Facilitating access to clinical data from a randomized controlled trial of complex adaptive design

Abstract

Objectives

To describe the steps taken in the creation of a Data Pack containing clinical data from the FOCUS4 randomized controlled trial in colorectal cancer, representing a case study of clinical trial data access and sharing.

Data description

The FOCUS4 trial recruited 1434 metastatic colorectal cancer patients between January 2017 and March 2020 from 94 sites across the United Kingdom (UK). The trial used an adaptive “umbrella” design, in which tumour samples from recruited patients underwent molecular testing, the results of which determined into which of multiple parallel randomized allocations they would be eligible to enter. Overall, 361 patients were successfully randomised into one of four “sub-trials”: FOCUS4-B (6 patients), FOCUS4-C (67 patients), FOCUS4-D (32 patients) and FOCUS4-N (254 patients). After the trial had concluded, we created a Data Pack containing pseudonymised data organised into registration data and follow up data, marked case report forms (CRFs), data dictionary, study protocols and statistical analysis plan. This will help researchers to easily understand the datasets for study replication, further research analysis or to integrate with other existing datasets or repositories.

Objective

Clinical trial datasets have enormous potential beyond the publication of the final trial report which generally summarizes the primary analysis. Onward sharing of trial data is now generally accepted to be of benefit to the research community, and some funders and journals now require that data be made available on reasonable request [1–4]. FOCUS4 was a randomized trial of complex “umbrella” design [5, 6] run by the UK Medical Research Council Clinical Trials Unit (MRC CTU) at University College London (UCL) to efficiently test the safety and efficacy of multiple molecularly-targeted therapies in metastatic colorectal cancer (mCRC) [6–10].

The MRC CTU is working on improving access to the clinical trials data where participants consented to have their data to be used in further research [3]. To facilitate this, pseudonymised dataset will be prepared, to be shared with researchers who have applied to either replicate the study, integrate these data with other datasets or carry out an in-depth analysis. Going forward, the aim is for this process to be applied to all trials once results and findings have been made available in respect to the a priori stated aims and objectives and the trial has been formally closed by the Sponsor. This process aims to make MRC CTU trial data findable, accessible, interoperable and reusable (FAIR) [8, 10–14].

Data description

The FOCUS4 trial recruited 1434 patients between January 2017 and March 2020 from 94 sites across the United Kingdom (UK). Recruited patients underwent molecular testing on material from their primary tumours, the results of which determined which of multiple parallel randomized allocations they would be eligible to enter. Overall, 361 patients were successfully randomised into one of four molecularly-stratified “sub-trials” within the umbrella structure, as well as the “non-stratified” sub-trial N [8, 10]. FOCUS4-D recruited 32 patients, and demonstrated that AZD8931 did not improve outcomes against the standard treatment for tumours of the “all wild-type” molecular profile [12]. FOCUS4-C recruited 67 patients, and demonstrated that AZD1775 was effective and safe among patients with tumours with RAS and p53 mutations (10). FOCUS4-B evaluated whether the impact of taking daily Aspirin improved the survival of patients whose tumours showed the PIK3CA gene mutation, but experienced poor recruitment (6 patients) and was unable to generate conclusive results (3). FOCUS4-A was never activated due to unavailability of the drugs for testing. FOCUS4-N was a “non-stratified” arm which ran concurrently throughout the lifetime of the trial, suitable for patients who were not eligible for the available sub-trials (11), or who declined to enter the sub-trial for which they were eligible. FOCUS4-N recruited 254 patients and demonstrated that capecitabine oral maintenance is effective and safe among patients responding to first-line therapy.

We organised the data into two sets, that is registration data and follow up data. In preparation for sharing, we pseudonymized the data to remove all identifying and identifiable participant information. We started by mapping all the variables in the dataset, determining the data types and assessing the potential risk of each to patient identification. The pseudonymization process included converting each participant’s randomization date to time zero, thereafter all other dates were calculated in days from the date of randomization. To permit assessment of trends over calendar time, we also converted date of birth into (non-identifiable) groups of age at the time of randomization. Individual recruiting site names were replaced by a grouping representing the facilities and experience of the site. Free text fields were redacted, as such fields have the potential to contain confidential information inappropriate for sharing.

Case report forms (CRFs) were available in PDF format, enabling mapping of values from the CRF to corresponding variables present in the datasets [15]. Each CRF question was mapped to its corresponding name in the dataset column, and the CRF electronic documents were annotated accordingly, including marking to show which values were available and which were not. This will enable researchers to quickly identify if a particular dataset will help answer their research questions without necessarily profiling the data first.

