Bioanalytical screening methods

Bioanalytical Screening of Feed and Food for elevated Levels of Dioxins and dioxin-like PCBs


1. Introduction

Analysis of feed and food for dioxins and dioxin-like compounds is an important section of European food safety, health and consumer protection. Results are used for making decisions on compliance or noncompliance with food safety legislation and for assessment of potential risks to public health. It is therefore paramount that the applied methods are well characterised, fully validated and documented to a satisfactory standard in order to yield reliable results.

Since their development in the 1990s, cell-based bioanalytical methods (bioassays) have effectively been applied in a number of environmental and research laboratories and in laboratories performing official control. According to EU legislation, certain criteria apply when checking samples for elevated levels of polychlorinated dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFs) and dioxin-like PCBs (DL-PCBs) within European official control of feed and food.

Such analytical criteria and requirements were first developed for screening and confirmatory methods in 2001 (1,2), and implemented in EU legislation in 2002 (3,4). They underwent substantial revision more recently by the network of the European Union Reference Laboratory (EU-RL) for Dioxins and PCBs in Feed and Food, Freiburg (Germany) and the National Reference Laboratories (NRLs), with special focus on method validation, cut-off concentrations and on-going quality control. The revised criteria were presented during the Dioxin 2010 conference in San Antonio, Texas (USA) (5) and adopted by European legislation (6,7) in 2012.

Previously, the analyst directly compared results from bioanalyses to the respective regulatory limits specified in Toxic EQuivalents (TEQ). However, correspondence between bioanalytical results, now expressed in Bioanalytical EQuivalents (BEQs), and the results from confirmatory gas chromatography / high resolution mass spectrometry (GC/HRMS), or gas chromatography / tandem mass spectrometry (GC-MS/MS), expressed in TEQs, may not always be in a one-to-one relationship. This observation stresses the need for evaluation of the BEQ/TEQ-ratio, which may be different for each sample matrix of interest. Before applying a bioanalytical method in routine screening, BEQ-based cut-off concentrations ensuring false-compliant rates below 5% must be established above which a sample is declared suspected to exceed the respective EU legal limits (8,9).

Further amendments as regards application of screening and confirmatory methods and the use of bioanalytical results were more recently proposed by the EU-RL/NRL network, and supported by the competent standing committees for adoption into EU legislation in June 2014 (10,11). These amendments comprise a revision of the “field of application” and the “classification of methods” and clarify, that numerical results from screening primarily give an indication of the range of TEQ-levels in case of follow-up by confirmatory analysis. However, numerical screening results are not suitable for evaluation of background levels, following time trends, estimation of intake, in concentrations, or for re-evaluation of action levels and maximum levels. For these purposes, GC/HRMS or GC/MS-MS confirmatory methods must be applied.

2. Bioanalytical Screening – The CALUX Bioassay

European legislation (10,11) permits the use of GC/MS-based or bioanalytical methods for screening of feed and food samples, for elevated levels of PCDD/Fs and DL-PCBs. Screening results are compared with a cut-off concentration, enabling the analyst to decide over sample compliance and to identify those samples requiring further investigation by confirmatory analysis. In addition, screening results may give a numerical indication of the PCDD/F- and DL-PCB-TEQ-levels in the sample. Expression of bioanalytical results as BEQs is particularly helpful for the analyst performing the follow-up by a confirmatory method, but mandatory during the initial validation process.

One interesting and frequently used cell-based bioanalytical screening method is the Chemically Activated LUciferase gene eXpression (CALUX) assay. It detects 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and structurally similar halogenated aromatic hydrocarbons (HAHs) based on their ability to activate the aryl hydrocarbon receptor (AhR) signalling pathway (12). CALUX utilizes recombinant mouse or rat hepatoma cell lines that contain a stably integrated AhR-responsive luciferase reporter gene plasmid. Exposure of these cells to extracts containing TCDD and/or related HAHs, results in a time-, dose- and chemical specific induction of luciferase expression. Upon addition of D-Luciferin, the substrate for firefly luciferase’s bioluminescence reaction, light is emitted which can be detected by a microplate luminometer. The level of reporter gene expression corresponds to the total concentration of TCDD-like AhR agonists in the sample.

