| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *Department of Pharmaceutics & Anesthesiology, University of Utah, Salt Lake City, Utah; Department of Anesthesiology, University of California at San Diego, San Diego, California; Cognetix, Inc., Salt Lake City, Utah; and Department of Anesthesiology, Stanford University, Palo Alto, California.
Address correspondence and reprint requests to Steven E. Kern, PhD, Department of Pharmaceutics, University of Utah, 421 Wakara Way #318, Salt Lake City, Utah 84108. Address e-mail to Steven.Kern{at}hsc.utah.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: For the bolus dose study, eight animals received a dose ranging from 16.7 to 1000 nmol under isoflurane anesthesia. Cerebral spinal fluid sampling for drug assay occurred up to 24 h. For the multiple day infusion study, three animals received infusions of 10, 40, and 160 µg/h respectively for 24 h at each rate. Cerebral spinal fluid sampling occurred during the infusion rate and the washout period after the 72 h of cumulative drug delivery. Data from the two study designs were modeled separately using NONMEM.
RESULTS: The results showed a biexponential disposition profile for both experiments with a rapid rate constant that was an order of magnitude greater than the slow rate constant. The bolus results showed a nonlinear dependence of the slow rate constant on administered dose due to the large bolus range used in the study.
CONCLUSION: These data, coupled with clinical pharmacology results, provide a basis for determining appropriate dosing strategies to achieve therapeutic intrathecal concentrations of Contulakin-G.
|
Introduction |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
An important issue in understanding the intrathecal pharmacokinetics of Contulakin-G relates to its physicochemical nature. Unlike most intrathecal agents used in clinical practice, the conopeptides are large charged molecules. These features will have a profound impact on the drug’s distribution after injection. Further, because Contulakin-G is a peptide, its elimination from the cerebral spinal fluid (CSF) may depend, at least in part, on the presence of peptidases and its limited ability to cross biological barriers.
The preclinical efficacy of Contulakin-G has led to an effort to develop it for human use. This report summarizes the intrathecal pharmacokinetics of Contulakin-G after intrathecal administration in beagles. Two experimental designs were conducted (bolus dosing and infusion dosing) to assess the acute and steady-state pharmacokinetics of this compound in the intrathecal space. Our intent was to characterize the intrathecal pharmacokinetics of Contulakin-G to permit the design of optimal dosing strategies during clinical development.
|
METHODS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Acute Bolus Study
Eight beagles (4 male/4 female) weighing approximately 15 kg each were purchased from a commercial supplier (Marshall Farms, North Rose, NY). The beagles were allowed a minimum of 5 days of environmental conditioning before the study. Animals were sedated with an IM injection of a combination of tiletamine hydrochloride 50 mg/mL/zolazepam hydrochloride 50 mg/mL (Telazol 1 mg) before induction of anesthesia. Anesthesia was maintained using 1% isoflurane in 100% oxygen administered via endotracheal tube Heart rate, arterial blood pressure, and oxygen saturation were continuously monitored during the anesthetic.
The intrathecal catheter insertion site was shaved and cleaned with betadine scrub. A spinal needle was inserted and an intrathecal catheter (Braun 24 ga., 0.1 mL priming volume, Spinocath, Bethlehem, PA) was passed through the needle in a cephalad direction until the tip resided at the L1–3 region. Placement was confirmed by measurement of insertion depth. A single 0.25 mL sample of CSF was withdrawn to confirm catheter patency. The catheter was positioned and sutured to the skin to prevent movement during the study.
A single bolus of Contulakin-G was administered to seven of the eight beagles (3F/4M). In the undosed beagle, catheter patency could not be achieved after placement and the animal was removed from the study before drug administration. Three different dose amounts were used, 1000, 50, and 16.7 nmol. Each was delivered in 1 mL of volume. Samples of CSF were made for a 24 h period from the catheter at the following timepoints relative to dose administration: 5, 10, 20, and 30 min, 1, 2, 4, 6, 8, 12, and 24 h. For each sample, a volume equal to the catheter dead-space was first removed over the course of approximately 30 s followed by the sample volume (125 µL). After the sample, a volume of saline equal to the catheter dead-space and the sample volume was injected into the catheter over approximately 30 s. This was attempted to keep the sampling from significantly altering the volume of the CSF throughout the experiment. All samples were frozen on ice and stored at –20°C until analysis by the study sponsor.
Each animal was kept under anesthesia for 4 h after study drug administration. After the 4 h time sample, the anesthesia was discontinued and the animal was allowed to recover. The remaining four intrathecal samples were made while the animal had recovered from anesthesia. After the 8 h sample, dogs were returned to the Animal Resource Center where the last two samples were drawn.
