Ariz. Admin. Code § 9-17-404.03

Current through Register Vol. 30, No. 50, December 13, 2024

Section R9-17-404.03 - Method Criteria and References for Chemical Analyses

A. In addition to the definitions in A.R.S. § 36-2801 and R9-17-101, the following definitions apply in this Section unless otherwise stated:

1. "Limit of quantitation" means the lowest concentration of an analyte that may be detected and the concentration of the analyte reliably and accurately determined.

2. "Mid-level standard" means a standard that is between the highest concentration and lowest concentration of standards containing the same substances that are used as a reference when testing for the concentration of an analyte.

3. "Response factor" means the ratio between a signal produced by an analyte relative to a signal produced by an internal standard at a specific concentration.

4. "Retention time" means the length of time taken by an analyte to pass through a chromatography column.

5. "Standard" means a sample of known concentration and containing specific substances that is used as a reference when testing for the concentration of an analyte.

B. To perform laboratory testing using chemical analytical methods for any of the analytes in Table 3.1, a laboratory may use:

1. An established national or international chemical method; or

2. A laboratory-developed method that was validated according to:

a. AOAC - Appendix K: Guidelines for Dietary Supplements and Botanicals, 2013, which is incorporated by reference, includes no future editions or amendments, and is available at http://www.eoma.aoac.org/app_k.pdf;

b. USDA - Guidelines for the Validation of Chemical Methods for the FDA FVM Program, 2nd Edition, April 2015, which is incorporated by reference, includes no future editions or amendments, and is available at https://www.fda.gov/media/81810/download; or

c. ICH - Validation of Analytical Procedures: Text and Methodology Q2(R1) 2005, which is incorporated by reference, includes no future editions or amendments, and is available at https://database.ich.org/sites/default/files/Q2_R1__Guide-line.pdf or https://www.fda.gov/regulatory-information/search-fda-guidance-documents/q2-r1-validation-analytical-procedures-text-and-methodology.

C. A technical laboratory director shall ensure that all instruments and equipment used for testing medical marijuana or a marijuana product by chemical analytical methods are:

1. Set up, tuned, and calibrated according to:

a. Manufacturer's acceptance criteria, or

b. Criteria validated according to subsection (B), as applicable;

2. Monitored and maintained according to AOAC - Guidelines for Laboratories Performing Microbiological and Chemical Analyses of Food, Dietary Supplements, and Pharmaceuticals, Appendix A: Equipment, August 2018, which is incorporated by reference, includes no future editions or amendments, and is available at https://www.aoac.org/aoac-accreditation-guidelines-for-laboratories-alacc; and

3. Applicable for the analytes to be tested.

D. A technical laboratory director shall ensure that for an initial demonstration of capability:

1. Before implementing a method or using a new instrument, at least four replicate reference samples including each analyte that are to be tested using the method or the instrument are:

a. Spiked into a clean matrix that is similar to the medical marijuana or marijuana product to be tested with a mid-level standard;

b. Taken through the entire sample preparation and analysis process;

c. Have a relative standard deviation of no more than 20%; and

d. Have an accuracy that meets the acceptance criteria in subsection (K)(2)(d);

2. Whenever a significant change to instrumentation or to a standard operating procedure occurs, the laboratory demonstrates, as specified in subsection (D)(1), that acceptable precision and bias can still be obtained by the changed conditions; and

3. Whenever a new laboratory agent who will be performing testing on medical marijuana or marijuana products is being trained, the laboratory agent demonstrates, as specified in subsection (D)(1), acceptable precision and bias.

E. For potency testing or testing for pesticides, fungicides, growth regulators, mycotoxins, or residual solvents, a technical laboratory director shall ensure that the retention time window for each analyte is established by using the absolute retention time for each analyte and internal standard from the calibration verification standard, prepared according to subsection (H) or (J) as applicable, at the beginning of the analytical sequence.

