CHMP Adopts Guideline on Genomic Sampling and Management of Genomic Data
On 14 September 2017, the Committee for Human Medicinal Products (“CHMP”) of the European Medicines Agency adopted ICH Guideline E18 (the “Guideline”) on genomic sampling and the management of genomic data. The Guideline takes effect on 28 February 2018.
The International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (“ICH”) developed the Guideline in acknowledgement of the growing awareness of, and interest in, genomic data generated from clinical studies. The ICH suggested that the absence of a harmonized guideline made it more difficult to conduct genomic research consistently in global studies. The fact that the CHMP has adopted the Guideline means that EU guidance on this subject is now aligned with the ICH standard.
The Guideline provides general principles for the collection and handling of genomic samples and management of genomic data. It also affirms broader principles, such as the need for informed consent and the protection of subjects’ privacy etc. The Guideline applies to both interventional and non-interventional clinical studies, irrespective of when the genomic research is carried out and whether it was envisaged in the study protocol. The ICH/CHMP intend the Guideline to be interpreted in accordance with the law and policies in each jurisdiction where genomic research takes place.
Key aspects of the Guideline include the following:
The Guideline strongly encourages taking genomic samples in clinical studies. This should be systematic, unbiased and subject always to consent. The number and quality of samples affects the reliability of the data produced: maintaining sample integrity is therefore important.
There should be documented standardized procedures for genomic sample collection, processing, transport, and storage so that specimens are suitable for their intended use and to ensure multi-site data are reliably comparable. This includes: collecting the appropriate type and volume of samples; having appropriate extraction and preservation methods (particularly with regard to the specimen type); maintaining specimen stability and minimising interference and degradation (including during transport and throughout storage).
The Guideline recommends long-term storage to enable future use as long as this is in line with subject consent, followed by sample destruction. This requires the proper curation and administration of a sample inventory.
Managing Genomic Data
It is important to select the appropriate platform and method to generate and manage genomic data for the intended purpose (e.g., whether data are for clinical decision-making or more exploratory use). The Guideline raises a number of technical points on data generation and analysis.
There is a focus on storing genomic data over the long-term (subject to consent), with the possibility to link to other data. This both maintains the integrity of the study and enables future use of the data.
Privacy and Confidentiality
Genomic data should be treated with the same high standards of confidentiality as other clinical data, in line with applicable privacy and data protection rules. The Guideline recommends single coding for genomic samples and data, highlighting the limitations of anonymizing genomic data.
Researchers should give thought to who accesses genomic data: research access may advance medical science overall; however researchers should take a risk-based approach, documented in appropriate policies and procedures.
The Guideline reminds researchers of the importance of obtaining informed consent (per ICG Guideline E6). Informed consent for the collection and use of genomic samples should permit a broad analysis and ideally allow for a broad application (e.g., in assay development, disease research, drug research, and pharmacovigilance).
The Guideline encourages researchers and sponsors to establish a mechanism to communicate data/results to subjects and/or clinicians, particularly where there is clinically actionable data.