Steady-State Infusion Study
A nylon vest (Lomir, Montreal, Canada) was placed on each of three dogs 48 h before scheduled intrathecal catheter placement surgery for acclimation in anticipation of continuous infusion using external pumps. Surgical preparation of the dogs for an intrathecal catheter was done approximately 72 h before dosing. Antibiotic (sulfamethoxazole–trimethoprim 240 mg tablet, 15–25 mg/kg, oral, twice daily) was given 48 h before surgery and for 48 h postsurgery. The dogs received atropine (0.04 mg/kg, IM) before sedation with xylazine (RompumÆ, 1.5 mg/kg, IM). Anesthesia was induced with inhaled isoflurane by mask. The animals’ tracheas were intubated after induction, and anesthesia was maintained under spontaneous ventilation with 1.0%–2.0% isoflurane and 50% O2/50% N2O (approximate values). Animals were continuously monitored for oxygen saturation, inspired and end-tidal values of isoflurane, CO2, N2O, and oxygen. Heart and respiratory rates were also monitored continuously during anesthesia.
Surgical areas were shaved and prepared with chlorhexadine scrub and solution. Using sterile technique, a skin incision was made and the incision draped. The cisterna magna was exposed by combined blunt and sharp dissection. The dura was exposed and a small incision (1–2 mm) was made and a PE-10 intrathecal dosing catheter (fabricated on site of polyethylene tubing and sterilized by E-beam irradiation) was inserted and passed caudally a distance of approximately 42 cm to a level corresponding to the L2–3 segment. A second catheter (PE-50) used for CSF sampling was also implanted to a distance of approximately 40 cm terminating at the region of T13. Confirmation of the appropriate placement of the catheters in the intrathecal space was judged by the free outflow of CSF. The catheters were then tunneled subcutaneously and caudally to exit at the left (sampling) or right (dosing) scapular region. Dexamethasone sodium phosphate (0.25 mg/kg, IM) was administered just after catheter placement. The incision was closed in layers with 3-0 Vicryl suture.
The nylon vest was again placed on the animal (after having been previously acclimated to the vest before surgery) and two infusion pumps (PANOMAT C-10, Disetronic Medical Systems, Plymouth, MN) were secured in a vest pocket. The pumps were then connected to the externalized end of the dosing and sampling catheter. A loose fitting canvas vest was placed over the nylon vest to prevent damage to the pump as needed. An infusion of 0.9% (w/v) Sodium Chloride for Injection, USP (Abbott Labs.) was initiated at approximately 150 µL/h to ensure catheter patency prior to infusion of Contulakin-G.
Contulakin-G was infused over 72 h, using three escalating doses of 10, 40, and 160 µg/h, each given over 24 h. The infusion rate was fixed at 30 µL/h over the 72 h, and the dose was escalated by increasing the concentration of Contulakin-G in the infusate every 24 h (i.e., 0.33 mg/mL over the first 24 h, 1.33 mg/mL from 24 to 48 h, and 5.33 mg/mL from 48 to 72 h). The infusion was terminated at 72 h, and the animals were observed during the subsequent 24 h washout period. Lumbar CSF samples were obtained just before, and 24 h after, the initiation of each infusion rate of Contulakin-G. Sampling was also performed at 5, 15, 30, 60, 120, 240, 360, 480, and 1440 min after termination of 160 µg/h infusions to assess clearance kinetics.
To obtain CSF samples, approximately 200 µL of fluid was removed from the catheter to account for dead-space. Subsequently 250 µL of CSF was removed, placed in a 1.5 mL centrifuge tube, and frozen on dry-ice. The volume of CSF removed (250 µL) was replaced with saline and the catheter dead-space was again filled with 200 µL saline. CSF samples were stored at –20°C before transfer to the study sponsor for assay.
CGX-1160 assay development, validation, and sample analyses were performed under contract by Tandem Labs (Salt Lake City, UT) using an internally validated assay procedure under Good Laboratory Procedure conditions. In brief, calibration standards were prepared for CGX-1160 over a range from 10 to 1000 ng/mL. In each analytical run, blank serum samples, with and without internal standard, were assayed in duplicate. In addition, for each dilution level, dilution control samples were run in triplicate for analysis. An internal standard (CGX041.001.17, monoglycosylated derivative of CGX-1160, provided by Cognetix) was added to all CSF and serum samples at a final concentration of 500 ng/mL. For concentration measurements, CGX-1160 samples were prepared by direct dilution and analyzed by liquid chromatography/ tandem mass spectrometry (API 3000 LC/MS/MS, AME Biosciences, Bedfordshire, UK). Liquid chromatography analyses were performed using a Metasil AQ HPLC column, 0.1% formic acid in mobile phases, and an isocratic method. The mass spectrometer was operated in the selected reaction monitoring (SRM) mode under optimized conditions for detection of positive ions formed by turbospray ionization of CGX-1160 and the internal standard.
