Immune potential in human uraemia. 1. Relationship of glomerular filtration rate to depression of immune potential.

Aspects of immune potential in uraemic subjects, categorized by glomerular filtration rate, were intercompared and compared with results obtained from a group of normal volunteers. Evidence is presented to show that depression of both cellular and humoral immune potential occurs with progressive reduction of glomerular filtration rate. Lymphocyte transformation testing to the non-specific mitogen PHA revealed a significant elevation of blastogenic response in uraemia after 96 hours of incubation.

Immune potential in human uraemia. 2. Changes after regular haemodialysis therapy.

Parameters of both humoral and cellular immune potential were measured in a group of patients with severe renal failure before and after three months' regular haemodialysis therapy. Evidence is presented of improvement in cellular immune potential and of a tendency of the response of lymphocytes to PHA to return to normal. No improvement in humoral responsiveness was demonstrable, and it is suggested that uraemic patients on regular haemodialysis may have an impaired capacity to establish new immunological memory.

Pharmacological evaluation of aerosolized cannabinoids in mice.

The reemergence on the debate of the use of marijuana for medicinal purposes has been the impetus for developing an acceptable delivery form of aerosolized cannabinoids. The goals of the present study were to: (1) develop and characterize the physical properties of an aerosolized form of Delta(9)-tetrahydrocannabinol (Delta(9)-THC), the major psychoactive constituent present in marijuana; and (2) assess the pharmacological effects of cannabinoid inhalation in mice. A Small Particle Aerosol Generator (SPAG) nebulizer, used to generate the aerosol, had an output of approximately 0.154 mg/l of aerosolized Delta(9)-THC with a 2.0 microm mass median aerodynamic diameter and a 2.2 geometric standard deviation (GSD). Virtually all the particles were less than 5.0 microm in diameter suggesting that they were sufficiently small to penetrate deeply into the lungs. Inhalation exposure to aerosolized Delta(9)-THC in mice elicited antinociceptive effects that were dependent on concentration and exposure time with an estimated Delta(9)-THC dose of 1.8 mg/kg. On the other hand, inhalation exposure to Delta(9)-THC failed to produce two other indices indicative of cannabinoid activity, hypothermia and decreases in spontaneous locomotor activity. The antinociceptive effects occurred within 5 min of exposure and lasted approximately 40 min in duration. The cannabinoid receptor antagonist N-(piperidin-1-yl)-5-(4-chlorophenyl)-1-(2, 4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide HCl (SR 141716A), but not naloxone, blocked these antinociceptive effects (AD(50)=0.09 mg/kg) indicating a cannabinoid receptor mechanism of action. Similarly, inhalation exposure to a water soluble cannabinoid analog, 3-(5'-cyano-1', 1'dimethylheptyl)-1-(4-N-morpholinobutyrloxy)-Delta(8)-te trahydrocann abinol (O-1057), produced antinociception that was blocked by SR 141716A. These results demonstrate that the development of an aerosolized form of cannabinoids for human medicinal use is feasible.

Prediction of drug residence times in regions of the human respiratory tract following aerosol inhalation.

A mathematical model was developed for predicting drug residence kinetics in various regions of the human respiratory tract (RT). The model allows for regional deposition of different dose fractions (following mouth inhalation of various particle sizes according to four popular breathing regimes). Predicted alveolar deposition was dependent on the mode of inhalation and breath-holding. Deposition in the ciliated airways, however, was largely unaffected by breath-holding and was at a maximum for aerodynamic diameters between 5-9 micron (slow inhalation) and 3-6 microns (fast inhalation). Selected mucociliary and absorption rate constants determined the durations (T) taken to deplete the initial deposition in a chosen lung region to a selected minimum dose fraction (MDF). Values of T for an MDF of 0.01 in the ciliated airways were dependent on aerosol size, mode of inhalation, and rate of dissolution. In the case of rapidly dissolving solutes, the maximum duration was short (1-2 h) and occurred at particle sizes and modes of inhalation which maximized deposition in the conducting airways. For less soluble particles, however, T in the same airways could approach 12 h due to a prolonged supply of particles from the alveolar regions. The optimal size distribution and the mode of inhalation for maximum duration differed substantially in each case. The model enables formulation of testable hypotheses relating to the extension of local drug residence in the RT following inhalation of therapeutic aerosols.

A novel dosing method for drug administration to the airways of the isolated perfused rat lung.

