A Technique for In Vitro Studying of the Permeability of the Spinal Cord Dura Mater.

We propose an in vitro method for studying permeability of spinal cord dura mater for components of autological serum using an original device. Sixty native samples of the spinal cord dura mater obtained from 12 mongrel dogs were used for testing of the device. The coefficient of permeability variation (V) for blood serum substances did not exceed 5% in most cases excluding lactate (V=8.03%). Analysis of spinal cord dura mater permeability in vitro for various substances using the developed device provides reproducible results with acceptable variability (5-10%).

Response of a clinical Escherichia coli strain to repeated cefquinome exposure in a piglet tissue-cage model.

BACKGROUND: In order to provide some basis for effective dosage regimens that optimize efficacy with respect to bacteriological and clinical cures, the in vivo activity of cefquinome against a clinical Escherichia coli (E.coli) strain (the minimum inhibitory concentration value for this strain equals to the MIC90 value of 0.25 µg/ml for 210 E.coli strains isolated from pigs) was investigated by using a piglet tissue-cage infection model. The aim was to elucidate the pharmacokinetic/pharmacodynamics (PK/PD) index associated with cefquinome efficacy, and then to identify the magnitude of the PK/PD parameter required for different degree of efficacy in clinical treatment. RESULTS: Tissue-cage infection model was established in piglets, and then the animals received intramuscular injection of cefquinome twice a day for 3 days to create a range of different drug exposures. The tissue-cage fluid was collected at 1, 3, 6, 9 and 12 h after every drug administration for drug concentrationdetermination and bacteria counting. Different cefquinome regimens produced different percentages of time during that drug concentrations exceeded the MIC (%T > MIC), ranging from 0% to 100%. Cefquinome administration at 0.2, 0.4, 0.6, 0.8, 1, 2 and 4 mg/kg reduced the bacterial count (log10 CFU/mL) in tissue-cage fluid by -1.00 ± 0.32, -1.83 ± 0.08, -2.33 ± 0.04, -2.96 ± 0.16, -2.99 ± 0.16, -2.93 ± 0.11, -3.43 ± 0.18, respectively. The correlation coefficient of the PK/PD index with antibacterial effect of the drug was 0.90 for %T > MIC, 0.62 for AUC0-12/MIC, and 0.61 for Cmax/MIC, suggesting the most important PK/PD parameter was %T > MIC. A inhibitory form of sigmoid maximum effect (Emax) model was used to estimate %T > MIC, and the respective values required for continuous 1/6-log drop, 1/3-log drop and 1/2-log drop of the clinical E.coli count during each 12 h treatment period were 3.97%, 17.08% and 52.68%. CONCLUSIONS: The data derived from this study showed that cefquinome exhibited time-dependent killing profile. And from the results of the present study, it can be assumed that when %T > MIC reached 52.68%, cefquinome could be expected to be effective against a clinical E.coli strain for which the MIC value is below 0.128 µg/ml (3-log drop of bacteria count can be achieved after six successive administrations for 3 days).

In vivo evaluation of mutant selection window of cefquinome against Escherichia coli in piglet tissue-cage model.

BACKGROUND: The resistance of cephalosporins is significantly serious in veterinary clinic. In order to inhibit the bacterial resistance production, the mutant selection window (MSW) hypothesis with Escherichia coli (E. coli) ATCC 25922 exposed to cefquinome in an animal tissue-cage model was investigated. RESULTS: Localized infection with E. coli was established in piglets, and the infected animals were administrated intramuscularly with various doses and intervals of cefquinome to provide antibiotic concentrations below the MIC99, between the MIC99 and the mutant prevention concentration (MPC), and above the MPC. E. coli lost susceptibility when drug concentrations fluctuated between the lower and upper boundaries of the window, which defined in vitro as the MIC99 (0.06 µg/mL) and the MPC (0.16 µg/mL) respectively. For PK/PD parameters, there were no mutant selection enrichment when T>MIC99 was ≤ 25% or T>MPC was ≥ 50% of administration interval. When T>MIC99 was > 25% and T>MPC was <50% of administration interval, resistance selection was observed. When AUC24 h/MIC99 and AUC24 h/MPC were considered, the mutant selection window extended from 32.84 h to 125.64 h and from 12.83 h to 49.09 h, respectively. CONCLUSIONS: These findings demonstrate that the MSW exists in vivo for time-dependent antimicrobial agents, and its boundaries fit well with those determined in vitro. Maintenance of antimicrobial concentrations above the MPC for > 50% of administration interval is a straightforward way to restrict the acquisition of resistance in this tissue cage model. This situation was achieved with daily intramuscular doses of 1 mg cefquinome/kg body weight.

