My Response to SAM

It's the standard, stupid

On cue, Smart Approaches for Marijuana (SAM)—the leading anti-marijuana group in the US—released its analysis of the HHS recommendation yesterday. Point by point, I respond. This is my sandbox, after all.

These responses are by no means exhaustive. Certainly, the below doesn’t spell out all the legal arguments—likely not even the best ones. This is a quick take. Also, most of what is below isn’t anything different than what we’ve argued in court in other cases or said elsewhere here On Drugs.

Okay, here it goes.

SAM Point #1: FDA changed the standard

SAM. In years past, the FDA, an agency of the HHS in charge of approving medicinal drugs, used an established five-factor analysis in determining whether Schedule I drugs have “currently accepted medical use.” In the current report supporting the recommendation to reschedule marijuana, the FDA changed these criteria to get its desired answer.

If marijuana had been held to the same standards as other drugs, it would not be deemed to have “currently accepted medical use,” due to the infeasibility of measuring all strains and the insufficient amount of existing research into its safety and efficacy. In fact, the FDA’s new two factor test explicitly states that studies used to support marijuana’s accepted medical use do not need to be controlled, a standard that was required in the five-factor test (save an evaluation done by a consensus of experts).

Response. In 1992, DEA—the agency that isn’t in charge of approving drugs for interstate marketing—created a five-factor test to determine whether drugs have a “currently accepted medical use.” The DEA Administrator responsible for that scheduling action stated, among other items, that “claims that marijuana is medicine are false, dangerous and cruel.”

In scheduling actions since, FDA has applied that 1992 test. According to the test, to show a “currently accepted medical use” five factors must all be met.

1. the drug's chemistry must be known and reproducible;

2. there must be adequate safety studies;

3. there must be adequate and well-controlled studies proving efficacy;

4. the drug must be accepted by qualified experts; and

5. the scientific evidence must be widely available

If this looks like the criteria needed for FDA approval, that’s no coincidence. Despite an appellate court expressly telling DEA it could not use FDA approval as a proxy for currently accepted medical use in 1987, DEA fashioned a test that looks exactly like FDA approval. Though possible to meet the five-factor test without FDA approval, it is not practical. Applying this test, marijuana has never been deemed to have a “currently accepted medical use.”

But as we’ve briefed many times, this five-part test is unlawful. It conflates a standard designed to assess “acceptable” medical uses (i.e., FDA approval) with what is written in the statute: what has been “accepted.” One former federal judge, Judge Paul Watford, reviewed our textual arguments and seemed to agree with us, stating: DEA “may well be obliged to initiate a reclassification proceeding for marijuana, given the strength of petitioners’ arguments that the agency has misinterpreted the controlling statute by concluding that marijuana “has no currently accepted medical use in treatment in the United States.”

SAM’s response illustrates the point precisely and gives away its game. It complains, for example, that it would be impossible to show marijuana has a “currently accepted medical use” due to the infeasibility of measuring “all strains.” How convenient. And yet, it is well-established that marijuana’s placement in Schedule I in 1970 was a tentative one. According to SAM’s view, however, this placement was actually permanent because it is impossible to show marijuana has an accepted medical use.

That is a telltale sign that the standard is wrong. It can’t be the case that Congress enacted a statute designed to permit rescheduling but in fact, made certain substances impossible to be rescheduled.

In the end, SAM is correct that HHS changed the standard. But the standard is and always has been wrong. That a wrong standard was once “established” does not make it correct. After all, the earth was once flat. SAM complains the two-part test is “made up.” DEA’s five-part test that preceded it was also made up. On some level, all agency tests interpreting statutes are made up. At least, the two-part test has grounding in the actual “currently accepted medical use” text.

SAM Point #2: Why did FDA change the test?

SAM. It is not clear why the FDA moved away from the five-factor test, unless the agency approached the rescheduling review with a pre-determined conclusion.

The five-factor test has been used for decades by Republican and Democratic Administrations. Additionally, the United States Court of Appeals for the District of Columbia Circuit examined and validated the test in Alliance for Cannabis Therapeutics v. Drug Enforcement Administration. The marijuana industry petitioners in that case even conceded that the five-factor test had no flaws, as explained in the opinion for the court.

Response. SAM could be right, i.e., HHS¹ approached the rescheduling review with a pre-determined conclusion. But one could say the same thing about the test itself and most of drug policy.

In the early 1970s, NORML petitioned to reschedule marijuana. Following three federal court proceedings, DEA was forced to put the question of whether marijuana had a currently accepted medical use to one of its own administrative law judges, Francis Young. The ALJ heard the evidence and concluded that marijuana had a currently accepted medical use. He recommended Schedule II.

The DEA Administrator then overruled the ALJ. In brief, the five-part test was a legal vehicle invented by DEA as to overrule the ALJ’s determination. DEA had a predetermined conclusion—Schedule I—and then created a legal test to keep marijuana trapped in Schedule I despite evidence of medical use. Then, thanks to a legal doctrine called Chevron deference—a doctrine that has federal courts defer to agency interpretations of statutes they administer but which may soon be going the way of the dodo—the appellate court never took a hard look at the test. That’s our established test folks.

SAM says Alliance for Cannabis Therapeutics signed off on the five-part test. That isn’t entirely correct. As we argued in Aggarwal (seriously, we’ve been arguing this issue in federal court for years) the line of judicial authority that leads to this five-part test becoming “established” is a comedy of judicial errors:

SAM’s point also appears to contain a demonstrable factual error. The “marijuana industry petitioners” in Alliance for Cannabis Therapeutics circa 1994 neither conceded the propriety of the test² nor were in a non-existent “marijuana industry.”

At bottom, the five-part test has been around for a while. But again, it has never been a lawful standard. DEA invented it in 1992 as pretext to keep marijuana in Schedule I. And at that moment in history, there existed no widespread acceptance among states.

Point #3: State medical marijuana programs as evidence that marijuana has currently accepted medical use.

SAM. The FDA now considers the existence of state medical marijuana programs as evidence that marijuana has currently accepted medical use.

A drug’s popularity among the public has never been used as a standard to determine medicine. Supreme Court Justice Steven Breyer in Gonzalez v Raich (545 U.S. 1 (2005), a case involving a medical necessity for the use of marijuana, opined during oral arguments that “medicine by regulation is better than medicine by referendum.” The Court’s 6-3 decision, which Justice Breyer joined, upheld the prohibition of marijuana for medical use under the Controlled Substances Act (CSA). The FDA’s novel standard in recommending marijuana’s rescheduling is rooted in a logical fallacy: some people say that marijuana is medicine, so marijuana must be medicine. The FDA is delegating its authority to determine that medications are safe and effective to popular opinion, a practice that not only exceeds the statutory authority of the agency but makes a mockery of the congressional intent of the Food, Drug, and Cosmetic Act to protect the public health. Following the FDA’s logic, psychedelic drugs, which are seeing a popular push for medical legalization, could also be considered medicine and be rescheduled due to shifting public opinion.

Response. You know someone grasps at straws when she’s citing an off-hand statement during oral arguments. SAM confuses “some people say marijuana is medicine” with 38+ states say marijuana is medicine. Nothing in the HHS recommendation supports a notion that “shifting public opinion” is evidence of accepted medical use.

What seems to continuously escape SAM is that the FDA and FDCA do not regulate the practice of medicine and that the CSA is not a copycat of the FDCA.

The purpose of the FDA and FDCA is not to protect the public health by regulating the practice of medicine but rather, to ensure that drugs are safe and effective for interstate marketing. Historically, FDA is a consumer protection agency. Of course, de facto, FDA approval impacts medical practice. But the primary responsibility to determine accepted medical practice resides with and has always resided with the states. Hence, physicians can prescribe medicines off-label, outside of approved indications.

