MED01-P1 — ChatGPT independent pass — RAW, verbatim

Received 2026-09-23 via Danny’s paste into the Claude session (ChatGPT chat reply, resumed-and-completed submission). Archived unedited below the line; nothing added, removed or reworded. Filed to git only under the §M1 independence firewall — NOT mirrored to Drive and NOT shown to Gemini until Gemini’s pass is filed. First pass to arrive for P1. Verification has not begun; nothing in this file is verified.

Filing notes (Claude): ChatGPT states prior project context WAS visible in its environment and that the two attachments were not mounted, so it retrieved the frozen brief and protocol from the user’s connected Drive itself. Both facts bear on independence and must be weighed in the critique. ChatGPT reports it could not obtain the required reproducible PubMed count (endpoint blocked) and did not estimate one.

Below is the resumed and completed MED01-P1 independent research pass. I have not saved or mirrored it to Drive; this is the frozen submission for Claude to receive first.

MED01-P1 — Independent Research Pass

Historical Evidence Landscape: Medical Cannabis & Cannabinoid Research, 1960–1999

Brief worked from: MED01-P1 Brief v1.0 — FROZEN 2026-09-20 Governing protocol: PROTOCOL-MEDICAL v0.5 BASELINE, git 7beaff6 MED01-wide evidence-search cutoff: 2026-09-20 Research pass: ChatGPT independent pass Scope: Evidence map, not an efficacy review; no medical advice.

Context-isolation disclosure

Prior MED01/project context was visible to me in the environment before this run, including earlier discussion of the project, its protocol development, and broad topic planning. I disregarded that material when selecting findings and sources for this pass. The substantive specification used here was the frozen MED01-P1 brief plus PROTOCOL-MEDICAL v0.5. Because the two intended attachments were not mounted in this conversation, I retrieved only those two exact frozen documents from the user’s connected Drive; I did not open the archive audit, earlier teaching material, the other researcher’s work, or another MED01 research pass.

This is not a systematic review. Searches were structured to map the literature, identify keystone studies, test specified historical claims, and verify source identity and support. “No study identified” below means exactly that: none was identified in the documented searches.

A. Volume and shape of the literature

The 1960–1999 literature was neither absent nor monolithic. It crossed chemistry, pharmacology, psychopharmacology, toxicology, psychiatry, respiratory medicine, oncology, neurology, ophthalmology, and eventually molecular biology. A later Scopus bibliometric analysis located 29,802 cannabis/cannabinoid articles and reviews across 1829–2021; medicine was its largest subject category. The United States contributed only 33 records in that analysis’s 1960s slice but 979 during 1970–79, and U.S. federal funding—NIH, HHS and increasingly NIDA—became conspicuous in the 1970s and 1980s. These figures are country-specific counts, not total global decade counts, and should not be mistaken for the requested worldwide totals. (PubMed Central (PMC))

That bibliometric record suggests four natural research phases within P1:

This phase structure is consistent with the period record itself: by 1999 Pertwee could describe discovery of the endogenous cannabinoid system as having “fueled renewed interest” in clinical potential, while the IOM simultaneously noted rapid basic-science growth alongside relatively sparse substantial clinical research. (PubMed)

Required reproducible database-count query

I attempted the following PubMed/MEDLINE query on 2026-09-23:

(cannabis[Title/Abstract] OR marijuana[Title/Abstract] OR marihuana[Title/Abstract] OR cannabinoid*[Title/Abstract] OR tetrahydrocannabinol[Title/Abstract]) AND ("1960/01/01"[Date - Publication] : "1999/12/31"[Date - Publication])

I attempted both the PubMed web search URL and NCBI ESearch with retmax=0. This environment blocked those executable/search URLs rather than returning the result set, so I did not obtain a live count and do not estimate one.

This is a protocol shortfall in this pass and is self-flagged. The verifier should rerun that exact query—or a frozen alternative chosen before looking at the answer—to supply the missing reproducible count. Older and non-English indexing, title/abstract terminology, delayed indexing, and papers named only for particular compounds all limit what such a count can represent.

B. Discoveries that redirected the field

The decisive change across P1 was not a single “medical cannabis discovery.” It was a chain that turned cannabis from a chemically variable plant drug into a tractable receptor-signalling research system.

Keystone extraction B1 — cannabidiol structure, 1963

Mechoulam & Shvo, 1963. Hashish—I. The structure of cannabidiol. Tetrahedron 19:2073–2078. DOI 10.1016/0040-4020(63)85022-X; PMID 5879214.

• Source type × design: primary × chemical/structural
• System/population: purified cannabis constituent; human population N/A
• Intervention/dose/route/duration/comparator: N/A — chemical structure work
• Outcome: chemical structure of cannabidiol
• Result: paper establishes the structure assignment reported in its title
• Access: bibliographic record; no abstract
• What it did not show: clinical effect or clinical safety

Identity verified live. (PubMed)

Keystone extraction B2 — Δ9-THC isolation/structure and synthesis, 1964–65

Gaoni & Mechoulam, 1964. Isolation, Structure, and Partial Synthesis of an Active Constituent of Hashish. JACS 86:1646–1647. DOI 10.1021/ja01062a046.

• Source type × design: primary × chemical/structural
• System: hashish constituent
• Human N: N/A
• Result: isolation, structural assignment and partial synthesis of the psychoactive constituent
• Full article: paywalled; title, authors, journal, pages and DOI verified live

Mechoulam & Gaoni, 1965. A Total Synthesis of DL-Δ1-Tetrahydrocannabinol, the Active Constituent of Hashish. JACS 87:3273–3275. DOI 10.1021/ja01092a065; PMID 14324315.

• Source type × design: primary × chemical synthesis
• Result: total synthesis
• Clinical inference: none

(ACS Publications)

These studies were enabling rather than therapeutic evidence: reproducible purified/synthetic THC made controlled pharmacology and dose-defined human work feasible.

Keystone extraction B3 — specific cannabinoid receptor, 1988

Devane et al., 1988. Determination and characterization of a cannabinoid receptor in rat brain. Mol Pharmacol 34:605–613. PMID 2848184.

• Source type × design: primary × mechanistic/in-vitro
• System: rat cortical P2 membranes and synaptosomes
• Probe: tritiated CP-55,940
• Comparator: displacement by Δ9-THC and related cannabinoids
• Outcome: saturable, reversible, stereoselective high-affinity binding and G-protein-linked behavior
• Result: fulfilled the authors’ criteria for a pharmacologically distinct high-affinity cannabinoid receptor
• Human effect: not tested

This directly challenged nonspecific membrane-action models. (PubMed)

Keystone extraction B4 — receptor localization and cloning, 1990

Herkenham et al., 1990. PNAS 87:1932–1936. DOI 10.1073/pnas.87.5.1932; PMID 2308954. Autoradiography using [³H]CP-55,940 demonstrated a conserved receptor distribution across mammalian brains, including human tissue, with high densities in basal-ganglia output nuclei, hippocampus and cerebellum. This was anatomical receptor evidence, not a human treatment study. (PubMed)

Matsuda et al., 1990. Nature 346:561–564. DOI 10.1038/346561a0; PMID 2165569.

• Source type × design: primary × mechanistic/molecular
• System: cloned cDNA and expression system; brain/cell-line RNA
• Outcome: sequence and functional expression of a G-protein-coupled cannabinoid receptor
• Result: expressed receptor reproduced expected cannabinoid pharmacology
• Clinical outcome: none

(PubMed)

Keystone extraction B5 — endogenous ligands, 1992–95

Devane et al., 1992. Science 258:1946–1949. DOI 10.1126/science.1470919; PMID 1470919.

• Source type × design: primary × mechanistic/biochemical
• System: porcine brain
• Substance: arachidonylethanolamide (“anandamide”)
• Outcome: structure, receptor binding, cannabimimetic bioassay
• Result: identified an endogenous compound binding the cannabinoid receptor
• Human clinical result: none

(PubMed)

Munro, Thomas & Abu-Shaar, 1993. Nature 365:61–65. DOI 10.1038/365061a0; PMID 7689702. Cloned a second cannabinoid receptor expressed principally in peripheral immune tissue rather than brain, later termed CB2. (PubMed)

Mechoulam et al., 1995. Biochem Pharmacol 50:83–90. DOI 10.1016/0006-2952(95)00109-D; PMID 7605349.

