A total of 15 articles corresponding to the same number of studies met our research criteria: 9 randomized double-blind placebo-controlled crossover studies, 3 randomized placebo-controlled trials, 1 prospective non-randomized single-arm clinical trial, 1 prospective cohort study with one year follow-up, and 1 single-dose open-label study (Figure 1). Their principal characteristics are reported in Table 2.

In the Netherlands, an experimental randomized placebo-controlled four-way crossover trial recruited 20 patients with chronic fibromyalgia pain and analyzed the analgesic effects of inhaled therapeutic cannabis. In the study, 100 mg each of three cannabis strains—Bedrocan (22.4 mg THC, 1 mg CBD), Bediol (13.4 mg THC, 17.8 mg CBD), and Bedrolite (18.4 mg CBD, 1 mg THC)—and placebo were given in a single day. There was no difference between the effects of active treatment and placebo on spontaneous pain evoked by electrical stimuli. Indeed, Bedrocan and Bediol caused a significant increase in the tolerance to pressure pain threshold. The most relevant effect was observed for the cannabis strain that contained high doses of THC and CBD (Bediol). When CBD was given with a very small dose of THC (Bedrolite, which mainly contains CBD), the analgesic effects were not superior to placebo. This result differs from those of studies in which patients with chronic pain reported beneficial effects with CBD treatment, probably related to improved anxiety, insomnia, and mood. Perhaps a single dose was insufficient to determine the analgesic effect or the dose was too low. With reference to adverse events, all three active treatments were associated with AEs related to the inhalation of cannabis, and the most common were drug high, dizziness, and nausea. There were no differences in the frequency of adverse events [17].

In Italy in 2018, a prospective non-randomized single-arm clinical trial analyzed data from 338 patients with fibromyalgia, headache, radiculopathy, and various forms of neuropathic pain. They received a daily dose of 5–40 mg/day of THC (many of the participants required 10 mg/day) as a decoction, corresponding to 28–210 mg of cannabis flos with 19% THC and 1% CBD for 12 months, and the intensity of pain was evaluated at follow-up visits. Among the patients, 33 stopped the study due to AEs, possibly due to the high percentage of THC in Bedrocan (19%), and 77 patients the study due to little benefit. The appearance of AEs was greater in the cannabis treatment groups than in the control, but they were transitory because they regressed after being interrupted. The most frequent were drowsiness and mental confusion, and other non-serious adverse events after termination of the study were worsening tachycardia, itching, diarrhea, gastralgia, nausea and vomiting, anal burning, increased depression, increased appetite, hallucinations, and muscle weakness. This study showed that treatment with cannabis is effective in reducing the intensity of pain, as measured by a visual analogue scale (VAS), and the disability caused by chronic pain as well as the resulting anxiety and depression, via the Hospital Anxiety and Depression Scale (HADS), without generating severe adverse events [18].

The safety profile of cannabis in the management of chronic non-cancerous pain was studied in a prospective cohort study, the Compass study. A 12.5% THC cannabis extract (CE) was dispensed with cannabis smoke to 215 subjects, in most cases for one year, with 216 nonsmoking subjects considered as a control group. The main outcome was the occurrence of serious adverse events (SAEs) and non-SAEs. The daily median dose was 2.5 g/d, and the recommended maximum limit was 5 g/d. No difference in the risk of SAEs between the two groups was detected, but in the cannabis group there was an increased risk of mild to moderate reactions, especially related to the nervous and psychiatric systems. The cannabis group was composed of 66% current cannabis users, 27% former cannabis users, and 6% cannabis naive users. The control group included 32% former cannabis users and 68% naive cannabis users. The most common SAEs in the cannabis group were abdominal pain, intestinal obstruction, and nephrolithiasis. However, none of the SAEs were definitely or likely related to cannabis. Two patients stopped the study due to SAEs, one with seizures considered possibly related to cannabis use and one for alcohol problems. Headache, nasopharyngitis, nausea, drowsiness, and dizziness were the most common non-SAEs reported. The cannabis group had also a higher rate of developing mild respiratory adverse events during the 12 months than the control group. In the study, the authors pointed out that cannabis users had an average decrease in FEV1 of 50 mL and an average reduction of 1% of the FEV1/FVC ratio after one year. As a secondary outcome, there was significant improvement in pain intensity and quality of life after one year for the cannabis group compared to control. In conclusion, the results of this study suggest that the adverse events with cannabis for medical use are modest and that an average dose of 2.5 g/d can be included in pain management programs safely with careful monitoring if conventional treatments have been considered inappropriate or inadequate [19].

