Medical Marijuana: Efficacy, Toxicity, and Legality

Medical Marijuana: Efficacy, Toxicity, and Legality

Dr. Wesley M. Durand

Dr. Tina Rizack

Wesley M. Durand, ScB, and Tina Rizack, MD, MPH

Article Highlights

  • A wide variety of formulations exist for medical marijuana use, including dried cannabis flowers, resins, extracts, and oils.
  • The legalization and use of cannabis for medical purposes is an evolving issue, and physician perspectives are likely evolving in tandem.
  • At present, there is limited but evolving support for the use of cannabinoids as anticancer therapies; the largest evidence base exists in gliomas.
  • A growing proportion of state legislatures have voted in favor of medical marijuana use, although tension between state and federal laws persists.

Cannabis use in Western medicine was initially popularized by Irish physician William B. O’Shaughnessy in 1840.1-3 It was included in the U.S. Pharmacopeia shortly afterward and was often prescribed for rheumatism, cholera, tetanus, rabies, and labor pain.1,4 In the early 20th century, however, cannabis was outlawed in several states.5 Access was further reduced with the passage of the Harrison Narcotics Tax Act in 1914, defining drug use as a crime de facto.1,6 The Marijuana Tax Act was passed in 1936, declaring nonmedical cannabis illegal.1,5 It was removed from the U.S. Pharmacopeia in 1942, justified by a lack of recognized medical value.4,5 The 1950s saw passage of the Boggs Control Act and Narcotics Control Acts codifying mandatory penalties for those caught possessing or distributing marijuana.1 Although laws were temporarily eased in the 1970s, the Reagan administration’s Anti-Drug Abuse Act (1986) reestablished a tough federal stance on marijuana use.1,7

The modern trend toward state-level legalization of medical marijuana started with California in 1996.1 Despite growing nationwide support for marijuana legalization, it remains classified as a Schedule I substance: one with high
potential for abuse and lacking a currently accepted medical use.8 Perhaps in response to this trend, in 2009 the Obama administration recommended against prosecution of marijuana for medical purposes.9,10 In 2013, the U.S. Department of Justice clarified that, although marijuana remained illegal on a federal level, it would presently refrain from challenging state-level legalization.11 As of January 2018, 29 states, the District of Columbia, and Puerto Rico permit medical marijuana use; an additional 18 states permit use of “low delta 9-tetrahydrocannabinol (THC), high cannabidiol (CBD)” products in medical circumstances (Fig. 1).12 Finally, although the National Institutes of Health provided $111 million for research on cannabinoid compounds in 2016, others have charged that the pace of cannabis research is hampered by restrictions on scientific marijuana acquisition and distribution.13,14

Cannabis Pharmacology


A wide variety of formulations exist for medical marijuana use, including dried cannabis flowers, resins, extracts, and oils.13 Cannabis oils may be particularly potent; evidence supports that oils may contain up to 75% THC, as compared to 20% in other formulations.15 Importantly, cannabis products may be composed of either THC, CBD, or both. THC is generally considered responsible for the psychoactive effects of cannabis, whereas CBD lacks such activity. This discrepancy may be attributable to differential binding affinity of CBD versus THC for the CB1 and CB2 cannabinoid receptors.13,16,17

Online purchasing is possible, and a recent JAMA research letter purchased and examined 84 cannabinoid products online and found that with respect to CBD, only 31% (+10%) were labeled accurately: 43% had more cannabinoids than labeled, whereas 26% had less or no active ingredient. THC was found in 21%. Oils were the most likely to be labeled accurately (45%), whereas vaporization liquids were the least likely.18

