Quality control and validation in flow cytometry

Introduction

Methods of diagnostic flow cytometry are commonly used primarily in the fields of immunology and hematology today. The cellular flow cytometric analysis, however, can also be applied in other fields, for example, in hemostaseology, transfusion medicine or pathology. But in most cases, routine diagnosis is the domain of diagnostic laboratories specialized in hematological and/or immunological diagnostics, and this chapter will focus on them. Today, most of these laboratories work with CE-marked test kits and ready-to-use reagents (in-vitro diagnostics). Nevertheless, validation and the correct implementation of quality controls place considerable demands on the laboratory in its clinical-diagnostic routines.

The sample in specialized hematological and immunological laboratories

The requirements for a specialized hematological or immunological laboratory have become very complex as a result of the diversity in the methods that have been established over time. In terms of quality assurance, one cannot only focus on flow cytometry samples, but must also take into account the different requirements.

The following types of tests are typically found in specialized hematological laboratories (in ascending order of specialization):

• The automated measurement of corpuscular blood components or corpuscular components from various special materials;

• The methods of cytomorphology and cytochemistry;

• Flow cytometry;

• Cytogenetics;

• The determination of colony forming units (CFUs);

• The methods of molecular diagnostics; and

• Next Generation Sequencing (NGS).

For specialized immunological laboratories, one must also add cellular functional assays, which are frequently designed so individually that routine protocols are generally available only rarely. The definition of a specialized hematology laboratory and the minimum requirements that must be put to such a laboratory are outlined in a position paper of the Laboratory Working Group of the German Society of Hematology and Oncology (DGHO) [1].

The spectrum of cytometrically analyzable specimens and their various corpuscular components is broad. For example, particles 0.5–50 μm in size can be analyzed. These include physiological blood cells, tumor cells, chromosomes, as well as algae and protozoa, but they play no role in specialized hematological or immunological laboratories. But of importance are also microvesicles and bead-based assays, which are used in the analysis of cellular functions in the multiplex evaluation of cellular products.

The various sample materials, such as peripheral blood, bone marrow, cerebrospinal fluid, various puncture fluids and suspensions from tissues create different requirements in terms of transport, storage and pre-analysis. The complex technology places special demands on environmental conditions, equipment specifications and the maintenance of equipment. The variety of methods and users will also require a consolidated training policy and careful documentation [1–4].

Given its high degree of complexity, flow cytometry has nowhere near achieved the same degree of standardization as other laboratory methods. However, specific points of flow cytometry have been added to national and international quality-relevant documents in recent years, so that there are already some binding targets in internal and external quality control as well as in the accreditation process.

The requirements for the analytic process can be seen in the respective documents and checklists (see below). The following section will discuss the main quality-relevant points.

Applied quality assurance in the flow cytometry laboratory

Different definitions of the term quality are found in the literature. It is very common to define quality as the degree of agreement between claims or expectations (target) about a product and its characteristics (actual). In the same vein, DIN EN ISO 8402 defines quality as “the totality of characteristics of an entity that bear on its ability to satisfy stated and implied needs.”

The quality standard DIN EN ISO 9000:2000ff goes one step further and describes quality as the ability of a set of inherent characteristics of a product, system or process to fulfill requirements of customers and other interested parties. The quality management system is defined as coordinated activities to direct and control an organization to improve the quality of products manufactured and services provided [5]. This very abstract definition is hard to apply to the requirements of a laboratory. But as a general rule, an accredited quality management serves to attain the following objectives:

• Meeting the demands of one’s own management;

• Use for marketing purposes;

• Improving quality and competence;

• Increasing customer satisfaction;

• Increasing legal certainty;

• Increasing the motivation of employees.

The long-term goal is to increase effectiveness and quality, which will benefit the patients, as well as the staff who implement the accreditation [6].

Accreditation in the flow cytometry laboratory according to DIN EN ISO 15189 and its requirements

While in most European countries the accreditation of medical laboratories is increasingly becoming obligatory, it is voluntary in Germany [20]. Many laboratories in Germany are still certified, as this allows for better comparability, facilitates international cooperation and, moreover, is often required by external clients.

The requirements of DIN EN ISO 15189 are the same for all laboratory areas and therefore also apply to flow cytometry. Special circumstances arise here not only in the context of the analytical process (see below), but also with respect to the general management requirements. The flow cytometry laboratory is often a separate entity in organizational terms. Quality management must take account of the special procedures in such a laboratory. Subcontracts are often associated here with complex interpretations and must be included accordingly when preparing the findings. The inclusion of cytometric raw data and findings in the laboratory’s IT is mostly labor-intensive and requires individual solutions. Ideally, flow cytometry data would be incorporated into a consolidated overall report at the end of the process.

