The role of the endocannabinoid system in cancer

Each cell suffers tens of thousands of molecular lesions per day, resulting in DNA damage that threatens genome integrity. Some mutations affect cell proliferation due to defects of certain genes, e.g., oncogenes, tumour suppressor genes, genes that control repair mechanisms, or the cell cycle. Eventually, this may lead to abnormal cell proliferation and cancer when DNA damage exceeds the repair capacity of the cell. Basically, the role of the endocannabinoid system (ECS) is to maintain tissue homeostasis. Many observations point to a dysregulation of the ECS in cancer; this can include variation in the signalling pathway, expression of receptors, enzymes, and/or concentration of endocannabinoids. Most studied are the cannabinoid receptors 1 and 2 (CB1, CB2) and their natural ligands, the endocannabinoids anandamide (AEA) and 2-arachidonoylglycerol (2-AG). Despite that our knowledge has considerably increased, the role of the ECS in cancer as well as the anticancer activities of cannabinoids, are far from being completely understood. In a number of tumour entities, canonical receptors such as the above mentioned CB1, CB2, and particularly GPR55 were found to be upregulated, although with sometimes conflicting results. Other receptors that have been reported to be expressed differently in cancer cells compared to normal cells are e.g., transient receptor potential cation channel subfamily members, melastatin type (TRPM8), vanilloid type (TRPV1, TRPV2), or ankrin type (TRPA1). Loss/reduced expression or increased expression of CB1 and possibly CB2 may accelerate tumour growth whereas increased levels of endocannabinoids AEA and/or 2-AG may increase apoptosis as they can activate not only CB1 and CB2 but many other receptors. Conversely, enzymes such as the fatty acid amid hydrolase (FAAH) or monoacylglycerol lipase (MAGL) decrease endocannabinoid levels and potentially promote cancers such as melanoma. Consequently, such imbalances compared to normal tissues may have prognostic value.

The role of cannabinoids in cancer

A large range of effects of cannabinoids on the tumour and tumour microenvironment respectively have been described in the past, notably inhibition of angiogenesis, metastasis, tumour proliferation, and tumour cell apoptosis. Although the exact mechanism of anticancer activities of cannabinoids is still a matter of intensive research, it is believed that substances such as CBD and THC modulate the ECS and oxidative stress conditions within cells. This seems to be a major mechanism as it induces instability of cell membranes, autophagy, mitochondrial fragmentation, and ultimately apoptosis of cancer cells. Cannabinoids and endocannabinoids target selectively and differently a wide range of receptors such as CB1, CB2, GPR55, TRPV1, TRPV2, TRPA1, TRPM8 known to play a role in cancer, and which are also differentially expressed amongst specific cancer cell lines. Effects are further modified, as cannabinoids modulate in vivo the level of endocannabinoids that have per se also anticancer properties [28, 29]. CBD as an example, increases levels of AEA by interfering with FAAH. As targets including FAAH are differently expressed in tumours, the effects of cannabinoids vary and may be cancer cell-specific. This explains why cannabinoids, but also extracts differing in their composition of cannabinoids, vary in their effects on cancer cell lines.

Moreover, cannabinoids (among other substances) affect tumour growth also by epigenetic modulation. Hypomethylation, as an example, is often found in tumours [30]. Hypo- and hypermethylation are complex mechanisms and can occur in parallel. On one hand, DNA methylation activates the formation of repair enzymes, but on the other, aberrant methylation can activate proto-oncogenes and inactivate tumour suppressor genes. As epigenetic changes are potentially reversible there is much hope that drugs (“epigenetic drugs”) can restore a normal epigenetic pattern and, perhaps, reduce the malignancy and/or growth of tumour cells. Noteworthy, CBD reduces the expression of inhibitor of DNA-binding-1 (ID-1), a transcriptional regulatory protein and key regulator of tumour cell invasiveness that is epigenetically activated in many tumours such as breast cancer and glioma [26].

In the native plant, cannabinoids are almost exclusively found in their acid forms such as CBD acid (CBDA) or THC acid (THCA); the respective decarboxylated, “neutral” cannabinoids such as CBD, THC, or CBN are, in fact, artifacts formed naturally by aging or artificially by heating. Overall, it seems that most cannabinoids have cytotoxic properties although they vary to a high degree. In head-to-head tests, CBD was generally the most effective phytocannabinoid, acid forms the least [25, 31, 32]. An in vitro study that assessed 12 different extracts on 12 cancer cell lines demonstrated that cytotoxicity differs considerably, whereby extracts with a high percentage of decarboxylated (neutral) cannabinoids (≥ 50% weight/weight) were generally more cytotoxic [13]. Pure THC, however, was less effective than the extracts. Only the LNCaP prostate carcinoma cell line was exceptionally sensitive to extracts containing natural phytocannabinoids in their “native”, acid form (e.g., THCA, CBDA, and others). As mentioned before, phytocannabinoids work through different pathways and receptors, the expression of which varies among cancer cell populations.