• System: canine intestine; transfected CB1/CB2 cells; mouse assays
• Substance: 2-arachidonoylglycerol
• Result: isolated endogenous 2-AG, demonstrated CB1/CB2 binding and cannabimimetic actions
• Human result: none

Sugiura et al., 1995. BBRC 215:89–97. DOI 10.1006/bbrc.1995.2437; PMID 7575630, independently demonstrated appreciable cannabinoid-receptor binding by 2-AG and proposed it as an endogenous brain ligand. (PubMed)

By 1999, therefore, the literature had moved from “what chemical in cannabis is psychoactive?” to a receptor-and-endogenous-ligand model. That conceptual transition is one of P1’s clearest natural phase changes.

C. What was studied in humans

Human research was much broader than a single medical-use program, but therapeutic trials were usually small, short, and product-specific.

1970s: the first concentrated therapeutic wave

Chemotherapy-induced nausea and vomiting

Sallan et al. conducted a randomized double-blind controlled comparison of oral Δ9-THC and placebo in chemotherapy patients. Twenty-two entered and 20 were evaluable; an antiemetic response occurred in 14/20 THC courses versus 0/22 placebo courses, and among completers in 12/15 versus 0/14, respectively. (PubMed)

Chang et al. then studied 15 osteogenic-sarcoma patients receiving high-dose methotrexate in a prospective randomized double-blind trial incorporating oral and smoked Δ9-THC. THC conditions produced fewer vomiting/retching episodes and less nausea than placebo; the abstract also reported a concentration-response relationship. The paper does not, from the accessible abstract alone, justify treating oral versus smoked cannabis as a clean route-comparison trial. (PubMed)

A later Cochrane reconstruction identified 23 randomized cannabinoid antiemetic trials conducted from 1975 through 1991, 19 of them crossover studies. Twelve tested nabilone and 11 dronabinol; nine included placebo comparisons and many used prochlorperazine or another contemporaneous antiemetic comparator. The review judged the majority vulnerable to allocation-concealment or attrition problems. This is retrospective evidence used to map the size and shape of the period literature, not to retroactively replace contemporaneous conclusions. (Cochrane)

Cancer pain

Noyes et al. gave placebo and 5, 10, 15 and 20 mg oral Δ9-THC double blind to ten patients with cancer pain. The 15- and 20-mg conditions produced greater pain relief than placebo, but sedation and mental clouding were prominent at those doses. A companion trial compared THC with codeine and characterized the analgesic effect of single oral THC doses as mild. (PubMed)

Intraocular pressure / glaucoma

Hepler and Frank’s 1971 JAMA report documented an acute intraocular-pressure response after marijuana smoking. The accessible PubMed record confirms human tonometry after cannabis exposure but provides no abstract-level sample or effect estimate. It should therefore be treated as an early physiological observation, not as proof that glaucoma had been successfully treated. (PubMed)

Asthma / bronchodilation

Williams, Hartley & Graham randomized 10 stable inpatient asthmatics on three successive days to one inhalation of placebo aerosol, 200 μg aerosolized THC, or 100 μg salbutamol, double blind. Salbutamol and THC both improved ventilatory function; salbutamol acted faster, while at one hour the study reported similar bronchodilatation. This was an acute, single-dose physiological study, not evidence about chronic asthma management. (PubMed)

1980s: neurological and pharmaceutical-cannabinoid work

Epilepsy

Cunha et al. first administered CBD 3 mg/kg/day for 30 days to eight healthy volunteers, with eight receiving placebo. In the clinical phase, 15 patients with secondary generalized epilepsy and temporal focus were randomized double blind to 200–300 mg/day CBD or placebo for up to 4.5 months while remaining on existing antiepileptic drugs. Four of eight CBD-assigned patients were reported almost free of seizures, three partially improved and one did not; seven placebo patients were unchanged and one improved. This is a small adjunctive trial with a narrow population and is not equivalent to a general epilepsy finding. (PubMed)

A later Cochrane review concluded that the very small historical CBD trials did not permit reliable general conclusions about cannabinoid efficacy in epilepsy and could not establish long-term safety. That later reassessment is kept separate from what the 1980 investigators themselves reported. (Cochrane Library)

Spasticity

Petro & Ellenberger used double-blind oral 5 mg, 10 mg THC and placebo conditions in patients with spasticity; the principal group contained nine cases. Ten milligrams reduced clinically measured spasticity compared with placebo, while responses in additional patients with other CNS lesions varied. The examiner identified THC conditions in seven of nine cases, underscoring likely blinding problems for psychoactive cannabinoid trials. (PubMed)

1990s: AIDS-related anorexia/wasting and renewed whole-cannabis interest

Beal et al. randomized 139 patients with AIDS-related anorexia and at least 2.3 kg weight loss to dronabinol 2.5 mg twice daily or placebo. Eighty-eight were efficacy-evaluable. Appetite improved by 38% from baseline in dronabinol patients versus 8% with placebo (P=.015); mean weight was stable versus a 0.4-kg placebo loss, a between-group difference that was not statistically significant (P=.14). Euphoria, dizziness and thinking abnormalities were among adverse events. The outcome therefore needs to stay separated: appetite and body weight were not the same endpoint and did not show the same comparative result. (PubMed)

ClinicalTrials.gov retrospectively contains pre-registry-period work such as NCT00000737, a 56-participant NIAID Phase I/II dronabinol/megestrol study in HIV wasting. The record was first submitted in November 1999 but was not posted until 2001, illustrating why absence from ClinicalTrials.gov cannot be used as evidence that a pre-2000 trial did not exist. (ClinicalTrials.gov)

D. Preparations and pharmaceutical development

The literature moved progressively from whole cannabis and crude/variable preparations toward purified or synthetic agents whose dose could be controlled.

Major preparation classes actually used

1. Smoked cannabis/marihuana — experimental human pharmacology, intraocular-pressure work, some antiemetic protocols and observational harms research.
2. Purified/synthetic Δ9-THC (dronabinol) — oral capsules or experimental oral THC in antiemesis, pain, spasticity and appetite.
3. CBD — oral purified cannabidiol in early anticonvulsant/pharmacokinetic work.
4. Nabilone — synthetic cannabinoid analogue developed as a pharmaceutical, especially antiemesis.
5. Levonantradol — investigational synthetic cannabinoid, oral or intramuscular, extensively explored for chemotherapy emesis but limited by adverse effects.
6. Other experimental analogues — including DMHP and by the late 1990s dexanabinol/HU-211.

Nabilone

Lemberger & Rowe’s 1975 Lilly-associated clinical pharmacology study administered single 1–5 mg doses and repeated 1–2 mg twice-daily regimens in humans, documenting dose-related CNS effects and tolerance to some subjective effects. (PubMed)

By 1979, a controlled comparison of nabilone with prochlorperazine reported fewer nausea and vomiting episodes with nabilone but more frequent/severe CNS adverse effects; the study acknowledged Eli Lilly support and drug supply. (New England Journal of Medicine)

Dronabinol / Marinol

FDA records state that the U.S. approved Marinol in 1985 for chemotherapy-associated nausea/vomiting in patients inadequately responding to conventional antiemetics and expanded its approval in 1992 to AIDS-associated anorexia/weight loss. DEA’s July 2, 1999 final rule records the regulatory chain: FDA approval May 31, 1985; FDA-approved dronabinol formulation transferred from U.S. Schedule I to Schedule II in 1986; and that formulation transferred to Schedule III in 1999. Pure dronabinol outside the approved formulation remained separately controlled. (U.S. Food and Drug Administration)

Nabilone / Cesamet

FDA’s retrospective official record states Cesamet (nabilone) was approved in the U.S. in 1985 for chemotherapy-related nausea/vomiting. The accessible period literature confirms human clinical pharmacology from 1975 and antiemetic trials in the late 1970s onward. (U.S. Food and Drug Administration)

Levonantradol

Levonantradol illustrates the distinction between “reached human testing” and “became a medicine.” Studies included intramuscular and oral Phase I/II antiemetic programs and randomized comparisons with metoclopramide, chlorpromazine and prochlorperazine. Some studies found substantial antiemetic activity, but CNS, orthostatic and injection-site adverse effects were frequent. It did not emerge in the period as an approved counterpart to dronabinol or nabilone. (PubMed)

DMHP and dexanabinol

DMHP had controlled human pharmacology by 1974, including intravenous administration and placebo conditions, but it remained investigational. (PubMed)

Dexanabinol/HU-211 had reached escalating intravenous pharmacokinetic testing in healthy volunteers by 1997; later Phase II injury-outcome work belongs to P2 if published after 1999. (PubMed)

The historical record therefore makes “cannabinoid,” “THC,” “synthetic cannabinoid,” and “cannabis” noninterchangeable categories.

E. Study quality: then and now

P1 contains genuine controlled trials, including randomized and double-blind studies, but modern appraisal would repeatedly flag small samples, crossover design without always-adequate reporting of carryover/washout, uncertain allocation concealment, attrition, subjective endpoints, incomplete prespecification, and especially compromised blinding from recognizable psychoactive effects.

The antiemetic literature is the clearest example. The retrospective Cochrane reconstruction found 19 of 23 trials were crossover designs, with most studies at risk of bias from allocation concealment or attrition. Psychoactive adverse effects were common enough that blinding plausibility is questionable even when “double blind” was stated. (Cochrane)

Trial sizes were often tiny: ten cancer-pain patients in Noyes; ten asthmatics in Williams; 15 epilepsy patients in Cunha; roughly nine principal spasticity patients in Petro. Sallan’s antiemesis trial had only 20 evaluable patients. These designs could establish signals and justify subsequent studies but should not be rhetorically inflated into population-wide conclusions. (PubMed)

By the 1995 Beal trial, multicenter parallel-group designs could be substantially larger, but attrition still mattered: 139 were randomized and only 88 were efficacy-evaluable. (PubMed)

No retrospective risk-of-bias score is assigned here to every trial because this pass is an evidence map. Where these studies enter the later ledger as load-bearing claims, full-text RoB 2 appraisal using the correct historical design variant should occur in verification.

F. Conditions of research

F1. Natural phase changes

The literature itself supports three particularly strong inflection points:

1964–65: Δ9-THC becomes chemically defined and synthesizable. (ACS Publications)

1970s: federal supply, synthetic standards and rapidly expanding U.S. funding create a large experimental pharmacology program. A 1969 Congressional Record entry reported NIMH marijuana research expenditures rising from $786,000 in FY1967 to $1.239 million in 1968 and $1.830 million in 1969, with $2.55 million proposed for 1970. (GovInfo)

1988–95: receptor discovery, cloning and endogenous ligands transform the mechanistic model. (PubMed)

F2. Law, scheduling, research access and supply

United States

Legal control did not make controlled research impossible. It created a regulated access system.

Contemporaneous congressional material documents the University of Mississippi marijuana project being established in 1968 as part of the national research program. By 1974 its cultivated material was described as routinely used in NIMH/NIDA-sponsored research. A 1970 congressional record described standardized crops being released through a special NIMH panel and government contracts with Arthur D. Little to produce kilogram quantities of purified Δ8- and Δ9-THC for animal and clinical research. (GovInfo)

A 1976 NIDA research review explicitly listed reliable standardized Mississippi cannabis, cannabinoid analytical methods, supply controls and synthetic THC as developments that enabled the field; it also acknowledged potentially therapeutic pharmacology. (GovInfo)

This did not mean access was frictionless. Schedule I research required regulatory authorization and dependence on controlled supply. But the existence of numerous U.S. human experiments after 1970 directly falsifies the strong claim that scheduling made human research legally impossible.

United Kingdom

By 1998 cannabis and cannabis resin were Schedule 1 substances under the Misuse of Drugs Regulations, so research required special Home Office licensing. The House of Lords described research as legally possible and noted new trials preparing licence applications, while some witnesses regarded the licensing layer as burdensome. Dronabinol and nabilone had different legal statuses. (UK Parliament)

F3. Geography and funding

The research record was international: Hebrew University in Israel drove much early cannabinoid chemistry and endogenous-ligand work; U.S. groups were central to pharmacology, clinical trials, receptor biology and federally supplied cannabis; Brazilian investigators produced important early CBD clinical work; British researchers contributed receptor pharmacology and policy/clinical reassessment; Sweden produced influential psychiatric epidemiology; Japan contributed independently to 2-AG work.

Bibliometric analysis nevertheless shows pronounced U.S. concentration, with the U.S. the most productive country in every decade examined and federal funders increasingly visible after 1970. (PubMed Central (PMC))

I did not identify a sufficiently rigorous grant-level dataset that lets me quantify “harms versus therapeutic research” across the entire 1960–1999 literature. I therefore do not assign a percentage or claim one side dominated the entire period.

G. Harms research

The proposition that the period studied only benefits is false; so is the proposition that it studied only harms. Harms research became a substantial parallel literature.

Tolerance and dependence

Jones, Benowitz & Bachman published Clinical studies of cannabis tolerance and dependence in 1976. PubMed verifies the clinical-trial identity and withdrawal/dependence indexing, although an abstract was unavailable in this search, so I do not extract effect magnitudes from it. DOI 10.1111/j.1749-6632.1976.tb49901.x; PMID 798533. (PubMed)

Psychiatric outcomes

Andréasson et al. followed 45,570 Swedish conscripts for 15 years. Cannabis use on more than 50 occasions was associated with subsequent schizophrenia diagnosis at RR 6.0 (95% CI 4.0–8.9) compared with non-use. This is observational association; the paper’s causal language cannot convert the design into proof that cannabis caused schizophrenia. (PubMed)

Respiratory outcomes

Tashkin et al. studied 28 healthy young experienced male users before and after 47–59 days of heavy ad-libitum marijuana smoking averaging 5.2 cigarettes/day at 2.2% THC. Several pulmonary-function measures fell modestly after the exposure period. (PubMed)

A later eight-year convenience cohort of 394 young adults found tobacco but not marijuana associated with accelerated FEV1 decline; the authors explicitly noted that this did not exclude other respiratory harms. (PubMed)

Roth et al. then used bronchoscopy, biopsy and lavage in 40 subjects and found greater airway inflammatory findings among marijuana, tobacco and combined-smoking groups than nonsmokers. These studies demonstrate why “respiratory harm” cannot be collapsed into one endpoint: airway inflammation, acute obstruction and longitudinal FEV1 decline are different outcomes. (PubMed)

Drug interactions

Benowitz et al. gave CBD 600 mg/day orally for 5–12 days to ten subjects. Hexobarbital oral clearance was 36% lower during CBD; systemic clearance in the four receiving both intravenous and oral hexobarbital was also 36% lower. This was direct human evidence, at a large CBD exposure, that cannabinoid-drug metabolic interactions were being investigated by 1980. (PubMed)

Reproductive, cardiovascular and cognitive domains

Period sources and the 1997 WHO report addressed endocrine/reproductive, developmental, brain/behavior, cardiovascular and other organ effects, but this pass did not extract a single sufficiently representative primary study from every subdomain. WHO’s report itself demonstrates that these domains were established parts of the harms research agenda by 1997. (TDR)

Balance of harms versus therapeutic attention: not quantifiable from the evidence I obtained. The 1969 NIMH record shows substantial federally financed harms/health research, while the 1970s–1990s clinical literature shows clear therapeutic programs. Without comparable publication or grant coding across both classes, saying either “research was mostly harms” or “mostly benefits” would exceed the evidence.

H. What the syntheses and official bodies said

WHO, 1997

World Health Organization, Cannabis: A Health Perspective and Research Agenda, WHO/MSA/PSA/97.4. This 50-page report reviewed current knowledge across brain/behavior, respiratory, endocrine/reproductive, development, immune and other health domains and included a therapeutic-use section plus future-research recommendations. The document is evidence that WHO was treating both harms and therapeutic questions as legitimate scientific subjects by 1997. (TDR)

UK House of Lords, 1998

The Science and Technology Committee stated that rigorous evidence was insufficient to establish conclusively whether cannabis itself had medical value, while judging the existing evidence sufficient to demand urgent trials for MS symptoms and chronic pain. It recommended moving cannabis from Schedule 1 to Schedule 2 to permit named-patient prescribing while retaining controlled-drug status. (UK Parliament)

The Government did not immediately adopt the central rescheduling recommendation. In December 1998 it stated that it wanted scientific evidence from clinical testing before allowing prescription and rejected the recommended Schedule 1→2 move at that stage. (UK Parliament)

U.S. Institute of Medicine, 1999

The White House Office of National Drug Control Policy commissioned IOM in January 1997 to evaluate medical benefits and risks. IOM’s 1999 report concluded that scientific data indicated potential therapeutic value of cannabinoid drugs, particularly THC, for pain, nausea/vomiting and appetite stimulation, while characterizing smoked marijuana as a crude delivery system that also delivered harmful combustion products. It recommended controlled short-term marijuana trials in selected circumstances and emphasized development of defined cannabinoid medicines and safer rapid-onset delivery systems. These are IOM’s conclusions, not this pass’s independent efficacy verdict. DOI 10.17226/6376. (NCBI)

The IOM’s language is useful historically because it rejects both extreme narratives available in 1999: neither “there is no medical signal at all” nor “whole cannabis was already proved as a general medicine” accurately describes its assessment.

