Standards for evaluating clinical trials 3
Analgesia 4
Nausea and Vomiting 9
Wasting Syndrome and Appetite Stimulation 17
Neurological Disorders 22
Glaucoma 38
Summary 42
Other Reports on Marijuana as Medicine 44
References 46


Chapter 4.
The Medical Value of Marijuana and Related Substances

Compared to most drugs, the accumulation of medical knowledge about marijuana has proceeded in reverse. Typically, during the course of drug development, a compound is first found to have some medical benefit. Following this, extensive tests are undertaken to determine the safety and proper dose of the drug for medical use. Marijuana, in contrast, has been widely used in the United States for decades. 165 In 1996, 68.6 million people or 32 percent of the United States population over 12 years old had tried marijuana or hashish at least once in their lifetime although only five percent were current users. 166

The data on the adverse effects of marijuana are more extensive than the data on effectiveness. Clinical studies of marijuana are difficult to conduct. Researchers interested in clinical studies of marijuana face a series of barriers. Research funds are limited, and there is a daunting thicket of regulations to be negotiated at the federal (FDA and DEA) and state levels (see chapter 5). Consequently, the rapid growth in basic research on cannabinoids contrasts with the scarcity of substantial clinical studies on medical uses.
This chapter is devoted to an analysis of the therapeutic value of marijuana and cannabinoids for specific symptoms associated with a variety of conditions. The risks associated with the medical use of marijuana were discussed in chapter 3. It should be noted that THC, the primary active ingredient in marijuana, is an FDA approved drug referred to as dronabinol and marketed under the brand name Marinol®. Marijuana is primarily advocated as a relief from the symptoms of disease, rather than a cure.
For the most part, the claims for medical uses of marijuana are for relief of symptoms, such as nausea, appetite loss, or chronic pain -- each of which can be caused by a variety of diseases or even by the treatments for disease. Therefore, this chapter is primarily organized by symptoms rather than diseases There are eight sections. The first six deal with specific symptoms and conditions and the last two summarize the medical benefits of marijuana and cannabinoids. The five sections on symptoms and conditions are as follows: pain, nausea and vomiting, wasting syndrome and appetite stimulation, spasticity and other neurological symptoms, and glaucoma.
The IOM study team received reports of over 30 different medical uses of marijuana, which is more than could be carefully reviewed in a report of this length; even more uses are reported elsewhere. 66, 67 For most of the infrequently mentioned medical uses of marijuana there are only a few anecdotal reports. This report reviews only the most prominent symptoms that are reportedly relieved by marijuana. However, many of those diseases not reviewed here share common symptoms, such as pain, nausea and vomiting, and muscle spasms which might be relieved by cannabinoid drugs.


Standards for evaluating clinical

Before evaluating individual clinical trials concerning the efficacy and safety of medical uses of marijuana and cannabinoids, it is useful to review the general qualities of clinical trials. Clinical trials involve groups of individuals in which different treatments are compared among different cohorts, or treatment groups. Such trials measure the efficacy of a medication, and are required by the FDA for approval of any new drug or new use of a drug (discussed further in chapter 5).
The degree of assurance that the outcome of a clinical trial is due to the treatment being tested depends on how well the trial is designed. Three important factors to consider in evaluating the design of a clinical trial are: sample selection, subjective effects, and effects that are independent of the treatment. For sample selection, it is important to ensure that patients are allocated to different treatment groups in such a way that the cohorts are not biased towards a particular treatment outcome. For example, the health status, gender, and ages of different cohorts should be equivalent. Subjective effects must be controlled because they influence experimental results in two important ways. First, a patient's expectation that a treatment will be effective can influence the degree of its effect (for example, in the control of nausea). Second, the investigator's expectation can influence his or her interpretation of the treatment effect (for example, when assessing the level of pain experienced by a patient). For these reasons, double-blinding, in which neither the subject nor the person who assesses the drug's effect is aware of the subject's treatment group, is particularly important in cannabinoid drug studies. Another important control for subjective effects includes the use of placebo drugs, which are inert substances, or the use of comparison drugs that have effects similar to the experimental drug. Finally, the quality of the experimental design depends on controlling for factors unrelated to the test drug that might nonetheless influence the treatment outcome. Sequencing effects are one example of such factor. For example, patients might react differently to the same medication depending on whether the medication was administered after an effective or an ineffective treatment. Likewise, compared to a patient whose symptoms are initially mild, a patient whose symptoms are initially severe might react differently to the same drug treatment. Because of the psychological effects associated with cannabinoid drugs, it is particularly important to consider how such side effects might influence the therapeutic value of the treatment. Conditions such as pain and nausea are especially susceptible to subjective influences. For example, depending on the individual, THC can reduce or increase anxiety; it is important to determine to what extent this 'side effect' contributes to the therapeutic effect experienced by the patient.
While double-blind, randomized, controlled clinical trials offer the highest degree of assurance of drug efficacy; such trials are not always feasible. For safety reasons, vulnerable population, such as children, older patients, and women of childbearing age, are often excluded from experimental drug trials. Nonetheless, such patients are part of everyday clinical practice. This challenge of integrating the ideal of standardized and rigorous processes for evaluating the effectiveness of treatments with everyday clinical practice has encouraged interest in single patient trials. 71


Methods for such trials have been established and tested in a variety of clinical settings, usually under everyday conditions. 70, 109, 162 They are particularly valuable when physicians or patients are uncertain about the efficacy of treatment for symptomatic diseases. Controls can be incorporated even in this kind of trial. Such trials can be double-blinded and can involve cross-over designs in which the patient is treated with alternating treatments, such as placebo-drug-placebo, or one drug followed by another drug Most importantly, as with any other clinical trial, a single patient trial should be designed to permit objective comparison between treatments.


Pain is the most common symptom for which patients seek medical assistance.5 Pain associated with either structural or psychophysiological disorders can arise from somatic, visceral, or neural structures. Somatic pain results from activation of receptors outside the brain and is transmitted back to the brain via peripheral nerves. Visceral pain results from activation of specific pain receptors in the gut. It is characterized as a deep aching or cramping sensation, but the source of the pain is often experienced at sites remote from the site of receptor activation, a phenomenon known as referred pain. Neuropathic pain results from injury to peripheral receptors, nerves, or the central nervous system. It is typically burning and the skin feels abnormally unpleasant when gently touched (allodynia), and often occurs in an area of sensory loss -- for example, post-herpetic neuralgia.
All of the currently available analgesic, or pain-relieving, drugs have limited efficacy for some types of pain. Some are limited by dose-related side effects, and some by the development of tolerance or dependence. Cannabinoids, or any new analgesic, could potentially be useful under one of the following circumstances:

· There are medical conditions or patients in which they are more effective than any currently available medication.