Overall, there is abundant literature demonstrating in vitro anticancer activities of cannabinoids and cannabis extracts. As such results have lower therapeutic evidence than animal models and observations in patients, these articles are excluded from the present overview. Moreover, effects in vitro do not always match results observed in vivo [24]. In the following, articles describing the effects of cannabinoids in animal models and/or patients with the brain, breast, colorectal, head and neck, haematological, hepatocellular, lung, pancreatic, ovarian, prostate, and skin cancers are summarised.

Brain tumours, glioma, and glioblastoma

There are over 150 different types of brain tumours. The most malignant (and most common in adults) is glioblastoma multiforme (also called glioblastoma or grade IV astrocytoma). Its incidence is around 2 to 4 per 100,000 and accounts for roughly 50% of the primary brain tumours. Within the various types of gliomata, glioblastoma ranks among the deadliest cancers with a mean overall survival of around 15 months; this is one of the shortest among cancers. Glioblastoma has a high degree of genetic heterogeneity which influences also prognosis. As an example, glioblastoma with an (epigenetic) O⁶-methylguanine-DNA methyltransferase (MGMT) promotor methylation does not produce the DNA-repair enzyme MGMT and shows a better response to standard chemotherapy with temozolomide (TMZ). The situation in glioblastoma is further complicated by the fact that molecular heterogeneity of cells seems to occur even within the same tumour. This contributes to tumour recurrence and treatment resistance [33].

There is a general consensus that high-grade glioma, including glioblastoma, express high levels of CB2 receptors whereby their expression positively correlates with tumour malignancy [34]. Levels of 2-AG can be elevated in surrounding tissues but the role of endocannabinoids is not entirely clear [35]. Patients with low MAGL expression (therefore higher 2-AG levels) had a slightly shorter overall survival than those with high MAGL expression [36].

A further receptor expressed in glioblastoma and also correlating with malignancy is GPR55. Conversely, levels of AEA seem to be decreased whereas CB1 and CB2 were elevated [35], or demonstrated no big differences between healthy and malignant cells, except isocitrate dehydrogenase (IDH)-wild type glioma compared to IDH-mutant. IDH-wild type glioma expressing CB1 may eventually better respond to agonists such as THC. TRPV1, another potential target of CBD, is under-expressed in glioma.

In contrast to many other substances including cannabinoid acids, decarboxylated cannabinoids penetrate the blood-brain barrier very easily [37]. Numerous publications described the effects of CBD and THC against glioma cells in vitro. Intriguingly, THC at submicromolar concentrations of 100 nmol/L increased the proliferation of glioblastoma U373-MG cells in vitro whereas much higher micromolar concentrations of 4 µmol/L or above induced apoptosis, indicating that the effect depends on concentration [38]. Similarly, increased viability of glioma stem cells occurred in vitro with moderate concentrations for CBD (below 10 μmol/L) and for CBG (below 5 μmol/L) but not with higher concentrations (10–15 μmol/L) [39]. The reason for this is not entirely clear but the lack of the influence of the natural antitumour immune system in such in vitro models may play a role. Studies in glioma cell lines have demonstrated that CBD triggered TRPV2-dependent Ca²⁺ influx, increasing the uptake of chemotherapeutic drugs such as TMZ, carmustine, and doxorubicin, and synergising with the drugs to induce apoptosis of glioma cells. The pro-apoptotic effect was greater than with the drugs alone, with no such effect in normal human astrocytes [40]. CBD most likely induces cell death (apoptosis) by upregulating reactive oxygen species (ROS), whereby apoptosis is dependent on the activation of TRPV2 [40].

The largest number of animal studies investigating cannabinoids in cancer has been on brain tumours. In vivo, locally delivered microparticles loaded with CBD (15 mg/kg) reduced the weight of U87MG cell-derived tumour xenografts in mice slightly more than THC (15 mg/kg) or their combination (7.5 mg/kg each) [41]. Other in vivo studies on brain tumour cell lines and cannabinoids are summarised below (Table 1).

In no other disease entity, so many animal experiments have been performed with cannabinoids as in brain tumours. Altogether, with the exception of only one study [42], results demonstrated a reduction of tumour growth in animals by cannabinoids administered mostly in the order of 15–25 mg/kg, more by CBD than THC. Combinations of CBD + THC or TMZ or triple combinations increased effects further. In one experiment complete eradication of tumour cells was observed in 1/5 animals after treatment with CBD (15 mg/kg i.p., 5 days per week for 28 days) [52]. Importantly, in another experiment that tested various combinations of CBD + THC in subcutaneous xenografts, no notable differences were observed in tumour growth inhibition with ratios of THC:CBD varying between 1:1 and 1:6 [47].

Although therapeutic experience with cannabinoids in patients with brain tumours is still very limited, all eight articles summarised below describe positive effects (Table 2).

