Quality Standards in State Programs Permitting Cannabis for Medical Uses

Abstract

Currently in the United States, there exists a patchwork of state-level laws and regulations surrounding cannabis use. Although cannabis (excluding hemp under the Agricultural Improvement Act of 2018, Public Law 115-334) is illegal at the federal level and is not FDA (U.S. Food and Drug Administration) approved for any indication, many states allow patients with qualifying conditions to register for their medical cannabis program (MCP). To better understand the quality of cannabis found in these programs, we collected laws, regulations, and guidance documents available on public state-run websites and compared them with current good manufacturing practices (CGMPs) applicable to finished drug products. CGMPs for human drugs contain minimum requirements for the methods, facilities, and controls used in manufacturing, processing, and packaging of a drug product to assure it is safe for use. Such a comparison will aid the development of consistent quality standards that could, in turn, improve the quality of a wide range of cannabis medical products in development that may be sold in the United States. States may likewise choose to have the cannabis and cannabis-derived products that fall within their MCP to follow quality-focused guidelines, such as those listed in CGMPs, to ensure the quality of these products and promote public health. This study further solidifies the need for standardized testing protocols and methodologies to keep consumers safe.

Introduction

Cannabis sativa L. is a plant that has been used by humans for fiber, food, and medicinal purposes for centuries, with the oldest known written record coming from the Chinese Emperor Shen Nung in 2727 B.C.¹ Presently, cannabis cultivated for medical and recreational use in the United States is illegal under federal law, yet numerous states have legalized and instituted their own medical and recreational programs. On the medical side, 34 states and the District of Columbia have fully operational medical cannabis programs (MCPs) as of June 30, 2021 (Supplementary Materials). It is estimated that > 5.3 million patients are registered in these state-level programs.²

There continues to be an increased public interest in studying cannabis for a variety of conditions such as seizure disorders, anxiety, and pain.³ The U.S. Food and Drug Administration (FDA) regulates a wide variety of medical products, including drugs, foods, and dietary supplements. For drug development, several submission types⁴ exist in the Center for Drug Evaluation and Research (CDER)⁵ for researchers who want to study cannabis in clinical trials for various conditions. When cannabis and cannabis-derived products (CCDPs) intended for use under human clinical trials are submitted, CDER treats these products like any other FDA-regulated product (i.e., subject to same authorities and requirements as FDA-regulated products containing any other drug substance).^6-8

Botanical products, such as cannabis, are heterogeneous in nature. Because of the uncertainty in their defined active constituents, control of botanical raw materials (BRMs)—including their authenticity and agriculture practices (e.g., growing, storage conditions, and processing methods)—is of particular importance in the manufacture of a botanical drug substance (BDS) for FDA approval. The main regulatory standard for ensuring pharmaceutical quality is the current good manufacturing practice (CGMP) regulation for human pharmaceuticals,⁹ and the manufacturing of a BDS should follow CGMPs^10,11

In part because of the challenging nature of these products, FDA has published guidance documents for cannabis products that discuss CGMPs. In 2020, FDA issued “Cannabis and Cannabis-Derived Compounds: Quality Considerations for Clinical Research,” guidance for industry.¹² This draft guidance, when finalized, will reflect FDA’s current thinking on topics relevant to clinical research on the development of human drugs containing CCDPs with a focus on manufacturing high-quality drug products. In addition, FDA’s Botanical Drug Development Guidance for Industry¹⁰ provides FDA’s current thinking regarding growing conditions and manufacturing quality controls for BRMs. Examples of BRM products include the whole hemp plant, seeds, and flowering parts of the cannabis plant.

Separately, states have also been creating standards for use when regulating cannabis products under state authorities. Although sharing a common goal of quality and consistency in products for consumers, there are differences between these standards and with the standards set out in the FDA guidance documents and CGMPs for drugs.