The efficacy and tolerability of inhaled cannabis was investigated in a short-term randomized double-blind control study conducted on 16 patients with pain caused by diabetic neuropathy. Each patient was exposed to four single doses of aerosolized placebo or a low dose (1% THC), medium dose (4% THC), or high dose (7% THC) of cannabis with 1% CBD. Baseline values were collected for basic spontaneous pain, evoked pain, and cognitive tests. Pain intensity and cognitive capacity were measured for 4 h. The weight of 400 mg of plant material for administration corresponded to 0, 4, 16, or 28 mg THC per dosing session. Each participant received a placebo or a cannabis dose with 1, 4, or 7% THC with an interval of 2 weeks between doses. Adverse events were feelings of euphoria and drowsiness, significantly relevant at high and medium doses compared to placebo. This study found a dose-dependent reduction in the intensity of spontaneous and evoked pain in response to cannabis inhalation in patients with diabetic neuropathy. There were significant differences in the levels of spontaneous pain between placebo and active doses (low, medium, and high doses) and between the high dose and the other active doses. There was also impaired performance on neuropsychological tests with the high dose [20].

An Israeli study in 2014 aimed to examine the pharmacokinetics, safety, tolerability, efficacy, and ease of use of a new portable thermal dose inhaler (tMDI) for cannabis by analyzing a cohort of eight patients suffering from neuropathic pain who were on a stable analgesic regime that included medicinal cannabis. Four had complex regional pain syndrome (CRPS), two had lumbosacral radiculopathy, one had pelvic neuropathic pain, and one had pain related to spinal cord injury. Patients received a single dose of 15.1 ± 0.1 mg of cannabis through the inhaler device. A blood sample was taken to evaluate delta-9-THC and 11-hydroxy-9 THC at baseline and 120 min after inhalation of cannabis. The drug used was cannabis flos (Bedrocan) containing 19.9% THC, 0.1% CBD, and 0.2% CBN. All patients were treated with inhaled cannabis by smoking 2 or 3 times a day. The monthly median dose used was 20–30 mg. In this study, the low dose of THC produced an analgesic effect on the various conditions that caused neuropathic pain. A single inhalation containing 3.08 ± 0.02 mg of THC raised the plasma level of delta-9-THC Cmax to 38 ± 10 ng/mL and provided a 45% reduction in pain intensity [21].

In a study published in 2016 in the USA, 42 participants with neuropathic pain caused by injury or spinal cord disease were recruited, and the analgesic efficacy of 400 mg cannabis administered via the Volcano vaporizer was assessed using placebo and doses of 2.9% or 6.5% THC (11.6 mg and 26.8 mg THC, respectively). The study was carried out in three 8 h sessions with a median interval of about 7 days between sessions. The patients’ pathologies included multiple sclerosis, cervical disc pathology, spinal cord cancer, occlusion of the vertebral artery, arachnoid cysts, and syringomyelia. Among the patients, 90% had previously used cannabis.

During the session, participants inhaled 4 puffs of cannabis or placebo, and 240 min later were asked to choose a dose ranging from 4 to 8 puffs. Prior to placebo administration, there was no significant difference in the pain rate between placebo and 2.9% and 6.7% THC doses. Then there was a significant dose effect on pain intensity. The post hoc Tukey test showed a step-by-step effect with the highest pain intensity in the placebo group and the lowest in the higher THC group. Recent cannabis did not affect the results. One hour after the first treatment dose and one hour after the variable phase, both active doses were associated with significantly lower pain than placebo. Pain relief persisted for another 2 h from the variable dose, but the impact on pain showed no distinction between upper and lower doses of THC. Both active doses did not affect allodynia, which is consistent with the lack of benefit of cannabinoid treatment in postoperative pain. Many of the psychoactive side effects were concentration-dependent. The highest THC doses have been associated with significantly higher levels of “desires”, appetite, difficulty remembering things, drunkenness, and confusion [22].

A phase III multicenter clinical trial was designed to investigate a standardized oral CE used for the symptomatic relief of muscle stiffness and pain in 277 adult patients with stable MS treated with CE or placebo twice daily for 12 weeks (2 weeks titration phase, 10 weeks maintenance phase). The active treatment was an extract from Cannabis sativa in soft gelatin capsules containing 2.5 mg THC and 0.8–1.8 mg CBD. Subsequent doses were individually titrated upward by 5 mg THC/day every 3 days up to 12 days to optimize the ratio between therapeutic effect and side effects, and to achieve a maximum daily dose of 25 mg THC. The primary outcome was based on an 11-point category rating score (CRS) measuring perceived changes in muscle stiffness and perceived relief from body pain, muscle spasms, and sleep disturbance as a secondary outcome.

