Development and validation of a sensitive solid-phase extraction and high-performance liquid chromatography assay for the novel antitumour agent CT2584 in human plasma

A HPLC assay and solid-phase extraction technique from human plasma has been developed and validated for the novel anticancer agent CT2584, 1-(11-dodecylamino-10-hydroxyundecyl)-3,7-dimethylxanthine, which has recently completed a phase I trial at the Christie Hospital, Manchester under the auspices of the CRC phase I/II committee. Following addition of CT2576, 1-(11-octylamino-10-hydroxylundecyl)-3,7-dimethylxanthine, as internal standard, a solid-phase extraction cartridge (100 mg cyanopropyl) was used to isolate the drug CT2584 from human plasma. Analysis was performed by reversed-phase chromatography. CT2576 was used as internal standard at a concentration of 4 μg ml⁻¹ for the quantification of CT2584 from plasma for the duration of this work. The lower limit of quantification for the drug CT2584 in buffer using this assay was found to be 0.0122 μM (0.008 μg ml⁻¹) and 0.048 μM (0.027 μg ml⁻¹) when extracted from human plasma.     

Quantitative determination of cefepime in plasma and vitreous fluid by high-performance liquid chromatography

An isocratic reversed-phase HPLC method was developed to determine cefepime levels in plasma and vitreous fluid. Cefepime and the internal standard cefadroxil were separated on a Shandon Hypersil BDS C18 column by using a mobile phase of 25 mM sodium dihydrogen phosphate monohydrate (pH 3) and methanol (87:13, v/v). Ultraviolet detection was carried out at 270 nm. The retention times were 4.80 min for cefepime and 7.70 min for cefadroxil. This fast procedure which involves an efficient protein precipitation step (addition of HClO₄), allows a quantification limit of 2.52 μg ml⁻¹ and a detection limit of 0.83 μg ml⁻¹. Recoveries and absolute recoveries of cefepime from plasma were 96.13–99.44% and 94–102.5% respectively. The intra-day and inter-day reproducibilities were less than 2% for cefepime at 10, 30, 50 μg ml⁻¹ (n=10).The method was proved to be suitable for determining cefepime levels in human plasma and was modified to measure vitreous fluid samples.     

High-performance liquid chromatographic analysis of AQ4N, an alkylaminoanthraquinone N-oxide

A simple, highly selective and reproducible reversed-phase high-performance liquid chromatography method has been developed for the analysis of the new anti-cancer pro-drug AQ4N. The sample pre-treatment involves a simple protein precipitation protocol, using methanol. Chromatographic separations were performed using a HiChrom HIRPB (25 cm×4.6 mm I.D.) column, with mobile phase of acetonitrile–ammonium formate buffer (0.05 M) (22:78, v/v), with final pH adjusted to 3.6 with formic acid. The flow-rate was maintained at 1.2 ml min⁻¹. Detection was via photodiode array performed in the UV range at 242 nm and, since the compounds are an intense blue colour, in the visible range at 612 nm. The structurally related compound mitoxantrone was used as internal standard. The validated quantification range of the method was 0.05–10.0 μg ml⁻¹ in mouse plasma. The inter-day relative standard deviations (RSDs) (n=5) ranged from 18.4% and 12.1% at 0.05 μg ml⁻¹ to 2.9% and 3.3% at 10.0 μg ml⁻¹ for AQ4N and AQ4, respectively. The intra-day RSDs for supplemented mouse plasma (n=6) ranged from 8.2% and 14.2% at 0.05 μg ml⁻¹ to 7.6% and 11.5% at 10.0 μg ml⁻¹ for AQ4N and AQ4, respectively. The overall recovery of the procedure for AQ4N was 89.4±1.77% and 76.1±7.26% for AQ4. The limit of detection was 50 ng ml⁻¹ with a 100 μl sample volume. The method described provides a suitable technique for the future analysis of low levels of AQ4N and AQ4 in clinical samples.     

