This article outlines why multiple testing modes are important, what to look for in third party test providers, and how a cultivator can evaluate these potential providers both at the outset and in an ongoing manner for accuracy and reliability.
Growing plants on an industrial scale brings both opportunities and challenges compared to small scale “home grow” operations. Cannabis is no exception to this, and large producers can achieve economies of scale and production levels far past what historical illicit grows could achieve. They also run increased risks inherent to all large-scale monoculture—it’s a prime ground for pathogen and pest issues which can decimate a crop. Lacking (in most jurisdictions) the chemical treatment tools and techniques traditional agriculture has available to combat these, effective screening of input materials and ongoing surveillance of crops to catch problems early are essential to risk mediation. Most of those strategies will rely on extensive third-party testing, and the accuracy of this is critical. Similarly, genetic testing for cultivar identification and assurance is an important foundational key—knowing incoming seedlings or cuttings are what’s expected at an early stage of growth avoids any unpleasant surprises after months of growth and resource utilization. This article outlines why these multiple testing modes are important, what to look for in third party test providers, and how a cultivator can evaluate these potential providers both at the outset and in an ongoing manner for accuracy and reliability.
With legalization of cannabis cultivation the shift from small illicit grows to massive greenhouse facilities provides potential for production efficiencies, but this doesn’t come without risks and challenges. Large scale monoculture in controlled, highly uniform environments provides a near perfect setting for the adaptation and emergence of pests including viruses, viroids, and bacterial or fungal microorganisms. The impact of these can range in severity from minor crop yield losses or alterations in desired chemotypic profile to catastrophic loss of entire crop cycles. Pest control challenges are further compounded for cannabis growers because in most jurisdictions very few pesticides are approved for use because of public health concerns and a lack of data regarding application of any pesticide on cannabis specifically. Effective pest management practices must instead rely on cultivation techniques such as tissue culture from meristem followed by ongoing testing supported by quarantine and culling as needed to mitigate risks. An underlying question is how to get this crucial testing done reliably. False negative results allow expansion of contamination until phenotypically visible (too late), and false positives can lead to culling of healthy crops (needless loss of revenue). What tests should a cultivator do in house, versus send out to a third party laboratory? How does a cultivator establish trust in third-party laboratory results? What factors influence test accuracy in either case?
Let’s tackle this first—as a cultivator, should you do in house pathogen testing or send samples out? There’s no “one size fits all” answer for this, so let’s review factors a cultivator should consider in making this choice.
The above process should be done for every target desired, and may give different answers for different pathogens. As a very general rule of thumb, antigen based testing can make sense in-house at low numbers as it requires little to no infrastructure, specialized staffing, or large capital outlays. Molecular testing will need all of these and unless the equipment and staff is used at near maximum capacity, it can’t compete with cost effectiveness of a commercial “core laboratory.”
Regardless of where tests are run, it’s important to understand a frustrating truth: no assay is perfect, and they all balance sensitivity (% true positives detected as positive) versus specificity (% true negatives detected as negative). A more detailed discussion of this topic is beyond our scope here but essentially, from that optimum balance, even small increases in sensitivity bring large increases in false positive rates. In reality, good tests are often in the 97–99% sensitive range. It’s important to understand the real-life implication of that—one in 30 to one in 100 “negative” results is actually positive.
If your considerations lead you to looking for external testing, it’s important to pick a service provider you can rely on. Don’t hesitate to ask them some questions such as these:
If your candidate laboratory can provide reasonable answers to these questions, that’s a great start. As people in the quality assurance (QA) field say though, “trust but verify.” Good next steps in initially validating a laboratory would be to send multiple blind-labeled replicates of samples (that is, the test laboratory won’t know them). If you send three replicates each of 10 samples (ideally, five each positive and negative), and all replicate sets are internally consistent, and sets report as expected, that’s excellent. If, however, it looks like someone’s flipping a coin to determine within or across sets what’s positive and what’s negative, your only possible conclusion can be that you can’t have much confidence in the results.
In addition to doing a process like this as an initial qualification test on a service provider, it’s worth doing on a smaller scale at intervals. Perhaps once a year, slip a few opaquely labeled known replicate samples into your test stream, and see if you’re continuing to get consistent valid results back.
These sorts of external quality Assurance processes are mandated norms in most medical testing. For what it’s worth to your operation, you should be using this approach, too.
This was covered above, but it’s so critical that it bears reiteration in a slightly different format. Regardless of whether you perform tests in house or send out, you will sometimes get wrong results.
False positive results are often the easiest to detect, if negative controls go positive. They’re also most easily prevented, as they usually arise from poor (and correctable) laboratory practice—cross contamination from other samples or positive controls being the most likely culprit, followed by cross-reaction to off target species.In this context, real-time (RT)-polymerase chain reaction (PCR) methods are far superior to obsolete endpoint gel resolved methods and highlights value of specificity tests in assay validation.
False negative results can be much more insidious, as they can arise from a number of causes. Some such as sample degradation (remember RNA instability) can be partially protected against with internal controls. Other causes, such as low level PCR inhibitors—enough to mask a low-level true positive but not enough to kill the assay positive control test—can go undetected. For molecular assays, mutations under (RT)-PCR primers (and probes, for probe based assays) can either block amplification or detection of targets, respectively, and do so in a way which looks fine to all reasonable controls. If you’re dealing with an RNA target, this is almost guaranteed to happen at some point and in some frequency of samples. Variability in pathogen load around the plant can be an issue too; if your sample isn’t from an actively infected section, it may test negative yet not accurately represent what’s going on elsewhere in the plant (and likely spreading). The key here is to remember, absence of proof is not proof of absence! Strictly speaking, assays of the type discussed here should be reported as “Detected/ Not Detected.” “Not Detected” is not synonymous with “not present,” and never can be. Failure to understand this nuance can lead to unpleasant consequences.
Regardless of test results, don’t ignore the evidence of your eyes. If a plant is visibly sick, no number of “Not Detected” test results will make it healthy or reduce risk of it transmitting something—maybe something novel you’re not testing for—to the rest of your grow. Isolate or destroy it early to minimize risk! Similarly, use some judgment in cases of positive results. While immediate quarantine of suspected positive plants is a great idea, if they don’t show overt disease, bear in mind molecular methods can detect non-viable pathogen remnants, or perhaps that variety has an effective resistance mechanism, and will continue to grow and flourish even in presence of low pathogen load. As cannabis pathogens and their behavior in large scale cultivation are better studied, all of this will gain further meaning wherein some pathogens will be known immediate critical detections (“there’s no such thing as a harmless XYZ”) as opposed to more opportunistic pathogens (“FGH is frequently seen in this context and at low levels doesn’t negatively impact growth”).
Effective testing is key to getting and maintaining healthy large scale cannabis grow operations. Consider your needs and resources in determining what to test in-house versus what to send out. Evaluate potential external test providers with questions and with blinded replicate samples. Understand real life limitations of testing methods, and be ready to retest or challenge results that don’t match observed reality.
Dr. John Brunstein is a leading expert on biotechnology, molecular biology, and clinically related topics including clinical trials, assay validation methods, and quality systems in regulated industries. He obtained his PhD in Biochemistry from the University of British Columbia, working on the molecular biology of viruses. He currently serves as a Scientific Advisory Board Member for Segra International, an agriculture technology company offering plant tissue culture, plant genomics, and pathogen detection services to accelerate the advancement of the cannabis industry. Learn more at segra-intl.com.
J. Brunstein, Cannabis Science and Technology 5(1), 40-43 (2022).