Pain represents the most common and debilitating consequence of spinal cord injuries (SCIs) and leads to poor quality of life (Burchiel and Hsu, 2001). Around the world, around 250,000 to 500,000 people suffer from SCI per year, and the annual global incidence is approximately 40–80 cases per million (Bickenbach et al., 2013). The lifetime costs per patient with SCI can reach $1.1–4.6 million (Ahuja et al., 2017). The treatment strategies for SCI patients may include control of blood pressure, use of corticosteroids, spinal immobilization, surgical intervention, anticoagulation prophylaxis, and sufficient pain management (Burchiel and Hsu, 2001; Wilson et al., 2013; Ahuja et al., 2017; Fehlings et al., 2017). Successful pain control in these patients could improve clinical outcomes, reduce hospital stay, lower the medical costs, and increase the quality of life (Burchiel and Hsu, 2001). Several multimodal analgesic agents, including opioids, gabapentinoids, a-adrenergic antagonists, antidepressants, and anticonvulsants, are widely used for pain management in SCI patients, but the overall effectiveness is suboptimal (Shiao and Lee-Kubli, 2018).

Cannabinoids, the active herbal compounds in cannabis, including tetrahydrocannabinol, dimethylheptylpyran, and parahexyl, have been present in Central Asia dating back 12,000 years (Andre et al., 2016; Crocq, 2020). Since then, the medicinal use of cannabis has been recorded in China, Egypt, Greece, and the Roman Empire. In China, cannabinoids were prescribed for anesthetic use from 221 B.C. to A.D. 220 (Crocq, 2020). In 1964, Raphael Mechoulam and Yechiel Gaoni identified THC in the cannabis sativa plant (Gaoni and Mechoulam, 1964). Subsequently, in 1967, Mechoulam R., Braun P. and Gaoni Y. synthesized THC (Mechoulam et al., 1967). Smoking or oral ingestion of cannabinoids may produce analgesic, antianxiety, antispasmodic, muscle relaxant, anti-inflammatory, and anticonvulsant effects (Andre et al., 2016). This medication may help SCI patients with pain reduction since cannabinoids may cause an antinociceptive effect by activating TRPA1, TRPV1, TRPV2, TRPV4, and G-protein–coupled receptors (Hill, 2015; Mlost et al., 2020). Currently, scientists are investigating the potential use of cannabinoids for pain reduction. However, cannabinoids are not risk-free. For example, in animal studies, the adverse effects or toxicity included neurotoxicity, hepatocellular injuries, developmental toxicity, embryo–fetal mortality, spermatogenesis reduction, organ weight alterations, male reproductive system alterations, and hypotension (Huestis et al., 2019). Another previous review study also reported the adverse effects of cannabinoids, including diarrhea, hepatic abnormalities, fatigue, vomiting, and somnolence, in humans (Huestis et al., 2019). Some studies have demonstrated that cannabinoids are effective for chronic pain, neuropathic pain, and spasticity, but the efficacy and safety of cannabinoids use in SCI patients have not been systematically evaluated. Therefore, in this study, we surveyed the existing literature to estimate the degree of pain relief and adverse events derived from cannabinoid use in SCI patients.

Our study did not demonstrate a better analgesic effect of cannabinoids than that of placebo in SCI patients. In fact, the adverse events were higher in the cannabinoid group than in the placebo group. The majority of these side effects involved the nervous system, such as dizziness, somnolence, and dysgeusia. To the best of our knowledge, this may be the first systematic review and meta-analysis to compare the analgesic effects and adverse events of cannabinoids in SCI patients.

The SCI-induced pain includes nociceptive (musculoskeletal or visceral), neuropathic (at level or below level), and other types of pain (Bryce et al., 2012). According to the International Association for the Study of Pain, neuropathic pain is defined as “pain initiated or caused by a primary lesion, dysfunction, or transitory perturbation of the peripheral or central nervous system” (Classification of chronic pain, 1986). Neuropathic pain can be classified into peripheral and central types. Central neuropathic pain occurs after spinal cord injury, stroke, and multiple sclerosis. Peripheral neuropathic pain commonly occurs in conjunction with diseases resulting in peripheral nerve damage such as cancer and diabetes. SCI causes both central and peripheral neuropathic pain and is usually challenging for clinicians to manage (Canavero and Bonicalzi, 2011). Canavero and Boncalzi reviewed the efficacy of cannabinoids and other medications for central neuropathic pain (Canavero and Bonicalzi, 2018). They concluded that cannabinoids were not better than the other commonly used drugs and do not support a role in the management of central pain. The common medications including opioids, gabapentinoids, and NSAIDS were also not beneficial for the treatment of central pain. Even for multiple sclerosis, which is considered a subtype of cord central pain, the in-depth review by Canavero and Bonicalzi did not find evidence for a major effect (Canavero and Bonicalzi, 2018).

