- Email: [email protected]
Pharmacokinetics in dogs after oral administration of two different forms of ascorbic acid
Research in Veterinary Science 2001, 71, 27±32 doi:10.1053/rvsc.2001.0480, available online at http://www.idealibrary.com on
Pharmacokinetics in dogs after oral administration of two different forms of ascorbic acid S. WANGz*, G. E. BERGEx, N. O. HOEMz, R. B. SUNDzy z
Department of Pharmacology, School of Pharmacy, University of Oslo, x Nordberg Small Animal Clinic, Oslo SUMMARY
The dog is able to synthesise ascorbic acid (AA), but is frequently given the vitamin in an attempt to improve health and performance. The pharmacokinetics of AA in this species, however, are not well studied. Using a selective analytical method and careful stability control, the pharmacokinetics of orally given AA was studied in 20 dogs, at two dosage levels (15 and 50 mg kgÿ1) and with two forms of supplement [crystalline AA and the vitamin C product Ester-C1 (Inter-Cal Corp., Prescott, AZ, USA)]. After oral administration, a rapid increase was found in the plasma level of AA, indicating a possible intestinal active transport mechanism in this species. The obtained Cmax and AUC values were found to increase in a non-linear fashion when the dose of AA was increased. The pharmacokinetic modeling of the elimination of AA was made difficult by a pronounced secondary peak appearing after about 9 hours. The comparison of crystalline AA and Ester-C1 did not indicate any significant differences in pharmacokinetic parameters between the two preparations of the vitamin. # 2001 Harcourt Publishers Ltd
ASCORBIC acid (AA) has been administered both orally and i.v. to dogs to improve health and performance, despite the fact that this species is able to synthesise the vitamin. In dogs, AA supplementation has been given to reduce the incidence of hip dysplasia (HD). Belfield (1976) studied puppies from parents with HD, and found that giving AA to the pregnant mothers and their puppies resulted in dogs free HD. Ascorbic acid has also been tried in the treatment of hypertrophic osteodystrophia (HOD) in dogs, with negative results (Grùndalen 1976, Teare et al 1979). Few publications have focused on monitoring plasma concentrations in dogs after AA administration, to evaluate the effect of AA supplementation in pharmacokinetic terms. Teare and co-workers (1979) studied plasma AA after oral or i.v. supplementation, as a part of a study on HOD in dogs. Schulze and co-workers (1992) compared the plasma concentration of AA after oral administration of L-ascorbate-2-polyphosphate or AA. However, the numbers of animals studied were small, and a general pharamacokinetic evaluation was not the main focus in any of those papers. Concerning the large number of methods for determination of AA and the variability seen in existing analytical methods, special attention should be paid to ensure reliable measurements, particularly regarding analytical selectivity. In addition, stability control must be ensured, as AA is easily degraded in samples and in
other aqueous solutions. This is particularly important for ambulatory reserch, as will often be the case in studies on domestic animals. A previously described method was used for AA determination based on HPLC and electrochemical detection, as well as stability data in samples from dogs (Wang et al 1995), and the method as well as the sampling and preparation procedures used in the present paper are based on this publication. Several products and derivatives of AA have been used for supplementation in dogs. L-ascorbate-2polyphosphate has been tried as a source of AA (Schulze et al 1992) to overcome the stability problem for ordinary AA in the diet. Ester-C1a (Inter-Cal Corp., Prescott, AZ, USA), a vitamin C product consisting of ascorbic acid, dehydroascorbic acid and calcium as well as minor amounts of smaller metabolites of AA, is another pharmaceutical formulation of the vitamin that has been given to dogs (Berge 1990). This formulation of vitamin C has been claimed to show a higher rate and degree of absorption compared to ordinary AA when given to rats (Bush and Verlangieri 1987), but no study has been performed to compare the two products in dogs. The main purpose of this work was to evaluate the pharmacokinetics of AA given orally to healthy dogs, using a rapid and selective analytical method for AA determination and careful stability control. Secondarily, we wanted to compare the kinetics of ordinary AA and Ester-C1a.
y
Deceased. *Corresponding author: Tùnsberg Hospital Pharmacy, Box 2168 Postterminalen, N-3103 Tùnsberg, Norway; Tel: 47 33 34 31 04; Fax: 47 33 31 34 67; E-mail: [email protected]
0034-5288/01/010027 06 $35.00/0
a
The product is also named C-flex1 when marketed as a veterinary supplement (powder). # 2001 Harcourt Publishers Ltd
28
S. Wang, G. E. Berge, N. O. Hoem, R. B. Sund
MATERIALS AND METHODS Animals The study included a total of 20 animals (eight German Shepherds, 10 Labrador Retrievers and two Riesenschnaussers) aged 1±10 years. All animals were healthy and had not been given supplementary vitamins prior to the study. Their regular feed was not declared to contain AA. The experiments were performed after an overnight fast, and no feed was given during the study period. Water was available ad libitum. All animals were placed under similar conditions and had about the same (low) activity level. Study design A total of 32 experiments were included in the study (Table 1). The kinetics of high-dose AA (50 mg kgÿ1) were investigated in six dogs (Group 1; two Riesenschnaussers, four Labrador Retrievers) given a single oral dose of crystalline AA (cAA) (Sigma Chemical Company, St. Louis, MO, USA). Six additional experiments (Group 2) were performed with a corresponding dose of Ester-C1 (Inter-Cal Corp., Prescott, AZ, USA). All high-dose experiments were cross-over designed to facilitate a better comparison of the pharmacokinetics of the two formulations of AA. The kinetics of low-dose cAA (15 mg kgÿ1) were investigated in eight dogs (Group 3; two Labrador Retrievers, six German Shepherds). In addition, eight experiments (Group 4; two Labrador Retrievers, six German Shepherds) were performed with a corresponding dose of Ester-C1. Eight of the 16 experiments at low dose were cross-over designed. The supplements were administered in gelatine capsules together with ca. 20 ml water. A wash-out period of 3 weeks was allowed between the series making the cross-over experiments. To study the diurnal variation in endogenous AA levels, plasma AA was monitored in two dogs (Labrador Retrievers) not given any supplement. The monitoring was repeated after 3 weeks. In all instances, blood samples for plasma AA determination were taken before supplement administration (if administered), at intervals during 12 hours and after about 24 hours. Blood sampling and ascorbic acid analysis Sampling and analytical procedure as well as stability data are described in detail elsewhere (Wang et al TABLE 1: Number of experiments included in the study (grouping based on form of supplement and dosage) Dosage in AA equivalents
