The Challenges and Workflow Dilemma of High-throughput Multi-residue Pesticide

Hello ladies and gentleman, my name is David hills from separation science. And I’d like to welcome you all to our latest webinar in collaboration with Thermo Fisher Scientific. Our topic today is high throughput multi residue pesticide screening program in a quality control environments, the challenges and workflow solutions. Let me introduce our presenter for today’s webinar. Kathy BANA Shecky is a method development supervisor at NOW Foods where she leads the development and validation of analytical methods. Katie has degrees in biotechnology and chemistry from William Paterson University. And prior to joining NOW Foods, she worked at the Institute for Food Safety and Health and Mars Snackfood us. She specializes in mass spectrometry applications and residue analysis. Katie has led the development and implementation of a routine pesticide residue monitoring program at NOW Foods and plays an integral role in growing the company’s analytical capabilities. Katie authored and co authored multiple publications and is continuously contributing to the scientific community in the area of research on natural products. And so, without further ado, I’ll turn the presentation over to Katie banner Shecky. Katie, thank you, David, for this kind introduction. And of course, for the opportunity to present the work we do up now. For those not familiar with the brand, NOW Foods is a manufacturer of dietary supplements, sports nutrition, personal care items and suits. NOW Foods is one of the top selling brands in the health food stores and online and award winning manufacturer and a leader in the fields of nutritional science and math development. We are continuously growing within the natural products industry. And we’re also contributing to the science community in the area of research on natural products. But the growth, the awards and the recognition are the effects of our carefully developed and implemented quality sensor systems. One of the most important aspects of formulating and manufacturing great quality and efficacy supplements is the quality control of the ingredients. So what makes a great botanical ingredient great quality botanical ingredient there’s a couple of key components that contribute to the control of botanical ingredient quality, first and foremost, establishing partnerships with dependable quality conscious suppliers. Also implementing good manufacturing practices is important in a reputable manufacturing companies also creating ingredients specifications which would list the plant species part and the potency of the chemicals that that is being claimed. But also testing botanicals for contaminants such as heavy metals and pesticides, plays a very very important role in controlling the quality of botanical ingredients. So how does quality control look like at now so we do have a very comprehensive vendor qualification probe program in place, which ensures that we only source the preeminent quality ingredients. We comply with 21 CFR one a lot of cGMP to make sure that the products that we manufacture are made according to the standards, the best standards in the industry. And something that I’m absolutely proud of is our world class labs and science teams to make sure that the products are safe from adulteration, contamination and ingredient substitution by implementing a very routine stringent testing program. As we learn the health promoting properties, botanical of botanicals and try to understand the mechanism of action we also learn that the active constituents are not the only ones we must be paying attention to. So besides health promoting components botanicals may contain contaminants such as pesticides which are more and more concerning. If we look at the society and people being more and more conscious about what they’re actually consuming, the widespread use of pesticides makes the appearance of the residues in conventional foods expected. However, we don’t want any residues present in the organic ingredients and more and more manufacturers as now food get questions regarding pesticide testing and the cleanliness of our ingredients that that are used in the manufacturing process. So the pressure overall from the society on manufacturers of dietary supplements and natural foods is is much stronger. Therefore, the need to develop and validate and implement methods for precise determination of pesticide residues in the ingredients becomes becomes a must. So what is the risk associated with pesticides being present in botanicals? Oh, people consume more dietary supplements than than ever. The nutritional content of consumed Foods has decreased drastically over the past decades. And it has a lot to do with the way POTUS is harvested for transport. So instead of eating a tomato that has low level of beneficial lycopene, it is just so much easier to swallow a capsule that has lycopene already purified concentrated and so much potent. It’s easy and it works. And if it improves, people’s overall well being, they will reach out for more. So naturally, if a botanical ingredient contains high pesticide residues, the exposure gets much higher. So the increased consumption as I mentioned above dietary supplements, which has become a prominent part of global culture could potentially expose customers to higher levels of pesticides. But also the more and more strict regulations globally, due to strong implications of the negative impact on human and animal health play a big role in in the risk assessment. From a company standpoint, the costs associated with organic ingredients compared to the conventional ingredients also plays a huge role because if we are paying for organic ingredient a premium price, we want to make sure that that ingredient is of the purity that the manufacturer or the vendor is claiming that we’re actually paying what we are getting. So to ensure the safety of our botanical ingredients, we have developed in house testing program to monitor the pesticide residues in our