This article reports the results of our review of 37 available MCP regulations (Supplementary Materials) from public state-run websites. This article is not intended to be a comprehensive review of state quality requirements. The goal of this study was to (i) understand what quality-related tests are written into state regulations for CCDPs, (ii) compare and discuss the written testing regulations to those outlined in the drug CGMPs, and (iii) provide considerations to facilitate the development of consistent high-quality products for potential use in clinical trials of drugs.

Methods

On June 30, 2021, state-run websites (departments of health, cannabis-regulating offices, etc.) were visited to collect MCP laws, rules, regulations, and/or guidance documents related to laboratory testing of CCDPs (Supplementary Table S1). For each program, the most recent version of their laws, rules, regulations, and/or guidance documents found available on the website was downloaded and saved. Results were found for 37 state-level MCPs, which were included in the final analysis.

To qualify, information regarding laboratory testing simply had to be present in the online document. There is one instance of a state’s pending legislation being included. Since this review is not intended to be a legal analysis, and after further investigation of the state’s program, this pending legislation was included here to further reflect the state-level thinking on the subject. This state’s program was not included in any further analysis, such as for pesticide (Fig. 2), microbial (Fig. 3), and heavy metal (Figs. 2 and 4) testing.

For each of the 37 programs, the identified tests included in their laws, rules, regulations, and/or guidance documents were summarized. From this summary, 10 different tests were identified across 37 programs: heavy metal, pesticide, microbial, mycotoxin, residual solvent, foreign matter, water activity, moisture content, cannabinoid, and terpene (Supplementary Materials, Definitions). Further analysis of the 10 tests was completed by collecting lists of analytes to be tested for, as well as action levels for each. The information was then compared with CGMPs for human drug products. Microsoft Excel was used for all data recording and analysis.

Overview of Quality Testing Found in State-Level MCPs

First, we compiled information on the types of quality-related testing found across state MCPs (n = 37). FDA has seen an intensified interest in clinical research using CCDPs sold within state MCPs. Therefore, it has become increasingly important to understand quality control, manufacturing, and testing regulations at the state level. We found most states have some combination of the following 10 tests written into their programs: heavy metal, pesticide, microbial, mycotoxin, residual solvent, foreign matter, water activity, moisture content, cannabinoid, and terpene. We commonly saw lists of contaminants and action levels (in quantifiable units) for each test written directly into program regulations.

Heavy metal testing is present in 33 programs (Fig. 1) and includes various metals, with the four most common being arsenic (As), cadmium (Cd), lead (Pb), and mercury (Hg). Most states list these four metals only, but we found a range from no to eight metals (Fig. 2). Pesticide testing is included in 36 programs (Fig. 1). State-level pesticide testing includes scanning for numerous insecticides, herbicides, plant growth regulators, etc. Some states provide their own list and action limits, ranging from 9 to upward of 60 pesticides (Fig. 2).

FIG. 1.

Laboratory testing of state-level CCDPs. The 10 tests listed in the figure have been identified from 37 state-level MCPs. These tests are important for determining the identity, strength, quality, and purity of CCDPs. Orange represents tests that are specific to CCDPs. The values listed on the right represent the number of state programs that had the test present in their regulations. No test was present in all 37 programs. One program did not include quality tests in their regulations. CCDPs, cannabis and cannabis-derived products; MCP, medical cannabis program. [Background image: Greola84/iStock/Getty Images Plus via Getty Images.]

FIG. 2.

Differences in pesticide and heavy metal testing between state MCPs. Pesticide (blue) and heavy metal (gray) testing from 37 state MCPs were further analyzed and 8 are represented in the figure. The values represent the number of pesticides or heavy metals that each program lists in their regulations. The programs vary greatly in the number of pesticides required for testing. The four metals commonly listed across the programs are As, Cd, Pb, and Hg, however, some list more (up to eight) or less (as low as zero) to be tested. As, arsenic; Cd, cadmium; Hg, mercury; Pb, lead.

A few states cite the U.S. Environmental Protection Agency’s standards set forth in The Code of Federal Regulations (40 CFR 180).¹³ Microbial testing is also found in 36 programs. Mycotoxin testing specifically is present in 28 programs. The last test found in most programs (n = 33) is for residual solvents (Fig. 1). The tests written into the fewest MCP regulations are foreign matter (n = 20), moisture content (n = 21), and water activity (n = 18) (Fig. 1).

