AOAC Guidance on FA Immunoassay Validation (August 2023)

AOAC INTERNATIONAL

STAKEHOLDER PROGRAM ON GLUTEN & FOOD ALLERGENS (GFA) GUIDANCE ON FOOD ALLERGENS IMMUNOASSAY VALIDATION

FIRST ROUND

(AUGUST 2023)

Delia Boyd AOAC INTERNATIONAL 2275 Research Blvd., Suite 300 Rockville, Maryland 20850 Tel: 240-801-8668 Ext. 126 Fax: +1-301-924-7089 Internet Email: aoac@aoac.org Web Site: www.aoac.org

Guidance on Food Allergen Immunoassay Validation 1 Food Allergen Working Group Draft for Public Comment, August 30, 2023

2 3 4 5 6 7 8 9

Contents 1. Scope

2. Applicability

3. Terms and Definitions

3.1. General

3.2. Quantitative Methods 3.3. Qualitative Methods

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

4. Required Method Information

4.1. General Method Information

4.2. Reporting Units

4.3. Calibrant

4.4. Test material preparation 4.5. Antibody description 5. Quantitative Method Validation 5.1. General Study Practices

5.2. Single Laboratory Validation Study (Method Developer Study)

5.2.1. Scope

5.2.2. Calibration Fit Study 5.2.3. Selectivity Study 5.2.4. Matrix Study 5.2.5. Robustness Study

5.3. Independent Laboratory Study (PTM)

5.3.1. Scope

5.3.2. Matrix Study 5.4. Collaborative Study

5.4.1. AOAC Requirements

5.4.2. Scope

5.4.3. Number of Laboratories

5.4.4. Test Materials 5.4.5. Data Analysis

5.4.6. Acceptance Criteria

6. Qualitative Method Validation 6.1. General Study Practices

6.2. Single Laboratory Validation (Method Developer Study)

6.2.1. Scope

6.2.2. Selectivity Study 6.2.3. Matrix Study 6.2.4. Robustness Study

6.2.5. High-Dose Hook Effect Study 6.3. Independent Laboratory Study (PTM)

6.3.1. Scope

6.3.2. Matrix Study 6.4. Collaborative Study

6.4.1. AOAC Requirements

48 49 50 51

6.4.2. Scope

6.4.3. Test Materials 6.4.4. Study Design

6.4.5. Data Analysis, Results Reporting, and Acceptance Criteria

6.4.6. Collaborator Comments 52 53 Table 1: Principal measurands, analytes and measurement systems for allergens 54 Table 2: Required Compounds for Selectivity Study for All Allergen Methods 55 Table 3: Required Compounds for Selectivity Testing for Target Allergen Test Methods 56 Table 4: Optional Compounds of Interest for Selectivity Testing 57 Table 5: Required Test Materials for Quantitative Study Designs 58 Annex A: Preparation of Food and Ingredient Testing Materials for Food Allergen Method Validation 59 Annex B: Preparation of Environmental Samples for Food Allergen Method Validation 60 Annex C: Statistical Methods: Data Analysis Guidance and Example Datasets 61

1. Scope 62 The purpose of this document is to provide AOAC International (AOAC) guidelines for method developers 63 conducting validation studies of immunoassay-based food allergen analysis methods, e.g., for 64 submission for AOAC Official Methods of Analysis (OMA) status and/or Performance Tested Methods 65 (PTM) certification. This document is not intended to describe requirements for laboratories using 66 commercial immunoassay methods for food allergen analysis, though it may assist these laboratories in 67 understanding the consensus-based validation approaches, terminology used, and information expected 68 to be provided by method developers. 69 2. Applicability 70 These guidelines are applicable to both quantitative and qualitative food allergen methods, of either a 71 proprietary or non-proprietary nature. 72 3. Terms and Definitions 73 3.1. General 74 3.1.1. Analyte: 75 3.1.1.1. chemical entity or entities* measured by the measurement system (see also ‘measurand’ 76 and narrative discussion). *which may be a marker(s) or surrogate(s), see Section 4.1.4 77 3.1.1.2. See also “measurand” definition and narrative examples (Section 4.1.4). See De Bievre 2013 78 for a detailed discussion of the difference between “analyte” and “measurand”.(1) 79 3.1.2. Enzyme-linked immunosorbent assay (ELISA): an analytical procedure characterized by the 80 recognition and binding of specific antigens by antibodies and signal generation by an enzyme- 81 substrate reaction 82 3.1.3. Food allergen: A food source known to have the ability to elicit IgE-mediated allergic reactions in 83 susceptible individuals, generally identified by the common name of the food, e.g., milk, egg, fish, 84 crustacean shellfish, soy, peanut 85 3.1.4. Food allergen material: A food ingredient or component derived from a food allergen 86 3.1.5. Matrix: Totality of components of a material system except the analyte (ISO 17511). For example, 87 the food, beverage, or environmental surface material to be included in the validation as per the 88 intended use of the method. 89 3.1.6. Measurand: The quantity intended to be measured (the specification of the measurand should be 90 sufficiently detailed to avoid any ambiguity). See also “analyte” definition and narrative examples 91 (Section 4.1.4) 92 3.1.7. Test material: A material used for method validation that either contains a target food allergen 93 present at a given concentration in the context of a food or environmental matrix or is a blank 94 matrix free of the target food allergen. 95 3.1.8. Spiked Test Material: A food matrix into which a food allergen material has been incorporated 96 after all relevant food processing operations have been completed (See Annex A for details). 97 3.1.9. Incurred Test Material : Prepared from a food matrix into which a food allergen material has been 98 incorporated prior to subjecting the matrix to a given food processing operation. 99 3.1.10. Qualitative method: Method of analysis whose response is either the presence or absence of the 100 analyte. 101

3.1.11. Quantitative method: Method of analysis whose result is the amount (mass or concentration) of 102 the analyte. 103 3.1.12. Reference material: material, sufficiently homogeneous and stable with respect to one or more 104 specified properties, which has been established to be fit for its intended use in a measurement 105 process (ISO Guide 30:2015) 106 3.1.13. Robustness: the ability of a method to resist significant changes in the final results when 107 reasonable deviations are made in the experimental conditions described in the procedure. 108 3.1.14. Sample: A small portion or quantity taken from a population or lot that is ideally a representative 109 selection of the whole. Samples are taken from lots for purposes of scientific examination and 110 analysis and are intended to provide characteristic information about the population, generally by 111 applying statistical calculations. (ISO 3534-1:1993) 112 3.1.14.1. Test portion: Portion of the test sample as prepared for testing or analysis, where the 113 whole quantity is used for analyte extraction at one time (ISO 16577:2022) 114 3.1.14.2. Analytical sample: The material from which the test portion is selected (after 115 grinding/homogenization if necessary 116 3.1.15. Selectivity: The degree to which the method can quantify the target analyte in the presence of 117 other analytes, matrices, or other potentially interfering materials 118 3.1.15.1. Cross-reactivity: Cross-reactivity is defined as a reaction to a material other than the target 119 analyte 120 3.1.15.2. Measurement Interference: A cause of significant bias in the measured analyte 121 concentration due to the effect of another component or property of the sample which may 122 result from nonspecificity of the detection system, suppression of an indicator reaction, or 123 inhibition of the analyte. (2) An interference can be endogenous, present in the sample, or 124 exogenous, introduced into the sample during the measurement process. 125 3.1.16. Technical replicate: One extracted test portion is analyzed more than one time 126 Note 1 to entry: Example for two technical replicates: An extract from a single test portion is 127 measured instrumentally multiple times. One test portion was extracted once, and the resulting 128 extract is pipetted in two wells of a microtiter plate. 129 3.2. Quantitative Methods 130 3.2.1. Bias: Difference between the expectation of the test results and an accepted reference value. Bias 131 is the total systematic error as contrasted to random error. There may be one or more systematic 132 error components contributing to the bias. (3) 133 3.2.2. Calibrant : a material used for calibration of a measurement procedure 134 3.2.3. Limit of detection (LOD): The lowest concentration or mass of analyte in a test sample that can be 135 distinguished from a true blank sample at a specified probability level (ISO 5725-1:1994). See 136 further details on how to determine LOD in Section 5.2.4.3.2. 137 3.2.4. Limit of quantification (LOQ): The lowest level of analyte in a test sample that can be reasonably 138 quantified at a specified level of precision (ISO 5725-1:1994). See further details on how to 139 determine LOQ in Section 5.2.4.3.2. 140

3.2.5. LOQ RSD : A limit of quantification with a specified relative standard deviation, expressed as a 141 percentage. For example, an LOQ 10 from a single laboratory validation would be the concentration 142 where the RSD i = 10%. In the case of an LOQ RSD estimated from a collaborative study, an LOQ 30 would 143 be the concentration where the RSD R = 30%. 144 3.2.6. Measurement range: the concentration range over which the target analyte can be reliably 145 quantified/detected 146 3.2.7. Recovery: The fraction or percentage of analyte that is recovered when the test sample is 147 analyzed using the entire method 148 3.2.8. Precision: Closeness of agreement between independent test/measurement results obtained 149 under stipulated conditions. (ISO 3534-2:2006) 150 3.2.9. Repeatability: Precision under repeatability conditions. (ISO 3534-2:2006). (Repeatability 151 Conditions: Observation conditions where independent test results are obtained with the same 152 method on equivalent test items in the same laboratory by the same operator using the same 153 equipment within short intervals of time. (ISO 3534-2:2006, with minor modifications) 154 3.2.10. Reproducibility: Precision under reproducibility conditions (ISO 3534-2:2006). (Reproducibility 155 Conditions: Observation conditions where independent test results are obtained with the same 156 methods on equivalent test items in different laboratories with different operators using different 157 equipment. (ISO 3534-2:2006, with minor modifications)) 158 3.2.11. Intermediate precision: precision under intermediate conditions (ISO 3534-2). (Intermediate 159 precision conditions: conditions where test results or measurement results are obtained with the 160 same method, on identical test/measurement items in the same test or measurement facility, 161 under some different operating condition.) 162 3.2.11.1. Note 1 to entry: There are four elements to the operating condition: time, calibration, 163 operator and equipment. 164 3.2.11.2. The specific and minimum conditions applicable for validation of food allergen 165 immunoassays are described in section 5.2.4.2.2 166 3.2.11.3. For the purposes of this document, the subscript notation "i" will be used to indicate terms 167 and estimators associated with intermediate precision. Estimation methods can be found in 168 section 5.2.4.2.3. 169 3.3. Qualitative Methods 170 3.3.1. Probability of detection (POD) : The proportion of positive analytical outcomes for a qualitative 171 method for a given matrix at a given analyte level or concentration. POD is concentration 172 dependent. 173 3.3.2. Claimed Detection Capability (CDC): An analyte concentration that demonstrates a POD of at least 174 0.90, with 95% confidence. This may be estimated across all matrices, or individually per matrix. 175 The claimed detection capability must be verified empirically during method validation. 176 3.3.3. Fractional recovery: Validation criterion that is satisfied when an unknown sample yields both 177 positive and negative responses within a set of replicate analyses. (5) 178 3.3.4. Lateral flow device (LFD): an analytical method characterized by use of an 179 immunochromatography platform for detection of specific target analytes 180 3.3.5. LPOD: composite POD pooled across laboratories. (6) 181

