Analysis of Contaminants in Food - FSAI

Seminar on Contaminants in Food FSAI

17th

November 2009

Analysis of Contaminants in  Food

Michael O’Keeffe

Scientific Committee FSAI

Analysis of Contaminants in Food

• Residue analysis – what is involved?

• Methods used

• Research advances

• New approaches

• Contaminants – what are found?

• Closing comments

Aims of Residue Analysis in Food•

to identify the residue status of food 

using the least number of  samples/tests1.

least interference with business (production, sale, shipment, etc.)

2.

lowest cost of testing

3.

highest assurance of food safety and/or compliance

Characteristics of Residue Analysis

Target Sample Result

Specific food item(carcase; milk tanker)

Part of the food item(tissue sample; milk sample)

Applicable only to the 

specific food item(sample/test method)

Specific batch of food(herd; milk silo)

Sample(s) from the batch(tissue sample(s); milk sample(s))

Applicable to the batch(sampling/sample/test method)

Survey of food(retail products)

Samples from multiple 

sources(multiple samples)

Reflecting status of food 

covered by survey(design/sampling/sample/test method)

Monitoring programme(national herd; milk production)

Samples covering the 

national production(range of samples)

Reflecting status of food in 

national production(scope/design/sampling/sample/test method)

HIGH

s u

rety

LOW

What Do Results Tell Us?Expected Occurrence in the Population based on Test Results (95%

confidence)

Number of 

Positive 

Results

Number of Samples

10 20 100 250 1000

0 0‐27 % 0‐20 % 0‐4 % 0‐1 % 0 %

1 0‐40 % 0‐23 % 0‐5 % 0‐4 % 0‐2 %

5 22‐78 % 10‐47 % 2‐11 % 3‐9 % 4‐7 %

•

If no positive in 297 random samples, occurrence   in population is ≤

1% (with 95% confidence)

• “No residue detected in sample”

–

means just that!

How are Results Interpreted?

• Maximum Residue Limit (MRL)‐

level of permitted drug allowed in a food

• Maximum Level (ML)‐

up to level of a contaminant allowed in a food

• Minimum Required Performance Limit (MRPL)‐

minimum content that must be detected

• Decision Limit (CCα)/Detection Capability (CCβ)-

method capability to identify and measure

• Limit of Quantification/Detection (LOQ, LOD)‐

measures of method sensitivity

F

O

O

D

M

E

T

H

O

D

Central Meat Control Laboratory 

State LaboratoryIrish Equine CentreAshtown Food Research Centre

Pesticide Control Service, DAFF

State Laboratory Pesticide Control Service

Marine Institute

Radiological Protection Institute

Public Analyst Laboratories

Approaches to Residue Testing

1.

Large numbers of samples for testing:‐

SCREENING tests for similar

contaminants

(aflatoxins; antimicrobials; steroids) 

‐

CONFIRMATORY tests

2.

Samples for testing for large numbers of  contaminants:

‐

multi‐residue CONFIRMATORY tests(antiparasitics; pesticides; dioxins)

Screening TestsAim:

rapid, inexpensive clearance of large numbers of samples as “negative”

• Immunoassays

-

Mycotoxins

(ELISA)

Veterinary Drugs

• Automated IAs

-

Veterinary Drugs

(Biosensor)

• Cell Receptor Assay

-

Dioxins

(e.g. CALUX)

• Inhibition Assays

-

Antimicrobials

(Six-plate; DELVO)

Screening Test –

Dipstick Immunoassay

sample extractsloading extract into sample well

immunoassay dipsticks

result shown as binding lines on pad

Screening Test – Biacore Biosensor

Biacore Biosensor ‐

operation

Reflected intensity

Incident light

Gold film

Resonance angle:

prism

Biacore Biosensor ‐

operation

Reflected intensityIncident

light

’

Change of resonance angle by altered surface

flow

Binding of analyte

Biacore Biosensor – measurement 

stable complex

regeneration

bind

ing

TIME (s)

