is no evidence for active placental transport of this drug. The relative impermeability of the placenta may afford some protection to the fetus from acipimox administered to the mother in vivo.
ACKNOWLEDGEMENTS We are grateful to Dr J. D. Paull and Mr S. Ziccone of the Department of Anaesthetics, Royal Women’s Hospital, Melbourne, Australia,for their assistance, to Dr G. W. Miialy for valuablediscussions and to FarmitaliaCarlo Erba for financial assistance.
REFERENCES Bawdon, R. E. & Fortunate, S. J. (1989) Determination of pentamidine transferin the in vitro perfused human cotyledon with high-performanceliquid chromatography.Atin3oe~~ of Obstetrkx und Gytre&gr, 160,759761.
Challier, J. C., Nan&kumanm, M. & Mondon, F. (1985) Placental transportof hexoses: A comparativestudy witb antipyrhteand amino acids. Pkma, 6,497-504. Cl&g, M. S., MIbaly, G. W., Morgan, D. J., Date, N. M., Hardy K. J. & SrbaIIwood, R. A. (1987) Low clearance of cimetidine across the human placenta.30~~1 of Phemumlogy and Eipenmental Z?w~4~eutics, 241, 1006-1009. Cl&g, M. S., Czuba, M. A., Mihaly, G. W., Morgan, D. J., Hyman, K. M., PauIl, J. D. & SmaRwo&, R. A. (1988) Mechanism of triamterenetransfer across the human placenta.30~~1 ofPhamam&gy and EE;ptimental nImpe#ics, 2%,1093-1097. Dpocis, J. WI&son,D., Ho&ins, I. A. & Levitz, M. (1988a) Placentaltransferof thiamine in the human subject:In vitro perfusion studies and maternal-cord plasma concentrations.American30umaZ of Obstetrics und Gynecology, 159,1435-1439.
Ghabrial et al: Transfer ofActpimoxAcross Human Placenta
661
Dancis, J., Lehanka, J. & Levitz, M. (1988b) Placental transport of riboflavin: Differential rates of uptake at the maternal and fetal surfaces of the perfused human placenta. American Journal of Obstetricsand Gynecology,158, 204-210. Contractor, S. F. & Stannard, P. J. (1983) The use of AIB transport to assess the suitability of a system of human placental perfusion for drug transport studies. Placenta, 4,19-30. Fortunate, S. J,, Bawdon, R. E., Baum, M. (1988) Placental transfer of cefoperazone and sulbactam in the isolated in vitro perfused human placenta. AmericanJournal of Obstetricsand Gynecology,159,1002-1006. Fortunate, S. J. & Bawdon, R. E. (1989) Determination of pentamidine transfer in the in vitro perfused human of Obstetricsand Gynecology,160,759cotyledon with high-performance liquid chromatography. Amticanjoumal 761. Fuccella, L. M., Goldaniga, G., Lovisolo, P., Maggi, E., Musatti, L., Mandelli, V. & Sirtori, C. R. (1980) Inhibition of lipolysis by nicotinic acid and by acipimox. ClinicalPharmacologyand Therapeutics,28, 790-794. Guerre-Milo, M., Rey, E., Challier, J, C., Turquais, J. M., d’Athis, Ph. & Olive, G. (1979) Transfer in vitro of three benzodiazepines across the human placenta. EuropeanJournal of ClinicalPharmacology,15,171-173. Mihaly, G. W. & Morgan, D. J. (1984) Placental drug transfer: Effect of gestational age and species. Pharmacology and Therapeutics,23,253-266. Mirkin, B. L. & Singh, S. (1976) Placental transfer on pharmacologically active molecules. In Perinatal Pharmacologyand Therapeutics, (Ed) Mirkin B. L. pp l-70. London: Academic Press. Nandakumaran, M., Gardey, C. L., Challier, J. C., Richard, M. O., Panigel, M. &Olive, G. (1981) Transfer of Salbutamol in the human placenta in vitro. DevelopmentalPharmacologyand Therapeutics,3,88-98. Penfold, P., Drury, L., Simmonds, R. & Hytten, F. E. (1981) Studies of a single placental cotyledon in vitro 1. The preparation and its viability. Placenta, 