As can be seen, treatments and patient population in these articles is extremely heterogenous. Cautiously, results suggest that cannabinoids, CBD, THC, and combinations including extracts achieved a treatment response in patients with brain tumours; in most cases, the concomitant chemotherapy was maintained. Three independent case series showed a benefit or an extension of overall survival with daily doses of CBD around 400 mg, and two with a THC:CBD combination. The publications on CBD include also observations with CBD of unknown purity in diffuse midline gliomas. Diffuse midline gliomas are aggressive paediatric brain tumours originating in the midline brain structures. Normally, they have a very short median survival of 10–11 months which seems, however, to double with concomitant CBD [49].

Pure THC has been used only once, as local administration in an experimental clinical trial in patients who were terminally ill; it was not aimed to study a therapeutic response. Other publications reported either results with cannabis of undefined composition or with a 1:1 combination of CBD with THC. Intriguingly, the study of Schloss et al. [58] favoured a 1:1 combination over an increased ratio of THC:CBD. To note, the experiments of Lopez-Valero et al. [47] favoured a higher ratio of CBD at the expense of THC.

Breast cancer

Breast cancer is the most common cancer in women (incidence 119 per 100,000), with varying molecular biologic characteristics. The most aggressive subtype, the so called triple-negative breast cancer (TNBC), is deficient in oestrogen receptor (ER), progesterone receptor (PR), and human growth factor receptor (HER2) expression. It accounts for 10–15% of breast cancers [60]. For this subtype chemotherapy and/or immunotherapy is in general the only therapeutic choice. Preclinical experiments suggest that cannabinoids, particularly CBD, may be effective against breast cancer. Repeatedly, a dose-dependent response to CBD has been reported on various breast cancer cell lines including TNBC cells. CBD is able to inhibit ID-1 gene expression in aggressive breast cancer cells, leading to the attenuation of tumour aggressiveness [26]. ID-1 is a regulator of differentiation/DNA binding and regulator of metastasis. Other effects of CBD include the inhibition of GPR55, known to be elevated in many malignant tumours including aggressive TNBC. Overexpression of GPR55 correlates generally with poor prognosis. In contrast, the role of the canonical receptors CB1 and CB2 is less clear. CB2 has been found to be overexpressed in HER2+ breast cancers as well as in gliomas and other cancer cell lines [61]. Opposite to previous reports, however, high expression seems to correlate with lower malignancy and better survival respectively [62]. CB2 is a receptor mainly found on immunocompetent cells such as leucocytes, monocytes, and macrophage-derived cells that are involved in tumour cell apoptosis [63].

Surprisingly, despite its medical importance and the number of encouraging in vitro experiments with various breast cancer cell lines demonstrating the cytotoxicity of cannabinoids, the number of animal experiments is rather limited, and studies in women are completely absent. The large majority of experiments have been done with triple-negative cancer cell lines, underlining the tumour-inhibiting effects of cannabinoids. It seems that effects increase with increasing doses. Animal studies are summarised in Table 3 below.

As can be seen, most of the 7 animal experiments were performed with CBD, with a remarkably wide range of dosages (1–10 mg/kg) administered daily or 2 to 3 times a week. Only two studies included THC whereby extracts were superior to pure THC [64]. CBD, administered before chemotherapy or in parallel, has been found to sensitize breast cancer and prostate cancer cells in vitro to cisplatin and paclitaxel and significantly increased chemotherapy-mediated apoptosis, suggesting a potential adjuvant role [70–72]. CBD administered before doxorubicin sensitized human triple-negative MDA-MB-231 cancer cells also in vivo.

Surprisingly and unexpectedly, one experiment demonstrated, in contrast to a similar, later experiment, a dose-dependent increase of the tumour mass (number and size of 4T1 tumour metastases in the lungs) after THC 25 mg/kg and even more so (almost twice as high) in BALB/c mice treated with THC 50 mg/kg (every other day for 18–21 days) [67]. BALB/c mice demonstrate Th2-biased immune responses and are relatively susceptible to infection and neoplastic diseases. With MCF-7 breast cancer cells, a widely used human breast cancer cell line positive for oestrogen, progesterone, and glucocorticoid receptors, it was demonstrated that THC-induced cancer cell growth most likely by interaction with cyclooxygenase-2 (COX-2), 17β oestradiol and aromatase [73]. No similar growth enhancing effect is known for CBD. Other factors that could favour tumour cell proliferation are hypoxic conditions [74]. Moreover, THC not only suppresses antitumour immunity, but animal models use genetically manipulated mice to induce breast cancers; this differs from the situation in humans.

In contrast to TNBC-cell lines, only few studies exist about the effects of cannabinoids on hormone receptor-positive (HR+) breast cancer cell lines as they are in general amenable to hormone therapy. Approximately 70% to 80% of breast cancers express ER and are therefore HR+. Furthermore, 65% of these cancers are also PR+.