The final two tests we examined, cannabinoid and terpene, are important for determining not only the quality, but also the identity of CCDPs. There are > 100 cannabinoids produced by the cannabis plant.¹⁴ In total, 36 out of 37 programs included cannabinoid testing in their regulations, with the most common cannabinoids listed for testing being tetrahydrocannabinol (THC), tetrahydrocannabinolic acid (THCA), cannabidiol (CBD), and cannabidiolic acid (CBDA). Cannabis also produces aromatic oils called terpenes, with > 100 identified naturally in the plant.¹⁴ Terpenes are important to identifying cannabis cultivars,¹⁵ as they contribute to the specific odor and flavor.

Only 14 programs list terpene testing in their regulations. Two programs provide a list of specific terpenes that products should be tested for, and three reference the cannabis inflorescence monograph published by the American Herbal Pharmacopeia (AHP).¹⁶ Terpenes and cannabinoids together create a distinct profile for the plant and identifying them is important for determining the quality of CCDPs.

Comparing State-Level Quality Testing to CGMP Requirements

To better understand the quality of CCDPs within state MCPs, we analyzed how closely the written regulations follow CGMPs. FDA’s portion of the Code of Federal Regulations is in Title 21, which interprets the Federal Food, Drug, and Cosmetic Act and related statutes. Parts 210¹⁷ and 211¹⁸ contain CGMP regulations that must be followed by drug manufacturers. We looked specifically at three considerations within Part 211 Subpart I, which focuses on laboratory controls. These are identity and strength of active ingredients, appropriate laboratory testing for microorganisms, and acceptance criteria for quality control sampling and testing.

As a general requirement, “laboratory controls shall include the establishment of scientifically sound and appropriate specifications, standards, sampling plans, and test procedures designed to assure that components, drug product containers, closures, in-process materials, labeling, and drug products conform to appropriate standards of identity, strength, quality, and purity.” We found some of the written state regulations would not be sufficient to support a submission to FDA for human clinical trials. Although our understanding of these regulations was based on their plain language, we understand that the requirements and implementation may vary from state to state.

Consideration 1: identity and strength of active ingredients

During review, FDA considers the final specifications for a drug product before release, which includes the identity and strength of each active ingredient.¹⁸ There are > 500 constituents naturally occurring in the cannabis plant,¹⁴ however, they may not occur simultaneously in all plant parts. This makes it very complicated for researchers to determine the active components in proposed products. FDA has stated in draft guidance that quantitative data regarding phyto-chemicals, including but not limited to cannabinoids and terpenes, are important to ensure the identity and strength of the final product.¹²

As mentioned previously, state MCPs generally listed testing for THC, THCA, CBD, and CBDA in their regulations. A few programs included additional cannabinoids such as cannabinol and cannabigerol. Currently within the United States, there is growing public interest in other cannabinoids, such as cannabidivarin, tetrahy-drocannabivarin, and cannabichromene.

Adequate characterization of CCDPs (e.g., chemical fingerprint) is critical to ensuring batch-to-batch consistency,¹² and quantifying these cannabinoids is useful so CCDPs can be further studied for possible effects in humans under clinical trials. In addition, including more cannabinoids in testing will provide more robust chemical identity profiles leading MCPs to have greater control over formulation of CCDPs. This approach will likely help increase product consistency and reduce product strength variations.¹⁹

Although there are many naturally occurring terpenes in cannabis, only 14 MCPs include terpene testing in their regulations. Cannabinoid and terpene content vary greatly between different cannabis cultivars and should be considered part of a product’s identity and quality specifications. Some terpenes are hypothesized to have potential adverse effects (e.g., beta-myrcene may be a possible carcinogen when administered orally).²⁰ One state even includes beta-myrcene on a list of chemicals known to the state to cause cancer or reproductive toxicity.²¹

Beta-myrcene can be the most abundant terpene in cannabis,²² and some individuals operating in the cannabis space hypothesize that the amount in a product will determine whether it exhibits an “energizing” or a “relaxing” effect.²³ Beta-myrcene is one example of the importance of terpene testing when determining the identity, and potentially the strength of CCDPs. As methodologies for identifying and quantifying terpenes continue to advance and become available,^24,25 testing for these compounds is becoming more accessible.