Pharmacokinetics and ex vivo pharmacodynamics of cefquinome in porcine serum and tissue cage fluids.

A tissue cage (TC) model was used to evaluate the pharmacokinetics and ex vivo pharmacodynamics of cefquinome after intravenous (IV) and intramuscular (IM) administration to piglets at 2 mg/kg bodyweight. The mean values of area under the concentration-time curve (AUC) were 21.28 (IV) and 21.37 (IM) µg h/mL for serum, and 17.40 (IV) and 16.57 (IM) µg h/mL for TC fluid (TCF), respectively. Values of maximum concentration (C(max)) were 6.15 µg/mL (serum) and 1.15 µg/mL (TCF) after IM administration. The elimination half-lives (t(1/2ß)) in TCF (10.63 h IV and 11.81 h IM) were significantly higher than those in serum (2.33 h IV and 2.30 h IM) (P<0.05). The values of AUC(TCF)/AUC(serum) (%) after IV and IM administration were 82.4% and 80.7%, respectively. The ex vivo time-kill curves were established for serum and TCF samples using Escherichia coli ATCC 25922. The minimum inhibitory concentration (MIC) and minimum bactericidal concentration values of cefquinome against E. coli were 0.030 and 0.060 µg/mL in Mueller-Hinton broth, and 0.032 and 0.064 µg/mL in both serum and TCF, respectively. The ex vivo growth inhibition data of TCF after IM administration were fitted to the sigmoid E(max) model; AUC(24h)/MIC was 35.01 h for bactericidal activity and 44.28 h for virtual eradication, respectively. The findings from this study suggest that cefquinome may be therapeutically effective in diseases of pigs caused by E. coli when used at a dose rate of 1.33 mg/kg administered every 24 h for organisms with MIC90⩽0.50 µg/mL.

Influence of oxytetracycline on carprofen pharmacodynamics and pharmacokinetics in calves.

A tissue cage model of inflammation in calves was used to determine the pharmacokinetic and pharmacodynamic properties of individual carprofen enantiomers, following the administration of the racemate. RS(±) carprofen was administered subcutaneously both alone and in combination with intramuscularly administered oxytetracycline in a four-period crossover study. Oxytetracycline did not influence the pharmacokinetics of R(-) and S(+) carprofen enantiomers, except for a lower maximum concentration (Cmax ) of S(+) carprofen in serum after co-administration with oxytetracycline. S(+) enantiomer means for area under the serum concentration-time curve (AUC0-96 h were 136.9 and 128.3 µg·h/mL and means for the terminal half-life (T(1/2) k10 ) were = 12.9 and 17.3 h for carprofen alone and in combination with oxytetracycline, respectively. S(+) carprofen AUC0-96 h in both carprofen treatments and T(1/2) k10 for carprofen alone were lower (P < 0.05) than R(-) carprofen values, indicating a small degree of enantioselectivity in the disposition of the enantiomers. Carprofen inhibition of serum thromboxane B2 ex vivo was small and significant only at a few sampling times, whereas in vivo exudate prostaglandin (PG)E2 synthesis inhibition was greater and achieved overall significance between 36 and 72 h (P < 0.05). Inhibition of PGE2 correlated with mean time to achieve maximum concentrations in exudate of 54 and 42 h for both carprofen treatments for R(-) and S(+) enantiomers, respectively. Carprofen reduction of zymosan-induced intradermal swelling was not statistically significant. These data provide a basis for the rational use of carprofen with oxytetracycline in calves and indicate that no alteration to carprofen dosage is required when the drugs are co-administered.

Development and validation of a tissue cage model of acute inflammation in the cat.