· They have a broad clinical spectrum of efficacy and a unique side effect profile that differs from other analgesics.

· They have synergistic interactions with other analgesics.

· They exhibit "side effects" which are considered useful in certain clinical situations.

· Their efficacy is enhanced in patients who have developed tolerance to opioids.

There have not been extensive clinical studies on the analgesic potency of cannabinoids, but data from animal studies indicate that cannabinoids could be useful analgesics. In general, cannabinoids seem to be mild to moderate analgesics. Opiates, such as morphine and codeine, are the most widely used drugs for the treatment of acute pain. But they are not consistently effective in chronic pain, they often induce nausea, and sedation and tolerance might occur in some patients. Recent research has made it clear that CB1, receptor agonists act upon pathways that


partially overlap with those activated by opioids but through pharmacologically distinct mechanisms (see chapter 2). This means that they would likely have a different side effect profile and perhaps additive or synergistic analgesic efficacy.

In light of the solid evidence that cannabinoids can reduce pain in animals (chapter 2), it is important to carefully re-evaluate the evidence concerning analgesic efficacy in humans, and to ask, what clinical evidence is needed to help us decide whether cannabinoids have any use in the treatment of pain?

Clinical studies with cannabinoids

There have been three kinds of studies on the effects of cannabinoids on pain sensitivity in human volunteers: 1) studies of experimentally-induced acute pain, 2) studies of post-surgical acute pain, and 3) studies of chronic pain. Overall, there have been very few studies -- only one since 1981 -- and they have been inconclusive.

Experimentally-induced acute pain

Early studies of cannabinoids in human volunteers using experimental pain models did not demonstrate consistent analgesia. In fact, three early studies testing THC on experimental pain caused by a variety of pain modalities electrical stimulation, tourniquet pain, and thermal pain -- resulted in an increase in pain sensitivity (hyperalgesia). 23, 88, 112
Other studies also failed to show an analgesic effect of THC, but they were not well-designed. Raft and coworkers 152 found no evidence for THC analgesia on pain thresholds and pain tolerance following electrical stimulation and noxious pressure. But their study suffers from two major methodological problems. First, they measured only the extremes of pain sensation -- threshold (the lowest intensity at which a particular stimulus is perceived as painful) and tolerance (the maximum intensity of pain that a subject can withstand). Most pain is experienced in an intermediate range, where effects on pain suppression are most detectable. Modern methods of pain assessment in humans typically use ratings of the intensity of the sensation of pain, a measure which has been found to be superior to assessing the effects of a drug on the extremes of pain. 196 The second problem was that they did not include a positive control; that is, they did not demonstrate the adequacy of their methodology by showing that an established analgesic, such as an opiate, was effective under their study conditions.
Clark and coworkers 23 tested smoked marijuana on thermal pain in human volunteers, and also failed to observe an analgesic effect. However, because of the study design the results are inconclusive. First, there was no positive control to demonstrate the adequacy of their methods; and second, the study subjects were habitual marijuana users. During the study they were hospitalized and allowed free access to marijuana cigarettes for a period of four weeks, consuming an average of 4 to 17 marijuana cigarettes per day. Pain was tested "approximately every one to two weeks." Thus it is quite likely that the subjects were tolerant to THC at the time of testing.


Cancer Pain

The most encouraging clinical data on the effects of cannabinoids on chronic pain are from three methodologically sound studies on cancer pain. Cancer pain can be due to inflammation, mechanical invasion of bone or other pain sensitive structure, or nerve injury. It is severe, persistent, and often resistant to treatment with opioids. In one study, Noyes and coworkers found that oral doses of THC in the range of 5-20 mg produced analgesia in patients with cancer pain. 141, 142 The first experiment was a double-blind, placebo-controlled study of 10 subjects using measures of both pain intensity and pain relief. 142 Each subject received all drug treatments: placebo, 5, 10, 15, and 20 mg THC in pill form each identical in appearance and given on successive days. Fifteen and 20 mg doses of THC produced significant analgesia. It is of interest that there were no reports of nausea or vomiting in this study. In fact, at least half of the patients reported increased appetite. With a 20 mg dose of THC, patients were heavily sedated and exhibited "depersonalization," characterized by a state of dreamy immobility, sense of unreality, and disconnected thoughts. Five of 36 patients exhibited adverse reactions (extreme anxiety) and were eliminated from the study. The mean age of the patients was 51 years, so they were probably not experienced marijuana smokers. Only one patient experienced this effect at the 10 mg dose of THC. A limitation of this study is that there were no positive controls -- that is, other analgesics that could provide a better measure of the degree of analgesia produced by THC.
In a subsequent larger single dose study, the same investigators reported that the analgesic effect of 10 mg of THC was equivalent to 60 mg of codeine, and 20 mg of THC was equivalent to 120 mg of codeine. 141 (Note that codeine is a relatively weak analgesic.) The side effect profile was similar, though THC was more sedating than codeine. In a separate publication, the same authors published data indicating that patients had improved mood, a sense of well-being and less anxiety.
These studies on cancer pain are consistent with the results using a nitrogen analog of THC. Two separate trials were reported: one comparing this analog to codeine in 30 patients, and a second trial comparing it to placebo or secobarbital, a short-acting barbiturate. 178 For mild, moderate and severe pain, this THC-derivative was equivalent to 50 mg of codeine and superior to placebo and to 50 mg of secobarbital.

Surgical Pain

Raft and coworker152 found no analgesic effect of THC on surgical pain, induced by tooth-extraction. However, that study suffers from several serious limitations: 1) the tooth extraction included treatment with the local anesthetic, lidocaine; 2) the pain during the procedure was assessed 24 hours later, and 3) there was no positive control. Levonantradol (a synthetic THC analogue) was tested in 56 patients with moderate to severe post-operative or trauma pain. 93 They were given intramuscular injections of levonantrodol or placebo 24 to 36 hours after surgery. To control for previous drug exposure, patients with a history of drug abuse or addiction, or those receiving an analgesic or anti-inflammatory drug, tranquilizer, sedative, or


anesthetic agent within 24 hours of the test drug were excluded from the study. On average, pain relief was significantly greater for the levonantradol-treated patients than for the placebo-treated patients. Because the authors do not report the number or the percent of people who responded, it is not clear whether this average represents consistent pain relief for all levonantradol-treated patients or whether some people experienced great relief and a few experienced none.

