LC-MS/MS Methods for the Effective Control of Veterinary Drugs in Raw Materials and Manufactured Products. A Food Industry Perspective Thomas Bessaire, Adrienne Tarres, Andrea Beck, Claudia Mujahid, Marie-Claude Savoy Perroud, Lucie Racault, Pascal Mottier, Cindy Bion, Thierry Delatour and Aurélien Desmarchelier. Nestlé Research Center, Lausanne, Switzerland North American Chemical Residue Workshop Naples (Florida, USA), July 24 th 2017
Control of Veterinary Drugs in Nestle Products is of Utmost Importance 1 billion Products sold Every Day 418 Factories 25 State-of-the-Art equipped Laboratories 200 000 Analyses per Day Source: Nestlé Annual Review, 2016 Factory Farms Worldwide Regulations New Communication Channels Antimicrobial Resistance Time to achieve 50 millions users (years) 40 20 EU, China, USA, Canada, Codex Alimentarius, etc. 2 Sources: Le Monde, L atlas du monde de demain, 2013 Le Monde, Dossiers & Documents, 427, 2013 0 1922 1936 1993 2001 2006 Year of launch
Effective Control of Vet. Drugs across the Nestlé Supply Chain 1- Background and Needs
Quality Testing along the Nestlé Supply Chain & Manufacturing Confirmatory Methods for Full Compliance Factory Lines Finished Product Warehouse Factory Raw Material Warehouse RM Specification & CoA External Supplier(s) Raw Material Collection Centers Rapid Methods for Effective Release Farms 4 The right method at the right place Mass spectrometry mandatory for full compliance testing
Regulatory Framework with 198 compounds 165 regulated 152 68 52 11 EU, China, USA, Canada, Codex Alimentarius, etc. 5 Regulations (Maximum Residue Limit MRL) drive the requirement in terms of limit of detection.
Relevant Food Commodities for Nestlé Milk-based Products Meat / Fish-based Products The «Usually-shown» Matrices Raw milk Fresh or cooked meat, fish and seafood The «Forgotten» Matrices Milk fractions e.g. skimmed milk powder, whey protein concentrate, lactose etc. Formulae with milk e.g. infant formulae, hypoallergenic formulae, etc. Infant cereals with milk Meat, fish and seafood powder e.g. shrimp, duck, pork, lamb, beef, etc. Infant cereals with meat tissues Baby foods in jar based on vegetables, meat/fish, pasta, cereals, etc. Fat / egg-based items In addition to common raw materials, Nestlé approach has to include complex matrices, e.g. processed ingredients and finished products. 6
Occurrence of Veterinary Drugs in Food In 2015 (FDA), 15 confirmed positive milk samples out of 1 912 samples tested. 0.78% 0.37% In 2014 (EFSA), 1 500 non-compliant samples out of over 425 000 samples. A high-throughput screening method to check sample compliance is well adapted to deal with shorter and shorter turn-around times. 7
Effective Control of Vet. Drugs across the Nestlé Supply Chain 2- Literature Review 1- Background and Needs
Mass Spectrometry Methods for Veterinary Drugs Publication year 2016 2011 2006. 245 articles in peer-reviewed scientific journals (1997-2016) 2001 9 1996 10 20 30 40 Number of publications
Mass Spectrometry Methods for Veterinary Drugs Publication year 2016 2011 2006. 245 articles in peer-reviewed scientific journals (1997-2016) 2001 10 1996 0 50 100 150 200 Number of analytes
Mass Spectrometry Methods for Veterinary Drugs Publication year 120 analytes 2016 2011 2006 only 15 methods capable to analyze 120 analytes and more 2001 11 1996 0 50 100 150 200 Number of analytes
Mass Spectrometry Methods for Veterinary Drugs Publication year 120 analytes 2016 2011 2006 Chen et al., 2016, J. Chromatogr. B LC-QqQ Egg, Milk, Muscle Robert et al., 2013, Food Addit. Contam. A LC-QqQ Egg, Honey, Milk, Muscle and only 2 methods applicable to at least 3 matrices! 12 2001 1996 An unique MS-based method demonstrating full compliance of vet. drugs in food does not exist so far 0 50 100 150 200 Number of analytes