Four cylindrical silicon tissue cages (TC, internal volume: 6.7 ± 0.11 cm(3)) were inserted subcutaneously in 29 young healthy cats. A mild inflammatory reaction was induced by intracaveal injection of 1 mL of a 2%λ-carrageenan solution. TC exudate was subsequently sampled at predetermined times (up to 120 h) to measure exudate leucocyte counts and the concentrations of protein and eicosanoids. TC remained in situ for 9-10 months and were well tolerated. Leucocyte counts peaked at 34 h (50.1 ± 57.6 × 10(3) cells/mm(3) ) and returned towards baseline after 72 h. Protein concentration increased from 26.2 ± 2.7 g/L to a peak of 35.9 ± 6.0 g/L at 12 h before returning to baseline at 48 h. Exudate prostaglandin (PG)E(2) concentration peaked at 24 h (11.7 ± 13.7 ng/mL) and returned to baseline by 120 h. Repeated collection of fluid from noninjected cages did not increase transudate PGE(2). Ketoprofen (2 mg/kg, subcutaneously) suppressed exudate PGE(2) at 24 h. The carrageenan-stimulated TC model is an ethical and novel means of investigating soft tissue inflammation in the cat, in which exudate PGE(2) acts as surrogate marker of cyclooxygenase-2 activity. This model will facilitate the investigation of in vivo pharmacokinetics and pharmacodynamics of anti-inflammatory drugs in this species.

A moderate aseptic local inflammation does not induce a significant systemic inflammatory response.

The aim of this prospective study was to characterize the response of the coagulation system to a defined sterile localized inflammatory process. Tissue cages were implanted subcutaneously in five healthy Beagles. After 9-10 weeks, local inflammation was induced by an injection of 0.5 ml 1% carrageenan. Serial samples of tissue cage fluid (TCF) and blood were collected at 10 time points (0-168 h). Nucleated cells (NC) of TCF were counted automatically to characterize local inflammation. C-reactive protein (CRP), leukocytes and coagulation variables (PT, aPTT, fibrinogen, factor VIII, antithrombin, protein C, protein S, and d-dimers) were determined in blood samples. Carrageenan induced a significant 32-fold increase of NCs in TCF (P<0.0001). A slight increase in leukocytes (P<0.0001) was observed. There was a significant 1.3- to 1.5-fold increase in protein C (P=0.0001) and protein S (P=0.0028). CRP, secondary hemostasis and fibrinolysis did not change. The mild increase from baseline in PC/PS, may reflect a physiological counter reaction.

Feasibility of the Ussing chamber technique for the determination of in vitro jejunal permeability of passively absorbed compounds in different animal species.

The aim of this study was to assess the feasibility of the Ussing chamber technique for the determination of the jejunal permeability of passively absorbed, high permeability model compounds (acetaminophen and ketoprofen) in different animal species. Additionally, electrophysiological measurements and histological examination of pre- and post-incubation tissue specimens were performed. Apparent permeability coefficients of turkey and dog jejunum were low and highly variable due to tissue fragility caused by differences in thickness of the remaining intestinal layers after stripping and resulting in severe damage. Pig and horse jejunum were markedly more suitable for permeability determinations and mild signs of deterioration were noticed after 120 min of incubation. Transepithelial electrical resistance and potential difference did not correlate well with the observed tissue damage. From these data, the Ussing chamber technique appears to allow for permeability measurements within a species, but seems unsuitable for interspecies permeability comparison. However, further validation of the method with low permeability compounds and actively transported compounds is needed.

Distribution of ceftiofur into Mannheimia haemolytica-infected tissue chambers and lung after subcutaneous administration of ceftiofur crystalline free acid sterile suspension.

The objective of this study was to evaluate the penetration of ceftiofur- and desfuroylceftiofur-related metabolites (DCA) into sterile and infected tissue chambers, lung tissue and disposition of DCA in plasma across four different sacrifice days postdosing. Twelve healthy calves were utilized following implantation with tissue chambers in the paralumbar fossa. Tissue chambers in each calf were randomly inoculated with either Mannheimia haemolytica or sterile PBS. All calves were dosed with ceftiofur crystalline free acid sterile suspension (CCFA-SS) subcutaneously in the ear pinna. Calves were randomly assigned to 4 groups of 3 to be sacrificed on days 3, 5, 7 and 9 postdosing. Prior to euthanasia, plasma and tissue chamber fluid were collected, and immediately following euthanasia, lung tissue samples were obtained from four different anatomical sites DCA concentration analysis. Results of our study found that, in general, DCA concentrations followed a rank order of plasma > infected tissue chamber fluid > noninfected tissue chamber fluid > lung tissue. Data also indicated DCA concentrations remained above the therapeutic threshold of 0.2 microg/mL for plasma and chamber fluid and 0.2 microg/g for lung tissue for at least 7 days post-treatment.

Investigation of in vitro transdermal absorption of fentanyl from patches placed on skin samples obtained from various anatomic regions of dogs.

