Incorporating automation in a cytogenetics laboratory: three practitioners’ perspectives on benefits and limitations

Abstract

Automation has been developed and continues to be refined for cytogenetics, including advances in the processing of samples, in analysis of conventional chromosome and fluorescence in situ hybridization (FISH) testing, and with artificial intelligence. Here we provide an overview of the various types of automation available to the cytogenetics laboratory and discuss possible benefits and limitations you may encounter based upon our collective experiences.

Introduction

Incorporating automation into the cytogenetics laboratory can be a rewarding, yet daunting task. Automative solutions exist for most processes for conventional chromosome analysis as well as fluorescence in situ hybridization (FISH). Here we provide an overview of the various types of automation available to the cytogenetics laboratory and discuss possible benefits and limitations derived from our cumulative practical experience, which you may encounter with their incorporation. Our commentary is not an exhaustive list of all available technology, rather tailored to our experiences in these various areas of automation. Our incorporation of automation spans four laboratory settings representing Northwell Health (a nonprofit integrated healthcare network that is New York State’s largest healthcare provider and private employer with two cytogenetics laboratories), Atrium Health (a large nonprofit hospital system with a cytogenetics laboratory), and The Ohio State University Wexner Medical Center (a large academic medical center with NCI-designated comprehensive cancer center with one cytogenetics laboratory). Our laboratories differ in test offerings, from a combination of constitutional and oncology, to oncology specialized, and volumes as shown in Table 1. Automation solutions for the cytogenetics laboratory span pre-analytic, analytic, and post-analytic areas of the laboratory. Here we will focus on some of the common types of automation available for cytogenetics laboratories at the pre-analytic and analytic stages. A list of available vendors is provided in Table 2.

Table 1.

Approximate laboratory chromosome and FISH test volumes in 2024 by sample types

	Chromosome Analysis				FISH Analysis
Institution	Oncology	Prenatal	Products of conception	Constitutional PB	Oncology FISH*	Prenatal aneuploidy FISH*	Postnatal FISH*
Northwell Health Lab 1	3,350	814 AF: 567 CVS: 247	1,300	Does not perform	26,800	3,300	Does not perform
Northwell Health Lab 2	230	Does not perform	Does not perform	3,100	2,500	Does not perform	200
Ohio State University	4000	Does not perform	Does not perform	Does not perform	21,000	Does not perform	Does not perform
Atrium Health	2400	Does not perform	Does not perform	760	32,400	Does not perform	40

*FISH volumes are calculated based on the number of probe loci. AF, amniotic fluid; CVS, chorionic villi sampling; PB, constitutional peripheral blood

Table 2.

Vendors offering automation for cytogenetics

Area of automation	Vendor (in alphabetical order)	Equipment
Automated suspension harvesting	ADS Biotec	HANABI P1 (16 tubes) Harvester
		HANABI P2N (24 tube) Harvester
	Tecan	Freedom EVO
Automated in situ harvesting	ADS Biotec	HANABI P1000 (65 Petri dishes or 40 chamber slides)
	Tecan	Freedom EVO
Automated cell separation	Miltenyi Biotec	MultiMACS Cell24 Separator Plus (24 samples)
	Stemcell Technologies	RoboSepTM-S (4 samples)
		RoboSep ™-16 (16 samples)
Slide dropping for chromosomes and/or FISH	ADS Biotech	HANABI P5 Metaphase Spreader (20 Slides)
		HANABI P7 Metaphase Spreader (72 slides)
	BioDot	CellWriter S (24 slides)
	Hudson Lab Automation	ARI Scorpion (24 slides)
	Tecan	Freedom Evo (48/50 slides)
Automated staining	ADS Biotec	HANABI S1020 Chromosome Staining System (batches of 20 slides, continuous loading)
	Fisher Scientific	MilliporeSigma™ Midas™ III (20/30 slides)
	SciGene	Little Dipper® (24 slides)
Automated slide scanner (chromosomes and FISH)	Applied Spectral Imaging	Slide Scanning System (99 slides)
	Abbott Molecular	BioView Encore (120 slides)
		BioView Duet-3 (50 slides)
	Leica Biosystems	CytoInsight GSL10 (10 slides)
		CytoInsight GSL120 (120 slides)
	Metasystems	Metafer Slide Scanning System (80 G-banded slides)
AI-assisted karyotype/FISH analysis software	Applied Spectral Imaging	HiBand (chromosomes)
		HiFISH
	Metasystems	Ikaros DNN
		Metafer RapidScore
	Abbott Molecular	BioView
Automated FISH hybridization processing	Biodot	CellWriter™ S (24 slides)
	Hudson Lab Automation	ARI Scorpion (24 slides)
FISH slide washing	Abbott Molecular	VIP2000
	SciGene	Little Dipper® (24 slides)

