THE JOURNAL OF INFECTIOUS DISEASES • VOL. 131, NO.3· © 1975 by the University of Chicago. All rights reserved.
MARCH 1975
Pharmacokinetics of Lincomycin and Clindamycin Phosphate in a Canine Model Richard B. Brown, Michael Barza, John L. Brusch, Yasuhiko Hashimoto, and Louis Weinstein
From the Infectious Disease and Anesthesiology Services, the New England Medical Center Hospitals; and the Departments of Medicine and Anesthesiology, Tufts University School of Medicine, Boston, Massachusetts
liver disease [5, 6]. That of lincomycin must also be altered in the presence of renal insufficiency [7]; whether or not the same is true for clindamycin remains controversial [6, 8]. The relation between excretion of various antimicrobial agents into bile and patency of the biliary tree has been studied extensively. Most antibiotics appear to require ductal patency for excretion [9-11]. However, lincomycin and clindamycin phosphate have not been well studied in this regard. Concentrations of clindamycin in gall bladder bile from patients with obstruction of the cystic duct were reported to be extremely low [12]. In this study the hepatic and renal pharmacology of lincomycin and clindamycin phosphate were compared in a canine model, and the effects of biliary obstruction on the behavior of each were studied. Although these were acutephase studies in dogs, the results suggest differences between the drugs that may have clinical relevance.
Lincomycin is effective against many grampositive bacteria [1]. Clindamycin, its 7-desoxy , 7-chloro derivative, possesses a similar antimicrobial spectrum but is also active against most pathogenic anaerobes [2, 3]. The parenteral preparation available for clincial use is the phosphate ester, which is not active in vitro; however, it is rapidly hydrolyzed in the body to the nonesterified, active form [4]. Throughout this paper the terms nonesterified and total clindamycin will be used; the latter refers to the sum of esterified and nonesterified drug. Because clindamycin and lincomycin are extensively eliminated by hepatic mechanisms, the dosage of each must be modified in patients with Received for publication June 28, 1974, and in revised form November 6, 1974. This work was supported in part by training grant no. 276 from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, and by a grant from the Upjohn Company, Kalamazoo, Michigan. We are grateful to Norman Peu and Carmen Keough for technical help and to Charles Gambel, Pat Smith, and Caroline Shanks for surgical and statistical assistance. We also acknowledge the guidance of Dr. Shiro Shimosato. Please address requests for reprints to Dr. Richard B. Brown, New England Medical Center Hospital, 171 Harrison Avenue, Boston, Massachusetts 02111.
Materials and Methods
Healthy mongrel dogs (15-35 kg) were observed for at least two days prior to surgery. All intake of food and water was discontinued 12 hr before 252
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on August 29, 2015
Lincomycin and clindamycin phosphate were studied in a canine model in which acute biliary obstruction was produced during iv infusion of antibiotic. Hepatic and renal extraction, biliary and renal excretion, and concentrations in liver and kidney were measured. Total and nonesterified clindamycin were assayed. The antibiotics were taken up by the liver at similar rates; however, the rates of excretion and concentration in bile were significantly higher for lincomycin than for clindamycin. Biliary obstruction did not affect the concentration of either antibiotic in canalicular bile. Lincomycin was extracted by the kidneys and excreted into urine at a much higher rate than was clindamycin. Concentrations of nonesterified clindamycin in the hepatic vein were higher than those in the portal vein, an observation suggesting metabolic activation within the liver. This relation was reversed by biliary obstruction. The results in this canine model indicate a greater role for the kidney in the disposition of lincomycin than in that of clindamycin, major differences between the rates of biliary excretion of the two agents, and a probable change in the metabolism of clindamycin produced by acute biliary obstruction.
