WO2005015166A1

WO2005015166A1 - Method and apparatus for automated swab sample analysis

Info

Publication number: WO2005015166A1
Application number: PCT/EP2004/008261
Authority: WO
Inventors: Vincenzo Rizzo; Katia Marcucci; Giuseppe Razzano; Rosaria Mariani
Original assignee: Pharmacia Italia S.P.A.
Priority date: 2003-07-25
Filing date: 2004-07-23
Publication date: 2005-02-17

Abstract

A method for the analysis of cleaning samples is described. These samples, delivered as adsorbates on a wiping surface like swabs, are produced in order to verify the absence of contamination for example in manufacturing equipment, after a specific cleaning procedure. The method is based on the automated extraction of swabs with a general purpose solvent by means of a liquid handler apparatus, the preparation of samples and standards in a well plate format, followed by HPLC/MS analysis of the assay of the active ingredient residues.

Description

TITLE OF THE INVENTION

METHOD AND APPARATUS FOR AUTOMATED SWAB SAMPLE ANALYSIS

BACKGROUND OF THE I NVENTI ON Field of the invention The invention addresses a method for the automated analysis of samples in the chemical industry, in particular in the pharmaceutical manufacturing. Background of the invention Cleaning procedures in the pharmaceutical industr y are crucial to avoid safety concerns that may result from cross -contamination among different pharmaceutical products or active ingredients. Critical cleaning processes are those that have a reasonable probability of generating contamination of the subse quently manufactured product if the involved surfaces are inadequately cleaned. For example, since 1963, the FDA has required that equipment be cleaned prior to use and this is one of the basic GMP requirements. It is also mentioned in more than one sectio n of 21 CF 211. The samples to be analysed may be also those deriving from cleaning procedures of the glove boxes, so as to avoid that dangerous contaminants could be present on these equipments during maintenance operations. The samples to be analysed can fiirther derive from other surfaces present in a pharmaceutical plant such as conditioning filters, walls, doors, safety suit and the like, in order to monitor environmental contamination by active ingredients. Validation of cleaning procedures involves organizational, technical and legal aspects. Specific areas to be taken into account are sampling methods, analytical methods, physical parameters and the selection of acceptance criteria.

The cleaning processes must be validated by sampling of critical s urfaces of manufacturing equipments after cleaning, followed by the assay of active ingredient residues on the relevant samples. For example, in a pharmaceutical industry plant it is normally required that the amount of detectable active ingredient in the cleaning samples must be less than 0.1 μg/cm², more preferably less than 0.01 μg cm ² of equipment surface area. For highly potent or mutagenic compounds, even lower limits are desirable and the analytical method sensitivity is often defining the practical ones. Swabbing is a widely used sampling technique. Swabs may be used dry or saturated with a suitable solvent to aid the solubilization and physical removal of surface residues. The selection of the swabbing agent can be done according to the solubilitie s of the active ingredients. In addition, the solvent should leave no toxic residues on equipment and should not leach extractables from swabbing material that could interfere with the active ingredients determination. The swabbed area may vary and should be accounted for in the calculations. When swab testing is used, the recovery of the sampling technique must be evaluated and documented. A representative surface is spiked with a known amount of the indicator substance and the amount recovered from swab s ampling is determined. The recovery should be greater than 50% and it must be taken into account when performing the calculations of residues. As stated before, the method used to analyze swab samples must be validated. The common practice is to manually extract swabs in a fixed volume of an appropriate solvent and then to analyze the solution with HPLC -UN, using method conditions and UN detection dependent on the analyte. As HPLC or spectrophotometric methods of analysis are often time consuming and subje ct to a number of interferences, other analytical techniques have been applied to cleaning validation such as total organic carbon analysis and overpressured layer chromatography. Surprisingly, in a survey of recent literature on this subject, we have not been able to find any application of LC -MS technique to cleaning procedures. Summary of the invention The present invention provides a new process for automated analysis of samples as obtained from an object, in particular from equipments, more particularl y from pharmaceutical equipments, to verify the absence of contamination from previously processed active ingredients (cleaning samples).

In general, it is desirable to have a simple and rapid way of detecting even very small amounts of active ingredients on objects. With an increased degree of sensitivity of the detection, it is, of course, possible to detect even smaller amounts of analyte on a given contaminated surface.