2, 149-154. Purves, R. D. (1989) Users Guide to Minim 1.6. Department of Pharmacology, Medical School, University of Otago, Dunedin, New Zealand. Schneider, H., Mohlen, K. H. & Dancis, J. (1979) Transfer of amino acids across the in vitro perfused human placenta. Pediatric Research, 13, 236240. Schneider, H., P&gel, M. & Dan&, J. (1972) Transfer across the perfused human placenta of antipyrine, sodium and leucine. American3oumal of Obstetricsand Gynecology, 114, 822-828. Schneider, H. Proegler, M., Sodha, R. & Dancis, J. (1987) A symmetrical transfer of a-aminoisobutyric acid (AIB), leucine and lysine across the in vitro perfused human placenta. Placenta, 8, 141-151. Shargel, L. Cheung, W. M. & Yu, A. B. C. (1979) High pressure liquid chromatographic analysis of antipyrine in journal ofPhamzaceutical.Sciences,68, 1052-1053. small plasma samples. Simon, W. (1972) Mathematical techniques for physiology and medicine. New York: Academic Press, pp. 52-56. Sirtori, C. R., Gianfunceschi, G., Sirtori, M., Bemini, F., Descovich, G., Montaguti, U., Fuccella, L. M. & Muscatti, L. (1981) Reduced triglyceridemia and increased high density lipoprotein cholesterol levels after treatment with acipimox, a new inhibitor of lipolysis. Atherosclerosis,38,267-271. Stuyt, P. M. J., Stalenboef, A. F. H., Demacker, P. N. M. & Van? Laar, A. (1985) Comparative study of the effects of acipimox and colifibrate in type III and type IV hyperlipoproteinemia. Atherosclerosis,55,s l-62. Taskinen, M. R. & Nikkila, E. A. (1988) Effect of acipimox on serum lipids, lipoproteins and lipolytic enzymes in hypertriglyceridemia. Atherosclerosis,69, 249-255.
Transfer of Acipimox Across the Isolated Perfused Human Placenta HAN-Y GHABRIAL, MARK A. CZUBA, CHERYL K. STEAD, RICHARD A. SMALLWOOD & DENIS J. MORGAN” University of Melbourne, Department of Medicine, Repatriation General Hospital, Melbourne, Victoria Victorian College of Pharmacy, Melbourne, Victoria, Australia at: Victorian College of Pharmacy, 381 Royal Parade, Parkville, Melbourne, Victoria, 3052 Australia n To tahom correspondence should be dressed
Paperaccepted 13.5.1991
SUMh%ARY The placental transfer of the new lipid-lowering agent, acipimox was investigated in the isolated perfused human placenta. Placentas obtained at caesarean section were perfused for 120 min, with both maternal and fetal circuits in closed recycling mode. Acipimox was added to either the maternal circuit alone eve experiments) or to both maternal and fetal circuits simultaneously @ve experiments) to achieve initial concentrations ofSug/ml. Antipyrine (IOpg/ml) and l-(14C)-leucine (250pM) were added in like fashion as reference compounds. Two hours after addition to the maternal circuit alone antipyrine was close to equilibrium across the placenta, but equilibration of acipimox was incomplete fetal/maternal ratio = 0.58 + 0.11). Maternal to fetal placental clearance of actpimox (0.80 + 0.18 ml/min) was 25per cent of antipyrine clearance. AJter simultaneous administration to both maternal and fetal n’rcuits the 1-(‘4C)-leucinefetal/maternal ratio was 1.44 It 0.13 at 120 min, whereas maternal and fetal concentrations of acipimox and antipyrine were at equilibrium for the duration of the experiment (fetal/maternal ratio of acipimox at 120 min = 1.10 f 0.06). This study shows that acapimox is transfered across the human placenta by da&ion at a slow rate. The low permeabilityof the placenta may afford some protection to the fetus fromacipimox administered to the mother in vivo.