F. A technical laboratory director shall ensure that:

1. The laboratory complies with the following requirements related to calibration and standards:

a. Except as specified in subsection (F)(1)(c), a minimum of:

i. Five standards are used for an average response factor or for a linear model,

ii. Six standards are used for a quadratic model, and

iii. Seven standards are used for a cubic model;

b. An X-value of zero is not included as a calibration point;

c. A calibration curve for heavy metal testing includes a minimum of three standards and a calibration blank;

d. One standard is less than or equal to the limit of quantitation;

e. The maximum allowable concentration in Table 3.1 for an analyte, with or without dilution, is less than the concentration of the highest calibration standard for the analyte; and

f. As applicable, a standard is created containing a concentration of specific analytes that is a dilution factor from the maximum allowable concentration in Table 3.1 for the analyte and is used when performing multiple runs on a sample, with or without dilution, to cover the range of maximum allowable concentrations in Table 3.1;

2. The acceptance criteria for testing is one of the following, as applicable:

a. The maximum relative standard deviation for the average calibration factor, for an external calibration model, or the response factor, for an internal calibration model, is no more than 20%; and

b. For linear and non-linear calibration models, the coefficient of determination (r²) is greater than or equal to 0.990 with no rounding;

3. For chromatographic testing methods using internal standards for calibration:

a. The relative retention time of each analyte to the internal calibration standard is within 0.06 units;

b. The areas of the peaks for the internal standards in any sample are between 50 and 200% of the area of the peak of a mid-level standard used for calibration; and

c. The internal standards:

i. Have retention times similar to the analytes being tested for,

ii. Do not interfere with any of the analytes, and

iii. Have similar chemical properties as the analytes being tested for;

4. For methods testing for heavy metals using internal standards, the internal standards:

a. Are appropriate for the analyte, and

b. Do not interfere with any of the analytes;

5. When using a selective ion monitoring technique for data gathering, the integration window includes the entire analyte peak; and

6. All standards included for calibration that are below the limit of quantitation have a signal-to-noise ratio of at least 3:1 according to ASTM E685-93, Standard Practice for Testing Fixed-Wavelength Photometric Detectors Used in Liquid Chromatography (2013), which is incorporated by reference, includes no future editions or amendments, and is available at https://web-store.ansi.org/Standards/ASTM/astme685932013.

G. To obtain an acceptable calibration, a technical laboratory director, for each calibration event:

1. May use any of the following options:

a. Perform instrument maintenance to optimize analyte responses, as long as all resulting calibration models meet the acceptance criteria appropriate for the analyte;

b. If the problem appears to be associated with a single standard:

i. Reanalyze that one standard, at the time of calibration and before any samples are analyzed, to rule out problems due to random error: and

ii. Recalculate and reevaluate the standard against the acceptance criteria;

c. Narrow the calibration range by replacing one or more of the calibration standards at the upper or lower ends of the curve;

d. Narrow the calibration range by removing data points from either extreme end of the range and recalculating the calibration function; or

e. Perform a new initial calibration according to subsection (F); and

2. May not:

a. Remove data points from within a calibration range while still retaining the extreme ends of the calibration range,

b. Use non-linear calibrations to compensate for detector saturation or to avoid proper instrument maintenance;

c. Use multiple points at the same calibration level if not also being done for all quality control samples, such as a sample required in subsection (K), and samples accepted for testing; or

d. Include calibration data from another calibration that was run at a different time.

H. A technical laboratory director shall ensure that, during each calibration event for initial calibration verification:

1. Standards are prepared either from a different source or from a different lot of standards from the same source than the source from which the initial calibration standards specified in subsection (F)(1) were obtained and must:

a. Be

a mid-level standard; and

b. Contain all analytes being reported to comply with R9-17-317(A)(5); and

2. The following acceptance criteria are used:

a. For potency testing, 80 to 120% recovery of true value;

b. For testing for pesticides, fungicides, growth regulators, mycotoxins, or residual solvents other than butanes, 70 to 130% recovery of the true value;

c. For butanes, 60 to 140% recovery of the true value; and

d. For heavy metal testing, 90 to 110% recovery of the true value.