Case reports and surveys

The few case reports of clinical analgesia trials for cannabinoids are not convincing. 89, 122 There are, however, anecdotal surveys that raise the possibility that, in certain patients with chronic pain syndromes with prominent spasticity, there might be a role for cannabinoids. A recent survey of over 100 patients with multiple sclerosis reported that a large number obtain relief from spasticity and limb pain (discussed further under section on multiple sclerosis). 29 Several paraplegic patients reported in an older survey that marijuana use relieved their phantom pain and headache. 42

Migraine Headaches

There is clearly a need for improved migraine medications. Sumatriptan (Imitrex) is currently the best available medication for migraine headaches, but fails to completely abolish migraine symptoms in about 30 percent of migraine patents. 120, 149 Marijuana has been proposed numerous times as a treatment for migraine headaches (reviewed by Russo 160), but there are almost no clinical data on the use of marijuana or cannabinoids for migraine. Our search of the literature since 1975 yielded only one scientific publication on the subject. This report presents three cases in which cessation of daily marijuana smoking was followed by migraine attacks, which is not convincing evidence that marijuana relieves migraine headaches. 44 In fact, the same result could have been found if migraine headaches were a consequence of marijuana withdrawal. While there is no evidence that marijuana withdrawal is followed by migraines, when analyzing the strength of reports such as these it is important to consider all the logical possibilities. Various individuals have claimed that marijuana relieves their migraines, but at this stage there are no conclusive clinical data or published surveys about the effect of cannabinoids on migraines.
There is, however, a possible link between cannabinoids and migraine which is suggested by the abundance of cannabinoid receptors in the periaqueductal gray (PAG) region of the brain. The PAG is part of the neural system that serves to suppress pain and is thought to be involved in the generation of migraines. 56 The link or lack thereof between cannabinoids and migraines might be elucidated by examining the effects of cannabinoids on the PAG. Recent results indicating that both cannabinoid receptor subtypes are involved in controlling peripheral pain, 16 suggest that this might be possible. Nonetheless, the PAG offers only a weak


anatomical correlate between cannabinoid effects on the brain and their possible role in migraine relief:

Conclusions and Recommendations

Basic Pain Research

A key question to address is whether there is any receptor selectivity for the analgesic efficacy of cannabinoids. Are the unwanted side effects (amnesia and sedation) caused by the same receptors in the same brain regions as those producing the analgesia? If the answer is yes, enhancing efficacy will not solve the problem of sedation. Similarly, are the pleasant side effects due to an action at the same receptor? Can the feelings of well-being and appetite stimulation be separated by molecular design? The recent results indicating that both cannabinoid receptor subtypes are independently involved in controlling peripheral pain 16 (discussed in chapter 2) suggest that this might be possible and that further research in this area is clearly warranted.
Further research into the basic circuitry underlying cannabinoid analgesia should be valuable. The variety of neural pathways that underlie the control of pain suggests that a synergistic analgesia "cocktail" might be effective. For example, Lichtman has shown the involvement of the a2 adrenoreceptor in cannabinoid analgesia. Perhaps a combination of a CB1 agonist and the oc2 agonist (e.g., clonidine) would provide enhanced analgesia with a lower side effect level.

Clinical Pain Research

Clinical studies should be directed at those populations of pain patients where there is a demonstrated need for improved management and where the particular side effect profile of cannabinoids holds promise of a decided benefit over current approaches. The following patient groups should be targeted for clinical studies of cannabinoids in the treatment of pain:

· Patients on chemotherapy, especially for the treatment of mucositis, nausea and anorexia

· Postoperative pain patients as an opioid adjunct to determine whether nausea and vomiting from opioids are reduced.

· Patients with spinal cord injury, peripheral neuropathic pain, or central post-stroke pain.

· Patients with chronic pain and insomnia.

· AIDS patients with cachexia, AIDS neuropathy, or any significant pain problem.

In any patient group, an essential question to be addressed is whether the analgesic efficacy of opioids can be augmented. The strategy would be to find the ceiling analgesic effect with an opioid (as determined by pain intensity and tolerability of side effects) and then add in a cannabinoid to determine if additional pain relief can be obtained. This would begin the investigation of potential drug


combinations. As with any clinical study on analgesic drugs, it will be important to investigate the development of tolerance and physical dependence. These are not, of themselves, reasons to preclude the use of cannabinoids as analgesics, but such information is essential to the management of many drugs to which patients develop tolerance or physical dependence.

A secondary question would be to determine whether THC is the only or the best component of marijuana for analgesia. How does the analgesic effect of the plant extract compare to that of THC alone? If there is any difference, then it will be important to identify which combinations of cannabinoids are the most effective analgesics.
In conclusion, the available evidence from both animal and human studies indicates that cannabinoids can produce a significant analgesic effect. One exception is the lack of analgesic effect in studies on experimentally-induced acute pain, but because of limitations in the design of those studies they were inconclusive. Further clinical work is warranted to establish the magnitude of this effect in different clinical conditions and to determine if the effect is maintained over time. Although the usefulness of cannabinoids appears to be limited by side effects, notably sedation, there are other effects, such as anxiolysis, appetite stimulation, and perhaps antinausea and antispasticity effects that should be studied in randomized, controlled clinical trials. It is this particular combination of effects that might warrant development of cannabinoid drugs for certain clinical populations.

Nausea and Vomiting

Nausea and vomiting (emesis) occur under a variety of conditions such as acute viral illnesses, cancer, radiation exposure, cancer chemotherapy, postoperative recovery, pregnancy, motion, and poisoning. Both are produced by excitation of one or a combination of trigger(s) located in the gastrointestinal tract, brain stem and higher brain centers (figure 4.1, Emesis-stimulating pathway). 129 There are numerous cannabinoid receptors in the nucleus of the solitary tract, a brain center that is important in the control of emesis. 83 84 Although the same mechanisms appear to be involved in triggering both nausea and vomiting, one can occur without the other. Much more is known about the neural mechanisms that produce vomiting than nausea, in large part because the act of vomiting is a complex behavior involving coordinated changes in the activity of the gastrointestinal tract, respiratory muscles and posture, whereas nausea is a sensation primarily involving higher brain centers and lacks a discrete observable action.108, 130 Most reports on the antiemetic effects of marijuana or cannabinoids are based on chemotherapy-induced emesis, they are the subject of the following section.