Acipimox (5-methylpyrazine carboxylic acid-4-oxide) is a new lipid-lowering agent (Fuccella et al, 1980) with a molecular structure similar to nicotinic acid. Acipimox appears to modify lipoprotein metabolism (Sirtori et al, 1981; Stuyt et al, 1985; Taskinen & Nikkila, 1988) and may be useful in patient type II and IV hyperlipoproteinaemia and hypertriglyceridaemia (Taskinen & Nikkila 1988). Such patient groups may include women of child-bearing potential, but there is no information concerning the likely exposure of the fetus to acipimox administered to the mother. 0143-4004/91/060653
+ 09 $05.00/O
0
1991 Baillike Tindall Ltd
654
Placenta(I 991), Vol. 12
This drug presents several interesting characteristics for the study of placental transfer. It is ionized at physiological pH (carboxylic acid group pKa = 3.25) and is relatively polar. These factors would be expected to lessen placental permeability for this drug (Mihaly & Morgan, 1984). However, acipimox resembles the vitamin nicotinic acid, and as active maternal to fetal transport occurs with other vitamins, such as thiamine (Dancis et al, 1988a) and riboflavin (Dancis et al, 1988b), acipimox may also be actively transported to the fetus. We have studied the placental transfer of acipimox in the isolated perfused human placenta and compared its rate of transfer with those of a freely diffusible molecule, antipyrine, and an actively transported amino acid, l-[‘4C]-leucine.
METHODS Perfusion techniques The perfusion technique was based on the method of Penfold et al (1981), as modified by Ching et al (1987, 1988). Briefly, placentas were obtained within 5 min of delivery by caesarean section and the fetal chorionic artery and vein supplying a peripheral cotyledon were cannulated with vinyl tubing and cleared of blood by flushing with warm, heparinized, Compound Sodium Lactate Injection B.P. at 3 ml/mm for 5 min. The placenta was placed on a platform with the perfused cotyledon exposed and supported by plastic mesh (3 mm grid) with the fetal side uppermost. Two blunt-ended cannulas (made from 19 gauge needles) were inserted 0.5 cm into the maternal decidual basal plate and the maternal perfusate flow rate was increased slowly to 14-15 ml/mm. Fetal flow rate was 7.5-8 ml/mm. The pet&ate consisted of Krebs-Ringer phosphate-bicarbonate electrolyte solution with added 30 g/l w/v dextran (MW 70 000), 1 g/I w/ v b ovine serum albumin and 1 g/l w/v glucose. The placenta and maternal (600 ml) and fetal (150 ml) per&sate reservoirs were housed in a thermostatically controlled cabinet at 37°C. Each circuit was pumped separately using an LKB peristaltic pump in a closed recycling mode. Perfusate passed from the reservoir, through a coarse filter (pore size approximately 1 mm), silastic membrane oxygenator and a bubble trap to the arterial inflow cannula. Indices of placental viability were maternal oxygen consumption > 1.5 ~mol/min, stable perfusate backpressure ~60 mmHg, with the absence (i.e. ~10 ml/h) of fetal pet&sate leakage, and active I-[‘4C]-leucine transfer from mother to fetus. Experimental design Maternal Administration. In five placental preparations the transfer of acipimox, 1-[‘4C]-
leucine and antipyrine was examined following administration to the maternal reservoir only. The maternal reservoir contained an initial concentration of 5 &ml acipimox (Farmitalia Carlo Erba, Melbourne, Australia), 20 ,~g/rnl antipyrine and 0.25 mu l-[‘4C]-leucine (specific activity 67,~Ci/nunol) (Amersham International, Buckinghamshire, UK). Per&sate (2 ml) was collected from the fetal and maternal circuits before and every 10 min after dosing for 120 min, and an equal amount of blank perfusate was added to replace that lost due to sampling. Maternal and Fe&Administration. This experiment was similar in design to that for maternal
dosing, except that in the five placental preparations used, both maternal and fetal reservoirs
Ghabrial et al: Transjir ofAcipimoxAcross Human Placenta
655