We then prepared a data dictionary, containing a list and description of all the variables that we made available [15]. The data dictionary also contains general descriptive notes such as the name and recruiting period of the trial, a list of recruiting sites, and contact information for the trial Principal Investigator(s) (PIs), details on registration and follow up data. Such details include variable names, data types, data labels, variable labels, associated CRFs, as well as variable format. The annotated CRF’s, data dictionary, study protocol and statistical analysis plan forms the data pack as shown in Table 1. The data pack makes it easy to understand the dataset beforehand which helps the researchers see if the dataset meets their requirements before they apply to access the data.

Table 1.

Contents of the data pack

Label	Name of data file/data set	File types (file extension)	Data repository and identifier (DOI or accession number)
Data file 1	Annotated CRF’s	Zip (.zip)	https://doi.org/10.5522/04/30920096 [15]
Data file 2	Study Protocol	PDF (.pdf)	https://doi.org/10.5522/04/30920096 [15]
Data file 3	Statistical Analysis Plan	PDF (.pdf)	https://doi.org/10.5522/04/30920096 [15]
Data file 4	Data dictionary	Spreadsheet (.xlsx)	https://doi.org/10.5522/04/30920096 [15]

Complex design features such as umbrella protocols are still relatively rare, although on the increase [5]. Therefore, researchers getting access to these datasets would be able to perform reproducibility analysis, which is part of knowledge sharing, training and advancements. This might help in planning and carrying out new trials with adaptive features, whether in the field of colorectal cancer or elsewhere. Getting access to these datasets also allows integration with other relevant clinical trial datasets, for example for evidence synthesis and for basic science discovery and validation work.

Limitations

Onward sharing of clinical trial data requires compromises to be made between accessibility to researchers, and protection of the individual participants such as via General Data Protection Regulation (GDPR) legislation. Standard procedures have been followed for the redaction or concealment of identifiable information, but it is possible that information of potential research value (for example, contained in free-text fields) may be lost as a result.

Conclusion

Complex design features such as umbrella protocols and use of multi-arm multi-stage (MAMS) approaches are still relatively rare, although on the increase [5]. Therefore, researchers getting access to these datasets would help in the reproducibility which is part of knowledge sharing and advancement. This might help in planning and carrying out new trials with adaptive features, whether in the field of colorectal cancer or elsewhere. Getting access to these datasets also allows integration with other relevant clinical trial or translational research datasets, for example for evidence synthesis. We have clearly stated how access to these datasets can be granted in the availability of data and materials section.

Abbreviations

CRF

Case report form

FAIR

Findable, accessible, interoperable and reusable

MAMS

Multiarm multistage

mCRC

Metastatic colorectal cancer

MRC CTU

Medical Research Council Clinical Trials Unit

PI

Principal investigator

UK

United Kingdom

Author contributions

Conceptualization: MS, VYE. Methodology: VO, DF, MS, VYE. Writing – original draft: VO. Writing – review and editing: VO, DF, LB, VYE, MS, TM. Supervision: DF.

Funding

FOCUS4 was jointly funded by the UK government (MRC/National Institute for Health Research Efficacy and Mechanism Evaluation Programme) and Cancer Research UK.

Data availability

Access to the data can be requested here: [https://www.mrcctu.ucl.ac.uk/our-research/other-research-policy/data-sharing/](https:/www.mrcctu.ucl.ac.uk/our-research/other-research-policy/data-sharing) . We have uploaded the data pack which give the researchers comprehensive information of the data on the UCL research data repository [15]. Any related queries on data access can be addressed to [mrcctu.datareleaserequest@ucl.ac.uk](mailto: mrcctu.datareleaserequest@ucl.ac.uk) . For more information on FOCUS4 trial, please visit [http://www.focus4trial.org](http:/www.focus4trial.org) .

Declarations

Ethics approval and consent to participate

The FOCUS4 Trial Programme was conducted in compliance with the approved protocol, the 1996 version of the Declaration of Helsinki, the principles of Good Clinical Practice (GCP), EU Directives 2001/20/EC Article 2 and 2005/28/EC and subsequent amendments, their implementation in national legislation in the UK by Statutory Instrument 2004/1031 and subsequent amendments, the UK Data Protection Act (DPA number: Z6364106), and the National Health Service (NHS) Research Governance Framework for Health and Social Care (RGF). This FOCUS4 Trial Programme and all current comparisons within it were approved by the NRES South Central Oxford – Panel C ethics committee and the Medicines and Healthcare products Regulatory Agency (MHRA). FOCUS4 is part of the NIHR Clinical Research Network (NIHR CRN) portfolio. The subsequent processing of participant data was done in accordance with consent forms signed by the participants, and that participant data was removed if they explicitly withdrew their consent.

Consent for publication

All the authors have read and approved the final manuscript for publication.

Competing interests

The authors declare no competing interests.