Leo E. Hollister
Veterans Administration Medical Center
and Stanford University School of Medicine,
Palo Alto, California
Copyright © 1986
by The American Society for Pharmacology
and Experimental Therapeutics
IV. Therapeutic Uses
For many centuries, cannabis was used as a treatment, but only during the 19th century did a particularly lively interest develop for exploiting its therapeutic potential. Cannabis was reported to be effective in treating tetanus, convulsive disorders, neuralgia, migraine, dysmenorrhea, post partum psychoses, senile insomnia, depression, and gonorrhea, as well as opium or chloral hydrate addiction. In addition, it was used to stimulate appetite and to allay the pain and anxiety of patients terminally ill with cancer (64, 121). However, the advent of modern pharmacology beginning in the 20th century discovered many other drugs more definitely effective in these disorders, with a subsequent decrease in the enthusiasm for cannabis as a therapeutic agent.
Advances in the chemistry of cannabis during the 1940s established tetrahydrocannabinol (THC) as the major active component. A semisynthetic THC-like material, synhexyl, was tested as a therapeutic agent during the late 1940s and early 1950s. Initial trials reported efficacy as an antidepressant and as a treatment for alcohol or opiate withdrawal, but subsequent clinical evaluations were negative (156,166).
The exact structure of THC was shown in 1964 to be delta-9-trans-tetrahydrocannabinol, which was soon synthesized. The relative abundance of this material permitted extensive laboratory and clinical studies from 1968 onwards. These studies have included potential therapeutic uses.
At the present time, a number of pharmaceutical houses have programs to develop cannabinoids as therapeutic agents. The major problem is to separate the specific desired pharmacological effect from the pronounced mental effects of cannabinoids. A number of reviews of the potential therapeutic uses of cannabis have been published recently (36, 92, 104). We will now discuss some indications of current interest.
A Antiemetic for Patients in Cancer Chemotherapy
Cancer chemotherapy, especially with the agent cisplatin, produces severe nausea and vomiting, which is extremely difficult to treat with ordinary antiemetic drugs, such as prochlorperazine. This complication is so severe that many patients forego effective cancer chemotherapy. The antiemetic effects of cannabis had been suggested as early as 1972 (6). THC was first tried as an antiemetic in a controlled trial comparing it with placebo in 20 patients undergoing cancer chemotherapy. Fifteen mg were given to some patients and 20 mg to the others in the form of gelatin capsules containing THC dissolved in sesame oil. The initial dose was started 2 h before chemotherapy and repeated 2 and 6 h later. Fourteen of the 20 patients in whom an evaluation could be made reported a definite antiemetic effect from THC, while none was observed from placebo during 22 courses of that drug (149).
Since then, studies have been largely confirmatory but not entirely so. Fifty-three patients refractory to other treatments were studied in an uncontrolled fashion. Ten had complete control of vomiting when THC was administered prior to chemotherapy and for 24 h thereafter. Twenty-eight had 50% or more reduction in vomiting, and only 15 patients showed no therapeutic effect whatsoever. However, four patients were dropped from the study because of adverse effects (113). Fifteen doses of 15 mg of THC were compared with 10-mg doses of prochlorperazine in a controlled cross-over trial in 84 patients. THC produced complete response in 36 of 79 courses, while prochlorperazine was effective in only 16 of 76 courses. Twenty-five patients received both drugs, of whom 20 preferred THC. Of the 36 courses of THC that resulted in complete antiemetic response, 32 were associated with mental effects characterized as a "high" (148). Another comparison between THC in 15-mg doses and prochlorperazine in 10-mg doses versus a placebo control was made in 116 patients who received p.o. doses 3 times a day. The THC regimen was equal to prochlorperazine, and both were superior to placebo. However, many patients who received THC found it to be unpleasant (55). A comparison of THC with placebo was made in 15 patients with each patient acting as his or her control. Fourteen of the 15 patients given THC obtained more relief of nausea and vomiting than from placebo during a course of high-dose methotrexate chemotherapy (28). Best results were obtained when plasma concentrations of THC were more than 120 ng/ml. Such concentrations would ordinarily be expected to produce rather definite mental effects, THC was compared with two other antiemetics, thiethylperazine and metoclopramide, in a controlled cross-over trial. No difference was found between the antiemetic effect of these three agents. However, adverse effects of THC were sufficiently greater than those from the other two drugs, which raised questions about its clinical utility (37). When THC was compared with prochlorperazine and placebo, the latter two treatments were not found to differ, but THC was superior to either one (131).