FIGURE 4.1 Emesis-stimulating pathways


Chemotherapy-Induced Nausea and Vomiting

The use of effective chemotherapeutic drugs has produced cures in some malignancies and retarded the growth of others. Unfortunately, nausea and vomiting are frequent side effects of these drugs. Nausea ranks behind only hair loss as a concern of patients on chemotherapy, and many patients experience it as the worst side effect of chemotherapy. The side effects can be so devastating that patients abandon therapy or suffer diminished quality of life. As a result, the development of effective strategies to control the emesis induced by many chemotherapeutic agents is a major goal in the supportive care of patients with malignancies.
The mechanism by which chemotherapy induces vomiting is not completely understood. Studies suggest that emesis is caused by stimulation of receptors in the central nervous system or in the gastrointestinal tract. This stimulation appears to be caused by the drug itself, a metabolite of the drug, or a neurotransmitter. 6 13, 36, 50 In contrast with an emetic like apomorphine, there is a delay between the administration of chemotherapy and the onset of emesis. This delay depends on the chemotherapy agent; emesis can begin anywhere from a few minutes after the administration of an agent like mustine to an hour for cisplatin.13
The most desirable effect of an antiemetic is to control emesis completely, which is currently the primary standard in testing new antiemetic agents (R. Gralla, IOM workshop). Patients recall the number of emetic episodes accurately, even with antiemetics that are sedating or that affect memory. 105 Thus the desired endpoint of complete control is also a highly reliable method of evaluation. The degree of nausea can be estimated through the use of established visual analog scales.a, 22, 59, 105
Another consideration for antiemetic drugs is that the frequency of emesis varies from one chemotherapeutic agent to another. For example, cisplatin causes vomiting in more than 99 percent of patients who are not taking an antiemetic, with approximately 10 vomiting episodes per dose of cisplatin; whereas methotrexate causes emesis in fewer than 10 percent of patients.59 36, 87 Among chemotherapy agents, cisplatin is the most consistent emetic known and has become the benchmark for judging antiemetic efficacy.. To date, antiemetics that are effective with cisplatin are at least as effective with other chemotherapy agents. Lastly, controlling for factors such as the influence of prior chemotherapy and balancing predisposing factors (i.e., gender, age, and prior heavy alcohol usage) among study groups is vital for reliability. Additionally, reliable randomization of patients and blinding techniques (easier when there are no psychoactive effects) are necessary to evaluate e control of vomiting and nausea.

a The visual analog scale is a continuous line representing all possible levels of a particular sensation. It is an estimation of a patient's subjective evaluation and not a true measurement. Patients select a point anywhere on the line to demonstrate the level of sensation they are experiencing, with one end representing one extreme, such as no sensations and the other end representing the opposite extreme, such as a maximum level of that sensation.


THC/Marijuana Therapy for Chemotherapy-Induced Nausea and Vomiting

Cannabinoids are mildly effective in preventing emesis in some patients receiving cancer chemotherapy. Several cannabinoids have been tested as antiemetics, including THC (both 9THC and 8THC) arid two synthetic cannabinoids, nabilone and levonantradol. In addition, smoked marijuana has been examined.

Antiemetic Properties of THC

The quality and usefulness of antiemetic studies depend on adherence to the methodological considerations outlined above. Many of the cannabinoid clinical experiences reported are not based on definitive experimental methods. In studies where THC was compared to placebo, THC was usually found to possess antiemetic properties. However, the chemotherapy drug used varied in most of the trials, and some of the studies consisted of small numbers of patients. In one study, THC was found to be superior to placebo for patients receiving chemotherapy with methotrexate, an agent that is not a strong emetic.19 The same investigators found only a weak antiemetic effect of THC in a small number of patients receiving anthracyclines, chemotherapy drugs that more likely to cause emesis than methotrexate.20
Other trials compared the efficacy of THC with that of prochlorperazine (Compazine®).145, 163 Compazine® was one of the more effective antiemetics available in the 1980s, but it was not completely satisfactory and the search for better agents continued. THC and prochlorperazine given orally showed similar degrees of efficacy, but those studies often used various chemotherapeutic agents, even when administered in combination, THC and prochlorperazine failed to stop vomiting in two-thirds of patients.54
In a carefully controlled double-blind study comparing THC and the antiemetic drug, metoclopramide, in which no patient had previously received chemotherapy so anticipatory emesis was not a factor, all patients received the same dose of cisplatin' and the subjects were randomly assigned to each group (THC or metoclopramide). Complete control of emesis occurred in 47 percent of those treated with metoclopramide, as opposed to 13 percent of those treated with THC.62 Major control (two or fewer episodes) occurred in 73 percent of the patients given metoclopramide compared to 27 percent of those given THC. Despite many flaws in experimental methods, those results suggest that THC reduces chemotherapy-induced emesis.19, 20, 54 164 The studies also indicate that the degree of efficacy is not high. In 1985. the FDA approved THC in the form of dronabinol for this treatment (discussed in chapter 5).
The THC metabolite, 11-OH-THC, is more psychoactive than THC but is a weaker antiemetic.123 Thus it might be possible to design antiemetic cannabinoids without the psychological effects associated with marijuana or THC. 8-THC is less


psychoactive than THC 153, but was found to completely block both acute and delayed chemotherapy-induced emesis in a study of eight children, aged 3-13 years. b Two hours before the start of each cancer treatment and every six hours thereafter for 24 hours, the children were given 8THC as oil drops on the tongue or an a bite of bread (18 mg/m2 body surface area). The children received a total of 480 treatments. The only side effects reported were slight irritability in two of the youngest children (aged 3.5 and 4 years). Based on the prediction that the THC-induced anxiety effects would be less in children than in adults, the authors used doses that were higher than those recommended for adults (5-10 mg/m2 body surface area).

Antiemetic Properties of Synthetic THC Analogues

Nabilone (Cesamet®) and levonantradol were tested in a variety of settings, with results similar to those with THC. While efficacy was observed in several trials, no advantage emerged for these agents. 179, 189 As with the THC trials, nabilone and levonantradol reduced emesis, but not as well as other available agents in moderately to highly emetogenic settings. Neither is commercially available in the United States.

Antiemetic Properties of Marijuana

Among the efforts to study marijuana a preliminary study was conducted in New York State on a group of 56 cancer patients who were unresponsive to conventional antiemetic agents.192 These patients were asked to rate the effectiveness of marijuana compared with results during prior chemotherapy cycles. Seventy-eight percent of patients in this survey rated marijuana as "moderately effective" or "highly effective." The authors concluded that marijuana had antiemetic efficacy, but its relative value was difficult to determine because no control group was used and the patient population was varied with respect to previous experiences, such as prior marijuana use and THC therapy.
A Canadian oncology group conducted a double-blind, cross-over, placebo-controlled study comparing smoked marijuana with THC given in pill form to 20 patients receiving a variety of chemotherapeutic drugs. 111 The degree of emetic control was similar, with only 25 percent of the patients achieving complete control of emesis. Thirty-five percent of the patients indicated a slight preference for the THC pills over marijuana; Twenty percent preferred marijuana and 45 percent expressed no preference.111

b Note that the authors of this study chose to use 8-THC because it is more stable and easier to produce than 9-THC; it does not follow from this particular study that marijuana, with its mixture of cannabinoids, should be a more powerful antiemetic than 9-THC.