Sulfalene concentrations in plasma and blood cells of Plasmodium falciparum malaria cases after treatment with metakelfin using high-performance liquid chromatography

A reversed-phase high-performance liquid chromatographic method using acetonitrile–methanol–1 M perchloric acid–water (25:9:0.8:95, v/v/v) at a flow-rate of 1.0 ml min⁻¹ on LiChrospher 100 RP 18 column (250×4 mm; 5 μm) with UV (254 nm) detection has been developed for the determination of sulfalene in plasma and blood cells after oral administration of the antimalarial drug metakelfin. Calibration curves were linear in the range 0.5–100 μg ml⁻¹. The limit of quantification was 50 ng ml⁻¹. Within-day and day-to-day coefficients of variation averaged 3.84 and 5.31%, respectively. Mean extraction recoveries of sulfalene from plasma and blood cells were 87.21 and 84.65%, respectively. Mean concentrations of sulfalene in plasma of P. falciparum cases on days 2, 7 and 15 were 44.58, 14.90 and 1.70 μg ml⁻¹, respectively; in blood cells concentrations of sulfalene were 7.77, 3.25 and 0.75 μg ml⁻¹, respectively, after oral treatment with two tablets (1000 mg) of metakelfin. Significant difference was recorded on day 2 for sulfalene concentration in blood cells of healthy and P. falciparum cases (t=9.49; P<0.001).     

Quantification of metronidazole in small-volume biological samples using narrow-bore high-performance liquid chromatography

A rapid, selective and sensitive HPLC assay has been developed for the routine analysis of metronidazole in small volumes of rat plasma, gastric aspirate and gastric tissue. The extraction procedure involves liquid–liquid extraction and a protein precipitation step. A microbore Hypersil ODS 3 μm (150×2.1 mm I.D.) column was used with a mobile phase consisting of acetonitrile–aqueous 0.05 M potassium phosphate buffer (pH 7) containing 0.1% triethylamine (10:90). The column temperature was at 25°C and the detection was by UV absorbance at 317 nm. The limit of detection was 0.015 μg ml⁻¹ for gastric juice aspirate and plasma and 0.010 μg g⁻¹ for gastric tissue (equivalent to 0.75 ng on-column). The method was linear up to a concentration of 200 μg ml⁻¹ for plasma and gastric juice aspirate and up to 40 μg g⁻¹ for tissue, with inter- and intra-day relative standard deviations less than 14%. The measured recovery was at least 78% in all sample matrices. The method proved robust and reliable when applied to the measurement of metronidazole in rat plasma, gastric juice aspirate and gastric tissue for pharmacokinetic studies in individual rats.     

Determination of sodium artesunate in plasma using ion-pairing high-performance liquid chromatography

A chromatographic method is described for the determination of sodium artesunate in plasma. This includes cetyltrimethylammonium bromide as a cationic pairing ion in a reversed-phase system using an octadecylsilica 100×4.6 mm I.D. 3 μm analytical column with a mobile phase of acetonitrile/acetate buffer at pH7. Column switching incorporating a 5 μm octadecylsilica 100×4.6 mm I.D. precolumn is used in addition to off-line solid-phase extraction for pretreatment of plasma samples in order to eliminate interference from endogenous components. Detection is by post-column derivatisation with 1.0 M methanolic KOH followed by UV detection at 289 nm. Calibration is linear over the range 100–1600 ng ml⁻¹ and the limit of detection is estimated as 20 ng ml⁻¹. Illustrative results are shown of the artesunate plasma levels determined by the proposed method following the administration of artesunate as tablets and as suppositories to healthy volunteers.     

Assay of R-apomorphine, S-apomorphine, apocodeine, isoapocodeine and their glucuronide and sulfate conjugates in plasma and urine of patients with Parkinson's disease

Analytical methods are described for the selective, rapid and sensitive determination of R- and S-apomorphine, apocodeine and isoapocodeine and the glucuronic acid and sulfate conjugates in plasma and urine. The methods involve liquid-liquid extraction followed by high-performance liquid chromatography with electrochemical detection. The glucuronide and sulfate conjugates are determined after enzymatic hydrolysis. For the assay of R- and S-apomorphine a 10 μm Chiralcel OD-R column is used and the voltage of the detector is set at 0.7 V. The mobile phase is a mixture of aqueous phase (pH 4.0)-acetonitrile (65:35, v/v). At a flow-rate of 0.9 ml min⁻¹ the total run time is ca. 15 min. The detection limits are 0.3 and 0.6 ng ml⁻¹ for R- and S- apomorphine, respectively (signal-to-noise ratio 3). The intra- and inter-assay variations are <5% in the concentration range of 2.5-25 ng ml⁻¹ for plasma samples, and <4% in the concentration range of 40-400 ng ml⁻¹ for urine samples. For the assay of apomorphine, apocodeine and isoapocodeine, a 5 μm C₁₈ column was used and the voltage of the detector set at 0.825 V. Ion-pairing chromatography was used. The mobile phase is a mixture of aqueous phase (pH 3.0)-acetonitrile (75:25, v/v). At a flow-rate of 0.8 ml min⁻¹ the total run time is ca. 14 min. The detection limits of this assay are 1.0 ng ml⁻¹ for apomorphine and 2.5 ng ml⁻¹ for both apocodeine and isoapocodeine (signal-to-noise ratio 3). The inter-assay variations are 5% in the concentration range of 5-40 ng ml⁻¹ for plasma samples and 7% in the concentration range of 50-500 ng ml⁻¹ for urine samples. The glucuronic acid and sulfate conjugates of the various compounds are hydrolysed by incubation of the samples with β-glucuronidase and sulfatase type H-1, respectively. Hydrolysis was complete after 5 h of incubation. No measurable degradation of apomorphine, apocodeine and isoapocodeine occurred during the incubation. A pharmacokinetic study of apomorphine, following the intravenous infusion of 30 μg kg⁻¹ for 15 min in a patient with Parkinson's disease, demonstrates the utility of the methods: both the pharmacokinetic parameters of the parent drug and the appearance of apomorphine plus metabolites in urine could be determined.     

