1. INTRODUCTION CONCEPT Title: Preparation and validation of lyophilised Page : 1 of 6 Reference materials are samples with a known concentration of analytes. It s the assignment of the Community Reference Laboratories (CRL) to prepare a bank of reference materials of lyophilised urinary samples from livestock. This bank is useful for the development and validation of an analytical method (BCR-RIVM report contract 5281-1994). It s necessary to evaluate the performance of a method on a selection of samples representative for the population to be tested. Reference materials are also important for quality control of analytical procedures. For purposes of validation of a procedure, certified reference materials (CRM s) are preferred, for quality control reference materials (RM s) preferably calibrated against a certified reference material (CRM). Criteria for RM s have recently been summarised by Stephany and Van Ginkel (3.6). Reference materials for within laboratory quality control purposes are now available for use. If the objective of the project is the preparation of certified reference materials the results of the homogeneity and stability study have to be compared with the criteria set by e.g. the EC/BCR. If the materials are considered suitable a coöperative study has to be organised. 2. SUMMARY This Standard Operation Procedure (SOP) describes the complete process of preparation and validation of lyophilised urinary reference materials for anabolic agents and veterinary drugs. The procedure is based on five steps: 1. Preparation of sample materials. 2. Filtration of the urine. 3. Filling of the vials. 4. Lyophilisation. 5. Quality Control. The procedure does not include co-operative and/or collaborative studies for certification of candidate certified reference materials. Except the homogeneity and stability studies described in this SOP the analytical characterisation of source materials are described in the SOP s ARO/113, ARO/114 and ARO/401. Before homogeneity and stability studies are performed it is necessary to determine the repeatability and reproducibility of the analytical procedure used for the material concerned and approximately at the same concentration as the reference materials. A minimal validation study on the method to be used concerns duplicate analyses on at least three different occasions. From these data the within assay variability and between assay variability (within laboratory reproducibility) can be calculated. As a rule the target concentration will be in the low µg/l (ppb) range. The repeatability for reference methods should at this level be better than 20% (% relative standard deviation (RSD)). Methods to be used for homogeneity and stability studies however should be significantly better. The within laboratory reproducibility should be better than 10% RSD. The results for analyses of samples of urine at different temperatures confirm the stability of these reference samples over the duration of the stability study. The stability data indicate that the prepared materials are sufficiently stable to allow their use as reference materials.
3. REFERENCES CONCEPT Title: Preparation and validation of lyophilised Page : 2 of 6 3.1 Multi residue analysis; anabolic agents, RIVM SOP: ARO/113, 97.10.20, revision 5, H.J. v. Rossum, P.W. Zoontjes and P.L.W.J. Schwillens. 3.2 The use of immunoaffinity chromatography in multi-residue and confirmation analysis of ß-agonists in biological samples; proceedings of the EC-workshop organised at RIVM, Bilthoven, The Netherlands; April 1991. L.A. v. Ginkel, R.W. Stephany. 3.3 Multi residue method using coupled-column HPLC and GC-MS for the determination of anabolic compounds in samples of urine. RIVM report nr.389002-035, September 1997, H.A. Herbold, S.S. Sterk, R.W. Stephany, L.A. v. Ginkel. 3.4 Control of reference materials, SOP ARO/429, 11.05.1998, S.S. Sterk. 