Neither study showed a clear advantage for smoked marijuana over oral THC, but neither reported data on the time course of antiemetic control, possible advantages of self-titration with the smoked marijuana, or the degree to which patients were able to swallow the pills. Patients with severe vomiting would have been unlikely to be able swallow or keep the pills down long enough for them to take effect. The onset of drug effect is much faster with inhaled or injected THC than it is for oral delivery.91, 114, 143 Although many marijuana users have claimed that smoked marijuana is a more effective antiemetic than oral THC, no controlled studies have yet been published that analyze this in sufficient detail to estimate the extent to which that this is the case.

Side Effects Associated with THC/Marijuana in Antiemetic Therapy

Frequent side effects associated with THC and marijuana include dizziness, dry mouth, hypotension, moderate sedation, and euphoria or dysphoria. The side effects seen with smoking marijuana include dizziness, dry mouth, confusion, and anxiety. 19, 20, 54, 111, 145, 163 179, 189 Dry mouth and sedation are the least troubling side effects to patients. Perhaps the most troubling side effects are orthostatic hypotension and dizziness, which could enhance the patient's distress.
There is disagreement as to whether the psychoactive effects of THC correlate with its antiemetic activity. In the prospective double-blind trial comparing THC and metoclopramide, the authors report no relationship between the occurrence of complete antiemetic control and euphoria or dysphoria. 62 In contrast, other investigators believe that such psychoactive effects (euphoria or dysphoria) are often associated with improved antiemetic control. 163 Nevertheless, consensus exists among most investigators that dysphoric effects are more common among patients who have not had prior experience with cannabinoids. An important and unexpected problem encountered in the New York State open trial with marijuana was the inability of nearly one-quarter of the patients to tolerate the administration of marijuana by smoking. 192 This intolerance could be due to inexperience with smoking marijuana and is an important consideration.

Therapy for Chemotherapy-Induced Nausea and Vomiting

Present Therapy

New classes of antiemetics have emerged over the last ten years. They have dramatically reduced the nausea and vomiting associated with cancer chemotherapy and literally transformed the acceptance of cisplatin by cancer patients. These new antiemetics, including selective serotonin type 3 receptor antagonists, substituted benzamides, corticosteroids, butyrophenones, and phenothiazines, have few side effects when given on a short-term basis and are convenient in a variety of clinical settings.
The most effective commonly used antiemetics are serotonin receptor antagonists (i.e., ondansetron, granisetron) with or without corticosteroids. 38, 60, 92, 147, 157
In a combination trial using dexamethasone (corticosteroid) and a serotonin


antagonist, complete control of acute cisplatin-induced emesis was observed in about 75 percent of patients. In tests where the chemotherapy was only moderately emetogenic, as many as 90 percent of the patients who received these combinations achieved complete control of emesis. Side effects of those antiemetic agents include headache, constipation and alterations in liver function, but are generally well tolerated by most patients. 14

Other commonly used antiemetics include phenothiazines (i.e., prochlorperazine [Compazine®], haloperidol) and metoclopramide. Metoclopramide is somewhat less effective than the serotonin antagonists, and has more side effects including acute dystonic reactions, drowsiness, diarrhea, and depress ions.14, 38 Side effects associated with phenothiazines include severe or acute dystonic reactions, hypotension, blurred vision, drowsiness, dry mouth, urinary retention, allergic reactions, and occasional jaundice. 14

SIDEBAR: Attitudes of Oncologists Toward Prescribing Marijuana

In the 1990s two groups of investigators conducted three surveys on the attitudes of clinical oncologists toward prescribing marijuana as an anti-emetic. These studies are arguably out of date, because the antiemetics that are available now are so much more effective than they were when these studies were conducted. Nonetheless, the studies merit attention because they are still frequently cited as evidence for or against the use of marijuana as an antiemetic.
The results of the two groups were contradictory. In 1994, by which time serotonin receptor antagonists (5-HT receptors had become available, Schwartz and Beveridge174 concluded that oncologists had little interest in prescribing marijuana to control emesis, whereas Doblin and Kleiman concluded in 1991 that interest was great.40 Since 1994 the two groups have debated in the literature as to which study represents the true sentiment among oncologist.39, 175, 180 In fact. there are numerous methodological differences between the two studies that might explain the different results.39, 175 Nonetheless, ultimately these studies are irrelevant. Both studies deal with perceptions rather than pharmacological realities based on well-designed outcome studies.180

The cost of effective antiemetic regimens can vary markedly depending on the agent, dose, schedule and route of administration. Overall, oral regimens cost less than intravenous regimens due to lower pharmacy and administration costs, as well as lower acquisition costs in many countries. Regimens costing (to the pharmacy) as low as $30 to $35 per treatment session have been shown to be effective.61 These costs reflect the treatment of acute emesis and delayed emesis, with the use of generic agents where available.
Major progress, generally not well-known to the public, in controlling chemotherapy-induced acute nausea and vomiting has been made since the 1970s. Patients receiving the most difficult to control emetic agents now have no more than a 20-30 percent likelihood of experiencing acute emesis,157 whereas in the 1970s the likelihood was nearly 100 percent despite antiemetics.59, 90 As has been seen, most


antiemetic studies with cannabinoids present methodological difficulties and are inconclusive. If one concentrates on the well-conducted trials, the evidence indicates that cannabinoids reduce emesis in about one quarter of patients receiving cancer chemotherapy Cannabinoids are not as effective as several other classes of agents, such as substituted benzamides, serotonin receptor antagonists and corticosteroids. The side effects associated with cannabinoid use are generally tolerable. As with cannabinoids, efficacy was apparent with smoked marijuana, but the degree of efficacy was no better than that seen with available antiemetic agents now considered to be marginally satisfactory. At present, the most effective antiemetic regimens are combinations of oral serotonin receptor antagonists with dexamethasone in single dose regimens given prior to chemotherapy. Neither multiple dose regimens nor intravenous antiemetics provide better control, and both add unnecessary costs. 63, 85

Future Therapy

Advances in therapy for chemotherapy-induced nausea and vomiting will require discovery of agents that work through different mechanisms than existing antiemetics, including the serotonin antagonists. Among the proposed new pathways, agents that involve the neurokinin receptors (NK-1 antagonists) appear to be the most promising.c In animal models, agents that block the NK-1 receptor prevent cisplatin-induced emesis. At the time of this writing, clinical trials with NK-1 receptor antagonists are underway (phase II or small phase III comparison studies) Preliminary results indicate that these agents have useful activity in both acute and delayed chemotherapy-induced emesis and are safe to administer orally.106, 137
It is theoretically possible, considering that the mechanism of cannabinoid action appears to be different than that of the serotonin receptor antagonists and of corticosteroids, that added to more effective regimens, THC might enhance control of emesis. Such combinations should aim to be as convenient as possible while having few added side effects. The critical issue is not whether marijuana or cannabinoid drugs might be superior to the new drugs, but rather whether there is a group of patients who might obtain added or better relief from marijuana or cannabinoid drugs.
Finally, even with the best antiemetic drugs, the control of nausea and vomiting that begins or persists 24 hours after chemotherapy remains imperfect. The pathophysiology appears different than that of acute emesis and it is more likely to occur with a strong emetic agent, hut it varies from patient to patient. Treatment to prevent this emesis requires dosing after chemotherapy as well as before.107

c Neurokinin receptors are found in brain and the gut and thought to be involved in motor activity, mood, pain, and reinforcement. They may well be involved in mediating gut sensations, including nausea.