High-performance liquid chromatography of the neuroactive steroids alphaxalone and pregnanolone in plasma using dansyl hydrazine as fluorescent label: application to a pharmacokinetic-pharmacodynamic study in rats

This report describes a rapid and sensitive analytical method for the quantification of the neuroactive steroids alphaxalone and pregnanolone in rat plasma using derivatization with dansyl hydrazine as fluorescent label. The method involves protein precipitation, alkaline derivatization and extraction of the compounds and internal standard pregnenolone with dichloromethane, followed by isocratic reversed-phase high-performance liquid chromatography on a 3-μm Microsphere C₁₈ column with fluorescence detection at wavelengths 332 nm and 516 nm for excitation and emission, respectively. The mobile phase consists of a mixture of 25 mM acetate buffer (pH 3.7)–acetonitrile (45:55, v/v for alphaxalone and 40:60, v/v for pregnanolone) with a flow-rate of 1 ml/min. The total run time was 35 min. In the concentration range of 0.010–10 μg ml⁻¹, the intra- and inter-assay coefficients of variation were less than 17% for both methods. In 50 μl plasma samples the corresponding limits of detection were 10 ng ml⁻¹ (signal-to-noise ratio=3). The utility of the analytical method was established by analyzing plasma samples from rats, which had received an intravenous administration of 5 mg kg⁻¹ alphaxalone or pregnanolone. Values for clearance, volume of distribution at steady state and terminal half life were 71.9 ml min⁻¹ kg⁻¹, 814 mg kg⁻¹ and 13.5 min for alphaxalone and 69.2 ml min⁻¹ kg⁻¹, 1638 ml kg⁻¹ and 27.8 min for pregnanolone, respectively. Due to its simplicity and sensitivity this method can be used on a routine basis for pharmacokinetic analysis of neuroactive steroids.     

New sensitive assay of vancomycin in human plasma using high-performance liquid chromatography and electrochemical detection

A method using reversed-phase high-performance liquid chromatography with electrochemical detection for the analysis of vancomycin in human plasma was developed. Chromatographic conditions included an octadecyl column, a mobile phase of acetonitrile–sodium phosphate buffer (pH 7) (12:88), a total run time of 12 min, and coulometric electrochemical detection at +700 mV. Linear detector response was found in the range 5–100 μg ml⁻¹ after a 1:80 dilution or from 0.5 to 50 μg ml⁻¹ after a 1:20 dilution of the samples. In both cases the correlation coefficient (r) of the calibration curve standard was better than 0.995. Vancomycin determination was based on a denaturation of plasma proteins with methanol, then a dilution with mobile phase was performed. Recovery of vancomycin from plasma was 103.1±3.9%, and no interference from commonly used drugs or endogenous compounds was observed. A significant correlation was shown with the EMIT assay (r=0.92, P<0.001) using clinical samples from children. This HPLC technique is simple, sensitive, rapid, precise, selective and requires only 100 μl of plasma for completion.     