3.5 Development of bovine muscle, liver and urine reference materials for zeranol, preparation, homogeneity and stability. M.O Keefe, A.Nugent e.a., J. Anal. Chem. (1997) 357: 1029 1034. 3.6 General overviews and sources of reference materials. Quality criteria for residue analysis and reference materials. Relationships between legal procedures and materials. Fresenius J. Anal. Chem. 338: 370-377. R.W. Stephany and L.A. v. Ginkel, (1990). 3.7 Bank of reference samples of blank urine from livestock. Fresenius J.Anal. Chem. (1998) 360:450-455, S.S. Sterk, F. v. Tricht, A. v. Soeren-Kieft, H.A. Herbold, R.W. Stephany and L.A. v. Ginkel. 3.8 Statistics for Analytical Chemistry, J.C. Miller, J.N. Miller, Ellis Horwood serie, 1992 (Second Edition). 4. MATERIALS AND EQUIPMENT Reference to a company and/or product is for purposes of identification and information only and does not imply approval or recommendation of the company and/or the product by the National Institute of Public Health and Environment (RIVM) to the exclusion of others which might also be suitable. Besides standard laboratory glassware and equipment is used: 4.1 Ampoules 83327, Müller 10 ml. 4.2 Ampoules 80391, Müller 15 ml. 4.3 Balance AE200, Mettler, and a PC. 4.4 Lyophilisor Delta II (Christ). 4.5 Filter, glass fibre? 135 mm, no 421057, Schleicher & Schuell. 4.6 Filter, separating gauze TG100,? 130 mm, no 423054, Schleicher & Schuell. 4.7 Membrane filter 0.45? m,? 142 mm, no10 404031, Schleicher & Schuell. 4.8 Membrane filter 0.2? m,? 142 mm, no 404131, Schleicher & Schuell. 4.9 Metal container, plus two metal gauze filters (large and fine)? 150 mm, available 1/10/90, Schleicher &Schuell. 4.10 Automatic Sample Preparation (ASPEC )XL, Gilson, The ASPEC system is programmed for filling vials. 4.11 Acetic acid, Baker, 6152. 4.12 Sodium hydroxide, Baker, 0402.
CONCEPT Title: Preparation and validation of lyophilised Page : 3 of 6 5. ANALYTICAL PROCEDURES. 5.1 Preparation of sample materials. Blank urine has to be mixed with incurred samples of urine, based on the results of the quantitative analyses and the desired concentration of the reference materials. 5.2 Filtration of the urine. It s important to filter the mixed urine through filters of different sizes, to remove solid contaminants from samples, e.g. hair, excrements. Generally a set filters consist of two glass filters (4.5), one membrane filter of 0.45?m (4.7), a fine membrane filter of 0.20?m (4.8) and two metal gauze filters (fine and large) (4.9), under positive pressure and in a metal container. Figure 1 shows the right order of the different filters. Figure 1. 5.3 Filling of the vials. The amber glass vials are filled with aliquots of urine by using the Aspec (4.10). The Aspec is programmed using a self-made program in basic, look further for instructions the manual of Aspec. In order to check the variation of the filled out amount and to control the Aspec, randomly weigh 10% of the total batch and after filling (up to a maximum of 20) and calculate the net fill weight. The coefficient of variation (CV) of the weight has to be? 1 %. After filling, the vials are fitted with a rubber stopper with holes (Figure 2). Figure 2. 5.4 Lyophilisation. Lyophilisation is performed with a Christ Epsilon 2 freeze-dryer. The lyophilisation process is started when the temperature of the sample material is below the eutectic temperature. For information about the operation of the freeze-dryer, see ARO/SOP 404. When the lyophilisation process is finished, the chamber is filled with nitrogen and the vials are closed automatically by pushing the stoppers (Figure 2) in the vials. The vials are fitted with an aluminium cap and labelled. The minimum information on the label, is name and sample number. The vials are stored in the dark at a temperature of 4 ± 2ºC.
