MR1251 - Safety and appropriateness of growth hormone treatments in Europe (SAGHE)

This Data Sharing Agreement permits the retention of the data provided under previous iterations of this Agreement for an interim period. This is a pragmatic approach to provide an active Agreement whilst enabling The Institute of Cancer Research to complete the necessary actions to enable a transition of the study data to the University of Manchester.

The following provides background information on the purpose of the original study:

Growth hormone treatment has been used since 1957 to treat growth hormone deficiency, and thereafter increasingly for short stature due to other causes. Initially hormone extracted from human pituitaries (p-hGH) was used, and then since the mid 1980s synthetic recombinant human GH (r-hGH). Concerns about long-term safety were initially caused by an outbreak of Creutzfeldt-Jakob disease consequent on prion infection of pooled pituitaries used to produce human GH, whose use was therefore discontinued in 1985. Subsequently concerns were raised by reports of apparent excesses of leukaemia and other cancers, and more recently of mortality in France although not in three other countries. There have been a number of studies investigating cancer and mortality risks in such patients. Raised risks of death and certain cancers have been reported inconsistently, but these findings have limitations and leave considerable uncertainty.

The SAGhE (Safety and Appropriateness of Growth Hormone Treatments in Europe) study was therefore initiated, funded by the EU, to provide a large-scale international collaborative cohort study of r-hGH treated patients with long-term follow-up for cancer incidence and mortality conducted independently of pharmaceutical companies. It is the largest and longest follow-up cohort study of growth hormone-treated patients with follow-up and analysis independent of industry and has formed a major international resource for investigating cancer and mortality risks in r-hGH patients.

The study is of cohort design, conducted in 8 European countries, with the design and conduct coordinated from the outset and analyses centralised at ICR, (hence all data for the cohort analyses are sent to ICR). With appropriate ethics and privacy committee approvals, the study identified in each country all resident patients who were born before 1991-1995 (the exact year depending on the country), who had been treated with r-hGH at paediatric endocrine clinics at any time since first use of the treatment in 1984 up to a date during 2007-9, depending on the country (or in France and Sweden up to 1997), and who had never been treated with human p-hGH and followed them for cause-specific mortality and cancer incidence.

This SAGhE study is a large collaborative project involving several UK and international organisations including:

• University College London

• University of Manchester

• Institute of Cancer Research

• Assistance Publique- Hôpitaux de Paris (France)

• Cliniques Universitaires St Luc (Belgium)

• Universität Leipzig (Germany)

• Dutch Growth Foundation (Netherlands)

• University of Rome (Italy)

• Karolinska Institutet (Sweden)

• University of Bern, Inselspital (Switzerland)

No patient-level data has been or will be shared with any of these other organisations of the SAGhE group under this Data Sharing Agreement.

The cohort consists of 24,232 patients, most commonly treated for isolated growth failure (53%), Turner syndrome (13%) and growth hormone deficiency linked to neoplasia (12%). Data were extracted from existing databases and from case-notes on demographic variables, parental heights, birth characteristics, results of growth hormone testing and additional endocrine deficiencies, and treatment of those deficiencies, at regular intervals height, weight, pubertal status, bone age, growth hormone dose and number of weekly injections and associated treatments that might interfere with growth (sex steroids, GnRH agonists, steroids), the use of cranial or total body irradiation, the diagnosis for which r-hGH was prescribed.

Follow-up of cohort members for incident cancers, death, emigrations, and other losses to follow-up was conducted by various methods depending on the country. In the UK it was via NHS Digital. Reported cancer diagnoses were validated by cancer registry data or from pathology reports. National cause-specific mortality data and population counts for the general population, to derive ‘expectations’ for mortality rates in the cohort, were obtained from death certificate-based statistics from national statistics offices. National site-specific cancer incidence data, likewise, were obtained from national cancer registries (the Netherlands, Sweden, UK, and Belgium from 2004 onwards) or where national cancer registration did not exist, from national estimates based on regional registry data (Belgium before 2004, France, Germany, and Switzerland).

The study’s Principal Investigator is an individual employed by University College London Hospitals NHS Foundation Trust (UCLH) who has a honorary contract with University College London (UCL). However, under this Agreement no individual at UCLH or UCL is permitted to exercise any influence over the purposes for which the data may be processed.

The purposes for which the data may be processed are as follows:

Confidential identifying details may be securely retained and not otherwise processed in anyway.

Data related to cancer incidence may be securely retained and not otherwise processed in anyway.

Pseudonymised record level patient data relating to mortality from which the identifying details have been separated may be securely stored and processed as required to support:

i) Finalising a paper about mortality across the 8 countries, including conducting any further analyses that may be requested by the journal ahead of its publication;

ii) Responding to questions, challenges, etc. to the content of the paper post-publication.