3.3.6. Sensitivity: Probability of a (+) response at a given concentration; POD(c) 182 4. Required Method Information 183 4.1. General Method Information 184 4.1.1. Scope 185 The information described in this section is required for all immunoassay-based food allergen methods, 186 including both quantitative and qualitative methods, submitted for AOAC OMA or PTM review. 187 4.1.2. Applicability Statement 188 Method developers must provide an applicability statement describing the method’s target analyte, 189 measurand, matrices within scope, and important limitations. 190 4.1.3. Standard Method Performance Requirements 191 If the method is intended to conform to an existing Standard Method Performance Requirements 192 (SMPR) document, the SMPR citation must be provided. In addition to the information described in this 193 document, methods submissions must provide any additional details mandated by relevant SMPRs. 194 4.1.4. Analyte and Measurand 195 4.1.4.1. The analyte and measurand must be clearly defined. 196 4.1.4.2. The term ‘analyte’ is better known and more widely used than ‘measurand’. ‘Analyte’, or the 197 name of a (bio)chemical substance or compound, are terms sometimes used for 'measurand', 198 incorrectly because the aspect of quantity is omitted. Two examples cited by Eurachem (8) 199 illustrate the difference between the two terms. The first is very simple, when analysing for 200 glucose in plasma, important for diabetics, the analyte is glucose and the measurand is ‘the 201 amount of substance (concentration) of glucose in plasma’. Another medical laboratory 202 example, a 24-hour urine protein analysis, which can help detect disease or other problems. 203 The analyte is protein and the measurand is ‘mass of protein in 24-hour urine’. Note the 204 protein is likely to be measured by a dye-binding method and reported in mg/100mL hence 205 the volume of urine collected must be known to calculate the mass of protein in the 24-hour 206 sample. Other examples are given by De Bièvre (1). 207 4.1.4.3. The International Vocabulary of Metrology, VIM (4), notes that changes in the measuring 208 system and the conditions under which the measurement is carried out might result in the 209 quantity being measured differing from the measurand-as-defined. This is particularly 210 important in allergen analysis . Moreover we need to have regard to and report sufficient 211 detail about the method and units to enable others to make sensible use of the reported data. 212 The following Table ( Table 1 ) illustrates the most common measurands, analytes, measuring 213 systems and units relevant to allergen analysis. 214

Table 1: Principal measurands, analytes and measurement systems for allergens 215 Measurand Analyte Matrix

Measurement procedure /system

Mass fraction (e.g., mg/kg) of total protein in a foodstuff Mass fraction (e.g., mg/kg) of total protein from the allergenic source (e.g., total milk protein) in matrix (e.g., cookies) Mass fraction (e.g., mg/kg) of total protein from the allergenic source (e.g., total milk protein) in matrix (e.g., cookies)

nitrogen (*)

Foodstuff

Kjeldahl or Dumas

Peptides (*)

LC-MS or LC-MS/MS

Foodstuff, but usually needs to be more specific as to the matrix

Protein (*)

Immunoassay (Relevant epitopes)

(*) A conversion between the analyte and the measurand is required and also between the allergenic food source 216 and its protein content. There remains work to be done to harmonize such conversion factors. There is a need to 217 specify the food commodity and its characteristics in sufficient detail to be of value to the customer and/or a risk 218 assessor. For example, if the quantity value obtained (measurand) depends on the nature of the calibrator, details 219 of the calibrator should be given. 220 4.2. Reporting Units 221 4.2.1. At a minimum for food, beverage, and non-surface environmental samples, method developers 222 must provide concentrations in units of: Mass of total protein from allergenic food per mass of 223 food or environmental sample (e.g., mg total peanut protein per kg food, mg total milk protein per 224 kg food, mg total peanut protein per kg rinse water, mg total peanut protein per L rinse water) 225 4.2.2. Method developers may provide additional concentration units if desired, e.g., mass of allergenic 226 food per mass of food (e.g., mg peanut per kg food) or mass of protein fraction per mass of food 227 (e.g., mg casein per kg food), but definitions must be provided for other units. 228 4.2.3. For environmental surface samples, method developers must provide amounts in units of mass of 229 total protein from allergenic food per defined swabbing area in accordance with a prescribed 230 swabbing protocol (e.g., µg total peanut protein per 100 cm 2 ) 231 4.3. Calibrant 232 Method developers must provide the following information on calibrants: 233 4.3.1. What is the calibrant for the method? 234 4.3.2. How was the concentration value of the calibrant assigned? 235 4.3.3. Is the calibrant made from raw or processed material? 236 4.3.4. Was the calibrant extracted or purified? If so, how? 237 4.3.5. How is the concentration of the calibrant expressed? 238 4.3.6. Is the calibrant commercially available? 239 4.4. Test material preparation 240

