general enquiries on this form should be made...

General enquiries on this form should be made to:Defra, Science Directorate, Management Support and Finance Team,Telephone No. 020 7238 1612E-mail: [email protected]

SID 5 Research Project Final Report

SID 5 (Rev. 3/06) of 17

NoteIn line with the Freedom of Information Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The SID 5 (Research Project Final Report) is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra website. A SID 5 must be completed for all projects.

This form is in Word format and the boxes may be expanded or reduced, as appropriate.

ACCESS TO INFORMATIONThe information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors.

Project identification

1. Defra Project code VM02139

2. Project title

Development and Validation of Confirmatory Methodology for the Determination of Aminoglycoside Antibiotics by LC-MS/MS

3. Contractororganisation(s)

Central Science LaboratorySand HuttonYorkYO41 1LZ

54. Total Defra project costs £ 241,277.00(agreed fixed price)

5. Project: start date................ 01 January 2004

end date................. 31 March 2007

SID 5 (Rev. 3/06) of 17

6. It is Defra’s intention to publish this form. Please confirm your agreement to do so...................................................................................YES NO (a) When preparing SID 5s contractors should bear in mind that Defra intends that they be made public. They

should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow.Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the SID 5 can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer.In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.

(b) If you have answered NO, please explain why the Final report should not be released into public domain

Executive Summary7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the

intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work.BackgroundThe Central Science Laboratory (CSL) was commissioned to undertake the ‘Development and Validation of a Confirmatory Method for the Determination of Aminoglycoside Antibiotics by LC-MS/MS’ in line with the requirements of Commission Decision 2002/657/EC. The original plan required the method to be validated for bovine muscle at CSL. However, during the life of the project, Defra requested that the method should be extended to include:

a related compound (spectinomycin), an additional matrix (bovine kidney), and an inter-laboratory validation of the procedure by a 2nd Laboratory (LGC Ltd).

ObjectivesThe main project objectives were:

To establish suitable Liquid Chromatography-Mass Spectrometry/Mass Spectrometry (LC-MS/MS) conditions, capable of confirming seven aminoglycosides (apramycin, gentamicin, kanamycin, neomycin, paromomycin, streptomycin and dihydrostreptomycin) in extracts from bovine milk and muscle at 25 µg kg-1 or less (1/2 the lowest MRL).

To develop and optimise an extraction procedure in order to achieve recovery and resolution of all seven aminoglycosides and spectinomycin (and to ensure interfering co-extractants for both bovine milk and muscle are minimised).

To validate the quantitative LC-MS/MS based confirmatory method in accordance with the Commission Decision 2002/657/EC.

To submit the developed methodology for publication in a peer-reviewed journal and provide the methodology in an internationally agreed format for possible distribution to other laboratories.

ResultsA robust confirmatory method has been developed for all target analytes. This method employs weak cation exchange for sample clean-up followed by LC-MS/MS to quantify and identify all seven aminoglycosides and spectinomycin in bovine muscle and kidney. The final method was validated according to Commission Decision 2002/657/EC [1] including:

Determination of CCα/β (decision limit/detection capability)

SID 5 (Rev. 3/06) of 17

and an assessment of: Method specificity Method selectivity Method ruggedness Analyte stability during storage

A number of critical steps in the method were identified including; the pH of the extract, the make/type of Solid Phase Extraction (SPE) cartridge used, and the temperature used during solvent evaporation. The final method was demonstrated to be selective to the target analytes with good intra-laboratory repeatability.

An assessment of the method’s applicability to other matrices was also performed. These data showed that this procedure is also capable of quantifying and confirming the presence of aminoglycosides in milk and honey.

A Standard Operating Procedure (SOP) was prepared in ISO 78/2 format and distributed to a number of other National Reference Laboratories for evaluation. These inter-laboratory assessments helped to prove the effectiveness and transferability of the developed method.

Options for future work include the extension of the confirmatory method to other matrices (e.g. eggs) and the inclusion of additional aminoglycoside compounds into the multi-residue procedure e.g. amikacin.

Project Report to Defra8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with

details of the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include: the scientific objectives as set out in the contract; the extent to which the objectives set out in the contract have been met; details of methods used and the results obtained, including statistical analysis (if appropriate); a discussion of the results and their reliability; the main implications of the findings; possible future work; and any action resulting from the research (e.g. IP, Knowledge Transfer).

The scientific objectives were as follows;

1. To establish suitable LC-MS/MS conditions, capable of confirming seven aminoglycosides (apramycin, gentamicin, kanamycin, neomycin, paromomycin, streptomycin and dihydrostreptomycin) in extracts from bovine milk and muscle at 25 μg kg-1 or less (1/2 the lowest MRL).

2. To develop and optimise an extraction procedure in order to achieve recovery and resolution of all seven aminoglycosides of interest and spectinomycin and ensure interfering co-extractants for both bovine milk and muscle are minimised.