Lincomycin and Clindamycin Phosphate
rates were determined, and specimens of blood were collected from the arterial and venous catheters every 15 min. ,With experience, it was possible to obtain all samples of blood for a given interval within 2 min. Care was taken to assure freedom of all specimens from anticoagulant. Urine and bile were collected over intervals of 30 min. Four groups of animals were studied. In two (three dogs each) the common bile duct remained unobstructed throughout 3.5 hr of continuous infusion of lincomycin or clindamycin phosphate. The same agents were administered to the other two groups (four dogs each) for 2 hr, after which the common duct was acutely ob~tructed. The infusion was continued for another 90 min. The period prior to obstruction was designated Part I of the experiment; that after obstruction was Part II. The duct was obstructed in the following manner. A manometer with its zero point at the level of the gall bladder was filled with antibiotic-free bile until equilibrium was reached. Equilibrium was virtually always reached at a level of 30 cm of bile within 20 min. The level remained constant throughout the rest of the experiment, a fact indicating that there was no leakage. During Part II bile was collected only at the end of the period so that the system would not be disrupted. Three portions were obtained: (A) bile contained in the manometer; (B) bile easily expressed from the cannula; and (C) bile obtained by milking the deepest portions of the common duct. At the end of Parts I and II, punch biopsies of the right kidney and wedge biopsy of the anterior-inferior edge of the left lobe of the liver were performed. The quantities of renal and hepatic tissue obtained were approximately 300 mg and 3 g, respectively. The specimens were frozen at - 20 C and were assayed for antimicrobial activity within'two days. The concentrations of antibiotic in fluids and tissues were assayed by an agar diffusion method with use of trypticase soy agar (Baltimore Biological Laboratories, Baltimore, Md.) seeded with approximately 105 Sarcina lutea/ml. Both total and nonesterified clindamycin were measured as described elsewhere [13]. Briefly, each sample was divided into two aliquots. That
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on August 29, 2015
the studies were initiated. Intubation was accomplished with use of iv gallamine triethiodide. Anesthesia was produced with iv chloraloseurethan and inhalation of 50% nitrous oxide with oxygen. The right jugular and femoral veins were catheterized with the animals in the supine position on a heated operating table. A midline abdominal incision was made from the xiphoid to the suprapubic area. The hepatic artery, portal and hepatic veins, and left renal artery were mobilized, and the cystic duct was ligated. Polyethylene catheters were inserted into the femoral artery and vein, left renal vein, and portal vein. With the liver gently retracted, a curved flexible catheter was inserted into a major branch of the hepatic vein and threaded downward 1 or 2 cm toward the liver. All catheters were stabilized by sutures and flushed with a heparin solution. Flow I-robes (Statham Flo-probe, Statham Co., Los Angeles, Calif.) were applied to the hepatic artery and the left renal and portal veins and were connected to an electromagnetic flow meter (M-400l, Statham). In addition, both ureters and the common bile duct were cannulated. The abdomen was draped with warm moist towels. Surgical preparation was completed in less than 2.5 hr. Careful hemostasis was preserved, and the surgical field was kept clean but not sterile. Adequate ventilation was assured by frequent measurement of arterial blood gases. Blood pressure, pulse rate, hematocrit, and urine output were measured; electrocardiographic monitoring was performed continuously. Fluids (glucose in water, saline, bicarbonate, and lactated Ringer's solution) were administered as required. Several studies were terminated because of poor urinary output or unstable vital signs; these data are not included in this report. Lincomycin or clindamycin phosphate (Upjohn Company, Kalamazoo, Mich.) dissolved in 0.85% NaCI was administered in a dose of 20 mg/kg per hr by continuous infusion into the jugular vein. Each antibiotic was diluted so that the infusion volume was 72 m1 per hr, and a Harvard pump (Harvard Apparatus Co., Dover, Mass.) was employed to insure constant delivery. Specimens of blood and bile (free of antibiotic) were obtained for use in studies of biliary obstruction and in assays. Vascular flow
253
Brown et al.