Although a considerable number of individual steps is involved from ligand binding to luciferase gene expression, this correspondence follows a classical ligand-receptor binding curve. When plotted in a semi-log graph, concentration-response data adopt a sigmoidal shaped pattern to which a 4-parameter, or better a 5-parameter logistic function, can be fit.

While GC/HRMS or GC/MS-MS methods look for selected congeners namely the 17 PCDD/Fs and 12 DL-PCBs, to which individual Toxic Equivalency Factors (TEFs) were assigned (13), CALUX detects all AhR-agonists present in a sample extract. Screening results may therefore be subject to interference from compounds structurally similar to the target analytes. This observation and several other aspects discussed elsewhere in more detail ultimately lead to the conclusion that bioanalytical results in BEQs cannot fully match TEQs resulting from GC/HRMS or GC/MS-MS analysis. However, BEQs may provide a reliable indication of the TEQ-levels present in a sample extract.

Bioanalytical screening methods include various fields of work and attention:

1. cell culturing and maintenance,
2. extraction of target compounds (PCDD/Fs and DL-PCBs) and, in case of fatty samples, of extractable lipids,
3. clean-up of the crude extracts on acidic silica,
4. additional clean-up by fractionated elution of PCDD/Fs, and of DL-PCBs, e.g. from an activated carbon/celite column,
5. exposure of genetically modified rat or mouse hepatoma cells cultured in microplate wells, to a TCDD-standard dilution series and to extracts of control samples and of unknown samples,
6. addition of luciferine and measurement of luciferase activity,
7. evaluation of raw data, fitting of 4- or 5-parameter equations to concentration- response data pairs, checking assay and sample related acceptance criteria,
8. evaluation of quality control data and results (control charts),
9. compliance assessment,
10. maintaining a QC data base with BEQ/TEQ results from confirmed samples

Bioanalytical results (BEQs) and WHO-TEQs

BEQs from bioanalytical methods may be close to yet not fully equal to TEQs measured with confirmatory methods. If both methods are validated according to legal requirements following internationally accepted guidelines, and provide reliable results, the following aspects may mainly contribute to observed discrepancies:

1. effects from differences between the toxic equivalency factor (TEF) and the relative potency (REP) value of each congener,
2. the fact that recovery correction for each congener with internal 13C-labelled standards is not possible in bioanalyses,
3. effects from co-eluting compounds altering the cell response or exhibiting even a cytotoxic impact on the cells,
4. the possible presence of other dioxin-like AhR-active compounds which, besides the 29 EU regulated PCDD/Fs and DL-PCBs, may have found their way into the final sample extract.

Hence, some samples may be declared from screening to be suspected to exceed MLs, or ALs, respectively, but then turn out compliant from confirmatory analysis.

Since EU regulatory limits are TEQ-based, it is essential to evaluate the extent of correspondence between results from both methods. The BEQ-TEQ relationship must therefore be assessed during an initial validation process (10,11) prior to use of the bioanalytical method by performing calibration experiments involving “blank” (low contaminated) samples spiked around the level of interest. Matrix-related cut-off concentrations (in BEQs) must be established ensuring a false compliant rate below 5%. These cut-offs mainly depend on the variability of bioanalytical results and of GC/HRMS-TEQs, the sensitivity of the bioanalytical method and the suitability of the control sample material used to compensate for any bias within acceptable apparent recovery ranges.

As mentioned above, numerical results of bioanalytical screening methods are not selective enough to be used for any quantitative assessment of concentrations in samples. Bioassays cannot identify and quantify the target compounds as required for quantitative methods according to the definition by IUPAC (15). However, as already stated, they may give, in BEQs, a valid numerical indication of the levels of PCDD/F-WHO-TEQs and WHO-PCB-TEQs present in the sample extract and ultimately in the sample.

3. New Criteria for Bioanalytical Methods

During a testing phase in 2007/2008, the Bioassay Research Unit of the EU-RL for Dioxins and PCBs in Feed and Food carried out a number of exploratory experiments with GC/HRMS-confirmed, well characterized food and feed samples. Commercially available protocols and cell systems obtained from various sources were utilized. Focus was on precision, recovery, sensitivity, and the working range of bioanalytical results expressed as the sum of PCDD/Fs and DL-PCBs. Some of these procedures, and the results obtained from various samples of interest at relevant concentrations and from “blank” (low contaminated) samples suggested that not all aspects of the then legally binding analytical criteria (3,4) were met. In one case, specific methods were not available at all for many of the matrices of interest. Hence, the EU-RL began focusing on method optimization and particularly on new developments related to the assay, sample extraction and clean-up procedures, and to raw data processing. Comprehensive concepts for method validation and continuous quality control were also developed and applied.

Thorough investigation and scientific exchange within the EU-RL/NRL-network finally led to the conclusion that the legal criteria (3,4) required re-evaluation and amendment. For this purpose, an EU-RL-headed expert working group was established in June 2008. Until July 2010, the analytical criteria underwent extensive revision. Some already existing criteria were re-evaluated, others newly introduced, and procedures established for an initial validation process, the calculation of matrix-related cut-off concentrations (in BEQs), and continuous quality control. Specification of numerical results in BEQs is now possible when based on full standard dose-response curves, and even required during the validation process. In routine analysis, BEQ values mainly serve as information for the analyst performing confirmatory analyses. All screening results above respective cut-offs must be followed-up by a confirmatory method producing the then required numerical information in TEQ.

Naturally, this concept requires a close cooperation of the bioassay lab with the laboratory running the confirmatory method (10,11) while the latter’s workload may considerably be reduced by sieving out most of the compliant samples up-stream.

Several important revisions were suggested to the European Commission in 2011 and adopted into legislation in March 2012. Requirements for bioanalytical methods and the reporting of bioanalytical results for PCDD/Fs and DL-PCBs were laid down in Commission Regulation (EU) No 252/2012 of 21 March 2012 (6) for food and in Commission Regulation (EU) No 278/2012 of 28 March 2012 (7) for feed. The criteria set out in these regulations are mandatory for laboratories screening food or feed samples within official control of the levels of PCDD/Fs and DL-PCBs, or for other regulatory purposes, to ensure compliance with the provisions in legislation.

Although Commission Regulations (EU) No 252 and 278/2012 were already quite definite about the use and practical value of bioanalytical results, it was considered important by the EU-RL/NRL-network to clarify even more in legislation the purpose for which bioanalytical methods may be applied. While it is fully clear and accepted that bioanalytical results cannot be quantitative, the terms “qualitative” and “semi-quantitative methods” and their various requirements may have led to some confusion.

The EU-RL/NRL-network therefore proposed to the European Commission further amendments with regard to method classification and the reporting of results. These were adopted into legislation by Commission Regulation (EU) No 589/2014 of 2 June 2014 (food) and Commission Regulation (EU) No 709/2014 of 20 June 2014 (feed). Applied methods are classified as “screening” and “confirmatory methods”, and screening results may still be reported qualitatively as “sample is compliant” or “sample is suspected to be noncompliant”. In addition, a numerical value in BEQs may be given as an indication of the concentration range for follow-up confirmatory analysis. During initial validation, however, BEQs must always be evaluated, e.g. for calculation of performance parameters such as matrix-related cut-off concentrations.

It is important to understand that a noncompliant result can only be reported as an outcome of a confirmatory method. On the other hand, the quality of reported numerical results well below the cut-off may be problematic, especially when dealing with low levels of interest and/or low levels of sample contamination.

Laboratories applying bioassays within official control, or for other regulatory purposes, must be accredited according to EN ISO/IEC 17025 by a recognized body operating in accordance with ISO Guide 58. Methods must be validated thereby providing evidence for compliance with EU legal criteria as given in Commission Regulations (EU) No 589 and 709/2014. The proficiency of the laboratory shall be proven by internal and external quality control measures. Continuous and successful participation in interlaboratory studies based on analyses of PCDD/Fs andDL-PCBs in the relevant feed / food matrices is mandatory.