6. QUALITY CONTROL. 6.1 Homogeneity study. CONCEPT Title: Preparation and validation of lyophilised Page : 4 of 6 6.1.1 Repeatability of the amount. Based on 20 ml vials (4.2), suitable for the lyophilisation of 10 ml aliquots of urine, a conventional lyophilisor can process approximately 600 till 800 vials on 4 different plates, yielding four batches A, B, C and D. From each batch, randomly 10% (up to a maximum of 20 vials) are selected for analyses to determine the batch homogeneity. In order to check the coefficient of variation (CV) of the weight of the lyophilised contents and to control the Aspec system (4.10): randomly (7.1.1) weighed 10% of the total batch (up to a maximum of 20), and after lyophilisation (6.4). Calculate the average weight and CV of the residue, criteria of the CV must be less than 2.5 %. 6.1.2 Significance tests. Analyse the concentration of the analyte in the vials and after lyophilisation for at least five vials. Calculate the yield of the analyte after lyophilisation process. Calculate the average value of the added amount and CV. With a significance test (T-test, see ref. 3.8) the influence of the lyophilisation process on the concentration is tested. The average amount after lyophilisation must be compared with the average amount lyophilisation, therefore the null hypothesis: X after? X The significance test (T-test) is useful to test whether or not X _ differs significantly from zero. The quantity calculated (T) is given by: T? X _ n after? Std after?? The mean of the differences between X after? X is X after?? The standard deviation of the differences is Std after? The null hypothesis is adopted that the variances are equal. In this case there are no influences of the lyophilisation process on the amount. If the null hypothesis is true then the variance ratio should be close to 1. Differences from 1 occur because of random variation, but if the difference is too great it can no longer be attributed to this cause, the null hypothesis is rejected. The critical values for T are a function of the number of degrees of freedom n (= no of analyses -1) and are listed in a table (as table A.1, page 216 of ref. 3.8). With a significance test (F-test, see ref. 3.8) it is tested the influence of the lyophilisation process on the standard deviation. The experimental variance (STD after lyophilisation) must be compared with the expected variance (STD lyophilisation), therefore the null hypothesis:? Std? after?? Std? must be tested, the only relevant alternative is:? Std? after?? Std? The F-test considers the ratio of the squares of the standard deviations. The quantity calculated (F) is given by: 2? F??? Stdafter?? Std? 2
CONCEPT Title: Preparation and validation of lyophilised Page : 5 of 6 The critical values for F are a function of the number of degrees of freedom n (= no of analyses -1) and are listed in a table (as table A3, page 217 of ref. 3.8). All other vials are stored at -70 o C until the stability of the materials at higher temperatures is adequately demonstrated. 6.2 Stability Study. 6.2.1 The degradation process. 1 Samples are selected randomly from all subbatches in sufficient number to identify each subbatch. Determine the concentration of the analyte over time at different temperatures to assess the degradation process. 6.2.2 Example of stability study. A batch of 40 vials is stored for a period of 1, 3, 6, 8 and 12 months. The ampoules are divided into four groups of different temperatures: - 10 ampoules at -70 ± 2ºC, - 10 ampoules at 4 ± 2ºC, - 10 ampoules at 20 ± 2ºC, - 10 ampoules at 37 ± 2ºC. After each storage time of 1, 3, 6, 8 and 12 months, 2 ampoules are removed at the storage location and stored at -70ºC. Determine the concentration of the analyte after 12 months and calculate the amount by a calibration curve and plot the results (amount against time). From these results can be determined the degradation process at the applied storing conditions. During the stability study the tolerance in temperature is? 3 o C. All samples are stored under controlled temperature conditions and under the exclusion of light. Other deviations are allowed provided that duration and temperature are recorded and included in the final evaluation of the stability. The protocol is summarised in a stability test sheet, identified by an experiment identification number. 1 In general decomposition at elevated temperatures can be described as a first order kinetic process. For each temperature a least squares regression analysis is made of: Ct C0? a? ln? t? C t = concentration at time t, C o = concentration at time t = 0, t i = time The value of a is related to the rate constant for decomposition? ug / l month?? a? k / The significance of k can be tested by testing (t-test as above) whether a deviates significantly from zero (ratio b yx /s a ).If significant rate constants are detected it is possible to combine the different values and come to a general description of the decomposition (Arrhenius): k = a x e [-E a /RT] This relationship between k and t can be linearised: lnk = lna - E a /RT which equals lnk =- (E a /R) x (1/T) + lna a linear relation between 1/T and lnk. If for a particular material A and E a are determined the rate constant k can be calculated at any desired temperature and stability can be estimated, e.g. in terms of half-life in relation to storage temperature. Especially for unstable materials this is of importance. i
CONCEPT Title: Preparation and validation of lyophilised Page : 6 of 6 6.2.3 Example-form: Stability test on Reference Material ARO D-&G : Analyte : Lot : Target concentration : Starting date (t=o) : t temperatures Date Month - 70?C - 20?C +4?C + 20?C + 37?C 1 3 6 8 12 Storage location: -70 o C : -20 o C : +4 o C : +20 o C : +37 o C : Storage data: