Cryptococcus neoformans brings a fatality to the medical Cannabis industry
How was it missed and does radiation help or hurt?
We have been predicting this sad event would unfold at Medicinal Genomics for some time. The risks are all over the literature but cannabis fatalities are rarely tracked by any federal registries due to its illegal nature. As a result our community loves to opine on how no one has ever died from cannabis but this assurance is resting on a common sampling frequency problem: no one is tracking these harms. As the time tested logic holds, “Absence of evidence isn’t evidence of absence”.
There is no VAERs for cannabis. The federal government is not keeping a ledger of these events and they were imposing massive social injustices against patients that admit to cannabis use including CPS taking children and pulling patients off of organ donor or recipient lists. Most of our history with cannabis, patients were motivated to hide their use.
For this reason much of the harms of cannabis are as hard to find as government funded literature finding its benefits.
Today that ended. Some one has finally tied the genomics of a fungal pathogen from a cannabis sample to the death of a patient and it didn’t happen with some black market boof, gas station hemp or recreational cannabis. It came from a medical grade dispensary in Pennsylvania.
This paper has created quite a stir in the cannabis community as this patient used Medical grade Cannabis from PA and the knee jerk reaction is to throw the baby out with the bath water.
“Screw testing, doesnt keep you safe anyway”.
So why did testing miss this?
PA has a 10,000 CFU/g regulation for TYM and is a state known to employ a lot of irradiation of the product. So how did this slip through?
It appears the traditional 48hr testing wasn’t sufficient. Sonication of the sample and 4 days of growth on specialized media eventually picked it up.
This was confirmed with ITS PCR and Sanger Sequencing. Many of the TYM qPCR tests in the market also target these ITS regions but they are designed to fail products at 10,000 CFU/g TYM and would likely pass this low CFU sample.
They then went on to sequence the fungal strain and compare it to a wild type C.neoformans biobank submission and found only 5 SNPs between the patient and the cannabis but ~42,000 variants from a different patients C.neoformans and ~8,700 SNPs from a Biobank C.neoformans control. This really pins down the exposure as not being an environmental coincidence but a direct Cannabis to patient route of exposure.
You will note the colony counts they found were loaded with other fungi and very few were C.neoformans suggesting a 10,000 CFU/g TYM test will never capture this risk and attempts to lower this to 1 CFU/g TYM will simply fail all cannabis. Species specific tests are really needed to target this risk and not fail all cannabis and divert it into less regulated hemp markets.
When we scan 1,244 Cannabis Whole genome shotguns, C.neoformans is at very low frequencies in the Medicinal Genomics cannabis Microbiome database (blue line below- note log scale on Reads Per Million).
C.neoformans is notoriously radiation resistant. It was found in the Chernobyl reactor site as the dominant colonizing fungi after meltdown.
This does pose several questions. Is the use of radiation selecting for radiation resistant microbes like C.neoformans that are more difficult to detect with traditional plating methods? PCR and sequencing can see it but species specific single CFU/g tests are going to be required to avoid failing every cannabis sample at 1 CFU/g TYM.
We do know the MGC TYM assay will amplify C.neoformans but if the sample is under 10,000 CFU/g it will still pass.
Many years ago, Agencourt Biosciences was hired by the Jet Propulsion Laboratory to sequence the microbiome of the Mars Rover before they sent it to Mars. They did not want to colonize Mars with Earth-borne life and contaminate the question of Life on Mars. During this project, they realized they were making their problem worse with continued radiation of the rover before launch. They were selecting for radiation resistant microbes and thus selecting for the very microbes most likely to survive the flight to Mars. They stopped doing this and instead relied on just cataloging what was present before launch.
We don’t know if radiation was used on these samples. This is why having this information on the label is important for studies like this. PA is known to use a lot of radiation on Cannabis but its not clear we can track that use to these particular samples.
When black market samples are tested compared to regulated markets we see cleaner cannabis, albeit problems still exist. Attempts to tighten the regulations on cannabis labs is often met with resistance from growers as the cannabis market is always competing with the hemp markets that require no testing and very stringent TYM testing regs have induced lab shopping in the cannabis markets.
To balance these issues, we think this is another example that underscores the need for the industry to move to species specific testing. We simply can not discern clinical risk with non specific tests like TYM set at arbitrary 10,000 CFU/g testing limits.
We are downloading this genome to see if we can find signature for gamma irradiation in the genome.