Effective Control of Vet. Drugs across the Nestlé Supply Chain 3- Analytical Challenges 2- Literature Review 1- Background and Needs
Stream 1 Stream 2 Management of Standard Mixes in Routine Environments Preparation and management of standard mixes = time-consuming task Standard Collection, CoA checking, weighing, solubilization, mixing, storage, etc. Systematically challenged during ISO 17025 accreditation and GLP assessment audits Stability data are required by guidelines for validation, e.g. EU 2002/657/EC or SANTE/11945/2015. Stability data in multiresidue approaches very often underestimated: High demanding work and pertinent acceptability criteria not clearly established An one-year stability for 152 individual stock solutions and working standard solutions (- 80 C, - 20 C and + 4 C) Discussion with suppliers to provide customized standard mixes Results to be published 14
Area Screening Approach in Multi-Residue Methods Screening method: Selection of samples with levels [ ] that exceed the Screening Target Concentration (STC), with a given certainty [ ]. The result of the screening analysis is either negative or suspect. [ ]. Results from samples exceeding the cut-off value shall be verified by a full re-analysis from the original sample by a confirmatory method (EC 2014/519). Results expressed either as < STC or Suspect Screening Target Concentration (STC) Maximum Residue Limit Cut-off value: Response, signal or concentration, obtained with the screening method, above which the sample is classified as suspect. 8.E+05 Matrix-matched calibration curve for Penicillin G in different samples 6.E+05 4.E+05 2.E+05 Fresh Salmon Infant Formula A Infant Formula B Whey Powder Infant Formula C Strong matrix effects are often observed for many analytes Absolute peak area can not be used as a cut-off value. 15 0.E+00 0 5 10 15 20 25 30 35 Concentration (µg/kg)
Screening Approach in Multi-Residue Methods UNSPIKE/SPIKE APPROACH: All samples are extracted in duplicate, i.e. as unspiked and spiked at 1x STC level. Relative comparison of peak areas between the unspiked and the spiked samples. Quickly identify negative sample Thiabendazole, Infant Formulae: < STC Quickly identify suspect sample Albendazole sulfoxide, beef powder: SUSPECT Unspiked Spiked 1x STC Unspiked Spiked 1x STC Decrease the False Positive (FP) rate Sulfanilamide, milk fat powder: < STC Easily identify potential False Negative (FN) results Imidocarb, fish powder: not analysable Unspiked Spiked 1x STC Unspiked Spiked 1x STC Enrofloxacin, duck powder: < STC Unspiked Spiked 1x STC 16
The Challenge of Hypoallergenic Infant Formula Analysis Small peptides (MW < 1 kda) generate huge matrix effects in sample extracts (HR-MS spectra). Blank Solvent MeOH:H2O (15:85) Pork Powder Extract Hypoallergenic Infant Formula Lower sensitivity observed since most compounds elute in the ion suppression area. 20 56 % of compounds are eluting in 3-6 min range MW profile of a hydrolysed whey protein: 10 17 0
Facilitate Data Treatment Development of an Internal IT Tool Manage data from different software Automatize and facilitate calculations Align and facilitate data sharing within the Nestlé analytical community Options often missing in supplier softwares (standard addition, reporting, etc.) 1) Select the Software where your data come from 2) Import your raw data 18 3) Select your identification criteria 4) Select the threshold level
Effective Control of Vet. Drugs across the Nestlé Supply Chain 4- Multi-Class, Multi-Residue Methods 3- Analytical Challenges 2- Literature Review 1- Background and Needs