RES

PO

NSE

(R

U)

concentration

Confirmatory TestsAim:

accurate, specific, quantitative results for samples screened as “query positive”

• LC (or GC) –

mass spectrometry

for Organic residues (Banned substances, Veterinary Drugs, Mycotoxins,

Dioxins, Pesticides, etc.)•

LC (or GC) –

other detectors

for Organic residues(permitted

Veterinary Drugs, Mycotoxins, Pesticides)

• Atomic absorption/emission spectrometry

for Heavy Metals

HPLC‐MS/MS

LC-MS/MS

CHROMATOGRAM OF MULTIPLE

RESIDUES

Residue Analysis supported by Research

Components of a Residue MethodMethod: Nitroimidazoles in Liver (2007)

Liver sample

LC-MS/MS

Repeat (x2)

Residue Extraction using “QuEChERS”

Dispersive SPE -

QuEChERS-

mixing of a sorbent with sample extract in a mini-centrifuge tube to retain matrix interferents, but not the analytes

•

No need for: SPE cartridges, apparatus, elution, evaporation

•

Less sorbent, less time, less cost

•

Better interaction of sorbent with sample extract for cleanup

• Fast, easy procedure

MixedExtract

Dispersive-SPE clean-up

Residue Extraction using “QuEChERS”

Developed for use in Pesticide analysis by USDA

More than 250 Pesticides extracted with 70-110% (mainly 90-100%) recovery

At AFRC Teagasc developed for anthelmintic drug residue analysis, combined with UPLC-MS/MS

• 38 drug residues included

• Good sensitivity achieved

• Replacing 3 methods

• Reducing analysis time (×

5)

• Reducing solvent waste (×

10)

Improvements in Chromatography

HPLC‐MS/MS Feature UPLC‐MS/MS

5 µm

0.3 ml/min

30°C

32 min

242

particle size

flow‐rate

temperature

run time

measurements

1.7 µm

0.45 ml/min

65°C

16 min

344

Example:Pesticides in Fruit 

and Vegetables

decreased         

run time

expanded     

scope

improved 

efficiency

increased 

sensitivity

more samples 

tested

Research guiding Testing

SARAF/v

1

Protein-bound Nitrofuran Residues

Liver

released, derivatised nitrofuran

metabolites

LC-MS/MS

ethyl acetate extraction (3 x 6 ml)

• Occur as bound residues in edible tissues, for long time after treatment

• Released by acid conditions of stomach as toxic metabolites (banned)

solvent washes acid incubation derivatisation

Test method targets bound residues:

New Approaches to Testing

Challenge – how to detect hormone abuse in animals?

‐ conventional testing of only 0.05% of animals

‐

use of low level hormone “cocktails”

and  “designer”

drugs

Solution –

blood screening for indicator metabolites

‐ proteins, creatinine, enzymes, electrolytes, urea, glucose

Confirmation by LC‐MS/MS analysis 

‐ testing of hair samples where hormones accumulate

“Indirect”

Testing

0 7 14 21 28 35 4225

50

75

100

125

150ControlTreated

**

Day

Cre

atin

ine

(m

ol/l)

0 7 14 21 28 35 4225

50

75

100

125

150ControlTreated** ** * *

Day

Cre

atin

ine

(m

ol/l)

Heifers treated with nortestosterone

Steers treated with oestradiol

“Indirect”

Testing

Heifers treated with nortestosterone

prediction success rate =

91.4%

Steers treated with oestradiol

prediction success rate =

95.7%

Ensuring the Quality of Testing• Method Validation and Performance Criteria 2002/657/EC

• Method Accreditation

ISO 17025 (INAB)

• Proficiency Testing

FAPAS

CRLs

What do the Results show?

http://nfrd.teagasc.ie

Dioxins + PCBs

“Background Samples (n = 24) “Potential impact”

Samples (n = 13)