I. End-of-era synthesis

I-1. What was reasonably established in humans by the end of 1999?

Here “reasonably established” describes the state of the period literature, not a present-day clinical recommendation.

Pharmaceutical cannabinoid antiemesis had the deepest randomized clinical record. By retrospective reconstruction, 23 randomized dronabinol/nabilone trials had been conducted from 1975–91; U.S. regulatory approvals existed for dronabinol and nabilone. Methodological and adverse-effect limitations remained. (Cochrane)

Dronabinol and appetite in AIDS-related wasting had a multicenter randomized placebo-controlled trial and FDA approval for the indication, although appetite and actual weight change showed different results. (PubMed)

Cannabinoid receptors and endogenous cannabinoid signalling were no longer merely hypothetical: receptor binding, CB1 cloning, CB2 cloning, anandamide and 2-AG had all been directly demonstrated, albeit predominantly in molecular/animal systems rather than as human treatment evidence. (PubMed)

I-2. Suggested but not established

Small controlled or preliminary human studies suggested signals in cancer pain, spasticity, asthma/bronchodilation, intraocular pressure/glaucoma, epilepsy and several other neurological indications, but small samples, short follow-up, psychoactive unblinding and product specificity prevented broad conclusions. (PubMed)

Preclinical evidence suggested antineoplastic and neuroprotective mechanisms, but the disease-modification claims remained non-human in the evidence located here. (PubMed)

I-3. Asserted but not traceably established in this search

I did not identify pre-2000 human evidence establishing that cannabinoids modified cancer progression, prevented neurodegenerative disease, or that a broad category called “medical cannabis” had already been proven effective across multiple diseases.

I also did not identify evidence making pharmaceutical-cannabinoid findings interchangeable with findings for smoked or whole-plant cannabis.

I-4. Open threads for P2

P2 should pick up: modern standardized whole-plant/extract trials; large MS/spasticity programs; nabiximols; expanded neuropathic/chronic-pain research; modern epilepsy CBD development; modern cannabinoid oncology studies; comparative delivery technologies; newer psychiatric/respiratory cohorts; pharmacovigilance; and the post-1999 resolution of the HHS neuroprotection patent and any clinical programs only begun by the end of P1.

Archive-derived audit probes — answered after the literature map

Q1. Isolation, structure and synthesis of major plant cannabinoids

Δ9-THC: Gaoni and Mechoulam’s 1964 paper verified isolation, structure and partial synthesis; Mechoulam and Gaoni’s 1965 paper verified total synthesis. (ACS Publications)

CBD: Mechoulam and Shvo’s 1963 paper verified its structure. (PubMed)

A 1965 Mechoulam/Gaoni paper also addressed cannabinolic, cannabidiolic and cannabigerolic acids, but I did not independently retrieve enough primary literature here to assign every “first isolation” claim for CBG, CBC, CBN and each minor phytocannabinoid. A later period review indicates dozens of cannabinoids were already known by 1980; exact priority claims should be verified individually rather than reconstructed from historical shorthand. (PubMed)

Q1 status: THC and CBD chronology verified; minor-cannabinoid “firsts” incomplete.

Q2. How cannabinoids act in the body

The best documented sequence in this pass is:

1988 — Devane et al., St. Louis University: saturable stereoselective cannabinoid binding in rat brain; receptor pharmacology. (PubMed) 1990 — Herkenham et al., NIMH: receptor autoradiographic localization, including human brain tissue. (PubMed) 1990 — Matsuda et al., NIMH: cloned and functionally expressed the CB1 receptor cDNA. (PubMed) 1992 — Devane/Hanuš/Mechoulam group: isolated anandamide from porcine brain. (PubMed) 1993 — Munro/Thomas/Abu-Shaar, MRC Cambridge: cloned peripheral CB2 receptor. (PubMed) 1995 — Mechoulam et al. and Sugiura et al.: 2-AG identified/proposed as another endogenous cannabinoid ligand. (PubMed)

Q3. Groups, institutions and countries

High-visibility nodes include Hebrew University/Israel; NIMH/NIH and U.S. universities; St. Louis University; University of Mississippi/NIMH/NIDA supply program; Brazilian groups associated with Carlini/Cunha; MRC Cambridge and other UK pharmacology groups; Karolinska/Sweden; and Japanese 2-AG researchers. Later bibliometrics show U.S. publication and funding concentration, but older/non-English indexing prevents interpreting this as a complete geography of all research. (PubMed Central (PMC))

Q4. Legal research supply

In the U.S., the University of Mississippi project was established in 1968 and developed standardized research cannabis; NIMH/NIDA controlled release to investigators. Synthetic THC was also manufactured under federal contracts. (GovInfo)

In the UK by 1998, cannabis research was possible under special Home Office licence despite Schedule 1 status. (UK Parliament)

This pass did not reconstruct every national supply regime worldwide.

Q5. Public-funder priorities and harms/therapeutics balance

A 1969 U.S. congressional record described rapidly expanded NIMH funding and research into brain, nervous-system, chromosomal, organ, long-term and behavioral effects. By 1976, NIDA’s own review also explicitly discussed possible therapeutic applications and clinical cannabinoid development. (GovInfo)

I did not find a comparable coded grant dataset supporting a numeric harms-versus-therapeutics balance for the whole era. Verdict: balance cannot be quantified from this search.

Q6. Was controlled human research legally possible?

Yes, demonstrably, in multiple jurisdictions. The United States produced controlled human work after Schedule I placement, including cancer-pain, antiemesis, asthma, CBD and spasticity studies; access required regulated study drug and approvals rather than being legally impossible. UK research likewise remained possible under special licence. (PubMed)

Q7. National/official commissions and expert bodies

Verified in this pass: WHO 1997; UK House of Lords 1998; U.S. IOM 1999. Earlier bodies including Canadian and U.S. commissions form part of the historical literature, but I did not obtain sufficiently detailed primary texts from each to reproduce all recommendations without secondary dependence. The three above are therefore the official syntheses on which I place most weight here. (TDR)

Q8. Claims that official cannabis reports were suppressed/destroyed

For the reports checked here, primary publication records exist: WHO’s report is institutionally archived; the House of Lords report was formally printed and remains in the parliamentary record; the IOM report was formally published by National Academies Press. I identified no primary documentation in these searches demonstrating confiscation or destruction of those reports. That does not prove no suppression episode ever occurred for any cannabis report. Broad suppression stories not tied to a specific primary record remain untraceable in this pass. (TDR)

Q9. Synthetic/pharmaceutical cannabinoids reaching human testing by 1999

Dronabinol/Δ9-THC: extensive 1970s–1990s human study; Marinol FDA approval 1985 for refractory chemotherapy emesis, 1992 AIDS anorexia; approved formulation Schedule II from 1986 and Schedule III from 1999 in the U.S. (GovInfo)

Nabilone: human clinical pharmacology published 1975 from Lilly-associated research; antiemetic studies followed; FDA retrospective record gives 1985 approval. (PubMed)

Levonantradol: investigational synthetic derivative, human antiemetic trials from the early 1980s; no approval identified in the P1 record. (PubMed)

DMHP: controlled human pharmacology documented in 1974; investigational. (PubMed)

Dexanabinol/HU-211: intravenous healthy-volunteer pharmacokinetic work documented by 1997; later therapeutic trial resolution belongs downstream where publication dates exceed P1. (PubMed)

This list may not exhaust every synthetic cannabinoid ever given experimentally to humans before 2000; that is a self-flagged completeness issue.

Q10. Head-to-head preparations/routes

The period did compare cannabinoid drugs with active antiemetics and tested multiple routes. Chang’s study incorporated oral and smoked THC, but the accessible abstract does not permit a clean isolated oral-versus-smoked treatment-effect estimate. Levonantradol was investigated orally and intramuscularly across separate trials. Nabilone was compared directly with prochlorperazine. (PubMed)

The planned UK trials described in 1998 were notable precisely because they proposed like-for-like oral cannabis extract vs dronabinol vs placebo, but those results were not yet available by the P1 boundary. (UK Parliament)

Q11. Disease-modification claims

Cancer

Munson et al., 1975, JNCI 55:597–602, DOI 10.1093/jnci/55.3.597, PMID 1159836:

• design: primary × animal/in-vitro
• model: mouse Lewis lung adenocarcinoma and leukemia models; isolated cells
• cannabinoids: Δ9-THC, Δ8-THC, CBN, CBD
• doses: oral THC/cannabinoids, including high mg/kg doses reported in abstract
• outcome: tumor growth/survival/cellular precursor uptake
• result: some cannabinoids retarded tumor growth in mouse models; CBD did not in the tested model
• humans: none

(PubMed)

I identified no corresponding pre-2000 human cancer disease-modification trial.

