Embracing personalized medicine, the European Medicines Agency (EMA) released a draft guidance document titled Draft Guideline on Good Pharmacogenomic Practice(“Guidance”) on good pharmacogenomic practice with the stated goal of developing and laying out appropriate genomic methodologies for use during the development and life-cycle of a drug. Past issues and problems that have been encountered in studies utilizing genetic variation in drug responses are reviewed and recommendations for further improvement are provided by the Guidance.

Through issuance of this Guidance, the EMA signals its support of the use of biomarkers to streamline the drug approval process and facilitate clinical trials while recognizing that there are currently no guidelines for good genomic practices. Guidance at page 3. For example, despite the numerous important variable genetic loci revealed in pharmacogenomics research, some clinical studies have resulted in “ambiguous findings,” which highlight the importance of correctly measuring, interpreting and translating pharmacogenomic data into clinical treatment. Guidance at page 5.

Examples of pitfalls observed in prior studies include the poor quality of employed analytics, lack of appropriate patient selection, lack of appropriate phenotype identification, lack of power in relation to the frequency of the genetic variation studied, irrelevant endpoints having been selected for the basis of the study, and failure to take into account the pharmacology of the drug in the design of the study. Guidance at page 5.

The Guidance also reviews and analyzes issues that arise when utilizing genomic data during the drug discovery process, such as phenotyping, the heterogeneity of tumor biopsies, the importance of quality DNA sequence design, quality aspects of the pharmacogenomics analysis, clinical study design, as well as the current use of biomarkers and their translation into clinical use, as well as the future dynamics of drug labels. The following highlights several recommendations and cautions from the Guidance.

Variability in Tumor Sampling

The tumor genome presents quality issues with samples obtained from tumor biopsies. Therefore, the heterogeneity of the tumor sample should be considered to minimize inter-sample differences. Guidance at page 7. For liquid biopsies (which involves the isolation of circulating DNA (ctDNA)), the Guidance notes that liquid biopsies may overcome some of the limitation of tumor heterogeneity, and it is therefore recommended that consideration be given to analyzing ctDNA in parallel with tumor biopsies.

DNA Sequencing Design

The Guidance notes that several issues require careful consideration when designing genomic sequence analysis. For example, genomic variations that are linked to functional importance should be studied and appropriate methods for DNA isolation should be employed for high quality samples. Critical sequencing results should be validated either by independent analysis or by re-sequencing. Care and caution should be followed when relying on published sequence databases and algorithms of relevance should be used when utilizing bioinformatics methods.

Pharmacogenomics Analysis

Quality issues can arise prior, during and after analytic analysis of the patient sample.  Care in sample preservation, labeling and transport is recommended.  With respect to the analysis itself, the Guidance recommends that a second independent testing method be used to validate the primary results. In situations where broad sequencing approaches are utilized, previous unknown mutations may be detected. The Guidance cautions that if follow-up studies are not instituted in these situations, irrelevant and /or incidental findings may be reported. Guidance at page 10. In addition, there is a stated need for uniformity in reporting biomarkers because the nomenclature of biomarkers is currently “heterogeneous in form and content.” Guidance at page 11.

Study Design

Genomic variation also needs to be considered in the design of clinical trials and is considered critical for a successful outcome. In the design of these trials, the Guidance recommends that analytical validity of any assay testing new genomic biomarkers be demonstrated as early as possible during clinical development so that “early clinical findings can be considered relevant to later clinical development.” Guidance at page 13.

Public Comments are Welcome

The Guidance notes that these recommendations apply to Marketing Authorization Applications for medicines for human use and they should be read in conjunction with other relevant EU and ICH guidelines and reflection papers. A non-limiting list of guidelines and reflection papers is noted on page 4 of the Guidance.

The public is invited to comment on the draft guidelines until September 16, 2016.