In vivo, a combined treatment with tamoxifen (2.5 mg/kg) plus a cannabis drug preparation (extract, with 45 mg/kg THC) administered three times a week for 4 weeks, reduced the tumour volume (HR+ T47D cells) more effectively (to about 180 mm³) compared to the cannabis drug preparation alone or tamoxifen alone (each about 250 mm³) and 45 mg/kg THC (about 400 mm³) vs. vehicle alone (about 500 mm³) [64]. Pure THC was the least effective treatment.

Tamoxifen and several other selective ER modulators (SERM) can act as inverse agonists on CB1 and CB2. Interestingly, AEA, the level of which is increased by CBD, and which is an agonist of CB1 and CB2, can inhibit the proliferation of ER+ MCF-7 and T-47D breast cancer cell lines [75].

As mentioned before, there is actually no clinical evidence evaluating cannabinoids in breast cancer patients. Kenyon et al. [22] claimed that four cases of breast cancer have been successfully treated with low dose CBD. However, the limited amount of data given in the article does not allow a more detailed description and conclusions.

Colorectal cancer

Colorectal cancer (CRC) ranks among the most common cancers. About 4 in 100 subjects will be diagnosed with CRC in their lifetime. CRC varies in its molecular, biological, and clinical features, and in its association with risk factors such as age and physical inactivity. Usually, cancer begins as a polyp, which is a non-cancerous growth that develops in the mucosal layer (inner lining) of the colon or rectum; polyps are common. Despite this high incidence, and numerous in vitro studies, the number of animal studies is still limited. No peer-reviewed publication on the use of cannabinoids in patients with CRC could be found.

CBD, but also CBG, THC, and CBDV exert cytotoxic effects on CRC cells, mostly by apoptosis, reducing the viability of colon cancer cells [3, 76, 77]. CBD induced the inhibition of DNA synthesis and apoptosis of human CRC cells by increasing the expression of B-cell lymphoma 2 (Bcl-2) homology 3 domain-only protein (Noxa) [77].

As in many other cancer cell lines and tissues, AEA as well as the expression of CB1 was found elevated in CRC [78]. AEA demonstrates an apoptosis-promoting effect per se; this may be interpreted as a self-defense mechanism. Promotor hypomethylation of the CB1 receptor gene (CNR1, inhibition of DNA methyltransferase) elevates transcription of the CNR1 gene and expression of CB1. Overexpression of GPR55 and also both CB1 and CB2 was correlated with poor prognosis in stage IV colorectal carcinoma whereby activation of CB2 has been reported to induce cancer cell proliferation [79–81]. As CBD and CBG are TRPM8 antagonists, their inhibiting effect in animal models sheds some additional light on other possible targets in CRCs [31].

The few animal studies with cannabinoids are summarised below (Table 4).

CBD and CBG are so far the only pure cannabinoids investigated in animal models, either to inhibit the formation of ACF, polyps, or tumours, or to inhibit the tumour growth of xenografts. With CBD (1 mg/kg or 5 mg/kg i.p., 3 times per week, over 4 weeks) an inverse dose-effect relationship was observed concerning the prevention of colon ACF by AOM [82], in contrast to a similar experiment using CBG (1 mg/kg or 5 mg/kg i.p., 3 times per week over 4 weeks) [31].

In the xenograft model, CBG inhibited tumour growth dose dependently with 1 mg/kg and 3 mg/kg daily, but an increase to 10 mg/kg did not increase the effect further [31]. In the second xenograft model, a CBD-rich extract (5 mg/kg i.p. per day, 65.9% CBD, 2.4% THC, CBG 1.0%, CBDV 0.9%, 0.3% CBDA) reduced tumour growth significantly on day 4 but no longer on day 7 [83].

Taken together, the evidence for a tumour reducing effect of cannabinoids in CRC is still very limited; experiences in men are completely missing.

Haematological cancers

Haematological cancers are a very heterogenous group of malignancies that begin in cells of the immune system or in blood-forming tissues such as the bone marrow. Among these cancers, diffuse large B-cell lymphoma is the most common haematological malignancy with an annual incidence of 7.9 per 100,000; chronic lymphocytic leukaemia (CLL), which like diffuse large B-cell lymphoma, is also a mature B-cell neoplasm, is next most common [84]. Leukaemia is a cancer of the blood cells and bone marrow, whereas lymphoma starts in the lymph system. Their incidences generally increase with age.

Human leukocytes such as neutrophils, monocytes, and T lymphocytes express CB1 and CB2 receptors; CB2 receptors are most abundant in eosinophils and B-lymphocytes, and unsurprisingly, also in lymphoma and leukaemia cells [85, 86]. They increase inflammation and can be targeted by cannabinoids. Agonists to CB1 and CB2 demonstrate antiproliferative as well as proapoptotic effects inducing cell death of CB1 or CB2 expressing leukaemia cells [87, 88].

Up to now, preclinical research has focussed mainly on solid neoplasms. Animal studies in mice that received murine EL-4 lymphoma cells are summarised below (Table 5).