Consideration 2: laboratory testing for microorganisms

Drug CGMPs require establishing test procedures to assure the quality of drug products. One area high-lighted in CGMPs is microbial testing.¹⁸ Microbial testing is especially important for botanical products. As mentioned, 36 MCPs we reviewed included some form of microbial testing in their regulations. What differed between programs were the types of tests listed, the action levels, and whether these varied among different types of products (e.g., inhalation vs. oral products) (Fig. 3). The two most common microbial tests we found written into state MCP regulations are for Escherichia coli and Salmonella.

FIG. 3.

Comparison of microbial testing across state MCPs reveals significant differences. The figure shows four different MCPs. Each program categorizes its products for microbial testing differently, with some being more detailed than others. In addition to categorizing products differently, MCPs also include a wide range of tests across product categories. In the four programs pictured, three to eight total tests were found across product categories listed for microbial testing. We also found inconsistencies in the action levels for each test for similar product categories. To demonstrate the complexity, the bold product categories in each box below are where a “cannabis brownie” would be categorized within each program. You can see how differently a brownie is characterized between programs. The table in the upper right shows the specific tests and action levels listed for a brownie undergoing microbial testing within each program. This highlights the vast differences in microbial testing for the same product across state-level MCPs. ^aDried usable cannabis not extracted; ^bextracted or processed cannabis product, that is, hash, bubble hash, rosin, kief; ^csolutions of cannabis in alcohol; ^dsolutions of cannabis in water; ^esolutions of cannabis derived by boiling in water for at least 15 min; ^fan alcoholic liquid extract made by percolation of cannabis so that 1 mL of the fluid extract represents 1 g of the cannabis; ^gdried cannabis to which boiling water is added immediately before consumption; ^hproducts infused with cannabis and other ingredients that are intended for use or consumption other than by smoking, including but not limited to edibles, ointments, and tinctures; ⁱproduced by extracting cannabinoids from cannabis. Includes concentrate consumed using a vaporizer delivery device or pressurized metered dose inhaler; ^jinhalable concentrate, which may comprise other ingredients inside a device that uses a heating element to create a vapor including, but not limited to, vaporizer cartridges and vaporizer pens; ^kremediation is the process of neutralization or removal of dangerous substances or other contaminants from regulated cannabis while changing the product type of the regulated cannabis; ^lincludes edible (chocolate, hard candy, brownie, beverages, etc.) and nonedible (lotions, balms, etc.) products; ^mextraction using ice water, rosin press, or dry ice (also includes kief); *these products are subject to additional testing because they may raise public health concerns, requiring additional safeguards and oversight.

Outside of these, regulations listed a combination of tests, including Aspergillus, total yeast and mold, total aerobic microbial count, total coliforms, and a handful of others. Two programs did not provide their own list and action levels and instead cited United States Pharmacopeia (USP) < 1111 > .²⁶ One program cited both USP < 1111 > and USP < 2023 > ²⁷ in their regulations. To better align MCP regulations with CGMP, it would be beneficial to, at minimum, follow the recommended tests and criteria listed in USP < 61 > ,²⁸ < 62 > ,²⁹ and USP < 1111 >.