This is the law. Just one year after Raich, the Supreme Court affirmed in Gonzales v. Oregon that medicine was traditionally supervised by States under their police powers. One quick read of Oregon upends the SAM understanding of the FDCA, CSA, and who regulates the practice of medicine. And, the notion that FDA has exceeded its statutory authority by looking to state practices is just not a serious argument—in court or otherwise.

In sum, SAM wrongly equates “popular opinion” with legislation and infrastructures built by entire states, the traditional regulators of medicine. Then, SAM says unelected career bureaucrats in Washington D.C. ought to determine what is medicine, not state governments.

Point #4: The FDA cherry picked Schedule I and II comparisons.

SAM. The FDA compares marijuana to a limited, hand-picked list of other controlled and noncontrolled substances (e.g., heroin, alcohol, cocaine), not all Schedule I drugs.

In the recommendation, the FDA measured marijuana’s potential for abuse by comparing it to a hand-picked selection of Schedule I, II, and III drugs. For example, the FDA compared marijuana to heroin, another Schedule I drug. The recommendation claims that because marijuana has a lower abuse potential than heroin, it shouldn’t be in the same category. Yet the FDA failed to compare marijuana to other Schedule I drugs, such as LSD. Comparing marijuana’s abuse potential against all Schedule I drugs would have allowed for a more rigorous analysis, but it would not have allowed the FDA to conclude that marijuana belongs in Schedule III. To qualify for Schedule III, a drug or other substance must have “potential for abuse less than the drugs or other substances in Schedules I and II” (21 USC 812(b)). Fifty years of data published by HHS show that marijuana does not meet this standard.

Response. SAM correctly identifies that FDA didn’t compare marijuana to all Schedule I drugs. That it did not is entirely unremarkable. No drug undergoing the scheduling process is compared to every drugs in a schedule. FDA/DEA typically compares a drug to a handful of drugs, such as this tryptamine evaluation from the Five Tryptamine proceeding, which tested or analyzed the tryptamines against some other hallucinogens like DOM—but not every drug in Schedule I.

Again, this is not serious argument. If SAM is trying to illustrate that the CSA and analysis under the CSA has an inherently arbitrary element, it succeeds. SAM may (unintentionally but persuasively) show why LSD and countless other less dangerous drugs do not belong in Schedule I and further, the core problem underlying schedules: drugs aren’t ranked based on danger. But the idea it wants to advance—that FDA should have been compared to all of the hundreds of drugs (some obscure) listed in Schedule I—is crazy.

Point #5: The studies suck.

SAM. None of the studies used by the FDA to justify its claim that marijuana is medicine support that conclusion.

The FDA determined that marijuana is acceptable for medical use for pain, nausea and vomiting, and anorexia. Only three studies were used to justify this claim. The first (University of Florida), concluded that results were “inconclusive or mixed.” The second (National Academies of Sciences & Medicine) relied primarily on a study for which the results were not statistically significant. For the third (Agency for Healthcare Research and Quality), the FDA concluded that the positive effects of marijuana in the study were small and that “the increased risk of dizziness, nausea, and sedation [from marijuana use] may limit the benefit” (page 27 of 252). Furthermore, some of these studies were with inhalable marijuana; prior FDA evaluations have excluded inhalable marijuana studies due to their unreliability and questionable practices.

Response. Though listed as the fifth point, this is SAM’s heaviest hitting argument. Nonetheless, it too, is misleading.

Later in the leaflet, SAM elaborates on the point:

The strongest evidence the FDA cites suggesting that marijuana assists with pain is the National Academies of Sciences & Medicine (NASEM) report from 2017, which concluded that there is “substantial evidence” that marijuana helps with pain (page 27 of 252). The NASEM report itself states that the most comprehensive study supporting this notion was Whiting et al. (2015). In the “Results” section of Whiting et al., the authors write, “Most trials showed improvement in symptoms associated with cannabinoids but these associations did not reach statistical significance in all trials.” The odds ratio (OR) interval for the “reduction in pain” category was 0.99-2.00, indicating that Whiting et al. did not find that marijuana reduced pain at a statistically significant value. Further, Whiting et al. had to issue four corrections to their paper after publishing incorrect values and graphs.