The staff must be trained appropriately. Likewise, a program for ongoing professional development (PD) must be established. This includes regular internal and external PD opportunities, the success of which must be checked. What does this mean in detail and where does one find appropriate specifications? The basic training for specialists or medical-technical laboratory assistants alone is not enough to meet the very specific requirements of a hematological laboratory. Accordingly, the laboratory should have on its staff a hematological medical-technical lab assistant or medical-technical lab assistant with equivalent qualifications. The diagnostic report should be drafted as a synopsis by a staff physician (laboratory physician and/or hematologist-oncologist, pathologist), and should always contain an assessment. Furthermore, the medical and technical staff of a specialist laboratory must have the necessary experience with the methods and procedures employed. The number of tests per month and year and should correspond to a minimum requirement [1]. According to 5.1.2 of DIN EN ISO 15189, the laboratory management must document the qualifications for the entire staff and for each position. The qualifications must reflect the proper education, training, necessary experience and proven skills, and must be appropriate for the tasks to be performed. The staff carrying out technical assessments of tests must have a relevant theoretical and practical background and experience. Regarding training measures and skills assessment, one can read in 5.1.5 and 5.1.6 that the laboratory must provide training measures to the entire staff in the following areas: a) the QM system, b) the work processes and methods assigned, etc. The staff must be under technical supervision during training measures at all times. The effectiveness of the training programs must be assessed regularly. Once the training has been completed, the laboratory will have to assess each employee’s capability of performing the assigned management or technical tasks in accordance with specified criteria.

Follow-up assessments must be carried out at regular intervals. Where necessary, the training will have to be repeated. As for the regularity and documentation of training and PD measures, it is stated in 5.1.8 and 5.1.9 that a program for ongoing professional development must be made available to employees who are involved in management and technical processes, and employees must participate in PD. The effectiveness of the PD program must be checked regularly. Employees must participate regularly in professional development or other activities associated with their profession. Records of relevant training and professional qualifications obtained, professional development and experience, as well as skills assessment of all employees must be maintained [3]. One could thus say that employees “owe” a duty to participate, while the employer has a duty to collect on such “debt”.

The facilities must be of adequate size and properly ventilated – cytometers generate heat, a situation that is compounded further by additional refrigerators, centrifuges, as well as, perhaps, safety workbenches and incubators.

Factors affecting flow cytometric analysis

Each step in the analytical process includes factors affecting the precision and accuracy of measurement results. Tables 1–3 show a list of the various factors. The factors, variables and corresponding solutions of that list are explained in detail in the following chapters.

Standardization and harmonization in flow cytometry

When using modern flow cytometers and properly validated antibody panels, highly reproducible measurements are possible. This generally involves two strategies to ensure optimal analyses: standardization and harmonization. Standardization is about the attempt, from the outset, to achieve a high degree of agreement through given antibodies with defined markings [21], while the objective of harmonization is to render the results from many platforms comparable [22]. The best known standardization activity is the Euroflow Initiative [21]. For most laboratories that receive variety of requests and frequently set up new panels, however, harmonization remains the only feasible solution. The flow cytometer requires an instrument that measures multiple parameters simultaneously for the setup and calibration of different types of standards:

• Reference standards;

• Compensation standards;

• Binding standards; and

• Calibration standards.

Reference standards often consist of microbeads having a defined fluorescence that corresponds to the excitation and emission spectra of cells labeled with antibodies of the corresponding conjugate. Compensation standards are also microbeads that may be used for compensation instead of cells. Binding standards bind a defined quantity of antibodies and thereby allow for a comparison of staining intensity. Calibration standards allow for the comparability of measurements relating to staining intensity and thus receptor density or expression level.

An example of the use of reference or calibration standards is the determination of antigen expression densities with (semi-)quantification. The specification of antigen expression densities includes, among other factors, the problem of dependence on equipment due to the different compensation capabilities of devices and the dependence on the antibody clone as well as the antibody concentrations used. Without (semi-)quantification, there is no comparability between laboratories in multicenter studies. In addition, the calculated standard values can only be used internally. One possible strategy is internal calibration by use of reference antigens (fresh blood or commercial cell products) with a known and relatively stable number of surface antigens (e.g. CD4) and external calibration by use of hard dye beads, such as Cyto-Cal™ Multifluor-Beads or Rainbow™-Beads. In the latter method, the determination of instrument-specific MESF (Molecules of Equivalent Soluble Fluorochrome) values is made possible by different concentrations of the fluorescent dyes, that is, different levels of brightness, via a calibration curve. The number of bound fluorochrome molecules on cell surfaces (antibodies bound per cell, ABC), for example, can be determined by means of QuantiBRITE™-Beads. A prerequisite for both types of standardization, as for all other types of measurements, is the compensation of the spectral overlaps, which is achieved with spectrally matched reference beads (e.g. CaliBRITE™ beads). [23] and [24] provide good overviews of this.