This list may not be inclusive of all available equipment or vendors

Pre-analytic

Harvesting automation

Harvesting cultures for cytogenetic analysis is a very manual and intensive multi-step process that can last for several hours (two to three hours on average) depending on the number of cultures. Our laboratories use this technology to harvest suspension cultures from oncology bone marrows (BMs)/neoplastic peripheral bloods (PBs), constitutional PBs, and products of conception (POCs) after cell detachment from flasks. Harvesting automation is also available for in situ cultures, such as amniotic fluids (AF), chorionic villus sampling (CVS), POCs, and solid tumors on glass coverslips inside small petri dishes. This method allows for a shorter turnaround time of culturing.

Benefits

• Reduces the technologist time spent in the wet laboratory, approximately 20 min hands on time with performing instrument runs (CM, personal experience); while the instrument is performing the harvest, the technologist can work on other tasks.

• Decreases variability in processing steps.

• Reduces tension among the technologists regarding complaints about quality.

• Lowers the risk of errors.

• Instruments can be set up with multiple processing protocols if the laboratory varies the timing of their harvest processing steps across various indications or specimen types.

• Reduces fixative odor in the laboratory.

• Maintenance is straightforward and not time-consuming.

• Downtime is infrequent.

• After optimizing run parameters, validation is simple.

• All the points mentioned for suspension cultures are valid for in situ culture automated harvesting except for reducing the fixative odor.

• Less handling of coverslips reduces the risk of breakage or scratching the coverslip.

Limitations

• Some sample types may not be suitable for processing on this type of instrument, such as bone biopsies which may have sizable particulate matter which can clog aspirators.

• These instruments typically only accept certain culture vessel types; therefore, the laboratory may have to switch to the acceptable culture vessels or perform a transfer step prior to loading on the instrument.

• Can increase the volume of reagents used, and waste generated from harvesting.

• The footprint of these instruments tends to be large, requiring cleared floor space.

• Moving and dumping a full waste receptacle can be difficult.

• Instrument may need up to 30 min to equilibrate after turning on prior to starting a run.

• Harvesting in situ is limited to tissues that attach on surfaces (AF, CVS, POCs, skin, tumors) and requires specific sized glass slides inside small petri dishes.

• Laboratories that want to process suspension cultures as well as in situ cultures would need two separate instruments.

Automated cell separation

Automated cell separation is primarily used in cytogenetic laboratories for the isolation of CD138⁺ plasma cells to improve sensitivity for detection of genetic aberrations in plasma cell neoplasms by FISH.

Benefits

• Automation can perform plasma cell separation for multiple samples simultaneously, which can reduce the technologist time spent in the wet laboratory.

• Utilizing the instrument reduces technologist’s hands-on time even when only a single sample is being processed.

• Instrument start up is simple.

• Decreases variability in processing steps.

• Lowers the risk of errors.

• Can be used jointly with immunology/flow cytometry/molecular laboratories.

• Relatively inexpensive type of automation.

• Small tabletop footprint.

• Records data from runs to assist with troubleshooting.

Limitations

• Typically used for plasma cell neoplasm samples so may be of limited benefit to laboratories with low plasma cell neoplasm volumes.

• Recommends at least 10% plasma cells in sample; specimens with lower plasma cell content will have variability in final sample purity.

• Cells may adhere to the side of the tubes and be difficult to dislodge.

Specimen dropping onto slides for chromosomes and/or FISH

Slide preparation from fixed cytogenetics pellets, commonly referred to as dropping slides, can be considered as much an art as a technical skill. Spreading of chromosomes from dropping is both temperature and humidity dependent which causes variation in chromosome preparations that affect the downstream steps of banding and analysis. Multiple slide-dropping instruments are available for cytogenetic processes; across our institutions we have experience with a variety of manufacturers in this area. Of note, instrumentation is available for providing environmental control while performing manual dropping, which will not be covered here.

Benefits

• Significant savings in technologist’s time.

• Some instruments are compatible with dropping for both conventional chromosomes and FISH.