were spiked to the above concentrations. Perfusate samples were taken from both maternal and fetal circuits before and every 15 min after dosing, for 120 min. In one study sampling was extended to 150 min. Protein Binding of Acipimox Binding of acipimox to bovine serum albumin in perfusate solution was assessed by equilibrium dialysis. Solutions of 0, 1,5 and 10 g/l (w/v) of albumin in perfusate were spiked to give a concentration of 5 pg/ml of acipimox. Two millilitres of this solution were dialyzed against protein-free perfusate (2 ml, pH 7.4) in perspex chambers separated by cellulose dialysis membrance (type 20, Union Carbide Corp, NY). Dialysis cells were incubated at 37°C for 12 h in an oscillating water bath. Unbound fraction was calculated as the ratio of acipimox concentration in buffer/acipimox concentration in dialyzed perfusate at equilibrium. Assays Aczpimox: To a screw-topped culture tube was added 1 ml of sample, 2 pg of sulphanilamide (in 100~1) and 1 ml of 1Mphosphoric acid. To this, 10 ml of ethylacetate/isopropanol(90: 10) was added, the tube was capped and vortexed for 3 min, and centrifuged. The organic layer was transferred to a tapered glass tube and evaporated to dryness. The residue was reconstituted in 150 ~1 of mobile phase of which 20 ~1 were injected into the HPLC. A standard curve was constructed by plotting the peak area ratio of acipimox to sulphanilamide versus known concentrations. The standard concentrations were 0,0.1,0.5, 1,2, 5 and 10 &ml. Quality control samples of known concentrations were analysed with each assay run. The HPLC system comprised an M6000A pump, a Novapak phenyl 4 pm radial compression module housed in a Z-module, a Lambda max 481 variable wavelength detector set at 264 nm and a Maxima 820 data station to collect and process the signal from the detector (all from Milliporen;Vaters, Milford, Massachusetts). The mobile phase was 25 mu di-potassium hydrogen phosphate buffer containing 40 ml of methanol, 5 mmol of tetrabutyl ammonium and 5 mmol triethylaminefl with pH adjusted to 6.8 using concentrated phosphoric acid. This was delivered at 3 ml/min. The coefficient of variation (precision) was 4.6 per cent (n = 6) at 1 pug/ml and 2.9 per cent (n = 6) at 5 pug/ml and accuracy was 7 per cent at 1 &ml and 4.2 per cent at 5 pug/ml. Antipyrine: Antipyrine was assayed by HPLC according to the method of Shargel et al (1979). Precision and accuracy were 6 per cent and 9.5 per cent, respectively, at a concentration of 6.25 pug/ml. l-[‘4C/-leucine: Aquasol (DuPont) (10 ml) was added to 200 ~1 of sample and counted on a Packard Scintillation counter, Model 3255. Quench and efficiency correction was carried out using external standardization. Calculations In maternal dosing experiments the transplacental clearance was calculated by fitting the following equation which describes equilibration between two compartments (Simon, 1972) to the fetal perfusate concentration data.
Placenta(1991), Vol. I2 cF,t
=
Where C,,t is fetal perfusate concentration at time t, VM and V, are maternal and fetal perfusate volumes respectively and Kis the transfer rate constant. Placental clearance (CL), which represents the ease of transfer of the substance across the placenta, is given by (Simon, 1972): cL
= W?M*~F) VM +
(2)
b
Kwas calculated by fitting equation (1) to.CF,t versus time data obtained from the maternal dosing experiments. Fitting was by least squares non-linear regression using the GaussNewton-Marquardt method as performed on the computer program Minim 1.6 (Purves, 1989). Recovery of the dose was calculated as follows: (3)
RESULTS Maternal administration Figure, 1 shows the increasing concentrations of antipyrine and acipimox in the fetal perfusate for the first 120 min. At 120 min the fetal concentration of antipyrine had almost reached a
0
20
40
60
60
loo
I2C
Time (min) Figure 1. Typical fetal perfusate concentration prohles for antipyrine (0) and acipimox (+) in a maternal administration study. The fit (solid line) acuw&tg to equation (1) is shown and yiekkd the f&wing parameters: antipyrine-A = 13.5pg/ml, K = 0.01% min-‘, CL = 2.94 ml&n, ? = 0.9%; and acipimox-A = 3.9O@nl, K = 0.0066 min-‘, CL = 0.99 ml/m& 2 = 0.979.