Form of supplement CAA
Ester-C1
None
Sum
8 Group 4 6 Group 2 ö
ö
16
ö
12
ö
8 Group 3 6 Group1 ö
4
4
Total
14
14
4
32
15 mg kg
ÿ1
50 mg kg ÿ1
1995). At the start of each experiment, all dogs were fitted with a catheter in Vena cephalica. Blood samples (5 ml) were collected into Vacutainer1 tubes containing heparin or EDTA. After centrifugation at least two 500 ml, aliquots of plasma were either immediately frozen (heparinised samples) in portable liquid nitrogen containers, or stabilised and deproteinised by perchloric acid with subsequent mixing, centrifugation and freezing in liquid nitrogen (samples treated with EDTA). The maximum time elapsed from sampling until freezing was 20 minutes. At the time of analysis, samples were thawed one by one and immediately prepared and analysed by a method based on high-pressure liquid chromatography and electrochemical detection (Wang et al 1995). Pharmacokinetic analysis Individual AA concentration±time curves were analysed by mono-exponential non-linear regression fits on the declining part of the curve and by non-compartmental pharmacokinetic methods. All dogs had substantial endogenous AA levels, and all measured concentrations were adjusted by subtraction of the endogenous start level. The elimination rate constant (kel) was estimated by a simple mono-exponential analysis. Non-linear regression fitting to a simple i.v. bolus injection model was utilised to extract an estimate of kel 1 from the declining part of the plasma concentration curves. Non-compartmental analysis was performed by calculating the area under the concentration±time curve (AUC) and the area under the first moment concentration-time curve (AUMC) by the liner trapezoid method. Because the curves reached close to zero level within the recorded period, most AUCs were not corrected for area by estimation of the final integral. Mean residence time (MRT), a measure for the mean time any molecule of the exogenous substance resides in the body after onset of absorption, was calculated as AUMC/AUC. For a bolus dose into a rapidly equilibrating (no profound distribution phase) system that can be described by simple first order kinetics the following relationship exists between MRT and the elimination (kel) and absorption rate (ka) constants: MRT
1 1 tlag kel ka
1
where tlag is the time delay of absorption, defined as the time before the plasma concentration increased by 5% over base level. After substituting these estimates of tlag and estimates of kel obtained from the mono exponential fits into the above expression, this relationship was used to estimate ka. Non-linear regression fitting was performed with the SAAM II pharmacokinetic analysis software, (SAAM Institute, University of Washington, USA). All data were fitted using relative weighting assigning a fractional SD of 010 to all data. The fitting procedure was evaluated by residual plots and comparison of total objective function value and the Akaike criterion.
29
Pharmacokinetics in dogs
Statistics Statistical analysis was performed by the S-PLUS 2000 software, MathSoft Inc. (USA). All statistical comparison between the groups were performed by nonparametric statistical methods, Mann±Whitney for non paired data and Wilcoxon signed-rank test for data from paired experimetns. A P-level less than 005 was considered statistically significant. RESULTS Initial plasma AA concentration (Cp0) was 402 68 mmol lÿ1 (n 32). In the experiments where supplement was given, plasma levels before and 24 hours after administration were 398 68 and 404 64 mmol lÿ1, respectively (n 28). The diurnal
FIG 1: The diurnal plasma regression, ^ ^ ^.
AA
plasma AA variation in animals not supplemented is shown in Figure 1. The fluctuation around the mean plasma AA level (412 mmol lÿ1) was 13% (n 4). The parameters obtained from the pharmacokinetic calculations are given in Table 2. All concentrations are relative to the individual pre-administration plasma AA concentration (Cp0), and calculated as the increase in plasma AA levels. Figures 2 and 3 show the relative plasma concentration±time curves for the dogs after p.o. administration of AA supplements (as cAA or Ester-C1) at the dosages of 15 and 50 mg kgÿ1, respectively. There was a notable variation in the plasma AA profiles after supplement administration. However, some structural features were prevalent: Tlag varied between 0 and 12 hours, and an apparent delay in absorption was observed in 12 of the 28
variation in animals not supplemented with vitamin C. Dog 1, ^*^; dog 2, ^&^; dog 3, ^^; dog 4, ^~±. Linear
TABLE 2: The pharmacokinetic parameters (mean SD) obtained after administration of crystalline AA (Group 1: 50 mg kg ÿ 1, Group 3: 15 mg kg ÿ 1) or Ester-C 1 (Group 2: 50 mg kg ÿ 1, Group 4: 15 mg kg ÿ 1) High dose (50 mg kg ÿ1)
Cmax (mmol l ÿ1)* Rel T max (hours) ÿ1 AUC (mmol hours l )* MRT (hours) kel (hours ÿ1) ka (hoursÿ1)
Low dose (15 mg kg ÿ1)
Group1
Group 2
Groups1 2
Group 3
Group 4
Groups 3 4
463(188) 36(155) 350(166) 64(18) 018(004) 077(028)
415(88) 28(184) 310(103) 62(17) 016(006) 117(069)
439(142) 32(05) 329(133) 63(17) 017(005) 097(055)
251(64) 271(111) 215(98) 73(27) 022(020) 099(029)
228(122) 301(137) 167(96) 69(22) 027(022) 092(030)
239(95) 28(12) 191(965) 71(24) 025(020) 096(029)
Cmax : Maximal plasma AA concentration after administration (pre-administration levels are subtracted). Rel T max : T max minus the estimated lag time. AUC: Area under the plasma ^concentration time curves (pre-administration levels are subtracted). MRT: Mean residence time, calculated as AUC/AUMC (area under the first moment concentration ^ time curve). kel: First order elimination rate constant, based on SAAM modelling. ka: First order obsorption constant based on estimate of MRT, lag time and kel. *For Cmax and AUC, the values obtained are based on the increase in plasma levels above the basal, pre-administration levels.