3. To validate the quantitative LC-MS/MS based confirmatory method in accordance with the Commission Decision 2002/657/EC

4. To submit the developed methodology for publication in a peer-reviewed journal and provide the methodology in an internationally agreed format for possible distribution to other laboratories.

Progress against each objective is reported below:

SID 5 (Rev. 3/06) of 17

1. To establish suitable LC-MS/MS conditions, capable of confirming seven aminoglycosides (apramycin, gentamicin, kanamycin, neomycin, paromomycin, streptomycin and dihydrostreptomycin) in extracts from bovine milk and muscle at 25 μg kg-1 or less (1/2 the lowest MRL).

This objective has been fully met and was reported to Defra in CSL REPORT FD 04/33 - Development and Validation of a Confirmatory Method for the Determination of Aminoglycoside Antibiotics by LC-MS/MS (First Milestone Report – issued February 2005).

In summary, initial efforts concentrated on identifying appropriate HPLC conditions (column and eluent programme) that (i) retained and separated the seven aminoglycosides and spectinomycin and (ii) were amenable for use with a mass spectrometer. HPLC conditions were developed and optimised in respect of peak shape, analysis time and resolution. The assessments included an examination of the influence of solvent composition, ion-pair concentrations and pH on both the chromatography and mass spectrometry. Both electrospray (ES) and atmospheric pressure chemical ionisation (APCI) were assessed and compared with respect to sensitivity, selectivity and suitability for quantification of the analytes, based on solvent standards. Product ion spectra resulting from collision-induced dissociation were examined and suitable ions selected for incorporation into multiple reaction monitoring (MRM) schemes in order to meet the criteria defined within Commission Decision 2002/657/EC. The transition ions selected are shown in Table 1. Table 1. LC-MS/MS transition ions for aminoglycosides and spectinomycin (in chromatographic retention time order).

Compound QuantificationChannel

(m/z)

ConfirmationChannel(s)

(m/z)Dihydrostreptomycin 584>263 584>221Streptomycin 582>263 582>221Spectinomycin 333>189 333>98Paromomycin 616>163 616>161

616>293Apramycin 540>378 540>217Kanamycin 485>163 485>324Gentamicin C2/C2a 464>322 464>160Gentamicin C1a 450>322 450>160Neomycin 615>161 615>163

615>203Gentamicin C1 478>157 478>322

This work demonstrated the capability of LC-MS/MS to detect and confirm the identity of the target aminoglycosides at concentrations equivalent to one half of the lowest MRL (i.e. 25 μg kg-1).

2. To develop and optimise an extraction procedure in order to achieve recovery and resolution of all seven aminoglycosides of interest and spectinomycin and ensure interfering co-extractants for both bovine milk and muscle are minimised

This objective has been met in full.

Using the developed LC-MS/MS conditions from Objective 1 a number of Solid Phase Extraction (SPE) clean extraction cartridges were evaluated with respect to analyte recovery from a range of sample matrices. The most suitable SPE cartridge was then used to develop the required confirmatory method. Details of method development phases are reported in the text below. The final validated method, in schematic format, is reproduced in the following embedded “pdf” document (double click to open).

During method development several aspects of the method were investigated and optimised. These experiments were undertaken using blank tissue (matrix) extracts spiked at or around the Maximum Residue Limit (MRL) as detailed in Table 2.

SID 5 (Rev. 3/06) of 17

Table 2. MRL’s or typical spiking concentrations used during method development.

Analyte MarkerMilk Muscle Kidney Honey

MRL (2377/90/EC) or spiking concentration , μg kg-

1

Dihydrostreptomycin DHS 200 500 1000 20Streptomycin Streptomycin 200 500 1000 20

Spectinomycin Spectinomycin 200 300 5000 300Paromomycin Paromomycin 500 500 1500 500

Apramycin Apramycin 1000 1000 20000 1000Kanamycin Kanamycin 150 100 2500 100Neomycin Neomycin B 1500 500 5000 500

Gentamicin Sum of C1, C1a, C2 and C2a 100 50 750 50

The individual parameters investigated included:

I. Extraction conditions (e.g. the possible need for exhaustive (sequential) extractions to maximise analyte recovery).

II. Sample clean-up conditions (e.g. pH and selection of SPE cartridges).

III. Approaches required for analyte quantification (e.g. the use of a series of “matrix-extracted” calibration standards versus the use of “matrix matched” calibrants and calibrants in solvent).

