As cannabis becomes a legitimate medical substance, its chemical consistency is critical for developing cannabis-based pharmaceutical products used in patient care for different medical purposes. Although cannabis strains were originally thought to be either ‘Sativa’ or ‘Indica’ cultivars, the extensive breeding of the cannabis plant has yielded cannabis into a hybrid chemotype of various strains. Licensed producers have used distinct agricultural processing practices to fabricate a variety of cannabis strains with heterogenous chemical compositions and concentrations [11]. This has resulted in numerous available strains and often significant batch-to-batch variation in the marketplace. These variations impact the potency, pharmacokinetics and physiological effects of cannabis-based products. Therefore, it is important to identify and classify medical cannabis strains by their chemistry (i.e., chemovar) rather than their imprecise breeding pattern or name. Given the lack of standardization of the chemical constituents within cannabis products, the clinical outcomes for one defined cannabis-based medicine cannot be generalized to other chemovars. Thus, the ability to demonstrate chemical consistency in cannabis products is essential for the quality production of safe and effective cannabis drug development to ensure reproducible effects.

The major advances in cannabinoid research are attributed to the discovery of the endocannabinoid system (ECS), which is composed of endogenous ligands (primarily anandamide and 2-Arachidonoylglycerol), enzymes and cannabinoid receptors (CBR) that are widely distributed throughout the human body [12,13,14]. The ECS plays vital regulatory roles in multiple physiological processes and behavioral pathways related to inflammatory and immune regulation, synaptic plasticity and neuroprotection, cancer progression, appetite and metabolism, thermogenesis, learning and memory, circadian rhymes, pain regulation, mood and behavior, reproduction, gametogenesis and heart function [15,16,17,18,19,20,21,22,23,24,25]. Recent studies have shown that the ECS may potentially undergo epigenetic modulation through various environmental and lifestyles factors that primarily target genes encoding for CBRs and that endocannabinoids may themselves induce epigenetic alterations [26,27]. Although ECS activity impairment has been implicated in several pathological conditions, it further emphasizes the relevance of the ECS as a potential therapeutic target for disease treatment [28,29].

To date, two CBR subtypes have been identified, CB₁ receptors (cloned in 1990) and CB₂ receptor (cloned in 1993), which differ in signaling mechanisms, tissue distribution, agonists/antagonists sensitivity and an amino acid sequence with CB₂ sharing only 44% of amino acid sequence identity with CB₁ [30,31,32]. The CBRs are a class of cell membrane receptors that belong to the G-protein coupled receptors (GPCRs) family with their endogenous ligands appearing to be from a family of anandamides [30] More recently, a third CBR known as GPR55 has also been proposed as a possible target for cannabinoid activity but this remains controversial [33]. CB₁ receptors are the most abundant GPCR, which are primarily expressed throughout the gastrointestinal tract and the central nervous system (CNS) on the axons and axon terminals of neurons predominantly located in the cingulate gyrus, prefrontal cortex, hippocampus, cerebellum and basal ganglia [34]. Activating the CB₁ receptors in the presynaptic terminal inhibits adenylate cyclase and the cyclic adenosine monophosphate (cAMP) and protein kinase A pathway, resulting in A-type potassium channel activation and subsequently inhibiting various neurotransmitters and neuromodulators [30,35]. CB₂ receptors are predominately expressed on immune cells and tissues, including the thymus, spleen, tonsils, lymphocytes and macrophages [30,31,36]. More recently, CB₂ receptors have also been shown to be present in the CNS (i.e., neurons of the hippocampus and microglial cells) in low concentrations [37]. Similar to CB₁, activating CB₂ inhibits adenylate cyclase and initiates mitogen-activated protein kinase, leading to increased intracellular calcium levels [30,35]. One of the functions of CB₂, along with CB₁, in the immune system is to modulate cytokine release [38]. Activating CB₂ also induces apoptosis, suppresses autoantibodies and cell proliferation and provides an overall inhibitory effect on inflammatory processes [39].