Cannabinoids can interact with the CB₁ receptors, CB₂ receptors, N-arachidonoyl glycine (NAGly) receptors, and opioid or serotonin (5-HT) receptors, which produce analgesic effects (Vučković et al., 2018). CB₁ receptors are commonly found in the brain, spinal cord, and peripheral nervous systems (Pertwee, 1997). CB₂ receptors are found in immune cells, and cannabinoids may induce anti-inflammatory and analgesic effects while acting on human immune cells (Galiègue et al., 1995). Previous studies reported that both CB₁ and CB₂ receptors could upregulate in the nervous structures when the nerve is damaged, explaining the advantages of cannabinoids for pain relief (Lim et al., 2003; Arevalo-Martin et al., 2016). This may be helpful in treating SCI pain, as the upregulated system may counteract the damage of the nerve structure (Arevalo-Martin et al., 2016). Cannabinoids also interact with various neurotransmitters and neuromodulators including acetylcholine, dopamine, γ-aminobutyric acid (GABA), histamine, serotonin, glutamate, norepinephrine, prostaglandins, and opioid peptides. Pharmacologic effects on movement and spastic disorders, which may present in SCI patients, are activated by the interactions with the GABAergic, glutaminergic, and dopaminergic transmitter systems (Musty and Consroe, 2002). However, a pilot study of only five participants reported that cannabinoids did not have a significant analgesic effect compared to that of placebo (Rintala et al., 2010). Similarly, our study showed that compared to placebo, the analgesic effects of the cannabinoids were not beneficial in treating SCI pain. However, this may have resulted from a lack of high-quality evidence. Only four studies were allowed for the meta-analysis. Wilsey et el. reported significant pain reduction in cannabinoids compared to placebo. However, the study was published in 2008 (Wilsey et el., 2008). In fact, the only study that concluded no statistically significant adverse event compared to placebo was the study NCT01606202 2018, which was only registered on Clinicaltrials.gov and never published, even after years (NCT, 2018).

The most common adverse events of cannabinoids involved the nervous system, constituting 27.5% (n = 49) in the cannabinoid group. Among these, the most common adverse events were dizziness (11.2%, n = 20), somnolence (7.3%, n = 13), and dysgeusia (8.4%, n = 15). and confusion (0.5%, n = 1) in the cannabinoid group. All the adverse events had higher prevalence in the cannabinoid group. Somnolence could be explained by the sleep-inducing effect of cannabinoids. Cannabinoids may decrease sleep onset latency, decrease waking after sleep onset, increase slow-wave sleep, and decrease REM sleep (Pivik et al., 1972; Feinberg et al., 1976). These effects are caused by the cannabinoids signaling on CB₁ receptors. Similarly, Huestis et al. reported that cannabinoids may cause severe adverse events, including somnolence, fatigue, diarrhea, vomiting, hepatic abnormalities, central nervous system inhibition, neurotoxicity, and hypotension (Huestis et al., 2019).

A previous meta-analysis from Aviram and Samuelly-Leichtag in 2017 explored the analgesic effect of cannabinoids in chronic pain. They reported that cannabinoids could relieve chronic pain, especially neuropathic pain (Aviram and Samuelly-Leichtag, 2017). However, the study did not focus on SCI-related neuropathic pain and pooled all chronic neuropathic pain in their analyses. A 2019 meta-analysis explored the relationship between the analgesic effects of different cannabinoids and neuropathic pain. However, they also did not specify the cause of the neuropathic pain in their analyses (Rabgay et al., 2020). Our meta-analysis focused on the analgesic effects and adverse effects of cannabinoids in SCI pain, and may therefore guide clinicians in the management of SCI.

There are few meta-analyses exploring the relationship between cannabinoids and neuropathic pain, which are not focused on SCI. The systematic search strategy to identify high-quality research allowed us to make an assessment of the study quality. However, the clinical heterogeneity among the included studies should be noted when interpreting our study findings. For example, some studies included SCI patients with peripheral neuropathic pain, while others included patients with central neuropathic pain. In addition, the included studies used various forms of cannabinoid drugs and treatment dosages, which may affect the certainty of our pooled results. However, in order to address this issue, we performed the meta-analysis using a random-effects model and applied the GRADE system to judge the certainty of the evidence.

Our systematic review and meta-analysis of RCTs suggested that cannabinoids, compared to placebo, have no clinically significant benefits for pain reduction among SCI patients but may have higher rates of adverse events, including dizziness, somnolence, and dysgeusia. Considering that the certainty of the evidence remains suboptimal due to the risk of bias, small sample sizes, and inconsistencies among the included studies, more RCTs are necessary to confirm our findings.