The main findings from each of these experiments are summarised below:

2.1 Extraction conditions (see step 3 in method schematic).

In this experiment a comparison of analyte recovery was made between the use of two sequential extractions (2 x 10 mL phosphate buffer) compared to a single extraction (1 x 20 mL of phosphate buffer). For the majority of compounds [streptomycin, dihydrostreptomycin, paromomycin, aparamycin, kanamycin, neomycin, gentamicin] there was no obvious difference between the two extraction procedures. However, for spectinomycin (which was included as an extension to the original proposal) a large reduction in measured concentration was observed when a single extraction was used (Table 3). This is indicative of a lower recovery for this analyte.

Table 3. Effect of extraction process on analyte recovery from bovine muscle. Measured concentration μg kg-1 (mean, n=5)

Analyte extraction conditionDouble (2 x 10 mL) Single (1 x 20 mL)

Streptomycin 640 540Dihydrostreptomycin 690 596Spectinomycin 346 140Paromomycin 646 605Apramycin 1110 1082Kanamycin 141 143Neomycin 668 693Gentamicin C1 69 65Gentamicin C2* 64 68Gentamicin C2A* 60 68Gentamicin C1A 52 62

Note (*) in some batches the gentamicin C2 and C2a isomers could be separated on the chromatographic column although this was not always the case. The MRL relates to the sum of all isomers. These measurements were made using matrix-matched standards (see section 2.4) .

Whilst the recovery of spectinomycin could be improved by the use of a second extraction this significantly increased the time taken to analyse a batch of samples. Therefore, to maintain cost-effectiveness of the final method, a single extraction was selected for all further work. The use of “matrix extracted” calibration standards

SID 5 (Rev. 3/06) of 17

(see paragraph 2.4 below) ensured that, whilst the analyte recovery for spectinomycin was relatively low, the absolute analyte concentration in a sample could be reliably quantified.

2.2 Sample clean-up ( see steps 8, 9 10 in method schematic): Prior to sample clean-up by weak cation exchange SPE the pH of the sample extract must be adjusted. In order to establish a suitable tolerance for this pH a series of experiments were conducted. Using bovine muscle extract, the pH was adjusted to half unit ranges (between 6.5 and 8.5) immediately prior to the SPE clean-up. A summary of these results are tabulated below (Table 4).

Table 4. Effect of varying pH of sample extract from bovine muscle prior to SPE.

Sample Identity MRL μg kg-1

pH 6.5 - 7.0 pH 7.0 - 7.5 pH 7.5 - 8.0 pH 8.0 - 8.5Mean conc.

μg kg-1

(n=6) %CV

Mean conc.

μg kg-1

(n=6) %CV

Mean conc.

μg kg-1

(n=6) %CV

Mean conc.

μg kg-1

(n=6) %CV

Dihydrostreptomycin 500 735 6 733 3 676 3 612 4Streptomycin 500 583 5 561 6 524 4 487 4Spectinomycin 300 198 20 320 6 303 6 323 8Paromomycin 500 568 20 634 5 612 9 564 7Apramycin 1000 1098 24 1264 9 1189 10 1067 13Kanamycin 100 96 27 110 12 110 10 105 15Neomycin 500 455 30 516 13 500 16 504 20Gentamicin 50 58 20 59 4 57 13 51 21

A greater degree of analytical variation (%CV) was observed in those replicates adjusted to pH 6.5-7.0 and pH 8.0-8.5. Therefore, a range in pH of 7.0-8.0 was considered optimal for this step in the method. This relatively wide tolerance means that pH indicator strips can be used to control this step, eliminating the need for a pH meter and thus improving method efficiency.

2.3 Evaluation of SPE cartridges and elution conditions ( see steps 11 - 17 in method schematic) The final method uses SPE based on weak cation exchange (BakerBond CBX). However, during the execution of this project, delays were experienced with obtaining supplies of this SPE cartridge from the manufacturer. To reduce the method’s reliance on this supplier “equivalent” products from alternative manufactures were evaluated. These cartridges were substituted directly into the method with no additional ‘method optimisation’. Additionally cartridges with an alternative retention mechanism, i.e. Hydrophilic Interaction Liquid Chromatography (HILIC), were assessed using test conditions provided by the manufacturer of these columns. Details of the SPE cartridges tested are as follows:

Weak cation exchanger – silica based Bakerbond Wide pore, CBX, 500 mg, 6 mL [part no 7217-16] (used in step 11 of final method) Waters Sep-Pak, Vac CM, 500 mg, 3cc [part no WAT020855] IST Isolute CBA, 500 mg, 6 mL [part no 520-0050-C] Supelco superclean LC-WCX, 3 mL [part no 57061] Varian HF Bond Elute CBA 500mg, 3 mL [part no 14102038]

Weak cation exchanger - polymeric Waters Oasis WCX 150 mg, 6cc, 30 µm [part no 186002498]

Alternative retention mechanism Sequant ZIC®-HILIC, 500 mg, 3 mL, [part no 2942-051]

The results of these evaluations are summarised in graphical format below (Figure 1).

Figure 1. Comparison of analyte recovery from different SPE cartridges.

