As observed thus far, Cannabis constitutes a pharmacologically rich plant harbouring a plethora of bioactive compounds. The multiplicity of substances within Cannabis, coupled with the multiplicity of receptors to which these substances potentially bind, creates the preconditions for complex interactions among them—sometimes synergistic, sometimes antagonistic. Beyond studies of phytocannabinoids as isolated entities, certain phytocannabinoids have been investigated clinically as the principal component of a Cannabis extract that has undergone no further processing for substance isolation. Such extracts are conventionally designated “full-spectrum”, as they encompass the “full spectrum” of substances produced by the specific chemovar. It is clarified that the notion of “full-spectrum” does not pertain to the entirety of substances potentially producible by the species Cannabis sativa L., but rather to the particular plant subjected to extraction. From a very early stage, it was observed that the action of isolated cannabinoids differs from that of full-spectrum preparations. This divergence manifests either as a tendency toward more adverse effects with isolated formulations or simply as a need for a multiple-fold higher dose of the isolated compound in order to achieve the same pharmacological effect that would be attained with a full-spectrum preparation. This phenomenon is attributed to the multiple synergies and antagonistic interactions within the full-spectrum context, which are absent from preparations of isolated cannabinoids. The totality of these synergistic and antagonistic interactions is currently termed the “entourage effect”. The term was introduced by the group of R. Mechoulam (Ben-Shabat et al., 1998) initially to describe the overall action of endocannabinoids in an in vivo environment, that is, together with all of the organism metabolites (in contrast with their isolated action in vitro). Today, it has acquired the meaning described above in relation to phytocannabinoids [Sanchez-Ramos, (2015); Russo, (2019)]. In French, the word “entourage” means “surroundings” or “environment” and often refers to a group of people who surround and support an individual. By analogy, one may readily envisage minor cannabinoids, terpenes, and flavonoids “supporting” and modifying the action of the primary cannabinoid, thereby constituting its entourage, which is evidently chemical in nature.
Blasco-Benito et al. (2018) examined preparations containing isolated THC in comparison with full-spectrum THC preparations and found that the latter were more effective against ER+/PR+, HER2+, and triple-negative tumors in cell culture systems. The same appears to be confirmed in clinical applications: for example, control of an epileptic syndrome typically requires 5–20 mg/kg of pure CBD [Perucca, (2017); Lazaridis et al., (2019); Chen et al., (2019)], whereas with full-spectrum galenic preparations this is usually achieved with one-half to one-fifth of that dose. The difficulty in interpretation here lies in the fact that observations with full-spectrum galenic preparations derive from unsystematized reports of outcomes in everyday clinical practice, whereas for isolated substances systematic studies are available. Similarly, Ferber et al. (2020) refer to the use of terpenes and cannabinoids in mood disorders and do not hesitate to designate the observed synergies as “entourage effect”, although they do not propose a specific mechanism.
Despite the fact that clinicians who practice cannabinoid therapy observe the “chemical entourage” phenomenon on a daily basis in their patients, the scientific community as a whole has not yet been convinced that the phenomenon exists. Those who dispute it consider that its existence is not supported by pharmacological research data and that advocacy in its favor is driven by commercial motives (Cogan, 2020). The opposing side returns the accusation of commercial motivation, arguing that researchers wish to restrict scientific cannabinoid therapy to pharmaceutical formulations based on isolated molecules and to synthetic cannabinoids, because only these two categories are eligible for patent protection and therefore have commercial interest and economic benefit. Studies up to 2020 have not identified, for example, any action of terpenes on cannabinoid receptors [Santiago et al., (20192); Finlay et al., (20203)] (see footnote 1), at least in the manner in which such action was sought. Nevertheless, they do not take a position against the existence of the phenomenon and state that potential actions of terpenes outside cannabinoid receptors must be investigated. In a more recent, comprehensive study by LaVigne et al. (2021), the terpenes α-humulene, geraniol, linalool, and β-pinene were found to exhibit cannabimimetic activity in experimental animals, which was partly inhibited by CB1R antagonists. It was also found that these four terpenes act in synergy with the synthetic cannabinoid WIN-55,212-2, enhancing its effect. The concentrations of the four terpenes that individually activate CB1R are high to very high (10–500 μM); nonetheless, these findings are sufficient as a basis for