Ghabrial et al: Transfer ofAcipimoxAcross Human Placenta
657
Time (min)
Figure 2. Mean z!z s.e.m. fetal/maternal ratios of perfusate I-[14C]-leucine (W) in the maternal administration studies.
concentrations
for antipyrine
(O), acipimox (+) and
plateau, whereas the fetal concentration of acipimox was increasing steadily. The ratios of fetal/maternal perfusate drug concentrations (F/M ratio) over this 120 min period are shown in Figure 2. At 120 min the F/M ratio for antipyrine was 0.94 + 0.11 (means + s.d.), indicating that fetal and maternal concentrations of this drug were very close to equilibrium. In contrast, at 120 min the F/M ratio for acipimox was only 0.58 f 0.11. Fetal perfusate concentrations of l-[14C]-leucine were still rising at 120 min, with the F/M ratio at this point being 1.02 k 0.12. Equation (1) fitted the fetal perfusate drug concentration profiles well and typical fits are shown in Figure 1. Parameter values and placental clearance values, calculated from equation (2), are shown in Table 1. The placental clearance of acipimox was 0.80 + 0.18 ml/ min, which was approximately one-quarter that of antipyrine (3.61 + 1.20 ml/min). The
Drug
Table 1. Kinetics of antipyrine,
acipimox and l-[‘4C]-leucine
120 min F/M ratio
K” (min-‘)
CLb (mLhin)
in maternal
dosing study Recovery W)
Clearance ratio’
Antipyrine
0.94 (0.11)
12.8 (0.79)
0.030 (0.010)
3.61 (1.20)
79.7 (4.92)
Acipimox
0.58 (0.11)
3.59 (0.47)
0.007 (0.002)
0.80 (0.18)
89.9 (11.8)
0.24 (0.08)
I-[14C]-leucine
1.02 (0.12)
(Z;
0.023 (0.004)
2.70 (0.45)
73.1 (18.5)
0.79 (0.18)
Means of five placental preparations with standard deviations in parenthesis. u From equation (l), A = dose/(VM + VF). b Placental clearance, from equation (2). ’ Clearance ratio of compound to that of antipyrine. d d/min/ml.
658
Plactwa (1991), Vol.12
0
15
30
45
60
75
90
105
120
Time (min)
Figure 3. Mean f s.e.m. fetal/maternal ratios of perfusate concentrations of antipyrine (Cl), acipimox (+) and I-[ 4C]-leucine (W) in maternal and fetal administration studies.
mean percentage of acipimox recovered in maternal and fetal perfusates at the end of the experiment [equation (3)] ranged from 73.7 per cent for l-[‘4C]-leucine to 89.9 per cent for acipimox (Table 1). Maternal and fetal administration Figure 3 depicts the mean F/M ratio in the experiments in which fetal and maternal drug concentrations were the same at the start. The F/M ratios of antipyrine and acipimox remained relatively constant over the 120 min whereas the F/M ratio of l-[‘4C]-leucine rose steadily throughout the experiment, reaching 1.44 f 0.13 at 120 min. At 120 min the F/M ratios of antipyrine (1.06 + 0.06) and acipimox (1.10 f 0.06) did not differ significantly, and their 95 per cent confidence intervals included the value of 1. At 120 min, the F/M ratio of l-[‘4C]-leucine was significantly greater than those of antipyrine and acipimox (P < 0.05). Placental viability Mean oxygen consumption by the perfused cotelydon was 5.08 f 0.56~moVmin in maternal administration experiments and 4.80 If: 1.67 pmol/min in dual administration experiments. Values were not corrected for placental weight as the whole placenta was not perfused. These values are similar to those obtained by us previously (Ching et al, 1987,1988). Protein binding No protein binding of acipimox was evident. After 12 h of dialysis the concentration of acipimox was similar in both chambers, regardless of the protein concentrations.