On line with current procedures, samples are prepared with swabbing one or more predefined surfaces of the equipment with a wad of adsorbing material wound around one end of a small stick (swab samples). The present invention provides a method for detecting surface contamination by an active ingredient by wiping the active ingredient of f the surface with the aid of a wiping surface, preferably a swab, extracting the active ingredient from the wiping surface with an appropriate extracting solvent, analyzing the resultant solution and optionally processing the obtained data, characterized in that: i) the extraction process of the wiping surface is performed in a fully automated fashion, and ii) LC-MS is used as analytical technique.

The investigated surface is preferably a portion of an apparatus used in the manufacturing of pharmaceutical products or active ingredients, or any surface within a building at or near the manufacturing facility, or a plate of_. any material, as needed for the simulation of each of the above surfaces during method development and validation.

Brief description of the drawings

Figure 1 is a schematic drawing of a 96 deep -well plate containing all samples, spiked samples and standards for the LC -MS analysis sequence. A case of 21 different sampling swabs is shown. Samples (Swl,2,...i) are in light grey color, spiked sample (Sw 1,2,... i+ spike) are in dark grey color and the standard samples are light up diagonal grey color. The _. other white ones are empty.

Figure 2 is the photograph of plastic test tube (A), special stopper (B); swab (C) and assembled set ready for auto matic extraction (D). Detailed description of preferred embodiments

In the method of the present invention, the fully automated extraction process of the wiping surface is performed with the help of an automatic liquid handler and of appropriate disposable material, part of the last one being also object of the present invention.

Moreover, the appropiate extracting solvent is chosen in view of the active ingredient to be analysed and of the analytical method. In the widely applicable method of the present invention the extraction of swabs is performed with a generic and powerful extracting solvent, independently of the nature of the active ingredient to be assayed. Besides the solvent should have poor volatility and low toxici y.

DMSO and isopropanol can be used to extract swabs, DMSO being preferred as a generic extraction solvent. In addition, the swabs extraction solvent must guarantee no interferences from swabbing material. The specificity tests performed on the clean swabs demonstrated the absence of interferences from Alpha Swabs® when extracted with DMSO. DMSO is preferred as generic solvent, since it makes the extraction process essentially independent of the nature of the active ingredient to be assayed.

The use of LC-MS as analytical technique ins tead of LC-UV greatly improves the detectable amount of active ingredient and avoids time consuming studies and possible interferences . As a matter of fact, the intrinsic specificity of the MS detection method greatly facilitates method development, and th e large sensitivity improvement offered with respect to UV generates an increased safety margin to the required detection limits. LC-MS sensitivity offers low detection limits, in the parts -per-billion range, facilitating the swab extraction and the sample dilution stages and providing safety margins in the working »₍ concentration range. ! One of the most important characteristics of the MS detector is its universality (i.e. its applicability to a wide range of compounds). In consideration of the main advantag e of the; ^■ present invention, that is a generic method that does not require a case -by-case evaluation, this feature is very valuable. The intrinsic specificity of the MS detection method greatly facilitates method development compared to a non -selective test. Product excipients, agents used in the cleaning process and compounds that may be introduced by the sampling procedure do not interfere with the key contaminant detection. Of the two commonly available direct spray ionization methods for MS, Electrospr ay Ionization (ESI) and Atmospheric Pressure Chemical Ionization (APCI), the former is preferred. Selected Ion Monitoring (SIM) was preferred to a full scan detection mode since the first affords a major sensitivity. For each compound the [M+H]+ ion may be monitored. Selected Ion Monitoring (SIM) scan type is preferable as it makes the method easily transferable to any mass analyzer. The selectivity and the sensitivity offered by the SIM detection mode are sufficient for this kind of application of the MS d etector. A specific fine tuning for each compound is not required. An additional control of the entire procedure can be used. As described in the experimental section, for each sample also a spiked solution can be injected. Every chemical interference (positive or negative) or instrumental deviation is kept under control using this strategy. When a sample falls outside the action lines, the process has to be investigated because it could be out of control. If a sample falls between action and warning lines, it does not give cause of concern but, if two such points are obtained consecutively, then the process should be investigated.