I. A technical laboratory director shall ensure that for the limit of quantitation:

1. The limit of quantitation is initially verified by the analysis of at least seven replicate samples, spiked with all analytes at the limit of quantitation, and processed through all preparation and analysis steps for each method;

2. The signal-to-noise ratio of the replicate samples in subsection (I)(1) is at least 5:1 according to ASTM E685-93, Standard Practice for Testing Fixed-Wavelength Photometric Detectors Used in Liquid Chromatography (2013), which is incorporated by reference, includes no future editions or amendments, and is available at https://webstore.ansi.org/Standards/ASTM/astme685932013;

3. The mean recovery of the replicate samples in subsection (I)(1) is:

a. For potency testing, ± 20% of the true value;

b. For testing for pesticides, fungicides, growth regulators, mycotoxins, or residual solvents, ± 50% of the true value; and

c. For heavy metal testing, ± 35% of the true value;

4. The relative standard deviation of the replicate samples in subsection (I)(1) is less than 20%;

5. The limit of quantitation is, as applicable, no greater than:

a. Half the maximum allowable concentrations for an analyte in Table 3.1;

b. For chlorfenapyr, cyfluthrin, or cypermethrin, the maximum allowable concentrations for the analyte in Table 3.1; or

c. 1.0 mg/g for each analyte for potency testing;

6. Any changes to specific sample amounts, dilutions, or volumes employed are reflected in the limit of quantitation stated on a sample report;

7. The signal-to-noise ratio in subsection (I)(2) is reverified each time the instrument used for testing is calibrated; and

8. Documentation of the current limit of quantitation is maintained for each analyte, matrix, and instrument.

J. Except as provided in subsection (P), a technical laboratory director shall ensure that for batch analysis:

1. Continuing calibration verification standards:

a. Are prepared and spiked with a mid-level concentration of all analytes from the same calibration standard source used to prepare the standards specified in subsection (F)(1); and

b. Have the following acceptance criteria:

i. For potency testing, 80 - 120% recovery of true value;

ii. For testing for pesticides, fungicides, growth regulators, or mycotoxins, or residual solvents other than butanes, 70 - 130% recovery of the true value;

iii. For butanes, 60 - 140% recovery of the true value; and

iv. For heavy metal testing, 90 - 110% recovery of the true value;

2. If internal standards are used in continuing calibration verification, the acceptability criteria of the internal standards is determined as follows:

a. For testing for pesticides, fungicides, growth regulators, mycotoxins, or residual solvents by mass spectrometry, if the area of the peak for an internal standard is different by a factor of two from the area of the respective standard in subsection (F)(1)(e), for the most recent initial calibration sequence, according to subsection (F):

i. The mass spectrometer is inspected for malfunctions and corrected, and

ii. Reanalysis of the continuing calibration verification meets acceptance criteria in subsection (J)(1)(b)(ii) before any samples are tested; and

b. For heavy metal testing:

i. The intensity of an internal standard is monitored for each analysis to ensure that the intensity does not vary by more than ± 30%, with respect to the intensity during the initial calibration in subsection (F); and

ii. If the intensity of an internal standard is outside the range also observed in the calibration blank required in subsection (F)(1)(c):

(1) Testing is stopped until the problem is corrected, the instrument is recalibrated, and the new calibration is verified;

(2) Reanalysis of the continuing calibration verification meets acceptance criteria in subsection (J)(1)(b)(iii) before any samples are tested; and

(3) The affected samples are retested; and

3. The frequency of continuing calibration verification is as follows:

a. For testing by a method other than mass spectrometry:

i. At the beginning of the test;

ii. After every 20 samples, not counting a quality control sample, such as a sample required in subsection (K); and

iii. At the end of the test; and

b. For testing by mass spectrometry:

i. At the beginning of the testing,

ii. After every 12 hours of running, and

iii. At the end of the run.