[AA]-time profiles 80.0
(a)
80.0
(b) G3-Exp13 G3-Exp14 G3-Exp17 G3-Exp19 G3-Exp21 G3-Exp23 G4-Exp15 G4-Exp16 G4-Exp18 G4-Exp20 G4-Exp22 G4-Exp24 G3-Exp25 G3-Exp26 G4-Exp27 G4-Exp28
Group3 Group4 70.0
60.0
[Ascorbic acid] (uM)
[Ascorbic acid] (uM)
50.0
40.0
30.0
70.0
60.0
50.0
40.0
30.0
20.0
20.0
10.0
10.0
0.0
0.0 0
5
10
15
0 5 20 25 Time after administration (hours)
10
15
20
25
FIG 2: The relative plasma concentration ^time curves after p.o. administration of AA 15 mg kg ÿ1 given as crystalline AA (Group 3) or Ester-C1 (Group 4). (a) Mean curves ( SD) for the two forms of supplement. (b) Individual curves. All plasma concentrations represent the increase in plasma levels above the basal, pre-administration levels.
[AA]-time profiles 80.0
(a)
80.0
(b) G1-Exp1 G1-Exp3 G1-Exp5 G1-Exp7 G1-Exp9 G1-Exp11 G2-Exp2 G2-Exp4 G2-Exp6 G2-Exp8 G2-Exp10 G2-Exp12
Group1 Group2 70.0
60.0
[Ascorbic acid] (uM)
[Ascorbic acid] (uM)
50.0
40.0
30.0
70.0
60.0
50.0
40.0
30.0
20.0
20.0
10.0
10.0
0.0
0.0 0
5
10
15
0 5 20 25 Time after administration (hours)
10
15
20
25
FIG 3: The relative plasma concentration ^ time curves after p.o. administration of AA 50 mg kg ÿ1 given as crystalline AA (Group 1) or Ester-C1 (Group 2). (a) Mean curves ( SD) for the two forms of supplement. (b) Individual curves. All plasma concentrations represent the increase in plasma levels above the basal, pre-administration levels.
Pharmacokinetics in dogs
experiments. The presence or degree of delay could be correlated to neither dose, form of supplement nor breed. In most animals the subsequent absorption phase was quite rapid, with a maximal concentration reached about 3 hours (Rel Tmax 30 14 hours) after estimated onset of absorption with a first order absorption constant ka estimated to 096 0409 hÿ1. In about one third of the experiments a clearly visible second peak was observed, appearing 6±10 hours administration, with a duration of 2±3 hours. Fitting the declining phase to a simple first order curve showed a rather high degree of model misfit, a phenomenon suggesting that not only the dogs with clearly visible second peaks showed this behaviour, but also the other dogs, but at a lower degree. The phenomenon was more prevalent in the high dose (50 mg kgÿ1) experiments in the Labrador Retrievers. No differences in kel values were seen between the different groups. All experiments taken together, the elimination rate constant kel was estimated to be 0214 016 hÿ1 (range 0053 to 0667). Pooling the data within the high-(Group 1 2) and low-(Group 3 4) dosage groups, the mean area under the plasma±AUC was increased from 191 mmol hÿ1 lÿ1 to 330 mmol hÿ1 lÿ1 when increasing the dose from 15 mg kgÿ1 to 50 mg kgÿ1. Such a lack of dose proportionality was also seen for the Cmax for the two dosages (mean 239 versus 439 mmol lÿ1). When comparing the two forms of vitamin C supplement (Group 1 versus Group 2, and Group 3 versus Group 4), no statistical significant differences were seen in the pharmacokinetic parameters, except for ka and tlag when perfroming a paired test between groups 1 and 2, when a borderline statistical significance was recorded. DISCUSSION Endogenous AA levels were recorded in all dogs, ranging from 265 to 499 mmol lÿ1 (402 68). These are in accordance with the values reported previously (Wang et al 1995). Basal plasma levels were recorded throughout 26 hours in four dogs, and it was found that the diurnal variation in plasma AA level was small. This finding was also confirmed by the small differences in AA levels found before (402 68) and at the end (404 64 mmol lÿ1) of each experiment. The pharmacokinetic parameters most sensitive to variations in basal plasma AA levels are Cmax and AUC. To minimise this effect when studying AA pharmacokinetics in the individual dogs, the baseline value was subtracted, thus calculating a `net' peak above the endogenous level. All Cmax and AUC values are therefore expressed as the increase in plasma AA levels above the endogenous level. The rapid increase in plasma AA after oral administration is in agreement with previous reports in dogs (Teare et al 1979; Schulze et al 1992). It is generally believed that for animals with the ability to synthesise AA the absorption follows the principles of simple
31
diffusion and is independent of sodium (Spencer et al 1963, Rose 1988). This is in contrast to non-AA-producing species like humans and guinea pigs where the absorption mechanism of AA occurs by a pH-dependent saturable active transport and relies on the presence of a carrier and sodium ions. Although the absorption of orally administered AA in horses (LoÈscher et al 1984), swine (Cromwell et al 1970) and cows (Knight et al 1941) seems to be poor, this does not imply that the same would be the case in all producing species. Ascorbic acid, being a very polar substance, would not be a probable candidate for passing into the systemic circulation by passive diffusion at the relatively high rated seen in this study. Thus, the data might suggest AA to be absorbed by a carrier-mediated mechanism even in dogs. Comparing AUCs calculated after low (15 mg kgÿ1) and high (50 mg kgÿ1) doses of AA, demonstrated a lack of linear dose proportionality. This was also independent of dosage form of supplement. Such behaviour might in general be explained by two factors: reduced systemic bioavailability or non-linear kinetics e.g. an increase in clearance as the concentration increases. For AA, both aspects might be involved. Generally, an increase in renal clearance will reduce AUC, also evident through an increased kel. However, Cmax should not be affected to any significant extent, thus a proportionate increase in Cmax would be expected when increasing the dose. A reduction in the bioavailable fraction will reduce AUC and Cmax, but kel will remain unaffected, assuming that Vd is constant. When analysing at the declining parts of the concentration±time curves, no clear indications of a changes in kel at high doses were found. Furthermore, the relative increase in AUC at high dosage (factor 173) is in fact very similar to the increase in Cmax (factor 183). Consequently, the disproportionate increase in AUC seen at high dosage in the present work is most probably caused by decreased bioavailable fraction of AA. The concentration±time curves often demonstrated a clear and in some instances quite substantial second peak, between 6 and 