犬におけるDigoxinおよびDigitoxinの薬物動態と実験的胆汁うっ滞の影響に関する研究 Pharmacokinetics of digoxin and digitoxin in the dog and the influence of experimental cholestasis on the pharmacokinetics
Pharmacokinetics of digoxin and digitoxin in the dog and the influence of experimental cholestasis on the pharmacokinetics
1785年，William Witheringはヒト医学臨床において，また1841年Delabere Blaineは獣医臨床で，ジギタリスの治療効果をはじめて紹介した。以来約200年を経過した今日においても，ジギタリスはうっ血性心不全の治療や心房細動の心室拍数コントロール上，重要な地位を占めている。しかしジギタリスは有効治療域が狭く，そのため副作用の発現に絶えず注意を払わなければならない薬物の1つである。また，獣医臨床では，ジギタリス剤の薬物動態や治療効果に関する実験的ならびに臨床的検討が主としてdigoxinで行われており，digitoxinとの比較研究が少ない。近年，臨床薬理学の進歩に伴い，家畜におけるジギタリスの薬物動態に関する知見も蓄積されつつあるが，不明な点も少なくない。とくに，薬物代謝の主要部位である肝の関与に関する検討が乏しい。本研究は，犬の実験的胆汁うっ滞時におけるdigoxinとdigitoxinの薬物動態から，両剤の代謝・排泄に対する肝の関与について検討したものである。成績のの概略を以下に述べる。 1．Digoxin投与による薬物動態の検討 健常雑種犬18頭を対照（C）群，総胆管結紮（L）群，phenobarbital前処置総胆管結紮（P）群の3群に分け，digoxin25μg／kgを1回静脈内投与した。P群には，肝ミクロゾームの薬物代謝酵素系を活性化することが知られているphenobarbitalを5mg／kg／dayの割合で1日1回，2週間連続投与し，その後に総胆管結紮を施している。C群は開腹して総胆管を確認しただけで腹壁を閉鎖したものである。いずれの群も術後5～6時間後にdigoxinを投与し，その後0．5，1，2，3，6，8，12，18，24，30，36および48時間目にヘパリン採血し，血漿digoxin濃度をradioimmunoassay（RIA）法で測定した。 各群の血漿digoxin濃度推移は，指数関数的に急速に減少する分布（α）相と，静脈内投与6～8時間以降の排泄（β）相とに分けられ，2－compartment open modelに一致し，次のような2つの指数関数の和として表わされた。 Ct＝A・e＾－α・t ＋ B・e＾－β・t Ct：静注t時間後の血漿中薬物濃度 A ：分布相におけるt＝0での外挿血漿中薬物濃度（外挿Y軸交点） B ：排泄相におけるt＝0での外挿血漿中薬物濃度（外挿Y軸交点） α ：分布相の消失速度定数 β ：排泄相の消失速度定数この式から求められた血漿digoxin理論濃度は，RIA法による実測値とよく一致した。 C群におけるdigoxinの薬物動態をみると，t1／2αは0．928±0．094（±SE）時間，またt1／2βは19．56±2．16時間であった。外挿法により求めた見かけの分布容量（Vd）は6．608±0．769l／kg，血漿濃度曲線下面積法による見かけの分布容量（Vdarea）は9．424±1．571l／kgであった。総体クリアランス（TBCL）は5．65±0．96ml／min／kgとなった。 L群の血漿digoxin濃度は，どの測定時においても他2群の値との間で有意差が認められなかった。薬物動態パラメータをC群と比較すると，t1／2αには有意差がなく，排泄（β）相における消失速度定数βは低値（p≦0．05）を示した。t1／2βはC群のそれ（19．56±2．16時間）との間に有意差を認めなかったけれども，25．18±1．53時間と延長した。中央（血液）コンパートメントから末梢（組織）コンパートメントへの移行速度定数K12は大きく（p≦0．05），Vdも増大した。TBLCはC群のそれより低かったが，有意差はなかった。 P群ではL群と同様な総胆管結紮を行なっても，t1／2βの延長およびVdの増大は認められず，C群とほぼ同じ値となった。L群と比較するとt1／2βは短縮（p≦0．05），Vdは減少（p≦0．01）した。また，TBCLはC群およびL群のそれとの間に有意差を認めなかったが，高値を示す傾向がみられた。 この実験期間中，どの犬の血中尿素窒素量および血清クレアチニン量には変化がなく，群間にも有意差を認めなかった。また，各群の体重kg当り尿量にも有意差がなく，これらの所見から腎機能に変化がなかったものと推定される。 一方，総胆管結紮をしたLおよびP群の血清総ビリルビン値，ALPおよびGPT活性値はC群のそれらに比較して高くなった。実験終了後の剖検および組織学的所見でも，胆汁うっ滞の証拠が明瞭であった。P群ではさらに，肝細胞の著しい肥大が観察され，薬物代謝酵素系の活性化が示唆された。 以上のごとく，実験的に作出した胆汁うっ滞犬に投与されたdigoxinの薬物動態には，組織分布の増大，血漿消失速度の遅延，t1／2βの遅延など，明らかに胆汁うっ滞の影響が認められた。さらにP群で分布容量の増大やt1／2βの延長が認められなかったことは，digoxinの代謝・排泄に肝が関与していることを間接的に示したものと考えられる。しかし，静脈内1回投与後の血漿中濃度は群間で差がなかったことから，この関与は主排泄経路としての腎よりも小さいと推測される。今回の実験は静脈内1回投与であるが，胆汁うっ滞をはじめ，他のタイプの肝障害が存在する場合の維持療法では，血中および組織内濃度が有意に上昇すると思われる。従って本剤の使用にあたっては，腎機能とともに肝機能に十分な配慮が必要と結論された。 2．Digitoxin投与による薬物動態の検討 健常雑種成犬15頭をC，LおよびP群に分け，digitoxin20μg／kgを1回静脈内投与した。その後の実験方法はdigoxinの場合と同じである。さらに別な雑種成犬10頭を，前記3群と胆管痩を設けたF群の計4群に分け，^<3>H標識digitoxin50μci／14kgおよび非標識digitoxin20μg／kgを静脈内投与した。投与後1，3，6，12および24時間目に採血した。血漿，尿および胆汁中の放射活性をジクロロメタン(CH_2 CL_2)溶性および非溶性分画にわけて計測した。 各群の血漿digitoxin濃度推移は，digoxinと同様2つの指数関数の和として表わされ，その理論濃度は実測値とよく一致した。C群のt1／2αは0．909±0．465時間，t1／2βは7．57±1．57時間であった。Vdは0．807±0．085l／kg，Vdareaは0．905±0．093l／kgであった。TBCLは1．56±0．45ml／min／kgとなった。 L群における各測定時間での血漿digitoxin濃度は，CおよびP群より有意に（p≦0．05）高い値を持続した。t1／2βは24．45±3．40時間に延長し（p≦0．05）VdおよびVdareaも低く（p≦0．01），C群のそれらとの間で有意差を認めた。 P群では総胆管結紮を行っているにもかかわらず，digitoxin血中濃度はC群とほぼ同じレベルで推移した。t1／2β，Vd，VdareaおよびTBCLもC群のそれらとの間に有意差を認めなかった。しかし，L群との比較では，これらパラメータに有意差が存在した。この実験期間における腎機能指標は，digoxinの場合と同様，変化がなかった。一方，血清総ビリルビン量はP群でのみ，ALPとGPT活性はLおよびP群で増高した。剖検および組織学的変化はdigoxin群のそれと基本的には等しかった。 ^<3>H標識digitoxin投与の結果，CH_2 Cl_2溶性分画（digitoxinおよび強心作用を有する脂溶性代謝産物）の放射活性が12および24時間後において，C，P，F群よりL群で高かった。C，L，F群の24時間尿における放射活性は投与量の15～20％で，その95％がCH_2 Cl_2不溶性分画（強心作用を失った水溶性代謝産物）にあった。P群の排泄量は他群よりも多く，主として不溶性分画であった。F群における胆汁の放射活性は投与量の7％で，その約85％は不溶性分画にあった。 以上の実験から，胆汁うっ滞時に投与されたdigitoxinは高い血漿濃度を維持し，t1／2βは正常時の約3倍に延長すること，phenobarbitalによる前処置はこうした変化を抑制し，強心作用を失った水溶性代謝産物の尿中排泄を増加させることなどがわかった。またこのものは胆汁中にも排泄されるが，その割合は小さいことも明らかとなった。ここから，digitoxinもまた，digoxinと同様，その代謝・排泄に肝の関与していることが明らかとなった。しかしt1／2βの延長度合はdigoxinよりもdigitoxinで大きく，また血漿濃度がdigitoxinでのみ有意に高値を持続したことから，胆汁うっ滞の薬物動態に及ぼす影響は，digitoxinのほうが大きいと考えられる。臨床上，胆汁うっ滞をはじめ他のタイプの肝障害がある場合，ジギタリス剤としてはその代謝・排泄に肝の関与が相対的に小さいdigoxinを選ぶべきである。 本研究では以上の如く，犬での実験がきわめて少なかったdigitoxinの薬物動態を明らかにするとともに，これまで推測に過ぎなかったdigoxinおよびdigitoxinの代謝・排泄に及ぼす肝の役割と胆汁うっ滞の影響を実証した。また両薬物の犬における動態を比較し，後者のt1／2βがヒトのそれにくらべて非常に短く，種差の存在することを明らかにした。これらの知見は，獣医学領域におけるジギタリス研究への新たな関心を喚起すると同時に，犬に対する臨床応用の理論的根拠を与えるものと思う。