Verification and validation in the flow cytometry laboratory

The terms verification and validation are central to the practice of quality assurance. Nevertheless, there is no uniform definition. According to DIN ISO 9000:2000, verification is the objective evidence that the requirements are met (“familiarization”). Validation under this norm is defined as the objective evidence that the requirements are met for a specific purpose [2]. These terms are also defined in the standard DIN EN ISO 15189 [25]. A more liberal interpretation might be: Verification and validation are both indispensable for process control. Both are independent procedures in the establishment, review, evaluation and documentation in the context of assessing whether a product or system complies with regulatory and/or technical standards. The “validation question” is: “Are you building the right thing?”, and the “verification question” is: “Are you building it right?” (see Table 4).

When a laboratory uses ready-to-use test kits and reagents marked IVD, it is possible to use them for diagnostics after a short familiarization phase. In flow cytometry, this is the case only for few, but often-used procedures, since most cases do not only involve prefabricated tests, but also their adaptation to the specific conditions at the laboratory. There are very good guidelines for implementing validation that comprise the most important elements [26]. The parameters that need to be collected, depending on the application, include the following entities relative to the type of test (also see Table 5):

• Accuracy, secured morphologically in hematology; most of the time, ten replicates are measures, while qualitative methods use 20 [27, 28];

• Specificity, the analytical specificity is mostly <5% [27]; clinical specificity is also important [28];

• Sensitivity, also includes detection limit, depends on device settings, measurement conditions and antibody titration [27]; this is especially important in connection with the measurement of CD34 and MRD [28];

• Precision, can be determined, for example, by way of peripheral blood, cell lines or blood standards and 20 replicates (recommended); for qualitative methods, at least three positive and negative ones each; a device comparison may also be required, which should come out to <10% or, worst case, <20% [27, 28];

• Linearity, this is particularly relevant for rare events (MRD, fetal erythrocytes) or quantitative measurements (CD64) [28];

• Stability, pre-analytical and post-processing stability are important to assess the robustness of the method; deviation should be under 20% [28].

• Matrix influences must be taken into account by validating all methods for each matrix used separately [28].

• Influences of anticoagulants on the stability must be validated, ideally on the basis of at least five samples at different times [28].

Validation is mandatory each time a new or modified method is introduced at the laboratory. Records should be highly structured and follow the scheme of validation plan, measured data, and validation record with release.

Key factors for valid flow cytometry data

Several factors are crucial to reliable flow cytometry diagnostics. Regardless of the system used, they are always part of it. An overview of relevant quality assurance measures can be found in Nebe et al. [13].

All cytometers undergo a daily system check. In diagnostic equipment, this tends to be software-based. When fluorescence-marked beads are used, the lasers, detectors and correct compensation are checked.

Compensation between detectors today is almost always software-based, because a manual correction multi-color fluorescence is very difficult to implement in day-to-day practice.

Appropriate control materials must be included for the analyses performed, if available. In general, stabilized blood is used. The composition to be expected is known at least for the basic parameters. Each working day, at least two control samples with target values in different concentration ranges are to be analyzed.

The parameters of the immune status and of the determination of CD34 stem and progenitor cells must be evaluated continuously according to the stipulations of RiliBÄK and/or the standard. Given the short batch life of many control preparations (<12 weeks), manufacturer limits generally apply (control preparations, such as CD-Chex^®, CD-Chex^® Low, Immuno-Trol™ cells, Immuno-Trol™ Low cells, BD Multi-Check™ Control, BD Multi-Check™ CD4 Low Control, CD-Chex^®CD34, BD™ Stem Cell Control, Stem Trol™ Control cells).

In hematologic samples for leukemia or lymphoma typing, this is possible only in few cases (CD Chex Plus, CD Chex BC, CD-Chex™ Select). Otherwise, control material will be replaced by plausibility checks to be performed within the analytical process.

The regular maintenance of equipment, compliance with all framework conditions of the analytical process (see below) and the processing by competent staff thus allow for the correct and clinically relevant analysis of the samples. In addition, the equipment used must be compared regularly [16].

The analytical process in the flow cytometry laboratory

The flow cytometry laboratory is typically integrated into the routine laboratory and interacts with the basic rules and structures, as well as work and analytical processes of other departments. The diagnostic process is comparable to that of other departments, but also has some special characteristics that are shown in Tables 1–3.

The selected anticoagulation matters in pre-analysis, and some tests have specific requirements, such as the calcium-based function test [29]. The samples must reach the laboratory within the time that is specified by the test manufacturers in the specification of their IVD. Generally, 24 h is the time frame that has proved practical for many immunophenotyping tests for pre-analysis purposes. Cell stabilizers may be used if necessary, but this must always be validated in each individual case. Identification of the sample is crucial to the correct allocation of the findings.