4.4.1. Method developers must provide complete information on the allergen materials and procedures used to prepare in house test materials to the ERP or volunteer expert. If detailed preparation techniques are perceived to be proprietary information, requests may be made for the ERP/volunteer expert for these data to remain confidential. Method developers must, however, provide information sufficient for end users to understand the allergen materials, test material composition (i.e., matrix formulation), and test material processing conditions in method 247 4.4.2. Specific guidance on allergen materials and test material preparation can be found in Annex A. 248 4.5. Antibody description 249 4.5.1. Method developers must provide information on the antibody/antibodies used. As much 250 information on antibody development and specificity as possible is strongly encouraged. 251 4.5.2. Information on the antibody must include the following: 252 4.5.2.1. Whether the antibody is monoclonal or polyclonal 253 4.5.2.2. What protein(s) does the antibody detect? Method developers must either provide 254 empirical confirmation of specificity (e.g., Western blot) or information on protein(s) used 255 for antibody development. 256 5. Quantitative Method Validation 257 5.1. General Study Practices 258 5.1.1. Method developers may prepare study test materials in house for the single laboratory validation 259 (method developer study), but all samples and test portions must be blind-coded and 260 randomized. Analysis conducted by the method developer must be performed by an independent 261 analyst without prior knowledge of the test materials undergoing analysis in a given sample set. 262 5.1.2. Ideally, all test materials for the independent laboratory and collaborative study should be prepared 263 by an external entity independent from the method developer. At least one incurred test material for 264 the independent laboratory and collaborative study must be prepared by an external entity independent 265 from the method developer. In situations where an independent entity is unavailable to prepare all 266 of the test materials for the independent laboratory and collaborative study, or their use is 267 impractical for all test materials, method developers may produce and distribute test materials as 268 long as detailed information is provided on procedures used to prevent bias (preparation, coding, 269 etc.), and justification is provided for failing to use an independent entity to prepare all of the test 270 materials. 271 5.2. Single Laboratory Validation Study (Method Developer Study) 272 5.2.1. Scope 273 A Single Laboratory Validation (SLV) study, also referred to as the Method Developer Study for the 274 purposes of the PTM program, is intended to evaluate the performance of a candidate method in the 275 following areas: (1) selectivity, (2) precision (repeatability and intermediate precision), (3) sensitivity, (4) 276 recovery, and (6) robustness. These studies are generally conducted within a method developer 277 laboratory. 278 5.2.2. Calibration Fit Study 279 5.2.2.1. Analyze calibration standards as they are included in the test kit. 280 241 242 243 244 245 246 validation reports.

5.2.2.2. Analyze at least four replicates of each concentration defined for the calibration curve. 5.2.2.3. Fit the calibration curve as described in method instructions and kit insert. Full descriptions must be provided with respect to performing the calibration function calculations, including any transformations conducted and the regression model used. Full calibration curve plots 5.2.2.4. From the calibration curve function, determine the calculated concentrations for each of the standards. Calculate the residuals for each non-zero concentration standard. Residuals are the difference between the observed value and the predicted value for each dependent variable in the calibration curve. (Residual = observed value - predicted value.) Residuals should be calculated from the instrument response. For most quantitative food allergen 5.2.2.6. Residuals should have random distributions and be centered on zero. If a non-random pattern is observed, the calibration function or measurement range may not be appropriate. 5.2.2.7. Residuals should generally also be <15% of the measured response, and up to 20% at the 297 The selectivity study is intended to provide information on potential sources of cross-reactivity and 298 interference. The information related to cross-reactivity and interference should be reported in the 299 validation report and in the package insert from the method developer. 300 5.2.3.1. Selectivity Panel Selection 301 5.2.3.1.1. Method developers must test their allergen detection method for cross-reactivity with 302 the target allergen in a variety of food commodities. Food commodities tested for 303 cross-reactivity should include a wide selection of foods and ingredients, particularly 304 those that are genetically similar to the target allergenic commodity and that are 305 likely to be analyzed for the presence of the target food allergen. (9) 306 5.2.3.1.2. For all food allergen analytes, all foods in Table 2 must be analyzed. For specific food 307 allergen analytes, additional matrices listed in Table 3 must be analyzed. Additional 308 foods and ingredients listed in Table 4 may be included in the study. 309 5.2.3.1.3. Method developers must make a good-faith effort to obtain the relevant foods listed 310 in Tables 2-4. If the method developer feels they are unable to secure any particular 311 food type, they should detail efforts to procure the food in the study report. 312 5.2.3.1.4. Method developers are advised to take extreme care in selecting, sorting, and 313 preparing samples for the cross-reactivity testing to avoid analyzing samples with 314 existing undeclared target food allergen residues. For example, it may be advisable 315 to obtain whole nuts, seeds, spices, etc. to facilitate hand sorting, washing/shelling, 316 and grinding/milling. 317 5.2.3.1.5. To identify possible allergen cross-contact, it is advisable to screen commodities with 318 an alternative method (ELISA, PCR or other). 319 5.2.3.2. Cross-Reactivity 320 5.2.3.2.1. Study Design 321 lowest non-zero calibration standard. 5.2.3. Selectivity Study 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 and calibration functions must be shown. methods, instrument response would be optical density (absorbance) values. 5.2.2.5. Plot the residuals versus concentration.

322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363

5.2.3.2.1.1. One test portion of each blank food material in the selectivity panel should be analyzed according to the entire method protocol to evaluate cross-reactivity. 5.2.3.2.1.2. Cross-reactivity testing should be based on the full-strength extracts, i.e., a sample of the item being tested for cross-reactivity should be extracted using the extraction buffer and procedure outlined in the method instructions, then analyzed at full strength to determine if it leads to a positive result. (9) 5.2.3.2.1.3. Certain concentrated food ingredients may require dilution (e.g., colors, spices, gums, etc.). Such ingredients may be tested at a 10% concentration in a matrix such as rice flour. If this type of dilution is conducted during method validation, corresponding procedures for specific food ingredients must be stipulated in the method protocol, kit insert, validation certificate, and

validation report.

5.2.3.2.1.4. In general, food items tested for cross-reactivity should be prepared as they

would normally be analyzed (raw or cooked).

5.2.3.2.1.5. If a positive result is obtained or it is outside the measurement range of the method, the extract must be diluted and rerun to characterize the extent of

the cross-reactivity.

5.2.3.2.1.6. Blank samples with a positive result (> LOD/LOQ) should first be repeated with a second lot of the cross-reactive sample, and if the result persists it may also be evaluated with an alternative method (PCR, Western blot, mass spectrometry, alternate ELISA, etc.) to verify whether the signal is the result of

cross-reactivity or a true positive due to cross-contact.