254
renal extraction was derived by multiplying the difference between the concentrations in renal venous serum and arterial serum by the plasma flow rate. For this purpose, the level of drug in the femoral artery was taken to represent that in the renal artery. The final value was doubled so that both kidneys were taken into account. Excretion. Biliary excretion (JLg per 30 min) for each interval was determined from the product of the concentration of antibiotic (JLg/ml) and the volume of bile (ml per 30 min). Urinary excretion (JLg per 30 min) via both ureters was measured in a similar manner. Percentage extraction. The percentage of antibiotic extracted by the liver was derived by expressing the difference between the concentration of drug in the portal vein (P) and that in the hepatic vein (H) as a percentage of the level in the portal vein, i.e., [(P - H) x 100]/P. The formula [(R - V) x 100]/R was used for the kidney; R and V represent the concentration of drug in arterial and renal venous serum, respectively. Statistical analysis. All analyses were done by the unpaired t-test. All results obtained after the 45-min time interval were used in comparing the two control groups. The effects of biliary obstruction on the concentration of antibiotic in serum extraction and on the percentage of drug extracted were assessed by comparison of results for the five I5-min intervals immediately preceding with those after obstruction. The cor.responding determination for rates of excretion of drug into urine was made by use of the three 30-min collections just before·and after interrup.,. tion of biliary flow. The same number of preand post-obstructive values was used to detect possible differences between the antibiotics. Results
Mean intravascular concentrations and flow rates. The concentrations of lincomycin and total clindamycin in the vessels of control animals were essentially the same as those in the animals that underwent obstruction; data for the latter group are shown in figures 1 and 2. The concentrations of lincomycin in renal and hepatic veins were consistently lower than those in the renal (femoral) artery and the portal vein. In contrast, levels of total clindamycin in the renal vein were similar to those in the portal vein and the femoral artery, and were higher than those in the hepatic vein. Levels of each drug in the
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on August 29, 2015
portion to be studied for total clindamycin activity was mixed with an equal volume of alkaline phosphatase in diethyl barbiturate solution (p H 9) and incubated at 37 C for 16 hr. Aliquots for assay of nonesterified clindamycin were mixed with equal volumes of 1.0 M phosphate buffer (P :fI 8), frozen at -20' C, and assayed at the same time as samples studied for total clindamycin. No distinction was made between nonesterified clindamycin and bioactive metabolites. Fresh standards containing known concentrations of antibiotic were prepared in NaCI as well as in pooled antibiotic-free canine bile (2% and 10% in NaCl) and serum. These, together with the specimens for study, were adsorbed onto filter paper disks and were assayed in triplicate. The disks were placed on the surface of agar plates and incubated at 37 C for 18 hr. Zones of inhibition of bacterial growth were measured, and the mean values were plotted on semilogarithmic paper. The lower limit of sensitivity of the assay was 0.325 JLg/ml for both drugs. Specimens of bile containing lincomycin were diluted 1:50 and compared with standards containing 2% bile; for clindamycin, a dilution factor of 1: 10 was used, and specimens were compared with the standards containing 10% bile. Urine containing either antibiotic was diluted 1:50 to produce interpretable zones. Biopsy specimens were immersed in a solution of 2.5% trypsin equal in weight to three times that of the tissue. The mixture was homogenized in a Waring blender for approximately 5 min, adsorbed onto filter paper disks, and plated. Standards for these specimens were prepared in 0.85% NaCl. In studies in our laboratory, no difference was demonstrated for either antibiotic between standards prepared in 0.85% NaCI and standards diluted in homogenates of tissue prepared in the manner described above. Methods of calculation. Extraction. The rate of extraction of antibiotic by the liver (JLg per 30 min) was calculated by multiplying the difference between the concentration of drug in the portal vein and that in the hepatic vein (JLg/ml) by the plasma flow rate (ml per 30 min) x 5/4. Plasma flow rate was determined by multiplying portal blood flow by (100 - hematocrit)/100. The factor of 5/4 was introduced because 20% of blood flow to the canine liver comes from the hepatic artery [14]. The rate of
255
Lincomycin and Clindamycin Phosphate
30
25
z o
..-
20
c:(
a:
-----....... -'-
~
z
LIJ U
15
Z
o
u
Femoral Artery Portal Vein Hepatic Vein Renal Vein
Biliary Obstruct ion
5
30
60
90
120
180
150
210
TIME (min)
various vascular sites were unaltered by biliary obstruction. Rates of flow through the various vessels remained stable in individual dogs and were unaffected by biliary obstruction. Mean values in animals with obstruction were 291-517 ml per min in the portal vein and 174-244 ml per min in the renal vein. There were no significant differences among the mean values for the four groups of animals. Hepatic pharmacology. In the control groups, rates of hepatic extraction, percentage extraction, biliary excretion, and biliary con-
centration of antibiotic were relatively stable after 1 hr. Extraction of total clindamycin in the controls was 40,600--65,400 f.Lg per 30 min, the percentage extraction was 24%-30%, and the rate of excretion was 34-180 f.Lg per 30 min. Biliary concentrations (5-46 f.Lg/ml) were only slightly higher than those in serum. The extraction and percentage extraction of lincomycin were similar to those values for total clindamycin; in contrast, both biliary excretion (1,000-7,500 f.Lg per 30 min) and concentration in bile (163-1,175 f.Lg/ml) were significantly higher for lincomycin than for clindamycin (P
MARCH 1975
Pharmacokinetics of Lincomycin and Clindamycin Phosphate in a Canine Model Richard B. Brown, Michael Barza, John L. Brusch, Yasuhiko Hashimoto, and Louis Weinstein
From the Infectious Disease and Anesthesiology Services, the New England Medical Center Hospitals; and the Departments of Medicine and Anesthesiology, Tufts University School of Medicine, Boston, Massachusetts
liver disease [5, 6]. That of lincomycin must also be altered in the presence of renal insufficiency [7]; whether or not the same is true for clindamycin remains controversial [6, 8]. The relation between excretion of various antimicrobial agents into bile and patency of the biliary tree has been studied extensively. Most antibiotics appear to require ductal patency for excretion [9-11]. However, lincomycin and clindamycin phosphate have not been well studied in this regard. Concentrations of clindamycin in gall bladder bile from patients with obstruction of the cystic duct were reported to be extremely low [12]. In this study the hepatic and renal pharmacology of lincomycin and clindamycin phosphate were compared in a canine model, and the effects of biliary obstruction on the behavior of each were studied. Although these were acutephase studies in dogs, the results suggest differences between the drugs that may have clinical relevance.
