AOAC SPIFAN Chlorate-01 Reviewer Form (November 2022)

First page Table of contents 1 Next page Last page

Animated publication

AOAC Stakeholder Program on Infant Formula and Adult Nutritionals

CHLORATES/PERCHLORATES REVIEWER FORMS (CHLORATE-01)

Wednesday, November 30, 2022

AOAC INTERNATIONAL 2275 Research Blvd., Suite 300 Rockville, MD, 20850 USA

dboyd@aoac.org 301.924.7077 x126

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method____Chlorate 01____ (Resubmission)

Title: Method for the Determination of Chlorate and Perchlorate in Baby Food, Baby Food Ingredients, Infant Formula, and Infant Formula Ingredients by Liquid Chromatography – Tandem Mass Spectrometry: Single Laboratory Validation Author: ADELLE K. NANCEKIVELL, ELSA M. LOTZ, LISA A. GRAHAM Reviewer Name: Reviewer 1 Please Provide a Brief Summary of the Resubmitted Method: Samples are extracted with acetonitrile and 2% formic acid in water then purified with a weak anion exchange SPE. The extracts were analyzed by LC MS/MS in multiple reaction monitoring (MRM) by electrospray ionization (ESI) in negative ion mode. Quantification is performed by isotopic dilution. Was Method Scope/Applicability Updated? Method was updated and authors optimized laboratory environment : Signal to noise was improved for the baby food and meat matrices using the 4 mL volume that is eluted from the SPE cartridge and avoiding 2 ml aliquot. The laboratory environment has been optimized by removing sources of contamination such as perchloric acid from the extraction area. All dispensers and equipment are cleaned regularly by being wiped with methanol.

Has the resubmitted method met the requirements outlined as per the ERP recommendation/comment(s)? All the ERP requirements were met except I could not find any update of the reference materials codes.

Please provide documentation: All information is included in Manuscript.

Provide General comments:

There are some stated chlorate high blank levels that if they are OK, I cannot understand how the spike level was done:  Baby Food - Cereal Based  Baby Food - Vegetable Based

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

 Infant Formula  Whole Milk Powder  Liquid Milk  Lactose The chromatograms of spiked samples have little bigger areas than blank samples. Is there an error in the written blank concentration levels?

• High blank level for infant formulas tested. Anyway, the sample levels are around the required 0,005 mg/kg level, and the chromatograms are ok.

• The authors mentioned that for three matrices (plant based infant formula, whey protein concentrate, and soy protein isolate) they could not find a suitable blank matrix, but the samples could have been spiked at least with lower levels of perchlorate to have more information. Typing: • In item “Procedures used for optimization and validation”, item (e) is part of item (d), and not a separate item. • Page 17 “All data meet the SANTE/11312/2021 requirements with both RSDr and RSDr being less than 20%.” The authors do not report intermediate precision in the new manuscript version, so, this sentence must be modified. • Table 9: Wrong units for the column “Spiked Level”. It must be mg/kg • Table 10: Looking at this table, it seems that the recovery mean, and recovery range were calculated at a concentration level of 0,005 mg/kg. But this is not true for the matrices that was unable to obtain blanks that contain less than 30% of the target LOQ. I suggest clarifying this. • Some results (recoveries and recovery ranges) are not exact. Minor changes but it would be better to amend them. • For whey protein concentrate, spiked level was 0.05 and not 0,005 mg/kg as stated in chromatogram.

Final Recommendation: I recommend moving the method to First Action.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method Chlorate-01 (Resubmission)

Title: Method for the Determination of Chlorate and Perchlorate in Baby Food, Baby Food Ingredients, Infant Formula and Infant Formula Ingredients by Liquid Chromatography – Tandem Mass Spectrometry: Single Laboratory Validation Author: ADELLE K. NANCEKIVELL, ELSA M. LOTZ, LISA A. GRAHAM; AsureQuality Ltd Reviewer Name: Reviewer 2 3 ), followed by extraction with acetonitrile and 2% formic acid in water. After centrifugation, the supernatant is subjected to a weak anion exchange solid phase extraction, eluted with 1.75% ammonia in acetonitrile then concentrated. The extract is analysed by LC-MS/MS in multiple reaction monitoring (MRM) by electrospray ionisation (ESI) in negative ion mode. Analytes are reported in terms of the free anion. Quantification is performed by isotopic dilution using 18 O 3 labelled chlorate and 18 O 3 labelled perchlorate as internal standards, against solvent-based calibration standards. Was Method Scope/Applicability Updated? No. Has the resubmitted method met the requirements outlined as per the ERP recommendation/ comment(s)? Please provide documentation: See summary below from the Jul 2022 ERP Meeting ERP Requirement Author Response Please Provide a Brief Summary of the Resubmitted Method: Samples are spiked with isotopically labelled internal standards ( 18 O

For infant formula provide additional example chromatograms at the LOQ level for different infant formula products For meat and baby food, try to improve signal to noise at the target LOQ level

