Last verified: April 2026

Chemotherapy-Induced Nausea and Vomiting — The Evidence

Chemotherapy-induced nausea and vomiting (CINV) was one of the earliest medical applications of cannabinoids, predating the modern medical cannabis movement by decades. Dronabinol (synthetic THC, brand name Marinol) received FDA approval for CINV in 1985 — making it the first cannabinoid approved as a pharmaceutical in the United States. Nabilone (Cesamet), a synthetic THC analogue, followed for the same indication.

NASEM rated CINV as having conclusive or substantial evidence of cannabinoid benefit. This rating was based on a substantial body of controlled trial data, much of it from the 1970s and 1980s when fewer antiemetic options existed. Cochrane-based systematic reviews have calculated a number needed to treat (NNT) of 3.4 for dronabinol in CINV — meaning that for every 3.4 patients treated, one achieves complete nausea control that would not have occurred with placebo alone.

The mechanistic basis is well-established. CINV is mediated through multiple pathways: the chemoreceptor trigger zone in the area postrema, the nucleus tractus solitarius, vagal afferents from the gut, and cortical anticipatory circuits. CB1 receptors are present throughout these emetic pathways, and THC’s agonism at CB1 suppresses the emetic reflex at multiple anatomical sites. Additionally, CBD’s activity at 5-HT1A serotonin receptors may contribute antiemetic effects through a mechanism distinct from THC’s CB1 agonism.

The Ondansetron Problem — Context Matters

The CINV evidence for cannabinoids must be interpreted in historical context. The comparison trials that established dronabinol’s efficacy were conducted primarily against older dopamine antagonist antiemetics — most commonly prochlorperazine (Compazine), metoclopramide, and chlorpromazine. Against these agents, cannabinoids demonstrated superior or equivalent efficacy.

However, the antiemetic landscape transformed with the introduction of 5-HT3 receptor antagonists, particularly ondansetron (Zofran, approved 1991), granisetron, and subsequently NK1 receptor antagonists like aprepitant (Emend). Modern CINV prophylaxis uses triple therapy — a 5-HT3 antagonist, an NK1 antagonist, and dexamethasone — which controls acute CINV in 80–90% of patients receiving highly emetogenic chemotherapy.

Cannabinoids have not been demonstrated to be superior to ondansetron or modern combination antiemetic regimens in head-to-head trials. The available evidence positions cannabinoids as adjunctive therapy for refractory CINV — an add-on for the 10–20% of patients who experience breakthrough nausea despite standard prophylaxis, particularly for delayed-phase CINV (occurring 24–120 hours post-chemotherapy) and anticipatory nausea (a conditioned response that develops after prior chemotherapy cycles and responds poorly to standard antiemetics).

This distinction matters because the claim that “cannabis treats chemo nausea” is technically accurate but clinically misleading if it implies cannabinoids should replace first-line antiemetics. The evidence-based position is that cannabinoids are a valuable second- or third-line option for patients who fail standard therapy — not a replacement for the ondansetron-aprepitant-dexamethasone backbone.

Cannabis-based medicines may be useful for treating refractory chemotherapy-induced nausea and vomiting. However, methodological limitations of the trials limit our conclusions.

Smith et al., Cochrane Database of Systematic Reviews 2015

Multiple Sclerosis Spasticity — The Measurement Paradox

MS spasticity is the other condition that received NASEM’s highest evidence rating, but the underlying data present a measurement paradox that illustrates a fundamental challenge in cannabinoid clinical research: the subjective and objective outcomes disagree.

The primary evidence comes from clinical trials of Sativex (nabiximols), a standardized oromucosal spray containing a 1:1 ratio of THC to CBD extracted from cannabis plants. Developed by GW Pharmaceuticals (now Jazz Pharmaceuticals), Sativex is approved in over 29 countries for MS spasticity — but notably, not in the United States. It is sprayed under the tongue or inside the cheek and delivers approximately 2.7 mg THC and 2.5 mg CBD per actuation.