Conclusions and Recommendations: Chemotherapy-induced nausea

Most chemotherapy patients are unlikely to want to use marijuana or THC as an antiemetic. In 1998, there are more effective antiemetic agents available than were available earlier. By comparison, cannabinoids are only modest antiemetics. However, since modern antiemetics likely act through different mechanisms, cannabinoids might be effective in people who respond poorly to currently used antiemetic drugs, or cannabinoids might be more effective in combination with the new drugs than either are alone. For both reasons, studies of the effects of adjunctive cannabinoids on chemotherapy-induced emesis are worth pursuing for patients whose emesis is not optimally controlled with other agents.
While some people who spoke to the IOM study team described the mood enhancing and anxiety reducing effects of marijuana as a positive contribution to the antiemetic effects of marijuana, one-quarter of the patients in the New York State study described earlier were unable to tolerate smoked marijuana. Overall, the effects of oral THC and smoked marijuana are similar, but there are differences. For example, in the residential studies of experienced marijuana users by Haney and coworkers, subjects reported that marijuana made them feel "mellow," 175 whereas comparable doses of oral THC did not.74 Such differences might be due to the different routes of delivery of THC, as well as the different mixture of cannabinoids found in the marijuana plant. As of this writing, no studies have been published that weigh the relative contributions of those different factors.
The goal of antiemetic medications is to prevent nausea and vomiting. Hence, antiemetics are typically given before chemotherapy, in which case, a pill is an effective from of drug delivery. However, in patients already experiencing severe nausea or vomiting, pills are generally ineffective, because of the difficulty in swallowing or keeping a pill down, and slow onset of the drug effect. Thus an inhalation (but, preferably not smoking) cannabinoid drug delivery system would be advantageous for treating chemotherapy-induced nausea.
Until the development of rapid onset antiemetic drug delivery systems, there will likely remain a subpopulation of patients for whom standard antiemetic therapy is ineffective and who suffer from debilitating emesis. It is possible that the harmful effects of smoking marijuana for a limited period of time might be outweighed by the antiemetic benefits of marijuana, at least for, patients for whom standard antiemetic therapy is ineffective and who suffer from debilitating emesis. Such patients should be evaluated on a case by case basis and treated under close medical supervision.

Wasting Syndrome and Appetite Stimulation

Wasting syndrome in acquired immune deficiency syndrome (AIDS) patients is defined by the Centers for Disease Control as the involuntary loss of more than 10 percent of baseline average body weight in the presence of diarrhea or fever of more than 30 days that is not attributable to other disease processes. 18 Anorexia, which is a loss of appetite, can accelerate wasting by limiting intake of the necessary


nutrients. Both wasting (cachexia) and anorexia are common end-stage features of certain fatal diseases such as AIDS as well as metastatic cancers. In AIDS, weight loss of as little as 5 percent is associated with decreased survival, and a body weight about one-third below ideal body weight results in death.103, 161

There are two forms of malnutrition: starvation and cachexia Starvation is the deprivation of essential nutrients and results from underlying causes such as famine or poverty, malabsorption, and eating disorders (e.g., anorexia nervosa). Starvation results in metabolic adaptations that deplete body fat before losses of lean tissue. Cachexia, in contrast, results from tissue injury such as trauma, infection, or tumor, and is characterized by a disproportionate loss of lean body mass, such as skeletal muscle. 98
The key distinguishing factor between starvation and cachexia is that the effects of starvation can usually be reversed by providing food regardless of the cause of starvation, while the effects of cachexia can be reversed only through control of the underlying disease, and at least for some patients drugs that stimulate metabolism, such as growth hormone or androgenic-anabolic hormones.

Malnutrition in HIV-Infected Patients

By 1997, there were more than 30 million people worldwide infected with human immunodeficiency virus (HIV), and this number is predicted to increase to almost 40 million by the year 2000. 128, 190 Malnutrition is common among AIDS patients and plays an independent and significant role in morbidity and mortality.99
Because treatment for malnutrition depends on whether it is caused by starvation or cachexia, one needs to know what effects HIV infection has on metabolic processes. The answer depends on the clinical situation and can be either or both.98
The development of malnutrition in HIV infection has many facets. Malnutrition in HIV-infected patients results in a disproportionate depletion of body cell mass,d total body nitrogen, and skeletal muscle mass -- consistent with cachexia.101, 198 Body composition studies show that the depletion of body cell mass precedes the progression to AIDS (i.e., falling CD4 lymphocyte counts), suggesting that malnutrition can be a consequence of the inflammatory response to the underlying viral infection, rather than a general complication of AIDS.146 In contrast, weight loss is often episodic and related to acute complications such as febrile opportunistic infections.115 Mechanisms underlying wasting in HIV-infected patients vary depending on the stage of HIV infection and on specific associated complications.
There are many reasons for decreased food intake among AIDS patients. These include mouth, throat, or esophageal infections or ulcers (oropharyngeal and esophageal pathology); adverse effects of medications;200 diarrhea; enteric infection,

d Body cell mass is the fat-free cellular mass. It is composed of the cells of the muscle and organs, plus circulating hematopoietic cells and the aqueous compartment of adipocytes. It is not fat, extracellular water, or extracellular solids (such as tendons, etc.).


malabsorption; serious systemic infection; focal or diffuse neurological disease; HIV enteropathy, depression; fatigue; and poverty. Nutrient malabsorption is often the result of microorganism overgrowth or infection in the gut, especially in the later stages of AIDS. 99, 159