Further requirements are:

- results shall be reported as yes/no-decision (sample is compliant / suspected to be noncompliant); in addition, a numerical value in BEQ may be given,
- false-compliant rates must be shown to be below 5 %,
- specific assay acceptance criteria apply (number of calibrators, quality of the curve fit, intra-assay variation, working range, reporting level, etc.),
- repeatability, expressed as relative standard deviation, shall not exceed 20%,
- within-laboratory reproducibility, expressed as relative standard deviation, shall not exceed 25%,
- apparent recoveries must be within given ranges,
- quality control schemes must be maintained: blank and positive control samples for indication of the background contribution to the results, and the apparent recovery, respectively, must be included in each sample series,
- BEQ results must be corrected for blank and recovery,
- a certain rate of compliant samples must be checked for possible presence of AhR- active compounds reducing the cell response, to demonstrate that those samples were not false compliant,
- matrix-related cut-off concentrations for decision over sample compliance, based on the ML, or AL, to be checked, must be established by the laboratory,
- all results above the cut-off must be confirmed by GC/HRMS or GC/MS-MS analyses; a noncompliant result can only be reported as an outcome of a confirmatory method,
- a certain proportion of compliant samples must be confirmed by GC/HRMS to demonstrate that those samples were really compliant (not false compliant).

4. Bioanalytical Method Optimization

Analytical methodology is continually improving, with focus on enhanced sensitivity, selectivity, precision, and recovery, increased throughput and/or ease of use. Eventually, an existing method may even be replaced with a better one. Demand for automation or laboratory robotics may also be a driving force if savings are expected. However, changing a bioanalytical method that tests for the potency of dioxin-like compounds on basis of molecular interactions requires extra care. A more sensitive method may respond to more impurities in the sample extracts leading to an increase or a decrease of the cell response. Increased selectivity may reduce the “apparent recovery” significantly for individual congeners.

Although BEQ results are not quantitative, the ratio between BEQs and TEQs (“apparent recovery”) is a prime consideration, because any bias in results may lead to false-compliant decisions. Finally, a reliable method must provide results with a run-to-run precision ensuring acceptable within-lab reproducibility of the results. Inaccurate, unreliable results are entirely inacceptable within the scope of efficient consumer health protection as they may ultimately lead to dramatic consequences.

CALUX technology as available from commercial companies including cell systems and protocols for assay- and sample-related procedures, and thereby generated bioanalytical results, do not meet all aspects of the existing legal criteria (status as of December 2014).

Discovered drawbacks related to sample processing include:

- efficiency of extraction and/or clean-up procedures for certain sample matrices of interest, and/or for handling concentration ranges of interest,
- methods for various EU-regulated matrices missing,
- methods poorly conceived / not reproducible for selective clean-up,
- curve models fitted to dose-response data inappropriate,
- correction for method blank and recovery not performed,
- positive control samples not included in sample series, not fully characterized, not suitable and/or not representative for sample matrices to be analyzed,
- procedures for calculation of cut-off concentrations not available,
- validation procedures according to EU legislation not available.

Assay-related drawbacks include:

- interpretation of assay parameters poor or missing,
- cell count for reproducible well conditions missing,
- poor fold induction values obtained from available transfected cells,
- poor stability of available transfected cells (usable only for ~ 20 passages).

Research work at the EU-RL is based on three decades of experience in analytical method development and continuous improvement in various analytical fields using widely accepted (normative) concepts and approaches. Before taking up the task of performing an unbiased state-of-the-art evaluation of methods and procedures submitted by various (commercial) providers, the EU-RL carefully reviewed relevant literature and used manifold opportunities for scientific exchange with bioanalytical experts from various research and routine laboratories.

Two major avenues are followed to establish reliable bioanalytical screening methods using genetically modified rat and mouse hepatoma cell lines as detection systems:

1. Analysis of PCDD/F-PCB-BEQs after extraction and clean-up of the crude extract on acidic silica, mainly using H4L1.1 rat hepatoma cells (16,17)
2. Separate analysis of PCDD/F-BEQs, and of DL-PCB-BEQs, after extraction, clean-up of the crude extract on acidic silica followed by fractionated elution from carbon/celite, involving H4L1.1 rat, H1L6.1 mouse, and highly sensitive “G3” H4L7.5 rat hepatoma cells (18)

Assay performance is assessed by fitting a suitable mathematical equation to the dose-response data acquired from exposure of the cells to a 2,3,7,8-TCDD standard dilution series followed by evaluation of a number of selected assay parameters. Assay and measurement parameters identified as of relevance are optimized. Not all unsatisfactory sample extraction and clean-up methods are re-developed by EU-RL from the ground up. It some cases, it may prove more beneficial to modify, optimize and validate methods available in public domain for producing results in compliance with EU legislation.