A novel method is described for the reproducible administration of known liquid quantities to the peripheral airways of the isolated perfused rat lung. The basis of the technique was to use a 25-microL metered dose of fluorocarbon propellant to expel liquid (as a coarse spray) from an intratracheal dosing cartridge into the airways, while simultaneously inflating the lungs with a fixed volume of gas. The methodology is illustrated by administration of 100-microL volumes of aqueous disodium fluorescein solutions to a series of lung preparations. The reproducibility and regional distribution of dosing were determined by dissection, homogenization, and fluorimetric assay. Even though the dye was distributed nonuniformly between the lung lobes, in a series of preparations, 65.9 +/- 4.8% of the recovered dose was still deposited in the lung periphery, the site from which absorption is believed to occur. The method will enable the study of airway-to-perfusate transfer kinetics for compounds administered in a variety of different liquid formulations.

Spinning-disk generation and drying of monodisperse solid aerosols with output concentrations sufficient for single-breath inhalation studies.

The air-driven spinning-disk aerosol generator was modified to allow the production of monodisperse dry spherical aerosols of disodium fluorescein (as model solute) in high output concentrations. Output concentrations were determined by filtration. Optical and aerodynamic size distributions were determined microscopically (after electrostatic precipitation) and by cascade impaction. The generator housing allowed the entrainment of 25-microns primary aqueous solution droplets in a 10-L X min-1 downward flow of dry, filtered air. Internal equipment surfaces were machined flush and polished to minimize aerosol losses. Primary droplets were dried within a stainless steel pipe encased in a tube furnace. Water vapor was removed by diffusion drying. Disk-driven air, satellite droplets, and additional dilution air were vented to waste without using a vacuum. Generator yields were increased by reducing the size of the satellite droplet extraction gap. Aerosols were generated reproducibly by delivering aqueous solutions at a rate of 0.2 mL X min-1 to the center of the disk and spinning at 1000 rps. Dry aerosols, with mass median aerodynamic diameters of 2, 4.9, and 9 microns, were produced in concentrations of 0.89, 5.48, and 54.6 micrograms X L-1 from aqueous solutions containing 0.0374, 0.584, and 3.4% solute by weight. Geometric standard deviations were less than 1.2 in all cases. Concentrations are several times higher than others in the literature and are suitable for single-breath inhalation studies of therapeutic aerosol deposition and effect.

Effects of heat treatment on the permeability of polyvinyl alcohol films to a hydrophilic solute.

Polyvinyl alcohol (PVA) films may be useful as release-controlling membrane systems. Untreated, they are readily permeable to water and hydrophilic drugs. Because heating has been used to increase crystallinity and thus reduce the solubility and swelling in water of PVA films, we have studied the effects of heat on the permeability of PVA films to a water-soluble drug marker. Heat treatment was varied in the temperature range 100-200 degrees C for 1 h. The effect of time of heating was studied at 100 degrees C for 0.5-160 h. After pre-equilibration with water (heat-treated membranes remained intact, untreated ones dissolved), membrane permeabilities to methylene blue in aqueous solution (37 degrees C) were determined in a rotating diffusion cell. Permeabilities decreased with increased heating times (0.98-0.039 cm X min-1 for 0.5-160 h at 100 degrees C, respectively). Heating in air or N2 produced similar results. Further dramatic decreases in permeability occurred with increasing pretreatment temperatures; membrane permeability fell by a factor of approximately 500 with increasing temperature in the range 100-200 degrees C. There was no evidence of decomposition at temperatures less than or equal to 190 degrees C for 1 h. Results were consistent with literature reports of heat-induced increases in crystallinity. Membranes were simple to prepare and permeability could be controlled without recourse to chemical manipulation.

Drug absorption from inhalation aerosols administered by positive-pressure ventilation. I: Administration of a characterized, solid disodium fluorescein aerosol under a controlled respiratory regime to the beagle dog.

An apparatus and novel method is described for administration of well-characterized inhalation aerosols, under strictly controlled respiratory regimes, direct to the respiratory tract (RT) of the beagle dog by positive-pressure ventilation. The method enables the study of systemic absorption kinetics of compounds delivered as inhalation aerosols as a function of the aerosol particle size and respiratory variables provided their intrinsic pharmacokinetics are linear. Aerosol characteristics are determined by sampling the aerosol at a point close to its entry to the endotracheally intubated animal. The chosen positive-pressure ventilatory regime, which is monitored as airway pressure and exhaled volume versus time, can be held constant for the aerosol administration period. The methodology is illustrated by administration of a solid polydispersed aerosol of disodium fluorescein. Resultant plasma concentrations (C) were determined as a function of time by sampling from an indwelling venous cannula. The pharmacokinetic analysis of resultant C versus time data, together with that from an intravenous control experiment, is described to determine the amount absorbed as a function of time. Following aerosol administration according to the chosen respiratory regime, fluorescein was rapidly absorbed from the RT. The methodology will enable systematic variation of the particle size and positive-pressure respiratory regime in order to determine effects on drug absorption kinetics.