Experiences in man are anecdotal only. They are summarised below (Table 6).

Neither pure CBD nor pure THC has been studied for treating patients with haematological cancers. Benefits have been reported in only two patients who used THC-Es. The other two articles reporting the use of THC or a combination with CBD had an experimental character. Data are inconclusive at present.

Head and neck squamous cell cancer

In Europe, the incidence of head and neck cancer (HNC), including lip/oral cavity, oesophagus, larynx, oropharynx, hypopharynx, salivary glands, and nasopharynx, is approximately 21.8 per 100,000; prevalence is higher in men. Quite half of the newly diagnosed patients are older than 65 years of age. About 90% of all HNCs are squamous cell carcinoma. Whenever possible, resection and treatment with cytotoxic drugs and radiotherapy is the strategy of choice. Recurrence is however common.

AEA effectively inhibited the proliferation of head and neck squamous cell cancer (HNSCC) cells, most likely via the production of receptor-independent ROS, whereas 2-AG did not [92]. Moreover, it was found that the genes CNR1 and CNR2 are upregulated in HPV+ HNSCC cancers compared with HPV– cancers. High expression of CB2 was significantly associated with reduced disease-specific survival [93]. Activation, (by agonists of CB1 and CB2), promotes in vitro the proliferation of HPV+ HNSCC cells whereas antagonists (e.g., rimonabant) inhibit proliferation [94]. As extracts or cannabis chemotypes differ in their composition, there are likely influences on responses; this may explain conflicting observations about cancers in cannabis users [95–97]. Results in animal models are summarised below (Table 7).

CBD has demonstrated time (48–72 h) and dose (0–15 μmol/L) dependent cytotoxicity against head and neck squamous cancer cells, and reduced significantly tumour volume in two xenograft models. Importantly, the combination treatment of cisplatin with CBD demonstrated a significant synergism [98]. This observation contrasts with THC in a study on HPV+ human head and neck squamous cell carcinoma UD-SCC-2, UPCI:SCC090, and UM-SCC-47 cells where very low concentrations of THC (1 nmol/L to 1 µmol/L) enhanced in vitro the proliferation, with the exception of 93VU147T cells (also HPV+) where 1 µmol/L THC suppressed tumour growth [94]. Enhancement of tumour growth was also seen in vivo in a xenograft model with UD-SCC-2 cells where nude mice received THC (3 mg/kg i.p., daily for 2 weeks). No similar effects are known for CBD.

Treatment experiences in man suffering from HNC are very limited (Table 8).

Kenyon et al. [22] described a further case of oropharyngeal cancer that has been successfully treated with low dose CBD. However, the limited amount of data given in the article does not allow a more detailed description and conclusions.

A recent study investigated the potential relationship between marijuana use and survival outcomes of HPV-related patients with proven p16-positive oropharynx squamous cell carcinoma [96]. Marijuana users were identified from a prospectively collected database of HNC patients. They were then case-matched on a 1-to-1 basis to patients who were non-marijuana users based on age, gender, and cTNM staging. No statistically significant difference between marijuana and non-marijuana users was found in terms of overall survival, disease-specific, disease-free, and metastasis-free survival after five years in contrast to much earlier reports [96, 100].

From that, it may be supposed that HNC differ in their molecular properties, and respond therefore differently to specific cannabinoids. As extracts or cannabis chemotypes vary considerably in their composition it is likely that this influences also responses.

Hepatocellular carcinoma

The incidence of hepatocellular carcinoma (HCC) varies widely between about 18 and 94 per 100,000, depending on risk factors [99]. After glioblastoma and pancreatic cancer, liver cancer is the most lethal tumour with an estimated 5-year survival of only 18% [101, 102].

As with other cancers, the ECS is disturbed also in HCC; levels of AEA are reduced [103]. A higher expression of CB1 and CB2 correlated with improved prognosis [104].

Animal models of HCC demonstrate that cannabinoids effectively reduce tumour growth (Table 9).

Until now, no report on patients with HCC treated with cannabinoids has been found. However, the University Medical Centre Groningen (UMCG), Netherlands, is to study the effect of cannabis oil on liver cancer patients based on two separate reports that patients with advanced liver cancer had seen their tumours shrink after using cannabis oil. Now, two and five years after their diagnoses, the tumours have completely disappeared and the patients are considered cured (https://bedrocan.com/umcg-studies-cannabis-oil-for-liver-cancer-patients-with-no-further-treatment-options/).

Lung

Lung cancer is the 2nd most common cancer worldwide (after breast cancer), the most common in men, and the 2nd most common in women with an age-standardized incidence rate (ASIR) of 22.4 per 100,000 (male: 31.5; female: 14.6). Lung cancers are classified as small cell lung cancer (SCLC, 13%) and non-SCLC (NSCLC, 84%). A major problem with lung cancers is that they are usually diagnosed in advanced stages when they have already spread to other organs; moreover, cancer cells easily develop resistance to common antitumour agents.

