AOAC SPIFAN Nutrients ERP Reviewer Forms (December 7, 2022)

First page Table of contents 1 Next page Last page

Animated publication

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method AMINO-06

Title: Single Laboratory Validation of determination of total taurine in infant formula and adult/pediatric nutritionals by Ultra High Performance Liquid Chromatography with Multiple Reaction Monitoring Mass Spectrometry . Authors: Geralt ten Kate¥, Martine van Gool¥, Bertus Popma¥, Gerrit van der Pruik¥, Mohèb Elwakiel‡ ¥Royal FrieslandCampina, Stationsplein 4, 3818 LE, Amersfoort, the Netherlands ‡DUCARES B.V., Trading as TRISKELION, Reactorweg 47-A, 3542 AD, Utrecht, The Netherlands

Reviewer Name: Reviewer 1

Summary of Method: Determination of total taurine in infant formula and adult/pediatric nutritionals by Ultra High Performance Liquid Chromatography (UHPLC) with Multiple Reaction Monitoring (MRM) Mass Spectrometry (MS). Powdered samples are reconstituted in water prior to further analysis. Liquid and reconstituted samples are diluted and extracted in Ultra High Quality (UHQ) water including 2-Aminoethane-d4-sulfonic acid as an internal standard (IS) and trichloroacetic acid (TCA). Deproteinized samples, due to the use of TCA, are filtrated. The filtrate is diluted and injected on a hydrophilic interaction liquid chromatography (HILIC) column. Taurine and the IS are separated from the matrix by the means of an aqueous ammonium formate gradient and selectively detected by MRM-MS.

Method Scope/Applicability: The method is applicable to determination of taurine in infant formula and adult/pediatric nutritionals.

General comments about the method: This method uses well-established methodology for analyzing total taurine using Ultra High Performance Liquid Chromatography (UHPLC) with Multiple Reaction Monitoring (MRM) Mass Spectrometry (MS).

Method Clarity: The method is well written and easy to follow.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Pros/Strengths: -The described method makes use of a selective detection method by MS together with the separation properties of a HILIC UHPLC column. -The robustness was tested against three different MS systems, two different LC columns, two different laboratories and three different analysts.

Supporting Data • General Comment:

Method Optimization: A calibration line is established on every analysis. The calibration line is analyzed before the analysis of samples and is based on seven concentrations from 0.03mg/L to 10.0 mg/L.

• Performance Characteristics:

Analytical Range: SMPR: 0.5-2500 mg/100 g Method: 0.5-110 mg/100 g

LOQ: SMPR: ≤0 .5 mg/100 g Method: 0.29 mg/100g

A method blank was spiked with a low concentration of taurine near the expected LOQ. Next to the spiked method blank a method blank was included to calculate the mean blank. A total of 24 measurements of spiked method blanks were performed over six days. The standard deviation of these analyses was calculated. The standard deviation was multiplied by three for the LOD (SLOD) and multiplied by ten for the LOQ (SLOQ). The LOD and LOQ were then determined by adding the SLOD and SLOQ to the blank mean.

Accuracy/Recovery : SMPR: 0.5-5.0 mg/100g 88-112% 5.0-150 mg/100g 90-110% 150-2500 mg/100g 93-107% Method: 92-105%

Reference materials NIST SRM 1869 and 1849a were used to assess the accuracy of the method based on their reference values of taurine. Both reference materials were analyzed on three days in triplicate. The determined accuracy of the method with NIST SRM 1849a is in good agreement with the reference value. The accuracy of the method with NIST SRM 1869 is lower than expected.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Precision (RSD r ): SMPR: 0.5-5.0 mg/100g

≤7%

5.0-150 mg/100g ≤5% 150-2500 mg/100g ≤3%

Method: 0.7 – 4.6 % 0.7 – 1.7%

All samples were analyzed in duplicate over six different days with two different analysts and two different LC-MS systems. The RSD of the repeatability in the analyzed samples was between 0.7-4.6%. The repeatability was also determined from spiked samples to obtain the precision at higher concentrations. Spiked samples were analyzed on three different days in triplicate. The RSD of the repeatability at spiked levels is between 0.7-1.7%. The RSD of the intermediate precision was determined over two different laboratories and was between 1,3 6,7% and in good agreement.

Reproducibility (RSD R ): ----

• System suitability: -The chromatographic system is equilibrated before starting a series of analysis. -Series of analysis are started with a set of six injections with a continuing calibration verification standard (CCV) followed by a system blank and the calibration standards T0-T6. Quality control -After the calibration standards start with a CCV and reinject after every ten analyzed samples. -The recovery of the CCV should be between 95-105%. -The ion ratio between the taurine qualifier ion and the taurine quantifier ion is determined. The ion ratio of a sample is allowed to deviate 15% from the average of the standards.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

1. Is the Validation Study Report in a format acceptable to AOAC? Yes. 2. Is the method described in sufficient detail so that it is relatively easy to understand, including equations and procedures for calculation of results (are all terms explained)? Yes 3. Are the figures and tables sufficiently explanatory without the need to refer to the text? Yes 4. Are all the figures and tables pertinent? Yes 5. Could some be omitted and covered by a simple statement? No 6. Are the references complete and correctly annotated? Yes, the references are completed. 7. Does the method contain adequate safety precaution reference and/or statements? Yes. Recommendation: Although the method complies with the performance parameters of SMPR 2014.013, I would like to mention that the SLV study was carried out with 6 samples provided by NIST and only one sample from SPIFAN. In this case, is it necessary to complete the SLV with samples from the SPIFAN kit?