Ghabrial et al: Transfer
ofAcipimox Across Human Placenta
659
DISCUSSION Antipyrine diffuses rapidly across the placenta, so that the limiting factor in transfer is placental blood flow rate (Schneider et al, 1972). Antipyrine is therefore a useful reference compound for assessing the placental permeability and transfer rate of other compounds. In our study the mean placental clearance of antipyrine was 3.61 + 1.20 ml/min. This is somewhat greater than found previously [2.15 f 0.11 ml/min, Ching et al (1987); 2.71 + 0.66 ml/min, Ching et al (1988), probably representing a greater degree of overlap of the maternal and fetal circulations studied]. The placental transfer of acipimox was appreciably slower than that of antipyrine (Figure 1). Fitting the equation describing diffusion of drug from the maternal to the fetal compartment to the maternal dosage data yielded a placental clearance value that was only 2.5 per cent that of antipyrine. Protein binding of acipimox to bovine serum albumin (l-10 g/l) was negligible, therefore placental transfer rate of acipimox would not have been affected by protein binding in the perfusate, which contained 1 g/l albumin. In maternal administration experiments the mean recoveries of antipyrine, acipimox, and l-[14C]-leucine at the end of the experiment, calculated using equation (3), ranged from 75-90 per cent (Table 1). The drug not accounted for was probably taken up by placental tissue, with the relatively high recovery for acipimox indicating a low affinity of this drug for placental tissue. Most drugs studied in vitro and in vivo have shown a rapid placental transfer from mother to fetus (Mihaly and Morgan, 1984). However, the placental clearance of acipimox was relatively low, being only 25 per cent that of antipyrine. A number of other drugs have been found to have a similarly low placental clearance, including clorazepate (Guerre-Milo et al, 1979), cimetidine (Ching et al, 1987), cefoperazone, sulbactam (Fortunato et al, 1988) and pentamidine (Fortunato and Bawdon, 1989). This indicates that the permeability of the placenta is relatively low for these drugs, although not as low as for inulin, for which the placenta is highly impermeable, having a placental clearance of only 7 per cent relative to antipyrine (Nandakumaran et al, 1981; Contractor and Stannard, 1983). The physicochemical properties of a drug that determine its placental permeability are primarily lipid solubility, degree of ionization in blood and molecular weight (Mirkin and Singh, 1976). Antipyrine is highly lipid soluble, unionized at physiological pH and has a molecular weight of 188.2. Acipimox (MW = 154.1) is a relatively polar compound (freely soluble in water, insoluble in ether) and with a pK, of 3.25 is fully ionized at physiological pH. Therefore, these latter two properties probably both contribute to the low placental permeability of acipimox. In contrast, the low placental permeability of cimetidine appears to be due primarily to its low lipid solubility only, because it is predominantly unionized at physiological pH (Ching et al, 1987). Acipimox is a structural analogue of the vitamin nicotinic acid. The mechanism of placental transfer of nicotinic acid has not been investigated, but the mechanism of placental transfer of several other vitamins, including thiamine (Dancis et al, 1988a) and riboflavine (Dancis et al, 1988b) involves active maternal to fetal transport. If there were also an active placental transport mechanism for nicotinic acid, acipimox might be a substrate for this process and this could augment the exposure of the fetus to maternally administered drug. The maternal and fetal administration experiments were designed to detect the presence of active transport of acipimox. Figure 3 shows that when maternal and fetal perfusate concentrations were equal at the start of the experiment, they remained equal for the 120 min duration of the experiment. This finding rules out the possibility of active transport of acipimox in our preparation. To