5.2.3.2.2. Data Analysis and Reporting

5.2.3.2.2.1. The absorbance or optical density (OD) values for all blank samples must be reported. The absorbance values for the following method standards analyzed with the blank samples also must be given: zero standard, first non-zero 5.2.3.2.2.2. The concentration for all blank samples that had an absorbance or OD above the limit of quantification of the method must be reported. Results falling between the LOD and LOQ must be reported as such. All results must be reported. If any analysis is repeated, both datasets must be reported, and a standard.

justification given for all repeat analysis.

5.2.3.2.3. Acceptance Criteria

5.2.3.2.3.1. Results of blank sample analysis are acceptable and not indicative of cross- reactivity if the extrapolated quantitative result of the blank sample is < LOQ.

5.2.3.2.3.2. Cross-reactivity reporting

5.2.3.2.3.2.1. If cross-reactivity is observed, corresponding information must be included in the applicability statement of the kit insert, validation report,

and validation certificate.

5.2.3.2.3.2.2. If the extrapolated quantitative result is between the LOD and LOQ, this must be described in the validation report but does not need to be stated

as cross-reactivity in the kit insert.

364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387

5.2.3.3. Interference

5.2.3.3.1. Study Design

5.2.3.3.1.1. One test portion should be spiked with the target analyte at a concentration that is three times the lowest LOQ claimed by the method to evaluate

interference.

5.2.3.3.1.2. Testing should be based on the full-strength extracts, i.e., a sample of the item being tested for cross-reactivity should be extracted using the extraction buffer and procedure outlined in the method instructions, then analyzed at 5.2.3.3.1.3. Certain concentrated food ingredients may require dilution (e.g., colors, spices, gums, etc.). Such ingredients may be tested at a 10% concentration in a matrix such as rice flour. If this type of dilution is conducted during method validation, corresponding procedures for specific food ingredients must be stipulated in the method protocol, kit insert, validation certificate, and full strength to determine if it leads to a positive result. (9)

validation report.

5.2.3.3.1.4. In general, food items tested for cross-reactivity should be prepared as they

would normally be analyzed (raw or cooked).

5.2.3.3.2. Data Analysis and Reporting

5.2.3.3.2.1. The calculated quantitative values for all spiked samples must be reported. 5.2.3.3.2.2. All results must be reported. If any analysis is repeated, both datasets must be

reported, and a justification given for all repeat analysis.

5.2.3.3.3. Acceptance Criteria

5.2.3.3.3.1. Spiked samples should render a result above the lowest LOQ.

5.2.3.3.3.2. The percent recovery should be calculated and reported for each tested food.

Table 2: Required Compounds for Selectivity Study for All Allergen Methods

388

Animal origin (raw and cooked forms, except as noted # )

Nuts and related (raw and roasted forms)

Beans and related legumes Dried Chickpeas

Other vegetables and related

Seeds stones Kernell Pine kernel

Pseudo- Cereals

Fruits and related

Cereals

Spices

Various Cocoa (powder) Ground, Roasted

Food additives

Buckwheat flour

Dried tomatoes

Food colors

Barley

Almond

Basil

Beef

Apple juice

Coffee Beans

Potatoes (flour)

Bovine skim milk (powder) # Chicken whole egg powder #

Pumpkin kernels

Orange juice

Corn flour

Millet flour

Green bean

Brazil nut

Chili

Carmine red

Red and white wine

Yellow orange turmeric curcumin Thickeners, stabilizers, gelling agents)*

Oat flour

Quinoa

Green peas

Pumpkin

Cashew

Cinnamon

Sesame

Oil/fat

Kidney beans Lentils Lupine

Sweet potato

Sunflower kernels

Rye

Rice flour

Coconut

Cloves

Chicken

Vinegar

Salt

Dried Tea Leaves

Spelt

Sorghum Teff flour

Hazelnut Pecan nut

Curry Garlic

Pork

Rape seed

Guar gum

Wheat Wheat starch

Turkey

Linseed Poppy seeds

Locust beam gum

Tapioca flour/starch

Humectants Lecithin from soybean

Peanut (raw)

Pine nuts

Turmeric Paprika (sweet) Mustard seeds (black/ brown/ yellow) Onion (powder)

Peanut (roasted)

Pistachio nut

Leavening Agents

Soy milk

Walnut

Whey protein concentrate

Soya flour

White beans

Pepper (black) Pepper (white)

pH control agents

Vegetable juice Salts Monosodium glutamate

389

*These compounds could be tested directly and/or at a 10% concentration.

Table 3: Required Compounds for Selectivity Testing for Target Allergen Test Methods 390 Target Allergen Almond Coconut Egg Milk Crustacean Fish

Mollusk

Peanut Kidney beans

Materials to be analyzed

Apricot (flesh & pit) Cherries (flesh & pit) Peach (flesh & pit)

Crustacean: crab, lobster, shrimp

Crustacean: crab, lobster, shrimp Mollusks: oysters, mussels, octopus, squid, scallops, clams Fish: Cod, salmon, tuna, sardine, another perciformes species (in addition to tuna), catfish, flounder, carp Sea urchin

Crustacean: crab, lobster, shrimp Fish: cod, salmon, stingray, tuna

Hazelnut Duck

Buffalo

Mollusks: oysters, mussels, octopus, squid, scallops, clams

Vanilla

Ostrich

Camel

Lentils

Dates

Quail

Cow

Cricket

Sea urchin

Paprika

Prunes (flesh & pit)