K. Except as provided in subsection (P), a technical laboratory director shall ensure that for batch analysis, which may contain no more than 20 samples accepted for testing:

1. A method blank, with a matrix similar to each type of sample matrix to be tested within the batch:

a. Contains the same internal standards as the samples in the batch,

b. Is prepared and tested with each batch, and

c. Produces results below the limit of quantitation;

2. Except as provided in subsection (R), a laboratory control sample and duplicate, with a matrix similar to each type of sample matrix to be tested within the batch:

a. Are prepared

with a mid-level standard;

b. Are spiked before extraction;

c. Are carried through all stages of sample preparation and included with each analytical batch; and

d. Have either the following acceptance criteria:

i. For potency testing, 80 - 120% recovery of true value;

ii. For pesticides, fungicides, growth regulators, mycotoxins, or residual solvents other than butanes, 70 - 130% recovery of the true value or according to control limits derived according to R9-17-404.05(B)(10);

iii. For butanes, 60 - 140% recovery of the true value or acceptance criteria within statistically derived limits developed by the laboratory; and

iv. For heavy metal testing, 80 - 120% recovery of the true value or acceptance criteria within statistically derived limits developed by the laboratory;

3. The relative percent difference for the laboratory control sample and duplicate for each analyte, calculated on the basis of concentration or amount, is no more than 20%; and

4. For all new matrix types to be tested, a matrix spike derived from a dispensary submitted sample:

a. Is prepared for each analyte in Table 3.1 with a mid-level standard;

b. Is carried through all stages of sample preparation and included with each analytical batch of up to 20 samples for each matrix type; and

c. Has either the following acceptance criteria or acceptance criteria within statistically derived limits developed by the laboratory:

i. For potency testing, 80 - 120% recovery of true value or according to control limits derived according to R9-17-404.05(B)(10);

ii. For testing for pesticides, fungicides, growth regulators, mycotoxins, or residual solvents, 70 - 130% recovery of the true value or according to control limits derived according to R9-17-404.05(B)(10); and

iii. For heavy metal testing, 75 - 125% recovery of the true value.

L. A technical laboratory director shall ensure that:

1. Except as provided in subsection (P), for potency testing or testing for pesticides, fungicides, growth regulators, mycotoxins, or residual solvents by mass spectrometry, the relative intensities of the characteristic ions agrees within 30% of the relative intensities of these ions in the reference spectrum; and

2. For heavy metal testing, the intensity of each internal standard is monitored for each analysis to ensure that the intensity does not vary more than ±30%, with respect to the intensity of the internal standard during the initial calibration specified in subsection (F).

M. A technical laboratory director shall ensure that:

1. In testing, by a method other than mass spectrometry, the resolution of chromatographic peaks is maintained so that the height of the valley between two chromatographic peaks is less than 50% of the lower peak height; and

2. For testing by mass spectrometry methods, the resolution of chromatographic peaks is maintained so that the height of the valley between two chromatographic peaks is less than 50% of the average of the two peak heights.

N. A technical laboratory director shall ensure that confirmation for testing for pesticides, fungicides, growth regulators, or residual solvents by a method other than mass spectrometry:

1. Is performed using:

a. A second column:

i. That has a stationary phase dissimilar to the stationary phase in the primary column, and

ii. From which the analyte is eluted in a different order than from the primary column;

b. A different instrument type, such as gas chromatography followed by mass spectrometry;

c. Gas chromatography with two different types of detectors; or

d. Other recognized confirmation techniques;

2. Meets the applicable criteria in subsections (D) through (M); and

3. Includes as part of the confirmation of the analyte:

a. An evaluation of the agreement of the quantitative values of the results from both methods of testing; and

b. Determination of the relative percent difference between the values.

O. If the relative percent difference between the values obtained according to subsection (N) is more than 40%, a technical laboratory director shall ensure that:

1. The chromatograms are checked to see if an obviously overlapping peak is causing an erroneously high result, and the chromato-graphic conditions are reviewed; and

2. Either:

a. If a problem is found with one of the tests, the result from the other test is reported; and

b. If there is no evidence of a chromatographic problem, the higher result is reported.

