Applying context of use to quantitative polymerase chain reaction method validation and analysis: a recommendation from the European Bioanalysis Forum

Since the advent of polymerase chain reaction (PCR) in 1983 [1], and the subsequent development of real-time or quantitative PCR (qPCR) [2], the technique has been used widely in research and has for many years been the gold standard in diagnostic testing for various viruses and bacteria. Since the late 2000s, qPCR assays have been adopted in the development of pharmaceuticals, including application in biopharmaceutical quality control (e.g., residual host cell DNA measurement), genomic analyses, determination of vector copy numbers, measurement of pharmacodynamic (PD) and biomarker (BM) end points.

With the rapidly growing discipline of translational science and medicine to expedite the discovery and development of pharmaceutical drugs, qPCR applications are increasingly used as biomarker assays to understand the biology and mechanism of action.

In particular, qPCR-based assays have become an important bioanalytical technique for gene therapy-based medicinal products (GTMPs or GT) and cell-based human medicinal products (CBMPs or cell therapy [CT]), collectively referred to as advanced therapeutic medicinal products or cell and gene therapies (CGT). In the remaining document CGT, CT or GT will be used as the common terminology.

qPCR & the European Bioanalysis Forum

The need to discuss qPCR practices and harmonization between bioanalytical laboratories came to prominence in the European Bioanalysis Forum (EBF) in early 2018 and a team was formed of qPCR experts from several EBF member companies. The team shared considerations and current practices with qPCR applications in the bioanalytical laboratory and collected experience on actual qPCR method evaluation, characterization and validation approaches applied. These considerations were presented at a workshop at the 11th EBF Open Symposium [3] in collaboration with the Japan Bioanalysis Forum, who provided feedback on their discussions and considerations on the topic [4]. With the formation of the EBF CGT team in 2019, here too bioanalysis by qPCR was an important discussion topic. The discussions in both EBF teams are the backbone on which this recommendation paper is built.

Although specific considerations for digital droplet PCR (ddPCR) are not included in this manuscript, many of the scientific considerations for qPCR discussed here are applicable to ddPCR.

Connecting qPCR with biomarker context of use

Considering qPCR applications within the field of translational science and medicine as biomarker assays to understand the biology of physiological and pathophysiological processes and the mechanism of action of therapeutic agents, such as altered gene expression, stimulated the team to reflect on the similarities to the challenges that industry encounters for BM assay development and validation. For BM assays, the concept of context of use (COU) for assay development, characterization and validation is becoming a well-established approach [5,6], with an understanding that the bioanalytical method validation (BMV) and regional guidance documents for PK assays [7–13] are often inappropriate in the outlined assessments, experimental design and the associated acceptance criteria when defining the requirements for a BM assay validation. Similar to BM assays, one current issue developing in the regulated bioanalytical community is that data from qPCR assays assessing biodistribution, exposure, migration, persistence, decay and shedding of a CGT, can be regarded as PK data, potentially used in interpretation of efficacy and safety. Hence, it can be erroneously interpreted that these qPCR assays require validation following existing regulatory guidance/guidelines for PK method validation [7–13].

The learnings and outcomes throughout EBF discussions on qPCR assays also illustrate how bioanalytical challenges resemble those for biomarkers, for example:

• There are many different COU for the various qPCR applications;

• With the wide range of different COU, each COU has its own performance requirements for the qPCR method;

• There is a desire for the harmonization of bioanalytical qPCR approaches, with harmonization of the terminology and definitions being the most imperative.