From ChatGPT5.o
Short answer: not really C→T. That’s more the UV “dipyrimidine” signature.
For gamma/ionizing radiation, the hallmark is indel‑heavy damage from double‑strand breaks, not a clean SNV pattern.
What to look for:
Excess small deletions (∼1–100 bp), especially 1–5 bp deletions, with microhomology at the junctions → matches COSMIC indel signature ID8 (error‑prone NHEJ/MMEJ).
More structural variants, notably balanced inversions and simple/complex rearrangements.
SNVs are weak/secondary: you may see some ROS‑type changes (e.g., C>A, T>G; SBS18/36) but not the UV‑like C>T at dipyrimidines.
Practical checklist for your C. neoformans data:
Compare indel spectra vs controls; test for enrichment of 1–5 bp deletions and microhomology usage at breakpoints (over random).
Quantify balanced inversions/DSB‑type SVs.
If you profile SNVs, check whether C>A/T>G (oxidative) nudges up, but don’t expect a crisp C→T transition peak.
So if gamma was used, expect a microhomology‑flavored deletion/SV signature, not a dominant C→T transition pattern.
Appendix.
Since some have asked if this risk is only for immonocompromised patients.
I have bad news.
C.neoformans first hit our radar when this immuno-competent patient got meningitis. Non-fatal but still worrisome.
This patient didn’t directly die from the C.neoformans but to treat C.neoformans many antifungals were required which are immunosuppressant and can lead to sepsis and pneumonia.
Yes — the same mechanisms that make that triple therapy (liposomal amphotericin B + flucytosine + high-dose fluconazole) risky for urosepsis also increase the risk for pneumonia, both bacterial and fungal, especially in the first few weeks of treatment.
How this regimen can predispose to pneumonia
1. Direct myelosuppression
Flucytosine is the main culprit — it can cause neutropenia and other cytopenias.
Neutrophils are critical for controlling bacterial and fungal lung pathogens.
When neutrophil counts drop, even commensal flora or hospital-acquired organisms can cause rapidly progressive pneumonia.
2. Indirect immunosuppression
Amphotericin B → nephrotoxicity → reduced clearance of flucytosine → higher marrow toxicity.
High-dose fluconazole → rare but possible leukopenia + microbiome disruption, allowing airway colonizers to overgrow.
3. Hospital & treatment environment
These patients are usually:
Hospitalized or have frequent visits.
Exposed to ventilators, nebulizers, or supplemental oxygen equipment.
In contact with healthcare-associated pneumonia (HAP) pathogens (Pseudomonas, Klebsiella, MRSA, etc.).
4. Specific pneumonia risks
Bacterial: Streptococcus pneumoniae, Haemophilus influenzae, Staphylococcus aureus, Pseudomonas.
Opportunistic fungi: Aspergillus spp., Candida (rare primary pneumonia), Mucorales in severely neutropenic patients.
Viral: Reactivation of herpesviruses or severe influenza due to reduced immune surveillance.
Clinical picture
Onset can be subtle in neutropenic patients — little sputum, mild CXR changes, but rapid decompensation possible.
Fever, hypoxia, or new infiltrates in this setting = empiric pneumonia coverage until proven otherwise.
Risk reduction steps
CBC monitoring at least twice weekly during induction therapy.
Adjust flucytosine dosing if renal function declines.
Strict pulmonary hygiene — incentive spirometry, mobilization.
Avoid unnecessary aerosol-generating devices unless sterilized.
Early empiric coverage if respiratory symptoms develop during neutropenia.
💡 Bottom line:
Yes — this combination can increase pneumonia risk, mostly through flucytosine-related neutropenia worsened by amphotericin’s renal effects. The high-risk window is the same as for urosepsis: when neutrophils are low and the patient is in a hospital environment with high exposure to opportunistic pathogens.
If you want, I can make you a side-by-side table showing the pneumonia vs urosepsis risk pathways on this regimen — it makes it easy to see how the same underlying drug effects lead to different infection sites. Would you like me to put that together?
One of my first literature reviews, in the early 1990s, discussed the benefits and risks of high gamma dose food irradiation.
Apart from lowering the nutritional value it not only failed to remove any pathogenic toxins, but it also selected for more pathogenic strains, just as per the reported case.
Was the individual in question cv19 injexed? I'll take my chances any day with cannabis as opposed to the toxic brew they call medicine.