botanical ingredients. The program is designed specifically with the focus on botanical matrices. And the method was developed and validated based on the sample document and below is a picture of the team. I call them a pasty team from the last is a shawl Jerry and Aman. Those that group is responsible for running the program up now suits and they’re doing a fabulous job. If I was familiar with pesticide analysis and botanical ingredients, the residue analysis and any plant based dried botanical is among the most difficult to perform due to the severe complexity of the matrix. But not only that most dietary supplements and ingredients come in the form of concentrated extracts, which creates an additional challenge because naturally those extracts are so much more concentrated. Thus the matrix and appearance becomes a much heavier also matrix specific sample preparation and sensitive instrumentation are essential in ensuring success in detecting small amounts of residues in complex matrices. What are our challenges? Now foods? We we have about three 100 botanical ingredients that are being used in manufacturing of our products. That is a lot of various botanicals. And if we are thinking about sample preparation, there’s a lot of various matrices to address. The for the analytical complexity of those botanical matrices complicates the development of a sample preparation that will actually fit all. Also, we’re sourcing our materials worldwide, we we need to keep our keep in mind that those ingredients have been exposed to different agricultural practices. Different parts of the world have different treatment allowances, de France, pesticides that are accepted to be used and different residues that are prohibited. So looking at the scope of the pesticides that we’re trying to screen for, we have to keep in mind that different regions different areas of the world, may may be exposed to different residues. As I said, the maximum residue limits are pretty well established for fresh produce and other foods. However, if we looked at dietary ingredients and botanical dry botanical materials, they’re the resources as far as maximum residue limits are not that well described. Therefore, some commodity in some commodities, interpretation of analytical results is very difficult because we cannot compare fresh produce action limit to a dry botanical action limits. So that’s why it is it is sometimes quite difficult to establish those acceptance criteria and botanical matrices. Once we decided to bring that ability and build the pesticide residue testing program in house, we had just a couple of simple goals, we wanted to be able to control the whole process from sample preparation to reporting on the result and have visibility of the of the process in house. As opposed to getting a number from from a contract laboratory. We also needed to customize the sample preparation to address those various matrices and mitigate those matrix interferences. But also, we want it to shorten the turnaround time and sometimes prioritize the analysis of samples that are just urgent that may affect our fill rates. The next step was defining a scope. Although most companies they opt for sending their samples to contract laboratory for pesticide residue testing, we decided to keep it in house and and define the scope of the pesticides that we were going to screen for, we have opted for using commercially available mixes to cover a broad range of pesticides most commonly seen and food supply or prohibited from from use. So for the GCM edible pesticides, we have created a list of 217 residues and for the LCM anabol pesticide group, the list is a little bit larger and contains 254 residues. What about 15% of overlap between the GC and the LC group. And if you look close enough, this is our separation on the LC and GC. So there’s a lot of different colorful peaks it’s it is a truly a beautiful separation. Next, we had to design a process that would fit in a cGMP environment. So that’s from the start was quite an undertaking because for those working in the area of residue testing things are not always black and white. Sometimes the deviation from a standard process is just necessary. So we had a couple of tasks to build and to make sure that the process is going to work. You efficiency when you had to develop an acquisition method and optimize it to cover the scope of the pesticide residues that we chose to look at, need also to develop a method for efficient data processing. To quickly assess the acquired data and conclude it a result, we also needed to design a functional reporting template that will allow us to quickly report the results and make it easy to review. And as I mentioned before, we needed to develop different methods for sample preparation to accommodate various matrices. And setting up quality control criteria is a very important step in setting up a functional process in the cGMP environment. When it came to instrumentation choices, we have done a market research. We wanted to choose something that was versatile and robust. So we’ve opted for the thermo scientific trace 1310 GC, that is equipped with the T SQ 9000 triple quadrupole mass spectrometer, the analysis of residues is done in Otter SRM mode. And we are also using the chameleon software for data acquisition and analysis. And this is our actual setup in the laboratory of our GC Mass Spec. Very brief overview of our GCMs ms method, this method allows us for determination of 217 pesticide residues. Currently, we’re using a 30 meter column, but we can easily switch to a 15 meter column for fast GC separation with unchanged the resolution and we have done that. And the resolution does not change maybe, except for a couple of Cypress runs that elude a little bit closer but other than that, it is a fast separation. And the maintenance of the hardware is very low. Because of the high robot robustness of the ion source. We really don’t have to bring that instrument down for maintenance often it is truly a workhorse. So, if you take a look on