Goat

Cockroach

Annatto Lupine beans

Fish: cod, salmon, tuna

Sheep

Frog legs

391 392 393

Table 4: Optional Compounds of Interest for Selectivity Testing

394

Other vegetables and related

Beans and related legumes

Seeds stones Kernell

Pseudo Cereals

Nuts and related

Animal origin Bovine serum albumin

Fruits and related

Cereals

Spices

Food additives

Thickeners, stabilizers, gelling agents

Black beans

Apricot kernels

Carob seedlings Apple fiber

Food colors

Cornstarch Amaranth

Carrots

Aniseed

Arrowroot flour

Annatto (achiote)

Kamut

Fava beans Cucumber Chestnut

Caraway

Casein

Dried apricots

Alginate

Garbanzo beans

Semolina

Macadamia Cardamom

Cod

Dried figs

Amaranth

Arrowroot

Pink peppercorn

Celery (powder)

Blue-spirulina (from algae) Green matcha green tea powder Pink beetroot powder Pitaya powder (from dragon fruit)

Soft wheat

Lima beans

Egg whites

Dried papaya

Carrageen

Pinto beans

Gelatin (bovine)*

Dried pineapple

Sumac

Coriander

Collagen

Gelatin (fish)*

Cumin

Grape juice

Pectin

Gelatin (porcine)*

fennel seed

Kiwi

Tapioca

Yellow pea flour

Sugar beet syrup

Purple acai berry

Fenugreek

Locust

Xanthan gum

Leavening Agents

Adzuki beans

Ginger

Mussels

Raisins

Saffron

Humectants

Marjoram

Mealworm

Beer

Romano Other related legumes

Nutmeg

Octopus

Emulsifiers

Ginger beer

pH control agents

Shrimp Whey powder

Konjac flour Lecithin from egg yolk

Fruit juice

Alternate sources of animal protein (i.e., crickets)

Salts Ascorbic acid Sulfites

395

*These compounds could be tested directly and/or at a 10% concentration.

5.2.4. Matrix Study 396 The matrix study is intended to provide data on precision (repeatability and intermediate precision), 397 sensitivity, and recovery. 398 5.2.4.1. Test Materials 399 5.2.4.1.1. Incurred test materials are required for evaluation of precision, sensitivity, and 400 recovery. See Annex A for description of best practices for incurred test material 401 preparation. 402 5.2.4.1.2. Minimum Number of Test Materials 403 5.2.4.1.2.1. At least 4 concentrations per matrix, including a zero/blank, must be included 404 in the study. One concentration should less than or equal to two times the 405 stated LOQ for the method. Other concentrations should span the calibration 406 range, e.g., at the middle of the calibration curve and upper end of the 407 calibration curve. 408 5.2.4.1.2.2. The minimum number of matrices and levels to be evaluated is dependent on 409 the parameter being estimated, as shown in Table 5.

410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436

Table 5: Required Test Materials for Quantitative Study Designs

Parameter Repeatability

Number of Matrices and Concentrations All matrices, 4 concentrations (blank, low, medium, and high) for each matrix All matrices, at least 3 concentrations (blank, low, medium) for each matrix All matrices, at least 3 concentrations (blank, low, medium) for each matrix All matrices, three non-blank concentrations (i.e., low, medium, and high)

Intermediate Precision

Sensitivity: LOD/LOQ Estimation

Recovery

5.2.4.2. Study Designs

5.2.4.2.1. A single, statistically valid study may be designed and utilized to provide estimates of precision (repeatability and intermediate precision), sensitivity, recovery, and lot-to- lot variability. Alternatively, individual studies may be designed for each performance parameter. Designs 1b and 2b below will provide sufficient data for all parameters in the Matrix Study and the lot-to-lot assessment required in the Robustness Study (5.2.5), if conducted on a sufficient number of test materials. 5.2.4.2.2. For food allergen immunoassays, intermediate precision study designs at a minimum must include multiple test portions , at least two test kit lots, and day/operator as a single confounded factor. Study designs given below may be used, but other designs

may also be able to give satisfactory data.

5.2.4.2.3. Precision: Repeatability and Intermediate Precision

5.2.4.2.3.1. Intermediate precision and repeatability can be estimated from one of several nested study designs. Depending on the design selected, results may be able to estimate repeatability, intermediate precision, variance from test kit lots,

437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472

variance from day/operator as one confounded variable, and variance from ELISA alone (i.e., well-to-well variance). In order for the nested designs to be capable of estimating repeatability, at least two test portions must be analyzed under repeatability conditions (i.e., conducted on the same day, by the same operator, with the same calibration and equipment). Under these conditions, the nested designs can estimate both intermediate precision and repeatability because repeatability is a variance component within intermediate precision, as expressed in the following equation, where s i 2 is the

intermediate precision variance, s lot

2 is the variance contributed by test kit lot,

s d/op 2 is the variance from the confounded factor of day and operator, and s r 2

is the repeatability variance:

2 = 2 + / 2 + 2

5.2.4.2.3.2. Repeatability estimates are required at four concentrations for each claimed matrix: blank, low, medium, and high levels, according to the claimed method

quantification range.

5.2.4.2.3.3. As intermediate precision estimates are used for the calculation of LOD and LOQ, estimates are required for all matrices, with at least three concentration

levels per matrix: blank, low, and medium.

5.2.4.2.3.4. Test kit lot variance (product consistency and stability) must be evaluated for at least one matrix using three test kit lots. This can be included in the estimation of intermediate precision (Designs 1b and 2b) or may be conducted separately (see Robustness Study, 5.2.5). Guidance on selection of test kit lots for product consistency and stability analysis can be found in

section 5.2.5.2.2.

5.2.4.2.3.5. Design 1a

5.2.4.2.3.5.1. Design 1a ( Figure 1 ) can be used to estimate (1) intermediate precision (which includes repeatability, test kit lot variance (with 1 degree of

freedom, df), and day/operator confounded variance) and (2)

repeatability.

