GOSH is the custodian of the most recent study dataset (“IMPORT” - 650 patients, 2012-2019). All of the trial/study datasets include: Full patient and tumour demographics, including associated congenital abnormalities, tumour stage, treatment received (surgery, chemotherapy, radiotherapy) with individual drug names and doses and additional pathological details on histological subtype (risk group), follow up for relapse and death. In addition, the most recent study – “IMPORT” (Improving Population Outcomes for Renal Tumours of childhood) contains data on the presenting symptoms that led to diagnosis of kidney cancer, an imaging archive of CT and MRI scans sent for central radiology review and biological samples (frozen tumour, blood, normal kidney) stored centrally on ~two thirds of cases. There is also tumour genomic copy number data available on about one third of the prior trial – SIOP WT 2001 (that ran in the UK, 2002-2011).

Childhood renal tumour dataset

This document describes the research dataset constructed through the Improving Population Outcomes for Renal Tumours of childhood (IMPORT) study from 2012 – 2019 (

The aim of this prospective clinical observational study is to achieve continuous improvement in short and long term outcomes for children with Wilms tumour and other rarer types of childhood renal tumours. The study collects clinical, molecular and imaging data to underpin the introduction of a more personalised approach to risk stratification – the process by which each child is assigned to one of several ‘standard of care’ treatment arms according to the predicted risk of their tumour recurring.

It is a live database, currently comprising data on 670 children with kidney tumours who are resident across the UK and Republic of Ireland. This represents >90% of all incident cases diagnosed in the population. Clinical data and patient samples are collected with explicit consent of their parent/guardian for use in research and for sharing with the relevant national cancer registration service according to where the child is resident at diagnosis.

Data are collected throughout the patient journey and are presented in tables with each data item as a comma separated value (CSV) corresponding to a data field captured through a case report form (CRF). These span the following timepoints from presentation to completion of treatment and then annual follow up or reporting of any events (relapse/death/second tumour).

F1 Registration information F2 Pre-operative chemotherapy (Unilateral tumours only) F3 Surgical form – operative findings F4 Pathology form – institutional pathologist F6 Post-operative chemotherapy F11 Cardiotoxicity reporting form F12 Bilateral Wilms tumour form – additional registration information (12a) and response assessment (12b) F15 Reference Pathology – “Central review” F17 Treatment of relapse or progression F18 Follow up form – for annual follow up reporting and to report any relapse or death. F20A End of treatment summary (Unilateral tumours) F20B End of treatment summary (Bilateral tumours)

Details of the content of each CRF are in the table below.

Most children (~90%) with Wilms tumour survive their cancer. Hence, an important aspect of the dataset is the planned linkage of each child’s data to their national cancer registration record. This will permit measurement of very long term health outcomes in adulthood by linkage to routine health care data utilising the approved information governance systems through which the UK’s national cancer registration services work with the NHS to ensure patient confidentiality.

The IMPORT study is one of five sequential clinical trials and studies that registered the majority of children diagnosed with kidney cancer in the UK and Republic of Ireland between 1980-2019 (over 3,000 patients). It is intended that these additional four closed trial datasets will be made available in the same way during 2020.

These trials’ primary endpoints, of relapse-free and overall survival according to clinically defined risk groups, have been published, with full clinical quality assurance completed. Over that 39 year period, risk stratification and treatment schedules have been refined many times, to reduce exposure to therapies that may cause long term side effects such as heart damage or second tumours. These adverse ‘late effects’ may take decades before they are first seen. Hence, these linked datasets are an important proof of principle about how very long term outcomes can be studied through linkage of clinical trials to routine health care data sources.

The IMPORT dataset will continue to be enlarged and enriched with tumour molecular information (copy number, genomic sequencing) and imaging review data (CT/MRI scans). These data items are not available for the previous closed trials.

Optional material if needed:

Plans for the linked dataset with the national cancer registration and analysis service (NCRAS) of Public Health England.

This project will create a new permanent data linkage between each patient’s trial identifier and data and their existing national cancer registration record, with all personally identifiable data being handled within the existing nationally approved information governance and data protection frameworks of the clinical trial units and the cancer registration service of each nation. The following technical description applies to Public Health England’s National Cancer Registration and Analysis Service (NCRAS). Access to the linked community prescription data is made available through PHE’s Office for Data Release.

NCRAS routinely collects and quality assures clinical information on all cases of cancer that occur in people living in England to construct a cancer registration record. The data collected by NCRAS comes from several sources including Multi-Disciplinary Team meetings and pathology reports. The patient identifiers collected facilitate linkage to other routinely collected datasets, that include Hospital Episode Statistics (HES) with details of all admissions, A&E attendances and outpatient appointments at NHS hospitals in England, mortality data, the Systemic Anti-Cancer Therapy and Radiotherapy datasets from all hospital providers and community dispensed prescriptions data.

Patient-level linkage to inpatient HES is routinely processed by NCRAS. PHE receives pseudonymised dispensed prescriptions data from NHS Business Services Authority, allowing patient-level linkage to the National Cancer Registration Dataset without revealing the identities of those without cancer. Prescription data is available from April 2015 to February 2019, providing a 3-4 year window of prescription drug usage by survivors of childhood renal tumours.

The project’s primary aim is addressed by calculating cumulative and relative risks of hospitalisation for specific organ dysfunctions from the HES-linked data set and relative risks of prescriptions for specific drugs compared to the general population rates. This newly created linked dataset can be refreshed for future follow up studies and linkage to additional datasets that may become available such as patient-reported outcome measures.

For HES-based analyses, follow-up will start at whichever is the later of 1 April 1997 and date of renal tumour diagnosis. For community prescriptions analyses, follow-up will start at whichever is the later of 1 April 2015 and date of renal tumour diagnosis. For both sets of analyses, follow-up will end at whichever is earliest of death, emigration, latest date for which linked death certificate data are complete, and the latest date for which data are complete in the linked databases for HES or community prescriptions as applicable.

Cumulative risk of hospitalisation for specific organ dysfunctions will be estimated from the HES-linked data set with death treated as a competing risk. Relative risk of hospitalisation for specific organ dysfunctions will be estimated by dividing the observed number in the HES-linked data set by the expected number obtained by multiplying the number of person-years accumulated in each age-sex-calendar year stratum in the study cohort by the general population rate in the corresponding stratum calculated from the entire HES database. Relative risks of prescriptions for specific drugs will be estimated in a similar way using the linked community prescription data set and the entire community prescription database.

The IMPORT study is one of five sequential clinical trials and studies that registered the majority of children diagnosed with kidney cancer in the UK and Republic of Ireland between 1980-2019 (over 3,000 patients). It is intended that these additional four closed trial datasets will be made available in the same way during 2020