Drug absorption from inhalation aerosols administered by positive-pressure ventilation. II: Effect of disodium fluorescein aerosol particle size on fluorescein absorption kinetics in the beagle dog respiratory tract.

Solid, polydispersed disodium fluorescein aerosols (MMDae = 1.1, 3.5, and 4.4 micron) were administered under the same respiratory regime, direct to the respiratory tracts of two beagle dogs by positive-pressure ventilation. Subsequent to aerosol administration, plasma fluorescein concentrations were determined after sampling from an indwelling cannula. The amount absorbed as a function of time was estimated from these and additional data collected from intravenous control experiments in the same animals. Fluorescein absorption from the respiratory tract was apparently a first-order process, the rate increasing directly with the bioavailable dose. First-order rate constants differed but appeared unrelated to aerosol particle size, possibly reflecting similarities in their regional deposition in the canine lung. The average value for the absorption half-lives in the dogs were 19.3 and 12.2 min, showing that even lipophobic solutes such as the fluorescein dianion, are absorbed extremely rapidly via the lung. In one dog, the rate constant for fluorescein absorption after intratracheal instillation of a solution of the disodium salt was within the range of those following aerosol administration. Possible explanations are discussed.

Preparation, characterization, and controlled release from coprecipitates of fluorescein and magnesium hydroxide.

Magnesium hydroxide was precipitated as a lyophobic sol in the presence of various concentrations of fluorescein sodium (3'6'-dihydroxyspiro[isobenzofuran-1(3H),9'-[9H]xanthen++ +]-3-one, disodium salt) ranging in molar equivalents between 0.1 and 2 times that of the hydroxide. Coprecipitates were washed and dried, and release of the dye and magnesium was determined (pH 7.4, 37 degrees C) from rotating disks. Release rates varied depending upon fluorescein content. The rate of dye release was retarded by less than or equal to 10(4) times that of fluorescein sodium alone, implying the existence of some form of solid association between the components of the coprecipitates. The presence of the dye in certain concentrations reduced magnesium hydroxide dissolution rates by a factor of three. Fluorescein dissolution rates, when expressed as percent release, passed through a minimum (coincident with the dye-induced reduction in Mg(OH)2 dissolution). Adsorption experiments revealed evidence for multiaffinity binding of fluorescein at the surface of freshly precipitated Mg(OH)2. Magnesium, fluorescein, and water contents of the coprecipitates were characterized by atomic absorption and UV spectroscopy and by thermogravimetric analysis. Fluorescein content increased in direct proportion to its initial concentration in solution. Controlled, but variable release of this easily assayed dye is possible by employing precipitates with different fluorescein contents.

Critical analysis of "flip-flop" phenomenon in two-compartment pharmacokinetic model.

Computer simulations were used to examine the effect of first-order absorption on the disposition of one- and two-compartment model drugs. Two-compartment systems that attain a clinically acceptable beta-phase after rapid intravenous injection were perturbed by introduction of drug via first-order absorption. The validity of perceiving such a system as a potential "flip-flop" model was tested by comparing the negative slopes of log-linear plasma-time profiles to known values for ka and beta for various values of ka, k12, k21, and k10. Although most log-linear plots showed excellent correlation coefficients (r2 greater than 0.996), their negative slopes (S) did not represent either ka or beta under various combinations. A similar consideration of the one-compartment model enabled a comparison to be made between the two systems. Maximum negative errors were observed for both one- and two-compartment drugs as ka leads to k2 or beta, respectively. The value for S provided a good estimate of the absorption rate constant, ka, when k2 greater than or equal 2ka (one compartment) or beta greater than or equal 2ka. The elimination rate constant (k2 or beta) could be obtained from S for all one-compartment and some two-compartment drugs when the value of ka was approximately twice that of k2 or beta. Large positive errors also were observed with certain two-compartment drugs where the ratio of the four rate constants apparently linearized a nonlinear plasma profile. Conditions wherein S may be expected to approach beta wherein S approaches ka are clearly defined.