5.2.4.2.3.5.2. Two test kit lots are used to analyze each test material. Two operators conduct analysis on two days for each test kit lot. For each day and lot, the assigned operator conducts extraction and analysis of three test

portions of the test material, with one ELISA measurement performed per

test portion.

473 474 475 476 477 478 479 480 481 482 483 484 485 486 487

Figure 1: Design 1a. Lot: test kit lot, TP: test portion, E: ELISA measurement. Design 1a can be

used to estimate intermediate precision and repeatability.

5.2.4.2.3.6. Design 1b

5.2.4.2.3.6.1. Design 1b ( Figure 2 ) can be used to estimate (1) intermediate precision (which includes repeatability, test kit lot variance (with 2 df), and day/operator confounded variance), (2) repeatability, and (3) lot-to-lot 5.2.4.2.3.6.2. Three test kit lots are used to analyze each test material. Two operators conduct analysis on two days for each test kit lot. For each day and lot, the assigned operator conducts extraction and analysis of two test portions of the test material, with one ELISA measurement performed per product consistency.

test portion.

488 Figure 2: Design 1b . Lot: test kit lot, TP: test portion, E: ELISA measurement. Design 1b can be used to 489 estimate intermediate precision, repeatability, and lot-to-lot product consistency. 490

491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506

5.2.4.2.3.7. Design 2a

5.2.4.2.3.7.1. Design 2a ( Figure 3 ) can be used to estimate (1) intermediate precision (which includes repeatability, test kit lot variance (with 1 df), day/operator confounded variance, and ELISA variance), (2) repeatability (which includes test portion and ELISA variance), and (3) ELISA variance. 5.2.4.2.3.7.2. In this instance the repeatability variance can be further split into test portion variance and ELISA variance as shown in the equation below,

where s r

2 is repeatability variance, s TP

2 is the variance attributed to test

2 = 2 + 2

portion, s ELISA

2 is the variance attributed to ELISA measurement variance:

5.2.4.2.3.7.3. Two test kit lots are used to analyze each test material. Two operators conduct analysis on two days for each test kit lot. For each day and lot, the assigned operator conducts extraction and analysis of three test portions of test material, with two ELISA measurements performed per

test portion.

507 Figure 3: Design 2a. Lot: test kit lot, TP: test portion, E: ELISA measurement. Design 2a can be used to 508 estimate intermediate precision, repeatability, and ELISA variance.

509 510 511 512 513 514 515 516 517 518 519 520 521

5.2.4.2.3.8. Design 2b

5.2.4.2.3.8.1. Design 2b ( Figure 4 ) can be used to estimate (1) intermediate precision (which includes repeatability, test kit lot variance (with 2 df), day/operator confounded variance, and ELISA variance), repeatability (which includes test portion variance and ELISA variance), (3) ELISA 5.2.4.2.3.8.2. Three test kit lots are used to analyze each test material. Two operators conduct analysis on two days for each test kit lot. For each day and lot, the assigned operator conducts extraction and analysis of two test portions of test material, with two ELISA measurements performed per variance, and (4) lot-to-lot product consistency.

test portion.

522 Figure 4: Design 2b. Lot: test kit lot, TP: test portion, E: ELISA measurement. Design 2b can be used to 523 estimate intermediate precision, repeatability, ELISA variance, and lot-to-lot product consistency.

524 525 526 527 528 529

5.2.4.2.3.9. If repeatability is conducted separately ( Figure 5 ), at least six test portions of each test material should be analyzed according to the entire method as written. Analysis should be conducted by one analyst on one day, using one

test kit lot (n=6 per test material).

530 531 532 533 534 535 536 537

Figure 5: Repeatability Only Design

5.2.4.2.4. Sensitivity: LOD/LOQ Estimation

5.2.4.2.4.1. In SLV studies for food allergen immunoassay methods, the LOD and LOQ will

be estimated using intermediate precision.

5.2.4.2.4.2. Data collected from analysis of incurred test materials for all matrices will be

used to model the relationship between analyte concentration and

538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577

intermediate precision. Data used must meet other method performance

criteria (e.g., recovery of 50-150%).

5.2.4.2.5. Recovery Assessment

5.2.4.2.5.1. Data collected for the purposes of precision evaluation may also be used for

the recovery assessment.

5.2.4.2.5.2. If conducted separately from the precision assessment, evaluate each incurred matrix with six independent analyses (test portions) per concentration level at a minimum of three non-blank concentration levels

covering the analytical range.

5.2.4.3. Data Analysis and Reporting

5.2.4.3.1. Precision

5.2.4.3.1.1. Nested Designs: Repeatability and Intermediate Precision

5.2.4.3.1.1.1. Data generated from nested designs, such as those as described above, should be analyzed by an ANOVA capable of providing estimates of

intermediate precision and repeatability.

5.2.4.3.1.1.2. Annex C contains full instructions, R code, and example datasets for the

study designs described in this guidance.

5.2.4.3.1.2. Repeatability Only

5.2.4.3.1.2.1. In a situation where a study design for estimating repeatability alone is selected, the mean, standard deviation, and relative standard deviation should be calculated for each test material (i.e., each matrix- concentration combination). Formulas for standard deviation and relative standard deviation, as defined in Appendix F (3), are as follows:

Repeatability standard Deviation (s r ): s r = [ Σ (x r – x̅ )

2 /n] 0.5

Repeatability relative standard deviation (RSD r ): RSD = s r × 100/ x̅

5.2.4.3.1.2.2. The study report must include the standard deviation and RSD values for each test material, and all repeatability estimates must meet requirements set forth in the relevant SMPR or established by the ERP. In the absence of an SMPR and ERP, acceptable RSD r values for food allergen

immunoassays are generally ≤20%.