I propose to move 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 Amino-06

Title: SINGLE LABORATORY VALIDATION OF DETERMINATION OF TOTAL TAURINE IN INFANT FORMULA AND ADULT/PEDIATRIC NUTRITIONALS BY ULTRA HIGH PERFORMANCE LIQUID CHROMATOGRAPHY WITH MULTIPLE REACTION MONITORING MASS SPECTROMETRY Author: Geralt ten Kate, Martine van Gool, Bertus Popma, Gerrit van der Pruik, Mohèb Elwakiel; Royal FrieslandCampina and TRISKELION Reviewer Name: Reviewer 2 Summary of Method: Powder samples are reconstituted with water. SIL-Internal standard is added while diluting the liquid and reconstituted samples with water and trichloroacetic acid. Following filtration and additional dilution the samples are analyzed via HILIC LC-ESI-MS/MS (QQQ) in positive mode using specific MRM transitions for taurine (quant and qual ions) and the SIL-IS (quant ion). Method Scope/Applicability: Applicable to IF and AN. Powder and liquid products with a range of hydrolyzation states (intact, partially-hydrolyzed, extensively-hydrolyzed) and ingredient bases (milk, soy) are included in the validation.

General comments about the method: Method Clarity:

The method is generally well described with clear method steps and sources for supplies and reagents provided.

Minor notes on method clarity:

• Throughout the method the solvent used to fill the volumetric flask is omitted. The assumption is that water is used, but stating it directly would clarify. • In Table 1 the Taurine Stock Concentration of T7 appears to have a typo. Based on the weight and volumes used the concentration should be changed from 5400 to 5000. Please check this value.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

• The method steps create 8 calibration levels (T1-T7 + T0), but Section M states the calibration is based on 6 concentrations. Is this a typographical error, or are only 6 levels used? • Lack of clarity on criterion for the initial 6 replicate CCV injections. Pros/Strengths: • Streamlined method will be easily adopted in labs with LC-MS/MS capabilities. • Anticipate enhanced selectivity, accuracy, and precision from specific transitions with isotopically labelled internal standard. Cons/Weaknesses • Daily preparation of mobile phase, working standard stocks, and calibration standards is cumbersome. • Mobile phase A and calibration stock creation could be more efficient. o The Mobile Phase uses a 2M stock, requiring 31.5 g of ammonium formate, then diluting 1mL into 1000 mL for the final 2mM concentration. As written, daily stock preparation is required. Is daily preparation of the mobile phase A stock required? Can the mobile phase be made without a stock (126 mg into 1 L)? o Are seven separate calibration stocks required? • Use of TCA, which is toxic, corrosive, and an environmental hazard. Supporting Data • General Comment: - Method Optimization: To increase method efficiency, the recommendation is to extend the stability of daily use solutions.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

• Performance Characteristics: Analytical Range:

The SLV report explains that taurine regulations limit the amount of taurine added to infant formula. The precision and spiking data extend beyond the expected taurine fortification level. SLV supports acceptable method range for this analyte.

LOQ:

The LOD and LOQ are presented in detail. LOQ meets SMPR.

Accuracy/Recovery:

Both SRM and spiking were used for accuracy assessment.

• SRM. NIST 1849a and NIST 1869 were analyzed in triplicate on three days. The results for both SRMs demonstrate acceptable method accuracy. • Spiking. Three powder matrices were spiked in triplicate at three levels on three days. The matrices included NIST 8261 (adult nutritional), NIST 10175 (infant formula, partially hydrolyzed soy), and NIST 1869 (infant formula, milk/whey/soy). The spiking levels are nominally 5 mg/100g, 50 mg/100g, and 100 mg/100g. All average recoveries and 79 of 81 individual recoveries were within the recovery SMPR range. The two individual recoveries outside the SMPR range were quite close to the upper limit at 111.8% and 110.8%. Overall, the spiking data indicates the method provides suitable accuracy. In aggregate, the presented accuracy data supports acceptable method performance for this validation parameter. No liquid/RTF samples are included in the accuracy assessment. The method procedure requires reconstitution of powders at 25 g + 200 g water. Including spiking into liquid products would augment the accuracy data. Precisionwas assessed using replicate analysis of innate levels in sevenmatrices for low taurine levels and three spiked matrices to evaluate higher taurine levels. • Innate Precision. Seven matrices were analyzed in duplicate on six days, including two analysts and two LC-MS systems. Samples included are shown below. All matrices met SMPR RSD r targets, with a range of 0.7% - 4.6%. Precision (RSD r ):

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

• Spiked Precision. Nominal spiking levels of 5 mg/100g, 50 mg/100g, and 100 mg/100g were completed in triplicate on three days using the following matrices: NIST 8261 (adult nutritional), NIST 10175 (infant formula, partially hydrolyzed soy), and NIST 1869 (infant formula, milk/whey/soy). This is the same data used for accuracy assessment (point #2 in Accuracy/Recovery section above). All RSD r met SMPR target (≤5%), with the range of 0.7% - 1.7%.

Reproducibility (RSD R ):

RSD R was not assessed, although intermediate precision (RSD iR ) from two different labs was presented for 6 of the 7 matrices with innate repeatability data. Sample SPIFAN DOMY545 (AN, RTF) was omitted from the RSD iR evaluation. The range of RSD iR for the tested samples is 1.3% - 6.7%, which is below the SMPR RSD R target. These data indicate the MLT RSD R will likely be within the SMPR target. Questions for authors: The between day precision for sample SPIFAN DOMY545 (calculated by reviewer) is 8.4 %RSD. This %RSD is higher than the other presented intermediate precision data. The higher %RSD for SPIFAN DOMY545 is driven by the lower results for Day 2, compared with the other precision days. The values for SPIFAN DOMY545 Day 2 replicated within day, which is why the repeatability was not impacted by the lower Day 2 results. What was the suspected cause for the lower Day 2 results for sample SPIFAN DOMY545? Why was sample SPIFAN DOMY545 eliminated from the inter-lab study?

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

• System suitability:

System suitability and QC requirements include retention time and asymmetry, ongoing CCV recovery, and taurine qualifier ion ratio, and R 2 requirements. Further detail on initial CCV, system blank, and method blank criteria is recommended.

1. Is the Validation Study Report in a format acceptable to AOAC?

Yes. 2. Is the method described in sufficient detail so that it is relatively easy to understand, including equations and procedures for calculation of results (are all terms explained)? Yes. 3. Are the figures and tables sufficiently explanatory without the need to refer to the text? Yes. 4. Are all the figures and tables pertinent? Yes. 5. Could some be omitted and covered by a simple statement? No. 6. Are the references complete and correctly annotated? Yes. 7. Does the method contain adequate safety precaution reference and/or statements? The method contains a standard safety precaution statement. Specific safety procedures for TCA are recommended, as this reagent is corrosive and a suspected carcinogen. Have less toxic protein precipitation procedures been evaluated? The SLV demonstrates the method meets the SMPR and is recommended for First Action status pending the clarifying points and questions shown below. Clarifying points and questions: • Adding clarification that water is used to bring flasks to volume. • For system suitability, additional detail for the system blank (how is “not detected” defined) and criteria for the method blank and initial 6 injections of the CCV. • Clarification of number of calibration standards used. The method describes 8 levels, but also states to use 6 levels. Recommendation:

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

• Check T7 Taurine Stock Concentration value. Should this value be changed from 5400 to 5000? • Specific safety procedures for TCA are recommended, as this reagent is corrosive and a suspected carcinogen. • Questions: - What was the suspected cause for the lower Day 2 results for sample SPIFAN DOMY545? - Why was sample SPIFAN DOMY545 eliminated from the inter-lab study? - Have alternative protein precipitation procedures using less hazardous reagents been evaluated?

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method Amino-07

Title: Analysis of Taurine in Infant Formula and Adult Nutritionals by Hydrophilic Interaction Liquid Chromatography-Tandem Mass

Spectrometry (HILIC-MS/MS) Author: Brendon D. Gill, Fonterra Reviewer Name: Reviewer 1

Method Scope/Applicability: The method is applicable to Taurine in Infant Formula and Adult Nutritionals

General comments about the method: The method is scientifically sound and utilizes commonly available instrumentation and materials. The method undergone comprehensive single laboratory validation using 12 samples that included a range of samples from SPIFAN II kit, commercial products and NIST SRMs. The method was shown to meet SMPR.

Method Clarity: Method is clearly written, full information about instrumentation, chemicals and supplies is provided. All steps of the analysis are clearly outlined.

Pros/Strengths: • Fast analysis

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

• Uses common instrumentation and chemicals • Uses isotope labelled internal standard

Cons/Weaknesses • Only applicable to Taurine so a different method needs to be used for other amino acids

Supporting Data • General Comment:

SLV evaluated performance characteristics outlined in SMPR as well as additional parameters such as method ruggedness. All data are clearly presented in the SLV report. Comparison with currently approved AOAC method 997.05 showed good agreement in results.

Method Optimization:

• Performance Characteristics:

Analytical Range: 0.27–2700 mg/100 g (meets SMPR)

LOQ: 0.14 mg/100 g (meets SMPR)

Accuracy/Recovery: 97.2–100.1 % (meets SMPR). Precision (RSD r ): 1.6–6.4 % (meets SMPR)

Reproducibility (RSD R ):

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

• System suitability: Check of repeatability of Taurine retention time as well as peak area ratio between Taurine and IS

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

1. Is the Validation Study Report in a format acceptable to AOAC? Yes 2. Is the method described in sufficient detail so that it is relatively easy to understand, including equations and procedures for calculation of results (are all terms explained)? Yes 3. Are the figures and tables sufficiently explanatory without the need to refer to the text? Yes 4. Are all the figures and tables pertinent? Yes 5. Could some be omitted and covered by a simple statement? No 6. Are the references complete and correctly annotated? Yes 7. Does the method contain adequate safety precaution reference and/or statements? Yes

Recommendation: I recommend this method as AOAC First Action Official Method for Taurine analysis in infant formula and adult nutritionals

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method AMINO-07

Title: Analysis of Taurine in Infant Formula and Adult Nutritionals by Hydrophilic Interaction Liquid Chromatography-Tandem Mass Spectrometry (HILIC-MS/MS) Author: Brendon D. Gill Fonterra Co-operative Group Ltd, PO Box 7, Waitoa 3341, New Zealand

Reviewer Name: Reviewer 2

Summary of Method: Following protein precipitation with Carrez solution, taurine is extracted and separated by hydrophilic interaction liquid chromatography (HILIC) with detection by tandem mass spectrometry (MS/MS) using multiple reaction monitoring (MRM). Stable isotope labelled (SIL) taurine internal standard is used for quantitation to correct for losses in extraction and any variation in ionization. Method Scope/Applicability: Applicable to the determination of taurine in infant formulas and adult/pediatric nutritional formulas. General comments about the method: Method Clarity: The method is well written, clear and easy to follow Pros/Strengths: • The described method requires equipment and reagents available in analytical laboratories. • The robustness of method was demonstrated with a Plackett-Burman trial. Supporting Data • General Comment: In the evaluation study were tested 6 samples of infant formulas from the SPIFAN, 2 NIST certified reference materials NIST 1849a, NIST 1869 and 4 samples purchased from local supermarkets. - Method Optimization : Linearity was evaluated by testing nine calibration standards through the analytical procedure on three different days. The range of taurine concentrations in linearity standards was 0.06–600 µg/mL, which is equivalent to a sample powder concentration of 0.27–2700 mg/hg RTF.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

• Performance Characteristics:

Analytical Range : SMPR: 0.5–2500 mg/hg Method: 0.27–2700 mg LOQ: SMPR: ≤ 0.5 mg /hg Method: 0.14 mg/hg

The LoD was estimated from 10 replicate analyses of an adult nutritional formula with a low concentration of taurine. The LoD was calculated as 0.04 mg/hg RTF and the LoQ is 0.14 m/hg RTF. Accuracy/Recovery: SMPR: 0.5-5.0 mg/hg 88-112%

5.0-150 mg/hg 90-110% 150-2500 mg/hg 93-107%

Method: 0.5-5.0 mg/hg 99.2% 5.0-150 mg/hg 102.3% 150-2500 mg/hg 101.4%

-Two certified reference materials, NIST 1849a, NIST 1869, were analized to evaluate the accuracy. The calculated p-values for both NIST 1949a and NIST 1869 indicate that no bias between the reference values and the measured values by the candidate method could be found. Recovery: Recovery was evaluated using an adult nutritional formula with a low concentration of taurine. Each replicate was spiked at three concentration levels: low (0.7 mg/hg), medium (7 mg/hg), and high (70 mg/hg). Spiked and unspiked samples were analysed in triplicate on three separate days.

Average method recoveries were estimated from 97.2–100.1 %

Precision (RSD r ): SMPR: 0.5-5.0 mg/hg

≤7% ≤5% ≤3%

5.0-150 mg/hg 150-2500 mg/hg

Method: 0.5-5.0 mg/hg 4.5% 5.0-150 mg/hg 1.8% 150-2500 mg/hg 1.8%

Precision as repeatability and intermediate precision was assessed by testing duplicate of 7 samples of different products on six separate days (5 SPIFAN samples, NIST SRM 1859 and NIST SRM 1869). The repeatability for the method ranged between 1.6–6.4 % RSDr.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Repeatability at the concentration ranges specified in SMPR 2014.013 were evaluated by spiking an adult nutritional formula at three concentration levels in the ranges: low (2 mg/hg RTF), medium 20 mg/hg RTF) and high (200 mg/hg RTF). Spiked samples, unspiked samples and spiked standards were analysed by the method as independent replicates eight times. The estimated values for repeatability at 2, 20, and 200 mg/hg RTF level were 4.5, 1.8 and 1.8 % RSDr respectively.

Reproducibility (RSD R ): -----

• System suitability: A system suitability test was performed at the beginning of every analytical run. - Run 3 times taurine calibration standard (CS3). - Repeatability as % RSDr of ratio of peak areas of taurine and 13 C 2 15 N taurine should be less than 2%. - Repeatability as % RSDr of retention time in seconds should be less than 5%. Quality Control (QC) checks were performed as part of routine analytical runs. - A QC sample should be analysed in duplicate with results plotted on a control chart. - A blank sample should be analysed at the beginning of each analytical run. - Check standards should be analysed regularly throughout run. 1. Is the Validation Study Report in a format acceptable to AOAC? Yes. 2. Is the method described in sufficient detail so that it is relatively easy to understand, including equations and procedures for calculation of results (are all terms explained)? Yes. 3. Are the figures and tables sufficiently explanatory without the need to refer to the text? Yes. 4. Are all the figures and tables pertinent? Yes. 5. Could some be omitted and covered by a simple statement? No. 6. Are the references complete and correctly annotated?

