Nov. 24 -- A series of cannabis recalls in short succession has industry members questioning the effectiveness of Maine's cannabis testing protocols.
Maine's Office of Cannabis Policy issued its first recall for contaminated recreational products in September. Two more recalls and one expansion swiftly followed suit and more are expected as the office ramps up auditing. All of the recalls were for failed mandatory yeast and mold tests.
On the heels of these recalls, which have contained little information about the circumstances for the failures, industry members are calling for an examination of the state's standards for yeast and mold, which they say are simultaneously too tight, with 10,000 colony-forming units per gram limit, and too broad, encompassing both dangerous and beneficial microbes.
Yeast and mold are the most common contaminants that cause a sample to fail a test, and mold is the hardest test to pass because it's so prevalent in the natural world. But it's important to monitor. According to the Office of Cannabis Policy, inhaling cannabis containing unsafe levels of mold can lead to sinus issues, allergies, headaches, dizziness and fatigue, as well as more serious infections.
Matt Hawes, co-founder of the Maine Cannabis Industry Association, wouldn't take a position on the testing standards -- he's "not a scientist," -- but he said it would be worthwhile to bring in an expert or consultant to look at the regulations.
"Before anyone races to any conclusions around the safety of these products, there needs to be a discussion around the current pass/fail thresholds for some of these analytes," he said. "I think we should be looking at whether we're actually dealing with unsafe product or just noncompliant product. Those could be two different things."
The recalled samples varied widely in the severity of contamination, but so far, the office has not released the numbers publicly. According to Vern Malloch, deputy director of operations for the state office, the two lowest counts were 11,000 and 34,000 colony-forming units per gram -- numbers that would have passed in a state like Connecticut, which has a 100,000-unit limit.
Others were more egregious. One sample came back at 490,000 units per gram, and "in two instances, the lab reported it was simply too numerous to count," Malloch said.
Malloch would not say which numbers correlated with which recall and said he could not provide the certificates of analysis from the labs until the office completed its investigation.
Since the recreational market launched in 2020, Maine labs have tested almost 44,000 samples with an average 6.7% failure rate.
The vast majority of samples failed for yeast and mold. Flower, or "usable cannabis," has a particularly high yeast and mold fail rate of around 15%.
John Hudak, director of the Office of Cannabis Policy, said previously that he stands by the testing regulations, as the same threshold is used by many other states and is the standard recommended by American Herbal Pharmacopoeia, a nonprofit educational organization dedicated to promoting the responsible use of herbal products and herbal medicines.
A WARNING SIGN
Of the 39 states that have a legal cannabis market (medical or recreational), 26 have a blanket test for yeast and mold. The majority, like Maine, have a limit of 10,000 colony-forming units per gram, but five states (Connecticut, Maryland, Michigan, Kentucky and Florida) have a significantly higher 100,000-unit ceiling, according to data from Medicinal Genomics, a cannabis testing, research and analytics company. Only Iowa, Utah, Delaware and Illinois had lower thresholds ranging from 100 to 1,000 colony-forming units.
It's hard to say what the "right" threshold is -- since cannabis is still federally illegal, there's limited research on the plant and no guidelines from an agency like the U.S. Food and Drug Administration.
States more or less copy each other's regulations and then they become status quo, said Yasha Kahn, vice president of marketing and data science for MCR Labs, one of the state's four licensed testing facilities. MCR also has locations in New York and Massachusetts.
"When we look at food, if it's visually moldy, that's when we complain, but with cannabis, since it's inhalable, it might be more dangerous to us, which is why the limit used needs to be lower than what's visible," he said.
Kahn didn't know exactly how many colony-forming units would create visible mold, but he said it was likely closer to 1 million.
"Nobody is going to find a study that says 12,000 or 22,000 is more dangerous than 21,000 or 9,000, but the purpose for this test is not to identify what's harmful and what is not," Kahn said.
Instead, total yeast and mold tests work as an indicator for a larger problem, like unsanitary growing conditions. He compared it to inspecting restaurants and finding urine on the kitchen floor. There may not be a study that says urine all over the kitchen floor has ever harmed someone, but seeing it is a quick way to identify a major problem within the facility.