supporting the existence of a “chemical entourage” phenomenon. In an even more recent in vitro study by Raz et al. (2023), conducted in a heterologous experimental system (see footnote 2), it is argued that the terpenes borneol, geraniol, limonene, linalool, ocimene, sabinene, and terpineol potentiate THC agonism at CB1R, even at particularly low concentrations (0.001–0.01 μM). These seven terpenes were found to be capable of autonomously activating CB1 receptors, but to 10–50 percent of the activation achieved by THC alone. Consequently, they may be characterized as weak agonists. The combination of certain terpenes with THC increases CB1R activation additively or multiplicatively (synergy) compared with THC alone. Synergy was not proportional to terpene concentration: optimization occurred at concentrations similar to those that occur naturally in Cannabis, that is, very low (contents <2.5 percent in dry flower). On the basis of the data from this study, certain terpenes may be characterized as modulators of THC action at CB1R. If all these findings are confirmed by additional studies, then there will no longer be any doubt about the “chemical entourage” phenomenon, at least with regard to the component that concerns terpenes. Raz et al. (2023) propose that, for a better therapeutic outcome, cannabinoid extracts may be enriched with terpenes selected according to the disease. It must be noted that, on one hand, the strength of the study is reduced by the heterologous nature of the experimental system and, on the other hand, synergies observed experimentally may, in vivo, be converted into antagonistic effects, because weak agonists may, in the presence of strong agonists, manifest as antagonists (Mackie, Devane, & Hille, 1993; see chapter 3).
In a recent electroencephalographic study employing artificial intelligence and focusing on the psychotropic effects of Cannabis, the “chemical entourage” phenomenon appears to be supported by the finding that inhalation of a full-spectrum preparation leads to a richer psychotropic experience compared with an isolated THC preparation, according to electroencephalographic rather than subjective criteria. This finding directly challenges the prevailing assumption that THC content alone determines the psychotropic outcome [Sharma, Haaz et al., (website)].
Nahler (2022) reviewed the action of cannabinoids in neoplastic diseases and found that full-spectrum preparations rich in CBD behave differently from corresponding preparations rich in THC. In the majority of in vitro experiments on neoplastic conditions, pure CBD was superior to CBD-rich extracts, whereas the opposite was observed for pure THC compared with THC-rich preparations. The same does not, however, apply in neurodegenerative and other diseases. This conclusion is open to multiple interpretations. The cause of the discrepancy is to be sought either in methodological issues or in the specific antineoplastic properties of cannabinoids, which may be either enhanced or attenuated in full-spectrum extracts, depending on their particular composition at any given time.
Furthermore, it is striking that, although Santiago et al. (2019) and Finlay et al. (2020) identified the mild agonistic activity of β‑caryophyllene at CB2R, which is in any case documented by other investigators (Fidyt et al., 2016), what is propagated in reviews on the subject is the assertion that terpenes in general do not act on classical cannabinoid receptors. Of course, if all or most terpenes were regular ligands of cannabinoid receptors, then they would either be categorized as cannabinoids or, more likely, there would be no particular interest in Cannabis, since numerous other plants would provide terpenes with cannabinoid properties. Therefore, this entire discussion belongs to the prevailing ultraconservative medical habit, which rests on the logical fallacy of denying the existence of phenomena solely because their existence has not yet been demonstrated, while suppressing (and/or concealing) the fact that their nonexistence has not been demonstrated either (see footnote 3). The research, which will inevitably be completed at some point, must turn toward the study of other targets, whether these are receptors or merely biochemical/pharmacological targets, because the chemical entourage phenomenon in cannabinoid therapy can have only three components.
The direct pharmacodynamic component: This concerns the effect of two or more substances on one and the same receptor. This depends on which G protein and which GTPases are activated via the receptor by the respective ligand and constitutes the phenomenon of preferential agonism (biased agonism; see chapter 3). It manifests either as antagonism, additive action, or synergy.
The indirect pharmacodynamic component: This concerns the functional interaction of two or more substances that act via different receptors, or via mobilisation of different second messengers by the same receptor6. At the functional level, the two initially independent actions are either antagonistic, additive, or synergistic.
The pharmacokinetic component: This concerns absorption, tissue distribution, metabolism, and excretion.