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Yes, they are completed. 7. Does the method contain adequate safety precaution reference and/or statements? Yes. The author mentions at the beginning of the method to take into account the safety conditions regarding the reagents used.

Recommendation: I recommend move 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 FLUOR-04 (Resubmission)

Was Method Scope/Applicability Updated? No: All infant/adult nutritional products. Has the resubmitted method met the requirements outlined as per the ERP recommendation/comment(s)? Please provide documentation:

SMPR requirements were met upon previous submission (Sept 2019). However, the results for multiple fluoride methods were inconsistent and the ERP was not able to determine accuracy without a CRM. Authors of all F methods were asked to obtain new kit samples and perform additional analyses to provide direct comparison of methods using common samples. The submission for these 5 new samples also meets SMPR 2014.016 with respect to LOQ, analytical range, recovery, and repeatability. Additionally, the results presented on these 5 new samples are consistent with other methods submitted using these same samples (FLUOR-05 and FLUOR-06) as shown in Table XIII (page 17 of SLV document).

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

With the concerns about accuracy removed (as much as possible without a CRM), the method authors have addressed the major concern of the ERP. Provide General comments:

Final Recommendation: I recommend this method for first action status.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method_Fluor-04 (10/2022) (Resubmission)

Title: Fluoride Determination in Milk, Soy, and Water-Based Products Using Ion-Selective Electrode and Direct Measurement Technique Author: Renée Erney and (Charles) Pete Black Reviewer Name: Reviewer 2 Please Provide a Brief Summary of the Resubmitted Method: Samples and standards are treated with concentrated hydrochloric acid to dissolve and release any bound fluoride from protein and/or minerals in sample. Concentrated sodium citrate buffer is added to complex interfering ions and adjust the pH and ionic strength. Fluoride is determined quantitatively using ion-selective electrodes by comparing its electrical potential to the potential of standards of known fluoride concentration. The potential is proportional to the activity of fluoride ions in the sample. Was Method Scope/Applicability Updated? Five SPIFAN powder samples, including some common samples that have been analyzed by other method developers, were included in the report. The Method scope and applicability meet the SMPR. Has the resubmitted method met the requirements outlined as per the ERP recommendation/comment(s)? Yes Please provide documentation: Some of the new data in the resubmission include Table 1: Repeatability and intermediate precision; Table II: Spike recovery data; Table V: LOQ calculation; and Table VI, results from different electrode. These new data demonstrate that the performance meet the SMPR requirements. Provide General comments: The report demonstrates that this method is a simple, robust, and reliable method for the fluoride in infant formula. Two ion-selective electrodes from two different vendors were tested with comparable results, which demonstrated the robustness of this method. The results for the common samples are in general in line with other methods’ results. There are some bias among the results from different methods. Since no certified value available for these reference materials, the bias is not considered a flaw to this particular method. However, there is one issue needs to be addressed: There is a question about the calculation equation (page 23). The units used in this equation are not in uniform. Specifically, the sample concentration from curve (term C) is a volume-based concentration because the working standard solutions were prepared or diluted by volume. The other terms (Z and X term) are in mass unit. These different units can introduce error that is not negligible. For example, in the sample prep procedure, about 50 mL of 1 M sodium citrate solution was used to make 100 mL of standard solutions

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

or about 100 mL of sample solutions. The density of the 1 M sodium citrate (according to some vendor’s specs) is about 1.1-1.2 g/mL. So, difference (or bias) between the volume-based concentration and the mass-based concentration, if no correction is made, could be as high as 10%. This can introduce a systematic error of up to 10% in the final results. This error would not affect the analytical performance evaluation in the linearity, repeatability/precision, recovery (from spiking), robustness, etc., but could affect the accuracy (compared to a true value). Final Recommendation: After the issue regarding the calculation formula is correctly addressed, it can be recommended to be the First Action method.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method Fluor-06 (Resubmission)

Title: Quantitative analysis of fluoride in infant formula and adult nutrition using headspace gas chromatography with a flame ionization detector Author: Mohèb Elwakiel, Rudie Vos, Emilie Bakker, Kafeelath Raimi and Leo van Stee, Triskelion Reviewer Name: Reviewer 1 Please Provide a Brief Summary of the Resubmitted Method: 0.3 g of IF sample + 2 mL of 3 M sulfuric acid + 10 µL of TMCS is placed into a 20 mL headspace vial. Fluoride is derivatized to TMFS and measured by GC-FID. Was Method Scope/Applicability Updated?

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

Yes.

Please provide documentation:

• Revise method with calibration standard preparation, system suitability, quality control procedures, and safety statement. o Calibration standard preparation procedures added to method (section 4.4). o System suitability added (section 4.8). This section includes duplicate control sample criteria (±15% from nominal value and ±15% between duplicates) and S/N (>10).  Additional recommended criteria include calibration (r 2 and residual of each calibration level) and chromatographic (resolution and asymmetry). o Quality control procedure is included with system suitability (section 4.8). o Safety statement has been added (section 4.3).  TMCS is toxic, corrosive, and highly flammable. At a minimum the safety statement should clearly call out the risks of TMCS. Adding statement to use TMCS in hood (per SDS sections 5.2 and 7.1) is recommended.

