In this study, we showed that, despite the complexity of cannabis pharmacology and the difficulty in precisely quantifying the results of treatment with cannabis, it is possible to assess the role of certain genetic polymorphisms in patients and to explain part of the great inter-individual variability observed in pain reduction. On average, the pain reduction in remaining patients (non-drop-out) at the end of the first month amounted to approximately 20%, but with great variability. Part of this was significantly associated with three polymorphic candidate genes: ABCB1, TRPV1 and UGT2B7. The three other genes, as the other variables (year and centres of recruitment, age, sex, disease, concomitant use of other painkillers and medication to treat the patient’s main disease, type of cannabis and quantity of THC and CBD taken) not only showed no significant association, but remained very far from significance. It is very likely that the associations observed were in part disguised by the constant adjustment of cannabis dosages, being approximately doubled on average, but reduced in certain cases, and by the fact that approximately 30% of patients changed their type of cannabis during the first month of treatment to the “titrate” type.

The rs1045642 polymorphism, also known as C3435T, of the MDR1/ABCB1 gene, which encodes a transporter protein (MRP1), produces a “silent” missense variant, which, however, determines the reduction in the expression of the transporter protein [15]. Patients with neuropathic pain treated with nortriptyline and TT homozygotes showed a lower pain reduction than others [16]. In this study, homozygous TT patients were less responsive to treatment with cannabis than the others. These results were in accordance with those obtained in a morphine treatment for chronic pain, as TT homozygous patients were less responsive than the others [12], confirming the role of this gene in the control of drug transport. Furthermore, these results were supported by the discovery that, in heavy cannabis users, THC plasma levels were significantly higher in carriers of at least one T allele [17]. An increase in the risk of neuropathy has also been reported after treatment with taxanes in TT homozygous patients [18]. It is interesting to note that this polymorphism has also been associated with variability in the cold pain threshold and in pain tolerance in healthy males [19], suggesting a role of the MDR1/ABCB1 gene in pain control, the pharmacological treatment aside.

The TRPV1 polymorphism (rs8065080), also known as c.1191A > G, is a common variant in the TRPV1 gene (transient receptor potential cation channel), which encodes, among others, vanilloid receptor subtype-1 and the capsaicin receptor. The polymorphism determined by rs8065080 was associated with salt sensitivity [20], the tolerance of cold, heat, pinpricks, homozygous carriers of the C allele with a higher pain tolerance [21] and with painful osteoarthritis of the knee [22]. In our study, CC homozygotes experienced a significantly greater pain reduction compared to the other genotypes. There is extensive evidence of the role of the TRPV1 gene in the therapeutic activity of cannabis. It has recently been demonstrated that TRPV1 channels are activated by cannabinoids [23], as by myrcene, contained in cannabis, which is recommended to obtain analgesic effects [24]. Cannabis has been successfully used in the treatment of epilepsy, and the inhibition of the transient receptor potential of vanilloid type 1 channels could be one of the mechanisms at the basis of CBD anti-seizure effects [25,26].

The UGT2B7 gene encodes UDP-glucuronosyltransferase-2B7, involved in phase II metabolism, conjugating a variety of compounds such as analgesics (morphine), carboxylic non-steroidal anti-inflammatory drugs (ketoprofen) and anti-cancer drugs (all-trans retinoic acid) [27]. The rs7438135 polymorphism has been proven to be involved in opioid withdrawal symptoms [28] and in the glucuronidation model of morphine [29]. In the present study, AA homozygous patients reported a significantly greater pain reduction compared to AG and GG heterozygotes. These results were supported by the evidence that oncological patients with an AA genotype are, generally, more sensitive (less tolerant) to opioids compared to those with AG genotypes [30].

The fact of having found three polymorphic genes which individually significantly influence pain decrease after treatment with cannabis poses the problem of the impact of their possible combinations in patients. As expected, patients simultaneously carrying the most favourable allelic combinations showed a greater pain reduction (∆VAS = 1.73) compared with those with a less favourable combination (∆VAS = 0.56). Considering genotype combinations, we could, therefore, group patients into good responders, intermediate responders and poor or non-responders. We are aware that this study included many variables, such as cannabis types, contents of active principles, different groups of diseases and patients’ overall clinical situations, the interactions of which were difficult to assess as they were not planned a priori, but could have important effects on the assessment of pain reduction. However, the results suggested that a patient’s genetic makeup is, at the moment, a significant descriptive factor of variability in the response to cannabis. Changes in cannabis dosages, doubled on average, which occurred during the first month and were empirically guided, were not associated with any factor, including genetic factors, thus, suggesting that the initial dosage, apparently low, was due to the caution of physicians who administered a drug which was not well known to patients who were treated for the first time. Furthermore, the association study showed that the A allele of the CNR1 gene, rs1049353, was a therapy drop-out risk factor. This gene encodes the most relevant cannabinoid receptor, the polymorphism of which was extensively studied in depression therapy [31,32].