Calculation of partition coefficient of an unstable compound using kinetic methods.

A stirred transfer cell containing equal volumes of light liquid paraffin and an aqueous phase at 37 degrees was used to demonstrate the feasibility of calculating the partition coefficient of an unstable compound by kinetic analysis. Cyclohept-2-enone was chosen since it is a neutral molecule and, therefore, should have a pH-independent oil-water partition coefficient, KD. Moreover, this cyclic alpha, beta-unsaturated ketone undergoes hydrogen-ion-catalyzed hydration but is sufficiently stable at neutral pH to determine KD. The model system chosen represents first-order transfer between the aqueous (C1) and organic (C2) phases with simultaneous, reversible, first-order hydration. The transfer constants k'12 and k'21, were determined at 37 degrees in the absence of degradation where asymptotic values for C1 agreed with the observed equilibrium values in nonkinetic partitioning studies. The first-order rate constants for hydration in 0.1 N HCl were determined at 37 degrees in the absence of the organic phase. Partitioning with simultaneous hydration when was studied using 0.1 N HCl and light liquid paraffin. Data were analyzed by nonlinear regression based on the equation for C1 as a function of time. The values for k'12 and k'21 from these experiments were comparable to the estimates obtained under stable conditions. This agreement demonstrates that simultaneous degradation and partitioning can be analyzed for k'12 and k'21, thus permitting calculation of the partition coefficient (i.e., KD = k'12/k'21) that would be observed if the drug were stable.

Simultaneous partitioning and hydrolysis kinetics of amoxicillin and ampicillin.

The kinetics of ampicillin and amoxicillin partitioning with simultaneous acid-catalyzed hydrolysis were studied in a stirred transfer cell containing isobutanol as the extract and aqueous hydrochloric acid (0.1-0.5 N) as the raffinate at 37 degrees. Biexponential data for the concentration in both the raffinate (C1) and the extract (C2) as a function of time were analyzed simultaneously by nonlinear regression to estimate the apparent first-order rate constant for transfer from hydrochloric acid to isobutanol (k '12), the reverse transfer constant (k '21), and the hydrolysis rate constant (k). Agreement between k values determined in the presence of simultaneous partitioning and those determined in the absence of partitioning (k app) verified the nonlinear estimates. Apparent partition coefficients, which represent the values that would be obtained in the absence of hydrolysis K'D = C1 infinity/C2 infinity), were estimated from K'D = k'12/k'21. During terminal monoexponential loss, where C1 approximately equal to Y'e-beta t and C2 approximately equal to Z'e-beta t, the kinetically controlled C2/C1 ratio (r) is described by [K'12/K'21-beta)], which decreases with decreasing kappa values until r approaches K'D. The difference between the terminal concentration ratio, r, and its corresponding partition coefficient, K'D, is a measure of the degree to which kinetic processes control distribution. Both ampicillin and amoxicillin showed kinetic control of the distribution ratios in 0.5 N HCl, where the hydrolysis rate constant was significant relative to the distribution rate constants. Ampicillin had r approximately equal to 1.74 and K'D approximately equal to 0.92; amoxicillin had r approximately equal to 0.95 and K'D approximately equal to 0.65. As the (K'12 + K'21/k ratio increased, the r values approached K'D so that in 0.1 N HCl, r approximately K'D = 0.33 for amoxicillin and r approximately 0.6 and K'D approximately 0.56 for ampicillin. In general, amoxicillin distribution rate constants (K'12 + K'21) were roughly twice those of ampicillin, whereas ampicillin K'D and r values were nearly double those of amoxicillin. Thus, the kinetic and thermodynamic rank orders are opposite. This result may have implications in drug design via molecular modification.

Thermodynamic dependence of interfacial transfer kinetics of nonionized barbituric acid derivatives in two-phase transfer cell.

