Critical quality attributes measured by flow cytometry

RELIABLE CELL CHARACTERIZATION: CRITICAL QUALITY ATTRIBUTES MEASURED BY FLOW CYTOMETRY

Defining the physical and biological characteristics that qualify as features that ensure acceptable quality of a therapeutic cell product (critical quality attributes) can be challenging as they should reliably and accurately reflect functionality, which – in turn – has to indicate its anticipated therapeutic effect after it’s been injected into the patient.

Flow cytometry can be used to quickly and accurately measure a variety of critical quality attributes of therapeutic cell products. The technique is particularly suited for measuring multiple features (sometimes up to 40) simultaneously on single cells for millions of cells in a short period of time. Critical quality attributes typically assessed by flow cytometry are: morphological features, membrane integrity, DNA content and expression, presence or absence of specific sets of extracellular and/or intracellular proteins. Specific sets of proteins on or in a cell, are generally identified by means of specific sets (panels) of fluorochrome-conjugated monoclonal antibodies raised against epitopes that are (ideally) uniquely part of the proteins of interest. Thus, proper panel design can allow for accurate measurements of highly specific cellular features. As more features are measured by multi-parameter flow cytometer, the accuracy of the quality assessment increases. Often, the more cellular features are considered for panel design, the higher the accuracy of the cell product characterization measurements.

In contrast to accuracy (i.e. close to the true value), most flow cytometry assays do not reach the level of reproducibility (i.e. low variability between measurements) that is required for manufacturing of therapeutic cell products. This is mostly due to variability introduced by a range of differences in sample preparation procedures, reagents, instrumentation and data analysis. For instance, differences in source material and differences in logistics and cell processing procedures for manufacturing of therapeutic cell products demand different sample preparation methods. Features of the targeted protein itself can demand specific sample preparation methods too (e.g. the need to permeabilize cell membranes for detection of intracellular proteins). Different sample preparation methods may require different reagents (i.e. antibodies) to detect a particular protein as chemicals used for fixation and/or permeabilization can alter an epitope such that it is no longer recognized by an antibody normally used on cells that have been handled differently. Sadly, awareness of sample preparation-induced alterations of epitopes is spars and antibody quality often receives the blame for unexpected assay results. Moreover, ambiguous language used in assay instructions can be a major source for errors that lead to unexpected assay results. It is not uncommon that significant dilution errors occur due to confusing language used in instructions, such as to ‘make a 1:1 dilution’ or ‘make a 1:2 dilution’, as both are sometimes interpreted as ‘1 part into a total of 2’. Antibody binding characteristics can be significantly different between a 1% formaldehyde and a 2% formaldehyde solution. Obviously, the boundaries of effective antibody binding in relation to sample preparation methods should be addressed in qualification studies. Aside from using clear, unambiguous language to minimize mistakes, qualification tests should also include possible conditions created due to (likely) human error (e.g. test “1 part in total of 2” as well as “1 part in total of 3”).

Between the wide range of cellular proteins and available antibodies conjugated to different fluorochromes for detection, there is a great number of combinations possible to identify one specific quality attribute. Unfortunately, specific combinations of different fluorochrome-conjugated antibodies are – in large part – determined by how it suits the technical capabilities of the instrument. In addition to the technical capabilities of an instrument, its performance has an impact on panel design as well. As shown in a collaborative study between NIST and the International Society for Advancement of Cytometry (Hoffman et al., Cytometry 2012; 81A: 785-796) performance can vary highly – even within the same make and model of instrument. Remarkably, younger instrument companies appear to ensure consistency in instrument performance more than established companies. Likewise, pronounced differences have been noted amongst a selection of reference materials designed for instrument setup.

Accordingly, when a critical quality attribute requires panel design, technical differences between flow cytometers generally cause different flow cytometry professionals to find different ways to (accurately) measure the same critical quality attribute. In characterization of CAR T cell products, for example, the panel may consist of 4 (anti-CAR, -CD3, -CD19, -CD56) or 10 (anti-CAR, -CD3, -CD4, -CD8, -CD19, -CD11b, -CD45RA, -CD45RO, -CD56, -CD62L) different antibodies in various combinations of fluorochromes, routinely depending on the technical specifications of the instrument.

Throughout the past roughly 25 years, a number of consortia has initiated the standardization and harmonization of antibody panels to reliably identify specific cell types for potential use as critical quality attributes in diagnosis of a variety of hematological diseases. A few of these initiatives are: the European Research Initiative (ERIC), the Immunoguiding Program of the Association for Cancer Immunotherapy (CIP), the EuroFlow Consortium, the ONE Study and, the Human Immunophenotyping Consortium (HIPC).

Another initiative worth mentioning is OMIP (Optimized Multicolor Immunofluorescence Panel). This publication format is applied by the scientific journal Cytometry-Part A, and aims to streamline panel design for multiparameter flow cytometry. An OMIP publication requires authors to submit documentation showing that their unique combination of reagents has been optimized relative to other combinations of reagents.

It is thanks to these kind of initiatives that great progress has been made in the accuracy at which specific cell types can be identified by flow cytometry. However, insufficient or lack of standards for sample preparation methods, reagents, instruments and data analysis, hinders improvements on reproducibility and generates large qualification studies in order to be able to define a critical quality attribute for a therapeutic cell product that can be reliably measured by flow cytometry.