OBJECTIVE: To investigate in vitro transdermal absorption of fentanyl from patches through skin samples obtained from various anatomic regions of dogs. SAMPLE POPULATION: Skin samples from 5 Greyhounds. PROCEDURE: Skin samples from the dogs' thoracic, neck, and groin regions were collected postmortem and frozen. After samples were thawed, circular sections were cut and placed in Franz-type diffusion cells in a water bath (32 degrees C). A commercial fentanyl patch, attached to an acetate strip with a circular hole, was applied to each skin sample. Cellulose strips were used as control membranes. Samples of receptor fluid in the diffusion cells were collected at intervals for 48 hours, and fentanyl concentrations were analyzed by use of high-performance liquid chromatography. RESULTS: Mean+/-SD release rate of fentanyl from the patch, defined by its absorption rate through the non-rate-limiting cellulose membrane, was linear during the first 8 hours (2.01+/-0.05 microg/cm2 of cellulose membrane/h) and then decreased. Fentanyl passed through skin from the groin region at a faster rate and with a significantly shorter lag time, compared with findings in neck or thoracic skin samples. CONCLUSIONS AND CLINICAL RELEVANCE: In vitro, fentanyl from a patch was absorbed more quickly and to a greater extent through skin collected from the groin region of dogs, compared with skin samples from the thoracic and neck regions. Placement of fentanyl patches in the groin region of dogs may decrease the lag time to achieve analgesia perioperatively; however, in vivo studies are necessary to confirm these findings.

Differences in polymorphonucleocyte function and local inflammatory response between horses and ponies.

REASONS FOR PERFORMING STUDY: Wound healing proceeds faster in ponies than in horses and complications during healing, such as wound infection, occur less frequently in ponies. Earlier studies suggested that this difference might be related to differences in the initial post traumatic inflammatory response. HYPOTHESIS: That polymorphonuclear leucocyte (PMN) function and profiles of humoral factors in local inflammatory processes are different in horses and ponies. METHODS: PMNs were isolated from venous blood of horses and ponies. Chemotaxis and reactive oxygen species (ROS) production was determined. Tissue cages were implanted in limbs and necks of horses and ponies and injected with carrageenan and, 3 weeks later, with LPS. In sequential samples of inflammatory exudate, the numbers of macrophages and PMNs and the production of PGE2, TNFalpha, IL-1, IL-6 and chemoattractants were determined. RESULTS: In vitro ROS production of PMNs was significantly higher in ponies than in horses, whereas in vitro PMN chemotaxis was significantly lower in ponies. In the tissue cages for both stimuli, the production of IL-1 and chemoattractants was significantly higher in ponies than in horses and remained so towards the end of the observation period in ponies. CONCLUSIONS: This study demonstrated a higher production of various inflammatory mediators by pony leucocytes. Despite the lower in vitro chemotaxis of pony PMNs, this higher in vivo production resulted in a stronger initial inflammatory response in ponies, as has been reported in studies on wound healing, through the attraction of leucocytes and triggering of the production of other cytokines. A stronger initial inflammation may promote healing by more rapid elemination of contaminants and earlier transition to repair. POTENTIAL RELEVANCE: Modulation of the initial inflammatory response might therefore be a valid option for therapeutic intervention in cases of problematic wound healing. Further, the intraspecies differences in leucocyte function may have an impact on many fields in equine medicine.

Tissue chamber model of acute inflammation in farm animal species.

A tissue chamber model of acute inflammation for use in comparative studies in calves, sheep, goats and pigs has been established and validated. Tissue chambers were prepared from silicon rubber tubing, of inner diameter 12.7 mm, length 115 mm and volume 15 ml, with 10 holes, each of 6mm diameter, at each end. In each animal two or four chambers were inserted at subcutaneous sites. Six weeks after implantation an acute inflammatory reaction in a single cage was generated by the intracaveal injection of 0.5 ml of 1% carrageenan solution. Serial samples of exudate (injected chamber), transudate (non-injected chamber) and blood were collected for measurement of exudate and transudate leucocyte count, prostaglandin (PG)E(2) concentration in exudate and serum thromboxane (Tx)B(2) concentration. In addition, skin temperature changes over exudate and transudate chambers were recorded. In all four species, carrageenan induced an acute inflammatory response, indicated by increases to peak values followed by return towards baseline in skin temperature, leucocyte count and PGE(2) concentration. For each of these variables in calves, sheep and goats the increases were significantly greater for exudate than for transudate. The degree of intra-species variation in each variable was acceptable. Marked inter-species differences were recorded: skin temperature rise was greatest in calves and least in sheep and goats; exudate PGE(2) concentration was increased in the order sheep>goat>pig>calf; serum TxB(2) concentration was increased in the order calf>goat>sheep>pig and exudate leucocyte count was increased to a greater extent in the pig than in the three ruminant species. The model has advantages over some previously described tissue chamber models of inflammation and will be suitable for use in comparative studies of inflammatory mechanisms and the pharmacokinetics and pharmacodynamics of anti-inflammatory drugs.