The mechanism of enhancement by fatty acid hydroperoxides of anaphylactic mediator release

Fragments of chopped lung from indomethacin treated guinea-pigs had an anti-aggregating effect when added to human platelet rich plasma (PRP), probably due to the production of prostacyclin (PGI₂) since the effect was inhibited by 15-hydroperoxy arachidonic acid (15-HPAA, 10 μg ml⁻¹). Both 15-HPAA (1–20 μg ml⁻¹ min⁻¹) and 13-hydroperoxy linoleic acid (13-HPLA, 20 μg ml⁻¹ min⁻¹) caused a marked enhancement of the anaphylactic release of histamine, slow-reacting substance of anaphylaxis (SRS-A) and rabbit aorta contracting substance (RCS) from guinea-pig isolated perfused lungs. This enhancement was not reversed by the concomitant infusion of either PGI₂ (5 μg ml⁻¹ min⁻¹) or 6-oxo-prostaglandin F_1α (6-oxo-PGF_1α, 5 μg ml⁻¹ min⁻¹). Anaphylactic release of histamine and SRS-A from guinea-pig perfused lungs was not inhibited by PGI₂ (10 ng - 10 μg ml⁻¹ min⁻¹) but was inhibited by PGE₂ (5 and 10 μg ml⁻¹ min⁻¹). Antiserum raised to 5,6-dihydro prostacyclin (PGI₁) in rabbits, which also binds PGI₂, had no effect on the release of anaphylactic mediators. The fatty acid hydroperoxides may enhance mediator release either indirectly by augmenting thromboxane production or by a direct effect on sensitized cells. Further experiments to distinguish between these alternatives are described in the accompanying paper (27).     

Fosfomycin determination in serum by capillary zone electrophoresis with indirect ultraviolet detection

Capillary zone electrophoresis with indirect ultraviolet detection was used for the determination of fosfomycin in serum. Running buffer consisted of a mixture of 200 mM sodium borate with 10 mM phenylphosphonic acid used as ultraviolet absorbing background electrolyte. Relationships between the pH of the buffer and the efficiency of the separation (migration times and selectivities) or the sensitivity of detection were investigated. The method was then validated over a 10–100 μg ml⁻¹ concentration range to be applied to further therapeutic drug monitoring. The choice of ethylphosphonic acid as internal standard is discussed. The specificity and the linearity of the technique are demonstrated. The inter-day precision was satisfactory with a relative standard deviation of less than 2%. Accuracy was calculated with a standard error near 0.5 and 18% for 100 and 10 μg ml⁻¹, respectively.     

Determination of

Separation in capillary electrophoresis is governed by various factors, including buffer type, buffer concentration, pH, temperature, voltage and micelles. Through proper adjustment of these parameters, nalidixic acid and its two major metabolites, 7-hydroxynalidixic and 7-carboxynalidixic, could be separated by micellar electrokinetic capillary chromatography using an electrophoretic electrolyte consisting of 50 mM borate buffer (pH 9) containing 25 mM sodium dodecyl sulphate and 10% acetonitrile. A linear relationship between concentration and peak area for each compound was obtained in the concentration range 0.15–100 μg ml⁻¹, with a correlation coefficient greater than 0.999 and detection limits in the 0.2–0.7 ng ml⁻¹ range. Intra- and inter-day precision values of about 0.8–1.2% RSD (n=11) and 1.3–2.0% RSD (n=30), respectively, were obtained. The method has been applied to the analysis of nalidixic acid and its two major metabolites in serum and urine with limits of sensitivity lower than 0.8 ng ml⁻¹.     

Separation and simultaneous determination of nalidixic acid, hydroxynalidixic acid and carboxynalidixic acid in serum and urine by micellar electrokinetic capillary chromatography

Separation in capillary electrophoresis is governed by various factors, including buffer type, buffer concentration, pH, temperature, voltage and micelles. Through proper adjustment of these parameters, nalidixic acid and its two major metabolites, 7-hydroxynalidixic and 7-carboxynalidixic, could be separated by micellar electrokinetic capillary chromatography using an electrophoretic electrolyte consisting of 50 mM borate buffer (pH 9) containing 25 mM sodium dodecyl sulphate and 10% acetonitrile. A linear relationship between concentration and peak area for each compound was obtained in the concentration range 0.15–100 μg ml⁻¹, with a correlation coefficient greater than 0.999 and detection limits in the 0.2–0.7 ng ml⁻¹ range. Intra- and inter-day precision values of about 0.8–1.2% RSD (n=11) and 1.3–2.0% RSD (n=30), respectively, were obtained. The method has been applied to the analysis of nalidixic acid and its two major metabolites in serum and urine with limits of sensitivity lower than 0.8 ng ml⁻¹.     