The analysis of actual samples for any active ingredient which is the object of the present invention is preceded by the method development. This method development includes the method validation regarding linearity, specificity, limit of detection, injections repeatability and recovery from stainless steel or other surfaces. The first four parameters are conventional experiments used to validate an analytical method. The recovery from stainless steel or other surfaces is a more critical parameter because it involves the extraction of the active ingredient from the swabs. The common practice is to determine the recovery from the s wabs and from the stainless , steel surface or other representative materials in independent investigations. Evaluating at ' first the recovery from the spiked plate is preferable, in order to obtain the numerical value to be used in the calculations on actual samples. Only if the recovery from the surface is less than 50% a further study to investigate the critical step (i.e. the recovery from swabs or the removal of the active ingredient from the plate) is necessary.

If the recovery from the swab, as obtained using the standard procedure, is not sufficient two adjustments can be made:

1. the first parameter to be taken into account is the chemical nature of the active ingredient. If the target compound is acidic or basic, the extraction can be performed with a solution of the solvent like DMSO supplemented with either a base like ammonia or an acid like formic acid. An independent determination of compound stability in these media is preferably required.

2. if the solubilization of the target compound cannot be improved as described under point 1. above, the number of extraction cycles of the liquid handler should be increased to obtain a higher recovery from swabs. If the removal of the target compound from the surface, for example a stainless steel plate, is the critical step, a simple adjustment of the pH of the solvent used to wet the swabs may prove sufficient. For example, since acidic solutions can improve the recovery of basic compounds, an acid solution, like 0.2% formic acid, may be added to the swabbin g agent in such cases. A high recovery value is also an evidence of chemical stability in these conditions.

The procedure for the treatment of actual samples (i.e. the swabs deriving from the sampling of manufacturing equipment surfaces) according to the p resent invention is consisting of three steps: extraction of the active ingredient from the swabs, analysis by LC-MS and optional data processing.

The extraction of the target compound from the swabs is performed as described in more detail herein below.

The chromatographic conditions must not be optimized for each active ingredient. The only ι analysis parameter to modify is the SIM scan range, which is chosen on the basis of the active ingredient mass spectrum (usually the protonated molecular ion is the mo st intense ion). The choice of the APCI interface, of the negative ion detection mode or fine tuning are required only in the very rare case when the active ingredient does not show a satisfactory response factor under the usual mass spectrometric conditio ns. Data analysis for the active ingredient simply requires the definition of the peak identification and integration settings in the automated processing method. The stepwise procedure for data processing and quantitation is identical for every active ing redient. The process of the present invention is adequate to investigate cleaning samples of any new active ingredient with only minor adjustments such as the fine tuning of the MS detector to the most suitable parameters and the supplement of the extracti ng medium (for example DMSO) with acidic or alkaline correctors, if needed. In particular, we found that good or excellent recovery is obtained with automatic extraction of swabs by means of DMSO; thus eliminating the need for the optimization of the extra ction solvent or reducing this process to two simple trials: pure DMSO and DMSO with a corrector (acidic or alkaline, depending on the nature of the active ingredient under investigation). The use of disposable materials in the extraction phase, as well as in the analysis, greatly simplifies the operator work: only two solutions of the active ingredient need to be prepared and very limited housekeeping work is required: no washing and rinsing of volumetric glassware, no vial labeling and crimping. Moreover, the use of a standard liquid handler and of an ordinary 96-well plate (Fig. 1) as sample container makes the sample preparation step easily exportable to other laboratories, without the need of purchasing expensive, dedicated instruments.

The analytical p art of the process is characterized by the well known advantages of LC -MS detection versus LC-UV detection: improved sensitivity, spectroscopy rather than chromatography based selectivity make method development easy, and permit the use of generic fast gradient based LC separation with very acceptable validation parameters. In particular, the improved limit of detection makes analysis intrinsically more robust; the reduced analysis time provides extra space for control samples and thus higher quality of the results and the simple structure of delivered data greatly facilitates final transformation into practical results. The method of the present invention can be therefore applied to several compounds involved in the pharmaceutical manufacturing processes.

As stated above, this method includes the extraction of residual active ingredient from swabs into an appropriate solvent like DMSO with adequate recovery, and a subsequent analysis with LC-MS. This analytical method is easily adaptable to any new active ingredient with a simple adjustment of instrumental parameters, requiring simple validation only for linearity, specificity, limit of detection and recovery from appropriate surfaces (i.e. stainless steel, etc.).