INTRODUCTION In 1785, William Withering utilized an extract from the foxglove plant, digitalis, in the treatment of dropsy. Withering also published his now famous book. An Account of the Foxglove and Some of its Medical Uses: With Practical Remarks on Dropsy and Other Diseases. Delabere Blaine, in 1841, presented evidence of favorable effects obtained with digitalis in certain dogs treated for ascites, but also observed that digitalis had no benefical effect on other dogs. Today, more than 200 years after the publication of Withering's famous book, the digitalis glycosides, digoxin and digitoxin, are primarily administered for the treatment of congestive heart failure and for the control of the ventricular rate in patients with atrial fibrillation. In recent years, due to the increased recognition of clinical heart diseases in domestic animals, digitalis glycosides have been used with increasing frequency in veterinary medicine. Rational therapy depends upon the accurate and complete knowlege of the drug's behavior within the body. Digitalis glycosides have relatively narrow margins of safety between their effective therapeutic dosage and toxic dosage and these drugs should be administered with caution to avoid possible toxicity. A considerable amount of information has been accumulated pertaining to the absorption, distribution, biotransformation and excretion of digitalis glycosides in animals, accompanying advances in clinical pharmacology. The metabolism and excretion of digoxin are more clearly understood than for the other digitalis glycoside, digitoxin. Better understanding of the pharmacokinetics of digoxin may be attributed to the much more extensive use of this drug in clinical practice and animal experiments than for digitoxin. However, there are still many problems concerning digitalis pharmacokinetics in veterinary medicine. It is known that the liver is an important site involved in the metabolism and excretion of these drugs, but the specific mechanisms and role of the liver pertaining to digitalis pharmacokinetics is not yet completely understood. It is not clear whether liver dysfunction may have a significant influence of digoxin and digitoxin pharmacokinetics in the dog. The present study was undertaken to evaluate the role of the liver in digoxin and digitoxin metabolism and excretion in dogs that have undergone ligation of the common bile duct, producing experimental cholestasis. Furthmore, the clinical effectiveness of digoxin and digitoxin were also evaluated and compared. MATERIALS and METHODSA. Digoxin Evaluation For this experiment, eighteen apparently healthy dogs were divided into three groups. After the dogs had been anesthetized with pentobarbital sodium, a midline anterior abdominal incision was made. The common bile duct was doubly ligated in seven dogs (L group). A group of three dogs was given phenobarbital for two weeks followed by surgical ligation of the common bile duct (P group). A control group, consisting of eight dogs received abdominal incisions, that were later closed leaving the common bile duct intact (C group). Digoxin was given by a single intravenous, to exclude the influence of variances in the absorption of the drug. All of the experimental dogs received 25 μg/kg of digoxin, five to six after the surgical operations. Venous blood samples were drawn in heparinized syringes from each animal prior to the operation and at 0.5, 1, 2, 3, 6, 8, 12, 18, 24, 30, 36 and 48 hours post intravenous injection of digoxin. The plasma was separeted and the samples were frozen for later digoxin analysis. The plasma digoxin concentration was determined in duplicate using a commercially available radioimmunoassay (RIA) kit.B. Digitoxin Evaluation For this experiment, fifteen healthy dogs were divided into three groups, as was the procedure digoxin evaluation: a control group (C group) of three dogs, a common bile duct ligation group (L group) of six dogs and a phenobarbital pretreatment group (P group) of six dogs that were given phenobarbital for two weeks prior to surgery involving ligation of the common bile duct. Digitoxin was given as a single intravenous administration at a dosage of 20 μg/kg. Multiple plasma samples were collected over a period just prior to digitoxin administration through 72 hours. Plasma digitoxin concentrations were also measured using the RIA technique. The other ten dogs were given tritium labeled (^<3>H-) digitoxin. Eight