Neuroprotection

Hampson et al., 1998, PNAS 95:8268–8273, DOI 10.1073/pnas.95.14.8268, PMID 9653176:

• design: primary × in-vitro
• model: rat cortical neuron cultures exposed to glutamate toxicity
• cannabinoids: CBD and Δ9-THC
• comparator/context: oxidative/neurotoxic injury assays
• result: reduced cell injury; antioxidant activity; receptor-independent neuroprotection in this model
• humans: none

(PubMed)

Thus the period supported preclinical hypotheses of disease modification, not human proof of cancer treatment or neurodegenerative disease modification.

Q12. Compassionate/individual-patient cannabis access

I located retrospective descriptions of a U.S. federal compassionate IND program but found conflicting dates across later institutional retellings and did not retrieve the primary opening/closure documentation needed to satisfy the brief’s standard for legal basis, exact dates and patient counts. I therefore do not freeze a precise start/end/patient-count claim here. This should be resolved during verification from FDA/PHS archival records.

Q13. Formal scheduling proceedings, 1970–99

The long U.S. marijuana-rescheduling proceeding culminated in DEA’s continuing Schedule I position. Later Federal Register summaries identify 57 FR 10499 (March 26, 1992), Marijuana Scheduling Petition; Denial of Petition; Remand, followed by judicial review. The administrative record also includes the well-known 1988 ALJ recommendation for Schedule II, but I did not retrieve the complete official 1988 decision from its original federal archive in this pass; therefore detailed quotations from that ruling are not used as candidate evidence. (GovInfo)

Q14. Pre-1960 background

One commonly repeated claim is firmly documentable: William Brooke O’Shaughnessy’s work on Cannabis indica and tetanus/convulsive conditions was published in the Calcutta medical literature and reproduced in the Provincial Medical Journal in 1843. The surviving primary text explicitly bears the title On the Preparations of the Indian Hemp, or Gunjah … and their Utility in the Treatment of Tetanus and other Convulsive Diseases. It establishes that cannabis was investigated and reported medically in nineteenth-century Western medicine; it does not by itself validate the efficacy claims by modern standards. (PubMed Central (PMC))

Public historical claims

P1. “There was almost no research on cannabis before legalization.”

Unsupported in the strong form. A large chemistry, pharmacology, harms and clinical literature existed well before modern legalization. Bibliometric data and the extensive 1970s–1990s primary record directly contradict “almost none.” What is supportable is that substantial modern clinical-efficacy programs were much smaller than basic/pharmacological research. (PubMed Central (PMC))

P2. “Drug scheduling made cannabis research impossible.”

Unsupported in the strong form; partly supported if restated as “created regulatory barriers.” Human and animal cannabis/cannabinoid research continued after U.S. Schedule I placement, supported through a federally controlled supply system. UK Schedule 1 likewise allowed research under special licence. (GovInfo)

P3. “Early research looked only for harms.”

Unsupported. Harms research was substantial, but contemporaneous therapeutic research included nausea/vomiting, pain, asthma, intraocular pressure, spasticity, epilepsy and appetite. NIDA’s 1976 review itself acknowledged therapeutic research. (GovInfo)

P4. “Medical benefits were already proven decades ago and then ignored.”

Partly supported only in narrow, product-specific senses; unsupported as a broad claim. Some pharmaceutical cannabinoid indications accumulated randomized evidence and regulatory approvals. Other indications rested on tiny or preliminary studies. Whole cannabis had not been established as a broad treatment by 1999; contemporary official reviews still called for better trials. (GovInfo)

P5. “Findings about pharmaceutical cannabinoids are findings about cannabis.”

Unsupported. Dronabinol, nabilone, levonantradol, smoked cannabis and plant extracts differ in chemical composition, dose, route and pharmacokinetics. Period and official sources treated them separately. (NCBI)

P6. “Government-commissioned reports reached conclusions that governments then disregarded.”

Sometimes partly supported, but too broad as a universal statement. The clearest documented example here is the UK: the 1998 House of Lords committee recommended rescheduling for medical prescription; the Government explicitly declined to accept that recommendation pending stronger clinical evidence. That is documented disagreement, not evidence that the whole report was “ignored.” (UK Parliament)

P7. “A government-held patent on cannabinoids proves the government knew of medical benefits.”

Unsupported as phrased. HHS-linked inventors Hampson, Axelrod and Grimaldi generated preclinical antioxidant/neuroprotection work. A U.S. patent family claims priority to April 21, 1998 and the U.S. application was filed April 21, 1999; the U.S. patent was not granted until 2003, outside P1. The underlying 1998 PNAS evidence was rat-neuron culture/chemical work, not a human clinical demonstration. A patent application establishes a claimed invention and its asserted utility, not clinical proof. (PubMed)

P1 handoff: patent grant/publication in 2003 belongs to P2.

Candidate claims for ledger review

CAND-01

Claim: CBD’s structure was published by Mechoulam and Shvo in 1963. Type: discovery. Evidence: Mechoulam & Shvo 1963, DOI 10.1016/0040-4020(63)85022-X; PMID 5879214; primary × chemical. Population/preparation/etc.: N/A—chemical structure study. Confidence: High. Would change mind: earlier primary publication establishing the same structure with priority. (PubMed)

CAND-02

Claim: Gaoni and Mechoulam published isolation, structure and partial synthesis of the principal psychoactive constituent in 1964 and a total synthesis in 1965. Type: discovery. Evidence: DOI 10.1021/ja01062a046; DOI 10.1021/ja01092a065, PMID 14324315; primary × chemical. Confidence: High for publications; historical “first” priority should remain bounded to verified sources. Would change mind: earlier directly comparable primary publication. (ACS Publications)

CAND-03

Claim: A high-affinity stereoselective cannabinoid receptor was experimentally characterized in rat brain in 1988. Type: discovery. Evidence: Devane et al. 1988, PMID 2848184; primary × mechanistic/in-vitro. System: rat brain membranes/synaptosomes; [³H]CP-55,940; cannabinoid displacement; no human outcome. Confidence: High. Would change mind: methodological evidence showing binding was not receptor-specific. (PubMed)

CAND-04

Claim: A cannabinoid receptor cDNA was cloned and functionally expressed in 1990. Type: discovery. Evidence: Matsuda et al., DOI 10.1038/346561a0; PMID 2165569; primary × mechanistic. Confidence: High. (PubMed)

CAND-05

Claim: Anandamide was isolated from porcine brain and identified as an endogenous cannabinoid-receptor ligand in 1992. Type: discovery. Evidence: Devane et al., DOI 10.1126/science.1470919; PMID 1470919; primary × biochemical/mechanistic. Confidence: High. (PubMed)

CAND-06

Claim: A second, predominantly peripheral cannabinoid receptor was molecularly characterized in 1993. Type: discovery. Evidence: Munro et al., DOI 10.1038/365061a0; PMID 7689702; primary × mechanistic. Confidence: High. (PubMed)

CAND-07

Claim: Independent 1995 publications established 2-AG as an endogenous lipid capable of cannabinoid-receptor binding/signalling. Type: discovery. Evidence: Mechoulam et al., DOI 10.1016/0006-2952(95)00109-D, PMID 7605349; Sugiura et al., DOI 10.1006/bbrc.1995.2437, PMID 7575630; primary × biochemical/mechanistic. Confidence: High. (PubMed)

CAND-08

Claim: In a 1975 randomized double-blind chemotherapy study, oral Δ9-THC produced antiemetic responses whereas placebo did not in the small enrolled cohort. Type: clinical-finding-of-the-period. Evidence: Sallan et al., DOI 10.1056/NEJM197510162931603; PMID 1099449; primary × RCT. Extraction: chemotherapy patients; N=22 entered/20 evaluable; oral Δ9-THC; dose not accessible in abstract; duration = chemotherapy courses; placebo comparator; vomiting/antiemetic response; 14/20 THC courses vs 0/22 placebo courses, completers 12/15 vs 0/14, P<.001. Confidence: High for abstract-reported comparison. Would change mind: full text materially altering denominator/design interpretation. (PubMed)

