December 1, 2008 (Vol. 28, No. 21)

Sophia Fox

As Use Increases There Is Need to Transform General Expectations into Practical Reality

Until “BioPharmaceutical Flow Cytometry 2008”, there had been little opportunity for users of flow cytometry in drug discovery/development to get together, share experiences, and discuss pitfalls and future requirements.

Over 100 delegates from around the world attended the recent one-day meeting hosted by AstraZeneca. Attendees included small- to mid-sized biotech companies, contract service organizations, global pharmaceutical companies, and flow cytometry (FC) technology manufacturers.

The meeting was the idea of Ruth Coldwell, Ph.D., associate principal scientist, clinical pharmacology and DMPK, and Bob Craggs, Ph.D., senior research scientist, discovery bioscience, both at AstraZeneca, who recognized an opportunity to bring drug discovery and development colleagues working with FC together to discuss common issues.

The event possibly represented the first of its kind to focus on the applications of flow cytometry in drug discovery and development, explained one of the meeting’s organizers, Rob Jepras, Ph.D., investigator in the molecular drug discovery department at GlaxoSmithKline (GSK).

“FC has been a commercially available technology for over 40 years but still has its roots firmly embedded in the clinical pathology/hematology laboratory and academic research bench.”

GlaxoSmithKline’s faith in flow cytometry as a screening tool for drug discovery led it to team up with Beckman Coulter to develop an automated plate-handling FC system capable of increasing throughput even further.

Benefits in DD and Development

“The technology is ideal for monitoring cell surface and intracellular events mediated by target-binding protein and cell-based drugs,” pointed out David Gordon, senior manager immunoanalysis at Aptuit. “FC allows you to simultaneously measure multiple analytes, demonstrate mode of drug action at the cellular level, and carry out cell subtype, counting, and distribution analyses.

“The technique generates semiquantitative or quantitative endpoints and can be used for applications ranging from simple binding assays to more sophisticated functional transcription response assays that evaluate linked biochemical processes in heterogeneous populations of primary cells.” The potential uses of FC thus range from drug discovery and early development mechanistic/efficacy assays to immunotoxicology, bioanalysis (PD markers), immunogenicity, and biopharmaceutical production, according to Gordon.

Despite the obvious utility of FC in a nonclinical setting, manufacturers also still view the hospital and academic markets as their bread and butter, Dr. Jepras said. “Innovation in instrumentation, reagents, and software is, therefore, still largely targeted at these markets.”

One of the major issues facing FC in drug discovery and development is the “lack of clearly defined regulatory guidance, both for assay development and with respect to validated software and informatics,” pointed out Garry Takle, Ph.D., vp of operations at WuXiAppTec.

“Flow is one of the most widely used methods for cell analysis and monitoring and really needs to be brought out of the academic and clinical labs and into the GLP/GMP arena. The meeting served to highlight how many companies are already using flow extensively, and the real need to start formalizing what the general expectations are for the use of FC and its associated data-management tools in a regulated environment.”

Scientists at AstraZeneca are spearheading the drive to bring the lack of guidance to the regulators’ attention. Dr. Coldwell and colleagues Jo Cunliffe, Ph.D., and Nicola Derbyshire, Ph.D., senior scientists, clinical pathology, are working to draft a paper that will hopefully promote further discussion within the industry and the regulatory bodies.

“Because there are few standardized procedures or methods for FC, and as the flow-derived data can be subjective, initial attempts to draft validation procedures will only be able to suggest minimum requirements,” Dr. Derbyshire suggested.

While industry is only too willing to help develop the foundations of a flow-cytometry validation package as a starting point for the regulatory authorities, the task won’t be easy, according to Mark Wing Ph.D., group director, clinical and translational sciences at Huntingdon Life Sciences. “Over the last five or six years, increasing numbers of our clients have been asking us to use FC in support of preclinical and clinical development programs, particularly in biomarker applications.

“About 20 percent of our biomarker services now use FC methods, but nobody has really come to grips with how you would go about developing a validation framework for FC assays.” The problem is compounded further, he pointed out, because there are no guidelines for biomarker assay development, whatever the method used.

“Whereas the FDA BMV guidelines clearly set out how you go about validating a method for PK analysis, biomarker assays present a greater challenge and typically follow the fit-for-purpose approach to validation. The absence of suitable reference material makes the assessment of accuracy for most FC assays challenging.

“For a PK method, negative samples would be spiked with purified drug, to assess the method’s ability to recover that drug. There aren’t readily available purified populations of cells that would allow you to assess the accuracy of a flow-based assay, however. That said, truly quantitative PK analyses are possible by FC when the reference material is available.”

With accuracy evaluation problematic for biomarker assays, assessing precision is key to in-house validation of an FC assay, Dr. Wing continued. “You, at least, want to demonstrate that your assay performs the same each time, regardless of how often it is carried out. It may be inaccurate, but as long as it is consistently inaccurate, that may be sufficient to derive meaningful data.”