The samples, calibration standards, and quality control samples were injected into the LC/MS/MS systematically. Quality control samples that spanned the high, medium, and low range of quantitation were run in duplicate with each analysis. Assay quantitation was based on linear regression analysis of the calibration curves. The lower limit of detection for the assay was 10 ng/mL.
Mixed Effects Pharmacokinetic Modeling
The data from the bolus study and the infusion study were modeled separately because of differences in experimental design. Pharmacokinetic modeling was done using NONMEM (Globomax, Rockville, MD). The data were fit to a biexponential disposition function to assess goodness of fit to the model results. For the bolus study, the CSF concentration was initially fit to the equation:
|
|
where D = Dose and the coefficients A and B are in units of 1/volume. For the infusion study, the CSF concentration data was fit to the equation:
|
|
where r = Rate of infusion and the coefficients (e.g., A/
) are in units of time/volume. The variable tinf refers to the time that a given infusion is running and t equals time. During the infusion t = tinf but after the infusion is stopped, t > tinf.
Secondary parameter values of intrathecal space clearance were calculated from the primary model parameters using the following relationship:
|
|
The interindividual parameter variability was modeled according to an exponential equation.
|
|
where 
i is the population estimate for the parameter P, Pi is the individual parameter estimate and the 
term is a normally distributed random deviation of the individual parameter estimate from the population value. The values are assumed to be normally distributed with mean zero and variance 
2 Residual intraindividual error was characterized by an additive model after the concentration data was log-transformed.
|
|
where the value for 
is assumed to be normally distributed with zero mean and variance of 
2.
The model was implemented in NONMEM using the first order conditional estimation method with interaction to estimate the population , 2, and . Addition of model parameters was considered significant if the model objective score, a log-likelihood estimator, decreased by more than 3.84 with the addition of another model parameter.
|
RESULTS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
Infusion Study Modeling
The raw data for the infusion study are shown in Figure 3. The data were reasonably well described by a biexponential profile. The model was also evaluated with dose as a covariate on the terminal exponential component (β). This covariate did not improve the model results for the infusion profile (no change in objective function) and was therefore not included in the final model estimates. Since the CSF concentrations of Contulakin-G were an order of magnitude lower in the infusion study than that seen in the bolus study, this result is not surprising. The results of the model predictions for the three animals studied in the infusion profile are shown in Figure 4.
|
|
The pharmacokinetic parameters for the bolus and the infusion studies are given in Table 1. The clearance for both experimental conditions is estimated for the models as well. For the bolus study model, the clearance is estimated assuming the 50 nmol dose of Contulakin-G. The parameter variance is greatest for the β phase coefficient term, which is relatively small in magnitude. The small relative magnitude of this parameter implies that the CSF concentration decreases rapidly after a bolus dose or termination of an infusion before the more gradual terminal elimination phase begins. The amount of Contulakin-G measured in the serum of all animals was generally negligible. The concentration in the serum during both studies was at least 1000 times less than that measured in the CSF and was often less than the assay detection limit of 10 ng/mL.
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
-conopeptide MVIIA and has been used in clinical trials for treatment of refractory pain in patients with cancer or Acquired Immune Deficiency Syndrome (5). Peptides must be administered in a manner that allows access to their neurological receptor targets. Intrathecal administration is required to administer peptides to targets within the central nervous system. The present work defines the pharmacokinetics of Contulakin-G after intrathecal administration. The lumbar intrathecal space is a small, poorly mixed, fluid volume. This space does not display a consistent flow pattern. Accordingly, local bolus delivery of large molecules (e.g., inulin, BDNF, or ziconotide) frequently displays high initial concentrations, suggesting dilution into a small volume (7,8). In dogs, this space typically reflects a volume of approximately 3 mL. This initial decrease in concentration is followed by a long slow clearance of approximately 60–120 min. Small lipid soluble molecules typically show a relatively more rapid decline, suggesting a rapid movement from the CSF, typically through the meninges, or into the parenchyma (9). In these cases, significant blood levels are typically observed. CGX 1160, is a peptide of 2069 D MW. Given its size and peptide structure, it would be expected to behave as the other nonlipohilic large molecules.