In vivo effect of Pasteurella haemolytica infection on bovine neutrophil morphology.

OBJECTIVE: To determine whether characteristic changes in neutrophil morphology caused in vitro by Pasteurella haemolytica leukotoxin (LKT) can be observed in vivo by electron microscopic examination of infected tissue chamber fluids and pneumonic lungs. ANIMALS: 7 mixed-breed beef calves. PROCEDURE: Tissue chambers were implanted subcutaneously in 3 calves and were inoculated with P haemolytica or phosphate-buffered saline solution. Chamber fluid samples, obtained at 8 and 32 hours after inoculation, were examined, using electron microscopy. Experimental pneumonia was induced in an additional 4 calves by transthoracic inoculation with P haemolytica. These calves were euthanatized at 6, 12, 24, and 36 hours after inoculation and lung sections were examined, using transmission electron microscopy. RESULTS: On examination, using transmission electron microscopy, neutrophils in lung sections and tissue chamber fluids had cytoplasmic and nuclear changes indicative of irreversible cell injury, including cell swelling, loss of plasma membrane ruffling, mitochondrial swelling, autolytic vacuolation, disruption of plasma membrane, nuclear pyknosis, karyolysis, and karyorrhexis. On examination, using scanning electron microscopy, leukocytes obtained from tissue chambers did not have their typical convoluted surfaces, but appeared rounded and swollen or shrunken with pitted surfaces. CONCLUSIONS: Pasteurella haemolytica-induced changes in neutrophil morphology in vivo were similar to those previously induced by in vitro exposure of neutrophils to LKT. Changes were suggestive of injury initiated by damage to the plasma membrane, which is consistent with the mechanism of action of pore-forming cytolysins. CLINICAL RELEVANCE: Pasteurella haemolytica LKT appears to be an important virulence factor in vivo; a fact that should be addressed in the development of vaccines.

Tissue chambers--a useful model for in vivo studies of cytokine production in the pig.

An in vivo tissue chamber model was developed to enable studies of local cytokine production and cellular events during inflammatory and immune reactions in the pig. Tissue chambers made of sialistic rubber tubing were surgically implanted in the subcutaneous tissue- and samples of tissue chamber fluid (TCF) and inflammatory cells were collected by aspiration with a syringe. To evaluate the model for local cytokine production, two cytokine inducers, polyribinosinic-polyribocytidylic acid (poly I:C) and fixed Aujeszky's disease virus infected PK15 cells (ADV-PK15), were injected into the tissue chambers and samples of TCF were collected 0, 4, 8, 12, 24 and 48 h post injection. Poly I:C injections induced local production of interferon-alpha (IFN-alpha) as well as tumor necrosis factor (TNF) in the TCF but kinetic differences in the production of the cytokines were noted. Poly I:C also induced an increase in cell numbers in the TCF, mainly due to increased neutrophil numbers. Injections of ADV-PK15 induced local IFN-alpha production in the TCF as long as the pigs were serologically negative to ADV. Immunofluorescence and in situ hybridization techniques could be applied for characterization of TCF cells. Moreover, cells recovered from the tissue chambers were viable and could be used in functional in vitro tests. Taken together, this tissue chamber model could prove very useful in in vivo studies of inflammatory/immune responses and cytokine production in the pig.

[The tissue cage in dogs--a pharmacologic model for the representation of plasma and tissue kinetics]. / Die Gewebekammer ("tissue cage") beim Hund--ein pharmakologisches Modell zur Darstellung von Plasma- und Gewebekinetik.