Post-column enzyme reactors for chemiluminometric detection of glucose, 1,5-anhydroglucitol and 3-hydroxybutyrate in an anion-exchange chromatographic system

A liquid chromatographic system consisting of a co-immobilized 3-hydroxybutyrate dehydrogenase-NADH oxidase reactor and an immobilized pyranose oxidase reactor in series and a chemiluminometer was developed for the simultaneous determination of glucose, 1,5-anhydroglucitol and 3-hydroxybutyrate in plasma. The enzymes were immobilized on toresylated poly(vinyl alcohol) beads. Separation was achieved on a TSK gel SAX column (40×4 mm I.D.) with an eluent of 50 mM NaOH containing 30 mM sodium butyrate. The hydrogen peroxide produced was detected by measuring the chemiluminescence emitted on admixing with luminol and potassium hexacyanoferrate(III). The calibration curves were linear from 0.8 to 500 μM (7 ng−4 μg) for glucose, from 0.8 to 400 μM (7 ng−3 μg) for 1,5-anhydroglucitol and from 1 to 700 μM (5 ng−4 μg in a 50-μl injection) for 3-hydroxybutyrate. The sample throughput was four per hour. The reactors were stable for at least ten days.     

Determination of ceftazidime in plasma using high-performance liquid chromatography and electrochemical detection: Application for individualizing dosage regimens in elderly patients

This study describes a sensitive HPLC–electrochemical detection method for the analysis of ceftazidime, a third-generation cephalosporin, in human plasma. The extraction procedure involved protein precipitation with 30% trichloroacetic acid. The separation was achieved on a reversed-phase column (250×4.6 mm I.D., 5 μm) packed with C₁₈ Kromasil with isocratic elution and a mobile phase consisting of acetonitrile–25 mM KH₂PO₄–Na₂HPO₄ buffer, pH 7.4 (10:90, v/v). The proposed analytical method is selective, reproducible and reliable. The assay has a precision of 0.2–15.1% (C.V.) in the range of 5–200 μg ml⁻¹. (corresponding to 0.5 to 20 ng of ceftazidime injected onto the column), and is optimised for assaying 50 μl of plasma. The extraction recovery from plasma was approximately 100%. The method was highly specific for ceftazidime and there was no interference from either commonly administered drugs or endogenous compounds. This assay was used to measure ceftazidime in elderly patients for therapeutic drug monitoring.     

Ammonium excretion by a mutant of the nitrogen-fixing cyanobacterium Anabaena siamensis

Anabaena siamensis isolated from rice fields in Thailand is a fast growing cyanobacterium with a high nitrogen-fixing activity. Mutant strains resistant to the l-glutamate analogue, l-methionine sulfoximine (MSX) were isolated by ethyl methanesulfonate mutagenesis. A stable mutant named A. siamensis SS1, which released ammonium to the medium, was studied further. In batch cultures the rate of ammonium production peaked at the early log phase and gradually decreased until the 4th day of growth when the cultures reached a density of 90 μg chl ml⁻¹. To obtain constant release of ammonium by SS1, continuous culture experiments were performed at a cell density of 5 μg chl ml⁻¹ and the following results were obtained: (1) growth rate as the parent (μ:0·123 h⁻¹) in the presence and absence of 500 μm MSX; (2) 48% GS transferase activity when compared with the parent; (3) ammonium excretion at a rate of 8 μmol (mg chl)⁻¹ h⁻¹ as measured up to 20 generations (120 h); (4) depressed nitrogenase activity; and (5) 30% higher nitrogenase activity than that of the parent. SS1 immobilized in alginate beads (5 μg chl ml⁻¹) exhibited values of glutamine synthetase and nitrogenase activity similar to those of free cells. However, ammonium excretion at the rate of 11·61 μmol (mg chl)⁻¹ h⁻¹ was obtained only up to 20 h after loading in bioreactors, due to the fast growth of SS1 as also occurred in batch cultures.     

Biodeterioration of some herbal raw materials by storage fungi and aflatoxin and assessment of Cymbopogon flexuosus essential oil and its components as antifungal

Deterioration of raw materials of six medicinal plants viz. Terminalia arjuna, Acorus calamus, Rauvolfia serpentina, Holarrhena antidysenterica, Withania somnifera and Boerhaavia diffusa was examined. Some of the contaminated raw materials were found to be deteriorated by toxigenic strains of Aspergillus flavus and contain aflatoxin B₁ (41.0–95.4 μg kg⁻¹) which is above the permissible limit. Essential oil of Cymbopogon flexuosus and its components was found efficient in checking fungal growth and aflatoxin production. C. flexuosus essential oil absolutely inhibited the growth of A. flavus and aflatoxin B₁ production at 1.3 μl ml⁻¹ and 1.0 μl ml⁻¹ respectively. The individual oil components were more efficacious than the Cymbopogon oil as such which emphasizes masking of their efficacy when combined together. Eugenol exhibited potent antifungal and aflatoxin inhibitory activity at 0.3 μl ml⁻¹ and 0.1 μl ml⁻¹ respectively. Eugenol was found superior over some prevalent synthetic antimicrobials and exhibited broad fungitoxic spectrum against some biodeteriorating moulds. Prospects of exploitation of the oil and its components as acceptable plant based antimicrobials in qualitative as well as quantitative control of biodeterioration of herbal raw materials have been discussed.     