Lincomycin is effective against many grampositive bacteria [1]. Clindamycin, its 7-desoxy , 7-chloro derivative, possesses a similar antimicrobial spectrum but is also active against most pathogenic anaerobes [2, 3]. The parenteral preparation available for clincial use is the phosphate ester, which is not active in vitro; however, it is rapidly hydrolyzed in the body to the nonesterified, active form [4]. Throughout this paper the terms nonesterified and total clindamycin will be used; the latter refers to the sum of esterified and nonesterified drug. Because clindamycin and lincomycin are extensively eliminated by hepatic mechanisms, the dosage of each must be modified in patients with Received for publication June 28, 1974, and in revised form November 6, 1974. This work was supported in part by training grant no. 276 from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, and by a grant from the Upjohn Company, Kalamazoo, Michigan. We are grateful to Norman Peu and Carmen Keough for technical help and to Charles Gambel, Pat Smith, and Caroline Shanks for surgical and statistical assistance. We also acknowledge the guidance of Dr. Shiro Shimosato. Please address requests for reprints to Dr. Richard B. Brown, New England Medical Center Hospital, 171 Harrison Avenue, Boston, Massachusetts 02111.
Materials and Methods
Healthy mongrel dogs (15-35 kg) were observed for at least two days prior to surgery. All intake of food and water was discontinued 12 hr before 252
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on August 29, 2015
Lincomycin and clindamycin phosphate were studied in a canine model in which acute biliary obstruction was produced during iv infusion of antibiotic. Hepatic and renal extraction, biliary and renal excretion, and concentrations in liver and kidney were measured. Total and nonesterified clindamycin were assayed. The antibiotics were taken up by the liver at similar rates; however, the rates of excretion and concentration in bile were significantly higher for lincomycin than for clindamycin. Biliary obstruction did not affect the concentration of either antibiotic in canalicular bile. Lincomycin was extracted by the kidneys and excreted into urine at a much higher rate than was clindamycin. Concentrations of nonesterified clindamycin in the hepatic vein were higher than those in the portal vein, an observation suggesting metabolic activation within the liver. This relation was reversed by biliary obstruction. The results in this canine model indicate a greater role for the kidney in the disposition of lincomycin than in that of clindamycin, major differences between the rates of biliary excretion of the two agents, and a probable change in the metabolism of clindamycin produced by acute biliary obstruction.