of these dogs were divided into three groups, the C, L and P groups, as per the conditions described above. In the remaining two dogs (F group), an external biliary fistula was prepared 5 to 6 hours prior to this study. All of these dogs received a single intravenous dose of 50 μCi/14 kg of ^<3>H-digitoxin and 20 μg/kg of cold digitoxin, as a carrier. In these four groups, blood samples from each animal were obtained just prior the operation and at 1, 3, 6, 12 and 24 hours after the intravenous injection of the drug. Urine specimens for these four groups and the bile sample from the F group were collected at 12 hour intervals following the administration of digitoxin and the total volume was also measured. The plasma, urine and bile samples were frozen to be analyzed later. The plasma, urine and bile were extracted using dichrolomethane (CH_2 Cl_2) to separate the cardioinactive water soluble metabolites from the CH_2 Cl_2-soluble cardioactive metabolites and the parent compound. The radioactivity of the CH_2 Cl_2-soluble and -insoluble fractions was counted using a liquid scintillation counter. Quenching was corrected using an automatic external standardization.C. Pharmacokinetic Analysis The natural logarithms of the plasma drug concentration, plotted on the Y axis, were plotted against time on the X axis. The drug concentration-time curve for each individual experimental group and for the group mean data were best fitted by a one- or two-compartment open model. Therefore, the plasma concentration of the drug, as a function of time, may be calculated by one exponential or the sum of two exponentials as in the following equation: Ct=Co･e^-k･t (Equation 1) Ct=A･e^-α･t + B･e^-β･t (Equation 2)where, Ct=the concentration of the drug at any time t Co=the concentration of the drug at time 0 k=the first-order rate constant for the overall elimination of drug from the body α=distribution (fast) phase rate constant β=elimination (slow) phase rate constant A=Y-axis intercept of the extrapolated distribution phase B=Y-axis intercept of the extrapolated elimination phaseThe concentration-time data was subjected to least squares regression analysis and the coefficient of correlation between the theoretical value from the equation and measured concentrations by RIA were calculated. The hybrid constants, A, B, α and β were estimated and used to calculated various pharmacokinetic parameters as follws: Elimination half-life (hr.) t1/2k = (In 1/2)/k Distribution phase half-life (hr.) t1/2α = (In 1/2)/α Elimination phase half-life (hr.) t1/2β = (In 1/2)/β Distribution rate constant (hr.^-1) (from the peripheral into the central compartment) k21 = (A･β + B･α)/(A + B) Elimination rate constant (hr.^-1) (from the central compartment) kel = a ･β /k_21 Distribution rate constant (hr.^-1) (from the central into the peripheral compartment) k_12 = α + β - k_21 - k_e1 Area under the drug concentration-time curve (ng･hr/ml) AUC=Co/k, or A/α+B/β Volume of the central compartment (l/kg) Vd_central = Dose/(A + B) Volume of the peripheral compartment (l/kg) Vd_perpheral = Vd_central･k12/k21 Total volume of the distribution (l/kg) Vd = Dose/Co, or Vd_central+Vd_peripheral Total volume of the distribution using the area method (l/kg) Vd_area = Dose/AUC･β Total body clearance (ml/min/kg) TBCL = Dose/AUC RESULTSA. Digoxin Evaluation The plasma digoxin concentration of the individual dogs, after a single intravenous digoxin injection, decreased rapidly (distribution phase) dering the first few hours and then more slowly (elimination phase) for the following 6 to 8 hours. This concentration-time curve fits a two compartment open model. Therefore, the kinetics of the concentration-time data for individual subjects and for the average data from each group was best described by a bi-exponential function (Equation 2). The mean correlation coefficients between digoxin assayed using RIA and the calculated digoxin value from the equation were 0.997 or more for each of the three groups, indicating an excellent fit. For the digoxin pharmacokinetics of the C group, the mean biological half-life of plasma digoxin in the distribution phase (t_1/2α) was about 1 hour and in the elimination phase (t_1/2β) was approximately 20 hours. The mean volume of the distribution by extrapolation (Vd) was 6.6 l/kg body weight. The apparent distribution volume for the central