CAND-09

Claim: The pre-1992 cannabinoid antiemetic literature comprised at least 23 randomized trials identified by a later Cochrane review, mostly crossover trials. Type: literature-shape. Evidence: Smith et al. 2015, DOI 10.1002/14651858.CD009464.pub2; sr-ma × rests on 23 RCTs. Population: adults receiving chemotherapy; nabilone/dronabinol orally; placebo or active antiemetics; nausea/vomiting. Confidence: High for the review’s corpus. Would change mind: eligible omitted RCTs or duplicate-study correction. (Cochrane)

CAND-10

Claim: Oral Δ9-THC produced dose-related analgesic signals in a ten-patient 1975 cancer-pain experiment, but higher doses also caused sedation and mental clouding. Type: clinical-finding-of-the-period. Evidence: Noyes et al., DOI 10.1002/j.1552-4604.1975.tb02348.x; PMID 1091664; primary × controlled clinical trial. Extraction: cancer pain; N=10; oral THC 5/10/15/20 mg; single-dose conditions; placebo; pain relief; 15/20 mg superior to placebo; sedation/clouding at those levels. Confidence: High within this small experiment. Would change mind: full-text design flaws changing comparative interpretation. (PubMed)

CAND-11

Claim: A small 1980 randomized double-blind adjunctive trial reported fewer seizures in several CBD-treated refractory-epilepsy patients than in placebo-treated patients. Type: clinical-finding-of-the-period. Evidence: Cunha et al., DOI 10.1159/000137430; PMID 7413719; primary × RCT. Extraction: secondary generalized epilepsy with temporal focus; N=15; CBD 200–300 mg/day oral; up to 4.5 months; placebo; background antiepileptics continued; seizure course/clinical condition; 4/8 CBD nearly crisis-free, 3/8 partially improved, 1 ineffective; 7 placebo unchanged, 1 improved. Confidence: Moderate because of tiny sample and background therapy. Would change mind: full-text evidence of allocation/reporting bias or reclassification. (PubMed)

CAND-12

Claim: A small double-blind 1981 human spasticity study found 10 mg oral THC reduced clinical spasticity versus placebo, with likely unblinding. Type: clinical-finding-of-the-period. Evidence: Petro & Ellenberger, DOI 10.1002/j.1552-4604.1981.tb02621.x; PMID 6271839; primary × controlled trial. Extraction: principal spasticity group N=9; oral 5/10 mg THC/placebo; acute administration; clinical tone/reflex/EMG outcomes; 10 mg significant P<.01; examiner correctly identified THC trials 7/9. Confidence: Moderate. Would change mind: full-text sequence/carryover evidence or outcome-selection concerns. (PubMed)

CAND-13

Claim: In a 1995 randomized AIDS-wasting trial, dronabinol improved the appetite outcome versus placebo but did not demonstrate a statistically significant weight advantage. Type: clinical-finding-of-the-period. Evidence: Beal et al., DOI 10.1016/0885-3924(94)00117-4; PMID 7730690; primary × RCT. Extraction: AIDS anorexia + ≥2.3 kg loss; 139 randomized, 88 efficacy-evaluable; dronabinol 2.5 mg oral BID; placebo; appetite/weight/mood/nausea; appetite +38% vs +8%, P=.015; mean weight stable vs -0.4 kg, P=.14. Confidence: High for reported comparative endpoints. Would change mind: full-text attrition analysis changing interpretation. (PubMed)

CAND-14

Claim: Aerosolized THC caused acute bronchodilation in a randomized double-blind ten-patient asthma study, but the trial did not address chronic asthma treatment. Type: clinical-finding-of-the-period. Evidence: Williams et al., DOI 10.1136/thx.31.6.720; PMID 797044; PMCID PMC470501; primary × RCT. Extraction: 10 stable inpatient asthmatics; THC 200 μg aerosol vs salbutamol 100 μg vs ethanol placebo; single inhalation on separate consecutive days; one-hour FEV1/FVC/PFR; both active treatments improved ventilation; salbutamol faster; similar at one hour. Confidence: High for acute physiology, low for any chronic inference. (PubMed)

CAND-15

Claim: Human cannabinoid-drug metabolic interaction was directly demonstrated by 1980 at high-dose CBD exposure. Type: harms-research. Evidence: Benowitz et al., DOI 10.1038/clpt.1980.139; PMID 7389248; primary × interventional. Extraction: N=10; CBD 600 mg/day orally, 5–12 days; hexobarbital pharmacokinetics; four received both IV and oral hexobarbital; oral clearance 36% lower on CBD; systemic clearance also 36% lower in IV subgroup. Confidence: High for tested interaction. Would change mind: replication demonstrating analytical error; no generalization to modern doses/drugs. (PubMed)

CAND-16

Claim: A 45,570-person Swedish cohort found heavy reported cannabis use associated with later schizophrenia diagnosis, but observational design did not establish causality. Type: harms-research. Evidence: Andréasson et al., DOI 10.1016/S0140-6736(87)92620-1; PMID 2892048; primary × observational. Extraction: Swedish conscripts N=45,570; self-reported cannabis exposure; 15-year follow-up; nonuser comparator; schizophrenia diagnosis; >50 occasions RR 6.0, 95% CI 4.0–8.9. Confidence: High for association; low for causal wording. (PubMed)

CAND-17

Claim: By the late 1990s, respiratory research showed marijuana smoking could be associated with airway inflammation even though one longitudinal cohort did not find accelerated FEV1 decline. Type: harms-research. Evidence: Roth 1998, DOI 10.1164/ajrccm.157.3.9701026, PMID 9517614; Tashkin 1997, PMID 9001303; primary × observational. Population: young habitual smokers/nonsmokers; bronchoscopy/biopsy/lavage or serial spirometry; smoking exposure; respiratory pathology/function. Confidence: High that endpoints differed and should not be collapsed. Would change mind: better matched longitudinal pathology studies. (PubMed)

CAND-18

Claim: U.S. federal agencies maintained a standardized cannabis research-supply program at the University of Mississippi from 1968, rather than prohibiting all research access. Type: regulatory. Evidence: 1974 Congressional Record; 1970 Congressional Record; official × administrative/historical. Confidence: High. Would change mind: primary contract records giving a materially different start date or role. (GovInfo)

CAND-19

Claim: FDA approved Marinol in 1985 for refractory chemotherapy nausea/vomiting and expanded it in 1992 to AIDS anorexia; the approved formulation moved to U.S. Schedule III in 1999. Type: regulatory. Evidence: DEA 64 FR 35928 (1999); FDA official history. Confidence: High. (GovInfo)

CAND-20

Claim: U.S. Schedule I status did not make cannabis/cannabinoid research impossible, although it imposed regulated access and supply constraints. Type: regulatory/literature-shape. Evidence: federal supply records plus post-1970 controlled human trials. Confidence: High for “not impossible”; burden magnitude requires separate analysis. Would change mind: legal authority showing cited trials were unlawful rather than federally authorized. (GovInfo)

CAND-21

Claim: The 1999 HHS-associated cannabinoid neuroprotection patent application rested on preclinical antioxidant/neuron-culture evidence, not demonstrated human therapeutic benefit. Type: synthesis-conclusion. Evidence: Hampson et al. 1998 PNAS; patent US09/674,028 filed 1999-04-21. Design: in-vitro rat neuronal culture + patent filing; humans none. Confidence: High. (PubMed)

CAND-22

Claim: The earliest disease-modification evidence located here for cancer and cannabinoid neuroprotection remained non-human through 1999. Type: absence. Evidence: Munson 1975 animal/in-vitro; Hampson 1998 in-vitro; documented human searches located no P1 disease-modification trial for those hypotheses. Confidence: Moderate because absence depends on search completeness. Would change mind: eligible pre-2000 human trial. (PubMed)

CAND-23

Claim: The 1998 UK House of Lords committee recommended urgent cannabis trials and medical rescheduling while explicitly acknowledging that rigorous evidence for cannabis itself was not yet conclusive. Type: synthesis-conclusion/regulatory. Evidence: House of Lords Ninth Report; official × synthesis. Confidence: High. (UK Parliament)

CAND-24

Claim: The 1999 IOM report concluded that cannabinoid drugs had therapeutic potential in several symptom domains while recommending further controlled research and distinguishing defined cannabinoids from smoked marijuana. Type: synthesis-conclusion. Evidence: IOM 1999, DOI 10.17226/6376; official × synthesis of clinical/preclinical literature. Confidence: High. (NCBI)

Era table

Areas for which no human study was identified in the documented P1 searches

This is not a statement that none existed.