Route of administration

Cannabis products can be inhaled, ingested, or administered topically, vaginally, or rectally.13,19 These diverse routes of administration are associated with distinct pharmacologic profiles. Evidence suggests that inhaled cannabis (typically smoked or vaporized) is associated with more rapid and predictable effects, although ingestion of cannabis yields more variable results.13 Smoke or inhaled cannabis act quickly, with usual onset in minutes, and lasts 2 to 4 hours. Pills and edible forms of cannabis take longer to achieve a therapeutic effect and last longer (4 to 12 hours). They can be perceived as not working and lead to increased doses and potential cumulative overdose. Novel, synthetic cannabinoid products are generally administered orally via capsules.13

FDA-approved cannabinoid products

At present, the U.S. Food and Drug Administration (FDA) has approved three cannabinoid products for medical use.13 Dronabinol is a synthetic THC isomer available in both oral capsular and liquid formulations. Both are indicated for anorexia associated with weight loss in adult patients with AIDS and treatment of chemotherapy-associated nausea and vomiting (CINV) in nonresponders to conventional antiemetic therapies.20 Nabilone is a synthetic cannabinoid derivative also available in an oral capsular formulation.21 Nabilone is indicated for second-line treatment of CINV, and its prescribing information indicates that the product is limited to refractory CINV due to “psychotomimetic reactions not observed with other antiemetic agents.”21

Several additional cannabinoid products may also be on the horizon. Ajulemic acid is an nonpsychoactive synthetic endocannabinoid analog that is being developed in a range of indications related to inflammation and fibrosis.13,21 Cannabidiol (an oral solution of CBD extract) was submitted for consideration by the FDA in late 2017.13 Nabiximol is an oromucosal spray with THC/CBD extracts; although it is currently marked ex-U.S., the company has yet to file for approval with the FDA.13,22 These latter two are notable in that they are composed of cannabis extracts rather than synthetic cannabinoids. The FDA’s response to these products will likely signal its receptivity to further, nonsynthetic cannabis derivatives.

Toxicity and adverse reactions

Medical cannabis exhibits a very favorable toxicity profile. The lethal dose of THC has been estimated to be far greater than amounts consumed in either medical or recreational use.3,23-25 The same is true for synthetic cannabinoids; monkeys administered nabilone at 2 mg/kg/day experienced no significant adverse effects.26

Nevertheless, cannabis use is associated with a variety of side effects. In the acute phase, these may include euphoria, altered judgment, paranoia, amnesia, difficulty concentrating, hallucinations, panic attacks, sedation, unsteadiness, dizziness, dry mucous membranes, and headache.27-29 There is also substantial evidence of an association between chronic cannabis use and the development of schizophrenia in a dose-dependent manner.30 Evidence of a link between marijuana use and other mental health disorders exists, but is less understood.13 Although many cannabinoids are used in treatment of nausea and vomiting, select patients may experience cannabinoid hyperemesis syndrome, characterized by persistent nausea, vomiting, and abdominal pain, and frequent hot bathing to relieve symptoms.31-33

The risk of substance dependence may also be considered a potential side effect of medical cannabis use. The risk of addiction among those experimenting with marijuana has been cited as 9%, and although the vast majority of medical cannabis patients may use marijuana daily, this statistic may be less relevant to these patients.29,34,35 Coffey et al’s 2002 study of Australian young adults observed DSM-IV cannabis dependence rates of 53% among those using cannabis “weekly or more often”; the proportion increased to 72% among those using cannabis “almost daily.”36  The most common dependence symptoms included persistent desire, unintentional use, and withdrawal. Further research on the long-term safety profile of medical cannabis use is warranted; such studies should ideally be conducted among both patients with intended long-term versus limited-term use (e.g., chronic pain vs. CINV, respectively).

Medical Cannabis: Patients and Providers

Who is using medical cannabis?

Several surveys have sought to understand demographics among patients using medical cannabis. In 2006, Reiman et al surveyed 130 patients from medical marijuana facilities in the San Francisco Bay Area.37 They observed a mean age of 40 years and male predominance (74%) among their sample. Medi-Cal (38%) was the most represented insurance type, followed by no insurance (24%), and 59% of those surveyed indicated an annual income of less than $20,000.