compartment (Vd_central) was 1.5 l/kg and for the peripheral compartment (Vd_peripheral) was 5.1 l/kg. Calculated from the area under the drug concentration-time curve, the distribution volume (Vd_area) was 9.4 l/kg. The total body clearance (TBCL) was 5.65 ml/min/kg for the C group. For the L group, the plasma digoxin concentration did not, at any time, differ significantly from observed values of the C group. There was no significant difference between the observed value obtained from this group and the C group, regarding the pharmacokinetic parameters of the distribution phase. However, the t_1/2β of about 25 hours, the Vd and Vd_peripheral tended to be prolonged or increased for the L group compared with the C group. The k_12, as the rate constant for the distribution from the central into the peripheral compartment was significantly larger for the L group than for the C group. The k_el, K_21, Vd_area and Vd_central were not significantly different between these two groups. The TBCL for the L group was lower than for the C group, although the difference failed to be statistically significant. For the P group, the pretreated with phenobarbital appeared to increase the activity of the hepatic microsomal enzymes. Despite this increased activity, however, the pharmacokinetic parameters for the distribution phase did not vary significantly from the values obtained from the C and L groups. The t_1/2β for this group, in spite of the common bile duct ligation, was not prolonged and the rate constants, Vd, Vd_central, Vd_peripheral and Vd_area were not increased. There was no significant difference between the P and C groups for these parameters. In contrast, the Vd and Vd_peripheral of the P group were significantly different when compared to the L group. Furthermore, the k_el, as the elimination rate constant from the central compartment, tended to be significantly increased than for the L group. Although the TBCL for the P group was higher than for the C and L groups, there was no statistically significant difference.B. Digitoxin Evaluation The plasma digitoxin concentration-time curve after a single intravenous digitoxin administration was also best described by a bi-exponential (Equetion 2), the same as for the digoxin pharmacokinetics. The mean correlation of the coefficients between the digitoxin values obtained by RIA assay and the theoretical digitoxin values derived from the equation were 0.986 or more for individual animals from each group. For the digitoxin pharmacokinetics of the C group, the mean value of t_1/2α was about 1 hour. The t_1/2β was calculated by using a least squares linear regression analysis from the plasma concentrations over the 48 hour period after the digitoxin dose, because the plasma digitoxin concentrations after 48 hours were not detectable. This mean value was approximately 8 hours. The Vd was 0.808 l/kg body weight. The Vd_central and Vd_peripheral were 0.404 l/kg and 0.403 l/kg, respectively. The Vd_erea was 0.906 l/kg. The TBCL of this drug was 1.57 ml/min/kg body weight for the C group. For the L group, the plasma digitoxin concentration maintained a significantly higher level than that did the C group during all of the sampling periods. However, the pharmacokinetic parameters of the distribution phase were not staistically different between the L group and the C group. The t_1/2β of the L group was about 24.5 hours this value was significantly longer than for the C group. Although the k_el of the L group was lower than the C group, the difference observed for this parameter and also for the other two rate constants failued to achieve statistical significance. The Vd and Vd_area were significantly smaller for the L group than for C group. The Vd_central and Vd_peripheral of the L group did not differ significantly from the C group, but the Vd_peripheral value for the L group was lower than the C group. The TBCL for the L group tended to be less than for the C group. For the P group, the plasma digitoxin concentration-time curve, after the intravenous digitoxin injection, for 4 of the 6 dogs and for the mean data were best descrived by a bi-exponential function (Equation 2). However, the data from the other two dogs in the P group was best fitted by a one compartment open model (Equation 1), because the plasma digitoxin concentration decreased in a mono-exponential pattern