The focus of this manuscript is qPCR bioanalysis in tissues and biofluids; methods used for biopharmaceutical product characterization and release are outside the scope. With this manuscript, the EBF propose that qPCR assays should be developed, characterized and validated for their COU to ensure that assays generate data with appropriate scientific confidence to support any decision necessary to efficiently move a prospective therapeutic through the development and approval stages, and ultimately improve the health and quality of life for patients around the world. The EBF views that the existing regulatory BMV guidance/guideline expectations written for PK assays using chromatographic and ligand binding assay technologies are generally not suitable for PCR technologies due to the fundamental differences of these analytical technologies from sample processing through to final assay readouts. Blinkered copying of BMV guidance experimental design and acceptance criteria to qPCR assays risks inappropriate assay design and may lead to false confidence in poor data and/or excessive use of time and resource. Instead, specific requirements should be defined for the COU, then the validation may then be designed to confirm that the performance characteristics of the assay are suitable and reliable for this intended use. Method validation cornerstones such as precision, (relative) accuracy, analytical measurement range, selectivity, specificity, stability and linearity of dilution will often be among the considerations.

We want to point out that other publications such as the Minimum Information for Publication of Quantitative Real-Time PCR Experiments (MIQE) guidelines are also relevant to give well established recommendations for experimental practice to allow reliable interpretation of qPCR/ddPCR results [14,15].

In summary, it is EBF recommendation that validation experimental design and associated acceptance criteria are specifically developed for each individual parameter and designed around the type of data and the decisions being made for an individual qPCR method. In continuation, the manuscript dives deeper into the scientific and regulatory background on which this EBF recommendation is built.

qPCR considerations in guidelines applicable to CGT

The limited information on regulatory expectations for qPCR assays included in the guidelines support the analytical laboratory in using an appropriate scientific approach for method characterization [16–26]. None of these guidelines refer to BMV.

For biodistribution studies and longterm persistence assessment of CGTs, the guidelines request the qPCR methods to be quantitative and sensitive and to specifically detect the vector sequence in animal and human tissues/fluids [19,20,22–24]. In the current US FDA guidance, a LOQ of ≤50 copies/µg genomic DNA with 95% confidence is requested to be demonstrated and documented for nonclinical biodistribution [24]. The FDA guidance also includes information that the samples should at least be run in triplicate for each tissue. To aid the interpretation of the qPCR assay results, one replicate of each tissue sample should include a spike of control DNA, including a known amount of the vector sequences. This in-study validation of the assay performance, using a spike control can also be referred to as qPCR inhibition assessment. It also can be interpreted as the qPCR alternative to perform an in-study validation to confirm analyte recovery and sensitivity of the qPCR assay in study samples. Guidance for qPCR assay sensitivity in biodistribution, persistence and integration studies for plasmid DNA vaccines are also available with a recommendation for a quantification of <100 copies of plasmid per µg of host DNA [22].

In the Draft ICH S12 guideline on nonclinical biodistribution assessment, a quantitative determination of the amount of genetic material of the GT product and, if regarded necessary, of the expression products is requested [20]. qPCR is regarded as the ‘gold standard’ for such an assessment and the limit of sensitivity as well as the reproducibility of the method is requested to be established and documented. A minimal sensitivity value, such as the 50 copies/μg genomic DNA, is not given in this new draft ICH guideline. To demonstrate the ability of the method to selectively detect the target sequence in different tissues and biofluids, the performance of spike and recovery experiments during method development is expected. A core panel of tissues/biofluids for biodistribution assessment as well as considerations on additional matrices is also provided in the draft ICH S12 guidance. The EBF wants to remind the audience that the ICH S12 is still in draft and should not be used for any regulatory purpose until final and adopted.

To assess shedding, qPCR methods are usually the first approach to measure the shed vector DNA, and in the case of replication incompetent or deficient products probably the only suitable method [19,23]. Shedding assays are requested to be sensitive, accurate, reproducible and specific. According to guidelines, the ‘fit-for-purpose’ assay should be qualified and meet minimal performance capabilities. Specificity and assay interference considerations are considered of special importance for qPCR assays used for shedding analysis due to the special nature of matrices like stool or saliva, which do not only contain host proteins and host nucleic acids but also the ones of the body’s natural flora. Recommendations given in the guideline include the use of an interference control to exclude shedding underestimation due to inhibition of the qPCR reaction and addition of a reference standard or an internal positive control to determine the extraction recovery.