No other processing of the data are permitted.

Under no circumstances may data be shared with or otherwise accessed by individuals employed by UCLH or UCL.

In accordance with the above stipulations, data may only be processed by individuals employed by the Institute of Cancer Research (ICR).

The ICR will be responsible for ensuring compliance with the above restrictions on processing and, as such, is the data controller for the purpose of this Agreement.

Processing the data as described will enable completion of publication of the results of this 8 country study. 3 papers are already published and the fourth is submitted.

The ICR has identified GDPR Articles 6(1)(e) (“a public task”) and 9(2)(j) (“research”) as the lawful basis for processing personal data for the purpose of this study. The ICR has determined that the processing is in the public interest for the safety and reassurance of patients treated with Growth hormone (GH) and their parents, relatives, and for the public to be reassured on whether children so-treated are being treated safely.

Under this Agreement, the data may be securely stored but not otherwise processed. No new data will be provided by NHS Digital under this Agreement.

The study data, including data provided by NHS Digital under previous agreements, are currently held by The Institute of Cancer Research.

The following provides background on the processing activities undertaken prior to this Agreement:

Identifiable data was shared with Health and Social Care Information Centre (HSCIC) to carry out the linkage between the study data and civil registration data. Participants records were ‘flagged’ with HSCIC. HSCIC notified the study team at the Institute of Cancer Research of participants’ deaths (date and cause) and cancer events when they occurred. Data was last supplied in March 2016.

There will be no new data flows under this Agreement. The data will be stored at the Institute of Cancer Research (ICR) and will only be accessed by individuals substantively employed by the ICR.

The identifying details and records relating to cancer incidence will be securely stored and not otherwise processed.

Pseudonymised records relating to mortality will be securely stored and only processed if new analyses are requested by the journal to which the final paper was submitted. In this case, the analyses will be conducted solely at ICR with no data flows or linkages and solely using data already held by the ICR.

Data processing will only be carried out by substantive employees of the ICR who have been appropriately trained in data protection and confidentiality.

This Agreement permits the secure retention of the data only and no other processing.

No new outputs will be produced under this Data Sharing Agreement.

Under the previous iteration of this agreement, 3 papers have already been published (Swerdlow, Cooke et al. 2015, Swerdlow, Cooke et al. 2017, Swerdlow, Cooke et al. 2019) and a fourth was submitted to The New England Journal of Medicine (NEJM) for review and acceptance. The NEJM declined to publish the paper.

The paper is now in the process of being reformatted for the Lancet with a view to send to the collaborators for their agreement in the coming weeks. Submission to the journal in planned by the end of January 2020, and if the journal are interested there may be one or more rounds of iteration with them over a 2-5 month period. If, however, they reject it at the first submission the ICR will submit to another journal, probably The Lancet Diabetes & Endocrinology, within a month of hearing back from them. If again rejected, new submission will probably be to the Journal of Clinical Endocrinology and Metabolism (JCEM), within a further month.

Subsequent to journal acceptance, the ICR will need to await publication, at a pace chosen by the journal, and then the data will be needed as noted in 5a (ii) to respond to any questions, challenges, etc. to the content of the paper post-publication. This will take several months, since publication does not happen immediately after acceptance, and responses from readers usually come 1-2 months after publication. In total, therefore, the ICR would estimate that 10 months is needed, in case submission has to be to more than one journal, followed by eventual publication and reader responses.

If necessary, further processing of the data may result in answering any questions raised by reviewers and journal editors in the review process, in order to obtain publication.

There may be subsequent presentations to meetings or conferences following publication of the data.

The published findings and any subsequent presentations will contain only data which is aggregated with any small numbers suppressed in compliance with NHS Digital’s published guidance and will contain no individual’s data.

This Agreement permits the secure retention of the data only and no other processing.

The benefits to patients, clinicians and the public are that the published findings contribute to the safety and reassurance of patients treated with Growth hormone (GH) and their parents, relatives, and for the public to be reassured on whether children so-treated are being treated safely.

The papers published to date have raised awareness of the following findings:

- Having examined the cancer risks in relation to growth hormone (GH) treatment, the results did not generally support a carcinogenic effect of r-hGH but it was highlighted that the unexplained trend in cancer mortality risk in relation to GH dose in patients with previous cancer, and the indication of possible effects on bone cancer, bladder cancer, and HL risks, need further investigation.

- Having examined meningioma risks in relation to GH treatment, the findings added to evidence of very high risk of meningioma in patients treated in childhood with GH after cranial radiotherapy but suggested that GH may not affect radiotherapy-related risk and that there is no material raised risk of meningioma in GH-treated patients who did not receive radiotherapy.

Not stated in the register.