and was not observed to be splitting into two phases (α and β phases). Therefore, for these two dogs, the pharmacokinetic parameters concerning the distribution phase could not be calculated. In spite of the fact that the dogs in the P group had the common bile duct ligated, as did the L group, the P group, with phenobarbital pretreatment, had significantly lower plasma digitoxin concentration than were found in the L group. There was no significant difference in plasma digitoxin concentration observed between the P and C groups. All of the pharmacokinetic parameters calculated in this study did not demonstrate a statiatically significant difference between the P and C groups. In comparison with the L group, the pharmacokonetic parameters of the distribution phase for the P group were not significantly different. The t_1/2β was significantly shorter for the P group than the L group. The k_el of this group tended to be higher than the L group. The Vd and Vd_area were significantly larger and Vd_peripheral tended to be larger for this group than the L group. The P group had a significantly larger TBCL than did the L group. In this experiment using ^<3>H-digitoxin, the plasma radioactivity of the CH_2 Cl_2-soluble fraction (digitoxin and its cardioactive metabolites) was higher for the L group than for the C, P and F groups at 12 and 24 hours after the administration of the ^<3>H-digitoxin. The plasma concentrations decreased grdually, after digitoxin administration, for all of the groups. These results were comparable to the results obtained from digitoxin evaluation data from RIA assays. In the radioactive analysis of the first 24 hours urine samples, 15～20 % of the radioactive intravenous digitoxin dose was excreted from the C, L and F groups, during the first 24 hour period. Of the radioactive digitoxin excreted into the urine, 95 % was the CH_2 Cl_2-insoluble fraction (cardioinactive water soluble metabolites). On the other hand, the mean urinary radioactive excretion of the P group, pretreated with phenobarbital, was 35 % of the intravenous digitoxin dose. This larger percentage of radioactive urinary excretion demonstrated by the P group, compared with the three other groups, was almost entirely composed of the CH_2 Cl_2-insoluble fraction. The t_1/2β of the CH_2 Cl_2-soluble fraction was calculated from the plasma concentration at 6, 12 and 24 hours after the single intravenous administration of ^<3>H-digitoxin by employing the least square regression analysis. For the C, P and F groups, the mean t_1/2β ranged from 9.7 to 11.0 hours and was much shorter than the mean value for the L group of 15.4 hours. In the F group, 7 % of the ^<3>H-digitoxin dose was excreted in the bile within the first 24 hours after the intravenous injection and 85 % of the radioactive material excreted into the bile was the CH_2 Cl_2-insoluble fraction. Therefore, most of the ^<3>H-digitoxin excreted into the urine by the four groups and into the bile of the F group was in the aqueous form.C. Biochemical Data and Histological Findings from Liver Examination For all of the groups in the digoxin and digitoxin experiments, the renal function, estimated from the plasma creatinine concentration and the blood urea nitrogen (BUN) were maintained within normal ranges during the entire experimental period. For the liver function tests during the preoperative period, 24, 48 and 72 hours after the digoxin or digitoxin dose, tje total bilirubin (T-Bil), alkaline phosphatase (ALP) and glutamic pyruvic transaminase (GPT) levels were determined using standard clinical laboratory techniques. In the digoxin evaluation, these liver function parameters increased gradually, following the common bile duct ligation and reached higher levels in the L and P groups than in the C group at 24 and 48 hours after the digoxin dose. In the digitoxin evaluation, the ALP and GPT levels gradually increased in the L and P groups and were higher in these groups than in the C group at 24 and 72 hours after the digitoxin dose. The T-Bil of the P group increased as did the other liver function test parameters. Both the C and L groups had almost the same T-Bil values, although the dogs from the L group had undergone common bile duct ligation. The histological examination was conducted using a light microscope after the conclusion of the experiment. The morphological features