Four (ESI)-LC-MS/MS Methods to Address the Needs «Multi-class» β-lactams Tetracyclines Aminoglycosides Generic method Polar compounds lost during acidic QuEChERS sample preparation Specific separation of epimers + LC conditions to avoid chelation Highly polar compounds requiring specific LC conditions. 105 compounds 23 compounds 10 compounds 14 compounds Agilent 1290 + Sciex QqQ 5500 Agilent 1290 + Sciex QqQ 5500 Agilent 1290 + Sciex QqQ 5500 Agilent 1200 + Sciex QqQ 5000 BEH C18, 2.1 x 100 mm, 1.7 µm BEH C18, 2.1 x 100 mm, 1.7 µm HSS T3, 2.1 x 100 mm, 1.8 µm Kinetex C18, 2.1 x 100 mm 2.6 µm Run Time: 16 min Run Time: 10 min Run Time: 11 min Run Time: 8 min A: H 2 O + 0.5 mm AF* + 0.1 % FA* A: H 2 O + 0.5 mm AF* + 0.1 % FA* A: H 2 O + 0.01 M oxalic acid A: H 2 O + 20 mm HFBA B: MeOH + 0.5 mm AF* + 0.1 % FA* B: MeOH + 0.5 mm AF* + 0.1 % FA* B: MeOH + 0.1 % formic acid B: MeCN Acidic QuEChERS-based Basic QuEChERS-based Liquid Liquid Extraction MIP-SPE + Low Temperature Partitioning *AF: Ammonium Formate *FA: Formic Acid
Multi-Class, Multi-Residue Method(s) to Cover 152 Compounds Antibiotics n = 104 -Lactams (23) Sulfonamides (21) Quinolones (18) Aminoglycosides (14) Tetracyclines (10) Macrolides (7) Amphenicols (3) Diaminopyrimidines (2) Lincosamides (2) Rifamycins (2) Aminocoumarins (1) Streptogramins (1) 21 Antiparasitics Benzimidazoles (14) Coccidiostats (11) Avermectins (6) Salicylanilides (4) Halogenated phenols (1) Imidazothiazoles (1) Organophosphorous comp. (1) Praziquantel Pyrantel n = 40 n = 5 n = 3 Anti-Inflammatories Tranquilizers Carazolol Chlorpromazine Xylazine Carprofen Diclofenac Flunixin Meloxicam Phenylbutazone
Intensity, cps Intensity, cps Four (ESI)-LC-MS/MS Methods to Address the Needs «Multi-class» β-lactams Tetracyclines Aminoglycosides 8.8e6 99 Veterinary Drugs in solvent at 1x STC level (equivalent to 0.3-15 µg/kg). 9.0e4 6 Avermectins in a salmon sample spiked at 1x STC level (5 µg/kg). 1 2 4 3 1- Emamectin B1a 2- Eprinomectin B1a 3- Abamectin B1a 4- Doramectin 5- Moxidectin 6- Ivermectin B1a 5 6 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 Time, min 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 Time, min
Intensity, cps Four (ESI)-LC-MS/MS Methods to Address the Needs «Multi-class» β-lactams Tetracyclines Aminoglycosides 3.4e5 16 23 β-lactams in a skimmed milk powder sample spiked at 1x STC level (ranging from 4-50 µg/kg). 6,7 9 11 12,13,14 15 17 19,20 1- Desacetylcefapirin 2- Amoxicillin 3- Tazobactam 4- Cefadroxil 5- Sulbactam 6- Aspoxicillin 7- Cefapirin 8- Cefquinome 9- Cefacetril 10- Cefalonium 11- Cefalexin 12- Ampicillin 13- Cefuroxime 14- Cefazolin 15- Cefoperazone 16- Ceftiofur 17- Piperacillin 18- Penicillin-G 19- Oxacillin 20- Penicillin-V 21- Cloxacillin 22- Dicloxacillin 23- Nafcillin 8 21 22,23 2,3 1 4,5 10 18 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 Time, min
Intensity, cps Four (ESI)-LC-MS/MS Methods to Address the Needs «Multi-class» β-lactams Tetracyclines Aminoglycosides 7.9e5 5 10 Tetracyclines in a meatbased baby food sample spiked at 1x STC (25 µg/kg). 2 3 5 1-4-epi-tetracycline 1 - Tetracycline 2-4-epi-oxytetracycline 2 - Oxytetracycline 3-4-epi-demeclocycline 3 - Demeclocycline 4-4-epi-chlortetracycline 4 - Chlortetracycline 5-6-epi-doxycycline 5 - Doxycycline 2 3 4 1 1 4 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 Time, min
Intensity, cps Four (ESI)-LC-MS/MS Methods to Address the Needs «Multi-class» β-lactams Tetracyclines Aminoglycosides 2.5e4 9 12, 13 4 10, 11 14 Aminoglycosides in a egg powder sample spiked at 1x STC (50 µg/kg). 1 6 7, 8 14 1- Spectinomycin 2- Hygromycin B 3- Streptomycin 4- Dihydrostreptomycin 5- Amikacin 6- Kanamycin A 7- Paromomycin 8- Apramycin 9- Tobramycin 10- Sisomycin 11- Gentamicin C1a 12- Gentamicin C2+C2a 13- Gentamicin C1a 14- Neomycin B 2 3 5 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 Time, min
Effective Control of Vet. Drugs across the Nestlé Supply Chain 5- Validation Process 4- Multi-Class, Multi-Residue Methods 3- Analytical Challenges 2- Literature Review 1- Background and Needs
Criteria Design Response (in %) Samples Validation Procedure Based on: Guidelines for the validation of screening methods for residues of veterinary medicines CRLs 20/1/2010. n > 60, including: Milk fractions, Infant formulae Infant cereals, Meat/seafood powders Fresh/cooked meat/seafood Meat-based baby foods Fat, Eggs 100% 80% Cut-off level: data collected for albendazole Response in Blank Samples Response in Spiked Samples (1 STC) Fortified at 0x, 1x, 2x STC 60% Three analysts involved Over a period of 15 days 40% Cut-off at 20% 20% Identification: RT < 0.2 min 2 transitions/compound 0% 0 10 20 30 40 50 60 70 Number of Samples (n=67) Validation: False suspect rate < 10 % False negative rate < 5 % 27
Validation Results «Multi-class» β-lactams Tetracyclines Aminoglycosides 67 samples 83 samples 62 samples 78 samples 105 compounds 23 compounds 10 compounds 14 compounds STC: 0.3 15 µg/kg STC: 4-50 µg/kg STC: 25 µg/kg STC: 50 µg/kg FP: 0 % - 2 % FP: 0 % - 2 % FP: 0 % FP: 0 % - 3 % FN: 0 % - 3 % FN: 0 % - 5 % FN: 0 % FN: 0 % - 3 % Clorsulon (FN 6%) Tulathromycin A (FN 6 %) 28
Effective Control of Vet Drugs Across Nestlé Supply Chain 6- Deployment in Nestlé Labs 5- Validation Process 4- Multi-Class, Multi-Residue Methods 3- Analytical Challenges 2- Literature Review 1- Background and Needs
Quality Control in Operation Vet. Drugs LC-MS/MS Methods Cergy, Vittel (France) Nunspeet (The Netherlands) Orbe (Switzerland) Padova (Italy) Rzeszow (Poland) Weiding (Germany) York (UK) Zhukovsky Shoham Tianjin Dublin St Louis Toluca Antigua Abidjan Dubai Moga Navanakorn Cabuyao Santiago São Paulo Clayville Shah Alam Singapore Mulgrave 30 In routine In implementation soon
Effective Control of Vet. Drugs across the Nestlé Supply Chain 6- Deployment in Nestlé Labs 5- Validation Process 4- Multi-Class, Multi-Residue Methods 3- Analytical Challenges 2- Literature Review 1- Background and Needs
Take Home Messages Uncontrolled occurrence of veterinary drugs in food is a health concern, particularly with regard to antimicrobial resistance as emphasized by WHO. Multiresidue analysis is needed for an effective control: o o o Matrix scope should cover raw materials, processed ingredients and finished products. Maximum Residue Limit (MRL) from various countries were considered to set STCs. Method performance should fit with throughput and positive rate for an as-low-as-possible analysis cost. LC-MS/MS multi-class, multi-residue methods (152 compounds, 4 methods) is being implemented to efficiently control the occurrence of vet. drugs along the Nestlé Supply Chain & Manufacturing. Never ending story: new compounds/matrices can be included upon needs. All these methods are about to be published in a peer-reviewed journal 32
THANK YOU FOR YOUR ATTENTION Thomas BESSAIRE Nestlé Research Center, Lausanne, Switzerland Thomas.bessaire@rdls.nestle.com
Four (ESI)-LC-MS/MS Methods to Address the Needs «Multi-class» β-lactams Tetracyclines Aminoglycosides 1-g test portion in duplicate: 1) Unspiked sample 2) Spike sample at STC Qualitative Screening Acidified water (0.1 % formic acid) Acetonitrile Solvent Extraction QuEChERS salt mixture (Na 2 SO 4, NaCl & Citrate salts) Shaking and Centrifugation Phase Separation Clean-up «most of them» (n=99) Clean-up Avermectins (n=6) d-spe sorbents (Na 2 SO 4 + PSA + C18) Shaking and Centrifugation d-spe sorbents (MgSO 4 + PSA + C18) Shaking and Centrifugation d-spe Clean-up Evaporation to dryness Reconstitution in methanol/water 15/85 Evaporation to dryness Reconstitution in methanol/water 80/20 Concentration Injection by LC-MS/MS Run A Injection by LC-MS/MS Run B
Intensity, cps Four (ESI)-LC-MS/MS Methods to Address the Needs «Multi-class» β-lactams Tetracyclines Aminoglycosides 6000 1 1- Wild Saumon 2- Infant Formula 1 3- Shrimp 4- Infant Formula 2 5- Whey Protein Concentrate 2 3 4 5 The zero tolerance Chloramphenicol is detected at very low level (0.3 µg/kg, EU MRPL) in all matrices. 4.00 4.10 4.20 4.30 4.40 4.50 4.60 4.70 4.80 4.90 Time, min
Intensity, cps Four (ESI)-LC-MS/MS Methods to Address the Needs «Multi-class» β-lactams Tetracyclines Aminoglycosides 1-g test portion in duplicate: 1) Unspiked sample 2) Spike sample at STC Phosphate buffer ph 9.2 - shaking Acetonitrile - shaking QuEChERS salt mixture (Na 2 SO 4 + NaCl) Shaking and Centrifugation d-spe clean-up (Na 2 SO 4 + C18) Shaking and Centrifugation Evaporation not to dryness Reconstitution in water Injection by LC-MS/MS Qualitative Screening Buffered Extraction Removal of Proteins Phase Separation d-spe Clean-up Concentration 3.4e5 1 2,3 4,5 6,7 8 9 10 11 12,13,14 15 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 Time, min 23 β-lactams in a skimmed milk powder sample spiked at 1x STC level (ranging from 4-50 µg/kg). 16 17 18 19,20 21 22,23 1- Desacetylcefapirin 2- Amoxicillin 3- Tazobactam 4- Cefadroxil 5- Sulbactam 6- Aspoxicillin 7- Cefapirin 8- Cefquinome 9- Cefacetril 10- Cefalonium 11- Cefalexin 12- Ampicillin 13- Cefuroxime 14- Cefazolin 15- Cefoperazone 16- Ceftiofur 17- Piperacillin 18- Penicillin-G 19- Oxacillin 20- Penicillin-V 21- Cloxacillin 22- Dicloxacillin 23- Nafcillin
Intensity, cps Four (ESI)-LC-MS/MS Methods to Address the Needs «Multi-class» β-lactams Tetracyclines Aminoglycosides 1-g test portion in duplicate: 1) Unspiked sample 2) Spike sample at STC Qualitative Screening 7.9e5 5 EDTA 0.1 M in water - shaking Buffered Extraction Acetonitrile - shaking Removal of Proteins 5 Freezing step (-20 C) Centrifuge Supernatant defatted with hexane Shaking and Centrifugation Evaporation to dryness Reconstitution in aqueous oxalic acid/methanol 70/30 Phase Separation Sample Clean-up Concentration 1-4-epi-tetracycline 1 - Tetracycline 2-4-epi-oxytetracycline 2 - Oxytetracycline 3-4-epi-demeclocycline 3 - Demeclocycline 4-4-epi-chlortetracycline 4 - Chlortetracycline 5-6-epi-doxycycline 5 - Doxycycline 1 1 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 Time, min 2 2 3 3 4 4 Injection by LC-MS/MS 10 Tetracyclines in a meat-based baby food sample spiked at 1x STC (25 µg/kg).
Intensity, cps Four (ESI)-LC-MS/MS Methods to Address the Needs «Multi-class» β-lactams Tetracyclines Aminoglycosides 2.5e4 9 12, 13 2-g test portion in duplicate: 1) Unspiked sample 2) Spike sample at STC EDTA 0.5 % in water + TCA 2% Shaking and Centrifugation Dilution with ammonium carbonate Centrifuge MIP (Molecularly Imprinted Polymer) SPE Condition / Load / Wash / Dry / Elute Qualitative Screening Extraction & Removal of Proteins Sample Dilution Sample Clean-up 1 2 3 4 5 6 10, 11 7, 8 14 1- Spectinomycin 2- Hygromycin B 3- Streptomycin 4- Dihydrostreptomycin 5- Amikacin 6- Kanamycin A 7- Paromomycin 8- Apramycin 9- Tobramycin 10- Sisomycin 11- Gentamicin C1a 12- Gentamicin C2+C2a 13- Gentamicin C1a 14- Neomycin B Injection by LC-MS/MS 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 Time, min 14 Aminoglycosides in a egg powder sample spiked at 1x STC (50 µg/kg).