the right hand side, I have a shortcut and a summary of the method that we are using for the GC ms analysis. And again, this one reflects a separation on the 30 meter column or separation, sample preparation for the GC analysis. I put a quick summary of the procedure, we are weighing approximately one gram of a sample and a 50 mil centrifuge tube the sample is then hydrated with 10 mils of water and extracted with 10 mils of acetyl nitrile containing internal standards. So this is basically a first step of caterers and then the mixture is being salted out with on hydrous magnesium sulfate and sodium chloride. So as I as I mentioned this, this is just a simple salting out procedure taken from captures, however, that upper portion of the extract is then cleaned up using an SPE procedure. So we’re using a combination of PSA and graphitized Carbon Black, for cleaning up of the botanical matrices. And after the cleanup procedure. The sample as is eluded from the SPE cartridge, evaporated and reconstituted until you were in before the analysis with the GCMS. This is not a procedure that we just came up with. This is a method that we have validated with with the FDA So doc Howard Hayward, and John Hwang have developed this this method and we were part of the multi laboratory validation program to validate this this method and It works really, really well on botanical matrices. So, this is just the reference, if anybody’s interested in details regarding this, this methodology, but this is basically what we are using for majority of botanicals as far as sample preparation goes for the GCMS analysis. So, for the LC ms analysis, we decided to go with the Vanquish H fields See, that is equipped with the Altos triple quad mass spectrometer. And similarly to the GCMS, we are detecting the analytes and our SRM mode. And once again, we are using our chameleon software for data acquisition and analysis. And the picture below shows our setups our setup is, is consisting actually of three instruments because we’re trying to use our instruments for different applications. So in combination with the Vanquish HPLC and the mass spec, we also have an IC setup for the glyphosate analysis. So very briefly, the LCMS method allows for determination of 254 pesticide residues and I, my previous slide had a separation showing although speaks the instrument sensitivity allows for a dilution of the sample. So sometimes, to mitigate matrix interferences, it’s just easier to dilute the sample and that’s exactly what we do. So on the right hand side, there is a table with a summary of our LCMS method. It’s, it’s a really nice separation, and we’re separating all the peaks within 16 minutes. But because of heavy matrix, our runtime is 20 minutes. So, here I have a summary of the sample preparation for our LC analysis. Similarly to the GC, sample preparation, or starting with the catcher’s salting out procedure. However, further past that procedure, we are using different sample cleanup procedures. Those may include this versus SPE cleanup, or direct injection, depending on what kind of matrix we’re dealing with. Clean up can can can vary. So, if we have a fatty material, sample preparation will will differ ever. So slightly. After the cleanup procedure, the LOI is evaporated or diluted with deionized water pseudo nitrile before the analysis. So, as you can see, we have different sample preparations for both LC and on GC. And I did not mention, but the sample preparation also includes spiking of the matrix and are all calibration curves are also constructed in the matrix. So we are using matrix match calibration curves, as well as internal standard correction for any possible analyte loss during the extraction. The method was validated. Using the Sante guidance document on analytical quality control and method validation procedures for pesticide residues analysis and Susan feet. We chose eight representative matrices for the validation. If we were to do it on all 300 matrices, I don’t think a lifetime would be enough to space them out and complete all that work. So we decided to do the validation on the representative sample set. Does the parameters included linearity limit of detection limit of quantitation analyte recovery and precision? And as I mentioned before extraction efficiency and matrix effects were corrected. With the use of internal standards, we have a different set of internal standards for LC and GC, that are LC or GC unmountable. So, for the software selection, we had a choice between a tray finder or chameleon. However, after reviewing both platforms, the group decided to choose chameleon to run belts instrument. And there were a couple of factors that made us decide to go with with the chameleon software. And I will explain our choice in just a little bit. Coming in and software is capable of running both LC and the GCMS. And the report templates can be created and consistent between both instruments, basically, for streamlining purposes. And the ease of use of software allows for much faster data processing, and reporting in a quality control environment. And quite frankly, who wouldn’t like Charlie, he’s so darn cute. So one of the biggest bottlenecks of any comprehensive time consuming analysis is the software ability to deliver functional necessary for creating streamline process. Having two options, you know, between trace binder and chameleon once, once again, we looked at various Bellet, Belize to review the port data of both platforms. And as I said before, chameleon was was our choice. And once the process of sample preparation and analysis was optimized, and validation under Billing, who had to think about the review and reporting process and how to design so everyone knows that the data review and data review, which contains hundreds and hundreds of residues is extremely tedious. I believe this is the biggest bottleneck of the whole process. So, again, identifying the features which make the process a little bit easier as critical. So one of those features which makes in my experience