On August 19, the City of Denver sent a notice to every licensed marijuana dispensary in the city, warning that investigators would be conducting random assessments at about 25 stores in the coming weeks "to evaluate contaminants in products on store shelves."
"Each sample will be tested for pesticides and total yeast and mold by a state- and ISO-certified marijuana testing facility. Results of their respective testing will be shared with each facility and will also be shared broadly within a write-up of results," the announcement read.
Although that write-up is still months away from being released, Westword learned of the tests and analyzed 25 Denver Department of Public Health and Environment dispensary inspection reports filed over a two-day span in September. Of the reports filed between September 9 and September 11, twenty listed at least one or more hold and quarantine orders for cannabis flower, shake or pre-rolled joints, an 80 percent failure rate. Each of those disciplinary actions was tied to plant matter testing above the maximum allowed for total yeast and mold.
The failed plant matter wasn't typically covered in white and gray mold like an old piece of bread, but can still carry potentially toxic fungi that isn't detectable by the naked eye.
DDPHE officials would not confirm whether all 25 of the tests conducted within that two-day period were part of the planned assessment of potential mold and yeast issues, though they acknowledge that some were.
"I would imagine we probably did some routine inspections during that time," DDPHE food and marijuana safety manager Abby Davidson says. "All of our enforcement in our work from the beginning of workings with marijuana facilities is public. So maybe some [reports connected to the assessment] went out before, maybe some went out after."
Still, Davidson adds, "we don't typically do 25 investigations in a week."
DDPHE officials stress that the random assessment inspections were part of a long-planned research project to learn more about the shelf life and packaging of marijuana that had been announced in August, and were not targeting any specific dispensaries. For that reason, when contacted by Westword about the tests, they declined to name the dispensaries whose products were recalled. (Westword found the names in city documents and contacted all of the dispensaries that had failed the testing for comment; seven responded.)
The majority of the dispensaries that failed microbial testing had gotten the flagged marijuana products from wholesale providers, which were not named in the reports. Every commercial grow must send samples of each harvest to state-licensed laboratories for microbial testing before their products arrive at the stores, according to Colorado Marijuana Enforcement Division regulations.
"While there are defined levels of acceptable microbials prior to leaving the grow facility, we are unaware of the existence of acceptable post-delivery microbial levels once the flower product leaves the grow and winds up on our shelves," says one Denver dispensary owner with two strains that failed the assessment on hold.
That's because there is no universal test for marijuana products once they hit shelves. As with food at restaurants, products sold at dispensaries are only inspected by the DDPHE during visits or investigations. But state and local regulators aren't necessarily trained to evaluate the products that marijuana dispensaries sell.
State regulations for marijuana contaminants hinge on what the laboratories are certified to test for.
Jacqueline Collins
"As the state tracking system (METRC) would show us, the product was tested/process-validated and okay to be sold to the public. Otherwise, if the product has been marked as failed, as a retail facility, we would not have even been able to accept it into our facility's system," says another dispensary owner with inventory that failed testing. "If we had any inclination that the product would not have passed testing, we would have not received or purchased the product wholesale. We are debating whether or not to take legal action against the vendor for this inconvenience and loss of business that we have experienced."
According to the DDPHE, the city is reviewing each of the failures on a case-by-case basis, using the state's seed-to-sale tracking system and other resources to track how and when the failed marijuana products became contaminated after passing MED testing.
"There are many different reasons why products may show up on the sale shelf contaminated. It's not that the dispensary that it was sent to had any hand, possibly, in contaminating the product. Or it could be that there were processes that happened after cultivation that maybe would've led to contamination," Davidson explains. "It's really hard to point any fingers until we're able to do our investigation and backtrack to how that product got to that dispensary."
Some of the dispensaries have destroyed the products that failed testing, while others have put them on a quarantined hold. Although the DDPHE will not be retesting the failed samples, it notes that businesses can retest them or re-mediate the plant matter, per MED rules, if they want to try to sell the products again or use them for extraction.
During the assessment, as many as six different strains of flower, trim or pre-rolls were tested at each dispensary. The DDPHE used state-certified marijuana labs and followed methodology from Denver Public Health that included the MED's total yeast and mold standards; initial DDPHE results showed the sample failure rate of flower, trim and pre-rolls tested at between 35 and 40 percent.