A theory was developed to describe interfacial transport kinetics of a series of drug homologs in a two-phase transfer cell. When tested, the theory held true for 5,5-disubstituted barbituric acid derivatives in a preequilibrated octan-1-ol = (pH 5) aqueous buffer system maintained at 37 degrees and stirred symmetrically at 50 and 100 rpm. Theoretical prediction of transfer kinetics was not possible in such a cell if the phases were stirred asymmetrically. For symmetric stirring, successful prediction of the transfer kinetics of any homolog in the series was possible from a knowledge of the partition coefficient and transfer kinetics of the parent compound, the partition coefficient of the homolog, and some easily determined system variables. The viscosity and density of the two phases and the phase-volume ratio were needed to define a system constant dependent on the solute diffusion coefficient, interfacial area, donor phase volume, and the boundary layer thickness for diffusion in the donor phase volume, and the boundary layer thickness for diffusion in the donor phase. A method is described to enable estimation of this constant from a knowledge of the transfer kinetics of the parent compound. The rank order of compounds in terms of their observed first-order transfer rate constants is shown to be dependent on the characteristics of the solvent system and stirring conditions employed, as well as on the physical chemistry of the solutes. The results are discussed in light of previously documented investigations.

An isolated perfused rat lung preparation for the study of aerosolized drug deposition and absorption.

The isolated, perfused rat lung preparation was modified to allow characterized solid aerosol delivery. Deposition and airway-to-perfusate transfer of disodium fluorescein from 3-4 micron dae solid aerosols were studied under different ventilatory regimes. The lungs inhaled from an aerosol stream of constant concentration via a tracheal cannula. Air displacement from a sealed artificial thorax housing the lungs provided the driving force for inhalation. The lungs were suspended in a physiologically normal position and both left and right sides of the heart were cannulated for constant rate perfusate flow. Fractional deposition was inversely proportional to respiratory frequency implying that sedimentation was the primary deposition mechanism. Increasing tidal volumes similarly enhanced the ratio of amount deposited/amount administered. Fluorescein transfer to the perfusate occurred from the lung regions containing intact vasculature, was apparent first-order, and independent of perfusate flow. The average rate constant for transfer was 0.057 +/- 0.02 min-1 (t1/2 = 12.2 +/- 4.2 min-1). The ratio of transferable amount/amount deposited appeared to indicate the depth of aerosol penetration. This increased at high respiratory frequency and tidal volume, while decreasing with increasing aerosol particle size. Potential applications of the model are discussed in the light of these results.

Preparation, characterization, and dissolution kinetics of two novel albuterol salts.

The possibility of producing slowly dissolving albuterol salts was investigated as a potential means of extending the duration of action of the drug following aerosol delivery to the lung. Albuterol adipate and stearate were precipitated from alcoholic solutions of albuterol and adipic or stearic acids, respectively. Differential scanning calorimetry and hot stage microscopy showed that albuterol adipate and stearate produced single melting endotherms at 182 and 116 degrees C, respectively, which were distinct from those of albuterol (158 degrees C), adipic acid (152 degrees C), and stearic acid (70 degrees C). The aqueous solubilities of albuterol free base, sulfate, adipate, and stearate were 15.7, 250, 353, and 0.6 mg . mL-1, respectively, at room temperature. Only the solubilities of the adipate and the stearate increased significantly when the temperature was elevated to 37 degrees C (452.5 and 1.4 mg . mL-1, respectively). With a rotating disk dissolution method, albuterol free base, sulfate, and adipate were found to have intrinsic dissolution rates of 1.1, 20.4, and 24.0 mg . min-1 . cm-2, respectively, in pH 7.4 phosphate buffer at 37 degrees C. Albuterol stearate dissolved much more slowly and in a nonlinear fashion; this was explained by the deposition of a stearate-rich layer on the dissolving surface of the compacted salt.

Drug-surfactant-propellant interactions in HFA-formulations.

The required replacement of chlorofluorocarbon (CFC) with hydrofluoroalkane (HFA) propellants has challenged formulators of pressurized metered dose inhalers in several major respects. Conventional (CFC soluble) surfactants are effectively insoluble in the major CFC replacement candidates, HFA 134 and HFA 227ea, in the absence of co-solvents. While these ethane and propane derivatives have comparable boiling points and vapor pressures to dichlorodifluoromethane (CFC 12), their increased polarity demands that formulators use either alternative (soluble) surfactants, or co-solvents along with traditional surfactants, in order to stabilize pressurized suspension products. The use of either approach is complicated by the existence of many competing patents and the fact that the science in the area is empirical; predictive theoretical approaches are frustrated by the lack of an adequate database. Technical developments in this area must also take into account the need to avoid crystal growth and/or adhesion of micronized, suspended drugs to internal container surfaces, problems which may be catalyzed by some combinations of surfactant type/concentration, vehicle(s) and physical form/type(s) of drug substance. For some drugs, it appears simpler to use co-solvents with HFA propellants to dissolve the drug, avoiding the need for suspension stabilization. This article presents an overview of the present state of the art with respect to the formulation of MDIs.