• Some instrumentation has multi-functionality and can be utilized to add FISH probes and/or add DAPI.

• Automating the process reduces tension among the technologists regarding complaints about quality.

• Can perform dropping of small volumes on multi-well slides.

• Can operate using barcode readers to know what patient to drop on what slide or what probe to drop in which well.

• Reduction in patient swap and probe swap errors.

• Some instruments provide tracking of samples run, probe lot numbers, and probe volumes used.

• Cell density optimization is available with some automation solutions.

• Some instrumentation can add coverslips.

Limitations

• Automated droppers typically need to achieve and maintain specific humidity and temperatures which may present a challenge in laboratory environments where these variables have wide fluctuations throughout the year.

• Instruments may require specific, or propriety slides, to function.

• Instruments may require specific types of sample tubes.

• Automated FISH droppers can have pipetting difficulty for runs that span multiple probe manufacturers due to differing viscosity of probes.

• Some instruments require specific pipette tips, which may not be compatible with high viscosity buffers.

• Minimum sample and probe volumes are required; samples with small pellets may not be adequate for dropping via an instrument.

• Validation can be cumbersome and challenging to establish parameters for optimal quality of chromosome spreads as well as FISH signals, which can also result in high consumable costs for validation.

• Space requirements are variable among the different manufacturers; some have larger floor space footprints versus tabletop instruments.

Automated staining

A variety of slide staining platforms are available. Many platforms designed for hematology and histology applications can be adapted for cytogenetics chromosome banding processes.

Benefits

• Reduces the technologist’s time spent on the bench by processing slides in large batches.

• Reduces variability in processing.

• Some instrumentation can reduce reagent amounts used.

• Instruments can store multiple programs for slide processing.

• Simple to clean and maintain.

• Relatively inexpensive type of automation.

• Some instruments can continuously load batches of slides.

Limitations

• Limited ability to adjust timing of banding steps on a day-to-day and patient-to-patient basis, otherwise time savings gained would be sacrificed.

• May have lower quality banding on a subset of slides due to sample variability.

• May require placement near a sink for water supply and waste drainage.

• Space requirements are variable among the different manufacturers.

Automated FISH hybridization processing

FISH hybridization workflow involves several steps, which although not complex in nature, can be very time consuming. This automation can streamline the FISH denaturing and hybridizing process.

Benefits

• Reduces time required to process FISH slides.

• Can eliminate some reagents from processing such as pretreatment, alcohol dehydration, and rubber cement.

• Eliminating some processing steps reduces technologist exposure to chemicals.

• Some instrumentation is compatible with regular slides and multi-well slides.

• Using multi-well slides reduces FISH probe amounts used which can lead to large cost savings.

• These instruments can handle multiple slides simultaneously.

• Can provide precise temperature, humidity, and timing for processing steps.

• Generally able to store multiple processing programs.

• Multi-function instrumentation which can perform specimen and probe dropping, cover-slipping and hybridization is available (see section on specimen dropping for chromosomes and/or FISH discussion above.)

Limitations

• May require significant optimization, particularly with multi-function instrumentation, to establish parameters for high quality FISH signals.

• Difficult probes or patient samples may perform suboptimal and require manual processing.

• Measuring slide slot temperatures can be challenging.

• If probes from different manufacturers are on the same run, optimization of denaturation conditions for these probe combinations would need to be performed.

FISH pre-treatment and/or post-hybridization washing

Similar to the hybridization process, pre-treatment and post-hybridization washing of slides are not complicated workflows but are repetitive and time consuming, which makes these tasks optimal for automation.

Benefits

• Time savings by washing batches of slides at once.

• Some instrumentation can be used for pre-treatment and post-hybridization washes.

• Some instrumentation has an “auto-start” option; add slides and once programmed temperature is reached the process begins.

• Most instruments have minimal maintenance and are easy to clean.

• Reduces variability in timing of processing steps.

• Improved safety for technologists as they are not manually transferring slides from scalding liquids.

• Able to store multiple programs.

• Validating the instrument is simple.

Limitations

• Some difficulty determining how to maintain temperature when washing multiple slides.

• Some instrumentation requires a computer to operate, which can complicate the process.

• Programs with high agitation may result in slide breakage.

• Ability to successfully remove coverslips is variable.