The majority of the chemical constituents of cannabis are produced by the female Cannabis sativa plant, which is classified into hundreds of varieties based on the composition of its fatty compounds known as phytocannabinoids [40,41]. All phytocannabinoids in cannabis are highly lipophilic molecules with low aqueous solubility [40,41]. Phytocannabinoids exert their effects on the ECS through multiple receptors, including CBRs, opioid and serotonin receptors, adrenergic receptors and GPCRs [12,42,43]. To date, approximately 568 compounds have been identified within the cannabis plant, of which roughly 120 are active phytocannabinoids that are synthesized in the secretory cells inside the glandular trichomes [40,41]. These naturally occurring phytocannabinoids share common chemical structural features (i.e., dibenzopyran ring and a hydrophobic alkyl chain) and are distributed among subclasses, including delta-9-tetrahydrocannabinol (Δ⁹-THC), Δ⁹-tetrahydrocannabinol (Δ⁸-THC), CBD, cannabigerol (CBG), Δ⁹-tetrahydrocannabivarin (THCV), cannabivarin (CBV), cannabidivarin (CBDV), cannabinol (CBN), cannabichromene (CBC), cannabinodiol (CBND), cannabielsoin (CBE), cannabicyclol (CBL), cannabichromene (CBC), cannabitriol (CBT) and over 30 other miscellaneous types that have undetectable levels in many commercial cannabis chemovars (Figure 1) [44]. Of these phytocannabinoids, the most chemically abundant and well-studied are THC and CBD. Although THC and CBD can be found in the dried cannabis plant, both are naturally produced in their carboxylic acid form, as tetrahydrocannabinol acid (THCA) and cannabidiol acid (CBDA), which are synthesized from a common precursor, cannibigerolic acid (CBGA) (Figure 2) [45,46].

One of the main challenges in addressing cannabis as a therapeutic is that, in contrast to traditional pharmaceuticals, cannabis contains several bioactive compounds and thus behaves as a complex mixture of active pharmaceutical ingredients (APIs). In addition to the extensively studied effects of THC and CBD, minor phytocannabinoids as well as non-cannabinoid phytochemicals such as CBC, CBG and beta-caryophyllene have also exhibited binding to cannabinoid receptors [47,48,49] and it remains to be determined how these activities contribute to or modulate the receptor activity of cannabis as a whole. Further complicating matters is the diversity of cannabis chemovars available to patients. The phytocannabinoid composition of cannabis is generally dominated by the THC and CBD compound families (i.e., THC, CBD and their acidic precursors THCA and CBDA) [50]. However, even among these major phytocannabinoids, significant variation in content is observed across cannabis varieties. In an analysis spanning 277 medical cannabis products from 25 licensed Canadian producers, Mammen et al. noted that THC content varied from 0.14% to more than 25% [51]. Meanwhile, the chemical quantification of 47 medical cannabis extracts and 12 cannabis oils by Yang and colleagues showed that CBD, THC, CBDA and THCA concentrations ranged from 0–51.7%, 0–73.1%, 0–60.3% and 0–76.1% (w/w), respectively, in the extracts and 0–8.9%, 0–4.6%, 0–31.7% and 0–15.3% (w/v), respectively, in the oils [52]. Variance in the chemical composition of medical cannabis products inevitably leads to variance in pharmacological responses when these products are consumed by patients [52].

Over the past few years, attempts have been made to deconvolute the chemical complexity of medical cannabis. Recently, Yang et al. pioneered a chemoinformatic approach, whereby the chemical profiles of medical cannabis extracts and cannabis oils were correlated to their corresponding in-vitro cannabinoid receptor activities. This regression analysis generated models which can predict the CB₁ and CB₂ activities of cannabis extracts and oils based on their concentrations of 4 major phytocannabinoids—CBD, THC, CBDA and THCA [52]. While such modeling at the in-vitro level has only limited use in a clinical context, it provided evidence that phytocannabinoids may behave differently when administered as part of an extract, compared to when administered as a pure compound. It is thus hoped that in addition to receptor activities, the inclusion of biomarkers and clinical markers such as cytokine levels and von Frey filament measurements as dependent variables could lead to more sophisticated modeling that accurately elucidates the relationship(s) between the chemical makeup of medical cannabis and their corresponding in-vivo clinical outcomes for ailments such as inflammation and chronic pain. Presently, healthcare professionals play an auxiliary role in the administration of medical cannabis, which is largely self-titrated by patients [53,54]. As such, more advanced and clinically relevant models could serve to guide physicians in their medical cannabis prescriptions, by narrowing down potential effective chemovars based on the type and severity of symptoms experienced by patients.

THC is the most prevalent phytocannabinoid in drug chemotypes, which is primarily metabolized in the liver through cytochrome (CYP) 3A4 and CYP2C9 [55]. Chemically, THC has a tricyclic 21-carbon structure without nitrogen and with two chiral centers in trans-configuration [12]. The decarboxylation of THCA to active THC occurs gradually in the dried cannabis plant but can be accelerated upon heating (i.e., smoking, baking or vaporization) [45,46]. This decarboxylation step is essential for obtaining higher potency and binding affinity at CBRs for achieving reliable pharmacological effects when cannabis and its derivatives are used for their therapeutic purposes [11,45,46]. THC, along with CBN and CBND, are known to induce a psychotropic effect [12]. This occurs as THC, being a CB₁ agonist and a weak CB₂ partial agonist, activates CBRs through the attenuation of neuronal mechanisms that stimulate receptor downregulation and inhibits the release of neurotransmitters, resulting in a psychotropic high [56]. Various ligand binding studies have indicated that THC has a high affinity to both CB₁ (Ki = 53 nM) and CB₂ receptors (Ki = 40 nM) but mainly binds to CB₁ receptors due to higher receptor affinity [57].