A number of in vitro studies with lung cancer cell lines have demonstrated that cannabinoids inhibit cell viability, inducing apoptosis, whereby the activation of CB1, CB2, and TRPV1 receptors plays a role. Mice lacking the CB2 receptor (CB2-knockout, CB2^–/– mice) or receiving a CB2 antagonist demonstrated a lower tumour burden than wild-type mice [107, 108]. It seems that CB2 plays a pro-oncogenic role in NSCLC; it is up-regulated in NSCLC tissues and up-regulation correlates with tumour size and pathological grading [109]. In contrast, high expression of the CB2 receptor gene was found to correlate with improved survival in lung cancer patients [110].

CBD has been shown to decrease dose-dependently the viability of various lung cancer cells, NSCLC cell lines (A549 and H1299) as well as SCLC cells (H69) [111]. CBD is also able to potentiate the effect of THC in vitro [112].

In animal models using nude mice, CBD (5 mg/kg i.p.) significantly reduced tumour size and lung metastatic nodules (from an average of 6 nodules to only 1 nodule) in an A549 xenograft tumour model [113, 114]. In another experiment with A549 xenografts in nude mice, either 15 mg/kg THC s.c., or 20 mg/kg CBN, or 40 mg/kg CBN per day was administered for 20 days. Tumour volume changes were significantly lower with THC (251%) or 40 mg/kg CBN (266%) than those of control mice (716%), but not significant with 20 mg/kg CBN (345%). The effect of CBN was thus dose-dependent [115]. Intriguingly, THC (5 mg/kg four times per week) increased considerably the volume of two different tumour xenografts in two weakly immunogenic murine lung cancer models [108]. Effects of cannabinoids seem to depend on the immune status. Animal lung cancer studies are summarised below (Table 10).

Taken together, animal models demonstrate that CBD (5 mg/kg) reduces lung cancer growth (NSCLC) in mice; THC and CBN are also effective.

At present, treatment in men is limited to a few case reports summarised below (Table 11).

An 81-year-old man with biopsy-confirmed adenocarcinoma of the lung was put on a regimen with 2% “CBD oil” (extract with a dose of 1.32 mg CBD twice daily) 11 months after diagnosis. The tumour was progressive at that time. CBD oil, which was the sole therapy, was increased to twice 6 mg per day after one week. CT-imaging four months later revealed near total resolution of the left lower lobe mass and a significant reduction in size and number of mediastinal lymph nodes (stable according to a CT control two months later) [119].

Another article [120] reports the case of a subject with terminal, also biopsy-confirmed lung cancer. The patient, a 53-year-old man, had a history of intense alcohol and drug abuse and repeated injuries to his spine after multiple car accidents. He suffered from very severe pain, insomnia, post-traumatic stress disorder (PTSD), anxiety, and depression, with a loss of bladder control in parallel. In a last attempt to improve his life, the patient joined Alcoholics Anonymous where one of his fellow members advised him to inhale vaporized cannabis oil. To his surprise, he was not only able to stop substance abuse but also his lung cancer disappeared within about three months of inhaling vaporized cannabis oils (composition unknown) on a daily basis. He died from cardiac failure about a year later [120].

A third article [121] describes the case of a woman in her 80s who was diagnosed with NSCLC. At diagnosis, her tumour was 41 mm in size, with no evidence of local or further spread, so was suitable for conventional treatment of surgery, chemotherapy, and radiotherapy. As the patient refused standard cancer treatment, she was placed under “watch and wait” monitoring, which included regular CT scans every 3–6 months. She was a smoker, getting through around a pack plus of cigarettes every week (68 packs/year). She also had mild chronic obstructive pulmonary disease (COPD), osteoarthritis, and high blood pressure, for which she was taking various drugs. Surprisingly, scans showed that the tumour was progressively shrinking, reducing in size from 41 mm to 10 mm 32 months later, equal to an overall 76% reduction in maximum diameter, averaging 2.4% a month. Discussions with the physicians revealed that the patient had taken 0.5 mL of “cannabis oil” two to three times daily since her diagnosis (20% CBD, 19.5% THC, 24% THCA, according to the supplier), i.e., approximately 200–300 mg of CBD and THC per day. The woman said she had reduced appetite since taking the oil but had no other obvious “side effects”. There were no other changes to her prescribed medications, diet, or lifestyle, and she continued to smoke throughout [121].

Taken together, these case reports suggest tumour-reducing effects of cannabis extracts in lung cancer patients, although neither the exact composition of extracts nor the exact daily dose is entirely clear.

Ovarian cancer

Data on ovarian cancer and cannabinoids is very limited; CB1 was overexpressed and correlated with disease severity in epithelial ovarian carcinoma [122]. CBD inhibited endometrial (ECC1) and epithelial ovarian cancer (Kuramochi) cell lines, with an inhibitory concentration (IC₅₀) between 2.5 and 20 nmol/L [123].