Analytic

Automated slide scanner for chromosomes

Slide scanning and manual image capturing of an adequate number of good quality metaphases for chromosome analysis is a very time-consuming process for the cytogenetics technologist. Utilizing automated image scanning equipment can eliminate this burden and improve quality and case turnaround time.

Benefits

• Significant technologist time is saved by removing manual scanning of slides and image capturing.

• Improved ergonomics for technologists by reducing time spent looking down through the microscope.

• Reduces equipment needs for light microscopes and cameras in the laboratory.

• Image capture can occur in off shifts to be ready for analysis at the beginning of first shift.

• Allows for capture of hundreds of images which can be useful in situations such as scanning for low-level clones or when extended analysis is needed.

• With some manufacturers, laboratories can customize capture settings such as quality of metaphases and number to acquire.

• Some instruments can have multiple capture program settings to optimize the process for various indications.

• Increases case transparency, all cells evaluated are readily available for review.

• After optimizing capture parameters, scanning is relatively easy to validate.

• Can perform barcode reading of slides which further reduces technologist hands on time and reduces errors.

Limitations

• Establishing initial capture parameters can be time consuming and require significant technical support from the manufacturer.

• Debris is typical captured even after training the system. Improving the recognition of metaphases using artificial intelligence is currently being developed by at least one manufacturer.

• Poor quality slides such as those with increased cell debris or stain precipitate can reduce ability to identify metaphases.

• Oiler problems, such as clogging, or excess application can be an issue.

• A scanner crashing can cause a significant bottleneck in testing.

• Automated slide scanners are directly tied to analysis software and exporting images into other software may be cumbersome.

• May result in increased data storage requirements.

• Most have a large tabletop footprint.

Artificial intelligence (AI)-assisted karyotype analysis software

This type of software utilizes AI algorithms to identify chromosomes within a metaphase spread and to organize them into a karyogram. Automating this process can significantly reduce manual work by performing segmenting, resolving overlaps, and arranging chromosomes from a metaphase spread within seconds. This piece of automation is ideal for cytogenetics laboratories with increasing volumes for chromosome analysis or those with staffing shortages looking to decrease technologist time spent in the analysis process [2, 4–6, 8].

Benefits

• Significant reduction in time spent preparing a karyogram.

• Reduces clicking and small repetitive moments for technologists improving ergonomics and technologist satisfaction.

• Time savings makes possible the creation of more than the minimum required karyograms, which can improve the case review process.

• Customizable automated chromosome enhancement settings can be built.

Limitations

• Complicates training new technologists and evaluating their performance on cases.

• Performance of the AI decreases with highly complex abnormal cases and with poor quality metaphases.

• Does not perform the analysis, just prepares the karyogram for technologists to more quickly reach the analysis step.

Automated slide scanner and analysis software for FISH

Incorporating automation for scanning and analyzing FISH slides has been shown to be valuable to some cytogenetics laboratories. With the right microscopy setup, FISH scanning can be performed using the same scanner automation as that used for chromosomes. Scanning allows for rapid capture of FISH interphase nuclei [3, 7].

Benefits

• Reduces time technologists need to spend sitting at a microscope and time spent in the dark.

• For some applications, image capture can occur in off shifts to be ready to be analyzed at beginning of first shift.

• Reduces equipment needs for fluorescent microscopes and cameras in the laboratory.

• Solutions are available for liquid and tissue-based FISH assays.

• Images can be fully analyzed by technologists or evaluated using a semi-automated approach which incorporates software suggested classifications with technologist review.

• Generally, settings are highly customizable, and laboratories can optimize capture, image enhancement, and analysis parameters for different panels.

• Users can select areas of interest or sort cells, such as sorting on cell size.

• During case analysis data is kept blinded between readers.

• Automated FISH scanning can be beneficial for training new FISH readers by permitting simultaneous viewing of cells by the trainer and trainee.

• Creates a transparent process with the ability to review how each cell was called by a technologist which may help resolve discrepancies between readers.

• Software can account for cutting and tissue thickness inconsistencies to improve images.

• Compares Hematoxylin and Eosin (H&E) slide to FISH slide for marking area of interest for scoring cells.

Limitations

• Oil clogging the scanner particularly when not used daily, results in the need to calibrate the slide and oiler.

• Low cellularity samples may produce long scan times.

• Photobleaching of fluorescence may occur when there are too few cells in the well prolonging the scanning process.

• May have difficulty sufficiently capturing and analyzing plasma cells due to their tendency to clump together.