See Chromatography file

See manuscript page 17: Signal to noise was improved for the baby food and meat matrices by blowing down the whole 4 mL volume that is eluted from the SPE cartridge. Validation data was obtained for these matrices using the improved procedure. The same procedure was trialled for the rest of the representative matrices with promising looking results. Blowing down the whole volume cuts down on a quantitative transfer, and tube and pipette tip for each sample, saving both time and plastic consumables.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

ERP Requirement

Author Response

Provide RSDs and mean recoveries for each analyte, matrix and concentration level combination (see Excel table) Determine LOQs according to the SMPR validation guidance at the lowest spiking level that meet the recovery and RSD requirements For baby food provide validation data for each representative matrix separately (cereal based, fruit/vegetable based & meat based) Label chromatograms with the actual spiking concentrations Update the reference material codes in the manuscript See summary below from the Mar 2022 ERP Meeting. ERP Requirement LC-MS/MS • Improve method to eliminate the observed interference. LC-MS/MS • Update SANTE guidelines reference & requirements for ion ratios and retention times. LC-MS/MS • Include information about column lifetime (number of sample injections). LC-MS/MS • Ensure there are correct transitions in the manuscript. Matrices for validation • Provide recovery & precision data for all representative matrices at the LOQ level (5 replicates). Matrices for validation • Provide recovery & precision data at 10x LOQ level (5 replicates) for at least one matrix (selected from the minimum representative matrices) category or sub-category. Matrices for validation • This validation data needs to be obtained using the final method conditions.

See manuscript Tables 10-13, page 25-26 *Whole milk powder recovery – one data point at 78% for 0.005 mg/kg perchlorate spike, one at 127% for 0.005 mg/kg chlorate spike See manuscript Tables 10, 12, 14-16, page 25-27

See manuscript Tables 9-13, page 22-26

See Chromatography file

See manuscript Table 17, page 28

Author Response

The chromatography appears to have been improved (based on review of many chromatograms submitted), although I do not see any change to the conditions. Statement in optimization section indicates that gradient elution was not recommended by column manufacturer and therefore not explored. SANTE guidelines updated (p 17 of manuscript). Acceptance limits included for retention times (Table 5) and ion ratios (Table 6). These have not changed from the March draft of the manuscript. Statement included in Selectivity section (p 13 of manuscript – 1000 injections). Described in Table 4 of manuscript. These have not changed from the March draft of the manuscript. The results in Tables 10 and 12 now includes all data points at the LOQ, 5 replicates, for almost all matrices, including at least one matrix per matrix category. The results in Tables 10 and 12 now includes data points at the LOQ and 10x LOQ for almost all matrices, including at least one matrix per matrix category.

I presume that the data presented here was collected using the final method conditions.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

ERP Requirement

Author Response

Specify the type of filter used.

Clarification added that SPE was optimized, and no filtration was necessary. SPE column described in materials section.

Address any other outstanding requirements highlighted by the ERP in December 2021 & March 2022 (see reviewer comments). Provide information about results calculated with calibration that is forced through zero & not forced through zero.

See below

Statement about forcing intercept through zero was removed. A statement was also added at the end of page 17 of manuscript (end of validation section) to indicate that curves were not forced through zero in this study.

See summary below from the Dec 2021 ERP Meeting, updated to include changes since the Mar 2022 Meeting. ERP Requirement Author Response

Request to see the chromatograms at the LOQ levels for each individual transition in each representative matrix, corresponding matrix blanks and all calibration solutions Clarify the calculation of the results and consider simplified formula based on internal standard ratios (an example provided to the method authors) Update the references based on reviewer comments and also ion ratio requirements Please reference Reviewer 2 comments for coverage of matrices. None of the matrix categories are fully covered. Based on the SMPR, in table 1 (footnote) please check requirements for validation. Specifically, infant/adult formula category is important. Whey protein concentrate is not the same as whey protein hydrolysate.

41 pages of chromatograms provided, covering all samples, calibrants, and blanks.

This appears to have been updated as recommended (manuscript page 12).

Updated to include SANTE guidelines as requested.

The table from my Dec 2021 review is included at the end of this form to assist in discussions at Jun 2022 meeting. Notes: • Validation was completed on a single baby food containing beef, rice, and vegetable. Is this adequate, or do they need to be separate samples? How would this sample be characterized (which matrix) if only used for one? • No LOQ can be reported for soy protein isolate and plant based infant formula due to high blank levels. However, other data looks acceptable. • No RSDrs are calculated per matrix – only overall for the study. While we have the data from the individual spikes and their recoveries, but I have not taken the time to try to calculate the individual RSDrs. Authors indicate that all RSDrs are < 20% per the

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

ERP Requirement

Author Response SMPR and SANTE guidelines but I would like to see the values. • Several samples had high recovery outliers and were excluded from calculations. Table 11 has been updated to include individual results for each matrix and concentration level. The results in Table 10 now includes data points at the LOQ and 10x LOQ for almost all matrices, including at least one matrix per matrix category. LOQs have been summarized in Table 9 Table 12 includes the incurred level for chlorate and perchlorate in each validation sample. The chromatography appears to have been improved (based on review of many chromatograms submitted), although I do not see any change to the conditions. Statement in optimization section indicates that gradient elution was not recommended by column manufacturer and therefore not explored. The chromatography appears to have been improved (based on review of many chromatograms submitted), although I do not see any change to the conditions. Statement in optimization section indicates that gradient elution was not recommended by column manufacturer and therefore not explored. Author Response