8 hours after administration, lasting for 2±3 hours. The event was seen in 11 of the 28 experiments, and as already pointed out, made it difficult to estimate kel even by non-compartmental analysis. This phenomenon has not previously been reported, however, concentration±time curves appearing in previous papers on kinetics of AA in dogs reveal that a similar second peak is indeed present, albeit not commented on, in the reports of Teare et al (1979) and Schulze et al (1992). This second peak is not believed to be an artifact caused by variation in the endogenous AA levels, as the blank series in this study did not reveal any such inconsistencies with time. The secondary peak might result from some secondary `absorption' process into the central compartment, thus superimposing the ongoing decline caused by elimination processes to produce a more linear plasma concentration profile. This phenomenon would then mask an underlying 1. order elimination process. Enterohepatic recycling of AA has in fact been demonstrated and studies in rats have indicated biliar excretion of AA after about 8 hours
32
S. Wang, G. E. Berge, N. O. Hoem, R. B. Sund
(Hornig et al 1973), which correlate well with the observations in this study. Ester-C1 is a vitamin C product consisting of 73 per cent AA, 8 per cent DHA, 10 per cent calcium as well as minor amounts of smaller metabolites of AA, mainly threonic acid and xylonic acid. The product has been claimed to be effective in the treatment of dogs with diseases in bone and cartilage (Berge 1990) and has been shown to have a higher absorption rate and reduced excretion rates in rats when compared to ordinary AA (Bush and Verlangieri 1987). In studies on human T-lymphoma cells (Fay and Verlangieri 1991) and mouse 3T3 fibroblasts (Fay et al 1994) one of the main degradation products in Ester-C1, threonate (1±2 per cent in Ester-C1), stimulated AA uptake into the cells, albeit at non-physiologic concentrations. The results presented in this study do not confirm any increased bioavailability for Ester-C1 as described in humans and rats compared to ordinary AA. Cross-over experiments were performed in an attempt to reduce the inter-individual differences in AA handling. When evaluating the differences between the two doses in a paired Wilcoxon signed-rank test a statistically significant difference was apparent for ka and tmax, but strangely also for lag time. On closer analysis however, given the fact that any such differences were not seen when testing in a non-paired manner, the small sample size (n 6) and the fact that another method for estimating ka [based on evaluating the expression tmax ln(ka/ke)/(ka ÿ ke)] did not show the same difference, there is hesitation to conclude that any such differences did actually register between the two preparations. Furthermore a paired t-test failed to show any such statistical significance. Also when comparing the individual pairs of data very similar values were found. The apparent clinical usefulness of supplementing dogs with AA, might allude a sub-optimal endogenous production capacity of this compound in dogs. In this context an interesting observation is reported in a study on the endogenous production of AA in different species (Chatterjee et al 1975), where dogs were found to have low production capacity compared to other AA producing species. The results from the present study, indicating an intestinal transport capacity for AA in dogs, might support a re-evaluation of the dogs as a selfprovided producer of this vitamin.
ACKNOWLEDGEMENTS The authors would like to thank The Norwegian Military Dog Training Centre and The Dog Training Centre of the Norwegian Association for Blinded for their interest and co-operation, and Mrs Lita Schram, School of Pharmacy, University of Oslo, for her excellent technical assistance. REFERENCES BELFIELD, W. O. (1976) Chronic subclinical scurvy and canine hip dysplasia. Vet Med Small Anim Clin 71, 1399±1403 BERGE, G. E. (1990) Polyascorbate (C-Flex), an alternative by ailments in the support- and movement structure in dogs. (Norwegian) Norwegian Veterinary Journal 102, 579±581 BUSH, M. J. & VERLANGIERI, A. J. (1987) An acute study on the relative gastrointestinal absorption of a novel form of calcium ascorbate. Res Commun Chem Path Pharmac 57, 137±140 CHATTERJEE, I. B., MAJUMDER, B. K., NANDI, B. K. & SUBRAMANIAN, N. (1975) Synthesis and some major functions of vitamin C in animals. Annals of The New York Academy of Sciences 258, 24±47 CROMWELL, G. L., HAYS, V. W. & OVERFIELD, J. R. (1970) Effect of dietary ascorbic acid on performance and plasma cholesterol levels in growing swine. J Anim Sci 31, 63±66. FAY, M. J. & VERLANGIERI, A. J. (1991) Stimulatory action of calcium Lthreonate on ascorbic acid uptake by a human T-lymphoma cell line. Life Sciences 49, 1377±1381 FAY, M. J. BUSH, M. J. & VERLANGIERI, A. J. (1994) Effect of aldonic acid on the uptake of ascorbic acid by 3T3 mouse fibroblasts and human T lymphoma cells. Gen Pharmac 25, 1465±1469 GRéNDALEN, J. (1976) Metaphyseal osteopathy (hypertrophic osteodystrophy) in growing dogs. A clinical study. Journal of Small Animal Practice 17, 721±735 HORNIG, D., GALLO-TORRES, H. Eo & Weiser, H. (1973) A biliary metabolite of ascorbic acid in normal and hypophysectomized rats. Biochim Biophys Acta 320, 549±556. È LOSCHER, W., JAESCHKE, G. & KELLER, H. (1984) Pharamacokinetics of ascorbic acid in horses. Equine Veterinary Journal 16(1), 59±65 KNIGHT, C. A., DUTCHER, R. A., GUERRANT, N. B. & BECHTEL, S. (1941) Destruction of ascorbic acid in the rumen of dairy cows. J Dairy Sci 24, 567±577 ROSE, R. C. (1988) Transport of water-soluble vitamins. Biochem Biophys Acta 947, 335±366. SCHULZE, J., BROZ, J. & LUDWING, B. (1992) Efficacy of L-ascorbate-2polyphosphate as a source of ascorbic acid on dogs. International Journal of Vitamin and Nutritional Research 63, 63±64 SPENCER, R. P., PURDY, S., HOELDTKE, R., BOW, T. M. & MARKULIS, M. A. (1963) Studies on intestinal absorption of L-ascorbic acid-1-C14. Gastroenterology 44, 768±773. TEARE, J. A., KROOK, L., KALLFELZ, F. A. & HINTZ, H. F. (1979) Ascorbic acid deficiency and hypertrophic osteodystrophy in the dog: A rebuttal. Cornell Vet 69, 384±401 WANG, S., SCHRAM, I. M. & SUND, R. B. (1995) Determination of plasma ascorbic acid by HPLC: Method and stability studies. European Journal of Pharmaceutical Sciences 3, 231±239