A 2014 study by Webb et al surveyed 100 patients in Hawaii, 97% of whom were prescribed cannabis for chronic pain.38 All patients had been certified for 1 year or less and were surveyed upon reapplication for use. The median age was 51 years, and patients reported a 64% average relative decrease in pain (7.8 out of 10 at baseline, 2.8 post-treatment). Additional reported benefits included reduction in stress and insomnia, and the vast majority of patients (71%) reported no adverse reactions.

Reinarman et al reviewed data from 1,746 consecutive admissions to a network of California medical cannabis assessment clinics from July to September of 2006.35 A relative underrepresentation of Latino patients was also observed (14% of the cohort, compared to 32% in California overall as of 2000 Census data). The authors hypothesized that this may be due to the “undocumented status of many Latinos in California.” The majority of patients reported that medical marijuana relieved pain (83%) and improved sleep (71%). Many patients reported substituting marijuana for another prescription medication (51%) or alcohol (13%). Importantly, most patients reported using medical cannabis daily (67%), and the vast majority reported administration via smoke inhalation at least some of the time (86%).

Physician opinions

The legalization and use of cannabis for medical purposes is an evolving issue, and physician perspectives are likely evolving in tandem. A 2005 study by Charuvastra et al found that 36% of a national sample of physicians believed that medically prescribed marijuana should be legal.39 In 2013, Adler et al published results from a poll of New England Journal of Medicine readers; in total, 76% of respondents favored use of medical cannabis.40 Among U.S. respondents, however, the results differed strongly by state. For example, whereas 96% (103 of 107) respondents from Pennsylvania had favorable views of medical marijuana, only about 1% of those from Utah (1 of 76) responded favorably. Notably, neither Pennsylvania nor Utah had legalized marijuana for medical use at the time of this survey.12 If accurate, these results suggest that federal-level efforts to legislate medical marijuana use will need to reconcile diverse regional perspectives.

Therapeutic Uses

Chronic pain

There is a large body of evidence supporting the efficacy of cannabinoids in the treatment of chronic pain. In 2015, Whiting et al published a meta-analysis of trials evaluating cannabinoids for a wide range of indications; chronic pain and CINV were among those with the greatest number of included studies.32 Studied cannabinoid modalities included nabiximols, smoked and vaporized THC, nabilone, THC oromucosal spray, dronabinol, and oral THC capsules, and almost all studies used a placebo comparator. Diagnoses included neuropathic pain, cancer pain, and diabetic peripheral neuropathy, among others. In aggregate, cannabinoids exhibited superior reduction in pain as compared to placebo. Although no differences were observed between diagnosis types or cannabinoid modalities, the ability to adequately detect differences may have been hampered by limited statistical power.

A separate meta-analysis by Andreae et al studied inhaled cannabis specifically, pooling data from five randomized controlled trials in chronic neuropathic pain.41 Using a hierarchical Bayesian model, they aggregated data from trials with differing designs and outcomes. The results indicated a statistically significant benefit of inhaled cannabis in the reduction of chronic neuropathic pain, with a number-needed-to-treat of 5.6 (95% CI [3.4,14]). Nevertheless, the authors caution that further long-term studies are required to evaluate the long-term efficacy and safety profile of inhaled medical marijuana.

