The eventual hope, van Beeman said in a phone interview Friday is to produce pills made from purified hemp extracts that people could take to lower the risk of being infected by COVID-19 after a known exposure to the virus. There’s also a possibility that those with compromised immune systems or who work in high-risk fields like health care could take them on an ongoing basis.

The approach holds particular promise because the cannabis compounds identified in the study — CBDA, CBGA, and THCA — are already known to be bioavailable through oral administration, and have extremely low risk profiles and long histories of safe human use compared to many more novel, synthetic medications. (The molecules can be thought of as the “raw” non-psychoactive precursors to their more familiar, de-acidified forms such as THC and CBD, which are created when marijuana is smoked or processed into edibles.)

However, many promising scientific breakthroughs fail to make the leap from petri dish to people. The efficacy of hemp supplements in preventing COVID-19 will have to be carefully studied in the real world before doctors start recommending them to patients.

And even if the supplements work, mass-manufacturing a pill that delivers consistent doses of the compounds could be tricky. Cannabinoid levels vary dramatically among different strains of hemp and marijuana, and they will have to be carefully extracted without applying heat, most likely using a solvent such as ethanol.

Another wrinkle: THC and its precursor THCA remain heavily restricted under the federal prohibition on marijuana, which prevented the Oregon researchers from fully testing its potential against the coronavirus even in the lab. Human trials of the drug — much less a commercially available supplement — would require various federal approvals, likely an onerous and time-consuming process. CBGA and CBDA can both be extracted from hemp, an ultra-low-THC variety of cannabis that was legalized nationally in 2018, but typically occur at lower concentrations in the plant.

Still, van Beeman said, “after so many decades of being prohibited from studying hemp, we were really pleased to be able to do that today.”

The origins of the study lie in advanced computer modeling of cellular-level interactions. Along with a new mass spectrometry technique for analyzing the cannabis molecules, the technology allowed van Beeman and his colleagues to quickly come up with a short list of natural compounds that were most likely to have a high affinity for binding with the “spike protein” of the coronavirus, in turn blocking it from infecting human cells. In addition to the cannabinoids, they found several other candidates in nature, including saponin compounds that occur in licorice.

Tafesse initially had no idea he was working with cannabis-derived compounds, an intentional step the researchers took to preserve the integrity of the experiment. He said was taken aback by the strong binding action of the relatively small and simple CBGA and CBDA molecules, which suggested modest doses could be effective at preventing infection, and compared them to currently available monoclonal antibody treatments that also work by blocking the protein spike of the virus.

“To me it was very surprising — I knew cannabis had some interesting compounds from a scientific point of view, but I wasn’t involved in that world at all,” he said in a phone interview Friday. “I never thought it would have an effect on [coronavirus] infections. There’s so much nature can provide us.”

Like van Breeman, however, he cautioned that cannabis is far from a cure-all for the contagious and deadly virus, and called on scientists to carefully design upcoming human trials so they account for vaccination status and other potentially confounding variables.

“My anticipation is that these compounds will not be as effective as a vaccine,” Tafesse said. “But for those subpopulations where the vaccine may not be working — for example, cancer patients and individuals with autoimmune diseases who are not able to produce antibodies even if they’re vaccinated — this could be extremely promising.”

Dan Adams can be reached at daniel.adams@globe.com. Follow him on Twitter @Dan_Adams86.