Accepted 15 May 2001
Pharmacokinetics in dogs after oral administration of two different forms of ascorbic acid S. WANGz*, G. E. BERGEx, N. O. HOEMz, R. B. SUNDzy z
Department of Pharmacology, School of Pharmacy, University of Oslo, x Nordberg Small Animal Clinic, Oslo SUMMARY
The dog is able to synthesise ascorbic acid (AA), but is frequently given the vitamin in an attempt to improve health and performance. The pharmacokinetics of AA in this species, however, are not well studied. Using a selective analytical method and careful stability control, the pharmacokinetics of orally given AA was studied in 20 dogs, at two dosage levels (15 and 50 mg kgÿ1) and with two forms of supplement [crystalline AA and the vitamin C product Ester-C1 (Inter-Cal Corp., Prescott, AZ, USA)]. After oral administration, a rapid increase was found in the plasma level of AA, indicating a possible intestinal active transport mechanism in this species. The obtained Cmax and AUC values were found to increase in a non-linear fashion when the dose of AA was increased. The pharmacokinetic modeling of the elimination of AA was made difficult by a pronounced secondary peak appearing after about 9 hours. The comparison of crystalline AA and Ester-C1 did not indicate any significant differences in pharmacokinetic parameters between the two preparations of the vitamin. # 2001 Harcourt Publishers Ltd
ASCORBIC acid (AA) has been administered both orally and i.v. to dogs to improve health and performance, despite the fact that this species is able to synthesise the vitamin. In dogs, AA supplementation has been given to reduce the incidence of hip dysplasia (HD). Belfield (1976) studied puppies from parents with HD, and found that giving AA to the pregnant mothers and their puppies resulted in dogs free HD. Ascorbic acid has also been tried in the treatment of hypertrophic osteodystrophia (HOD) in dogs, with negative results (Grùndalen 1976, Teare et al 1979). Few publications have focused on monitoring plasma concentrations in dogs after AA administration, to evaluate the effect of AA supplementation in pharmacokinetic terms. Teare and co-workers (1979) studied plasma AA after oral or i.v. supplementation, as a part of a study on HOD in dogs. Schulze and co-workers (1992) compared the plasma concentration of AA after oral administration of L-ascorbate-2-polyphosphate or AA. However, the numbers of animals studied were small, and a general pharamacokinetic evaluation was not the main focus in any of those papers. Concerning the large number of methods for determination of AA and the variability seen in existing analytical methods, special attention should be paid to ensure reliable measurements, particularly regarding analytical selectivity. In addition, stability control must be ensured, as AA is easily degraded in samples and in
other aqueous solutions. This is particularly important for ambulatory reserch, as will often be the case in studies on domestic animals. A previously described method was used for AA determination based on HPLC and electrochemical detection, as well as stability data in samples from dogs (Wang et al 1995), and the method as well as the sampling and preparation procedures used in the present paper are based on this publication. Several products and derivatives of AA have been used for supplementation in dogs. L-ascorbate-2polyphosphate has been tried as a source of AA (Schulze et al 1992) to overcome the stability problem for ordinary AA in the diet. Ester-C1a (Inter-Cal Corp., Prescott, AZ, USA), a vitamin C product consisting of ascorbic acid, dehydroascorbic acid and calcium as well as minor amounts of smaller metabolites of AA, is another pharmaceutical formulation of the vitamin that has been given to dogs (Berge 1990). This formulation of vitamin C has been claimed to show a higher rate and degree of absorption compared to ordinary AA when given to rats (Bush and Verlangieri 1987), but no study has been performed to compare the two products in dogs. The main purpose of this work was to evaluate the pharmacokinetics of AA given orally to healthy dogs, using a rapid and selective analytical method for AA determination and careful stability control. Secondarily, we wanted to compare the kinetics of ordinary AA and Ester-C1a.
y
Deceased. *Corresponding author: Tùnsberg Hospital Pharmacy, Box 2168 Postterminalen, N-3103 Tùnsberg, Norway; Tel: 47 33 34 31 04; Fax: 47 33 31 34 67; E-mail: [email protected]
0034-5288/01/010027 06 $35.00/0
a
The product is also named C-flex1 when marketed as a veterinary supplement (powder). # 2001 Harcourt Publishers Ltd
28
S. Wang, G. E. Berge, N. O. Hoem, R. B. Sund
MATERIALS AND METHODS Animals The study included a total of 20 animals (eight German Shepherds, 10 Labrador Retrievers and two Riesenschnaussers) aged 1±10 years. All animals were healthy and had not been given supplementary vitamins prior to the study. Their regular feed was not declared to contain AA. The experiments were performed after an overnight fast, and no feed was given during the study period. Water was available ad libitum. All animals were placed under similar conditions and had about the same (low) activity level. Study design A total of 32 experiments were included in the study (Table 1). The kinetics of high-dose AA (50 mg kgÿ1) were investigated in six dogs (Group 1; two Riesenschnaussers, four Labrador Retrievers) given a single oral dose of crystalline AA (cAA) (Sigma Chemical Company, St. Louis, MO, USA). Six additional experiments (Group 2) were performed with a corresponding dose of Ester-C1 (Inter-Cal Corp., Prescott, AZ, USA). All high-dose experiments were cross-over designed to facilitate a better comparison of the pharmacokinetics of the two formulations of AA. The kinetics of low-dose cAA (15 mg kgÿ1) were investigated in eight dogs (Group 3; two Labrador Retrievers, six German Shepherds). In addition, eight experiments (Group 4; two Labrador Retrievers, six German Shepherds) were performed with a corresponding dose of Ester-C1. Eight of the 16 experiments at low dose were cross-over designed. The supplements were administered in gelatine capsules together with ca. 20 ml water. A wash-out period of 3 weeks was allowed between the series making the cross-over experiments. To study the diurnal variation in endogenous AA levels, plasma AA was monitored in two dogs (Labrador Retrievers) not given any supplement. The monitoring was repeated after 3 weeks. In all instances, blood samples for plasma AA determination were taken before supplement administration (if administered), at intervals during 12 hours and after about 24 hours. Blood sampling and ascorbic acid analysis Sampling and analytical procedure as well as stability data are described in detail elsewhere (Wang et al TABLE 1: Number of experiments included in the study (grouping based on form of supplement and dosage) Dosage in AA equivalents