Cannabinoids are known to enhance the potency of opioids, and when used synergistically they can allow for reduced doses of opioids and better pain control. When used alone, cannabinoids have shown efficacy in neuropathic pain but appear to be less effective when used alone for other types of pain. The 2017 National Academies of Sciences, Engineering, and Medicine report concluded that adults with chronic pain treated with cannabis or cannabinoids are more likely to experience a clinically significant reduction in pain symptoms, adding that the effect is modest.13

Chemotherapy-induced nausea and vomiting (CINV)

A large number of studies have assessed the efficacy of cannabinoids in the treatment of CINV. Whiting et al’s recent meta-analysis analyzed 28 studies with 1,772 total patients, ultimately finding that patients receiving dronabinol or nabiximols had significantly greater odds of exhibiting a complete nausea and vomiting response, as compared to placebo (OR 3.82, 95% CI [1.55, 9.42]). A recent Cochrane review similarly found substantial evidence supporting efficacy of cannabinoids over placebo in CINV (complete absence of nausea and vomiting RR 2.9, 95% CI [1.8, 4.7]).42

A 2001 meta-analysis by Tramèr et al found that cannabinoids (including nabilone, dronabinol, and levonantradol) were more effective as first-line antiemetics as compared to prochlorperazine, metoclopramide, chlorpromazine, thiethylperazine, haloperidol, domperidone, or alizapride (complete control of nausea and vomiting RR 1.38, 95% CI [1.18, 1.62]).43 These effects, however, were not observed among patients receiving chemotherapy with very low or very high emetogenic potential. Further, evidence from crossover trials suggested that patients strongly preferred cannabinoids. However, Smith et al’s recent Cochrane review found differently; although they observed a trend toward superior efficacy of cannabinoids over prochlorperazine, the results were not statistically significant.42 Additionally, the authors noted the lack of studies comparing cannabinoids to 5-HT3 antagonists and NK-1 inhibitors, indicating the need for additional studies of cannabinoid efficacy against modern antiemetic agents.

Sleep disorders

Whiting et al’s meta-analysis identified two large-scale randomized controlled trials studying cannabinoids in sleep disorders specifically, reporting significant improvement in the sleep quality with nabilone as compared to both placebo and amitriptyline.13,32 A more recent trial by Carley et al compared dronabinol to placebo in the treatment of moderate-to-severe obstructive sleep apnea.44 Treatment with dronabinol (10 mg/day) was associated with significantly improved apnea/hypopnea index and Epworth Sleepiness Scale as compared to placebo.

Additional studies suggest that these results may be generalizable to improved sleep among patients without sleep disorders. Whiting et al analyzed 19 chronic pain and multiple sclerosis trials that included sleep-related outcomes.32 In total, patients treated with cannabinoids (including nabiximols, nabilone, dronabinol, THC/CBD capsules, and smoked THC) exhibited superior sleep quality as compared to placebo.

Anticancer agent

At present, there is limited but evolving support for the use of cannabinoids as anticancer therapies; the largest evidence-base exists in gliomas. A wide range of preclinical in vivo and in vitro studies have observed that cannabinoids exhibit substantial antitumoral activity through a wide range of mechanisms, including necrosis, apoptosis, cytotoxicity, antiproliferation, and oxidative stress-induced cellular damage.45 At present, clinical evidence is limited to a single phase I trial conducted on surgery- and radiotherapy-refractory glioblastoma multiforme patients; Guzman et al administered intracranial THC to nine patients, observing minimal adverse effects.46 A reduction in tumor growth was observed in two patients.


Medical cannabis is an evolving and potentially divisive issue. A growing proportion of state legislatures have voted in favor of medical marijuana use, although tension between state and federal laws persists. Cannabis is associated with a number of potential adverse reactions and risk of dependence; addressing these issues may be crucial for those seeking broader cannabinoid utilization. Nevertheless, medical cannabis has demonstrated efficacy as a treatment for a wide range of conditions, including chronic pain, CINV, and sleep disorders. It may also have potential as an anticancer agent, although the evidence for this effect is currently limited. Additional research on the medical use of cannabinoids is strongly warranted. The pace of such research is likely to accelerate as states continue to approve cannabis for medical purposes. 

About the Authors: Mr. Durand is a medical student at Brown University. Dr. Rizack is a medical oncologist at South County Hospital and a clinical assistant professor of medicine of obstetrics and gynecology at the Alpert Medical School of Brown University.