"The reality is, maybe a little bit is fine because a mouse ran through or a fly blew through and peed a little bit. That's not a problem, but if the entire floor is covered in mold or urine, that's where we need to discuss. We don't know exactly how dangerous this is, but we do know there's a problem."
GOOD AND BAD MOLDS
There are thousands of species of mold, many of which are harmless. Some are beneficial for the plant. Others, though, are very dangerous. The problem is that total yeast and mold tests don't differentiate between good and bad.
That's why some argue for species-specific tests.
Mark Barnett, a cannabis business owner and policy director for the Maine Craft Cannabis Association, said that without speciation, "there's no conversation to be had about testing."
"If these were pesticide and heavy metal failures or the introduction of some toxic agent, then I would be concerned about the supply chain and these companies that have seen the recalls and failures," Barnett said. "But living plants have yeast and mold on them."
Malloch, at the Office of Cannabis Policy, said he's only recently started hearing questions about testing for specific mold species.
"It's not something that's done now and it's not the kind of thing where we have a position on it," Malloch said. "We're always interested in updating the standards for keeping folks safe, but that is not something that's part of the testing program right now."
Twenty-four states test for the presence of various species of Aspergillus, a type of mold that can be dangerous or even fatal in high doses, particularly for immunocompromised patients. Maine does not have a separate test for Aspergillus, but if a sample fails for yeast and mold, the product will be retested for certain compounds called mycotoxins, which are produced by certain molds, including Aspergillus.
Eight of the states that test for Aspergillus, including California and Vermont, do not have a separate yeast and mold requirement. Arkansas and Washington don't test for either.
Neither test is perfect, Kahn noted. Total yeast and mold could account for beneficial molds that cause the test to fail, or it could result in passing a sample that had dangerously high amounts of Aspergillus, but not above the total threshold. And an Aspergillus test would miss other potentially dangerous microbes. Plus, there's the additional cost that would add to an already expensive process.
TESTING THE LIMITS
Christopher Altomare, co-founder of Nova Analytic Labs, another of the state's testing facilities, is trying to figure out the right balance.
"Basically, we want to answer the question, is the 10,000 limit appropriate? And based on the molds that are commonly found in cannabis, does that limit make sense? Do we need to have limits based on specific pathogenic strains?" he said.
Altomare and others at the lab have already done some research. For example, a study of over 300 cannabis samples from across the state identified 48 species of mold, and the species quantity varied widely across counties.
But there are still too many unknowns for a sweeping policy change.
Malloch, the deputy director at the Office of Cannabis Policy, said that aside from the important health and safety questions, there are still gaps in the information, like how many species the office would need to add tests for, or if the labs would need to buy new equipment or hire new staff. With a higher limit, business owners would be more likely to pass a yeast and mold test, but how much more would the tests cost?
"We have to talk to the Legislature about the fiscal impact on licensees (if) they're doing more detailed testing," he said.
There are costs associated with a lower limit, too. If a sample fails, a licensee can choose to have it tested again. Or, providing it didn't fail for something like pesticides or heavy metals, the cannabis can be remediated -- blasted with radiation or ozone to kill the bad microbes -- before testing again.
Some business owners have started irradiating their cannabis before the testing process to reduce the risk of failure.
Cannabis flower and trim that is remediated or treated with radiation or ozone needs to be labeled as such, following a law that went into effect this month.
A LACK OF TRANSPARENCY
Industry members have also complained about the lack of transparency surrounding the recalls and testing in general.
The state has so far only released vague information, including the strain, batch number and location of the affected plant, but not the results that led to the failure, how many individual items were sold and how the product reached the market.
The Office of Cannabis Policy publishes quarterly and annual data about the number of tests completed, the fail rate and nature of the failures.
Maine does not, however, release lab-specific testing data or certificates of analysis, which critics say can cover up fraud within the testing industry.
Some labs, for example, may artificially inflate potency or decrease yeast and mold results to court more clients in an effort known as "lab shopping."
Malloch said the Office of Cannabis Policy has examined the certificates of analysis but found no evidence of lab shopping.
Altomare and Kahn, however, are not convinced and argue the data should be public because it protects the labs, the consumers and the state agency, provided everyone is doing what they're supposed to.
If the Office of Cannabis Policy is going to demand more accountability by conducting more audits, it's reasonable to ask the office to make sure the state's regulations are both appropriate and being followed by all parties, Altomare said.