SID 5 (Rev. 3/06) of 17

Figure 1 shows that, when a single acidic elution step is employed (step 16 in method), the optimum SPE cartridge for use is the Bakerbond CBX with recoveries > 60 % for all analytes. Recovery of each analyte is achieved using the Varian CBA cartridge but, with many recoveries < 30 %, this was not deemed adequate for use within this project without further optimisation.

Some additional method development was then undertaken and the elution from the SPE cartridge was further investigated for the one polymeric and four out of five of the silica based cation-exchange cartridges. In this case a two-stage elution procedure was employed [10 % ammonia in methanol (3 mL) followed by 10 % acetic acid in methanol (3 mL)]. These data are presented in Figure 2.

Figure 2. Comparison of analyte recovery from different SPE cartridges using a two-stage elution.

In comparison to the single step elution there was a notable increase in analyte recovery for the majority of analytes for the alternative SPE cartridges. In contrast the BakerBond CBX cartridge showed a decrease in analyte recovery for some analytes. Despite the improved data from the alternative SPE cartridges, these results indicated that the original BakerBond CBX column (and a single elution stage with 10% v/v acetic acid in methanol) remained the preferred option for sample clean-up.

SID 5 (Rev. 3/06) of 17

One slight disadvantage of using 10% v/v acetic acid in methanol to elute extract from the SPE was the time taken to evaporate samples prior to determination by LC-MS/MS (step 17 in method). In order to reduce this time, and thus improve the sample throughput, other acid/methanol combinations were investigated. This study used the more volatile formic acid at three different concentrations (5 %, 10 and 15 % v/v formic acid in methanol). Data from this experiment are presented in Table 5.

Table 5. Variation in SPE elution solvent: Effect on method quantification / repeatability using bovine muscle extracts.

Compound MRL μg kg-1

10% v/v Acetic acid

5% v/vFormic acid

10% v/v Formic acid

15% v/v Formic acid

Mean conc.µg kg-1

(n=5) %CV

Mean conc.µg kg-1

(n=5) %CV

Mean conc.µg kg-1

(n=5) %CV

Mean conc.µg kg-1

(n=5) %CVDihydrostreptomycin 500 501 5 502 6 523 6 542 9

Streptomycin 500 488 4 527 7 520 6 530 8Kanamycin 100 109 5 100 11 104 6 106 6

Paromomycin 500 497 8 468 17 504 11 534 9Apramycin 1000 1030 8 900 18 977 11 1032 8

Gentamicin C1a

50

53 10 51 23 52 15 56 9Gentamicin C2 51 10 51 25 54 16 57 8Gentamicin C1 45 6 46 25 51 16 56 13

Gentamicin C2a 49 9 45 18 49 20 53 10Neomycin 500 485 6 458 20 498 15 542 10

This experiment showed that method repeatability was slightly better when using 10% v/v acetic acid in methanol as the SPE elution solvent. Therefore, as the sample evaporation time was found to be only slightly faster when formic acid was employed, the 10% v/v acetic acid in methanol was retained as the preferred SPE elution solvent.

Finally, a study was conducted to investigate the possible need to further condition the BakerBond CBX columns with a buffer solution before loading sample extract (i.e. between steps 11 and 12 in the method schematic). These experiments showed that sequential column conditioning with methanol and then water resulted in good quality data. Therefore, additional column conditioning stages with buffer solution(s) were considered unnecessary.

2.4 Evaluation of procedures for method calibration (comparison of matrix extracted calibration standards versus matrix matched calibrants and calibrants in solvent)

The measurement of analytes by LC-MS/MS is often affected by the presence of co-extractives or “matrix effects”. To evaluate this parameter different calibration standards were produced so that solvent standards could be compared to “matrix-extracted” and “matrix-matched” calibration standards in bovine kidney, muscle and milk. “Matrix-matched calibration standards” were prepared by extracting and cleaning-up a known blank sample using the developed SPE method and then spiking the final sample extract immediately prior to measurement. When these data are compared to corresponding solvent standards the “matrix suppression “of the LC-MS/MS measurement can be evaluated. “Matrix-extracted calibration standards” were prepared by spiking a series of known blank samples with increasing amounts of aminoglycosides prior to extraction. Calibration curves prepared using this method will automatically correct the data for analytical recovery and matrix suppression. Also the true recovery for the method can be estimated by dividing the LC-MS/MS response (peak area) obtained for the matrix-extracted calibration standard by the equivalent matrix-matched calibration standard.

An example of these data is shown below (Figure 3). Results for all analytes/matrices are presented in Appendix 1.

Figure 3. Comparison of LC-MS/MS response (peak area) for three types of method calibration.