660
Pkumta (1991), Vol. 12
demonstrate that the preparation was capable of active transport, we also included equal concentrations of l-[14C]-leucine in maternal and fetal perfusates at the start of the experiment. l-leucine is considered to be a useful indicator of active maternal to fetal placental transfer (Schneider et al, 1987). Figure 3 shows that in the maternal and fetal administration experiments, the F/M ratio of I-[‘4C]-leucine reached 1.44 by 120 min. Although some metabolism of the leucine during the 120 min period cannot be excluded (Schneider et al, 1987) the accumulation of l-[14C]-leucine in the fetal perfusate reflects active transport of this substance. This F/M ratio of 1.44 is similar to values obtained previously by us and others (Ching et al, 1987, 1988; Challier et al, 1985; Schneider et al, 1979). For drugs that rapidly diffuse across the placenta, equilibrium between maternal and fetal blood concentrations of drug will be rapid. Therefore, the degree of fetal exposure will depend on the equilibrium between maternal and fetal blood rather than the rate of placental transfer (Mihaly and Morgan, 1984). The relative maternal and fetal blood concentrations at equilibrium are determined primarily by maternal and fetal factors (e.g. clearance, blood pH, plasma protein binding) and the placenta plays an insignificant role for such drugs (Mihaly and Morgan, 1984). In contrast, the maternal administration studies with acipimox, which diffuses slowly across the placenta, showed that there is a substantial delay to equilibrium between maternal and fetal drug concentrations. The potential in vivo consequences of this slow equilibrium are that peak fetal plasma drug concentration will be lower than the maternal peak concentration during maternal treatment (Mihaly and Morgan, 1984), thereby attenuating the fetal effects of the drug. In conclusion, this study shows that the placental transfer of acipimox is slow and consistent with free diffusion between the maternal and fetal compartments. There is no evidence for active placental transport of this drug. The relative impermeability of the placenta may afford some protection to the fetus from acipimox administered to the mother in vivo.
ACKNOWLEDGEMENTS We are grateful to Dr J. D. Paull and Mr S. Ziccone of the Department of Anaesthetics, Royal Women’s Hospital, Melbourne, Australia,for their assistance, to Dr G. W. Miialy for valuablediscussions and to FarmitaliaCarlo Erba for financial assistance.
REFERENCES Bawdon, R. E. & Fortunate, S. J. (1989) Determination of pentamidine transferin the in vitro perfused human cotyledon with high-performanceliquid chromatography.Atin3oe~~ of Obstetrkx und Gytre&gr, 160,759761.
Challier, J. C., Nan&kumanm, M. & Mondon, F. (1985) Placental transportof hexoses: A comparativestudy witb antipyrhteand amino acids. Pkma, 6,497-504. Cl&g, M. S., MIbaly, G. W., Morgan, D. J., Date, N. M., Hardy K. J. & SrbaIIwood, R. A. (1987) Low clearance of cimetidine across the human placenta.30~~1 of Phemumlogy and Eipenmental Z?w~4~eutics, 241, 1006-1009. Cl&g, M. S., Czuba, M. A., Mihaly, G. W., Morgan, D. J., Hyman, K. M., PauIl, J. D. & SmaRwo&, R. A. (1988) Mechanism of triamterenetransfer across the human placenta.30~~1 ofPhamam&gy and EE;ptimental nImpe#ics, 2%,1093-1097. Dpocis, J. WI&son,D., Ho&ins, I. A. & Levitz, M. (1988a) Placentaltransferof thiamine in the human subject:In vitro perfusion studies and maternal-cord plasma concentrations.American30umaZ of Obstetrics und Gynecology, 159,1435-1439.