and the team’s experience, life much easier is the Smart Link features. So this slide shows the smart smart link peak presentation for a quick assessment of data, which can be easily customizable, depending on one’s preferences. So when I reviewed data and I need to look quickly at what was acquired whether there are any any hits in the batch, I look at my Smart Link overlaid peaks. For example, Jerry prefers to to look at them stacked. So the ability to customize views, depending on one’s preference is absolutely fabulous with within this software. So all information necessary to determine the data quality and the compound presence or absence is in one place. We really don’t need to switch between different windows to access information that we need for for quick assessment of the data and I pointed out to two locations to views where a compound having a green checkmark meets requirements and simply hovering over that green checkmark compound shows the meeting requirements which are preset in a data processing method. The same with a compound that does not meet the preset requirements just hovering over the red checkmark shows the reason why the compound does not meet the requirements. So this feature is great for quick assessment of component presence of absence and on the spot evaluation of the quality control criteria that we’re preset in on In the data processing methods so when a compound is detected, we need to report it and once again the ease of accessing various views and information. With a switch of view, it’s fantastic within within this software package, because in one view, I can see the response to the peak my standard curve and ion ratios, as well as the results. What the team also likes about this particular feature is the ability to customize the tables to include results, additional results, additional evaluation of the data with an addition of a column, the software is Excel based, so any modifications can be made on the on the fly with the results pretty much instantaneously. Once the data is reviewed, a custom report can be generated. And we have easily created a report template from existing reporting functions to fit our needs. So this is an example of our internal report that we use for reporting pesticide residues. And it contains a table with our probable hits quantitation non detected compounds. And creation of that particular report template was was very easy. It is thoughtfully customizable. And it’s really user friendly. There’s no coding necessary, there is a lot of options as it comes to functions, which are quite intuitive and allow building reporting templates that fit the needs of the customer. So we were able also to include a header for reviewers for secondary review, once again to fit our cGMP environment. But even with the tools that we have available, the process is quite difficult to place under the cGMP umbrella, we still have strict documentation requirements. To support strong quality systems, we have to have defined acceptance criteria. And again, comprehensive data review and approval process creates additional challenge. So how do we make this all work, we need to be able to allow flexibility and the controlled environment by clearly defining expected deviation. So we know sometimes if it comes to data interpretation for pesticide residues, as I mentioned before, things are not always black and white, we need to be able to deviate from a design path. But in order to approve that deviation, we have to explain and define what kind of deviations are acceptable. We are using software features to streamline the data and review and report process to optimize the workflow. We all know that this is a very laborious workflow. And in order to make it work, we have to have clearly defined way of doing things. I also create defined work structure within the group. So those three people that are responsible for the sample preparation, the analysis and reporting. They need to have a clearly defined responsibilities to make this work. They work as a team and they should be able to communicate within and divide and conquer. And we are striving for continuous improvement of the process. We’re fairly be new to it, especially in our environment. So we’re identifying areas in which we can improve processes that we can implement to make, once again, this, this whole workflow more and more functional. One of the areas that we see an opportunity to improve on is the integration of the process with data management system. So for further automation, and workflow, streamline, integration of the developed process with the lens system is very, very beneficial. So the combination of the powerful software capability with the data management system will allow us eventually for a smooth workflow, management and visibility throughout the whole process. Once again, from the very beginning to starting from a sample preparation, through the analysis and reporting. And a summary, we’ve built and implemented a pesticide residue testing program utilizing two analytical platforms, the GC and SMS and our cmsms. For high throughput workflow to screen a total about 500 pesticide residues in botanical matrices. The sample preparation was optimized for groups of botanical matrices, and targeted towards matrix mitigation. And this process can be implemented in a quality control environment for monitoring pesticide residues and botanical matrices. Just very quickly, I wanted to shout out to the pasty team, the shawl Jerry and Arman for doing a wonderful job. thank you to Aaron and Alan for allowing me to to do the work that we do up now. And of course, to thermo scientific, for letting me talk about the wonderful work that we do have now foods. Well, thank you, Katie, for that excellent presentation. We now have time for the question and answer sections. So Katie, the first question we have today is what regulatory limits does now food use when defining acceptance criteria. So when defining acceptance criteria, we look at global allowances and global MRLS. We also look