SID 5 (Rev. 3/06) of 17

The majority of analytes showed the same pattern between the different calibration curves. As expected a greater instrument response is observed for solvent calibrants compared to matrix-matched equivalents, indicating matrix associated ion-suppression effects in the LC-MS/MS. For this reason, it was concluded that calibration curves must be constructed using matrix. However, to correct for analyte recovery and ion-suppression at the same time, it was decided that “matrix-extracted calibration standards” were the most suitable approach for this method. This approach for analyte calibration was of particular benefit when quantifying spectinomycin due to its lower recovery (when compared to the aminoglycosides).

[Note: Isotopically labelled internal standards could also be used to compensate for analyte recovery and ion suppression but they are not currently available for the compounds included in this project].

Objective 3. To validate the quantitative LC-MS/MS based confirmatory method in accordance with the Commission Decision 2002/657/EC.

This objective has been achieved in full.

The aim was to validate the quantitative LC-MS/MS based confirmatory method (from objective 2) in accordance with Commission Decision 2002/657/EC for both bovine muscle and bovine kidney. The validation exercise thus included:

I. Determination of CCα/CCβ and repeatabilityII. RuggednessIII. Specificity IV. Analyte stability, andV. An inter-laboratory assessment of the method at LGC Ltd.

3.1 Determination of CCα/CCβ (Decision limit/Detection capability)

Commission Decision 2002/657/EC permits a number of approaches for determining CCα and CCβ. In this study we chose to employ the calibration curve procedure according to ISO 11843 (2). In this case blank material is to be fortified around the permitted limit (i.e. MRL) in equidistant steps. Therefore, CCα and CCβ measurements in both bovine muscle and bovine kidney were undertaken by extracting seven blank samples spiked at 3 concentrations (0.5, 1.0 and 1.5 times MRL) on each of 3 separate days. Tables 6 and 7 present the data obtained for bovine muscle and kidney respectively.

SID 5 (Rev. 3/06) of 17

Table 6. CCα and CCβ concentrations (µg kg-1) obtained by CSL for bovine muscle.

Compound CC CCMuscle MRL

(g kg-1)

Dihydrostreptomycin 541 581 500

Streptomycin 553 607 500Spectinomycin 680

(356*)1059(412*)

300

Paromomycin 554 608 500

Apramycin 1127 1254 1000

Kanamycin 117 134 100

Neomycin 614 729 500

Gentamicin 66 81 50*Data produced excluding Day one data (see below)

Table 7. CCα and CCβ concentrations (µg kg-1) obtained by CSL for bovine kidney.

Compound CC CCKidney MRL

g kg-1


Streptomycin 1104 1208 1000

Spectinomycin 6868 8735 5000


Apramycin 23318 26635 20000

Kanamycin 2939 3379 2500

Neomycin 6617 8234 5000

Gentamicin 909 1067 750

The above data demonstrate that the developed method is capable of quantifying and confirming all target analytes and that the CC values are typically within 10 – 20 % of the MRL concentration for each analyte/matrix combination. This is typical for a method that is in control.

The data obtained for spectinomycin was more variable with CC values of >140 % of the MRL. The exact cause of the variability for this analyte is not known. However, the known lower recovery for this analyte may have had an effect on overall method precision. To further investigate this issue both the intra- and inter-batch data used to calculate CC were evaluated (see Tables 8 and 9 below). Data obtained for spectinomycin in bovine muscle during Day 1 was atypical with %CVs of 30 – 53%. The cause of this extra variation on Day 1 is unknown and is difficult to explain as these experiments were conducted by spiking samples with mixed solutions containing all analytes. Since there was no known scientific reason to exclude data, the CCfor spectinomycin in bovine muscle of 680 µg kg-1 (Table 6) was calculated using all 63 data points from Table 8. However, for comparison, an additional CCwas calculated by eliminating all Day 1 data and using the 42 data points from Days 2 and 3. This provided a much improved CCin bovine muscle)of 356 µg kg-1 compared to the MRL of 300 µg kg-1.

Table 8. Bovine muscle intra- and inter-batch repeatability data (µg kg-1)

Dihydrostreptomycin Streptomycin Spectinomycin Paromomycin Apramycin Kanamycin Neomycin Gentamicin