In summary, it would appear that THC has definite antiemetic effects, that these are comparable to many other commonly used antiemetic agents such as prochlorperazine, thiethylperazine, and metoclopramide, but that the major disadvantage of the drug is the mental effects produced by the doses given.
A synthetic homolog of THC, nabilone, was developed in 1972 and has been tested extensively for antiemetic activity. Across-over study comparing nabilone with prochlorperazine in 113 patients revealed significantly greater response rates following nabilone therapy. However, side effects from nabilone were also more common (77). Although it was hoped that nabilone separated the antiemetic effects from the mental effects of THC, this goal was not fully achieved. Levonantradol and BRL 4664 are two other synthetic THC homologs which showed antiemetic effects in open studies (43, 154). The exact role of synthetic homologs of THC as antiemetic agents remains to be determined.
Currently, a large amount of data on the clinical use of THC as an antiemetic is being accumulated in therapeutic situations monitored by the Food and Drug Administration. Unfortunately, this massive amount of clinical experience has no control, so that it may be impossible to conclude more than what is already known. Meanwhile, extremely promising results have been obtained with larger than usual i.v. doses of metoclopramide. When this drug was compared with prochlorperazine and placebo, it was more effective than either, the only disturbing side effect being sedation (59). The doses used of metoclopramide were 1 mg/kg i.v. before treatment with cisplatin (perhaps the most emetic anticancer drug) and several times after treatment. Protection was total in 48% of courses and major in another 23% (157).
This experience with metoclopramide suggests that the whole issue of the antiemetic effects of THC may become moot, as there are other drugs such as domperidone, which may also be effective in this situation.
Discovery of the ability of cannabis to lower intraocular pressure was more or less fortuitous. Intraocular pressure was measured as part of a multifaceted study of the effects of chronic smoking of large amounts of cannabis. Intraocular pressure was found to decrease as much as 45% in 9 of 11 subjects, 30 min after smoking (75). Lowered intraocular pressure lasted 4 to 5 h after smoking a single cigarette. Its magnitude was unrelated to the total number of cigarettes smoked. the maximal effect on intraocular pressure was produced by the amount of THC absorbed in a single cigarette containing 19 mg of THC. When patients with ocular hypertension or glaucoma were tested 7 of 11 showed a fall of intraocular pressure of 30%. Confirmatory evidence was obtained from a trial in which i.v. injection of THC in doses of 22 ug/kg and 44 ug/kg produced an average fall in intraocular pressure of 37%, with come decreases as much as 51% (40). Many experiments done in rabbits using various routes of administration, including instillation of cannabinoids into the eye, have confirmed the ability of cannabis to reduce intraocular pressure.
Topical administration would be especially desirable for treating glaucoma as with other drugs used for this purpose. Smoking cannabis or taking THC i.v. would be totally unsuitable for patients with glaucoma. Rabbits have been used traditionally for studying eye medications. The lipid solubility of THC has been overcome by using mineral oil as the vehicle for its instillation into the eye. The degree of lowering of intraocular pressure is at least as great as that with conventional eye drops, such as pilocarpine, and the duration of effect is often longer. Some minimal systemic absorption of the drug occurs when it is applied to the conjunctivae, but it is of no consequence in producing mental effects. Other cannabinoids besides THC, such as cannabinol or 8-alpha- and 8-beta-11-dihydroxy-delta-9-THC, have also produced this effect in rabbits (62). These agents have no mental effects, which makes them of considerable interest for therapeutic use.