5. Method Performance: Validation and Quality Control

Within the scope of establishing strong EU-wide standards for routine and reference methods within official control, the EU-RL for Dioxins and PCBs in Feed and Food evaluates performance and suitability of CALUX bioassay technology as available on the European market as a permanent task. Research work includes testing of methods available from commercial companies and from public domain, optimization of the latter when desirable and promising, new developments, enhanced raw data processing, and new concepts for method validation and QC. Included in the EU-RL’s evaluations are, besides commercially available assays, genetically modified rat and mouse hepatoma cells available on a non-commercial basis from University of California Davis (USA).

Well-characterized and fully validated methods are essential to acquire results that can be interpreted unambiguously. Bioanalytical methods are constantly undergoing changes and improvements and as such are often at the cutting edge of the technology. Each bioanalytical method has its own characteristics, which may vary from analyte to analyte and/or between groups of analytes. In response to this observation, an EU-RL-headed expert group has suggested specific performance criteria and validation requirements for screening of PCDD/Fs and DL-PCBs in feed and food samples within official control (5,6,7). Each laboratory must demonstrate fitness-for-purpose of the bioanalytical method prior to use. Specified run acceptance and quality control criteria apply when performing screening analysis.

Validation of Bioanalytical Methods

Before new criteria entered into force in 2012, bioanalytical results were directly compared to regulatory limits which are given in TEQs. However, correspondence between BEQs from bioanalytical screening and TEQs from GC/HRMS confirmatory analyses may not be in a one-to-one relationship, depending on physical properties of the sample matrix, recovery losses, congener patterns, differences between TEF and REP values, and the properties of the sample material used for recovery control and correction. These observations underline the need to evaluate this correspondence for each sample matrix, or matrix group, of interest, during an initial validation process, before application of the method in routine (10,11).

The EU-RL assesses bioanalytical method performance in a two-step procedure (19):

1. Initial Validation (before application in routine): Evaluation of basic method performance by matrix-matched calibration using “blank” samples spiked on various levels, followed by calculation of the initial cut-off value.
2. Performance Re-evaluation: Calibration experiments with compliant and noncompliant GC/HRMS-confirmed incurred samples, taking into account the variability of matrix properties and congener patterns; actual false-compliant rates are evaluated.

Initial Cut-off Concentrations

Before a bioanalytical method is applied in routine, matrix-specific BEQ-based cut-off concentrations must be established, ensuring a false-compliant rate below 5%. BEQ-results beyond the cut-off suggest that TEQ-concentrations may exceed the respective legal limits in the sample. Two alternative concepts developed at EU-RL were adopted in recent revisions of legal requirements (6,7,10,11), providing a statistically sound estimation of these important parameters.

The first, comprehensive approach evaluates the correspondence between BEQs and TEQs from results of matrix-matched calibration experiments. These involve low contaminated authentic samples spiked around the level of interest, e.g. in the range of 0 – 2x ML. Spiking levels are plotted vs. bioanalytical results corrected for blank and recovery. Linear regression is performed and performance parameters including the cut-off concentration are calculated. The cut-off equals the lower branch of the prediction interval calculated around the regression line at the level of interest.

The second, abbreviated approach allows a simplified estimation of the cut-off as the lower endpoint of the distribution of bioanalytical results obtained from multiple analyses of a representative low contaminated sample spiked at the level of interest under intermediate precision conditions. 2/3 ML can also be used as preliminary cut-off concentration after it could be demonstrated that the actual false-compliant rate is smaller than 5%. However, depending on method performance, this may lead to a larger fraction of false noncompliant results.