Marijuana, THC, and in HIV-Infected Patients

Despite their frequency of use, there is little published information about the effectiveness of marijuana or cannabinoids for the treatment of malnutrition and wasting syndrome in HIV-infected patients. The only cannabinoid evaluated in controlled clinical studies is THC, or dronabinol. Short-term (6 weeks? and long term (1 year) therapy with dronabinol was associated with an increase in appetite and stable weight, and in a previous short-term (5 weeks) clinical trial in five patients, dronabinol was shown to increase body fat by one percent.9, 10, 182 In 1992, the FDA approved THC, under the trade Marinol®, as an appetite stimulant for the treatment of AID related weight loss. Megestrol acetatee (Meg ace), an appetite stimulant, is more effective than dronabinol in stimulating weight gain, and there is no additive effect of dronabinol when used in combination with megestrol acetate.187 HIV/AIDS patients are the largest group of patients who use dronabinol. However, some reject dronabinol because of the intensity of neuropsychologic effects, an inability to easily titrate the oral dose, and the delayed onset together with the prolonged duration of its action.3 There is evidence to suggest that cannabinoids modulate the immune system (see chapter 2, section on Cannabinoids and the Immune System), which might be a problem in immunologically-compromised patients. No published studies have formally evaluated the use of any of the other cannabinoids in AIDS wasting or as an appetite stimulant.
Anecdotes abound that smoked marijuana is useful for the treatment of HIV associated anorexia and weight loss.24, 66 Some individuals report a preference for smoked marijuana over oral THC because it gives them the ability to titrate the effects, depending on how much they inhale. In controlled laboratory studies on normal, healthy adults, smoked marijuana was shown to increase body weight, appetite, and food intake.49, 121 Unfortunately, there have been no controlled studies of the effect of smoked marijuana on appetite, weight gain and body composition in AIDS patients. At the time of this writing, Donald Abrams at the University of California at San Francisco (UCSF) is conducting the first clinical trial to test the safety of smoked marijuana in AIDS patients, and the results are not yet available.
A major concern with marijuana smoking in HIV-infected patients is that they might be more vulnerable than other marijuana users to immunosuppressive effects of marijuana, or to the exposure of infectious organisms associated marijuana plant material (see chapter 3, Marijuana Smoke).

e Megestrol acetate is a synthetic derivative of progesterone that can stimulate appetite and cause significant weight gain when given in high doses (320 to 640 mg/d) to AIDS patients.58


Therapy for Wasting Syndrome in HIV-Infected Patients

Present Therapy

Generally, therapy for wasting in HIV-infected individuals has focused on appetite stimulants. Few therapies have proven successful in the treatment of the AIDS wasting syndrome. The stimulant studied most is megestrol acetate, which has been shown to increase food intake about 30 percent over baseline for reasons that remain unknown. Its effect in producing significant weight gain is dose-dependent, but the majority of the weight gain is fat tissue, not lean body mass. Although the findings are still preliminary, anabolic compounds, such as testosterone or growth hormone, might be useful in preventing the loss of, or help in restoring, lean body mass in AIDS patients. 11, 46, 68, 173 Additionally, enteral and parenteral nutrition have been evaluated and shown to increase weight, but again, the increase is more due to body fat than to lean body mass. 100, 102
Encouraging advances in the antiviral treatment of HIV infection coupled with developments in the prophylaxis and therapy of opportunistic infections have recently changed the outlook for the long-term health of HlV-infected individuals. Death rates have halved, and the frequency of serious complications, including malnutrition, has fallen markedly. 98, 135

Future Therapy

The primary focus of future therapies for wasting in HIV-infected patients is to increase lean body mass as well as appetite. Active systemic infections are associated with profound anorexia which is believed to be mediated by cytokines that stimulate inflammation through their actions within and outside the brain. 134 Cytokine inhibitors such as thalidomide have been under investigation as potential treatments to increase lean body mass and reduce malnutrition. Even though cannabinoids do not appear to restore lean body mass, they might be useful as adjunctive therapy. For example, cannabinoids could be used as an appetite stimulant in patients with diminished appetite who are undergoing resistance exercises or anabolic therapy to increase lean body mass. Additionally, cannabinoids could be beneficial for a variety of effects, such as increased appetite while reducing the nausea and vomiting caused by protease inhibitors, as well as the pain and anxiety associated with AIDS.
Considering current knowledge about malnutrition in HIV infection, cannabinoids, by themselves, will not likely be a primary therapy for this condition, but might be useful in combination with other therapies. Specifically, the proposed mechanism of action of increasing food intake, would most likely be ineffective in promoting an increase in skeletal muscle mass and functional capacity, the goal in the treatment of cachexia in AIDS patients (D. Kotler, Columbia University, personal communication, 1998).


Malnutrition in Cancer Patients

Malnutrition in cancer patients compromises their quality of life and contributes to the progression of their disease. Approximately 30 percent of Americans will develop cancer in their lifetimes, of whom two-thirds will die as a result of their disease. 5 Depending on the type of cancer, 50-80 percent of patients will develop cachexia, and up to 50 percent of these patients will die, in part, as a result of cachexia. 12, 41 The cachexia appears to result from the tumor itself, and cytokines (proteins secreted by the host during an immune response to tumor? are likely important factors in this development. Cachexia does not occur in all cancers, but generally occurs at the late stages of advanced cancers, such as pancreas, lung, and prostate.
The only cannabinoid evaluated for treating cachexia in cancer patients is dronabinol, which has been shown to improve appetite and promote weight gain. 58 Present treatments for cancer cachexia are similar to that for cachexia in AIDS patients. These treatments are usually indicated at late stages of advanced disease, and include megestrol acetate and enteral and parenteral nutrition. Megestrol acetate stimulates appetite and promotes weight gain in cancer patients, although the gain is mostly fat mass (reviewed by Bruera 1998 15 ). Both megestrol acetate and dronabinol result in dose-related side effects that can be troublesome for patients: megestrol acetate can cause hyperglycemia and hypertension, and dronabinol can cause dizziness and lethargy. Cannabinoids have also been shown to modulate the immune system (see chapter 2, Cannabinoids and the Immune System)' which could be contraindicated for certain cancer patients (both the chemotherapy and the cancer can be immunosuppressive).
Future treatments will probably depend on the development of methods that block cytokine actions, and the use of selective 32-adrenergic receptor agonists to increase muscle masses. 15, 77 Additionally, treatments for cancer cachexia will most likely need to identify the patients' individual needs, in that some patients might need only a cytokine inhibitor, while others could benefit from combined approaches, such as an appetite stimulant and ,B2-adrenergic receptor agonists. In this respect, cannabinoids such as THC might prove useful as part of a combination therapy as an appetite stimulant, antiemetic, analgesic and anxiolytic, especially for patients in late stages of the disease.

Anorexia Nervosa

Anorexia nervosa, a psychiatric disorder characterized by distorted body image and self-starvation, affects an estimated 0.6 percent of the United States population, with a greater predominance in girls than boys. 5 The mortality rate is high, and response to standard treatments is poor.
THC appears to be ineffective in treating this disease. In one study it caused severe dysphoric reactions in 3 of 11 patients. 69 One possible explanation for the


dysphoria is that THC increases appetite and thus intensifies the mental conflict between hunger and food refusal.14 malnutrition.199 Furthermore, such patients may have underlying psychiatric disorders (schizophrenia, depression) in which cannabinoids might be hazardous (see chapter 3 Psychological Harms).