2000 2004 2006 2007 2000 2004 2006 2007

Mean 0.49 0.37 0.45 0.39 Mean 0.48 0.41 0.51 0.50

Max 0.95 0.66 1.06 0.52 Max 0.83 1.06 1.83 1.51

EU Regulation (1881/2006):  6 pg WHO‐TEQ/g fat 

Dioxins + Dioxin-like PCBs in Milk (pg WHO-TEQ/g fat) –

EPA Surveys

BUT! –

see Rhod Evans story on Dioxins

RadioactivityDoses to Heavy Seafood Consumers (*) due to Radiocaesium

(*200 g fish plus 20 g shellfish per day)

Conclusion: “it is safe to eat seafood landed at Irish fishing ports”

Dose of radiation from naturally-occurring polonium-210 in seafood = 148 µSievert

Marine Biotoxins

Shellfish

Amnesic Shellfish 

Poisoning (ASP) toxins

Azaspiracid (AZP) toxins Diarrhetic Shellfish 

Poisoning (DSP) toxins

Samples < ML > ML*(%)

Samples < ML > ML#

(%)Samples < ML > ML$

(%)

Mussels 112 111 1(0.9%)

1621 1402 219(13.5%)

1612 1604 8(0.5%)

Scallops 541 505 36(6.7%)

(*20 µg/g domoic acid) (#0.16 µg/g AZP equivalents) ($0.16 µg/g okadaic acid)

Pesticides

Monitoring Programme  for Fruit and Vegetables 2003‐2007

Profile:  some 900 samples of ≈

75 commodities tested each 

year(approximately 25% domestic and 75% imported)

Year: 2003 2004 2005 2006 2007

Pesticides tested 89 118 148 153 292

Pesticides found 45 53 75 72 103

Testing:

Results:  44% (2003) – 59% (2007) samples contain pesticides

3% to 5% of these pesticides are at levels >MRLs

Are Pesticides a Food Safety Issue?

Pesticide intake per kg body weight

PesticideADI

(mg/kg 

bw/d)

Mean 97.5 percentile    

of intake

99.9 percentile    

of intake

mg % of 

ADI

mg % of 

ADI

mg % of 

ADI

Iprodione 0.06 0.00015 0.25 0.00075 1.24 0.00665 11.1

Thiabendazole 0.1 0.00070 0.70 0.00458 4.58 0.02577 25.8

Cyprodinil 0.03 0.00007 0.24 0.00016 0.53 0.00029 0.96

Chlorpyriphos 0.01 0.00013 1.26 0.00075 7.5 0.00181 18.1

Estimated Exposure of Children to 4 Pesticides, based on Food Consumption

Conclusion: intake of pesticides by children are below ADI values

Veterinary Drugs ‐

Antimicrobials

Veterinary Drugs ‐

Antiparasitics

Increased incidence of anthelmintics found with improved testing:

Incidence of Non‐compliant Samples

2005 2006 2007 2008

none  1 ×

ovine 2 ×

milk 1 ×

bovine5 ×

milk1 ×

ovine1 ×

poultry

Application of new method QuEChERS extraction coupled with UPLC-MS/MS

Veterinary Drugs ‐

Anticoccidialss

Coccidiosis serious in poultry

Nicarbazin –

a very effective drug

Maximum limit: 200 µg/kg in liver

Research leading to Improved Practices

BUT! –

very difficult to eliminate

Veterinary Drugs –

Industry Self‐monitoring

Directive 96/23/EC (S.I. 143/2007) applies for Foods of Animal Origin

Bovine Ovine Porcine Poultry Milk

Antimicrobials

2329/0

Β‐Agonists 

4121/0

Trenbolone

4183/0

Antimicrobials 

3147/0

Anticoccidials

379/0

Antimicrobials 

29025/0

Chloramphenicol

501/0

Sedatives      

161/0

Antimicrobials 

2840/1

Anticoccidials

1160/30

Antimicrobials 

524569/2108

Pesticides 

383/0

2008 Self-monitoring