Form of supplement CAA
Ester-C1
None
Sum
8 Group 4 6 Group 2 ö
ö
16
ö
12
ö
8 Group 3 6 Group1 ö
4
4
Total
14
14
4
32
15 mg kg
ÿ1
50 mg kg ÿ1
1995). At the start of each experiment, all dogs were fitted with a catheter in Vena cephalica. Blood samples (5 ml) were collected into Vacutainer1 tubes containing heparin or EDTA. After centrifugation at least two 500 ml, aliquots of plasma were either immediately frozen (heparinised samples) in portable liquid nitrogen containers, or stabilised and deproteinised by perchloric acid with subsequent mixing, centrifugation and freezing in liquid nitrogen (samples treated with EDTA). The maximum time elapsed from sampling until freezing was 20 minutes. At the time of analysis, samples were thawed one by one and immediately prepared and analysed by a method based on high-pressure liquid chromatography and electrochemical detection (Wang et al 1995). Pharmacokinetic analysis Individual AA concentration±time curves were analysed by mono-exponential non-linear regression fits on the declining part of the curve and by non-compartmental pharmacokinetic methods. All dogs had substantial endogenous AA levels, and all measured concentrations were adjusted by subtraction of the endogenous start level. The elimination rate constant (kel) was estimated by a simple mono-exponential analysis. Non-linear regression fitting to a simple i.v. bolus injection model was utilised to extract an estimate of kel 1 from the declining part of the plasma concentration curves. Non-compartmental analysis was performed by calculating the area under the concentration±time curve (AUC) and the area under the first moment concentration-time curve (AUMC) by the liner trapezoid method. Because the curves reached close to zero level within the recorded period, most AUCs were not corrected for area by estimation of the final integral. Mean residence time (MRT), a measure for the mean time any molecule of the exogenous substance resides in the body after onset of absorption, was calculated as AUMC/AUC. For a bolus dose into a rapidly equilibrating (no profound distribution phase) system that can be described by simple first order kinetics the following relationship exists between MRT and the elimination (kel) and absorption rate (ka) constants: MRT
1 1 tlag kel ka
1
where tlag is the time delay of absorption, defined as the time before the plasma concentration increased by 5% over base level. After substituting these estimates of tlag and estimates of kel obtained from the mono exponential fits into the above expression, this relationship was used to estimate ka. Non-linear regression fitting was performed with the SAAM II pharmacokinetic analysis software, (SAAM Institute, University of Washington, USA). All data were fitted using relative weighting assigning a fractional SD of 010 to all data. The fitting procedure was evaluated by residual plots and comparison of total objective function value and the Akaike criterion.
29
Pharmacokinetics in dogs
Statistics Statistical analysis was performed by the S-PLUS 2000 software, MathSoft Inc. (USA). All statistical comparison between the groups were performed by nonparametric statistical methods, Mann±Whitney for non paired data and Wilcoxon signed-rank test for data from paired experimetns. A P-level less than 005 was considered statistically significant. RESULTS Initial plasma AA concentration (Cp0) was 402 68 mmol lÿ1 (n 32). In the experiments where supplement was given, plasma levels before and 24 hours after administration were 398 68 and 404 64 mmol lÿ1, respectively (n 28). The diurnal
FIG 1: The diurnal plasma regression, ^ ^ ^.
AA
plasma AA variation in animals not supplemented is shown in Figure 1. The fluctuation around the mean plasma AA level (412 mmol lÿ1) was 13% (n 4). The parameters obtained from the pharmacokinetic calculations are given in Table 2. All concentrations are relative to the individual pre-administration plasma AA concentration (Cp0), and calculated as the increase in plasma AA levels. Figures 2 and 3 show the relative plasma concentration±time curves for the dogs after p.o. administration of AA supplements (as cAA or Ester-C1) at the dosages of 15 and 50 mg kgÿ1, respectively. There was a notable variation in the plasma AA profiles after supplement administration. However, some structural features were prevalent: Tlag varied between 0 and 12 hours, and an apparent delay in absorption was observed in 12 of the 28
variation in animals not supplemented with vitamin C. Dog 1, ^*^; dog 2, ^&^; dog 3, ^^; dog 4, ^~±. Linear
TABLE 2: The pharmacokinetic parameters (mean SD) obtained after administration of crystalline AA (Group 1: 50 mg kg ÿ 1, Group 3: 15 mg kg ÿ 1) or Ester-C 1 (Group 2: 50 mg kg ÿ 1, Group 4: 15 mg kg ÿ 1) High dose (50 mg kg ÿ1)
Cmax (mmol l ÿ1)* Rel T max (hours) ÿ1 AUC (mmol hours l )* MRT (hours) kel (hours ÿ1) ka (hoursÿ1)
Low dose (15 mg kg ÿ1)
Group1
Group 2
Groups1 2
Group 3
Group 4
Groups 3 4
463(188) 36(155) 350(166) 64(18) 018(004) 077(028)
415(88) 28(184) 310(103) 62(17) 016(006) 117(069)
439(142) 32(05) 329(133) 63(17) 017(005) 097(055)
251(64) 271(111) 215(98) 73(27) 022(020) 099(029)
228(122) 301(137) 167(96) 69(22) 027(022) 092(030)
239(95) 28(12) 191(965) 71(24) 025(020) 096(029)
Cmax : Maximal plasma AA concentration after administration (pre-administration levels are subtracted). Rel T max : T max minus the estimated lag time. AUC: Area under the plasma ^concentration time curves (pre-administration levels are subtracted). MRT: Mean residence time, calculated as AUC/AUMC (area under the first moment concentration ^ time curve). kel: First order elimination rate constant, based on SAAM modelling. ka: First order obsorption constant based on estimate of MRT, lag time and kel. *For Cmax and AUC, the values obtained are based on the increase in plasma levels above the basal, pre-administration levels.
