For each sample in a clinical trial, the audit trail can extend to many thousands of entries, making data recording a challenge. In the UK, The Liverpool Cancer Trials Unit decided to replace its paper trails with a more efficient system – Dr William Greenhalf and John Boother describe how
To maintain both the safety and rights of clinical trial participants and to ensure the reliability and robustness of the data generated clinical trials are tightly regulated. In the EU the relevant statutes are the EU Clinical Trials Directive (20001/20/EC and 2005/28/EC). New EU Clinical Trials Regulations were approved by the European Parliament in April 2014 and are expected to become law throughout the EU in 2016.
In the UK, enforcement of the resulting statutes is delegated to the Medicines and Healthcare products Regulatory Agency (MHRA), which has full powers to inspect all records and arrest any individual considered in potential breach of UK law (or EU regulations). The MHRA has published guidance for the maintenance of regulatory compliance in laboratories that perform the analysis and evaluation of samples from clinical trials.1 Compliance with Good Clinical Practice (GCP) guidelines for the laboratory allow clinical labs to ensure the safety and efficacy of data and that it is repeatable, reliable, auditable and can be easily reconstructed and audited in a research setting.
Approximately 40% of clinical trials in the EU are being conducted by academics, foundations, hospitals, or research-networks (often referred to as ‘non-commercial sponsors’), according to the European Commission Public Health Department.2 In the UK, Experimental Cancer Medicine Centres (ECMCs) were established following a joint initiative from the UK government via the National Institute of Health Research and the UK charity, Cancer Research UK. The aim was to offer laboratory support to any clinical trial aimed at the cure or relief of cancer. ECMC status was awarded to Liverpool in 2007. The Liverpool ECMC worked with the Liverpool Cancer Trials Unit (LCTU), also funded by Cancer Research UK and which co-ordinates national and international cancer trials.
GCP laboratory facilities were established in many of the ECMCs, including Liverpool. The Liverpool GCLP facility collects, stores and analyses samples from trials, and all samples that could have an impact on the evaluation of trial outcomes are covered by legislation and should be amenable to MHRA inspection.
While the paper-based recording systems and computer-based environmental monitoring system originally used at the Liverpool GCLP facility met all the regulatory requirements, any internal audit of paper records could take many months for any study and would be prohibitively expensive. The challenge was to find a replacement for paper records that would be cost effective, could be externally inspected and provide a transparent audit trail.
The challenge was to find a replacement for paper records that would be cost effective, could be externally inspected and provide a transparent audit trail
With modern laboratory information management systems (LIMS) capable of tracking clinical trial collection kits/supplies, managing and tracking specimens, managing laboratory testing work flows and recording quality performance, creating laboratory reports and managing study-specific documents, the Liverpool GCLP facility worked with Autoscribe, a developer of LIMS, to establish a system to meet the rigorous requirements for clinical trials. Autoscribe’s Matrix Gemini LIMS was configured using its OneTime configuration tools to log each step in the workflows for each trial and store the log in a readily accessible format that allows an audit in minutes rather than days.
The Matrix Gemini solution is FDA CFR21 Part 11-ready and allows a Microsoft SQL server database to be created and updated, with audit points for each sample being added from workflows in the LIMS which mimic the wet lab processes used in the trials.
Since the wet lab workflows rely on the manufacture of kits with all the components necessary for the collection, processing and storage of samples, for each study kits are built in silico in parallel with their manufacture in the lab. Each kit is logged to a batch table on a password-protected structured query language (SQL) server database and each component of the kit is logged to a sample table. The kit must be manufactured in silico before the label for the kit can be printed. Quality control for each component of the kit (confirmation of its presence in the kit and correct labelling) must be carried out before the in silico kit can be stored or dispatched to site.
The printing of labels acts as the link between the in silico world and the real world. If the steps on the LIMS had not been completed there would not be any labels to stick on the tubes
The labels on kit components allow complete patient anonymity but give a link between the laboratory and clinical data. They also give an audit trail from research data back to the patient consent form, showing whether the data was collected legally, blind to trial outcomes and in line with patient wishes. This can be achieved without compromising patient confidentiality. When the fax sheet is received by the LCTU a patient code is generated to link the kit to the clinical details, this code is passed to the GCLP facility and logged on the LIMS. When the samples are received in the GCLP facility, the time is logged on the LIMS and batch numbers of reagents to be used in the immediate processing of the sample are also entered.
Centrifugation steps must be completed in silico, in parallel with the wet lab processing of the sample. The code for the specific instrument used is logged; this can be matched to the service record of the machine which is also recorded on the LIMS.
Any aliquot of a sample is recorded on the LIMS, generating new sample codes for each aliquot, which are used on labels printed from the LIMS. The database record of the aliquot carries a code for the parent sample, any combination of samples is also recorded by the generation of a new sample code and an update of the parent sample records to include the code of the daughter sample. In this way an audit trail is produced allowing any sample to be traced back to its original kit component and hence to the patient code and (via the LCTU) to the patient’s consent form.
Any assay carried out on the sample is matched to a defined ‘Test’ on the LIMS that records each step in the standard operating procedure and gives a test component that can be matched to entered test results in a sample results table.
Storage of each sample is recorded as a specific predefined location, with no option to store more than one active sample in the same location
Entry of the data from the test can be done manually, but is more commonly via comma-separated value (CSV) files, which allow data to be saved in a table format and which are generated from the equipment used in the analysis. The CSV file will match with a data file on the analysis machine that can be inspected to confirm data entry is an honest representation of the data generation.
Storage of each sample is recorded as a specific predefined location, with no option to store more than one active sample in the same location (for example, a unique code made up of the freezer name, the drawer in the freezer, the box in the drawer and the co-ordinate within the box). When the in silico workflow demands the removal of a sample from a location that location is released, but obviously a record remains of the original occupancy, once again allowing a full audit trail.
The Liverpool Cancer Trials Unit
Matrix Gemini is fully controlled by a non-hierarchical password access, showing who has made each entry and limiting access to those areas of the system appropriate for each user. The LIMS offers audit tools for quality assurance of all the data entered. For inspection the SQL database is available with every entry on the in silico kit preparation, shipment, sample receipt, processing, storage and testing. These records can be matched with selected laboratory book entries and equipment records. In addition, the flexibility offered by the OneTime configuration tools means that the system can be readily modified in the event of any future changes to the official statutes.
Liverpool was awarded status as only the second academic institution with Early Phase Clinical Trial accreditation in the UK
Following inspections by the MHRA that included Matrix Gemini, Liverpool was awarded status as only the second academic institution with Early Phase Clinical Trial accreditation in the UK. On the basis of this accreditation Liverpool has since attracted national and international clinical trials funded by companies such as Novartis, Boehringer, Merck, Kael GemVax and various charities including Cancer Research UK.
Trials served by Matrix Gemini span Phase 1 through to Phase 3. The number of clinical trials conducted in the EU between 2007 and 2011 fell by 25%1 and yet despite this background of an overall decrease in trial activity, turnover from clinical trials in Liverpool has increased three-fold since adopting the Matrix solution.
Dr Bill Greenhalf is Operational Director of the Liverpool GCLP facility and Reader at the University of Liverpool.
John Boother is Managing Director, Autoscribe