Continuous Quality Control

Connecting the initial validation process to the re-evaluation of method performance after some time of routine use, quality control forms the third pillar ensuring state-of-the-art performance and cut-off effectiveness. Complying with legal requirements (10,11) the EU-RL maintains a comprehensive QC-system based on, yet not limited to:

- matrix-matched GC/HRMS-characterized control (“reference”) samples,
- inclusion of one procedural (reagent) blank and one recovery control sample in each sample series,
- verification of acceptance criteria for the assay and for the sample run,
- monitoring and evaluation of quality control data over time (QC charts),
- checking routine samples for the presence of AhR-active compounds lowering the cell response thus leading to false-compliant results,
- GC/HRMS confirmation of all suspected samples and of a fraction of samples declared compliant,
- maintenance of a QC-database (BEQ/TEQ results) for each sample matrix / matrix group, for re-evaluation of method performance and cut-off,
- verification of false-compliant rates to be smaller than 5%,
- continuous and successful participation in interlaboratory (PT) studies for the sample matrices of interest.

The QC-database maintained at the EU-RL Bioassay Research Unit for each sample matrix / matrix group of interest includes bioanalytical and GC/HRMS results, from approximately 1300 samples, both confirmed noncompliant and compliant. Method performance and cut-off efficiency are frequently being re-evaluated whenever sufficient new data are included for each sample matrix.

Actual α- and ß-errors are assessed by applying the initially (previously) calculated cut-off to GC/HRMS-confirmed compliant and noncompliant results collected in the QC-database. This is important because the fraction of false-compliant results must be below 5%, to prove efficiency of the cut-off. The rate of false-noncompliant results should be reasonably small to keep bioanalytical screening upstream to the GC/HRMS technology advantageous.

6. Reporting of Bioanalytical Results

For decision over sample compliance, or a suspected noncompliance, the analyst compares the bioanalytical result to a pre-established cut-off concentration. Samples with a result below the cut-off are declared compliant, while samples with a result exceeding the cut-off must be further investigated by confirmatory analysis. If the sample turns out to be indeed noncompliant, analysis will be repeated. If the result from a second analysis proves noncompliance, the duplicate mean value will be reported. It is important to note that a noncompliant result can only be reported based on confirmed results expressed in TEQs.

Therefore, results from bioanalytical screening can only be reported as “the sample is compliant” or “the sample is suspected to be noncompliant”. In addition, a numerical indication of the WHO-PCDD/F- and WHO-PCB-TEQ-levels in the sample may be given in bioanalytical equivalents (BEQ).

7. State of Play

EU-RL’s Bioassay Research Unit has so far developed, optimized, validated and re-evaluated bioanalytical methods for 22 analyte group/sample matrix combinations of interest, based on more than 1300 GC/HRMS-confirmed samples from official control and from follow-up analyses during contamination incidents. Sample extraction, clean-up and assay-procedures and parameters applied are described elsewhere :

1. PCDD/F-PCB-BEQs: fat (bovine), meat (bovine), liver (bovine, sheep), fish meat, fish oil (as food, and as feed), hen’s eggs, cow’s milk fat, human milk (WHO study), cow’s whole milk, vegetable oil
2. PCDD/F- and DL-PCB-BEQs (separately analyzed): fat (bovine), meat (bovine), meat (porcine), liver (bovine, sheep), fish meat, fish oil (as food and as feed), hen’s eggs, cow’s milk fat, human milk (WHO study), cow’s milk

EU-RL’s bioanalytical screening methods fully comply with legal requirements10,11, providing state-of-the-art performance and a high level of reliability when checking compliance with maximum and action levels. ML-based false-compliant rates are way below the tolerable 5% mark. False noncompliant rates are within an acceptable 20 - 40 % range of those samples suspected to be noncompliant, and within a much smaller fraction of all screened samples.

Evaluated concepts and results from EU-RL’s research work performed in the field of bioanalytical methods were published in a number of scientific papers (5,21,22,23,24,25,26).

9. Summary, Outlook

Method validation and QC data collected at EU-RL demonstrate that the new legal requirements for application of bioanalytical methods are achievable for the CALUX bioassay technology in practice. On the other hand, they represent a strong driving force for thorough step-by-step method optimization providing an effective framework for attaining a high level of performance. CALUX bioassays, if validated and applied according to the new criteria may hence serve as a valuable tool for pre-selection of samples suspected to exceed the respective levels of interest, thereby considerably relieving GC/HRMS laboratories of their workload.