Three out of four pivotal Sativex trials demonstrated statistically significant improvement on patient-reported spasticity outcomes — primarily the Numerical Rating Scale (NRS), where patients rate their spasticity from 0 to 10. The average improvement was approximately 1.2–1.5 points on the NRS compared to placebo, with approximately 40–50% of patients achieving the pre-defined threshold of ≥20% improvement.

However, when spasticity was measured using the Modified Ashworth Scale — an objective, clinician-administered assessment of muscle tone that physically tests resistance to passive limb movement — the results were not statistically different from placebo. In other words, patients consistently reported feeling less spastic, but when a physician physically examined their muscle tone, the measurement showed no significant change.

Interpreting the Subjective-Objective Split

This discrepancy has been interpreted in multiple ways. Advocates argue that patient experience is the clinically meaningful outcome — if patients feel less stiff, can move more easily, and report improved quality of life, the objective Ashworth Scale may simply be too crude to capture the relevant improvement. Spasticity is a complex phenomenon involving velocity-dependent increases in tonic stretch reflexes, and the Ashworth Scale measures only one component (resistance to passive movement) in a single moment.

Skeptics argue that the subjective improvement may reflect the psychoactive effects of THC — patients feel better because they are mildly intoxicated, not because their underlying spasticity has physiologically improved. THC produces relaxation, mild euphoria, and altered body awareness, all of which could plausibly reduce perceived spasticity without changing the actual muscle tone. This interpretation is supported by the observation that blinding in Sativex trials was imperfect: patients receiving the active spray likely recognized the cannabis effects.

The regulatory response has been pragmatic. Agencies like the UK’s MHRA and the EMA approved Sativex based on the patient-reported outcomes, reasoning that patient experience is itself a valid therapeutic endpoint — particularly for a symptom like spasticity where functional improvement matters more than a physical examination score. The FDA, which requires more stringent efficacy demonstrations, has not approved Sativex, though this may relate to additional factors beyond the Ashworth Scale question.

Other Nausea and Spasticity Applications

Non-chemotherapy nausea: Evidence for cannabinoids in nausea beyond CINV is substantially weaker. There are no large RCTs of cannabinoids for postoperative nausea, gastroparesis-associated nausea, or pregnancy-related nausea. The mechanistic rationale is present (CB1 receptors in emetic pathways are not specific to chemotherapy), but the clinical data are limited to small series and case reports. Notably, the paradoxical condition of cannabinoid hyperemesis syndrome (CHS) — in which chronic cannabis use causes cyclical vomiting — demonstrates that the relationship between cannabinoids and nausea is bidirectional and dose/duration dependent.

Spinal cord injury spasticity: Limited evidence suggests potential benefit for spasticity from spinal cord injury, but the data are far less developed than for MS. Small trials and case series have reported subjective improvement, but no large pivotal trial equivalent to the Sativex MS studies has been conducted for spinal cord injury populations.

Appetite stimulation: Dronabinol holds a separate FDA approval for appetite stimulation in AIDS-associated anorexia, based on trials showing improved appetite and modest weight stabilization. NASEM rated the appetite evidence as limited — below the CINV rating — reflecting smaller trial sizes and less consistent results. The mechanism involves CB1 activation in the hypothalamic appetite circuits, particularly the arcuate nucleus, and peripheral CB1 effects on gut motility and ghrelin release.

Clinical Bottom Line

Cannabinoids are effective antiemetics for CINV, but their role in 2026 is as adjunctive therapy for refractory cases, not as first-line treatment. Patients failing standard ondansetron-aprepitant-dexamethasone prophylaxis have the strongest rationale for cannabinoid addition. For MS spasticity, Sativex reliably improves patient-reported outcomes, which is clinically meaningful even if the Ashworth Scale does not move — though patients should understand that the mechanism may involve subjective perception as much as objective muscle tone change. Both indications have legitimate evidence bases, but neither represents a scenario where cannabinoids are the best available first-line therapy.