[AA]-time profiles 80.0
(a)
80.0
(b) G3-Exp13 G3-Exp14 G3-Exp17 G3-Exp19 G3-Exp21 G3-Exp23 G4-Exp15 G4-Exp16 G4-Exp18 G4-Exp20 G4-Exp22 G4-Exp24 G3-Exp25 G3-Exp26 G4-Exp27 G4-Exp28
Group3 Group4 70.0
60.0
[Ascorbic acid] (uM)
[Ascorbic acid] (uM)
50.0
40.0
30.0
70.0
60.0
50.0
40.0
30.0
20.0
20.0
10.0
10.0
0.0
0.0 0
5
10
15
0 5 20 25 Time after administration (hours)
10
15
20
25
FIG 2: The relative plasma concentration ^time curves after p.o. administration of AA 15 mg kg ÿ1 given as crystalline AA (Group 3) or Ester-C1 (Group 4). (a) Mean curves ( SD) for the two forms of supplement. (b) Individual curves. All plasma concentrations represent the increase in plasma levels above the basal, pre-administration levels.
[AA]-time profiles 80.0
(a)
80.0
(b) G1-Exp1 G1-Exp3 G1-Exp5 G1-Exp7 G1-Exp9 G1-Exp11 G2-Exp2 G2-Exp4 G2-Exp6 G2-Exp8 G2-Exp10 G2-Exp12
Group1 Group2 70.0
60.0
[Ascorbic acid] (uM)
[Ascorbic acid] (uM)
50.0
40.0
30.0
70.0
60.0
50.0
40.0
30.0
20.0
20.0
10.0
10.0
0.0
0.0 0
5
10
15
0 5 20 25 Time after administration (hours)
10
15
20
25
FIG 3: The relative plasma concentration ^ time curves after p.o. administration of AA 50 mg kg ÿ1 given as crystalline AA (Group 1) or Ester-C1 (Group 2). (a) Mean curves ( SD) for the two forms of supplement. (b) Individual curves. All plasma concentrations represent the increase in plasma levels above the basal, pre-administration levels.
Pharmacokinetics in dogs
experiments. The presence or degree of delay could be correlated to neither dose, form of supplement nor breed. In most animals the subsequent absorption phase was quite rapid, with a maximal concentration reached about 3 hours (Rel Tmax 30 14 hours) after estimated onset of absorption with a first order absorption constant ka estimated to 096 0409 hÿ1. In about one third of the experiments a clearly visible second peak was observed, appearing 6±10 hours administration, with a duration of 2±3 hours. Fitting the declining phase to a simple first order curve showed a rather high degree of model misfit, a phenomenon suggesting that not only the dogs with clearly visible second peaks showed this behaviour, but also the other dogs, but at a lower degree. The phenomenon was more prevalent in the high dose (50 mg kgÿ1) experiments in the Labrador Retrievers. No differences in kel values were seen between the different groups. All experiments taken together, the elimination rate constant kel was estimated to be 0214 016 hÿ1 (range 0053 to 0667). Pooling the data within the high-(Group 1 2) and low-(Group 3 4) dosage groups, the mean area under the plasma±AUC was increased from 191 mmol hÿ1 lÿ1 to 330 mmol hÿ1 lÿ1 when increasing the dose from 15 mg kgÿ1 to 50 mg kgÿ1. Such a lack of dose proportionality was also seen for the Cmax for the two dosages (mean 239 versus 439 mmol lÿ1). When comparing the two forms of vitamin C supplement (Group 1 versus Group 2, and Group 3 versus Group 4), no statistical significant differences were seen in the pharmacokinetic parameters, except for ka and tlag when perfroming a paired test between groups 1 and 2, when a borderline statistical significance was recorded. DISCUSSION Endogenous AA levels were recorded in all dogs, ranging from 265 to 499 mmol lÿ1 (402 68). These are in accordance with the values reported previously (Wang et al 1995). Basal plasma levels were recorded throughout 26 hours in four dogs, and it was found that the diurnal variation in plasma AA level was small. This finding was also confirmed by the small differences in AA levels found before (402 68) and at the end (404 64 mmol lÿ1) of each experiment. The pharmacokinetic parameters most sensitive to variations in basal plasma AA levels are Cmax and AUC. To minimise this effect when studying AA pharmacokinetics in the individual dogs, the baseline value was subtracted, thus calculating a `net' peak above the endogenous level. All Cmax and AUC values are therefore expressed as the increase in plasma AA levels above the endogenous level. The rapid increase in plasma AA after oral administration is in agreement with previous reports in dogs (Teare et al 1979; Schulze et al 1992). It is generally believed that for animals with the ability to synthesise AA the absorption follows the principles of simple
31
diffusion and is independent of sodium (Spencer et al 1963, Rose 1988). This is in contrast to non-AA-producing species like humans and guinea pigs where the absorption mechanism of AA occurs by a pH-dependent saturable active transport and relies on the presence of a carrier and sodium ions. Although the absorption of orally administered AA in horses (LoÈscher et al 1984), swine (Cromwell et al 1970) and cows (Knight et al 1941) seems to be poor, this does not imply that the same would be the case in all producing species. Ascorbic acid, being a very polar substance, would not be a probable candidate for passing into the systemic circulation by passive diffusion at the relatively high rated seen in this study. Thus, the data might suggest AA to be absorbed by a carrier-mediated mechanism even in dogs. Comparing AUCs calculated after low (15 mg kgÿ1) and high (50 mg kgÿ1) doses of AA, demonstrated a lack of linear dose proportionality. This was also independent of dosage form of supplement. Such behaviour might in general be explained by two factors: reduced systemic bioavailability or non-linear kinetics e.g. an increase in clearance as the concentration increases. For AA, both aspects might be involved. Generally, an increase in renal clearance will reduce AUC, also evident through an increased kel. However, Cmax should not be affected to any significant extent, thus a proportionate increase in Cmax would be expected when increasing the dose. A reduction in the bioavailable fraction will reduce AUC and Cmax, but kel will remain unaffected, assuming that Vd is constant. When analysing at the declining parts of the concentration±time curves, no clear indications of a changes in kel at high doses were found. Furthermore, the relative increase in AUC at high dosage (factor 173) is in fact very similar to the increase in Cmax (factor 183). Consequently, the disproportionate increase in AUC seen at high dosage in the present work is most probably caused by decreased bioavailable fraction of AA. The concentration±time curves often demonstrated a clear and in some instances quite substantial second peak, between 6 and 8 hours