CBD is the second most prevalent phytocannabinoid in drug chemotypes that is metabolized by the cytochrome P450 pathway by several isozymes [55]. It has a bi-cyclic 21-carbon structure with a double bond in the terpene ring [12]. The decarboxylation of CBDA to active CBD also occurs gradually but can be expedited upon heating [11]. CBD is a non-psychotropic phytocannabinoid and at high doses can modulate the intoxicating effects of THC through CB₁ dependent mechanisms despite having low CB₁ binding affinity [58]. This observed CB₁ antagonism may be related to CBD’s ability to behave as a negative allosteric modulator of THC at CB₁ [59]. CBD is an antagonist of CBRs that only indirectly interacts with CBRs at high concentrations and has been reported to have relatively weak binding affinities for CB₁ (range, 4.35–10 nM) and CB₂ (range, 2.4–10 nM) [11]. Unlike THC, studies have proposed that CBD also binds to other non-CBR types of receptors with low efficiency, including serotonin 1A receptor, vanilloid receptor 1 and adenosine A2A receptors [12,42,43].

Cannabis also contains other organic compounds, such as flavonoids and terpenoids, which may contribute to cannabis chemical activity. Flavonoids impart cannabis’ color while terpenoids are responsible for the taste and aroma of cannabis [60]. Both compounds vary substantially across different cannabis chemovars. Over 200 terpenoids have been reported with approximately 50 cannabis terpenes routinely encountered in North America chemovars [60,61]. In the majority of cannabis samples, terpenoids consist of <1% of cannabis with the most common terpenes being myrcene, pinene, terpinolene, limonene, β-caryophyllene and linalool (Figure 3) [60,61]. Although both flavonoids and terpenoids can be bioactive, minimum research has been conducted on their physiological and pharmacological effects in humans. Collectively, when active and inactive cannabis metabolites interact to affect receptor potencies of the active constituents, this is referred to as an “entourage effect.” It is believed that the entourage effect can influence the subjective experience or clinical effects of cannabinoids by non-cannabinoid compounds [60,61]. This suggests that cannabis chemovars may differ in their clinical applications due to their unique combination of cannabinoid and non-cannabinoid components.

From a chemical perspective, exogenous CBR ligands can be categorized into three primary types—phytocannabinoids, endocannabinoids and synthetic cannabinoids (SC) (Figure 4). As mentioned above, phytocannabinoids are the natural plant-derived compounds found in the cannabis plant. Endocannabinoids are naturally occurring cannabinoid molecules that are synthesized in the living organism [62]. The two main and most well-studied endocannabinoids are anandamide and 2-Arachidonoylglycerol, which partially overlap some of the behavioral and molecular characteristics of THC [13,14,63]. Following the isolation and characterization of THC, SC derivatives were also developed. SC, such as dronabinol, nabilone and nabiximols are laboratory fabricated structural THC analogues that have gained widespread attention for their utilization in various disease aliments as they closely mimic the effects of THC on the ECS [62]. Although SC are considered chemical relatives to cannabis, differences do exist among SC and plant-based cannabis. First, THC in natural cannabis is considered a partial agonist to CBRs, whereas SC are full agonists with higher affinity for CBRs [64]. Second, SC are produced as either oral capsules or sprays, which alter their pharmacokinetics and pharmacodynamics [65]. For instance, dronabinol, an oral synthetic THC capsule, has poor solubility and its high first-pass hepatic metabolism accounts for its poor bioavailability with only 10–20% of the administered dose reaching systemic circulation [65] Dronabinol and its active metabolite, 11-hydroxy-delta-9-tetrahydrocannabinol (11-hydroxy-THC), also have a slow absorption rate with peak plasma concentrations being obtained at 2–4 h after consumption that declines steadily over several days [65]. The compound 11-hydroxy-THC can also be more potent than THC and produced in larger quantities, which can heighten SC psychoactive and adverse effects [66,67]. In contrast, inhaled or vaporized cannabis is rapidly absorbed and fast-acting as it bypasses gastrointestinal absorption and hepatic metabolism [66,67]. Third, aside from THC, SC lacks all other active ingredients found in cannabis (e.g., terpenoids) [64]. Given this, SC may be less effective and contribute to an increased psychoactive effect as it does not contain elements to moderate these chemical effects and possibly prevents the entourage effect. SC products also contain additives and chemical preservatives, which may alter their intended effects [64].