Antifungal activities of anise oil, lime oil, and tangerine oil against molds on rubberwood (Hevea brasiliensis)

The antifungal activities of anise oil, lime oil, and tangerine oil against molds identified from rubberwood surfaces (Aspergillus niger, Penicillium chrysogenum, and Penicillium sp.) were investigated. The broth dilution method was employed to determine the minimal inhibitory concentration (MIC) and minimal fungicidal concentration (MFC) using the concentration of essential oils between 20 and 200 μl ml⁻¹. Inhibitory effects of the essential oils against those molds on rubberwood were also examined by means of the dip treatment and vacuum impregnation treatment. It was found that the MIC and MFC values for each treatment on agar were identical for all conditions examined. Anise oil was the strongest inhibitor with the MIC and MFC of 40 μl ml⁻¹ against Penicillium sp. and A. niger, and 60 μl ml⁻¹ against P. chrysogenum. Lime oil and tangerine oil were also effective against those molds at higher concentrations of 100–180 μl ml⁻¹. All essential oils at the MIC and MFC provided a protection from mold growth on rubberwood for at least 12 weeks at storage conditions of 30 °C with 100% RH.     

Specific high-performance liquid chromatography assay for determination of rifabutin plasma concentration following Extrelut column extraction

A specific, precise and accurate assay for determination of rifabutin in human plasma using Extrelut column extraction was developed and validated. Rifabutin concentrations were calculated with a standard curve ranging from 5 to 800 ng ml⁻¹, using a split-curve approach. Chromatographic peaks were separated by means of a 5 μm Symmetry Shield RP8 using a KH₂PO₄ (0.05 M) buffer–acetonitrile mobile phase. Detection wavelength was set at 275 nm. Chromatography was carried out at room temperature (20–25°C). The limit of quantification was 5 ng ml⁻¹. The recovery was over 71%. The intra-day precision of the assay was 5, 7, and 1% while the inter-day precision was 11.2, 8.1, and 5.8% at concentrations of 30, 150 and 500 ng ml⁻¹, respectively. The accuracy ranged from 99 to 108%. Forty of the drugs most commonly administered to HIV-positive patients were found not to interfere with the assay. The assay has been used in a comparative study of rifabutin pharmacokinetics in HIV-positive patients with or without wasting syndrome.     

Human serum albumin chromatography by Cibacron Blue F3GA-derived microporous polyamide hollow-fiber affinity membranes

An affinity dye ligand, Cibacron Blue F3GA was covalently attached onto commercially available microporous polyamide hollow-fibre membranes for human serum albumin (HSA) adsorption from both aqueous solutions and human plasma. Different amounts of Cibacron Blue F3GA were incorporated on the polyamide hollow-fibres by changing the dye attachment conditions, i.e. initial dye concentration, addition of sodium carbonate and sodium chloride. The maximum amount of Cibacron Blue F3GA attachment was obtained at 42.5 μmol g⁻¹ when the hollow-fibres were treated with 3 M HCl for 30 min before performing the dye attachment. HSA adsorption onto unmodified and Cibacron Blue F3GA-derived polyamide hollow-fibre membranes was investigated batchwise. The non-specific adsorption of HSA was very low (6.0 mg g⁻¹ hollow-fibre). Cibacron Blue F3GA attachment onto the hollow-fibres significantly increased the HSA adsorption (147 mg g⁻¹ hollow-fibre). The maximum HSA adsorption was observed at pH 5.0. Higher HSA adsorption was observed from human plasma (230 mg HSA g⁻¹ hollow-fibre). Desorption of HSA from Cibacron Blue F3GA derived hollow-fibres was obtained using 0.1 M Tris–HCl buffer containing 0.5 M NaSCN or 1.0 M NaCl. High desorption ratios (up to 98% of the adsorbed HSA) were observed. It was possible to reuse Cibacron Blue F3GA derived polyamide hollow-fibre without significant decreases in the adsorption capacities.