Patients treated with cannabis were divided into groups based on the severity of their symptoms of muscle stiffness and pain (low vs. high) and use of drugs (yes or no) to fight muscle spasms and pain. The level of relief, as reported by patients, was higher in the group treated with cannabis, detected after 4, 8, and 12 weeks, with the greatest difference observed in those not using antispastic drugs (37.9% cannabis vs. 16.3% placebo). When the titration period was completed, 87% of patients in the placebo group and 47% in the cannabis group took the maximum dose of 25 mg every day, and the percentages were lower at the end of the treatment period (24.5% vs. 69.4%). In the cannabis group, 33 subjects (21%) and 9 patients (6.7%) were withdrawn from or suspended treatment because of AEs. About 95% of AEs detected in all treatment groups were mild or moderate and transitory, and were not present at the end of the treatment period. SAEs were reported by 7 patients (3 patients reported urinary infections) in the cannabis group (4.9%) and 3 patients in the placebo group (2.2%). The rate of AEs was higher during the titration period in the cannabis group (75.5%). The most common AEs were related to the nervous system (71.3%) and gastrointestinal system (41.3%). AEs were more frequent in cannabis patients compared to the placebo group; they included vertigo, attention disorder, equilibrium disorder, drowsiness, dry mouth, nausea, fatigue, weakness, diarrhea, urinary infection, confusion, and falls [23].

In another publication, after randomization to placebo or smoked cannabis (4% THC), 30 participants with MS were evaluated through 8 visits over a period of 2 weeks. Patients were asked to smoke once daily for 3 days, with an 11-day washout period between treatments. Each dose was an average of 4 puffs per cigarette. The sample was composed of 63% women with an average age of 50 years old; 70% of the participants were undergoing disease-modifying therapy, and 60% were taking antispasticity agents. Most of the participants (80%) had previous recreational experience with cannabis, and 33% had used cannabis within the previous year. Those who smoked cannabis had reduced patient scores for spasticity using the modified Ashworth scale by an average of 2.74 points and on VAS by 5.28 points more than placebo. It is worth mentioning that in this study, participants began with relatively low levels of pain. Smoking cannabis did increase patient perception of “highness” by 5.04 points more than placebo. Five patients withdrew from treatment due to adverse events: two patients felt uncomfortably “high”, two had dizziness, and one had fatigue [24].

In a double-blind placebo-controlled crossover study of 39 patients with neuropathic pain, inhalation of 10.3 mg of vaporized THC, divided into 2 sessions and separated by a 2 h interval, was associated with a 31 and 25% reduction in pain intensity at 3 and 5 h, respectively. Increasing the THC dose to 28.2 mg produced an equianalgesic response that remained stable when monitored at the same time intervals (3 and 5 h). The AEs were minimal, reversible, and well tolerated. Seven patients felt a sensation of light-headedness in the first minutes after inhalation, which regressed rapidly [25].

Doses of 0, 2.5, 6, and 9.4% THC were used in a population of patients with different types of neuropathic pain. Compared to placebo, a single inhalation with a low dose, 25 ± 1 mg cannabis containing 9.4% THC, administered 3 times a day for 5 days, was associated with an average Cmax of 45 ng/mL and a decrease of 11.4% in average daily pain intensity. Adverse events were THC concentration-dependent [26].

The effects of smoked cannabis were studied in patients affected by HIV-associated distal sensory predominant polyneuropathy (DSPN) and pain resistant to other analgesic drugs in a placebo-controlled double-blind crossover trial. Participants were treated with cannabis containing 1–8% THC or placebo 4 times a day for 5 consecutive days over a period of 2 weeks. After 2 weeks of washout, each group received the other treatment. Changes in pain intensity were evaluated together with possible modifications in mood and daily functioning. Cannabis produced the greatest pain relief in comparison to placebo. Both treatments ameliorated mood and daily functioning. Two participants were withdrawn from the study due to safety issues. In particular, one subject, naïve to cannabis, showed an acute psychosis with smoked cannabis. Another participant reported a severe but transitory cough caused by cannabis. Other minor AEs were a transitory increased cardiac rate, concentration deficit, drowsiness, fatigue, more prolonged sleep duration, sense of thirst, and reduced salivation [27].