Yet, a word of caution seems appropriate regarding claims that implementation and application of a bioassay be “simple” and accomplished within just a few weeks or, even more daring, within just a few days, followed by a throughput of several hundred samples per week. Performing the required matrix-related validation procedures and QC schemes take time and attention, depending, of course, on the laboratory’s expertise and resources. Laboratories applying screening methods must establish a close cooperation with a laboratory applying the confirmatory method, for quality control and for confirmation of the analytical result of suspected samples. Therefore, a bioanalytical laboratory can never operate successfully and on a sound legal basis within official feed and food control when only established as a “stand-alone unit”.

Future focus at EU-RL will be on evaluation and validation of bioassays for screening of few remaining yet significant food matrices of interest for PCDD/Fs and DL-PCBs, such as milk products and food for infants and young children. The EU-RL further aims at evaluating bioanalytical methods for screening of various (complex) feed samples, for which MLs and ALs apply for dioxins and PCBs, as well.

10. References

1. Malisch R, Baumann B, Behnisch PA, Canady R, Fraisse D, Furst P, Hayward D, Hoogenboom R, Hoogerbrugge R, Liem D, Papke O, Traag W, Wiesmuller T (2001); Harmonized quality criteria for chemical and bioassays analyses of PCDDS/PCDFS in feed and food. Part 1: General Considerations, GC/MS-Methods. Organohalogen Compounds 50: 53-58
2. Behnisch PA, Allen R, Anderson J, Brouwer A, Brown DJ, Campbell TC, Goeyens L, Harrison RO, Hoogenboom R, Van Overmeire I, Traag WA, Malisch R(2001); Organohalogen Compounds 50: 59-63
3. Commission Directive 2002/69/EC laying down sampling methods and methods of analysis for official control of dioxins and dioxin-like PCBs in foodstuffs, OJ L 209, 06.08.2002, p. 5
4. Commission Directive 2002/70/EC establishing requirements for the determination of levels of dioxins and dioxin-like PCBs in feedingstuffs, OJ L 209, 06.08.2002, p. 15
5. Hoogenboom LAP, Hädrich J, Eppe G, Goeyens L, Elskens M, Malagocki P, Scippo ML, Vanderperren H, Windal I, Kotz A, Denison MS, Malisch R (2010); Revised EU-Criteria for Applying Bioanalytical Methods for Screening of Feed and Food for Dioxins and Dioxin-like PCBs. Organohal. Compounds 72: 1800-1905
6. Commission Regulation (EU) No 252/2012 of 21 March 2012 laying down methods of sampling and analysis for the official control of levels of dioxins, dioxin-like PCBs and non dioxin-like PCBs in certain foodstuffs and repealing Regulation (EC) 1883/2006, OJ L 84, 23.03.2012, p. 1
7. Commission Regulation (EU) No 278/2012 of 28 March 2012 amending Regulation (EC) No 152/2009 as regards the determination of the levels of dioxins and polychlorinated biphenyls, OJ L 91, 29.03.2012, p. 8
8. Commission Regulation (EU) No 1259/2011 of 2 December 2011 amending Regulation (EC) No 1881/2006 as regards maximum levels for dioxins, dioxin-like PCBs and non dioxin-like PCBs in foodstuffs, OJ L 320, 03.12.2011, p. 18
9. Commission Regulation (EU) No 277/2012 of 28 March 2012 amending Annexes I and II to Directive 2002/32/EC of the European Parliament and of the Council as regards maximum levels and action thresholds for dioxins and polychlorinated biphenyls, OJ L 91, 29.03.2012, p. 1
10. Commission Regulation (EU) No 589/2014 of 2 June 2014 laying down methods of sampling and analysis for the control of levels of dioxins, dioxin-like PCBs and non-dioxin-like PCBs in certain foodstuffs and repealing Regulation (EU) No 252/2012 (food)