SID 5 (Rev. 3/06) of 17

spike conc. MRL 500 500 300 500 1000 100 500 50

0.5

x M

RL

Day 1 n=7

Mean 256 247 97 248 486 51 256 30%CV 5 7 53 6 6 9 11 7

Day 2 n=7

Mean 255 245 153 242 484 49 249 26%CV 6 5 12 9 8 10 17 15

Day 3 n=7

Mean 283 274 166 258 508 54 259 18%CV 11 9 11 12 11 15 11 8

All data n=21

Mean 260 251 138 248 490 50 246 25%CV 7 7 31 8 7 10 12 23

1.0

x M

RL

Day 1 n=7

Mean 499 516 615 449 889 105 547 50%CV 3 4 50 7 5 7 16 5

Day 2 n=7

Mean 517 487 296 478 956 106 509 49%CV 5 5 5 8 9 10 9 8

Day 3 n=7

Mean 490 459 322 47 920 98 452 31%CV 5 4 13 7 9 12 17 11

All data n=21

Mean 511 505 408 467 938 104 482 43%CV 5 5 56 8 9 9 19 22

1.5

x M

RL

Day 1 n=7

Mean 770 788 619 726 1442 153 756 74%CV 4 7 30 3 3 7 18 4

Day 2 n=7

Mean 775 724 412 717 1435 154 755 72%CV 1 3 4 3 4 4 8 3

Day 3 n=7

Mean 751 703 499 767 1559 156 751 50%CV 9 8 8 13 13 12 15 16

All data n=21

Mean 782 766 509 745 1493 157 729 65%CV 5 7 27 9 9 9 15 19

Table 9. Bovine kidney intra- and inter-batch repeatability data (µg kg-1)

Dihydrostreptomycin Streptomycin Spectinomycin Paromomycin Apramycin Kanamycin Neomycin Gentamicin

spike conc. MRL 1000 1000 5000 1500 20000 2500 5000 750

0.5

x M

RL

Day 1 n=7

Mean 499 513 2276 693 9282 1167 2418 364%CV 5 7 7 4 5 5 11 10

Day 2 n=7

Mean 504 514 2210 713 9494 1206 2517 424%CV 6 6 11 8 8 8 11 11

Day 3 n=7

Mean 541 464 2605 729 12406 1281 2488 335%CV 10 25 13 4 6 5 5 12

All data n=21

Mean 515 497 2363 711 10394 1218 2474 374%CV 8 15 13 6 15 7 9 15

1.0

x M

RL

Day 1 n=7

Mean 1018 1026 4805 1405 20468 2366 4611 802%CV 2 3 5 3 5 7 12 9

Day 2 n=7

Mean 1011 1017 6050 1440 22675 2584 5912 839%CV 4 6 27 7 4 10 22 13

Day 3 n=7

Mean 1046 1020 5123 1668 24314 2785 5355 907%CV 6 9 12 6 6 7 8 8

All data n=21

Mean 1025 1021 5326 1505 22486 2578 5293 849%CV 5 6 21 10 9 10 18 11

1.5

x M

RL

Day 1 n=7

Mean 1349 1228 7790 2071 28992 3402 6296 1044%CV 6 4 6 9 11 11 14 8

Day 2 n=7

Mean 1436 1473 10636 2178 35202 3962 9266 1227%CV 9 10 11 11 11 10 12 7

Day 3 n=7

Mean 1426 1540 6426 2325 3180 3684 7184 1251%CV 10 13 9 9 7 9 8 9

All data n=21

Mean 1404 1414 8284 2192 31998 3683 7582 1174%CV 9 14 24 10 12 11 20 11

3.2 Method Ruggedness

To check method ruggedness and identify any critical points within the developed method several additional experiments were undertaken including:

An evaluation of the solvent evaporation temperature (step 17 in method schematic) An assessment by a second analyst

SID 5 (Rev. 3/06) of 17

An assessment of the methods applicability to additional matrices – milk and honey.

Results of these experiments are reported below:

3.2.1. An evaluation of the solvent evaporation temperature (step 17 in method schematic)

A comparison was made between solvent evaporation temperatures and the recovery (or loss) of analyte. In these experiments multiple extracts were spiked at or around MRL concentrations and then evaporated at three different temperatures (40, 50 and 60 °C). These data (Table 10) showed that, for a number of analytes, there was an apparent decrease in recovered analyte with increasing temperature. Therefore, the Standard Operating Procedure (SOP) includes a warning to keep the evaporation temperature to 40 +\- 5 °C in order to minimise analyte loss at this stage of the analysis.

Table 10. Effect on analyte recovery due to changes in variation in solvent evaporation temperatures (bovine muscle extract).


40°C 50°C 60°CMean conc.

µg kg-1

(n=6) %CV

Mean conc.µg kg-1

(n=6) %CV

Mean conc.µg kg-1

(n=6) %CVDihydrostreptomycin 500 698 8 672 7 671 7

Streptomycin 500 569 6 541 10 497 9Spectinomycin 300 357 6 290 12 208 22Paromomycin 500 486 12 463 8 445 12

Apramycin 1000 966 11 944 8 905 12Kanamycin 100 104 10 96 8 84 16Neomycin 500 576 15 457 17 322 45

Gentamicin 50 52 14 49 11 49 16

3.2.2. An assessment by a second analyst

Batches containing known blank matrix was spiked at three levels (0.5, 1.0 and 1.5 times MRL) were extracted, by a second analyst, for both bovine kidney and muscle. Table 11 provides a summary of the data generated by the second analyst at CSL. These data compare well in terms of precision to that of the primary analyst (see Tables 8 and 9) and confirmed that the method was sufficiently rugged to transfer between analysts within the same laboratory.