In order to perform the automatic extraction procedure of the present invention, we have found an appropriate stopper capable to keep the swab in the upright position without preventing the movement of the liquid handler tips, so that the position of the swab throughout the automatic extraction proc edure is fixed. This special stopper (the B piece in Figure 2) is a further object of the present invention. It may be prepared so that it fits any ordinary test tube and contains two openings; one opening to tightly fix the swab stick, and the other hole to accommodate the pipetting ends of the automatic liquid handler. Preferably, the stopper of the present invention fits any ordinary 16 mm diameter test tube and contains a small opening to tightly fix the swab stick, and a wider hole to accommodate the disposable 1 mL tips of a liquid handler. More preferably, the wider opening is circular, and the small opening fixing the swab stick has the same shape of the stick, for example ellyptic.

Extraction procedure Swab extraction is preferably performed using as solvent DMSO. With the chosen disposable tubes, preferably polypropylene plastic tubes, a volume is administered which is sufficient to completely cover the part of the swabs where the active ingredient is adsorbed. The program sequence of the liquid ha dler performs the following operations: prepare the analytical well plate with distributing the correct amount of buffer or other solvent into the wells where samples, spiked samples and calibration solutions will be placed; dispense the target volume of extraction soIvent^DMSO) in the tubes, preferably polypropylene tubes; ' for each tube extract the active ingredient absorbed on swabs with repeated cycles of aspiration and dispensing, each followed by a pause, such as one minute pause; distribute samples (typically 0.1 mL solution) from the tubes into the appropriate wells of the analysis plate; prepare calibration solutions (typically 4), and spiked samples with dispensing appropriate volumes of a stock solution of the investigated active ingredient; thoroughly mix all solutions in wells with repeatedly aspirating and dispensing the entire contained volume in disposable tips.

Preferably, the number of extraction cycles is six and, as stated above, it may be extended above six in order to obtain the best r ecovery from swabs. Each cycle is composed of a solution mixing step followed by a wait time, which allows for a better solubilization of the active ingredient. The optimal number of extraction cycles may be experimentally determined after testing differen t numbers of cycles.

The calculated concentration for each sample Csw , is used for the determination of the amount of the active ingredient on the manufacturing equipment surface: QSi = (Cswi x DV x 0.001) / (R x SwA) Eq. 1 wherein: QSj = amount of the target compound on the swabbed part of the equipment (as μg/cm ² of equipment surface area) Csw; = calculated concentration for the i ^tb sample (as ng/mL)

DV = dilution volume (the total volume where the content of a given swab is distributed) R = recovery from model surface

SwA = manufacturing equipment sampled area (usually 100 cm ²) In order to validate the present method and to assess its linearity and repeatability tests, different working solutions and standard solutions can be prepared by well kno wn procedures. In particular, working solutions are used for the recovery determination by means of a) preparation of the surfaces and b) swabs preparation and extraction.

The following example illustrates the invention without limiting it. ₍

Examples ₍ , Determination of residual content of two cytotoxic drugs firinotecan hvdrochloride and idarubicin hvdrochloride. on stainless steel surfaces with equal or better than 0.2 ng/cm ² and 0.7 ng/cm detection limits, respectively.

EXAMPLE 1

Preparation of the standard solution. About 2.4 mg, exactly weighed, of each compound were transferred in a 100 mL flask and dissolved with a suitable volume of DMSO in order to obtain a 24 μg/mL solution. The exact quantity of DMSO to be added is determined by weight assuming a density of 1.101 g /mL.

EXAMPLE 2

Preparation of the standard solutions for linearity and repeatability tests. The standard solution (see above ) was diluted 1 :50 in DMSO in order to obtain a 480 ng/mL intermediate standard solution (solution A). The solutions for the linearity test (A 1 , ... , A7) and for the repeatability test (A8) were prepared according to the scheme reported in Table

1. A 10 mM ammonium acetate buffer pH 5 was used for the dilutions. Table 1. Composition of the solutions for the linearity and re peatability tests

EXAMPLE 3 ; ^{' " '}

Analytical method : ■ -- -- - -

HPLC-MS analyses were conducted on a Finnigan LCQ Deca ion trap mass spectrometer equipped with an Electrospray (ESI) ion source. The mass spectrometer was directly connected to a Spectra System P4000 HPLC pump (Thermo Separation Products), equipped with a 20 PAL Autosampler (CTC Ana lytics). For sample injection a Vici injector, equipped with a 10 μL loop was used. The system was controlled by the Excalibur Software 1.2 version. Details of the chromatographic method are reported here below. Column: X-Terra MS C is, 3.5 μm, 4.6 x 30 mm, Waters