An extract of nonpsychoactive components of cannabis whose composition is still uncertain has been tested both alone and in combination with timolol eye drops in patients with increased intraocular pressure. The effects of the two agents are additive and are said to be effective when other measures have failed (177). BW 146Y, a synthetic THC homolog, has been given p.o. to glaucomatous patients. Unfortunately, mild orthostatic hypotension and subjective effects were noted in addition to reduced intraocular pressure (167).
No psychoactive component of cannabis can be considered as a feasible therapeutic agent in this situation. Intraocular pressures, although they are reduced acutely, have not been shown to be reduced following long-term treatment, nor has there been any demonstration that visual function is preserved by the use of cannabinoids in glaucoma. Some of the problems associated with the development of cannabinoids as treatment for glaucoma have already been cited (61). The exploitation of cannabinoids for treatment of glaucoma will require much further developmental work to ascertain which cannabinoid will be lastingly effective and well tolerated. The potential benefits could be great, as present-day glaucoma treatment still does not prevent blindness as often as it might. If the effects of cannabinoids are additive to those of other drugs, the overall benefit to patients may be greater than is currently possible with single drugs.
Smoking of material estimated to deliver 12 mg of THC increased sensitivity to an electric shock applied to the skin (78). Single p.o. doses of 10 mg and 20 mg of THC were compared with codeine (60 mg and 120 mg) in patients with cancer pain. A 20 mg dose of THC was comparable to both doses of codeine. The 10 mg dose, which was better tolerated, was less effective than either dose of codeine (129). THC given i.v. in doses of 44 ug/kg to patients undergoing dental extraction produced an analgesic effect, which was less than that achieved from doses of 157 ug of diazepam per kg i.v. Several of these patients actually preferred placebo to the dose of 22 ug of THC per kg because of anxiety and dysphoria from the latter drug (139).
The apparent paradox is that THC both increases and decreases pain. Traditionally, aspirin-like drugs, which work peripherally by inhibiting the synthesis of prostaglandins, are used to treat pain derived from the integument. The initial mental stimulation from THC might increase sensitivity to this kind of pain. Visceral pain, such as that of cancer patients, is usually treated by opiates, which have both peripheral and central sites of action. Recent evidence suggests that opiates may act directly on pain pathways in the spinal cord as well as reducing the effect that produces pain. Cannabis could conceivably modify the effective response. Thus, when the two types of pain are distinguished from each other, the apparent paradox is solved.
THC, nantradol, and nabilone shared some properties with morphine in chronic spinal dog model. Latency of the skin twitch reflex was increased, and withdrawal abstinence was suppressed. Naltrexone did not antagonize these actions, suggesting that they are not mediated through opiate receptors (56). Levonantradol i.m. was compared with placebo in postoperative pain, and a significant analgesic action was confirmed. No dose-response relationship was observed, and the number of side effects from levonantradol was rather high (89). It seems unlikely that any THC homolog will prove to be analgesic of choice, when one considers the present array of very effective new analgesics of the agonist-antagonist type. It is too early to be sure, however.
D. Muscle Relaxant
Patients with spinal cord injuries often self-treat their muscle spasticity by smoking cannabis. cannabis seems to help relieve the involuntary muscle spasms that can be so painful and disabling in this condition. A muscle relaxant or antispastic action of THC was confirmed by an experiment in which p.o. doses of 5 or 10 mg of THC were compared with placebo in patients with multiple sclerosis. The 10 mg dose of THC reduced spasticity by clinical measurement (135). Such single small studies can only point to the need for more study of this potential use of THC or possibly some of its homologs. Diazepam, cyclobenzaprine, baclofen, and dantrolene, which are used as muscle relaxants, all have major limitations. A new sleetal muscle relaxant would be most welcome.