11. Commission Regulation (EU) No 709/2014 of 20 June 2014 amending Regulation (EC) No 152/2009 as regards the determination of the levels of dioxins and polychlorinated biphenyls (feed)
12. Denison MS, Nagy SR(2003); Activation of the aryl hydrocarbon receptor by structurally diverse exogenous and endogenous chemicals. Annu. Rev. Pharmacol. Toxicol. 43: 309–334. doi:10.1146/annurev.pharmtox.43.100901.135828. PMID 12540743
13. Van den Berg M, Birnbaum LS, Denison M, De Vito M, Farland W, Feeley M, Fiedler H, Hakansson H, Hanberg A, Haws L, Rose M, Safe S, Schrenk D, Tohyama C, Tritscher A, Tuomisto J, Tysklind M, Walker N, Peterson RE (2006). Toxicol. Sci. 93 (2): 223-241
14. Goeyens L, Hoogenboom R, Eppe G, Malagocki P, Vanderperren H, Scippo M-L, Windal I, Baeyens W, Denison MS and Hädrich J. (2010); Discrepancies between Bio-Analytical and Chemo-Analytical Results have a non-negligible Message. Organohalogen Compounds 72: 964-967
15. IUPAC. Compendium of Chemical Terminology, 2nd ed. (the "Gold Book"). Compiled by A. D. McNaught and A. Wilkinson. Blackwell Scientific Publications, Oxford (1997). XML on-line corrected version: (2006-) created by M. Nic, J. Jirat, B. Kosata; updates compiled by A. Jenkins. ISBN 0-9678550-9-8. doi:10.1351/goldbook.
16. H4L1.1 (H4IIe) rat hepatoma cells, available from BioDetection Systems (BDS), Amsterdam (NL)
17. H4L1.1 rat (clone different from ref. 16), H1L6.1 mouse and H4L7.5 rat hepatoma cells, available from Prof. M S Denison, University of California Davis (USA)
18. He G, Tsutsumi T, Zhao B, Baston DS, Zhao J, Heath-Pagliuso S, Denison MS. (2011); Third-generation Ah receptor responsive luciferase reporter plasmids: Amplification of dioxin-responsive elements dramatically increases CALUX bioassay sensitivity and responsiveness. Toxicol. Sci. 123: 511–522
19. Haedrich J, Hoogenboom LAP, Eppe G, Goeyens L, Elskens M, Malagocki P, Scippo ML, Vanderperren H, Windal I, Kotz A, Denison MS, Malisch R. (2012); Principles of method validation and quality control: Compliance with the new EU criteria for bioanalytical screening of feed and food. Organohalogen Compounds 74: 193-196
20. Haedrich J et al. (2016); in preparation
21. Hädrich J, Eppe G, Goeyens L, Hoogenboom LAP, Malagocki P, Scippo M-L, Vanderperren H, Windal I, Denison MS, Stumpf C, Kotz A and Malisch R (2010); New Cut-Off Values for Applications in Bioanalytical Screening: Decision over Sample Compliance with Legal Limits set by the European Union for PCDD/Fs and dioxin-like PCBs. Organohalogen Compounds 72: 1806-1809
22. Vanderperren H, Hädrich J, Eppe G, Goeyens L, Hoogenboom LAP, Malagocki P, Scippo M-L, Windal I and Lekens M (2010); Application of the XDS-CALUX Bioassay in Routine: Semi-Quantitative Screening Using AL-BEQ Cut-Off Values. Organohalogen Compounds 72: 701-704
23. Elskens M, Baston DS, Stumpf C, Haedrich J, Denison MS, Baeyens W and Goeyens L (2011); CALUX measurements: Statistical Inferences for the Dose Response Curve. Talanta 85: 1966– 1973
24. Haedrich J, Podestat U, Stumpf C, Kotz A, Wahl K, Malisch R (2012); Bioanalytical screening of fish muscle tissue for dioxins and dioxin-like PCBs: Results from application of three different extraction methods. Organohalogen Compounds 74: 189-192
25. Haedrich J, Stumpf C, Podestat U, Kotz A, Wahl K, Malisch R (2012); Considerations on the working range in bioassay dose-response curves: Curve fit and assay background response. Organohalogen Compounds 74: 177-181
26. Haedrich J, Kotz A, Malisch R (2014); Bioanalytical Methods Performance: Practical Experience with the New EU Criteria, Potentials and Limitations. Organohalogen Compounds 76: 438-441