Determination of pharmacokinetic data from plasma often is not sufficient in order to predict or to explain the clinical efficacy of drugs, e.g., in non steroidal antiinflammatory drugs (NSAID) with short elimination half lives in plasma a discrepancy appears between elimination half life and clinical effect. We have studied the tissue kinetics of a drug having a long elimination half life (naproxen) in comparison with phenylbutazone and flunixin, having a short elimination half life in dogs after oral application. Tissue fluid was obtained by using so called "tissue cages" implanted laterally in the neck. An inflammatory response was produced by injecting 1-2 ml of 2% carrageenan into the chamber. The drug concentrations were determined throughout the experiments both in plasma and tissue fluid. Naproxen concentrations (elimination half life of about 72 h in dogs) achieved equilibrium between tissue fluid and plasma concentration. In the case of phenylbutazone, concentration in tissue fluid declined with a definitely longer half life than in plasma and exceeded plasma concentrations after some hours. Flunixin concentrations in plasma and exudate attained equilibrium after 2-5 h. Of clinical importance is the finding that after this time flunixin accumulated in inflammatory exudate and the concentration in exudate exceeded that in plasma two to ten times, then declined with an almost identical half-life. The number of leukocytes in exudate as a measure of inflammation was reduced during 2 days after phenylbutazone, 3-4 days after flunixin and 5-7 days after naproxen. This allows therapy with a lower dosage than calculated on the basis of plasma concentrations in the case of phenylbutazone and flunixin, possibly also with other NSAIDs having a short plasma half life.

Characterization of a sterile soft-tissue inflammation model in thoroughbred horses.

This paper describes the use of subcutaneously-placed tissue chambers as a sterile soft-tissue inflammation model in Thoroughbred horses. Acute, non-immune inflammation was initiated by injecting a sterile lambda carrageenan solution into a tissue chamber. This model was used to study the temporal changes in oxygen and carbon dioxide tensions, pH, bicarbonate, protein, albumin, prostaglandin E2 (PGE2) and leukotriene B4 (LTB4) concentrations, cell counts and differential counts in tissue fluid from inflamed tissue chambers and control chambers. Skin temperatures over control and inflamed chambers were also compared. Carrageenan-induced inflammation resulted in significant increases in tissue-fluid carbon dioxide tension, leucocyte count, albumin, and PGE2 and LTB4 concentrations. It also resulted in a significant decrease in tissue fluid pH and HCO3-concentration. Inflammation did not result in significant changes in tissue-fluid protein concentration, differential cell counts or skin temperature over the chambers. The use of this type of tissue chamber is well-suited for studying the pathophysiology of a self-contained, non-immune inflammatory process. The model described in this paper could prove to be very useful in studies of the distribution of anti-inflammatory drugs and the effects of such drugs on various aspects of the inflammatory process.

Pharmacokinetics of ceftazidime in sheep and its penetration into tissue and peritoneal fluids.

Serum, tissue and peritoneal fluid concentrations of ceftazidime were studied in ewes after intravenous, intramuscular and subcutaneous administration at 50 mg kg-1 bodyweight. Tissue and peritoneal cages were implanted in the animals studied. After intravenous bolus administration, the mean serum concentration versus time profile was best described by a two-compartment open model. The distribution rate constant (alpha) was 3.5 +/- 1.1 h-1 and the half-life (t 1/2 alpha) 0.22 +/- 0.09 hour. The elimination rate constant (beta) was 0.43 +/- 0.04 h-1 and half-life (t 1/2 beta) 1.6 +/- 0.2 hours. The area under the curve was 275.7 +/- 84.0 micrograms.ml-1 h. The volume of distribution as steady state was 356.1 +/- 208.0 ml kg-1. The penetration ratio into tissue fluid was 62.6 +/- 15.1 per cent and into peritoneal fluid 61.1 +/- 16.5 per cent. After intramuscular injection, the elimination half-life was 1.7 +/- 0.2 hours, the area under the curve was 228.7 +/- 43.3 micrograms.ml-1 h. and the elimination rate constant was 0.42 +/- 0.05 h-1. The penetration ratio into tissue fluid was 68.5 +/- 37.3 per cent and into peritoneal fluid 73.3 +/- 34.4 per cent. After subcutaneous injection, the elimination half-life was 1.8 +/- 0.5 hours, the area under the curve was 231.8 +/- 65.6 micrograms.ml-1 h. and the elimination constant was 0.41 +/- 0.10 h-1. The penetration ratio into tissue fluid was 47.2 +/- 3.5 per cent and into peritoneal fluid 58.1 +/- 15.6 per cent.(ABSTRACT TRUNCATED AT 250 WORDS)