Table 11. Repeatability data produced by second analyst (for bovine muscle and kidney).

mat

rix


spike @ 0.5 x MRL spike @ 1 x MRL Spike @ 1.5 x MRLMean conc. measured

(n=7)μg kg-1 %CV

Mean conc. measured

(n=7)μg kg-1 %CV

Mean conc. measured

(n=7)μg kg-1 %CV

Dihydrostreptomycin 500 319 3 584 3 918 3

SID 5 (Rev. 3/06) of 17

mus

cle

Streptomycin 500 252 4 446 2 703 3Spectinomycin 300 189 10 319 9 490 10Paromomycin 500 254 3 459 5 702 3


Gentamicin 50 31 3 53 6 77 6

kidn

ey

Dihydrostreptomycin 1000 393 4 720 8 932 8Streptomycin 1000 591 3 1107 5 1472 8

Spectinomycin 5000 3176 4 5816 8 7530 8Paromomycin 1500 801 4 1606 3 2269 9


Gentamicin 750 425 13 865 9 1145 15

3.2.3. An assessment of the methods applicability to additional matrices – milk and honey

In addition to bovine kidney and bovine muscle, the applicability of the method was also tested for milk and honey. As no MRL’s exist for honey, the spiking levels used for this work are as detailed in the table below (Table 13). Data from these additional experiments (Tables 12 and 13) demonstrated that this confirmatory method is suitable for use for these two additional matrices.

Table 12. Method ruggedness: Data from the analysis of spiked milk samples.

Compound

Milk

MR

L or

(S

piki

ng

conc

entra

tion)

(μg

kg-1)

0.5 x MRL(*) 1.0 x MRL(*) 1.5 x MRL (*)

Mean conc. measured

(n=7)μg kg-1 %CV

Mean conc. measured

(n=7)μg kg-1 %CV

Mean conc. measured

(n=7)μg kg-1 %CV

Dihydrostreptomycin 200 121 4 214 9 307 18Streptomycin 200 94 4 175 11 254 17

Spectinomycin 200 111 16 274 27 411 15Paromomycin (*) 500 253 4 442 6 605 22

Apramycin (*) 1000 521 5 941 8 1282 23Kanamycin 150 80 12 161 12 236 28Neomycin 1500 725 14 1403 19 2180 32

Gentamicin 100 58 4 101 9 137 22(*) No MRL, therefore spiking concentration quoted for Apramycin and Paromomycin

Table 13. Method ruggedness: Data from the analysis of spiked honey samples.

Compound

Honey

Spi

king

leve

l (S

L) μ

g kg

-1 0.5 x SL 1.0 x SL 1.5 x SL 2.0 x SLMean conc. measured

(n=7)μg kg-1 %CV

Mean conc. measured

(n=7)μg kg-1 %CV

Mean conc. measured

(n=7)μg kg-1 %CV

Mean conc. measured

(n=7)μg kg-1 %CV

Dihydrostreptomycin 20 12 2 24 6 34 6 51 6

SID 5 (Rev. 3/06) of 17

Streptomycin 20 10 5 20 8 28 6 41 7Spectinomycin 300 159 3 316 3 445 6 628 3Paromomycin 500 262 3 487 5 675 5 999 13

Apramycin 1000 516 3 908 4 1251 7 1857 16Kanamycin 100 56 4 96 6 132 7 193 15Neomycin 500 273 3 391 10 571 19 891 34

Gentamicin 50 35 1 49 6 67 13 95 22

3.3 Method Specificity

Commission Decision 2002/657/EC states that it is of prime importance that, “interference, which might arise from matrix components, is investigated”. In this study twenty different untreated samples of both bovine muscle and bovine kidney were extracted and analysed using the optimised method. No measurable matrix interferences at the retention times of the eight analytes were observed in the 20 different samples of each matrix - thus demonstrating the applicability and specificity of the method across each matrix type.

3.4 Analyte Stability

Commission Decision 2002/657/EC also requires an investigation of analyte stability in sample, sample extracts and calibration standards in solvent. To meet this requirement a 20-week stability study was conducted whereby analyte in solvent, sample extracts, and spiked animal tissue were tested at regular intervals. These experiments, included:

a) Solvent stability – analytical standards were prepared in solvent (water) at two concentrations; a stock standard at 1 mg ml-1 and a spiking solution for each analyte at a concentration appropriate to their respective MRLs (between 20 and 400 µg ml-1). Aliquots were stored under four conditions – room temperature in daylight, and room temperature, + 4 °C and – 20 °C in the dark between 0 and 20 weeks. At each time point six replicate aliquots were removed from each storage condition, diluted to within the calibration range and measured against freshly prepared solvent standards.

