Sample preparation, including aliquoting, dilution, and extraction, often accounts for the majority of the analytical time in the laboratory (1,2). But in mass spectrometry, sample preparation is a key step in isolating, enriching, or concentrating the analyte of interest. It also helps remove interferences in order to minimize ion suppression. As a result, the decision to automate or even semi-automate sample preparation for mass spectrometry can be difficult.
Is automated sample preparation right for your lab?
The first step in any automation process is understanding the lab’s needs and challenges. Laboratorians should carefully analyze all their current procedures for sample preparation for each analyte (i.e. protein precipitation, solid-phase extraction, liquid-liquid extraction, filtration, and dilution). This workflow analysis should map out the technologist’s time associated with each sample preparation step along with costs of the labor, supplies, and reagents.
Laboratorians also need to evaluate the strengths and weaknesses of their current sample preparation processes objectively to identify whether changes are needed prior to automation. If the lab’s current analyte recovery or ion suppression is a problem, then it needs to look at new sample preparation methods rather than automate a bad process. Once the lab chooses an acceptable method, it can begin to see which steps in the process should be automated. The lab should also clearly define its goals and metrics for the automation project and make these clear to vendors.
Another important decision is whether the lab needs complete automation (walk-away) solutions or partial automation of sample preparation. Laboratorians must identify which tasks to automate (e.g. uncapping of samples, pipetting, mixing, incubation, extraction, centrifugation, plate sealing/unsealing). Leaders should look at systems that can accommodate their current volume and throughput, but are scalable for future growth or new applications.
Additional items to consider include integrating automated sample preparation with the laboratory information system, project cost, and physical or engineering restrictions (e.g. footprint, humidity, and waste). In the end, laboratory directors will need to balance the costs of automation with their goals to determine if automation is practical.
What are the main advantages and disadvantages of automating sample preparation for mass spectrometry?
The need for automation usually revolves around several key areas: improving productivity and sample throughput, eliminating or reducing error-prone manual procedures, improving personnel safety, saving costs by reallocating laboratory staff freed from manual tasks, and improving consistency for precision, accuracy, and turn-around-time. Automating sample preparation can accomplish all of these goals. It also minimizes a technologist’s risk for ergonomic injuries and exposure to solvents and other biohazards while standardizing sample preparation and improving an assay’s reproducibility.
However, automation is not amenable to all applications and sample types. Pediatric samples and low-volume irretrievable sample types like cerebrospinal fluid or meconium are difficult for any automation system and may still require an individual, manual process.
What are key factors to consider when deciding between automation options?
Labs primarily have opted for task-targeted or stand-alone options. Stand-alone robotic liquid-handling (RLH) systems offer a large variety of footprints in a wide range of prices, from bench top to 8-feet long floor-standing models. Labs may best achieve simple task-targeted automation using very economical, small stand-alone systems that can pipet, mix, or transfer samples from 96-well plates.
Another consideration is using disposable versus fixed pipetting tips. Disposable tips offer the advantage of no carry-over, but come at an added cost. Fixed tips work well with a variety of solvent types and transfer sample directly from tubes with pierceable caps. However, they require washing and careful validation to minimize carry-over.
The deck layout configuration and options on RLH systems are also numerous. While it may be possible to automate every step in the sample preparation process, laboratorians will have to decide if the cost is really justified. For example, does centrifugation, plate sealing/unsealing, and/or incubation need to be on the RLH deck to achieve walk-away automation? Or are offline manual solutions (i.e. centrifugation of a plate) more cost-effective with minimal impact on labor?
Similarly, adding incubators or centrifuges may cost more than double the price of stand-alone equipment that already exists in the lab while offering only minimal gains in productivity. Ultimately, each lab will need to carefully evaluate all the options and select the one that makes the best sense for its operations.
What benefits have you seen from automating sample preparation for mass spectrometry?
In the clinical mass spectrometry laboratory at the Mayo Clinic, we use a wide range of sample preparation automation: from small, inexpensive systems like the Integra Viaflo Assist, to large floor-standing models like the Hamilton Star or Starlet. These systems have enabled us to improve turnaround times and throughput by >25%, improve the imprecision of analytes (e.g. levetiracetam % coefficient of variation at 5 mcg/mL went from 5.7% to 2.8% post automation), and save labor and supply costs by up to $1.73/sample.
Importantly, these automated systems free technologists from non-value added tasks that are highly repetitive and lead to injuries. Overall, they have enabled us to continue to grow and expand while consistently meeting patient needs with high quality results.
What features of sample preparation automation do you want to see in the future?
As mass spectrometry continues to be incorporated into all areas of the laboratory and becomes part of total laboratory automation, I would like to see one comprehensive informatics system. This would manage and bi-directionally interface between the hospital/laboratory information systems and all the pre-analytical, analytical, and post-analytical systems—from sample preparation to mass spectrometers to storage and retrieval.
Such comprehensive software could receive samples into the laboratory, schedule samples for setup, coordinate, and control analysis—and with autoverification rules—directly release results into the electronic medical record. This software also would track sample locations, and send samples to storage, or retrieve samples for add-on testing, depending on whether adequate volume remained and the correct specimen type and analyte-specific stability requirements were met. The integration of sample preparation directly with mass spectrometry would benefit from bi-directional interfacing between multiple vendor platforms to allow the most efficient operation and gains.
I would also like to see increased throughput on automated RLH devices to match the speed of some of the newer mass spectrometer-based systems (i.e. Agilent Rapidfire, Bruker MALDI HTS, Phytronix Laser Diode Thermal Desorption), whose analytical times are now less than 10 seconds per sample. As technology continues to expand, clinical laboratories will need to determine when the automation of sample preparation makes sense for their operations.
References
1. Zheng N, Jiang H, Zeng J. Current advances and strategies towards fully automated sample preparation for regulated LC-MS/MS bioanalysis. Bioanalysis 2014;6:2441–59.
2. Pan J, Zhang C, Zhang Z, et al. Review of online coupling of sample preparation techniques with liquid chromatography. Anal Chim Acta 2014;815:1–15.
Paul Jannetto, PhD, DABCC, FACB, MT(ASCP), is a senior associate consultant at the Mayo Clinic in Rochester, Minnesota, where he is co-director of the clinical mass spectrometry laboratory. +Email: [email protected]
CLN's Focus on Mass Spectrometry is sponsored by Waters Corporation.