Who’s the cherry-picker how? The NASEM report itself says “[t]here was moderate-quality evidence to support the use of cannabinoids for the treatment of chronic pain and spasticity” and “[b]ased on the GRADE approach, there was moderate-quality evidence to suggest that cannabinoids may be beneficial for the treatment of chronic neuropathic or cancer pain (smoked THC and nabiximols) and spasticity due to MS (nabiximols, nabilone, THC/CBD capsules, and dronabinol).”

Not surprisingly, that’s similar to what FDA said:

Our review of published systematic reviews and meta-analyses shows most authors concluded there is some benefit with marijuana in the treatment of pain conditions, generally ranging from low to moderate effect based on low to moderate quality of evidence.

Agencies can give different weight to different bodies of evidence. In this case, HHS gave certain evidence more than it has in the past. With certain admin law strictures, that is an agency’s prerogative.

Still, there’s a point here. As I noted previously: the science has developed, but not too much or in a material way. That necessarily means that the legal standard is the reason for the change, not the science or data.

Top to bottom, the analysis in the above article—published months before HHS released the redacted analysis—remains spot on. You should read it. Indeed, I called my shot:

Before we see the recommendation from FDA/HHS, I want to highlight the following point from our Sisley briefs and call my shot:

I haven’t seen the HHS recommendation yet. But how much do you want to bet that the Schedule III recommendation is based on some variant of what we identified above? After all, we know the following:

• Marijuana hasn’t been approved by the FDA;

• The science on marijuana as medicine hasn’t changed;

• Marijuana still doesn’t otherwise meet the five-part test criteria;

• What has changed is that many more states have medical marijuana; and

• FDA asked for this state data during its rescheduling review.

We brought Sisley and urged rescheduling in 2020 (and still today) because we knew then that this was quite clearly the best, realistic legal change the cannabis industry and community could obtain based on politics and how the law actually works… The bottom line is that we’re both Gretzky over here. We’re skating where the puck is going to be, not where it has been.

It remains true that one can’t say, for example, that since 2016, the field made a discovery or significant finding regarding marijuana’s medical value that wasn’t available or known in 2016. What developed is acceptance among states. So while the HHS analysis cites newer evidence, in my reading, it doesn’t point to some previously unknown scientific understanding as the basis for rescheduling. Instead, it puts forward a new legal standard and then views the body of evidence though the lens of that standard—as agency’s often do.

None of that means the HHS recommendation is wrong or cherry-picks evidence. But it does mean that this process amounts to less “follow the data and science” and more “follow the law and then reevaluate the data and science.” And there isn’t anything wrong with that.

Conclusion

Let me end with a sincere complement. While I obviously disagree with SAM’s points and analysis, its analysis does show a deep awareness/engagement with the factual/legal issues—engagement that is often absent in analysis I read. It is not banal and is worth talking about.

SAM is upset that the standard for assessing “currently accepted medical use” changed away from a standard that is impossible for medical marijuana to meet and chalks it up to politics. It is probably right. But that doesn’t mean the new legal standard is wrong. Indeed, looking to state acceptance to determine “currently accepted medical use in treatment in the United States” hews far closer to the original meaning of the text than the five-part test.

The legal lesson here is that which a federal agency giveth with discretion can taketh away with discretion. Within the bounds of the law, agencies can change policies and interpretations, including based on political concerns. This is one reason we have elections, after all—to elect a President whose administration matches our political priorities. Here, political winds changed and President Biden told HHS to take another look. So, policy and interpretation changed to embrace that reality. It’s that simple.

1. HHS changed the test, not FDA. FDA then applied the test.

2. In fact, in the 1994 case, the petitioners asked the court to review the propriety of the test and it refused to do so under “law of the case.” SAM appears to take a line from the case completely out of context. The court said: “petitioners concede in their briefs that the new [five part] standard has corrected the flaws we identified in ACT.” Thus, what was conceded wasn’t the propriety of the test but that the test corrected flaws the court had identified in a prior opinion.