at the acceptance criteria that are defined in CFR and because the acceptance criteria for botanical matrices are not very well defined as compared to fresh produce. And because NOW Foods distributes and sells products worldwide, we want to make sure that our acceptance criteria are in line with global allowances therefore we implement the most stringent acceptance criteria if it comes to two pesticide residues, but you already have our residues have action limits of 10 parts per billion or lower? Thank you, Kathy. We have a second question here, which is why is the sample preparation for LC and GC analysis different when most labs use one extraction to run both techniques? That is correct majority of laboratories prepare to sample using straight up cutters sample preparation and then they split the acetyl nitrile extract analyzed by LC and GC. And for streamlining purposes, that may work a little bit better. However, this is not an appropriate process if it comes to botanical matrices, and the analysis of complex samples that we deal with on a daily basis. Because the GC analysis performs better when using solvents like Tony Wayne, and the method that was developed and validated by Dr. Hayward and dr. long work works beautifully, we decided to implement that in our laboratory. But for the LLC, we decided to stick with what works better for the LLC. In regards to the sample preparation and have those samples in a pseudo nitrile water, it does require to prep the sample twice. However, the results and the recoveries of analytes that we look at for either GC Mass Spec or LC mass spec is much more much more intuitive. And we can we can recover more residues that way, and we can mitigate matrix interferences much better utilizing two procedures as opposed to just one. Okay, thank you very much, Katie, I think we have a quite an important question next, which is what happens to an ingredient which has detected pesticide residues above the tolerance level. We return it to the vendor, that ingredient is always rejected, we notified a vendor of what we found. We are a very transparent company. And we want to make sure that those issues are communicated with with vendors. Things happen, and we do see those cases, but in every case of the sort of material is rejected and never used in manufacturing of our products. Thank you, Kathy. We have another question here, which is how many internal standards are included in your LC, ms ms method for the pesticides. For the LC Ms. method, we are using a set of seven deuterated internal standards that cover different groups of pesticides. So Katie, the next question we have is, during sample preparation, you use dispersive SPE for sample cleanup, can you talk about the recovered results for this? So as I mentioned before, for both the GC and LC, we are using slightly different sample preparation, but both involve our sample cleanup using an SPE technique. So for the GC, once again, because the method is targeted specifically for botanical matrices, our recoveries are really good. It all depends on the matrix that we’re dealing with and some matrices, even with this optimized methodology, they are they’re still difficult and we see higher losses, but those losses are never higher than 5% of the group that we are looking at recoveries are typically within the 75 to 120%. With the GC we have to get a little bit more creative. Depending whether the matrix is high and fat or protein, we have to utilize a different DSP II cartridge. And once again, we see sometimes a higher losses if the matrix is is more difficult. And sometimes those losses are not really associated with the type of SPE that we are using, but with the matrix interference and in which case, we need to dilute the sample to see a little bit more through the matrix. But as I mentioned before, sometimes things are not clear cut if it comes to sample preparation analysis of such difficult matrices. So so a certain level of creativity is necessary to to be able to overcome those issues. Thank you, Katie. Our next question is, what did you consider when choosing instrumentation for this job. When we looked at what’s available on the market and different vendors, we wanted an instrument that will basically become a workhorse in our environment. But we also looked at something that will be versatile. Pesticides are not the only analysis that we are performing in house. We have a lot of different methods that require instrumentation LC or GCMs back instrumentation. So we wanted an instrument that will be easily adaptable to different methodologies. And that’s exactly what we got with both LC and GC. As I showed earlier on the picture, the LC ms that we’re using for pesticide residue analysis is also utilized for the glyphosate analysis with the simple addition of ice. So that’s exactly what we were looking for. And, and we’re really happy with our choices. Great. Thank you very much, Katie. And I think the final question we have today is, what aspect of having the in house pesticide testing program is most vital, valuable from a company’s perspective? Being able to control the process for sure, from addressing the sample preparation or difficulty, through the analysis, being able to always go back and look at the data and rationalize through things when when issues arise, and be able to report data ourselves. But also the the turnaround time, sometimes we need those results really quickly. We are manufacturing facility and prioritizing samples that are that are of a higher urgency within our environment. It’s something that we that we needed to be able to do in house. Okay, well, once again, Katie, thank you very much for your presentation today. If any, welcome, if any other questions do come in. I’m sure Katie would be happy to respond to my email afterwards. Thank you also to our scientific partners in this webinar being Thermo Fisher Scientific for all their technical input and support. And with that, I’d like to thank you all for logging in and listening today. Thank you, and goodbye.