Also important in this context is the evaluation of samples on the basis of acceptance criteria. As DIN EN ISO 15189 states in 5.4.6, authorized staff must verify the obtained samples to ensure that they meet the relevant acceptance criteria for the required test(s) [3, 25]. According to the Immunology checklist, the identity and integrity of samples must be checked “appropriately”. Another requirement involves a set of written criteria for rejecting unacceptable samples or for the special treatment of suboptimal samples [4].

An adequate quantity of samples, that is, cell number, and a high vitality of cells are necessary for diagnostics. Today’s test procedures mostly involve whole blood without washing steps. Antibodies and fluorochromes are selected as part of the validation and corrected if necessary. Finally, it is the gating strategy that determines the correct measured results. Nowadays, the use of an absolute cell count measurement in the cytometer is recommended, that is, a single-platform technology [13].

Taking advantage of the age-based reference intervals, the flow cytometry findings report is then created in an indication-specific manner, but this requires access to the original plots [4].

External quality controls and laboratory comparisons

The implementation of external quality assurance in the form of external quality assessment schemes is mandatory in Germany, both under RiliBÄK [2] and DIN EN ISO 15189 [3, 25]. external quality assessment schemes or inter-laboratory tests serve as external quality and plausibility checks in comparison to the findings of other laboratories and specialist areas (inter-laboratory comparison). RiliBÄK makes participation in one external quality assessment scheme per quarter mandatory for the parameters in Table B1 a to c (RiliBÄK). If a user has exceeded the permissible deviation (column 5) of Table B1 a to c, he must, according to RiliBÄK, clarify the causes and – where he is capable of doing so – eliminate them. The flow cytometric parameters CD4 T-cells and CD8 T-cells are listed in RiliBÄK Table B 2-2 together with the manual differential blood count under qualitative laboratory tests. External quality assessment schemes are required every 6 months, or every quarter in the case of blood smears. The need for a new classification within RiliBÄK was already emphasized under “Guidelines of the German Medical Association (“RiliBÄK”)” [13–15].

Since there is no recognized reference method in the field of immunophenotyping, an evidence-based foundation is to be used for the future definition of coefficients of variation used in internal and external quality control [10]. In the cited paper, the coefficients of variation (CV) of lymphocyte subpopulations from 14 laboratories were examined. One conclusion of the authors was that the CV for round-robin testing of CD4 could be set to <15% (normal CD4 concentration) or <20% (low CD4 concentration) [13]. Another conclusion of the authors was that the reference institutions would have to be told to establish reference methods and ensure greater transparency in the representation of the quality of external quality assessment schemes over the long term.

National and international proficiency testing (PT) organizations offer external quality assessment schemes for flow cytometric tests. Table 6 provides an overview of external quality controls currently available.

The German PT organizations INSTAND e.V. and the Reference Institute for Bioanalysis are reference institutions named by the German Medical Association. Both institutions have been accredited by the German Accreditation Body (DAkkS) under DIN EN ISO/IEC 17043 as providers of external quality assessment schemes. The current test programs are available on the respective websites [30, 31]. The providers use different test materials (samples), such as fresh blood, stabilized blood, control material, listmode data on a CD and digital image material. This may be quite demanding in terms of logistics. Fresh blood samples for external quality testing are taken from donors the day before, divided into aliquots, labeled, and sent to the scheme participants by overnight courier. The duration of shipment must be documented and be traceable in case of complaints. The inter-laboratory test participants will be informed that the material should be measured as soon as possible upon arrival. A reordering of samples is not possible in such cases. The preparation and quality assurance of external quality assessment samples, as an entire process, is subject to quality control and has to be strictly documented.

As can be seen in Table 6, external quality assessment schemes are nearly covering the entire spectrum of flow cytometric analysis. Here, according to the current board resolutions of the sector committee on medical laboratory diagnostics, external quality control is still mandatory for test methods for which there are no national external quality assessment schemes [16]. This can either be in the form of an inter-laboratory comparison or, alternatively, can be performed as participation in an international external quality assessment trial [16]. It should be emphasized that external quality testing of equipment manufacturers and test providers cannot replace independent external quality assessment schemes. The frequency of conducting inter-laboratory comparisons should be based on the frequency of round-robin tests. If possible, an inter-laboratory comparison involving at least two samples (different cell concentrations, positive/negative, etc.) should be carried out twice a year [3, 16]. Like the results of round-robin tests, the results of inter-laboratory comparisons must be documented and evaluated. In the event of failure, appropriate measures of problem analysis must be taken [3].

Common quality problems

Improvements to the flow cytometer have led in recent years to devices that provide stable and reliable measurements. Every day that they are started up, they require quality checks, and document relevant quality information. Still, audits and quality circles encounter typical problems that can and should be avoided.

The following Tables 7–9 give an overview of the most common quality problems.