• Human cannabinoid treatment demonstrating cancer disease modification.
• Human CBD/THC treatment demonstrating prevention or modification of Alzheimer disease, Parkinson disease, stroke injury or other neurodegenerative disease.
• Large, long-duration randomized trials of smoked/whole cannabis for chronic disease.
• Robust pediatric CBD epilepsy programs comparable to the post-2010s evidence base.
• Human trials establishing therapeutic effects of the endogenous cannabinoids anandamide or 2-AG themselves.
• Controlled clinical evidence showing that the 1998 cannabinoid antioxidant/neuroprotection findings translated into human benefit before 2000.
• A period-wide randomized evidence base permitting “cannabis” to be treated as one standardized intervention.

Reference list

Status key: VL = verified live for identity and accessible claim support; VL-abstract = identity + abstract-level support only; period = 1960–1999 primary/period source; retrospective = later source about period.

1. Mechoulam R, Shvo Y. 1963. Hashish—I. The structure of cannabidiol. DOI 10.1016/0040-4020(63)85022-X; PMID 5879214. VL-bibliographic · period. (PubMed)
2. Gaoni Y, Mechoulam R. 1964. Isolation, Structure, and Partial Synthesis of an Active Constituent of Hashish. DOI 10.1021/ja01062a046. VL, access limited to bibliographic/first-page metadata · period. (ACS Publications)
3. Mechoulam R, Gaoni Y. 1965. A Total Synthesis… DOI 10.1021/ja01092a065; PMID 14324315. VL-bibliographic · period. (PubMed)
4. Devane WA et al. 1988. Determination and characterization of a cannabinoid receptor in rat brain. PMID 2848184. VL · period. (PubMed)
5. Herkenham M et al. 1990. Cannabinoid receptor localization in brain. DOI 10.1073/pnas.87.5.1932; PMID 2308954; PMCID PMC53598. VL · period. (PubMed)
6. Matsuda LA et al. 1990. Structure of a cannabinoid receptor and functional expression of the cloned cDNA. DOI 10.1038/346561a0; PMID 2165569. VL · period. (PubMed)
7. Devane WA et al. 1992. Isolation and structure of a brain constituent that binds to the cannabinoid receptor. DOI 10.1126/science.1470919; PMID 1470919. VL · period. (PubMed)
8. Munro S et al. 1993. Molecular characterization of a peripheral receptor for cannabinoids. DOI 10.1038/365061a0; PMID 7689702. VL · period. (PubMed)
9. Mechoulam R et al. 1995. Identification of an endogenous 2-monoglyceride… DOI 10.1016/0006-2952(95)00109-D; PMID 7605349. VL · period. (PubMed)
10. Sugiura T et al. 1995. 2-Arachidonoylglycerol… DOI 10.1006/bbrc.1995.2437; PMID 7575630. VL · period. (PubMed)
11. Sallan SE et al. 1975. Antiemetic effect of delta-9-THC… DOI 10.1056/NEJM197510162931603; PMID 1099449. VL-abstract · period. (PubMed)
12. Chang AE et al. 1979. Δ9-THC as an antiemetic… DOI 10.7326/0003-4819-91-6-819; PMID 293141. VL-abstract · period. (PubMed)
13. Noyes R Jr et al. 1975. Analgesic effect of Δ9-THC. DOI 10.1002/j.1552-4604.1975.tb02348.x; PMID 1091664. VL-abstract · period. (PubMed)
14. Noyes R Jr et al. 1975. Analgesic properties of Δ9-THC and codeine. DOI 10.1002/cpt197518184; PMID 50159. VL-abstract · period. (PubMed)
15. Hepler RS, Frank IR. 1971. Marihuana smoking and intraocular pressure. PMID 5109652. VL-bibliographic; full effect inaccessible · period. (PubMed)
16. Williams SJ et al. 1976. Bronchodilator effect… DOI 10.1136/thx.31.6.720; PMID 797044; PMCID PMC470501. VL · period. (PubMed)
17. Cunha JM et al. 1980. Chronic administration of cannabidiol… DOI 10.1159/000137430; PMID 7413719. VL-abstract · period. (PubMed)
18. Petro DJ, Ellenberger C. 1981. Treatment of human spasticity with Δ9-THC. DOI 10.1002/j.1552-4604.1981.tb02621.x; PMID 6271839. VL-abstract · period. (PubMed)
19. Beal JE et al. 1995. Dronabinol as a treatment for anorexia…AIDS. DOI 10.1016/0885-3924(94)00117-4; PMID 7730690. VL-abstract · period. (PubMed)
20. Benowitz NL et al. 1980. CBD–hexobarbital interaction. DOI 10.1038/clpt.1980.139; PMID 7389248. VL-abstract · period. (PubMed)
21. Andréasson S et al. 1987. Cannabis and schizophrenia. DOI 10.1016/S0140-6736(87)92620-1; PMID 2892048. VL-abstract · period. (PubMed)
22. Tashkin DP et al. 1976. Subacute effects of heavy marihuana smoking… DOI 10.1056/NEJM197601152940302; PMID 1244507. VL · period. (PubMed)
23. Tashkin DP et al. 1997. Heavy habitual marijuana smoking does not cause an accelerated decline in FEV1 with age. PMID 9001303. VL-abstract · period. (PubMed)
24. Roth MD et al. 1998. Airway inflammation in young marijuana and tobacco smokers. DOI 10.1164/ajrccm.157.3.9701026; PMID 9517614. VL-abstract · period. (PubMed)
25. Munson AE et al. 1975. Antineoplastic activity of cannabinoids. DOI 10.1093/jnci/55.3.597; PMID 1159836. VL-abstract · period. (PubMed)
26. Hampson AJ et al. 1998. Cannabidiol and Δ9-THC are neuroprotective antioxidants. DOI 10.1073/pnas.95.14.8268; PMID 9653176; PMCID PMC20965. VL · period. (PubMed)
27. Lemberger L, Rowe H. 1975. Clinical pharmacology of nabilone. DOI 10.1002/cpt1975186720; PMID 1204278. VL-abstract · period. (PubMed)
28. Herman TS et al. 1979. Superiority of nabilone over prochlorperazine… DOI 10.1056/NEJM197906073002302; PMID 375088. VL · period. (PubMed)
29. Cronin CM et al. 1981. Antiemetic effect of intramuscular levonantradol… DOI 10.1002/j.1552-4604.1981.tb02572.x; PMID 7028793. VL-abstract · period. (PubMed)
30. Brewster ME et al. 1997. Clinical pharmacokinetics of escalating IV dexanabinol… PMID 9314087. VL-bibliographic/abstract-limited · period. (PubMed)
31. WHO. 1997. Cannabis: A Health Perspective and Research Agenda. WHO/MSA/PSA/97.4. VL · period · official. (TDR)
32. UK House of Lords Science and Technology Committee. 1998. Ninth Report, HL Paper 151. VL · period · official. (UK Parliament)
33. Institute of Medicine. 1999. Marijuana and Medicine: Assessing the Science Base. DOI 10.17226/6376. VL · period · official synthesis. (NCBI)
34. Smith LA et al. 2015. Cannabinoids for nausea and vomiting… DOI 10.1002/14651858.CD009464.pub2; PMID 26561338. VL · retrospective · sr-ma. (PubMed)
35. Ng JY, Chang YK. 2022. A bibliometric analysis of the cannabis and cannabinoid research literature. DOI 10.1186/s42238-022-00133-0. VL · retrospective · bibliometric. (PubMed Central (PMC))
36. ClinicalTrials.gov NCT00000737. NIAID, HIV-wasting/dronabinol-megestrol. VL · retrospective registry record for period trial. (ClinicalTrials.gov)
37. DEA. 1999. 64 FR 35928, rescheduling FDA-approved dronabinol formulation Schedule II→III. VL · period · official. (GovInfo)
38. Hampson/Axelrod/Grimaldi HHS patent family; U.S. application 09/674,028, filed 1999; US6630507B1 granted 2003. VL · period filing / post-period grant. (Google Patents)
39. O’Shaughnessy WB. 1843 republication, On the Preparations of the Indian Hemp… PMCID PMC2490410/PMC2490264. VL · pre-period primary background. (PubMed Central (PMC))

No retraction or expression-of-concern flag surfaced on the PubMed pages examined. I did not run a complete Crossref/Retraction Watch integrity audit; PROTOCOL-MEDICAL assigns that full integrity check to verification.