Table 11 – validation results have to be provided for each representative matrix and concentration level individually Validation should be done at the LOQ and at 10x LOQ as the minimum. Follow the SMPR for LOQ determination. Concentration in blanks used for validation needs to be provided Provide summary tables for the LOQs

ERP Recommendations

Consider improving chromatography to achieve narrower and symmetrical peak shapes (peak symmetry below 1.5)

If isocratic elution is used, consider a gradient elution

Provide General comments: The validation data has been updated and clarified significantly. Thank you! Final Recommendation: I recommend this method for First Action status.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Matrix Table from Dec 2021 Review, Updated Jun 2022 and Nov 2022 Matrix Category Matrix Sub-Category Minimum Representative Matrices

SMPR Met?

Additional Validation Data Needed for

Baby food

Cereal-based Fruit/vegetable-based Meat-based Cereal-based Fruit/vegetable-based Meat-based

YES

Baby food ingredients Infant/adult formulas

YES

Milk-based Plant-based

YES

Animal-based milk powdered protein sources

Whole milk powder Whey protein concentrate

YES 1

Plant-based protein sources

Soy protein isolate

YES YES YES YES

Liquid milk

Bovine Sheep Oil/fat

Infant/adult formula ingredients

Fat-based

Carbohydrate-based

Lactose Maltodextrin Oligosaccharides

Mineral- and vitamin-based (premixes)

Any

YES

1 One data point at 78% for 0.005 mg/kg perchlorate spike, one at 127% for 0.005 mg/kg chlorate spike; mean recoveries were within SMPR

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method Chlorate-01 (Resubmission)

Was Method Scope/Applicability Updated?

Has the resubmitted method met the requirements outlined as per the ERP recommendation/comment(s)?

Yes

Please provide documentation:

Provide General comments: The method makes use of standard instrumentation usually available in labs dealing with chemical contaminants. Validation study covered all matrices listed in the SMPR at the appropriate level. Chromatographic peaks are adequate. 1) Validation data have been sorted according to ERP comments and are much clearer for readers. they would still need some minor changes to make the information available even clearer:

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Some captions of tables do not reflect what the tables contain and are a bit confusing. I suggest considering the following:

Table 9: individual recovery values on samples used during validation.

Table 10 : Mean recoveries, RSD for chlorate at 0.005 mg/kg Table 11: Mean recoveries, RSD for chlorate at 0.05 mg/kg

Table 10 and 11 can be combined

Same for tables 12 and 13 as regards of perchlorate

2) Some clarifications on the approach used for validation would be needed.

In one hand, for matrices where blanks containing more than 30% of the target LOQ, the LOQ was estimated based on extrapolation of linear regression of S/N ratio and concentration to a theoretical S/N 10. According to tables 14 and 15, matrices validate with this approach were: plant based infant formula, whey protein concentrate, and soy protein isolate (with incurred values, according to table 9, at 0.1304 mg/kg chlorate, 0.0124 mg/kg chlorate and 0.2474 mg/kg chlorate & 0.0059 mg/kg perchlorate, respectively). In the case of whey protein concentrate (target LOQ is 0.01 mg/kg), therefore the incurred value is < 30% target LOQ (0.013), and therefore it could have been validated with spiking/recovery experiments. On the other hand, according to table 9 other matrices contained incurred amounts higher than 30 % LOQ, and were validated using spiking/recovery experiments.

- Cereal based baby food: 0.0299 mg/kg chlorate (30 % target LOQ = 0.0015 mg/kg) - Vegetables based baby food: 0.0138 mg/kg chlorate (30 % target LOQ = 0.0015 mg/kg) - Liquid milk: 0.0167 mg/kg chlorate (30 % LOQ = 0.0015 target mg/kg) - Sheep milk: 0.0183 mg/kg perchlorate (30 % LOQ = 0.0015 target mg/kg)

A clarification on the approach used for validation and the criteria for selecting this approach in each case should be provided.

Final Recommendation:

Recommended for first action after discussion and clarification on the LOQ estimation

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method____Chlorate -1_________ (Resubmission)

Please Provide a Brief Summary of the Resubmitted Method: Updated method has more well-defined parameters and relevant data from previous versions.

Was Method Scope/Applicability Updated? Yes

Has the resubmitted method met the requirements outlined as per the ERP recommendation/comment(s)? Yes

Please provide documentation:

Provide General comments: According to the data and c-grams provided, it indicates that 0.0005 mg/kg quantitation of Perchlorate achieved. Can you please provide c-grams for this level of detection? Is it possible to see blank values for both internal standards to prove that 18 O enriched IS does not degrade overtime losing its efficacy? - Reference to USEPA method 332.0 and 331.0 for drinking water.

Final Recommendation:

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

I recommend the method for next steps with these minor proof of concept data

Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15 Page 16

Made with FlippingBook flipbook maker