Effect of energy on propylene glycol aerosols using the capillary aerosol generator.

The CAG is being developed for pulmonary drug delivery. Liquids are pumped, heated and vaporized by the CAG, whence they nucleate and condense to form aerosols. This study characterized the effect of energy on the aerosolization process. With increasing energy, the CAG produced an increasing fine particle fraction (FPF) until "optimal aerosolization" was achieved between 40 and 45 J; this energy range agreed with that theoretically required to vaporize the dose of PG. Further increases in energy above this optimal range did not improve PG's aerosolization efficiency. Based on the energy, FPF and temperature profiles, it was possible to deduce the nature of the liquid flow-boiling during aerosol generation. The aerosol particle size went through a minimum, as energy was increased through the "optimal range." In the "energy excess" region, where additional energy increased PG vapor temperature and velocity, droplet sizes were increased primarily due to changes in the nucleation rates and supersaturation ratios affecting the nucleation and condensation processes occurring within the vapor jet. The in vitro MMAD of the PG aerosol changed as a function of the applied energy, suggesting that for any pharmaceutical application, the choice of applied energy is critical to deposition profile of the aerosol.

The quantitative determination of aspirin and its degradation products in a model solution aerosol.

Formulation of pressurized aerosol solutions in propellants for inhalation requires the use of high quantities of surfactants to solubilize the drug. Due to the lipophilic nature of these surfactants, analytical difficulties are created for those wishing to quantify the drug and its degradation products. In order to quantify drug and degradation products by LC it is necessary to separate surfactant and analytes prior to chromatography. To illustrate a typical situation, a method was developed for the analysis of acetylsalicyclic acid (approximately 2.5 x 10(-3) M) and its major degradation products (salicylic acid, acetylsalicylsalicylic acid and salicylsalicylic acid) solubilized in trichloromonofluoromethane (CFC-11) containing 10(-2) M sorbitan trioleate (Span 85). Surfactant extraction problems were reviewed experimentally. The presentation of all analytes and the surfactant, dissolved in hexane, to silica solid phase extraction columns, followed by elution in a polar solvent, was found to be an efficient way of separating this lipophilic surfactant from the analytes. The final assay employed propellant evaporation, reconstitution of the non-volatiles in hexane, normal phase solid phase extraction (recoveries of 100 +/- 10% were observed for all analytes), elution and dilution with mobile phase, and reversed-phase liquid chromatography (Econosphere C8 5 microns, 4.6 x 250 mm). The assay utilized a mobile phase of water, methanol, tetrahydrofuran and 1 M phosphoric acid with ultraviolet detection at 275 nm. Using external standards, linear calibration curves of peak height versus concentration were obtained for all analytes in the expected concentration ranges (r > 0.991). As it is described, the assay had a relative standard deviation of < or = 3.7% for all analytes.

High-pressure liquid chromatographic assay for the determination of thyrotropin-releasing hormone and its common metabolites in a physiological salt solution circulated through the isolated perfused rat lung.

A specific and sensitive assay for TRH (L-pyroglutamyl-L-histidyl-L-proline amide), 3H-TRH (L-proline 3,4-3H(N), histidyl-3-3H(N)), and four possible metabolites of TRH, present in the recirculated perfusate of an isolated perfused rat lung preparation, was developed. Unlike previous methods, the method developed does not require extraction of the analytes from the biological matrix. The crude sample was adjusted to a pH of 3.2 with concentrated trifluoroacetic acid and injected on to a PRP-1 (polystyrene divinylbenzene) column (10 microns, 25 cm x 4.6 mm i.d.). The mobile phase was 10% v/v acetonitrile and 90% v/v 0.75 g l-1 1-hepantanesulfonic acid in 0.004 M trifluoroacetic acid, adjusted to a pH of 2.4 with concentrated NaOH. The flow rate was 0.5 ml min-1 and the analytes were detected by UV absorption at a wavelength of 26 nm and by radiochemical detection utilizing a liquid scintillation counter. The nominal retention times for L-PRO, L-PRO-NH2, TRH, cyclo(HIS-PRO) and TRH-OH were 4.0 +/- 0.9, 10.0 +/- 0.2, 15.5 +/- 0.4, 19.2 +/- 0.5 and 25.3 +/- 0.5 min respectively. The assay performs well in terms of precision and accuracy as indicated by linear regression and intra-assay variability analysis.