Self-flagged weak points

1. Required independent database count missing. The exact PubMed query is supplied, but this interface rejected both the PubMed result URL and NCBI ESearch endpoint. No count was invented.
2. Not all old articles were full-text accessible. Several candidate claims therefore rest on verified PubMed abstracts, and full-text RoB 2 extraction belongs in verification.
3. Minor phytocannabinoid priority is incomplete. THC and CBD chronology is strong; “first isolation/structure” priority for every other phytocannabinoid was not adequately resolved.
4. Compassionate IND dates/patient numbers remain unresolved. Later institutional accounts differ, and I did not obtain the primary PHS/FDA records required by Q12.
5. Synthetic-cannabinoid inventory may not be exhaustive. Dronabinol, nabilone, levonantradol, DMHP and dexanabinol were verified; additional short-lived experimental analogues may have reached humans.
6. Research-bias balance cannot be quantified. I found strong evidence for both federally supported harms research and therapeutic programs, but no comparable period-wide coded publication/grant dataset that justifies a ratio.
7. Global geography remains index-biased. English/PubMed/Scopus-visible research is much easier to retrieve than Soviet/Eastern European, Asian, Latin American and other older non-English literature.
8. Q8 suppression claims are incompletely tested globally. Publication/access is verified for the official reports examined; no universal conclusion about all alleged suppressed cannabis reports follows.
9. Q13’s 1988 ALJ scheduling recommendation needs original-document retrieval. Later Federal Register history confirms the subsequent DEA proceeding, but the original hearing decision itself was not available through the searches I completed.
10. Historical terminology is unstable. Older indexing sometimes uses cannabis, marihuana, THC, cannabinol derivatives or a drug’s specific development code; any one search string will miss some relevant material.

Exact search log

Search date: 2026-09-23 Cutoff honored for retrospective literature: 2026-09-20.

Databases/sites actually consulted

PubMed/MEDLINE pages via live web retrieval; PubMed Central; Cochrane Library; ClinicalTrials.gov; WHO; U.S. GovInfo/Federal Register; FDA; UK Parliament/Hansard; National Academies/NCBI Bookshelf; Google Patents; journal publisher pages.

Core PubMed count query attempted

(cannabis[Title/Abstract] OR marijuana[Title/Abstract] OR marihuana[Title/Abstract] OR cannabinoid*[Title/Abstract] OR tetrahydrocannabinol[Title/Abstract]) AND ("1960/01/01"[Date - Publication] : "1999/12/31"[Date - Publication])

Result count: unavailable — search endpoint inaccessible through this environment. No substitute count estimated.

Principal literature-search strings actually run

Ng Chang 2022 cannabis bibliometric analysis 1829 2021 Scopus 29802 decade 1970s NIH NIDA DOI 10.1186/s42238-022-00133-0

Mechoulam Shvo 1963 cannabidiol structure PMID 5879214 DOI 10.1016/0040-4020(63)85022-X

Gaoni Mechoulam 1964 isolation structure partial synthesis delta-1 tetrahydrocannabinol JACS DOI 10.1021/ja01062a046

Mechoulam Gaoni 1965 total synthesis delta-1 tetrahydrocannabinol PMID 14324315 DOI 10.1021/ja01092a065

Devane 1988 determination characterization cannabinoid receptor rat brain PMID 2848184

Matsuda 1990 structure cannabinoid receptor functional expression Nature PMID 2165569 DOI 10.1038/346561a0

Devane 1992 anandamide Science PMID 1470919 DOI 10.1126/science.1470919

Munro 1993 peripheral cannabinoid receptor Nature PMID 7689702 DOI 10.1038/365061a0

Mechoulam 1995 2-arachidonoylglycerol endogenous cannabinoid ligands canine gut PMID 7605349 DOI 10.1016/0006-2952(95)00109-D

Sugiura 1995 2-arachidonoylglycerol endogenous cannabinoid receptor ligand brain PMID 7575630 DOI 10.1006/bbrc.1995.2437

Sallan 1975 antiemetic delta 9 tetrahydrocannabinol randomized double blind 22 patients PMID 1099449 DOI 10.1056/NEJM197510162931603

Chang 1979 delta 9 THC oral smoked antiemetic 15 osteogenic sarcoma PMID 293141 DOI 10.7326/0003-4819-91-6-819

Noyes 1975 analgesic effect delta 9 tetrahydrocannabinol 10 cancer patients PMID 1091664

Williams 1976 bronchodilator THC asthma Thorax 10 patients PMID 797044

Cunha 1980 chronic administration cannabidiol healthy volunteers epileptic patients PMID 7413719

Petro Ellenberger 1981 treatment human spasticity THC multiple sclerosis PMID 6271839

Beal 1995 dronabinol anorexia AIDS 139 randomized PMID 7730690

Andreasson 1987 cannabis schizophrenia Swedish conscripts 45570 PMID 2892048

Tashkin 1976 heavy marijuana smoking pulmonary function PMID 1244507

Roth 1998 marijuana airway inflammation PMID 9517614

Benowitz 1980 cannabidiol hexobarbital clearance PMID 7389248

Munson 1975 antineoplastic activity cannabinoids mice Lewis lung adenocarcinoma

Hampson 1998 cannabidiol neuroprotective antioxidant PNAS PMID 9653176

Cochrane 2015 cannabinoids chemotherapy nausea vomiting 23 randomized controlled trials 1975 1991

site:clinicaltrials.gov dronabinol AIDS anorexia NCT00000737

Government/regulatory/history searches

Institute of Medicine Marijuana and Medicine Assessing Science Base 1999 ONDCP commissioned cannabinoid drugs pain nausea appetite

House of Lords Science Technology Ninth Report Cannabis Medical Use 1998 reschedule Schedule 2 trials multiple sclerosis chronic pain

WHO Cannabis a health perspective and research agenda 1997 WHO MSA PSA 97.4

FDA dronabinol Marinol approved 1985 1992 nabilone Cesamet 1985 testimony official

DEA dronabinol Schedule III 1999 Federal Register July 2 1999

site:govinfo.gov University of Mississippi 1968 marijuana research NIMH contract cannabis

site:govinfo.gov Arthur D. Little tetrahydrocannabinol NIMH 1970 marijuana research

site:govinfo.gov 57 FR 10499 marijuana scheduling petition denial 1992

site:publications.parliament.uk government response House of Lords cannabis 1998 medical use rescheduling

US6630507 cannabinoid antioxidants neuroprotectants 1999 Hampson Axelrod Grimaldi patent

O'Shaughnessy 1840 1843 Indian hemp cannabis primary

Cochrane and registry access

Cochrane Library pages for historical cannabinoid antiemesis and epilepsy were retrieved. The 2015 antiemesis review’s search had identified 441 records and included 23 RCTs, but those figures are the Cochrane review’s search, not my own database count. (DOI)

ClinicalTrials.gov was searched for historical dronabinol studies; NCT00000737 was retrieved and explicitly shows a November 1999 first submission and 2001 first posting. (ClinicalTrials.gov)

Material not accessible or not completed

• Executable PubMed/NCBI ESearch result-count endpoint.
• Full text for several paywalled 1960s–1980s primary papers.
• Complete primary records for the U.S. compassionate IND program.
• Original 1988 DEA ALJ rescheduling decision.
• Comprehensive non-English database searching.
• An exhaustive list of every experimental synthetic cannabinoid entering humans before 2000.
• Full Crossref/Retraction Watch integrity audit.

Search-stopping rationale

Searching stopped after convergence across five independent evidence layers: period primary chemical/mechanistic papers; period human clinical studies; period harms studies; period governmental/institutional records; and later bibliometric/systematic syntheses used only to map and cross-check the earlier literature. Additional searches were increasingly returning already identified keystone papers or downstream post-2000 work.

The unresolved areas above are explicitly retained as weaknesses rather than inferred away. The largest remaining completeness risks are older non-English literature, minor phytocannabinoid priority claims, the compassionate-access administrative record, and obscure investigational synthetic cannabinoids.

Frozen-submission statement: This is my complete MED01-P1 independent pass. I have not seen the other researcher’s pass. I would not revise this submission after seeing it; any later corrections should enter the verification/corrections workflow rather than silently altering this raw pass.

This is now suitable to hand to Claude as the first frozen raw pass without putting it on Drive.