Only one article has been found, concerning the treatment of ovarian cancer with cannabinoids in an 81-year-old woman with metastatic, low grade, ovarian carcinoma, accidentally diagnosed during surgery (Ca-125 value 77 U/mL, resected tissue oestrogen and PR+) [124]. A CT scan of the chest, abdomen, and pelvis demonstrated multiple mesenteric soft tissue masses ranging from 7 mm to 7 cm and omental carcinomatosis. A 5.8 cm solid right adnexal mass and 3.3 cm solid left adnexal mass were also identified together with lymphadenopathy along the left common iliac vessels and the left pelvic sidewall. Chemotherapy was proposed but declined. At this time, two months after surgery, the Ca-125 value had dropped to borderline 46 U/mL. Laetrile tablets (500 mg orally four times per day) and CBD oil (1 drop sublingually each evening) were started. Assuming a volume of at least 35 µL per drop and a concentration of at least 10% CBD, 1 drop contained at least 3.15 mg CBD. One month later, the Ca-125 value had dropped to 22 U/mL. This treatment was maintained. Repeated CT imaging in the following months showed a dramatic reduction in her disease burden, with near-complete resolution of all previously identified lesions seven months after surgery. Ca-125 values remained low around 12 U/mL. The authors related the dramatic decrease to the intake of CBD [124]. However, it should be noted that the Ca-125 value had already dropped from 77 at the time of the surgical intervention to 46 two months later when CBD was started; the resolution of lesions may not be due entirely to CBD.

Pancreatic cancer and pancreatic ductal adenocarcinoma

Pancreatic cancer is, like glioblastoma, an orphan disease with an incidence of around 4 to 13 per 100,000 (National Cancer Institute, US); it ranks among the most malignant forms of cancer. Unfortunately, in most cases, the cancer is diagnosed late, usually when it has already spread. Its incidence varies considerably and may be underestimated as it increases with age [125, 126]. Survival time is rather short with a mean of around 4 to 9 months after diagnosis, and correlates negatively with age [127, 128]. Younger age at diagnosis and a resectable tumour has a better prognosis. The 5-year survival rate is about 5 to 12%. Pancreatic cancer is commonly resistant to most of the available chemotherapeutic drugs.

As with other cancers, the ECS participates in the defense against cancer growth. It has been observed that 2-AG suppresses pancreatic cancer cell proliferation and tumour growth in vitro and in vivo [129]; this 2-AG-induced antiproliferative effect is CB1-receptor dependent. Another receptor that obviously plays a role is GPR55 which is increased in human pancreatic ductal adenocarcinoma (PDAC) specimens [130].

Only few animal experiments with natural cannabinoids have been published. They are summarised below (Table 12 below).

In a mouse model of PDAC pharmacological blockade of GPR55 with CBD, GEM (a standard treatment), and CBD plus GEM increased the rodent lifespan compared to vehicle (mean survival 25.4 days, 27.8 days, 52.7 days, and 18.6 days respectively), with many of the signalling pathways involved in reducing PDAC cell cycle progression and cell growth identified [130]. Most interestingly, whereas pure THC, pure CBD, and pure GEM have demonstrated a benefit in terms of reduced tumour growth or longer survival, a study using an extract containing both cannabinoids, THC and CBD in a ratio of 1:6 did not demonstrate a significant impact on the tumour volume; even worth, the highest dose of 10 mg/kg extract showed a more pronounced tumour growth than that of the negative control [27] Overall, the best effect was achieved with a combination of high CBD and GEM.

Experiences in humans with phytocannabinoids are rare. Only one small case series was found that included a total of nine consecutive patients with pancreatic cancer who received CBD, most of them as comedication to standard chemotherapy; two patients have been treated only with CBD [132]. CBD was usually administered in an oral daily dose of 400 mg. Five patients received also low dose THC for improving appetite. The mean overall survival of these nine patients was 11.5 months (median 11 months), thus about twice as long as expected from historic data (Table 13).

Prostate cancer

Prostate cancer is the second most frequent carcinoma in males after lung cancer. The incidence increases dramatically with age and is 1% to 2% after the age of 75 years [133]. Prostate cancer presents as two subtypes, androgen-sensitive (androgen-dependent) and androgen-independent prostate cancer. Androgen deprivation therapy (ADT) is the treatment of choice for the palliation of men with androgen-sensitive disease. Androgen-independent prostate cancer is observed when cancer advances despite primary hormone therapy. Most of the patients initially responding to ADT will eventually develop castrate resistance. It seems that a protein called TRAF4 is responsible for this conversion of androgen-sensitive prostate cancer cells into castration-resistant cells; TRAF4 is frequently overexpressed in advanced prostate cancers [134]. Most intriguingly, a very recent study demonstrated in silico that theoretically CBD, its acid CBDA, and THC could act as androgen receptor (AR) inhibitors whereas CBDV and cannabinodiol (a natural derivative of CBN produced by photochemical conversion) could act as 5α-reductase inhibitors among other minor phytocannabinoids [135]. 5α-reductase inhibitors such as finasteride have been known for a long to reduce serum prostate-specific antigen (PSA) levels. Therefore, the observation that marijuana use was inversely associated with PSA levels deserves further investigation [136].