• Validation of reconstitution procedure. o Data for slurry included in section 5.4. • Clarify Table 1 for Analytical range and LOQ.

o Table 1 from initial submission was updated in the resubmission with a summary table containing the validation overview. Analytical range and LOQ were described in section 4.9. • Add information on method specificity and the impact of acid treatment on bound fluorine.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

o Specificity addressed in section 5.3. The impact of acid treatment on bound fluoride is discussed in section 4.1. • Describe how innate fluoride concentrations were obtained for the spike experiments. o Innate fluoride concentrations are described in section 5.4. • Explain the calibration curves used for each spiking level. o Details of the calibration curves used for each validation experiment are described in section 5.1. • Add chromatograms of mid-level calibration, matrix without F - spike, and matrix with F - spike. o Chromatograms are shown in section 5.3.

Provide General comments:

All SMPR criteria are met.

Once concern is that compared to alternative candidate methods, this assay utilizes reagents that have a greater health risk (TCMS reaction with water) and more complex analytical instrumentation (HS-GC-FID). These factors will impact the selection of the final fluoride method. In the Principle section, the report describes the derivatizing reagent as trimethylsilylchlorosilane. The more common name for this reagent is trimethylchlorosilane. The recommendation is to utilize the more common name for this reagent moving forward. Final Recommendation: This method meets all the criteria in the SMPR and is recommended for First Action status, with the following considerations:

• Additional system suitability criteria including:

o calibration (r2 and residual of each calibration level) o chromatographic (resolution and asymmetry) • Augmenting the safety statement to clearly describe the risks of TMCS and consider adding

statement to use TMCS in hood (per SDS sections 5.2 and 7.1) • Utilize the term trimethylchlorosilane in future publications

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method: Fluor-06 (Resubmission)

Title:

Quantitative analysis of fluoride in infant formula and adult nutrition using headspace gas chromatography with a flame ionization detector

Author:

Mohèb Elwakiel, Rudie Vos, Emilie Bakker, Kafeelath Raimi and Leo van Stee

Reviewer Name: Reviewer 2

Please Provide a Brief Summary of the Resubmitted Method: A single laboratory validation study was performed for the method “Quantitative analysis of fluoride in infant formula and adult nutrition using headspace gas chromatography with a flame ionization detector” Infant formula samples are dissolved in 3M sulfuric acid followed by derivatization of free fluoride trimethylfluorosilane using trimethylchlorosilane in an GC headspace vial. The sample is then analyzed by HS GC-FID.

Was Method Scope/Applicability Updated? No change in scope

Has the resubmitted method met the requirements outlined as per the ERP recommendation/comment(s)? 1. Revise method with calibration standard preparation, system suitability, quality control procedures, and safety statement Sufficient details to standard calibration, system suitability, and a safety statement. A description of the use of QC sample spiked with fluoride was described for this SLV study. 2. Validation of reconstitution procedure Comparison detailed between direct powder sampling and slurry sampling to SMPR dilution (25/200g) and report no difference in results. 3. Clarify table 1 for analytical range and LOQ Detailed response describing method analytical range and LOQ given. 4. Add information on method specificity and the impact of acid treatment on bound fluorine Description given of role of acid on acid-labile organic compounds (not expected in product) and the use of sulfuric acid to reduce column degradation. 5. Describe how innate fluoride concentrations were obtained for the spiked experiments Samples describing the addition of fluoride to each samples provided. 6. Explain the calibration curves use for each spiking level Calibration curves satisfactorily described. 7. Add chromatograms of mid-level calibration, matrix w/out fluoride spike and matrix with fluoride spike

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Chromatograms provided. 8. Experimental data which focuses on the appropriateness of using TMCS in an aqueous environment. Method authors have addressed this concern satisfactorily. 9. If possible, compare method to the Willard-Winter method No comparison to other methods provided.

Please provide documentation: n/a

Provide General comments: Method authors have provided information requested for by ERP in its review July, 2022

Final Recommendation:

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method HMO -02 (Resubmission)

Title: Determination of Seven Human Milk Oligosaccharides (HMO) in Infant Formula and Adult Nutritionals

Author: T. Bénet, N. Frei, V. Spichtig, D. Cuany, S, Austin Reviewer Name: Reviewer 1

Please Provide a Brief Summary of the Resubmitted Method: The determination of seven human milk oligosaccharides (2’FL, 3FL, 3’SL, 6’SL, DFL, LNT and LNnT) in infant formula and adult nutritionals is described. All oligosaccharides are labeled at their reducing end with 2 aminobenzamide (2AB), separated by means of liquid chromatography and detected by using a fluorescence detector. Consideration of interfering components such as, lactose-3’-phosphate, maltodextrins, galactooligosaccharides (GOS), polydextrose and fructooligosaccharides is taken and solutions for limiting interference are described. The method has been validated in a single laboratory on infant formula (e.g. cow, goat, soy milk based) and adult nutritionals. Recoveries were in the range of 90.9-109% in most cases. Relative repeatabilities (RSD r ) were in the range 0.1-4.2%. Was Method Scope/Applicability Updated? No, the scope and applicability of the method was not changed or updated, as it was not required. The method is applicable on most infant formula and adult nutritionals. However, there are some limitations in regard to 6’SL determination in soy-based formula and LNT/LNnT in elemental formula.