of a normal liver were observed in tissue speciemens taken from the dogs of the C group. In the L and P groups, including the L group from the digitoxin expoeriment, bile plugs were found in the intrahepatic ducts and canaliculi. The Kupffer cells and hepatocytes contained bile pigment and indications of bile stasis was detected. Furthermore, for dogs from the P group, pretreated phenobarbital, an inducer of hepatic microsomal drug metabolizing enzyme activity, in both of these experiments, the liver preparations demonstrated varying degrees of liver-cell hypertrophy or what are called "induction cells" under the light microscope. DISCUSSIONA. Digoxin Evaluation The digoxin pharmacokinetics for the control group, C group, are in agreement with data that has been previously reported. In humans, digoxin is excreted after administration primarily by the kidney and also in the bile. The largest fraction of the drug and its metabolites that were excreted into the urine and into the bile were the unchanged, original glycoside and also its lipid soluble cardioactive metabolites. The water soluble cardioinactive metabolites of digoxin composed only a small fraction of the compounds that were excreted. The concomitant administration of phenobarbital, as an inducer of hepatic microsomal drug metabolizing enzyme activity, did not affect the digoxin t_1/2β for humans. Therefore, digoxin undergoes insignificant metabolism in humans, the influence of the liver on digoxin metabolism and excretion in humans appears to be limited. Furthermore, it has been repoted by many authors that in patients with hepatic diseases (alcoholic cirrhosis, acute and chronic hepatitis), the blood digoxin concentration and the urinay excretion of digoxin were unaltered. Therefore, digoxin can be administered with relative safety in patients if their renal function is normal. In the dog, digoxin is excreted both in the urine and in the bile, as is the case in humans. Moreover, the excretion volume of digoxin and its metabolites into the urine and bile is almost the same for both species. However, the ratio of the cardioinactive water soluble metabolites of digoxin to the total excreted volume of digoxin and all of its metabolites was much higher in dogs than in humans. For the dogs, the excretion volume of the bile was almost totally composed of cardioinactive water soluble metabolites of digoxin, in contrast to humans. The liver is a primary site involuved in the metabolism of many drugs. It is known that phenobarbital induces hepatic microsomal drug metabolizing enzyme activity. There are some reports that in normal dogs, phenobarbital pretreatment has shortened the digoxin t_1/2β, because the metabolism and excretion of the digoxin in dogs may be affected by the liver. Therefore, the influence of liver diseases on the digoxin pharmcokinetics for dogs appeares to be greater than in humans. In the present study of digoxin in the dog, the apparent volume of distribution was increased and the t_1/2β tended to be prolonged for dogs had undergone experimental cholestasis, due to common bile duct ligation. Furthermore, the t_1/2β in dogs with phenobarbital pretreatment followed by the common bile duct ligation was shortened, because of the enhancement of the hepatic microsomal drug metabolizing enzyme activity. Althought there are some studies that suggest the pharmacokinetics of digoxin appears to be less susceptible to liver influence, the results from this experiment that the liver can significantly influence the metabolism and excretion of digoxin in dogs. However, the effect of the liver may be less than the kidney, as plasma digoxin concentrations did not vary significantly between dogs, whether or not they had undergone surgical cholestasis, throughout this experiment. In clinical situations, it has been speculated that digoxin concentrations in blood and tissues may increase gradually and the incidence of digitalis toxicity caused by high concentrations may be increased in dogs that have liver diseases during their digoxin maintenance therapy. Accordingly, dogs who require digoxin should be evaluated carefully regarding not only their renal function but also their liver function, and precautions should be taken when digoxin is administered to patients with cholestasis or other hepatic disorders.B. Digitoxin Evaluation The digitoxin pharmacokinetics