Not all of the marijuana industry is on board with this approach. The CEO of one of Colorado's largest dispensary chains, which had shake and pre-rolls fail microbial testing, calls the DDPHE's method "questionable."
"It is important to note that the tests collected by DDPHE were not taken as part of an enforcement action or as a result of any complaints received about our products," he says. "DDPHE collected samples from several Denver dispensaries as part of an assessment of the marijuana industry that is based on questionable scientific principles."
In July 2018, when the DDPHE first announced plans for a baseline assessment of dispensary product contamination at a meeting of the city's Cannabis Health and Safety Advisory Committee, the Marijuana Industry Group — a trade group that represents dozens of Colorado-based marijuana product manufacturers and dispensaries — argued that the MED's minimum standard for total yeast and mold testing, also used by the DDPHE, was too strict and broad, criticizing the process's lack of specificity when identifying the overall amount of yeast and mold colonies.
Davidson agrees that not all the yeast and mold that might show up in a total testing count are necessarily toxic to humans, adding that "to figure that out, you'd have to do some speciation of the mold."
When the city announced the methodology it had chosen for the baseline assessment in early 2019, the testing standards did not include any yeast or mold speciation. According to DDPHE division director Danica Lee, the department focused on the certified testing processes available at the labs, and specificity still isn't part of those processes when examining marijuana plant matter; the labs are set up to comply with MED testing.
"The challenge there is that it isn't part of the MED required process for the lab. The lab systems aren't there," Lee says of more detailed yeast and mold testing. "There is ongoing dialogue about how meaningful those thresholds are. They are what we have to go on, because it's what's in MED rule, but that's an ongoing conversation between industry and regulators alike."
Although scientists know that some yeasts and molds in marijuana are toxic to humans when smoked or ingested, it's still unknown at which point the levels become harmful.
Authors of a 2011 study published by the National Library of Medicine explain that "both tobacco and marijuana are commonly contaminated with fungi," which possibly leads to chronic pulmonary aspergillosis, a severe lung infection, when smoked. However, the test subjects of those studies smoked five to 22 joints per day for decades, and the study's results were still unsure whether mold spores survive the burning process or whether exposure came from "handling the marijuana rather than smoking it."
"Research into the pathogenesis of both marijuana-related lung [infection] and chronic pulmonary aspergillosis are needed to proportionate causal blame in cases such as these," the study reads. Another study published the same year by the NLM calls for more research into potential risk of smoking moldy marijuana and tobacco.
In September, the MED implemented mandatory testing for mycotoxins, a specific microbial that can be harmful when inhaled — but the new requirement was only for concentrate extracted from marijuana plant material that had already failed total yeast and mold testing, and didn't extend to cannabis flower and trim.
Commercial marijuana recalls over mold concerns have dramatically increased since 2017, going from virtually non-existent to the leading cause for marijuana recalls over the past two years, surpassing banned pesticide use. In 2018, around 15 percent of Colorado's marijuana flower and trim failed microbial testing — up from approximately 10 percent in 2017.
On October 14, the DDPHE announced a voluntary recall from Bonsai Cultivation, one of the state's largest wholesale cannabis providers, over failed yeast and mold samples; Bonsai's cannabis and infused products with concentrate extracted from it had been distributed to 144 dispensaries. This recall was one of a handful of city and state mold recalls since 2017; it was not related to the DDPHE's random testing.
Why the increase in moldy pot? In 2014, when recreational sales began, the MED only mandated testing for potency, then moved into pesticides by 2015. Commercial marijuana wasn't tested for yeast and mold until 2017, after enough state testing labs became certified to do so.
The DDPHE views total yeast and mold contamination as another growing pain for the legal pot industry — and one that likely was present in black market and licensed sales before lab testing identified it.
"Any time you have a brand-new industry that is completely unprecedented and comes online as quickly as it did in Colorado, it takes a lot of time and kicking systems of rules and regulations to get things functioning optimally," Lee says. "We can reassure consumers that there are very meaningful assessments that are made at the industry level, and that we follow up. It is an ongoing challenge, and we're not done tweaking everything."