Ghabrial et al: Transfer ofActpimoxAcross Human Placenta
661
Dancis, J., Lehanka, J. & Levitz, M. (1988b) Placental transport of riboflavin: Differential rates of uptake at the maternal and fetal surfaces of the perfused human placenta. American Journal of Obstetricsand Gynecology,158, 204-210. Contractor, S. F. & Stannard, P. J. (1983) The use of AIB transport to assess the suitability of a system of human placental perfusion for drug transport studies. Placenta, 4,19-30. Fortunate, S. J,, Bawdon, R. E., Baum, M. (1988) Placental transfer of cefoperazone and sulbactam in the isolated in vitro perfused human placenta. AmericanJournal of Obstetricsand Gynecology,159,1002-1006. Fortunate, S. J. & Bawdon, R. E. (1989) Determination of pentamidine transfer in the in vitro perfused human of Obstetricsand Gynecology,160,759cotyledon with high-performance liquid chromatography. Amticanjoumal 761. Fuccella, L. M., Goldaniga, G., Lovisolo, P., Maggi, E., Musatti, L., Mandelli, V. & Sirtori, C. R. (1980) Inhibition of lipolysis by nicotinic acid and by acipimox. ClinicalPharmacologyand Therapeutics,28, 790-794. Guerre-Milo, M., Rey, E., Challier, J, C., Turquais, J. M., d’Athis, Ph. & Olive, G. (1979) Transfer in vitro of three benzodiazepines across the human placenta. EuropeanJournal of ClinicalPharmacology,15,171-173. Mihaly, G. W. & Morgan, D. J. (1984) Placental drug transfer: Effect of gestational age and species. Pharmacology and Therapeutics,23,253-266. Mirkin, B. L. & Singh, S. (1976) Placental transfer on pharmacologically active molecules. In Perinatal Pharmacologyand Therapeutics, (Ed) Mirkin B. L. pp l-70. London: Academic Press. Nandakumaran, M., Gardey, C. L., Challier, J. C., Richard, M. O., Panigel, M. &Olive, G. (1981) Transfer of Salbutamol in the human placenta in vitro. DevelopmentalPharmacologyand Therapeutics,3,88-98. Penfold, P., Drury, L., Simmonds, R. & Hytten, F. E. (1981) Studies of a single placental cotyledon in vitro 1. The preparation and its viability. Placenta, 2, 149-154. Purves, R. D. (1989) Users Guide to Minim 1.6. Department of Pharmacology, Medical School, University of Otago, Dunedin, New Zealand. Schneider, H., Mohlen, K. H. & Dancis, J. (1979) Transfer of amino acids across the in vitro perfused human placenta. Pediatric Research, 13, 236240. Schneider, H., P&gel, M. & Dan&, J. (1972) Transfer across the perfused human placenta of antipyrine, sodium and leucine. American3oumal of Obstetricsand Gynecology, 114, 822-828. Schneider, H. Proegler, M., Sodha, R. & Dancis, J. (1987) A symmetrical transfer of a-aminoisobutyric acid (AIB), leucine and lysine across the in vitro perfused human placenta. Placenta, 8, 141-151. Shargel, L. Cheung, W. M. & Yu, A. B. C. (1979) High pressure liquid chromatographic analysis of antipyrine in journal ofPhamzaceutical.Sciences,68, 1052-1053. small plasma samples. Simon, W. (1972) Mathematical techniques for physiology and medicine. New York: Academic Press, pp. 52-56. Sirtori, C. R., Gianfunceschi, G., Sirtori, M., Bemini, F., Descovich, G., Montaguti, U., Fuccella, L. M. & Muscatti, L. (1981) Reduced triglyceridemia and increased high density lipoprotein cholesterol levels after treatment with acipimox, a new inhibitor of lipolysis. Atherosclerosis,38,267-271. Stuyt, P. M. J., Stalenboef, A. F. H., Demacker, P. N. M. & Van? Laar, A. (1985) Comparative study of the effects of acipimox and colifibrate in type III and type IV hyperlipoproteinemia. Atherosclerosis,55,s l-62. Taskinen, M. R. & Nikkila, E. A. (1988) Effect of acipimox on serum lipids, lipoproteins and lipolytic enzymes in hypertriglyceridemia. Atherosclerosis,69, 249-255.