One of the first therapeutic uses suggested for cannabis was as an anticonvulsant. Such an effect was documented experimentally many years ago (112). Subsequent studies in various animal species have validated this action. THC in cats temporarily reduced the clinical and electrographic seizure activity induced by electrical stimulation of subcortical structures (175). Mice were protected by cannabidiol against maximal electroshock seizures but not against those caused by pentylenetetrazole. Its profile of activity more resembled that of phenytoin than that of THC (170). THC and cannabidiol both potentiated the anticonvulsant effects of phenytoin against electrically induced seizures in mice. The two cannabinoids in combination produced the most effect (29). Kindling involves the once-daily application of initially subconvulsive electrical stimulation to culminate in generalized convulsive seizures. THC given chronically to rats prevented the kindling effect (174).
Clinical testing has been rare, despite all these various lines of evidence supporting an anticonvulsant effect of cannabinoids. Better control of seizures following regular marijuana smoking was reported in a not very convincing single case (39). Fifteen patients not adequately controlled by anticonvulsants were treated with additional cannabidiol in doses of 200 or 300 mg or placebo. Cannabidiol controlled seizures somewhat better that the addition of placebo (25). Cannabidiol has little if any psychoactivity, making it a good candidate for this use.
F. Bronchial Asthma
A general study of the effects of marijuana on respiration revealed a bronchodilating action in normal volunteer subjects. Marijuana smoke was calculated to deliver 85 or 32 ug of THC per kg. A fall of 38% in airway resistance and an increase of 44% in airway conductance occurred in the high-dose group. The low-dose group showed lesser changes, but they were still significant as compared with baseline. The sensitivity of the respiratory center to carbon dioxide was not altered by either dose, indicating no central respiratory depression (172).
Asthma was deliberately induced by either inhalation of methacholine or exercise in asthmatic patients. They were then treated with inhalation of placebo marijuana, of saline, of isoproterenol, or of smoke derived from marijuana containing 1 g of THC. Both marijuana smoke and isoproterenol aerosol effectively reversed both methacholine- and exercise-induced asthma, while saline and placebo marijuana had no effect (160). Aerosols of placebo-ethanol, of THC (200 ug) in ethanol, or of salbutamol (100 ug) were tested in another study of ten stable asthmatic patients. Forced expiratory volume in 1-s forced vital capacity, and peak flow rate were measured on each occasion. Both salbutamol and THC significantly improved ventilatory function. Improvement was more rapid with salbutamol, but the two treatments were equally effective at the end of 1 h (181).
Both delta-8 and delta-9-THC have bronchodilating effects, while neither cannabinol nor cannabidiol has such actions. Thus, this action resides only in the psychoactive material. No evidence of tolerance to this effect developed over 20 days of continual administration (58). The treatment of asthma is much more chronic; further studies of tolerance will be needed.
Some patients might experience bronchoconstriction following THC. Doses of 10 mg p.o. produced mild and inconsistent bronchodilator effects as well as significant nervous system effects. One patient of the six studied developed severe bronchial constriction (1). Mild but significant functional impairment, predominantly involving the large airways, was found in 74 regular smokers of cannabis. Such impairment was not detectable in individuals of the same age who regularly smoked tobacco (64).
THC would be best administered by aerosol for this purpose, but whether effective doses would avoid the mental effects is uncertain. The mechanism of action by which THC increases airway conductance may be different from the usual beta-adrenergic stimulants. Resistance to repeated applications of beta-adrenergic stimulants does occur. Another type of bronchodilator might help some patients. The recent introduction of highly effective steroid aerosols, such as beclomethasone, meets that need to a considerable extent.