b) Extract stability – at week 0 a series of blank samples were extracted and spiked post extraction at concentrations equivalent to the MRL. These extracts were stored under the same conditions as the solvent stability samples. At each time point an aliquot was removed from each and measured against freshly prepared matrix-extracted standards.

c) Matrix stability – at week 0 a series of known blank animal tissue samples were spiked at MRL concentrations and stored at –20 °C. At each time point six aliquots were extracted and analysed against a matrix-extracted calibration curve (prepared at the time of extraction)

The results from these stability measurements are shown graphically in Appendix 1. These graphs plot the mean measured concentration for each time point (after dilution onto the calibration range) along with error bars (+/- 2 SD). Overall, these data demonstrated that all of the compounds tested are stable for at least four weeks in solvent, matrix extract and spiked bovine muscle and kidney. Some small fluctuations in the concentrations measured do exist after this time point but, by visual inspection, these changes appear to be negligible. Formal statistical analysis would be required to confirm this opinion. This formal assessment of the data is not a requirement of 2002/657/EC but, if required, could be conducted by a statistician at a later date.

Objective 4. To submit the developed methodology for publication in a peer-reviewed journal and provide the methodology in an internationally agreed format for possible distribution to other laboratories.

The majority of this objective has been met.

The method was written in ISO78/2 format and circulated to a number of other National Reference Laboratories (NRLs) within Europe including RIKILT(NL), Tallinn-VFL (Estonia), and LGC Ltd (UK).

SID 5 (Rev. 3/06) of 17

T-VFL has converted this method from an LC-MS/MS method to a UPLC-MS/MS procedure. This NRL has obtained good CC data (data not shown) and has also evaluated the method for another matrix (egg) and an additional aminoglycoside (tobramycin).

The inter-laboratory validation of the method at LGC was delayed until May/June 2007 due to supply problems with the SPE cartridges. Initially LGC also had trouble with analyte “carry-over”. To help solve this issue LGC made some minor changes to the method and divided their validation batches into two sub-sets. LGC’s CCα and CCβ data (as calculated by CSL) are presented in Table 14.

Table 14. CCα and CCβ concentrations (µg kg-1) obtained by LGC Ltd for bovine kidney

Compound CC CCKidney MRL

g kg-1


Streptomycin 1414 1828 1000

Spectinomycin 6394 7787 5000


Apramycin 23653 27306 20000

Kanamycin 3331 4162 2500

Neomycin 6413 7826 5000

Gentamicin 933 1116 750

At the end of their validation study LGC used this new method to successfully confirm the presence of an aminoglycoside in a sample taken under the National Surveillance Scheme – thus demonstrating the importance of method transfer within R&D projects

At the time of writing this report a publication on this method (and associated validation data) has not been prepared. The current plan is for the data from all laboratories to be combined into a comprehensive paper for peer review.

ConclusionsA LC-MS/MS method has been successfully validated in accordance with Commission Decision 2002/657/EC for the quantification and confirmation of the aminoglycosides dihydrostreptomycin, spectinomycin, paromomycin, apramycin, kanamycin, neomycin, gentamicin, and spectinomycin in bovine muscle and kidney. The method is also applicable for the determination of these analytes in milk and honey.

The method is found to be robust to a range of small changes and the critical control points have been identified and highlighted within the Standard Operating Procedure. The method was shown to be repeatable (intra-laboratory reproducibility data) and was successfully transferred and validated by a number of other laboratories using different instrumentation.

An SOP has been produced in ISO 78/2 format and distributed to several laboratories in anticipation of a cross-laboratory validation and improvements of the method, which will in turn lead to a publication in an appropriate journal.

Further WorkOpportunities for further work include:

Continued investigation into the use of alternative SPE cartridges including full optimisation of the SPE conditions (load, wash and elution steps),

Determination of CC in other matrices eg honey and poultry products including egg, Extension of the list of analytes to include other aminoglycosides e.g. amikacin, hygromycin, and

sisomycin).

Appendix 1. (double click to open)

SID 5 (Rev. 3/06) of 17

References to published material9. This section should be used to record links (hypertext links where possible) or references to other

published material generated by, or relating to this project.1. Commission Decision 2002/657/EC of 12 August 2002 implementing Council Directive 96/23/EC concerning the performance of analytical methods and the interpretation of results. Official Journal of the European Communities L22,1 8-36.

2. International Organisation for Standardisation (ISO 11843) http://www.iso.org/iso/en/CatalogueDetailPage.CatalogueDetail?CSNUMBER=1096

3. Poster submission to EuroResidue VI, May 2008 (in collaboration with RIKILT)

SID 5 (Rev. 3/06) of 17

http://www.iso.org/iso/en/CatalogueDetailPage.CatalogueDetail?CSNUMBER=1096

general enquiries on this form should be made...

Documents