Lincomycin and Clindamycin Phosphate
rates were determined, and specimens of blood were collected from the arterial and venous catheters every 15 min. ,With experience, it was possible to obtain all samples of blood for a given interval within 2 min. Care was taken to assure freedom of all specimens from anticoagulant. Urine and bile were collected over intervals of 30 min. Four groups of animals were studied. In two (three dogs each) the common bile duct remained unobstructed throughout 3.5 hr of continuous infusion of lincomycin or clindamycin phosphate. The same agents were administered to the other two groups (four dogs each) for 2 hr, after which the common duct was acutely ob~tructed. The infusion was continued for another 90 min. The period prior to obstruction was designated Part I of the experiment; that after obstruction was Part II. The duct was obstructed in the following manner. A manometer with its zero point at the level of the gall bladder was filled with antibiotic-free bile until equilibrium was reached. Equilibrium was virtually always reached at a level of 30 cm of bile within 20 min. The level remained constant throughout the rest of the experiment, a fact indicating that there was no leakage. During Part II bile was collected only at the end of the period so that the system would not be disrupted. Three portions were obtained: (A) bile contained in the manometer; (B) bile easily expressed from the cannula; and (C) bile obtained by milking the deepest portions of the common duct. At the end of Parts I and II, punch biopsies of the right kidney and wedge biopsy of the anterior-inferior edge of the left lobe of the liver were performed. The quantities of renal and hepatic tissue obtained were approximately 300 mg and 3 g, respectively. The specimens were frozen at - 20 C and were assayed for antimicrobial activity within'two days. The concentrations of antibiotic in fluids and tissues were assayed by an agar diffusion method with use of trypticase soy agar (Baltimore Biological Laboratories, Baltimore, Md.) seeded with approximately 105 Sarcina lutea/ml. Both total and nonesterified clindamycin were measured as described elsewhere [13]. Briefly, each sample was divided into two aliquots. That
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on August 29, 2015
the studies were initiated. Intubation was accomplished with use of iv gallamine triethiodide. Anesthesia was produced with iv chloraloseurethan and inhalation of 50% nitrous oxide with oxygen. The right jugular and femoral veins were catheterized with the animals in the supine position on a heated operating table. A midline abdominal incision was made from the xiphoid to the suprapubic area. The hepatic artery, portal and hepatic veins, and left renal artery were mobilized, and the cystic duct was ligated. Polyethylene catheters were inserted into the femoral artery and vein, left renal vein, and portal vein. With the liver gently retracted, a curved flexible catheter was inserted into a major branch of the hepatic vein and threaded downward 1 or 2 cm toward the liver. All catheters were stabilized by sutures and flushed with a heparin solution. Flow I-robes (Statham Flo-probe, Statham Co., Los Angeles, Calif.) were applied to the hepatic artery and the left renal and portal veins and were connected to an electromagnetic flow meter (M-400l, Statham). In addition, both ureters and the common bile duct were cannulated. The abdomen was draped with warm moist towels. Surgical preparation was completed in less than 2.5 hr. Careful hemostasis was preserved, and the surgical field was kept clean but not sterile. Adequate ventilation was assured by frequent measurement of arterial blood gases. Blood pressure, pulse rate, hematocrit, and urine output were measured; electrocardiographic monitoring was performed continuously. Fluids (glucose in water, saline, bicarbonate, and lactated Ringer's solution) were administered as required. Several studies were terminated because of poor urinary output or unstable vital signs; these data are not included in this report. Lincomycin or clindamycin phosphate (Upjohn Company, Kalamazoo, Mich.) dissolved in 0.85% NaCI was administered in a dose of 20 mg/kg per hr by continuous infusion into the jugular vein. Each antibiotic was diluted so that the infusion volume was 72 m1 per hr, and a Harvard pump (Harvard Apparatus Co., Dover, Mass.) was employed to insure constant delivery. Specimens of blood and bile (free of antibiotic) were obtained for use in studies of biliary obstruction and in assays. Vascular flow
253
Brown et al.