Mobile phase A: water with 0.1 %(v/v) formic acid

Mobile phase B: acetonitrile with 0,1% (v/v) formic acid

Flow rate: 1.2 mL/min

Ratio between total flow rate and flow to MS detector: 1 : 6

Elution: Time (min) %Mobile phase A 0.0 90 5.0 10 5.1 90 6.3 90

For the two compounds the electrospray with positive ion detection was used as ionization mode. A generic tuning of the LC/MS instrument was applied: spray voltage = 4 kV, heated capillary temperature = 340 °C. Ion trap parameters were as follows: inject waveform type

2 on; automatic gain control on; maximum isolation time 150 ms; 3 microscans per scan; ion targets for MS ^ϊ 5x10⁷ and for SIM 2x10⁷. For each compound a Single Ion Monitoring

(SIM) scan type was chosen: irinotecan MH ⁺ = 587, mass range 586.5 - 588.5; idarubicin

MH⁺ = 498, mass range 497.5 - 499.5.

Data analysis was automatically performed using the program LCQuan, included in

Xcalibur data system. The stepwise procedure for quantitativ e analysis included peak i identification, peak integration, quantitation, reviews of all results and reports generation. __ 1 The LCQuan routine accomplishes quantitation by comparing the sample measured peak area to the calibration curve. ^l

EXAMPLE 4

Linearity test and limits of detection The results for solutions (Al,..., A7) of the two compounds are reported in Table 2.

The calibration data were fitted to a linear calibrati on curve (y = a + bx) using the least squares regression method, with equal weight for all calibration data points. The results (Table 3) confirm good linearity in the range 1 - 100 ng/mL for irinotecan.HCl and in the range 5 - 100 ng/mL for idarubicin.HCI . ,-

Data of Table 2 provide also an estimate of the limit of detection (LOD, based on S/N > 3) for the two compounds: LOD for idarubicin.HCl = 5 ng/mL

LOD for irinotecan.HCl < 1 ng/mL

EXAMPLE 5

Repeatability

A solution with the concentration of 80 ng/mL (see solution A8 in Table 1) was analyzed six times for an assessment of method repeatability. Results are collected in Table 4 and show that relative standard devi ation (RSD) is maintained below 3% for both compounds at this concentration level.

EXAMPLE 6

Preparation of the working solutions

About 2 g, exactly weighed, of each compound were transferred into a 100 mL volumetric flask and totally dissolv ed with methanol in order to obtain a 20 μg/mL working solution (solution Wl). A volume of 5 mL of solution Wl was transferred into a 50 L volumetric flask and brought to volume with methanol in order to obtain a 2 μg/mL working solution (solution W2).

EXAMPLE 7

Swabs preparation and extr ction for the recovery determination .

By means of a suitable micropipette for volatile liquids, 50 μL of working solution Wl were distributed on four different swabs. After about 30 minutes at room temperature the swabs were dried for one hour at 40°C, The dried swabs were inserted in the plastic tubes and fixed in the upright position by mean of the special stoppers. The extraction was automatically performed by the liquid handler with the addition of 3 mL DMSO to each tube. Separately, each well of the analytical plate was prepared with 0.3 mL of 10 mM ammonium acetate buffer pH 5, and then 0.1 mL of DMSO extract was added and finally mixed with the buffer. Therefore, the final dilution volume of the swab content was 12 m L.

EXAMPLE 8

Recovery from swabs.

For each compound four swabs were prepared and extracted as described above . The solutions (Sw i) were analyzed with HPLC -MS method. The recovery for each swab was calculated according to the following equation:

Recovery ( %) = (Area S j Area Dl) x 100 Eq.2 wherein:

Area Sw. = peak area obtained for injection of the solutions Sw i. Area Dl = peak area obtained for injection of a standard solution (Dl) containing the same amount of active ingredient as distributed on each s wab (1 μg) in the same final volume

(12 mL) as obtained after extraction and dilution for analysis.

For each compound the recovery was calculated as the mean value of the recoveries obtained for the four swabs. The extraction solvent was DMSO (3 mL). Two types of ^[ swabs were tested: the Large Alpha Swab TX 714A suitable for wide areas sampling and the Alpha Swabs^® TX 761 good for small area sampling. Both types are made of knitted polyester.