Beside these limited assay requirements, the aforementioned guidelines do not specifically consider details around assay validation or validation reports [19,20,22–24]. The guidelines only indicate that the methods need to meet minimal performance capabilities or need to demonstrate capability to specifically detect the vector sequence in the different matrices.

In conclusion, considering the current regulatory environment has very limited guidance for qPCR-based assays, it is the responsibility of the scientific expert community to help shape a sustainable regulatory environment building on scientific excellence and prevent undue scope creep by copying requirements from guidelines not written with qPCR in mind.

COU considerations for the decisions to be taken from the qPCR data

It is well known that qPCR assays can be used for different purposes in nonclinical and clinical studies. Hence, and comparable to biomarker assays, a simple and straightforward ‘one-size-fits-all’ classification of qPCR assays may be difficult or even impossible to design. While qPCR assays have been commonplace in the CMC space for quite some time, they are often relatively new to bioanalytical laboratories and newly recruited bioanalytical qPCR staff may mistakenly be trained to use BMV standard operating procedure’s rather than using a best scientific practice mindset. Again, in line with our recommendation on biomarker assay validation [5,6], EBF strongly encourages bioanalytical laboratories to involve the stakeholders and qPCR experts, in other words, those having experience with qPCR analysis in research and diagnostic settings, in the dialogue.

The first and most important consideration when establishing a qPCR assay for tissues and biofluids is to understand the anticipated use of the assay and its data. A few areas of application, including our view on the method requirements looks like this:

• Gene and cell therapy (GCT), for example, viral vectors, viruses, RNA, DNA plasmids, CAR-T cells and stem cells

  • Assessment of biodistribution, migration and persistence analysis, for example, to support PD and toxicological studies (nonclinical)

    • Method requirement: accurate and sensitive quantification of the vector DNA or RNA over its whole analysis range

  • Shedding analysis, especially in clinical studies to assess environmental risks

    • Method requirement: main focus on the LOD; when has the GT DNA cleared from the matrix and when can the patient stop preventive measures

• Gene expression and biomarkers, for example, efficacy (PD) end points, gene regulation and expression, target gene knock out, genotyping, liver damage and circulating tumor cells

  • Method requirements: diverse; from qualitative to quantitative, absolute versus relative quantification, including for example normalization to a housekeeping gene, baseline or amount of DNA

It should be clear from the above high-level overview of potential areas of application, assay readouts and decisions taken on the qPCR assay data that tunnel-visioned application of BMV guidance should not be the default approach, and it illustrates the importance of considering the concepts of COU.

Examples of different COU for qPCR assays are given in Table 1. More in-depth bioanalytical considerations will be given in a follow-up publication building further on our current recommendation which will be published at a later timepoint.

Clearly, for each qPCR application, communication with relevant stakeholders needs to be established to define the COU of the assay and thus to be able to plan for an appropriate qPCR method development, characterization, as well as appropriate validation and sample analysis strategy. The considerations should include the decisions which will be made based on the qPCR data and how data will be reported.

Conclusion

The objective of this manuscript is to highlight important points to consider when establishing and validating qPCR assays, and to promote the recommendation that validation of such methods should be driven by scientific principles and the intended use of the data (context of use), principles which apply independent of the technology platform used or the analyte. In presenting our extensive industry experiences, we also hope to engage in constructive conversation with regulatory bodies to ensure current and future expectations for qPCR are based on solid scientific principles appropriate to the technology platforms and the decisions being made with the data. It is not the intention of EBF to request specific guidance(s) for bioanalytical validation of qPCR methods used to support CGT biodistribution, shedding studies or biomarker investigation. The EBF team aims to continue to work to share experience including recommendations for how to use COU for qPCR assays and identify areas where harmonization may be possible/desirable. In continuation, the EBF is committed to engage with all stakeholders and regulatory bodies to share our experience and our vision on applying qPCR in the regulated bioanalytical laboratory at conferences or via scientific publications, including evaluation of case studies and considerations for assay methodology, characterization and sample analysis.