The results from the bolus study in dogs reveal the presence of a nonlinearity with respect to terminal elimination rate constant. By incorporating the dose into the estimate of the β term, a statistically better estimate of the population parameters was obtained to fit the measured bolus dosing data. This nonlinearity was most evident after the largest dose, which was 20–60 times larger than the other doses given in the study. We do not know if this nonlinearity will have any significance in humans. It is important to recognize that this dose was only administered to one animal in the study, and that the value for the coefficient was very small, implying that the contribution of this nonlinearity would only show up if a large dosing range was given with this agent. Further, because of the desire to eliminate the influence of bolus volume on the CSF spread of the bolus doses used in this study, we chose to alter the concentration such that the largest bolus dose was hyperbaric compared to the lower doses (baricity approximately 1.0204 mg/mL). This may have contributed to this dose-dependent nonlinearity found in this study.
In contrast to the bolus study, the infusion study showed no significant improvement in the model fit by incorporating a nonlinearity in the estimate of the β term. The infusion study concentrations were in the range of those seen with the lower bolus doses, which likely accounts for the lack of evidence of nonlinearity. As shown in the accompanying article by Allen et al., analgesic efficacy occurs at concentrations found during the lower bolus dose and infusion studies (1). Thus the nonlinearity from the large bolus dose may not have any pharmacologic relevance, and since it occurred in only one animal, may not be repeatable.
Given that the intrathecal space is very small in the dog (approximately 2–3 mL), the normal assumptions that pertain to compartmental modeling regarding instantaneous and homogenous distribution of drug into the volume of distribution are likely to be unrealistic. For this reason, the data were not modeled using a compartmental approach. Further, we sought to minimize the influence of CSF sampling on the results by replacing the volume removed from CSF with saline after each sample. While this may maintain a relatively constant volume condition, it will also necessarily disturb the fluid distribution within the CSF. The influence this may have had on our results is unknown.
Wermeling et al. reported similar values for the terminal rate constant of the conopeptide ziconotide after intrathecal delivery (10). These investigators did not measure the initial concentrations rapidly enough to characterize the rate constant that describes the initial redistribution of the peptide from the intrathecal site of administration. The rapid decline of the initial intrathecal concentrations after the cessation of drug administration implies that the peptide either diffuses throughout the CSF or is bound to tissue in the intrathecal space, so that the concentration in the CSF is decreased.
It is important to note that unlike traditional compartmental pharmacokinetics, where drug is administered into a large well-mixed initial volume, intrathecal pharmacokinetics can vary greatly due to anatomical differences and experimental influences of sampling small volumes of fluid from a contained, small volume space. It is easy to imagine that the very act of sampling drug may affect its local distribution, and thus interpretation of intrathecal pharmacokinetic results should always be considered with these caveats in mind. The data show us that after administration in the intrathecal space, the concentration of drug at the injection site decreases in a biphasic manner with a residence time on the order of hours. Further clinical studies will be required to determine if this temporal profile matches any clinical effect that occurs with these agents.
The present study did not measure any pharmacodynamic end-points. Therefore, we are not able to identify whether the analgesic effect of Contulakin-G follows the intrathecal CSF concentrations directly, or whether there is hysteresis between the time course of CSF concentration and the time course of drug effect.
Measurement of serum concentration of Contulakin-G during this study after intrathecal administration revealed negligible concentrations of the peptide in the bloodstream. This implies that Contulakin-G may be metabolized within the CNS or, once it enters the circulation, it is rapidly bound by tissues or metabolized.
In summary, the pharmacokinetics of the peptide Contulakin-G can be described by a biexponential disposition function after both bolus and infusion administration. The kinetics show a rapid initial redistribution phase followed by a slow terminal elimination phase. The study did not characterize the time course of drug effect, nor the rate at which Contulakin-G diffuses to the target site in the spinal dorsal horn. Based on the likelihood that larger molecules diffuse slowly though neural tissue, it is likely measurements of drug effect will reveal considerable hysteresis between the time course of CSF concentration and the time course of drug effect. Future studies in human subjects will be used to determine clinically effective concentrations and to understand the relationship between dosing in humans and the results reported here.
|
Footnotes |
|---|
Supported by funding from Cognetix, Inc., Salt Lake City, Utah.
Drs. Kern, Shafer, and Yaksh have all served as consultants to Cognetix.
Drs. Shafer and Yaksh were recused from all editorial decisions related to this manuscript.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