A study showed higher expression of CB1 receptors in androgen-sensitive as well as androgen-independent prostate cancer cell lines compared to normal prostate epithelial cells; the over-expression of CB1 and TRPV1 correlates with increased grades of prostate tumours [137]. Basically, prostate cancer cell lines used in preclinical studies can be divided into AR+ (LNCaP and 22RV1) and AR– (DU-145 and PC-3). Intriguingly, receptors are expressed differently in specific cancer cell lines: 22RV1 which is AR+, only expresses CB1 while DU-145 (AR–) only expresses CB2. Though CB1 and CB2 can be found in both LNCaP and PC-3, their levels are much more prominent in PC-3. TRPV1 is expressed in all four prostate cancer cell lines, with the highest expression found in DU-145 cells, whereas TRPM8 receptor has been found in LNCaP cells [137, 138]. In vitro, CBD can inhibit the expression of the AR in AR+ cell lines [139]; CBD is also a TRPM8 antagonist. In vivo, results are actually restricted to only two studies that are summarised below (Table 14).

The effect of CBD (150 mg/kg per day), cisplatin (50 mg/kg per day), Cannabis sativa extract (200 mg/kg per day, “complete cannabis extract”, composition unknown) or vehicle, injected peritoumorally for 10 days in combination with siRBBp6 gene therapy (RBBp6 antibodies), was studied in mice (BALB/c nude mice, s.c. PC3-xenograft). The protein RBBp6 is implicated in cell cycle regulation and is supposed to promote tumour genesis. Tumour sizes were reduced with both, CBD and cisplatin, by approximately 90% and significantly more than by the cannabis extract [140].

CBD potentially downregulates the expression of PSA, VEGF, and pro-inflammatory cytokines [138]. In the only study that analysed the effects of CBD on PSA in 18 patients with biopsy-proven adenocarcinoma of the prostate after localized therapy, who received up to 800 mg CBD once daily for 90 days, the large majority (88%) had the stable biochemical disease, i.e., an increase in baseline PSA of < 25% [141].

Kenyon et al. [22] described very briefly one case of prostate cancer that has been successfully treated with low dose CBD. However, the limited amount of data given in the article does not allow a more detailed description and conclusions.

The effect of cannabis/cannabinoids on other urogenital cancers is still a matter of debate. A recent study found that previous use of cannabis was a significant protective factor in women against renal cell carcinoma and bladder cancer. In men with a history of tobacco smoking, previous cannabis use was a significant protective factor for prostate cancer. A history of cannabis use had a null effect on rates of testicular cancer [142].

To sum it up, cannabinoids possibly have effects on prostate cancer, or may even reduce it, however, the amount of data is at present insufficient for any conclusions.

Skin cancer and melanoma

Skin cancer is a common disorder. In addition to melanoma, there are two other major types of skin cancer, basal cell carcinoma and squamous cell carcinoma; relatively rare types are Kaposi sarcoma or cutaneous T-cell lymphoma.

Melanoma is a highly heterogenous and frequent type of tumour. The incidence varies widely between about 100 per 100,000 (coloured people) and 600 per 100,000 (Caucasians). Melanoma ranks among the cancers with the highest mutational burden that implicates also a high probability of primary resistance to any pharmacological therapy. Mutations are favoured due to exposure to mutagenic ultraviolet radiation (UV).

In vitro studies on various melanoma cell lines suggest that cannabinoids target CB1, TRPV1, and PPAR-α receptors, and induce apoptosis in a dose-dependent way. CBD and the combination of CBD + THC can reduce cell viability in different melanoma cell lines in vitro. The inhibitory effect of CBD plus THC was stronger than that of either drug alone [143, 144]. The reduction of the viability of melanoma cell lines seems to be mediated mainly by CB1, TRPV1, and PPAR-α receptors. Some melanoma cell lines such as A375 melanoma cells express high levels of FAAH, COX-2, CB1, TRPV1, and GPR55 genes [145]. Unsurprisingly, AEA cytotoxicity was potentiated by FAAH inhibition.

The few animal studies that exist are summarised below (Table 15).

Although the number of experiments is very limited, in vivo models demonstrate that pure, single cannabinoids, CBD as well as THC in a dosage between 5 mg/kg and 15 mg/kg b.w. have a strong antitumour effect that can be increased by combinations. The MEKi showed the strongest growth inhibition [144].

No publication on the treatment of melanoma in men has been found so far.

Basal cell carcinoma is the most common skin cancer in the world but has a very low mortality rate. For that, it is often not included in cancer statistics. Similar to squamous cell carcinomas, it is a descent from epidermal keratinocytes. Only two case reports have been found; they are summarised below (Table 16).