for the control group of dogs, C group, from the present study are in agreement with data that has been previously reported. In humans, although digitoxin excreted into the urine and bile, the ratio of the cardioinactive water soluble metabolites of digitoxin to the total excreted volume is less for both the urine and bile. Furthermore, almost all of the previous studies concluded that the blood concentrations and t_1/2β of digitoxin in patients are not affected by various liver diseases (cirrhosis, acute and chronic hepatitis). It has also been suggested that patients with chronic active hepatitis have a shortened t_1/2β compared with control subjects, because of the enhansment of the digitoxin metabolism and excretion. It can be stated from these results that the influence of the liver on digitoxin pharmacokinetics for humans does not appeare to be significant and the digitoxin elimination is not impaired by various forms of hepatic diseases, although the concomitant administration of phenobarbital may reduce the t_1/2β of digitoxin. In the dog, digitoxin is excreted into the urine and bile, as is the case in humans. However, the volume of digitoxin and its metabolites excreted in dogs is much greater than in humans. Moreover, for dogs the largest percentage of the excreted total volume is comprised of the cardioinactive water soluble metabolites of digitoxin. It appears that digitoxin that is administered to dogs undergoes a significant degree of hepatic biotransformation and, thereafter, the unchanged original glycoside and its cardioactive and cardioinactive metabolites are excreted into the urine and the bile. Therefore, the influence of the liver on digitoxin metabolism and excretion in the dog seems to be much greater than in humans. On the other hand, there are some reports that the t_1/2β of digitoxin in dogs, even with the most severe liver disease, is not prolonged and also that the pretreatment with phenobarbital, as an inducer of hepatic microsomal enzymes, does not alter the t_1/2β of the dogs. In the present study of digitoxin in the dog, the plasma digitoxin concentration maintained a significantly higher level in the group of dogs that had the common bile duct ligation than for the control group, throughout the experiment. The t_1/2β was about three times longer for the dogs with experimental cholestasis than for the control dog group. Furthermore, the dogs that received phenobarbital pretreatment, followed by the common bile duct ligation, did not havae increased plasma digitoxin levels or a longer t_1/2β than the control dog group. For the group of dog that was pretreated with phenobarbital, the volume of the cardioinactive water soluble digitoxin metabolites that were excreted into the urine was increased. These results indicate that the liver can significantly influence digitoxin metabolism and excretion in dogs. Although liver damage plays a role in the alteration of the pharmacokinetics of digitoxin in dogs, the influence of biotransformation might differ depending on the type of liver disease and the degree of liver damage. In clinical situations, blood and tissue digitoxin concentrations may be expected to increase even more in dogs with liver diseases during maintenance therapy and digitalis toxicity is a serious concern. Accordingly, prior to digitoxin therapy, patients should be evaluated carefully, specifically focusing on the status of their liver function. Digitoxin should be used with caution in dogs with cholestasis or other liver diseases.C. Comparison Between Digoxin and Digitoxin Digitalis glycosides consist of the basic steroid-type nuleus (cyclo-pentanoperhydrophenanthrene nucleus) to which is attached an unsaturated lactene ring at carbon atom 17 (C-17) and three glucose molecules at C-3. The digoxin and digitoxin molecules differ only in one position as digoxin has a hydroxyl (OH) group at C-12 in the steroid-type nucleus, while digitoxin lacks OH group at this position. This small difference in the structure significantly influences the water and lipid solubility, the extent of plasma protein binding and the rate of gastrointestinal absorption. Digoxin, with the OH group at C-12 and C-14 in a steroid-type nucleus, is more polar, less successfully bound to the plasma proteins and is absorbed to a lesser degree. On the other hand, digitoxin, with a OH group only at C-14, is