P. A technical laboratory director may release testing results that are scientifically valid and defensible, according to R9-17-404.06(B)(3), with the following data qualifier notations if:

1. The target analyte detected in the calibration blank required in subsection (F)(1)(c) or the method blank specified in subsection (K)(1) is at or above the limit of quantitation, but the sample result:

a. For potency testing, is below the limit of quantitation - B1; or

b. When testing for pesticides, fungicides, growth regulators, mycotoxins, heavy metals, or residual solvents, is below the maximum allowable concentration in Table 3.1 for the analyte - B2;

2. The limit of quantitation and the sample results were adjusted to reflect sample dilution - D1;

3. The relative intensity of a characteristic ion in a sample analyte exceeded the acceptance criteria in subsection (L)(1) with respect to the reference spectra, indicating interference - I1;

4. When testing for pesticides, fungicides, growth regulators, mycotoxins, heavy metals, or residual solvents, the percent recovery of a laboratory control sample is greater than the acceptance limits in subsection (K)(2)(c), but the sample's target analytes were not detected above the maximum allowable concentrations in Table 3.1 for the analytes in the sample - L1;

5. The recovery from the matrix spike in subsection (K)(4) was:

a. High, but the recovery from the laboratory control sample in subsection (K)(2) was within acceptance criteria - M1,

b. Low, but the recovery from the laboratory control sample in subsection (K)(2) was within acceptance criteria - M2, or

c. Unusable because the analyte concentration was disproportionate to the spike level, but the recovery from the laboratory control sample in subsection (K)(2) was within acceptance criteria - M3;

6. The analysis of a spiked sample required a dilution such that the spike recovery calculation does not provide useful information, but the recovery from the associated laboratory control sample in subsection (K)(2) was within acceptance criteria - M4;

7. The analyte concentration was determined by the method of standard addition, in which the standard is added directly to the ali-quots of the analyzed sample - M5;

8. A description of the variance is described in the final report of testing according to R9-17-404.06(B)(3)(d)(ii) - N1;

9. The relative percent difference for the laboratory control sample and duplicate exceeded the limit in subsection (K)(3), but the recovery in subsection (K)(2) was within acceptance criteria - R1;

10. The relative percent difference for a sample and duplicate exceeded the limit in subsection (O) - R2; or

11. The recovery from initial or continuing calibration verification standards is greater than the acceptance limits in subsection (H)(2) or (J)(1)(b) as applicable, but the sample's target analytes were not detected above the maximum allowable concentrations in Table 3.1 for the analytes in the sample - V1.

Q. A technical laboratory director shall include in the final report of testing, according to R9-17-404.06(B)(3)(d)(iii), the following data qualifier notations if:

1. Sample integrity was not maintained - Q1;

2. The sample is heterogeneous, and sample homogeneity could not be readily achieved using routine laboratory practices - Q2; or

3. Testing result is for informational purposes only and cannot be used to satisfy dispensary testing requirements in R9-17-317.01(A) or labeling requirements in R9-17-317 - Q3.

R. For batch analysis of samples to determine potency, a technical laboratory director may check precision by using either a duplicate laboratory control sample or a duplicate sample prepared from the medical marijuana or marijuana product being tested, according to requirements in subsections (K)(2) and (3).

S. A technical laboratory director shall ensure that the reporting units for:

1. Pesticides, fungicides, growth regulators, heavy metals, or residual solvents are in parts per million (ppm);

2. Mycotoxins are according to R9-17-404.04(I)(4); and

3. Potency are:

a. In either:

i. Percent (w/w) relative to the bulk plant material or marijuana product, as applicable; or

ii. Number of milligrams per designated unit; and

b. For:

i. Total tetrahydrocannabinol, the sum of tetrahydrocannabinolic acid (THC-A), multiplied by 0.877, and delta-9-tetra-hydrocannabinol ([DELTA]9-THC); and

ii. Total cannabidiol, the sum of cannabidiolic acid (CBD-A), multiplied by 0.877, and cannabidiol (CBD).

Ariz. Admin. Code § R9-17-404.03

Adopted by final exempt rulemaking at 26 A.A.R. 734, effective 4/2/2020. Amended by exempt rulemaking at 26 A.A.R. 2848, effective 10/15/2020. Amended by exempt rulemaking at 26 A.A.R. 2991, effective 11/1/2020. Amended by final exempt rulemaking at 27 A.A.R. 111, effective 1/15/2021. Amended by final rulemaking at 29 A.A.R. 2396, effective 10/1/2023.