Has the resubmitted method met the requirements outlined as per the ERP recommendation/comment(s)? The authors provided answers and proceeded to corrections in the method as requested/suggested from the ERP and reviewers.

Please provide documentation:

Provide General comments: In general, the method is well written and relatively easy to understand. The authors have updated the method and corrected what was suggested by the ERP and reviewers. Several tables and figures are supporting the text and particularly the validation work. In addition to that, additional supporting material was provided – as it was requested. The necessary method performance requirements as obtained by the single laboratory validation

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

study are provided, discussed and compared with those required by the SMPRs. With some exceptions recoveries, relative repeatability (RSD r ) and reproducibility (RSD R ) are within the ranges defined by the SMPRs.

Final Recommendation: After the reviewing the authors’ comments, the resubmitted document and supplementary material provided, I recommend moving to a first action method.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method___HMO-02__________ (Resubmission)

Title : Determination of Seven Human Milk Oligosaccharides (HMO) in Infant Formula and Adult Nutritionals Author: Thierry Bénet, Nathalie Frei, Véronique Spichtig, Denis Cuany, Sean Austin Reviewer Name: Reviewer 2 Please Provide a Brief Summary of the Resubmitted Method: A method for the quantitative determination of 7 HMOs. Method utilizes 2-AB derivatization chemistry combined with LC-FL separation and detection. Selected interferences may be removed with one or more enzymatic treatments. Was Method Scope/Applicability Updated?

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

Mostly

Please provide documentation:

• Many comments addressed in author provided .pdf (Nestle_HMO-02-responsetoreviews.pdf) • 6’-SL does not meet LOQ in soy matrices • DFL does not meet LOQ in extensively hydrolyzed formula and bovine formula with FOS, but it is still withing proposed analytical range. • Potential high bias for LNnT and LNT in Elemental formulas at low concentrations (may impact effective LOQ in that matrix) • DFL may not meet LOQ in formulas containing polydextrose Provide General comments: I’m still concerned by the R 2 being <0.999 for 2 of the 7 HMOs (6’-SL and LNnT), would a different calibration model give a more reproducible fit? However, even given this disclaimer the observed standard residual error is generally low (generally <5% with two exceptions), therefore it appears the impact is generally small. Final Recommendation: With some clarification on scope limitations (i.e., LOQ in selected matrices as noted above) I recommend this method for first action

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method_ HMO-04

Title: Validation of an LC-MS method for the analysis of milk oligosaccharides Author: Triskelion Reviewer Name: Reviewer 1

Summary of Method: Powdered, or liquid samples are derivatized with 2-AB and separated by reversed phase chromatography. The separated derivatized HMOs are quantified by MS utilizing a custom synthesized 13 C labeled SIL-2AB IS. Method Scope/Applicability: Determination of 7 HMOs from 2.5-500 mg/100g (as-fed basis) in milk and milk powders. General comments about the method: Not clear how purity of HMO standards is addressed. Materials sourced from Carbosynth are anticipated to range from ~90-98% purity. 6’-SL and 3’-SL are supplied in the form of sodium salts, these details need to be included in the method. This becomes critical when assessing samples rather than spikes derived from the same material used to calibrate the instrumentation. The method needs some editing to comply with AOAC format as well as clarification of exact procedures in order to setup/replicate. Editorial suggestions: • Please note, 3’-FL is not equivalent to 3-FL. The ‘ designation indicates which end of lactose the fucose is attached to. The correct designation is 3-FL.

• Units misspelled in the Figure 01 text • Please provide example chromatograms

Method Clarity:

Some additional detail would be helpful to setup/run this method.

Pros/Strengths: • Method meets SMPR criteria for accuracy within the ranges studied.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Cons/Weaknesses • IS not added until the last step of the prep, this is not optimal for method ruggedness. • IS appears to not be commercially available. • Method does not meet SMPR LOQs for 6’-SL, 3’-SL, and LNT • Method not evaluated for top of SMPR range for 3-FL • All analyte RSDr exceed SMPR limits at 2.5 mg/100g and LNT and DFL not met at 500 mg/100g. • DFL standard prep adds error due to assumptions around anticipated amount of material present in vial is exact. • Need to resolve apparent 21% disconnect observed for LNnT in QC E2 sample • 6 of 7 HMOs studied have at least one curve out of six with r 2 <0.995. Is the current calibration model the best fit for the instrument response?

Supporting Data • General Comment: I wish more matrices were evaluated against this method to test the ruggedness of the technique.