5.2.4.3.2. Sensitivity: LOD, LOQ

5.2.4.3.2.1. LOD will be estimated using a hypothesis test approach, with α = β = 0.05. The relationship between observed concentration and intermediate precision standard deviation must be taken into account in the estimation of LOD (also referred to as a precision profile estimation method for LOD). Full instructions

for the calculations to estimate LOD are in Annex C.

5.2.4.3.2.2. LOQ estimation will be based on the relationship between concentration and

intermediate precision standard deviation. Full instructions for the

calculations to estimate LOQ are in Annex C.

5.2.4.3.3. Recovery

578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607

5.2.4.3.3.1. Percent Recovery = (Experimental concentration)/(Expected concentration) x

100

5.2.4.3.3.2. The expected concentration for each test material should be calculated from

the incurred concentration, accounting for any mass changes during processing operations (e.g., moisture loss during baking).

5.2.4.4. Acceptance Criteria

5.2.4.4.1. If an applicable SMPR is available for a method, the SLV study data must meet the

corresponding criteria.

5.2.4.4.2. In the absence of an applicable SMPR, an expert review panel will evaluate the study data according to their expert opinions. With respect to recovery, while ideal mean recovery values are from 80-120%, values of 50-150% are acceptable. (9)

5.2.4.4.3. LOQ

5.2.4.4.3.1. The LOQ must be greater than or equal to the LOD.

5.2.4.4.3.2. The RSD i at the LOQ must be less than or equal to the RSD i in the relevant SMPR (or the RSD R if an RSD i is not listed). If there is no SMPR available for a

particular food allergen, RSD i at the LOQ must be ≤ 30%.

5.2.4.4.3.3. If a method developer has an LOQ claimed as part of the method design (e.g., the lowest non-zero calibrant), the estimated LOQ from the SLV (which meets the SMPR requirements for maximum RSD i ) must be less than or equal to the claimed LOQ of the kit, within statistical tolerances. If the estimated LOQ from the SLV is greater than the claimed LOQ of the kit, the method developer must revise the LOQ claimed in the test kit insert and validation reports to meet the

precision requirements for LOQ.

5.2.4.4.3.4. In the validation reports and test kit inserts, the method developers must indicate the actual RSD i value estimated for the LOQ of the kit as part of the

LOQ information. For example:

5.2.4.4.3.4.1. LOQ 15 , for a method where the existing LOQ claimed by the kit had an

estimated RSD i of 15% in the SLV

5.2.4.4.3.4.2. LOQ 30 , for a method where the LOQ was set based on the SLV outcome

and a maximum RSD i of 30%

5.2.5. Robustness Study 608 The Robustness study is intended to provide information on (1) robustness and (2) product stability and 609 consistency. 610 5.2.5.1. Acceptable Test Materials 611 5.2.5.1.1. Spiked matrices are acceptable for test kit lot-to-lot stability analysis and robustness 612 analysis (except when varying extraction conditions). See Annex A for description of 613 best practices for spiked matrix preparation. 614 5.2.5.1.2. Incurred matrices may also be used for the robustness study. If sufficient quantities of 615 incurred matrices have been prepared for the matrix study, these samples may also 616 be used for the robustness studies (i.e., separate incurred matrices are not 617 required). 618

619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661

5.2.5.2. Study Design

5.2.5.2.1. Robustness

5.2.5.2.1.1. The robustness of the method should be investigated by performing experiments in which specific parameters are changed to determine the impact on the experimental result. In particular, the effect of deviations in incubation times, reagent volumes, extraction conditions (time and temperature) should be investigated. It is recommended that deviations for 5.2.5.2.1.2. Full factorial or screening factorial designs, such as Plackett-Burman designs, may be appropriate depending on the number of parameters varied. (10, 11) 5.2.5.2.1.3. Analysis should be conducted on a minimum of one claimed matrix type. 5.2.5.2.1.4. For each of a minimum of two concentration levels (including one blank test material), two test portions should be analyzed at each combination of time and volume be investigated at ±5% or more, and incubation temperatures tried at ±3°C or more. (9) 5.2.5.2.2.1. If the test method is sold as a kit or device prepared in lots or batches, a product consistency and stability study must be performed to ensure that the performance of the product is consistent from lot-to-lot and over time under normal storage conditions for the shelf life of the product. Lot-to-lot consistency and product stability can be measured in the same set of experiments. As specified in Section 5.2.4.2, lot-to-lot consistency and product consistency can also be assessed in the context of nested designs for intermediate precision estimation that utilize at least three lots of test kits. 5.2.5.2.2.2. The shelf life should include the stability of all the reagents provided with the test kit, ideally through real-time testing of reagents under normal storage conditions. Accelerated stability testing at higher than normal storage temperatures can also be used to estimate stability. An expiration date for each test kit should be clearly indicated, along with appropriate conditions for 5.2.5.2.2.3. A minimum of three separate product lots must be evaluated. The product lots should span the shelf life of the kit. For example, if the kit shelf life is 12 months, an approximately 12-month-old kit, six-month-old kit and recently produced kit should be evaluated. Alternatively, accelerated aging may be used if kits at the end of their shelf life are not available yet - if this is done, then lot-to-lot stability should still be performed across 3 recent lots. Kits should be aged using increased temperature storage as described in ASTM F1980-16 or CLSI EP25-A. Real time data is needed for validations such as AOAC Official Method applications, and prior to the first AOAC Performance storage before use. (9) 5.2.5.2.2.4. If conducted separately from the matrix study, the product consistency and stability study should be conducted on at least one matrix, at three concentrations (blank, low, and high). Five test portions should be analyzed analysis conditions. Tested Method renewal.

5.2.5.2.2. Product Stability and Consistency