In another placebo-controlled double-blind crossover study, 38 patients with central or peripheral neuropathic pain syndrome were asked to smoke cannabis with 7% THC or 3.5% THC or a placebo during three 6 h experimental sessions. The cumulative dose at each session was 9 puffs (2 puffs after 1 h, 3 puffs after 2 h, and 4 puffs after 3 h). The primary endpoint was based on VAS pain intensity before and after smoking marijuana. Cannabis produced an analgesic effect with cumulative dosing that began to reverse within 1–2 h after the last dose. The doses of 3.5 and 7% THC were equianalgesic at every time point, with no differences between the two over time. No significant differences in outcome were observed between the different pain conditions. “Feeling high” and “feeling stoned” were more prevalent in the active treatment groups and highest in the 7% THC group. “Feeling impaired” was higher in both treatment groups, but no significant difference was found between the higher and lower THC doses. Feelings of confusion, sedation, and hunger were also higher in the two active treatment groups. No mood changes were observed. A general cognitive decline was evident in both treatment groups, with greater cognitive impairment in the high-dose group [28].

In another study, 50 patients with painful HIV-associated sensory neuropathy were assigned to smoke either cannabis with 3.57% THC or placebo cigarettes 3 times daily for 5 days. Those who smoked cannabis had reduced daily pain by 34% compared with 17% in the placebo group, and 52% of patients in the cannabis group reported >30% pain reduction compared to 24% of patients in the placebo group. No serious AEs were reported in this study. Although there were few minor AEs, side effect ratings were higher in the cannabis group than the placebo group for anxiety, sedation, disorientation, confusion, and dizziness [29].

In the Cannabinoids in Multiple Sclerosis (CAMS) study, a clinical trial of 15 weeks that recruited 630 participants, 211 participants received an oral CE, 211 received THC, and 206 received placebo. The primary outcome was the effect of cannabis on the degree of spasticity as measured by the Ashworth scale. The active pharmacological treatment was a synthetic THC capsule (Marinol) and a cannabis extract containing 2.5 mg THC, 1.25 mg cannabidiol, and <5% other cannabinoids per capsule. The dose administered was based on a scheduled titration of 5 weeks up to a maximum possible dose of 25 mg daily. In those patients with spasticity alone, cannabinoid treatment was not helpful in achieving an improved Ashworth score. However, patients who presented spasticity and pain together experienced an improvement in spasticity by 61% (n = 121), 60% (n = 108), and 46% (n = 91) among those who received CE, synthetic THC, and placebo, respectively. This was regarded as a subjective rather than objective clinical effect. There were reports of SAEs in all groups, but they were more frequent in the placebo group. There were also minor AEs reported: frequent episodes of dizziness, light-headedness, or dry mouth in the active groups. There were some differences between groups in gastrointestinal side effects: constipation was more frequent in the cannabis extract group, and diarrhea was reported more in the active groups and not in the placebo group. Increased appetite was also a side effect in treatment groups, although with low frequency: four in the CE group, six in the THC group, and one in the placebo group [30].

A randomized double-blind placebo-controlled twofold crossover study evaluating the safety, tolerability, and efficacy of synthetic oral THC and Cannabis sativa plant extract was conducted with 16 patients with MS and severe spasticity (10 with secondary progressive MS and 6 with primary progressive MS). Each patient received the following 3 treatments for 4 weeks: synthetic THC (Dronabinol), Cannabis sativa extract, and placebo. During the first 2 weeks, study medications (THC and extract) were administered in twice daily doses of 2.5 mg THC or plant extract containing the same level of THC. If well tolerated, the dose was elevated to 5 mg twice a day for the next 2 weeks. There was a 4 week washout period between treatments. The primary outcome was a change in VAS score for pain. All patients completed the study. Six patients had used cannabis before, none on a regular basis. Both THC and plant-extract capsules were well tolerated. No SAEs were reported. AEs were more common during plant extract treatment. Five patients reported increased spasticity during plant extract treatment. One AE was rated as severe acute psychosis lasting for 5 h after the scheduled dose increase of plant extract. No clinically relevant changes were observed on physical examination or in hematology or chemistry measurements. Because of the limited sample size, no definite conclusions were reached, but the results of this study suggested there was no therapeutic benefit with either THC or plant extract treatment. The route of administration was cited as a possible explanation for the lack of efficacy. THC is absorbed reasonably well from the gut, but the process is slow, with large variations between and within individuals. A second possible explanation could be the prescribed dose. However, even at this dose, the number of AEs, especially during plant extract treatment, was rather high, suggesting that a higher dose might not be well tolerated [31].