Method Optimization:  Evaluate fit of current calibration model  Consider adding IS earlier in prep  Clarify how standard purity is addressed (including standard obtained in salt forms)

• Performance Characteristics:

Analytical Range: 2.5-500 mg/100g (as-fed basis)

LOQ: 2.5 mg/100g

Accuracy/Recovery: 90.1-104%

Precision (RSD r ): 2.5-9.7% RSD

Reproducibility (RSD R ): While results are included it appears that data was only generated in a single lab. Results provided range from 3.3 to 15% RSD

• System suitability:

r 2 >0.990, or r>0.995 Standards may not deviate more than 15% from calibration curve

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

1. Is the Validation Study Report in a format acceptable to AOAC? No 2. Is the method described in sufficient detail so that it is relatively easy to understand, including equations and procedures for calculation of results (are all terms explained)? No 3. Are the figures and tables sufficiently explanatory without the need to refer to the text? Yes 4. Are all the figures and tables pertinent? Yes 5. Could some be omitted and covered by a simple statement? No 6. Are the references complete and correctly annotated? Yes 7. Does the method contain adequate safety precaution reference and/or statements? No Recommendation: I do not recommend this method for first action status as it does not cover the SMPR LOQs for 3 of the 7 HMOs and it does not meet RSD r for all 7 HMOs at the bottom of the validated range (2.5 mg/100g). Additionally, the method ruggedness is unclear as very few matrices have been evaluated. I have additional concerns about the apparent 21% disconnect observed in QC E2 for LNnT. Lastly, it does not appear that one of the reagents used is commercially available which would severely limit the ability of the method to be generally adopted.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method HMO-04

Title: Validation of an LC-MS method for the analysis of milk oligosaccharides Author: Triskeliion Reviewer Name: Reviewer 2

Summary of Method: A UPLC-MS/MS method for quantification of multiple human milk oligosaccharides (HMO) in milk powders is described. Stable isotope labeled 2-aminobenzamine (2-AB) was used for the preparation of HMO internal standards. The HMOs were derivatized by means of 2-AB labeling. The method has been validated in a single laboratory on 3 types of products (two commercially available infant formula and one unknown product). Method Scope/Applicability: Determination of human milk oligosaccharides in milk powders and milk. Seven HMO are quantified. However, it is not clear whether the method was initially designed for 2’FL and LNnT quantification and during the progress of the work additional HMOs were added in the General comments about the method: The UPLC-MS/MS approach for HMO quantification is very interesting. Especially the use of the multiple reaction monitor (MRM) technique that allows quantification of compounds within complex matrices. The highly specific MRM technique has been widely applied in absolute quantification of proteins and metabolites and - although limited – also in HMO quantification (e.g. Xu et al. J. Nutr. 2017 ). One of the advantages is that the applied technique can easily be expanded to other target compounds, if necessary. Despite this, several points should be addressed in the current method. Among other, the method needs editing to comply with AOAC format, as well as additional information regarding sample preparation, analysis steps, data calculations, supporting figures, safety precautions, additional references etc. should be provided.

Method Clarity: The method is not easy to read and to follow the work performed.

Pros/Strengths: • Determination of more than one HMO in one run

• UPLC-MS/MS and particularly the MRM technique is advantageous for complex matrices. • 2-Picoline-borane replaced sodium cyanoborohydride in the reductive amination step. • Easy to extend when additional HMOs are of interest.

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Cons/Weaknesses • 13 C labeled 2-AB Internal standard must be synthesized as it is not commercially available. • Method requires instrumentation that may not be available at laboratories. • Method performance requirements (based on SMPR) were not met for all HMO. To my understanding, the acceptance criteria were mainly based on the 2’FL and LNnT SMPRs.

Supporting Data • General Comment:

Not sufficient supporting data were provided. The method includes several tables where results are presented. However, validation raw-data that could be used to re-calculate and double-check the data were not provided.

Method Optimization:

Not provided.

• Performance Characteristics: In paragraph 2.3 of the method only the draft SMPR for 2’FL and LNnT are mentioned, although for all discussed HMOs are SMPRs published and are available. To my understanding, the method has been developed based on the 2’FL and LNnT. The Method Performance Criteria of the other HMOs should be considered. Analytical Range: No information provided in the text. Should be clearly indicated and discussed based on the SMPRs. Provided was only the calibration range (0.1 – 75 µ g/mL) LOQ: Reported is 2.5 HMO mg/100g milk with set criterion of 4mg/100g milk. The LOQ requirement is not met for 6’SL, 3’SL and DFL. Information about the required LOQ for each HMO is provided in the published SMPRs. Accuracy/Recovery: The reported recovery was between 90.1 – 97.2 %, 96.5-103% and 98.5-104% for the low, middle and high level QC A samples. However, regarding the low levels in HMO, the recovery acceptance criterion (85-110 %) was based on the concentration range of 2.5-20mg/100g. For a number of HMOs the SMPR request a lowest limit that is lower than 2.5g/100g. It is advisable to consult the published SMPR.

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 Page 17 Page 18 Page 19 Page 20 Page 21 Page 22 Page 23 Page 24 Page 25 Page 26 Page 27 Page 28 Page 29 Page 30 Page 31 Page 32 Page 33 Page 34 Page 35 Page 36 Page 37 Page 38 Page 39 Page 40 Page 41 Page 42 Page 43 Page 44 Page 45 Page 46 Page 47 Page 48 Page 49 Page 50 Page 51 Page 52

Made with FlippingBook flipbook maker