THC does not differ from conventional hypnotics in reducing rapid eye movement (REM) sleep (136). THC in doses ranging from 61 to 258 ug/kg produces in normal subjects increments in stage 4 sleep and decrements in REM sleep, but without the characteristic REM rebound which follows chronic treatment with hypnotics. When THC was administered p.o. as a solution in doses of 10, 20, and 30 mg, our subjects fell asleep faster after having mood alterations consistent with a "high." Some degree of "hangover" the day following was noted from larger doses (42). Another sleep laboratory study showed that a dose of 20 mg of THC given p.o. decreased REM sleep. After four to six nights of use, abrupt discontinuation of THC produced mild insomnia but not marked REM rebound (52). REM rebound may not be apparent after low doses of THC. However, very high doses (70 to 210 mg) reduced REM sleep during treatment and were followed by marked REM rebound after withdrawal (48).
The sleep produced by THC does not seem to differ much from that of most currently used hypnotics. Side effects before sleep induction as well as hangover effects make the drug less acceptable than currently popular benzodiazepines. It seems unlikely that THC will supplant existing hypnotics in treatment of insomnia.
H. Miscellaneous Uses
1. Hypertension Orthostatic hypotension occasionally follows use of THC (5). A dimethylheptyl side-chain derivative has more profound and constant effects on blood pressure. In man, this compound showed a marked orthostatic hypotensive effect, as well as tachycardia and some mental symptoms resembling THC. While the latter are less than from THC in proportion to the blood pressure-lowering effect, a definite separation of pharmacological effects has not been attained (106).
Effective antihypertensive drugs have been one of the outstanding achievements of pharmacology over the past 30 years. A new antihypertensive based on orthostatic hypotension, perhaps the least desirable mode of lowering blood pressure, is hardly very enticing (8). The issue seems hardly worth pursuing further.
2. Abstinence syndromes due to central nervous system depressants Synhexyl, the first THC homolog to be synthesized, was tested as a treatment for withdrawal reactions from opiates and alcohol with little evidence of efficacy. Withdrawal symptoms experienced by rats following morphine pellet implantation, followed by subsequent injection of naloxone, were reduced by THC. Cannabidiol, without any direct effect itself, augmented the action of THC (79).
This relatively weak effect of cannabinoids in opiate dependence is unlikely to be of clinical use. Detoxification programs using methadone have been highly successful and acceptable.
3. Antineoplastic activity Both the delta-8 and delta-9-THC isomers, as well as cannabinol, have some antineoplastic effect on transplanted lung tumors in animals, as well as on tumors in vitro (125). THC may have a general ability to reduce the synthesis of nucleic acids, which may account for reported immunosuppressant effects as well. Many agents are available that inhibit nucleic acid synthesis, so the possibility that THC or other cannabinoids might be advantageous seems rather unlikely.
4. Antimicrobial action Both THC and cannabidiol inhibit and kill staphylococci and streptococci in vitro at concentrations of 1 to 5 ug/ml (173). Such concentrations are well above those reported for use of THC in man, even at the highest tolerated doses. Thus, this effect seems to have little practical application.
5. Migraine This indication has not been studied systematically in recent years, although it has a long history. In one patient I treated, the mental effects sought socially caused the patient to abandon treatment. Innumerable successful treatments for migraine have been reported at one time or another.
6. Appetite stimulant Most antipsychotic agents will stimulate appetite, but few other drugs do. THC as compared with ethanol and dextroamphetamine produced a variable response on appetite, both in fed and fasted subjects. The majority had increased appetite and food consumption as compared with placebo (80). Anorexia nervosa might be helped by an appetite stimulant. A test of the presumed appetite-stimulating properties of THC in patients with anorexia nervosa failed over a 4-week period. Doses of THC ranged between 7.5 and 30 mg/day and were compared with 30 mg of diazepam per day and placebo. Three of the 11 patients treated with THC experienced severe dysphoria (65).
7. Alcoholism Cannabis users are said not to drink, but most do. The prospect of changing an alcoholic into a cannabinolic has some appeal. A study showed that cannabis was not very attractive to alcoholics. Little difference in retention occurred among those given no medication, or a cannabis cigarette, or disulfiram or the combination of the cigarette and disulfiram (143).