a relatively nonpolar digitalis glycoside and is therefore nearly completely absorbed across the gastrointestinal menbrane after oral administration. Also digitoxin binds much more successfully to the plasma proteins. It has been demonstrated that there are significant differences between digoxin and digitoxin pharmacokinetics. In the dog, the t_1/2β for digoxin is approximately 20 to 30 hours and the t_1/2β for digitoxin is significantly shorter, from 6 to 14 hours. The results from the present study concerning the relationship between the t_1/2β of digoxin and digitoxin are in agreement with previously published experimental results. The dog had a longer t_1/2β for digoxin than for digitoxin. The t_1/2β for digoxin for humans and dogs are almost the same. However, in humans the t_1/2β of digitoxin is from 4 to 10 days, and for humans the t_1/2β for this drug is much longer than for digoxin. The pharmacokinetics of digitalis, especially digitoxin, in the dog are significantly different from those in humans. Concerning the volume of distribution, there are differences between these two digitalis glycosides. The volume of digoxin distribution is about 10 times greater than for digitoxin. In clinical practice both digoxin and digitoxin have both benefits and disadvantages. The comparative study of digoxin and digitoxin is very important for the evaluation of suitable digitalis therapy. There have been few clinical evaluations reported concerning digitoxin and also very few clinical and experimental comparisons between these two digitalis drugs. In this section, a comparisons between the influence of the liver on digoxin and digitoxin pharmacokinetics will be discussed. The influence of experimental cholestasis on digoxin pharmacokinetics differs from digitoxin. Consequently, it appears that there are differences in the role of the liver and its effect on these two drugs. The t_1/2β of both digoxin and digitoxin for dogs was prolonged due to the experimental cholestasis and the t_1/2β was shortened for dogs that had experienced cholestasis elicited by pretreatment with phenobarbital which increased the activity of the hepatic microsomal drug metabolizing enzymes. However, the prolongation of the t_1/2β was much larger for digitoxin than for digoxin. Furthermore, the plasma digitoxin concentration was significantly higher in dogs with cholestasis than in the group of control dogs. In the digoxin experiment, there was no statistically significant difference the plasma digoxin concentrations observed between these two groups. From the data, the influence of the liver on digitoxin pharmacokinetics appears to be much more significant than the liver's effects on digoxin in canines. It has been postulated that digitoxin, which is a highly bound plasma protein and also lipid soluble, may undergo more complicated pharmacokinetics in canines as a result of various pathological conditions. These conditions can influence the plasma protein concentration, e.g., hypoproteinemia caused by the exacerbation of congestive heart failure or liver diseases, or plasma protein movement due to ascites, in addition to the direct influence of liver disorders. Therefore, for the clinical selection of a specific digitalis glycoside, it is suggested that digoxin may be preferable to digitoxin for attainment of optimal digitalis therapy in dogs with cholestasis or other hepatic diseases. The pharmacokinetics of digoxin and digitoxin in dogs are very complex as is the case n humans. Comparative studies of these two drugs are scant and confusion may result from the extrapolation of the pharmacokinetic data obtained from human experimental evaluations and then applied to dogs, due to species differences. Increased clinical and experimental investigations focusing on these two digitalis glycosides during maintenance therapy is suggested, evaluating both normal, healthy canines and comparing these results with data obtained from dogs exhibiting various stages of differing pathological conditions. This additional data will assist in establishing optimal digitalis doseage schedules and to determine the clinical usefulness of both of these drugs. Future studies should include comparisons between digoxin and digitoxin pharmacokinetics, since these drug's biotransformation and excretion in canines is very complex and differs significantly from humans.