• May have difficulty capturing signals with dense tissue.

• Abnormal cells may end up in the “unclassified bin” which must be checked for every slide.

• Difficulty in validating the counting of signals and discerning abnormal cells with unusual signal patterns.

• Can have long training and adjustment period for technologists to feel comfortable using the system.

• Technologists may have difficulty adapting to Z-stack limitations compared to the ability of a microscope to focus up and down throughout a cell.

General considerations

Most automation comes at a high cost, from thousands to hundreds of thousands of dollars requiring capital request and approvals. Some vendors may offer outright purchase of an instrument, trial periods, rental programs, or per use charge (reagent rental). Some institutions may not permit equipment or reagent rentals and instead will only sign contracts for equipment purchases. For justification of upfront capital costs consider the offset of savings on downstream labor from reducing hands on time, simplifying processes, and reducing sample repeats and errors. Also consider any cost savings from reductions in equipment, reagents, and consumables. Instrumentation requires preventative maintenance, and many vendors require service and support contracts or charge repair costs, which can be costly and will need to be factored into annual budgeting. Most institutions have some type of oversight process that requires risk assessment for automation. This may involve evaluating how instrumentation accesses and handles Protected Health Information (PHI), determining if the vendor will have remote access to the instrument, and establishing how servers and data storage are configured and where they will be housed. Therefore, it is imperative to include information services (IS) early in the discussion otherwise new equipment may sit for months while IS determines if it meets storage and security requirements for the institution.

For instrumentation with computer connectivity or laboratory information software (LIS) integration, operating systems, LIS updates, or changes to vendor software may result in incompatibilities between these systems. Upgrades to vendor software may come at additional cost and will also require re-validation. Many types of laboratory automation are influenced by the capabilities of the laboratory’s LIS, for example, barcode reading is important for fully capitalizing on all the features of various automation solutions. We recommend determining LIS capability to barcode and evaluating compatibility with the automated equipment of interest. Beyond barcoding, some instrumentation can be integrated to communicate with LIS, for example to populate sample data to the instruments, provide tracking of runs, or to populate results into reports. This type of functionality may incur additional development costs and require collaboration between the vendor, LIS builders and technical support.

The capacity of the instrument may limit its usefulness in some labs, processing small volumes may not see a high rate of return on investment with some instruments, whereas for large volume laboratories more than one run per day may be required. Alternatively, the lab may choose to purchase more than one instrument to accommodate volumes. Besides volume-based time savings, other considerations may factor in the purchase of an instrument, such as the need to improve the quality of metaphases or FISH hybridization, to increase technologists’ morale, and to help address staffing shortages, which continues to be a challenge for our field [1].

Furthermore, the laboratory will need to consider either equipment redundancy or maintaining protocols, competency, and space for manual processes and procedures in the event of instrument downtime. Space is also a consideration when purchasing instrumentation, not only the dedicated footprint of the instrument, but where to best place the instrument to optimize workflow, access to emergency power and internet access, and environmental factors such as additional heat or noise generated by the instrument.

Last, but not least, for automation to succeed in the laboratory, technologist buy-in is paramount. It can be beneficial to have technologists participate in demonstrations and provide input during the selection of new automation. Thorough training and dedicated time for technologists to become familiar with and adapt to new automation will help to ensure a successful launch. Technologists’ job satisfaction can be greatly increased with automation by decreasing tedious work in the laboratory.

Automation solutions continue to be developed, refined, and adopted by cytogenetics laboratories [3, 7]. Further development of AI to improve metaphase recognition, chromosome segmentation and identification in poor quality and complex abnormal metaphases, flagging potentially abnormal chromosomes, improving image quality, and accurate automated scoring of FISH across a range of probes and sample types would be welcome. Beyond these advances discussed for traditional cytogenetic testing, automation also continues to be developed for a range of other companion technologies, including microarray, optical genome mapping, and next generation sequencing. Integrated user-friendly solutions for comprehensive collation of results across these modalities is another area worthy of continued development. As cytogenetics laboratories evolve to respond to increasing volumes, staffing shortages, and providing faster turnaround time, the further development and refinement of automation will be a key avenue to help address these challenges.

Author contributions

All authors contributed equally in preparing this manuscript.

Data availability

No datasets were generated or analysed during the current study.

Declarations

COI

NC serves on the scientific advisory board of ADS Biotech, CM none, JT none.