Richard Spero, PhD
CEO at Redbud Labs.

Talk to someone who has developed a sample-to-answer cartridge, and you’ll hear horror stories about how

challenging and expensive it was. Chances are, the protagonist of those horror stories is sample prep. It doesn’t have to be this way. By planning well and choosing the right tools, you can minimize the risk of implementing on-cartridge sample prep for your diagnostic or life science assay.

The source of sample prep surprises

Sample prep means different things in different assays. It could be as simple as adding a dried reagent or as complex as library prep, but it’s usually a shorthand for everything that happens upstream of your core technology. As in: “We take a raw sample, prepare the sample, then use our technology.” See how the word “prepare” sweeps a pile of technical risk under the rug?

Sample prep risk doesn’t present itself until late in development. In the early days of the project, you can use off-the-shelf sample prep methods: centrifuges, magnetic beads, water baths, rotators, and pipettes. But most of these tools can’t be used on a cartridge. Others, like magnetic beads, require an experienced developer to manage the transition. This is the source of sample prep surprises: when you transition benchtop methods to a microfluidic cartridge, overall assay performance typically drops.

Simply put, your early benchtop experiments might lead you to overestimate the performance you can get from your final product. When this happens, you’ll end up in a process like this:

  1. Develop your assay using benchtop sample prep methods
  2. Implement a microfluidic version of the sample prep method on a cartridge
  3. Test the assay with cartridge-based sample prep and discover performance has dropped
  4. Iterate the on-cartridge prep and/or re-optimize your core technology to work with lower-quality prep
  5. Repeat steps 3 and 4 until the project runs low on time or budget

Ultimately, product teams are forced to make concessions that curtail the success of the product—or worse, cause it to fail entirely.

To minimize surprises, get to know your sample prep

To avoid this fate, start by understanding your sample prep tolerances. A tolerance is the allowable amount of variation in a key parameter of your assay. It’s the “plus or minus” on your specification.

How do you measure tolerances? You intentionally degrade performance. For example, if you’re using a centrifuge to spin down blood, back off on the purity of the plasma until you know how red blood cell contamination affects your limit of detection. If you’re re-suspending a dried reagent, back off on the mixing to see what it does to your assay precision.

Knowing your tolerances won’t eliminate surprises, but it will expose problems sooner. For example, if you know your assay can tolerate up to 5% hematocrit, then you can evaluate the performance of your on-cartridge plasma separation module the first time you use it, without ever connecting it to your downstream modules. If you’re working with an outside cartridge engineering firm, including tolerances on your specification will also help you establish the milestones for the project.

You’ll notice that this approach doesn’t change the process mapped out above. It only gets you to step 3 sooner, meaning you’ll have more time and money to iterate through steps 3 and 4. The more iterations you have in your development cycle, the better your chances of resolving issues that arise.

To eliminate surprises, go microfluidic earlier

Obviously, it would be better to avoid sample prep surprises in the first place, which you can only do by using microfluidic tools from the beginning.

Start testing off-the-shelf microfluidic tools from your earliest days of product development. There’s usually more than one approach to sample prep for a given assay. Choose the microfluidic-native methods where you can. They may be unfamiliar or disorienting, but they will ensure you’re optimizing your final, cartridge-based assay, instead of your prototype, benchtop assay. Most importantly, you’ll quickly learn what methods are unworkable and which tools you can trust.

For example, imagine you have a reagent that will need to be stored dry, on-cartridge. The performance of that reagent depends on how you dry it, resuspend it, mix it, and incubate it. Instead of working with that reagent in a microfuge tube, dry it into a microfluidic chamber. To re-suspend the reagent, try pumping it back and forth, then compare against a microfluidic mixing chip.

Or perhaps your assay includes purification. If you want to try magnetic beads, start the reaction with the beads dried inside a microfluidic chamber instead of pipetting them. Then compare this method against a microfluidic sample prep chip.

The bottom line is that you can only avoid sample prep surprises if you develop your assay using microfluidic methods early in the process. It’s important to know that off-the-shelf microfluidics is a new frontier. Historically, microfluidic cartridges have been completely custom, but this is starting to change.

As you test microfluidic methods, you’ll find many are unwieldy for rapid assay development and early prototyping. You’ll also find a few that are just as useful at the bench as your pipettes and tubes.

At Redbud Labs, we have a name for microfluidic tools that translate effortlessly from the bench to the final product: we call them cartridge-ready. The more of these tools that you can integrate into your early product development, the fewer changes you’ll need to make during cartridge integration, and the more likely you’ll be to maintain your assay performance as you finalize your product design.