An additional indication in favor of the existence of the chemical entourage phenomenon arises from the toxicology of synthetic cannabinoids, which flood the market under various names such as Spice, K2, Magic Gold, ZoHai Rx, Probation Kush, and others. These preparations contain no CBD, no terpenes, and no flavonoids, while they are full agonists at CB1/2R and other receptors (GPR55, TRPs, etc.). Their agonism in brain cells and other organs is both full and unmodulated, because the multitude of accompanying compounds that we term the chemical entourage is absent; hence they exert destructive effects that are not observed with Cannabis use (Spaderna, Addy, and D’Souza, 2013).
The phenomenon of the chemical entourage remains, in many of its aspects, inadequately studied. One such aspect is that extracts from different Cannabis chemovars, with the same basic composition (for example, content of the principal cannabinoids CBD and THC), inhibit the growth of different cancer cell lines in vitro. Similar differences are observed in autistic patients, in pain syndromes, in epileptic syndromes, and so forth, even when only the batch of the preparation is changed while the gross composition remains the same. The unresolved question in this context is the following: How many and which constituents of a Cannabis extract are necessary for the extract to preserve its properties (for example, to be cytotoxic to certain cancers but not to others, or to control the epileptic seizures of a specific patient but not of another, and so on)? Questions of this kind are being addressed by the group of D. Meiri at the Technion Research Institute in Israel (Berman et al., 2018). With so many substances interacting, linear prediction, using finite parameters and large numbers, is feasible only up to a point. The stochastic approach being pursued at Technion may yield conclusions of practical applicability [lecture by D. Meiri, 2018 (YouTube, after the second minute)].
In conclusion, studies evaluating Cannabis constituents have shown that many of them could be utilised therapeutically within a comprehensive approach to various diseases—particularly chronic, degenerative, and malignant conditions—without the exclusions of the past, which were grounded solely in prejudice and excessive conservatism. Cannabis offers certain advantages in the management of symptoms in such patients. Acceptable indications already include several epileptic syndromes, neuropathic and cancer-related pain, spasticity in multiple sclerosis, chemotherapy-induced nausea and vomiting, and possibly autistic disorders, post-traumatic stress disorder, selected manifestations of dementias and neurodegenerative disorders, and the anorexia and cachexia of patients with AIDS. It provides the advantage of addressing all of these with a single preparation and with a low likelihood of addictive and/or dependence-producing consequences (certainly lower than that of opioids). A potential advantage is that, with appropriate management—empirical thus far—it might also exert disease-modifying effects. It has the disadvantage that cannabinoids are less potent than specialised antiemetics, analgesics, hypnotics, and so forth.
Beyond the therapeutic action of phytocannabinoids, which is indisputable at the preclinical level and partly at the clinical level thus far, there are also their accompanying substances in Cannabis and in other plants: terpenes and flavonoids. These do not yet appear to be taken sufficiently seriously with respect to their therapeutic value. However, traditional ethnobotanical knowledge and the signals emerging from existing studies already stimulate interest in their inclusion in clinical research, since certain of these substances seem able to contribute to tissue protection by various mechanisms and/or to provide pharmacokinetic support to chemotherapeutic agents, in addition to their participation in the “chemical entourage” phenomenon of cannabinoids.
Footnote 1: The study by Santiago et al. (2019) has the methodological limitation of searching for possible terpene effects only in terms of modulation of GIRK channels, whereas modulation could also be manifested in any other manner, via various receptors or independently of them. The study by Finlay et al. (2020) has excluded orthosteric and allosteric binding of terpenes to classical cannabinoid receptors. Remaining is the examination of all other receptors (within and outside the endocannabinoid system) and the investigation of any direct and indirect pharmacological interactions with the multitude of Cannabis compounds, which it does not address.
Footnote 2: Heterologous means that cannabinoid receptors were expressed in a cell line that does not naturally produce them. Typically, the genes for the CB1 receptor are introduced into a host cell (often a cell type that is easy to culture and handle, such as a human embryonic kidney cell line or HEK cells). This system allows researchers to study receptors in isolation from other variables that are present in their natural environment and are expected to complicate the intended observation.
Footnote 3: Example of sophistry: The proposition “It has not been proven that Cannabis constitutes a therapeutic option for malignant tumors” is presented as equivalent to the proposition “It has been proven that Cannabis does not constitute a therapeutic option for malignant tumors”, while the proposition “It has not been proven that Cannabis does not constitute a therapeutic option for malignant tumors” is almost universally ignored.