254
renal extraction was derived by multiplying the difference between the concentrations in renal venous serum and arterial serum by the plasma flow rate. For this purpose, the level of drug in the femoral artery was taken to represent that in the renal artery. The final value was doubled so that both kidneys were taken into account. Excretion. Biliary excretion (JLg per 30 min) for each interval was determined from the product of the concentration of antibiotic (JLg/ml) and the volume of bile (ml per 30 min). Urinary excretion (JLg per 30 min) via both ureters was measured in a similar manner. Percentage extraction. The percentage of antibiotic extracted by the liver was derived by expressing the difference between the concentration of drug in the portal vein (P) and that in the hepatic vein (H) as a percentage of the level in the portal vein, i.e., [(P - H) x 100]/P. The formula [(R - V) x 100]/R was used for the kidney; R and V represent the concentration of drug in arterial and renal venous serum, respectively. Statistical analysis. All analyses were done by the unpaired t-test. All results obtained after the 45-min time interval were used in comparing the two control groups. The effects of biliary obstruction on the concentration of antibiotic in serum extraction and on the percentage of drug extracted were assessed by comparison of results for the five I5-min intervals immediately preceding with those after obstruction. The cor.responding determination for rates of excretion of drug into urine was made by use of the three 30-min collections just before·and after interrup.,. tion of biliary flow. The same number of preand post-obstructive values was used to detect possible differences between the antibiotics. Results
Mean intravascular concentrations and flow rates. The concentrations of lincomycin and total clindamycin in the vessels of control animals were essentially the same as those in the animals that underwent obstruction; data for the latter group are shown in figures 1 and 2. The concentrations of lincomycin in renal and hepatic veins were consistently lower than those in the renal (femoral) artery and the portal vein. In contrast, levels of total clindamycin in the renal vein were similar to those in the portal vein and the femoral artery, and were higher than those in the hepatic vein. Levels of each drug in the
Downloaded from http://jid.oxfordjournals.org/ at University of Manitoba on August 29, 2015
portion to be studied for total clindamycin activity was mixed with an equal volume of alkaline phosphatase in diethyl barbiturate solution (p H 9) and incubated at 37 C for 16 hr. Aliquots for assay of nonesterified clindamycin were mixed with equal volumes of 1.0 M phosphate buffer (P :fI 8), frozen at -20' C, and assayed at the same time as samples studied for total clindamycin. No distinction was made between nonesterified clindamycin and bioactive metabolites. Fresh standards containing known concentrations of antibiotic were prepared in NaCI as well as in pooled antibiotic-free canine bile (2% and 10% in NaCl) and serum. These, together with the specimens for study, were adsorbed onto filter paper disks and were assayed in triplicate. The disks were placed on the surface of agar plates and incubated at 37 C for 18 hr. Zones of inhibition of bacterial growth were measured, and the mean values were plotted on semilogarithmic paper. The lower limit of sensitivity of the assay was 0.325 JLg/ml for both drugs. Specimens of bile containing lincomycin were diluted 1:50 and compared with standards containing 2% bile; for clindamycin, a dilution factor of 1: 10 was used, and specimens were compared with the standards containing 10% bile. Urine containing either antibiotic was diluted 1:50 to produce interpretable zones. Biopsy specimens were immersed in a solution of 2.5% trypsin equal in weight to three times that of the tissue. The mixture was homogenized in a Waring blender for approximately 5 min, adsorbed onto filter paper disks, and plated. Standards for these specimens were prepared in 0.85% NaCl. In studies in our laboratory, no difference was demonstrated for either antibiotic between standards prepared in 0.85% NaCI and standards diluted in homogenates of tissue prepared in the manner described above. Methods of calculation. Extraction. The rate of extraction of antibiotic by the liver (JLg per 30 min) was calculated by multiplying the difference between the concentration of drug in the portal vein and that in the hepatic vein (JLg/ml) by the plasma flow rate (ml per 30 min) x 5/4. Plasma flow rate was determined by multiplying portal blood flow by (100 - hematocrit)/100. The factor of 5/4 was introduced because 20% of blood flow to the canine liver comes from the hepatic artery [14]. The rate of
255
Lincomycin and Clindamycin Phosphate
30
25
z o
..-
20
c:(
a:
-----....... -'-
~
z
LIJ U
15
Z
o
u
Femoral Artery Portal Vein Hepatic Vein Renal Vein
Biliary Obstruct ion
5
30
60
90
120
180
150
210
TIME (min)
various vascular sites were unaltered by biliary obstruction. Rates of flow through the various vessels remained stable in individual dogs and were unaffected by biliary obstruction. Mean values in animals with obstruction were 291-517 ml per min in the portal vein and 174-244 ml per min in the renal vein. There were no significant differences among the mean values for the four groups of animals. Hepatic pharmacology. In the control groups, rates of hepatic extraction, percentage extraction, biliary excretion, and biliary con-
centration of antibiotic were relatively stable after 1 hr. Extraction of total clindamycin in the controls was 40,600--65,400 f.Lg per 30 min, the percentage extraction was 24%-30%, and the rate of excretion was 34-180 f.Lg per 30 min. Biliary concentrations (5-46 f.Lg/ml) were only slightly higher than those in serum. The extraction and percentage extraction of lincomycin were similar to those values for total clindamycin; in contrast, both biliary excretion (1,000-7,500 f.Lg per 30 min) and concentration in bile (163-1,175 f.Lg/ml) were significantly higher for lincomycin than for clindamycin (P