Consideration 3: acceptance criteria for quality control sampling and testing

Determination of adequate data-based acceptance criteria (i.e., action levels) is important to assure the quality of drug products.¹⁸ Having specific action levels for each quality test helps guarantee that batches of drug products are safe for use. Our analysis focused on heavy metal testing on CCDPs in state MCPs. Heavy metal testing was written into most (n = 33) MCPs. The regulations listed between four and eight metals, with the most common being As, Cd, Pb, and Hg. This is likely because these four metals are listed in the AHP Cannabis Inflorescence Quality Control Monograph¹⁶ as well as USP < 561 > articles of botanical origin.³⁰

We looked specifically at how action levels for these four metals compared with those listed in USP < 232 > .³¹ USP < 232 > specifies limits for the amounts of elemental impurities in drug products, drug substances, and excipients. This chapter suggests risk-based approach to limit 15 metals whose action levels depend on route of administration (ROA): oral, inhalation, or parenteral. We found only 12 state programs (36%) having specified action levels for inhalation products versus oral products, even though nearly all programs (n = 33) allow both ROAs. Only three of these programs (25%) matched USP < 232 > for both ROAs.

In total, 7 of the 12 programs (58%) matched USP < 232 > guidelines for inhalation products and 6 matched for oral products. Four programs (33%) matched USP < 232 > guidelines for inhalation products and USP < 2232 > ,³² which specifies criteria for elemental impurities in dietary supplements, for oral products. The remaining 21 (64%) programs only provided a single set of criteria for all ROAs or did not provide a list of specific heavy metals. Overall, testing for these four heavy metals in inhalation and oral products was lacking in the majority of MCPs (64%), inconsistent across programs when present, and did not always align with USP recommendations (Fig. 4).

FIG. 4.

Sample of MCP action levels for heavy metal testing compared to USP < 232 >. We compared a small sample of programs to USP < 232 > limits for a) inhalation and b) oral products. Green indicates state criteria matching at or below USP < 232 > limits. Red indicates state criteria exceeding USP < 232 > limits. When we analyzed MCPs with heavy metal testing listed (n = 33), only 12 programs specified acceptance criteria for both inhalation and oral products, even though nearly all 33 allow both ROAs. The remaining 21 programs only provided a single set of criteria for all ROAs or did not provide a list of specific heavy metals. If only one set of criteria was provided and the MCP allows inhaled products, we compared their action levels to USP < 232 > for inhalation.

One way to facilitate the production of cannabis products that are made to the standards set in drug CGMPs is to follow the recommendations in the FDA Botanical Guidance.¹⁰ This guidance explicitly lists pesticide and heavy metal tests as quality tests to be performed on each batch of drug substance, and asks for the analytical procedures used, the test results, and the proposed acceptance criteria. For heavy metals, the guidance recommends drug developers use the tests outlined by USP. This recommendation is echoed in the FDA Draft Cannabis Guidance.¹² Our analysis identified some areas in which MCPs could more closely align their regulations with drug CGMPs.

Conclusion

There is increased public interest in using CCDPs from state-regulated dispensaries in human clinical trials. In this study, we examined testing regulations within state-level MCPs and compared them with the federal CGMP requirements for drug products used in clinical trials. Our goal is to highlight the differences between the regulatory approaches related to product manufacturing and quality and promote consistency where possible. Our analysis illustrates the variances between MCPs and drug CGMPs. We, as scientific and medical professionals, note that our review was not intended to be a legal analysis, and primarily focused on only the testing sections of these regulations.

Therefore, our analysis might not paint the complete picture of how regulations are implemented in each program. Importantly, although compliance with CGMPs may help ensure the quality and consistency of products, following CGMPs does not ensure that the products are safe or effective for their intended use. The best way to explore the therapeutic potential of these products is through FDA’s investigational new drug program.

Abbreviations Used

AHP

American Herbal Pharmacopeia

As

arsenic

BDS

botanical drug substance

BRMs

botanical raw materials

CBD

cannabidiol

CBDA

cannabidiolic acid

CCDPs

cannabis and cannabis-derived products

Cd

cadmium

CDER

Center for Drug Evaluation and Research

CGMPs

Current Good Manufacturing Practices

FDA

U.S. Food and Drug Administration

Hg

mercury

MCP

medical cannabis program

Pb

lead

ROA

route of administration

THC

tetrahydrocannabinol

THCA

tetrahydrocannabinolic acid