EXAMPLE 9 Preparation of the stainless steel plates for the recovery determination , By means of a suitable micropipette for volatile liquids, 500 μL of working solution W2 were distributed on each one of three stainless steel plates (10x10 cm). The plates were dried at room temperature and, after solvent evaporation, each plate was rubbed with a swab previously wetted with an appropriate solvent. EXAMPLE 10 Recovery from stainless steel plates For each compound three square stainless steel plates, 10 cm side, were prepared as described above. After solvent evaporation the plates were rubbed with the Large Alpha Swabs^® previously wetted with ethanol in the case of irinotecan, ethanol + 0.2 % formic acid in the case of idarubicin. The swabs were extracted according to the above described method. The solutions (S_p]) were analyzed with HPLC -MS as reported above. The recovery for each plate was calculated according to the following equation: Recovery (%) = (Area S _pi/ Area Dl) x 100 Eq. 3 wherein: Area S_Pj = peak area obtained for injection of the solutions S _pj Area Dl = peak area obtained for injection of a standard solution (Dl) containing the same amount of active ingredient as distributed on the pla te (1 μg) in the same final volume (12 mL).

The results are reported in Table 6. For both compounds the recovery exceeds the critical threshold of 50 % and is considered acceptable

EXAMPLE 11

Determination of minimum detectable amounts for idarubicin.HCl and irinotec an.HCl absorbed on stainless steel plates. With the use of Eq. 1, the minimum detectable amount ( MDA, μg/cm²) of irinotecan.HCl or idarubicin.HCl absorbed on stainless steel plates can be calculated, once the analytical detection limits (LOD) are used in place of the analytical concentration Csw,-. The sampling surface SwA = 100 cm², and the dilution volume DV = 12 mL are typical values for these determinations. MDAidarubi n = (LOD idarubicin x 12 mL x 0.001) / (R idarubicin x 100 cm²) = 0.0007 μg/cm² or

0.7 ng/cm

MDAirinotecan = (LOD irinotecan X 12 mL X 0.001) / (R irinotecan 100 Cm²) < 0.0002 μg/cm ² Or

0.2 ng/cm²

Claims

CLAIMS 1. A method for detecting surface contamination by an active ingredient by wiping the active ingredient off the surface with the aid o fa wiping surface, extracting the active ingredient from the wiping surface with an appropriate extracting solvent, analyzing the resultant solution and optionally processing the obtained data, characterized in that: i) the extraction process of the wipi ng surface is performed in a fully automated fashion, and ii) LC-MS is used as analytical technique. 2. A method according to claim 1 characterized in that the appropriate extracting solvent is selected from dimethylsulfoxide and isopropanol. 3. A method according to claim 1 characterized in that the wiping surface is a swab.

^■ 4_* A method according to claim 3 characterized in that the swab used for wiping the active ingredient off the surface is wetted with an appropriate solvent. 5. A method according to claim 1 characterized in that the surface is a portion of an apparatus used in the manufacturing of pharmaceutical products or active ingredients, or any surface within a building at or near the manufacturing facility, or a plate of any material, as needed for the simulation of each of the above surfaces during method development and validation. 6. A method according to claim 4 characterized in that the pH of the solvent used to wet the swabs is adjusted. 7. A method according to claim 1 characterized in t hat the extraction is performed with a solution of dimethylsulfoxide optionally supplemented with either a base or an acid. 8. A stopper capable to keep in the upright position a swab used for detecting surface contamination by an active ingredient, witho ut preventing the movement of liquid handler tips, characterized in that it fits any ordinary test tube and contains two openings, one to tightly fix the swab stick, and the other one to accommodate the pipetting ends of the automatic liquid handler. 9. A method according to claim 1 characterized in that the fully automated extraction is performed with a liquid handler apparatus having program sequence performing the following operations: prepare the the analytical well plate with distributing the correct amount of buffer or other solvent into the wells where samples, spiked samples and calibration solutions will be placed; dispense the target volume of extraction solvent in the swab -containing tubes, fit with the stopper described in claim 8; for each tube extract the active ingredient absorbed on swabs with repeated cycles of aspiration and dispensing, each followed by a pause; distribute samples from the tubes into the appropriate wells of the analysis plate; prepare calibration solutions, and spiked sam pies with dispensing appropriate volumes of a stock solution of the investigated active ingredient; thoroughly mix all solutions in wells with repeatedly aspirating and dispensing the entire contained volume in disposable tips.

10. A method according to cl aim 9 characterized in that the number of extraction cycles of the liquid handler is six or more.