Current treatments include psychological techniques to overcome emotional or behavioral problems and dietary intervention to reverse the malnutrition.199 Pharmacological treatments such as antidepressants have been used in addition to psychotherapy, but tend to lack the desired level of efficacy.34 Recently, alterations in a gene for one of the serotonin receptors have been identified in some patients with anorexia nervosa. 47 The possibility of a genetic component presents a pathway for the development of new drugs to treat this disease.


The profile of cannabinoid drug effects suggests that they are promising for treating wasting syndrome in AIDS patients. Nausea, appetite loss, pain, and anxiety are all afflictions of wasting and all can be mitigated by marijuana Although there are medications that are more effective than marijuana for these problems, they are not equally effective for all patients. Thus we recommend the development and clinical testing of a rapid onset (that is, within minutes) form of THC for such patients. We do not recommend smoking. The long-term harms from smoking make it a poor drug delivery system, particularly for patients with chronic illnesses.
Terminal patients raise different issues. For those patients, the medical harms of smoking are of little consequence. For terminal patients suffering debilitating pain or nausea and for whom all indicated medications have failed to provide relief, the medical benefits of smoked marijuana might outweigh the harms.

Neurological Disorders

Neurological disorders are those that affect the brain, spinal cord, or peripheral nerves and muscles in the body. Marijuana has been proposed most often as a source of relief for three general types of neurological disorders: muscle spasticity, particularly for multiple sclerosis patients and spinal cord injury victims; movement disorders, such as Parkinson's disease, Huntington's disease, and Tourette's syndrome; and epilepsy. Note that marijuana is not proposed as a cure for such disorders, merely that it might relieve certain associated symptoms.

Muscle Spasticity

Spasticity is the increased resistance to passive stretch of muscles and increased deep tendon reflexes. Muscles may also contract involuntarily (flexor and


extensor spasms). In some cases these contractions are debilitating and very painful, and require therapy to relieve the spasms and associated pain.

There are numerous anecdotal reports that marijuana can relieve the spasticity associated with multiple sclerosis or spinal cord injury, and animal studies have shown that cannabinoids affect motor areas in the brain -- areas that might influence spasticity 55, 82, 132, 171

Multiple Sclerosis

As the name implies, "multiple sclerosis" (MS) is a condition in which multiple areas of the CNS are affected. Many nerve fibers become demyelinated,f some are destroyed, and scars form (sclerosis) resulting in plaques scattered throughout the white matter of the CNS. MS exacerbations appear to be caused by abnormal immune activity which causes inflammation and myelin destruction in the brain (primarily the periventricular area), brain stem or spinal cord. The demyelination slows or blocks transmission of nerve impulses and results in an array of symptoms such as fatigue, depression, spasticity, ataxia (inability to control voluntary muscular movements), vertigo, blindness, and incontinence. Approximately 90 percent of MS patients eventually develop spasticity. With an estimated 2.5 million MS patients worldwide, spasticity is a major concern for many patients and physicians.136 Spasticity is variably experienced as muscle stiffness, muscle spasms, flexor spasms, or cramps, muscle pain or ache. The tendency for the legs to spasm at night (flexor spasms) can interfere with sleep.
Marijuana is frequently reported to reduce the muscle spasticity associated with this disease.66, 125 In a mail survey of 112 MS patients that regularly use marijuana, patients reported that spasticity was improved and the associated pain and clonus decreased.29 However, a double-blind placebo-controlled study on postural responses in 10 MS patients and 10 normal volunteers indicated that marijuana smoking impairs posture and balance in MS patients as well as normal volunteers.65 Nevertheless, the 10 MS patients felt that they were clinically improved.
The subjective improvement, while intriguing, does not constitute unequivocal evidence that marijuana relieves spasticity. Survey data do not measure the degree of placebo effect, estimated to be as great as 30 percent in pain treatments.124, 133 Furthermore, surveys do not separate the effects of marijuana or cannabinoids on mood and anxiety from spasticity.
The effects of THC on spasticity were evaluated in a series of three clinical trials testing a total of 30 patients.25, 150, 191 They were 'open trials,' which means the patients were informed before treatment that they would be receiving THC. Based on patient report or clinical exam by the investigator, spasticity was less severe after the THC treatment. However, THC was not effective in all patients and frequently caused unpleasant side-effects. Spasticity was also reported to be less severe in a single case study after nabilone treatment (figure 4.2). 119

f Myelin is the lipid covering that surrounds nerve cell fibers and facilitates the conduction of signals along nerve cells and ultimately between the brain, spinal cord and the rest of the body.


FIGURE 4.2 Effect of nabilone on multiple sclerosis symptoms



Figure 4.2 shows the results of an e-of-1 trial in which a 45-year-old man with multiple sclerosis was given 4-week treatments alternately with placebo and nabilone, which he received every second day.119 In this study design, the patient served as both experimental subject and control; that is, his treatment sequence was: nabilone-placebo-nabilone-placebo. This pattern of alternating treatment reduces the possibility that the observed changes are unrelated to the drug and are not simply due to other factors that changed with the passage of time. The results of this study are consistent with the possibility that THC might relieve spasticity, but this study, although more rigorous than many self-report studies for psychoactive substances, still has problems.
First, although the patient could not distinguish the treatments at the time of taking the pill, he felt sedated after the nabilone. Thus it is not possible to know how much the expectation of relief contributed to his perception of relief: Second, the study measured his perception of pain in which spasticity is an important, but not the only, factor. It is not possible to know to what extent nabilone affected the perception of pain versus the stimulus that generated the pain -- in this case, involuntary muscle contractions. Because it is unaffected by conscious control, the frequency of nocturia is clearer evidence of the effect of THC, although it might also represent how well the patient slept. This single case is intriguing, but not definitive proof that THC can reliably relieve spasticity.


In general, the abundant anecdotal reports are not well-supported by clinical data (summarized in table 4.1) But this is more due to the limitation of the studies than to negative results. There are no supporting animal data to encourage clinical research in this area, but there are also no good animal models for spasticity in MS. Without an appropriate model, studies to determine how marijuana or THC might relieve spasticity cannot be conducted. Nonetheless, the survey results suggest that it would be useful to investigate the potential therapeutic value of cannabinoids in relieving symptoms associated with MS. Such research would require the use of objective measures of spasticity such as the pendulum test.g Since THC is mildly sedating, it is also be important to distinguish this effect from anti-spasticity effects in any such investigations. Mild sedatives such as Benadryl® or benzodiazepines would be useful controls for studies on the ability of cannabinoids to relieve muscle spasticity. The regular use of smoked marijuana, however, would be contraindicated in a chronic condition such as MS.

g The pendulum test is an objective and accurate measure of MS-induced spasticity. It is done by videotaping a patient who lies supine on a table with his or her leg extending off the edge. The leg is dropped and the resulting motion is mathematically analyzed by computer to provide a quantitative measure of spasticity.