after administration, lasting for 2±3 hours. The event was seen in 11 of the 28 experiments, and as already pointed out, made it difficult to estimate kel even by non-compartmental analysis. This phenomenon has not previously been reported, however, concentration±time curves appearing in previous papers on kinetics of AA in dogs reveal that a similar second peak is indeed present, albeit not commented on, in the reports of Teare et al (1979) and Schulze et al (1992). This second peak is not believed to be an artifact caused by variation in the endogenous AA levels, as the blank series in this study did not reveal any such inconsistencies with time. The secondary peak might result from some secondary `absorption' process into the central compartment, thus superimposing the ongoing decline caused by elimination processes to produce a more linear plasma concentration profile. This phenomenon would then mask an underlying 1. order elimination process. Enterohepatic recycling of AA has in fact been demonstrated and studies in rats have indicated biliar excretion of AA after about 8 hours
32
S. Wang, G. E. Berge, N. O. Hoem, R. B. Sund
(Hornig et al 1973), which correlate well with the observations in this study. Ester-C1 is a vitamin C product consisting of 73 per cent AA, 8 per cent DHA, 10 per cent calcium as well as minor amounts of smaller metabolites of AA, mainly threonic acid and xylonic acid. The product has been claimed to be effective in the treatment of dogs with diseases in bone and cartilage (Berge 1990) and has been shown to have a higher absorption rate and reduced excretion rates in rats when compared to ordinary AA (Bush and Verlangieri 1987). In studies on human T-lymphoma cells (Fay and Verlangieri 1991) and mouse 3T3 fibroblasts (Fay et al 1994) one of the main degradation products in Ester-C1, threonate (1±2 per cent in Ester-C1), stimulated AA uptake into the cells, albeit at non-physiologic concentrations. The results presented in this study do not confirm any increased bioavailability for Ester-C1 as described in humans and rats compared to ordinary AA. Cross-over experiments were performed in an attempt to reduce the inter-individual differences in AA handling. When evaluating the differences between the two doses in a paired Wilcoxon signed-rank test a statistically significant difference was apparent for ka and tmax, but strangely also for lag time. On closer analysis however, given the fact that any such differences were not seen when testing in a non-paired manner, the small sample size (n 6) and the fact that another method for estimating ka [based on evaluating the expression tmax ln(ka/ke)/(ka ÿ ke)] did not show the same difference, there is hesitation to conclude that any such differences did actually register between the two preparations. Furthermore a paired t-test failed to show any such statistical significance. Also when comparing the individual pairs of data very similar values were found. The apparent clinical usefulness of supplementing dogs with AA, might allude a sub-optimal endogenous production capacity of this compound in dogs. In this context an interesting observation is reported in a study on the endogenous production of AA in different species (Chatterjee et al 1975), where dogs were found to have low production capacity compared to other AA producing species. The results from the present study, indicating an intestinal transport capacity for AA in dogs, might support a re-evaluation of the dogs as a selfprovided producer of this vitamin.
ACKNOWLEDGEMENTS The authors would like to thank The Norwegian Military Dog Training Centre and The Dog Training Centre of the Norwegian Association for Blinded for their interest and co-operation, and Mrs Lita Schram, School of Pharmacy, University of Oslo, for her excellent technical assistance. REFERENCES BELFIELD, W. O. (1976) Chronic subclinical scurvy and canine hip dysplasia. Vet Med Small Anim Clin 71, 1399±1403 BERGE, G. E. (1990) Polyascorbate (C-Flex), an alternative by ailments in the support- and movement structure in dogs. (Norwegian) Norwegian Veterinary Journal 102, 579±581 BUSH, M. J. & VERLANGIERI, A. J. (1987) An acute study on the relative gastrointestinal absorption of a novel form of calcium ascorbate. Res Commun Chem Path Pharmac 57, 137±140 CHATTERJEE, I. B., MAJUMDER, B. K., NANDI, B. K. & SUBRAMANIAN, N. (1975) Synthesis and some major functions of vitamin C in animals. Annals of The New York Academy of Sciences 258, 24±47 CROMWELL, G. L., HAYS, V. W. & OVERFIELD, J. R. (1970) Effect of dietary ascorbic acid on performance and plasma cholesterol levels in growing swine. J Anim Sci 31, 63±66. FAY, M. J. & VERLANGIERI, A. J. (1991) Stimulatory action of calcium Lthreonate on ascorbic acid uptake by a human T-lymphoma cell line. Life Sciences 49, 1377±1381 FAY, M. J. BUSH, M. J. & VERLANGIERI, A. J. (1994) Effect of aldonic acid on the uptake of ascorbic acid by 3T3 mouse fibroblasts and human T lymphoma cells. Gen Pharmac 25, 1465±1469 GRéNDALEN, J. (1976) Metaphyseal osteopathy (hypertrophic osteodystrophy) in growing dogs. A clinical study. Journal of Small Animal Practice 17, 721±735 HORNIG, D., GALLO-TORRES, H. Eo & Weiser, H. (1973) A biliary metabolite of ascorbic acid in normal and hypophysectomized rats. Biochim Biophys Acta 320, 549±556. È LOSCHER, W., JAESCHKE, G. & KELLER, H. (1984) Pharamacokinetics of ascorbic acid in horses. Equine Veterinary Journal 16(1), 59±65 KNIGHT, C. A., DUTCHER, R. A., GUERRANT, N. B. & BECHTEL, S. (1941) Destruction of ascorbic acid in the rumen of dairy cows. J Dairy Sci 24, 567±577 ROSE, R. C. (1988) Transport of water-soluble vitamins. Biochem Biophys Acta 947, 335±366. SCHULZE, J., BROZ, J. & LUDWING, B. (1992) Efficacy of L-ascorbate-2polyphosphate as a source of ascorbic acid on dogs. International Journal of Vitamin and Nutritional Research 63, 63±64 SPENCER, R. P., PURDY, S., HOELDTKE, R., BOW, T. M. & MARKULIS, M. A. (1963) Studies on intestinal absorption of L-ascorbic acid-1-C14. Gastroenterology 44, 768±773. TEARE, J. A., KROOK, L., KALLFELZ, F. A. & HINTZ, H. F. (1979) Ascorbic acid deficiency and hypertrophic osteodystrophy in the dog: A rebuttal. Cornell Vet 69, 384±401 WANG, S., SCHRAM, I. M. & SUND, R. B. (1995) Determination of plasma ascorbic acid by HPLC: Method and stability studies. European Journal of Pharmaceutical Sciences 3, 231±239
Accepted 15 May 2001