Assessing precision and reproducibility brings its own problems, he admitted. “The analytes, i.e., cells, aren’t particularly stable. Whereas with immunoassay or mass-spectrometry techniques, your analytes tend to be stable and can be frozen for repeated analyses over a number of days if necessary, cells don’t freeze well. The use of stabilized cell preparations as surrogates for the real analyte may well be the best way to get around this.”

Despite these issues, a number of companies are successfully using FC in an increasing range of applications. “We have successfully used FC to support GMP quality control of biopharmaceuticals, to assess drug potency in different species, and to measure biomarkers in preclinical and clinical trials as part of submissions to the regulatory authorities,” Dr Wing said.

Indeed, as long as you can adequately demonstrate what the technology can achieve by showing evidence and defining boundaries, there should be no real hindrance to submitting FC-derived data as part of a regulatory package, suggested Gordon. Aptuit has been using FC technology for the last couple of years, primarily for preclinical work in areas such as species specificity.

“When evaluating the effects of a drug in a nonhuman species, it is vital to ensure that your animal is relevant and reflects the activity of the drug in humans,” he explained. “Recent European guidelines on strategies to identify and mitigate risks for first-in-man studies highlighted the need to demonstrate the relevance of an animal model. With increasing numbers of protein-based therapeutics such as antibodies coming into development, FC technology has come into its own as a method of choice for examining whether the extent of binding of a protein-based candidate to human and animal cells is similar.”


Dr. Derbyshire stressed the need for controls as a means to help standardize methods and monitor assay variability, whatever FC application is undertaken. Commercial/in-house quality control (QC) material can be used to assess assay performance and accuracy over time, whereas the use of control animals can help to establish baseline data and aid in the interpretation of a meaningful change in toxicology studies.

“For our immunotoxicology work we use control animals every time we run an assay, and our immune cell function assays incorporate a positive QC. It’s important to carry out a battery of control tests to help ensure your FC assay is as reproducible and precise as possible and not just rely on historical or pooled data.”

Variety of Applications

Validation issues aside, FC is proving a robust and high-content method for a wide range of application. Dr. Derbyshire’s work in the clinical pathology department at AstraZeneca is centered on immunotoxicology, where FC is just one of a number of techniques in the analytical arsenal.

“Although FC is not necessarily used routinely, we do apply the technology if, for example, specific issues have been flagged for a compound. For example, when there are indications of drug-related changes in white cell counts or inflammatory responses or when prospective patients will be immunocompromised. In these cases, FC can answer specific questions, such as whether there is proliferation of certain subsets of cells, whether immune cell markers are up- or downregulated, and whether immune cells have been activated or cell function altered.”

At WuXiAppTec, FC is used primarily for services in biomarker analysis, but it is also exploited as an analytical tool in biomanufacturing, to monitor cell expression and clonal changes.

“We have a quality system in place that allows us to carry out GLP and GMP FC, to support clinical trials right through to Phase III,” Dr. Takle said. “Our intent is to leverage our existing client relationships as a contract manufacturing organization for cellular therapies, and expand our footprint into clinical trial support for these types of drugs and others using FC as a key analytical methodology.”

One of the major advantages of FC is that it can evaluate multiple analytes in heterogeneous populations of primary cells. “We can interrogate mixed populations of cells such as a blood sample and look at a number of markers in these multiple cell types after drug administration,” Dr. Jepras added. “This means we can look at the effect of a drug on a number of targets in just one cell type, or in a native population of different cell types, and still get a quantitative result. Exploring intracellular signaling, for example, would conventionally involve lysing a uniform cell line population and looking for phosphorylation using ELISA-type techniques.

“The method works because you get a nice large signal window, cell number isn’t a problem, and you can screen thousands of compounds. This approach is virtually impossible with primary cells, however, where the number of cells is invariably limited. Using FC, you can identify which cells you want to look at, sideline the rest, and still get a big signal window.”

The combination of multiplexing and cell-sorting capabilities with new bead-based resources for evaluating soluble analytes, and plate-based instrumentation for higher throughput, are expanding the potential uses of FC even further.

GSK’s faith in FC as a screening tool for drug discovery led it to team up with Beckman Coulter to develop an automated plate-handling FC system capable of increasing throughput even further. “The lack of available assays for some of our kinase-target programs led us to harness FC as a highly sensitive approach for detecting certain phosphorylation events,” Dr. Jepras explained. “We increased throughput by moving from tube-based to plate-based CF, but it was still necessary to add automation to allow the analysis of a more meaningful numbers of plates”.

Despite increases in throughput achieved by automation, flow technology still has a way to evolve before the capability to screen whole libraries of compounds will be achieved, he maintained.

“The addition of automated plate handling will not be enough. One option is to push flow-cytometer manufacturers to make their instruments better and faster, with better data-management and analysis capabilities. This will throw up new hurdles in terms of cell and liquid handling, however. Alternatively, we will need to look at the potential to implement new technologies such as microfluidics and/or completely new approaches to signal detection.”

Previous articleExecutive Compensation in Troubled Times
Next articleAlkermes Regains Commercialization Rights to Vivitrol in U.S.