By Dr. Benjamin Caplan, MD | Board-Certified Family Physician, CMO at CED Clinic | Evidence Watch
Dealing with a condition like this?
Dr. Caplan has worked with 30,000+ patients on conditions like this. A consultation starts with your specific situation — not a generic protocol.
Book a consultation →A small Italian study of 58 chronic pain patients found that a common gene variant involved in dopamine metabolism (COMT 680T>C) was statistically linked to how much THC reached the bloodstream, and that the brain-health protein BDNF correlated with cannabinoid plasma levels. However, the study is too small, lacks a control group, and did not correct for multiple statistical comparisons, meaning these results should be treated as early hypotheses rather than established clinical facts.
Small Study Links Genetic Variant and Brain Protein Levels to Cannabis Drug Exposure, But Results Are Preliminary
A cross-sectional analysis in 58 chronic pain patients finds the COMT 680T>C dopamine-metabolism variant associated with delta-9-THC plasma levels, and neuroprotective biomarkers BDNF and neurofilament light chain correlate with cannabinoid concentrations, though the study’s small size, absence of a control group, and uncorrected multiple comparisons mean all findings remain hypothesis-generating only.
#28
Limited Direct Relevance
Intriguing pharmacogenomic signal but far too preliminary for any change to clinical cannabis prescribing or dosing.
Medical Cannabis
Chronic Pain
BDNF Biomarker
COMT Gene Variant
The promise of pharmacogenomics-guided cannabis prescribing is that clinicians could one day predict who will respond to a given cannabinoid regimen, who may need dose adjustments, and who faces elevated risk of adverse effects. Currently, however, the evidence base for such precision dosing in cannabis medicine is remarkably thin. Identifying genetic variants and circulating biomarkers that reliably predict cannabinoid plasma exposure would represent a meaningful step toward individualized therapy, particularly in chronic pain populations where treatment response varies enormously and titration remains empirical.
Cannabis-based medicines produce highly variable plasma concentrations across individuals, and the sources of that variability remain poorly characterized. This Italian study sought to determine whether common genetic polymorphisms in enzymes related to cannabinoid and catecholamine metabolism, alongside circulating neuroprotective and inflammatory biomarkers, might explain some of that inter-individual variation. The investigators enrolled 58 cannabis-naive adults with neuropathic or chronic pain at a single pain clinic in Turin, collected trough blood samples at confirmed steady state after at least 15 days of once-daily medical cannabis use, and measured cannabinoid concentrations by UHPLC-MS/MS alongside BDNF and neurofilament light chain (NFL) by ultrasensitive Simoa assay and IL-10 by ELISA.
Among the genetic variants examined, the COMT 680T>C polymorphism (rs4680, Val158Met) showed a statistically significant association with delta-9-THC plasma levels (p=0.017). BDNF and IL-10 concentrations differed significantly between patients using oral decoction versus inhaled cannabis (p=0.004 and p=0.009, respectively), and both BDNF and NFL correlated with cannabinoid plasma concentrations. However, these p-values were not corrected for the multiple comparisons performed, substantially inflating false-positive risk. There was no untreated control group, effect sizes and confidence intervals were not reported in the available text, and the full methods section was truncated. The authors themselves describe the study as preliminary and explicitly call for replication in larger, more rigorously designed cohorts before any clinical inferences are drawn.
I appreciate the direction this group is headed. The idea that a patient’s COMT genotype might partly explain why two people on identical cannabis regimens end up with very different plasma THC levels is biologically compelling and, if replicated, could have real clinical utility. The analytical platforms used here are genuinely rigorous. But the gap between an uncorrected p-value of 0.017 in 58 patients and a validated pharmacogenomic dosing tool is enormous. Without correction for multiple testing, without a control arm, and with heterogeneous routes and formulations mixed together, these associations remain fragile. I would not want anyone to see this and conclude that COMT genotyping should guide cannabis dosing today.
In my own practice, I still rely on careful clinical titration, patient-reported outcomes, and close follow-up rather than pharmacogenomic panels for cannabinoid prescribing. I do find BDNF and neuroinflammatory markers intellectually interesting as potential tracking biomarkers for neuropathic pain patients on cannabis, but we are nowhere near being able to use them for real-time dosing decisions. The right response to this paper is to watch for the larger replication study, not to order new tests.
For clinicians following the pharmacogenomics-of-cannabis literature, this study sits squarely in the hypothesis-generating phase. It is among the first to combine validated ultra-high-performance liquid chromatography with mass spectrometry for cannabinoid quantification and Simoa-based neurodegenerative biomarker measurement in a medical cannabis pain population. That methodological rigor distinguishes it from lower-quality observational work, but the study design itself remains fundamentally limited by its cross-sectional structure. The observed COMT association, while pharmacologically plausible given dopamine-endocannabinoid pathway crosstalk, does not establish that COMT genotype drives THC metabolism or exposure in a causal sense.
From a safety and drug-interaction standpoint, the findings raise no immediate prescribing concerns, but they do underscore how poorly we understand cannabinoid pharmacokinetic variability. Clinicians should be aware that COMT inhibitors (used in
