The medical marijuana controversy rages on. Is it a "medicine?" Does it work? Is it safe? Are claims of medical benefits merely a ploy for legalization?

The FDA weighed in several months ago by endorsing a multi-agency study that found "no animal or human data supported the safety or efficacy of marijuana for general medical use." This enraged those who claim that cannabis is an appropriate treatment for ailments from nausea and vomiting to muscle spasticity and intractable pain. They accused the FDA of elevating politics over science -- more specifically, over the conclusions of a 1999 report from the Institute of Medicine (IOM), a branch of the prestigious National Academy of Sciences.

It also rubbed Charley Hooper ("The FDA's Marijuana Problem," TCS Daily, 18 August) the wrong way. However, his arguments are typical of the specious arguments in favor of using smoked marijuana -- as opposed to purified, standardized drug preparations -- for medicinal purposes. For example, he asserts that the FDA's statement, "no sound scientific studies supported medical use of marijuana . . . for general medical use," conflicts with the IOM report and other findings. The operative phrase here is for general medical use. No reputable group has made that claim.

In fact, the FDA's position both makes sense and is consistent with the requirements of the Federal Food Drug and Cosmetic Act. And in spite of claims to the contrary by cannabis supporters and much of the media, it is also consistent with the 1999 IOM report. The IOM's experts rejected the idea that crude herbal (usually smoked) cannabis had been shown to be a safe and effective medication for various medical conditions, concluded that there is "little future in smoked cannabis as a medically approved medication," and emphasized that smoked plant material is a crude drug delivery system that exposes patients to a significant number of harmful substances. They recommended smoked cannabis only for short term use (less than 6 months), and only for patients who suffer from debilitating conditions like intractable pain or vomiting, who have failed on all other therapies, and who are under the close supervision of a physician and an institutional review board-type process. Finally, they predicted that "if there is any future of marijuana as a medicine, it lies in its isolated components, the cannabinoids and their synthetic derivatives," and called for clinical trials to develop "rapid-onset, reliable and safe delivery systems."

The IOM's analysis is far from an endorsement of crude cannabis, in whatever form, as a safe and efficacious medicine that should be made available to patients for a wide variety of medical conditions, as is permitted in eleven states.

In the context of the IOM report and federal law, the FDA's position was perhaps inevitable. First, federal law requires that to be marketed, a drug must have been judged safe and effective by experts who have evaluated evidence obtained from well-controlled clinical trials.

Second, there is the question of what constitutes "evidence". Contrary to the implications of the news reports, it is not an amorphous collection of anecdotal reports and patient testimonials, but rather hard data arising out of carefully designed preclinical and clinical trials. Although there are some recent data from small safety and efficacy trials using smoked cannabis, as the IOM pointed out such trials are merely a first step towards the development of a suitably defined and tested pharmaceutical.

That brings us to a critical third point: What is a "medicine?" In order for a company to sell a drug, and a physician to prescribe it, it must be standardized by composition, formulation, and dose, have been tested for a particular medical condition in rigorous trials, and be administered by means of an appropriate delivery system.

From these, as it were, first principles, it is clear that smoked marijuana will have great difficulty in meeting the required scientific and legal standards. There is insufficient evidence that smoked cannabis is a safe and effective medicine (leaving aside the question of whether the federal soldiers in "the war on drugs" have obstructed clinical testing). Different cannabis strains vary radically in cannabinoid composition and contaminants; plant materials may be contaminated with fungi, bacteria, pesticides, heavy metals and other substances; and there is no safe and reliable delivery system for crude cannabis products.

All of this provides an answer to the question posed by Hooper in his article, "If the synthetic versions are so good, why hasn't the FDA embraced the natural version?" It's not standardizable, so it's not a medicine. Moreover, the synthetic versions are not, in fact, all that good. Both smoked marijuana and the synthetic versions of delta-9-tetrahydrocannabinol (THC) mentioned by Hooper are poorly tolerated by many patients for chronic use. But, as discussed below, there seems to be a better alternative on the horizon.

Finally, there is no justification for treating cannabis differently from other pharmaceutical raw materials. Other plant-derived drugs -- morphine, codeine, and taxol, among many others -- became available only after successfully passing through the FDA review process. The FDA recently issued a guidance document setting forth the path that botanically-based products must take in order to gain regulatory approval, and cannabis must meet those requirements. If physicians and patients are ever to have meaningful access to cannabis' therapeutic potential, crude plant material should serve only as a substrate, the first step in the development of a modern medicine.

The FDA's position on medical marijuana does not forsake science in favor of politics, nor do regulators appear to be negatively disposed toward cannabis as the source of medicines. In January, they approved an Investigative New Drug submission for a product called Sativex, a cannabis-derived drug that has been approved in Canada for the treatment of neuropathic pain in multiple sclerosis and that, although not yet fully licensed, is available by prescription in both Spain and the United Kingdom. More than 1,500 patients are currently using it for a variety of serious conditions under the supervision of their physicians.

After reviewing the data, the FDA agreed to pivotal late-stage (Phase III) clinical trials of Sativex in the United States. The product has been standardized and tested in accordance with modern pharmaceutical standards. It is composed of a fixed ratio of cannabinoids (tetrahydrocannabinol and cannabidiol, a non-psychoactive cannabinoid, in a 1:1 ratio) and is administered by means of an oral spray that delivers the drug through the mucosa of the mouth. These elements appear to enlarge the "therapeutic window," better enabling patients to seek symptomatic relief without experiencing the kind of "high" that many view as an undesirable side effect.

Although it may be the presence of cannabidiol in Sativex that improves the risk/benefit profile, that compound is almost entirely absent from most herbal cannabis in the United States, which has been selected and bred to enhance the levels of tetrahydrocannabinol, THC, for recreational use. Another possible explanation for the minimal psychoactive effects is the spray delivery method, which prevents THC blood levels from rising too rapidly.

The availability of drugs like Sativex should (but won't) end the rancorous debate over medical marijuana in a way that would both benefit patients and satisfy the legal requirement that marketed medicines must be proven safe and effective. Even if it did, the issue of whether marijuana should be legalized as a recreational drug would remain.

Meanwhile, FDA officials must ensure that the testing and potential approval of cannabinoid-containing drugs are not hindered by political agendas or other nonscientific considerations, inside or outside the agency. For the benefit of patients in need, this is something about which the FDA., the "war on drugs" components of the government and other interested parties should strive to agree.

Henry I. Miller, a physician and fellow at the Hoover Institution, headed the FDA's Office of Biotechnology from 1989 to 1993. Barron's selected his most recent book, "The Frankenfood Myth..." one of the 25 Best Books of 2004.