LATe TreatmENT Related Toxicity in Melanoma (LATENT)

Overview

Recent improvements in advanced melanoma treatment with immunotherapy have dramatically improved patient survival. Longer survival however has come at a cost of toxicity. Short term side effects can occur in \>50% of patients undergoing immunotherapy treatment; however, many long-term survivors are also living with serious consequences of these treatments which may be under reported in literature. Data regarding long term toxicities, from these treatments is lacking and an area of important unmet clinical need. Therefore, in collaboration with the Clatterbridge and Christie's teams, the investigators propose to retrospectively analyse the nature, incidence, frequency, and severity of immune related toxicities in around 400 patients who received immunotherapy for advanced melanoma with ongoing durable responses to treatment of at least 3 years. The investigators will set up a collective anonymized database and record this information through review of electronic medical records of patients that meet the eligibility criteria. The investigators will also review the patterns of use of long-term immunosuppression and assess the need for specialist referrals for managing late side effects. The investigators hope that this data will help us address gaps in the management of long-term survivors by identifying areas of need and establishing a coordinated evidence based multidisciplinary service to provide personalised, risk stratified long term follow up.

LATENT will be a retrospective non-interventional analysis of pre-existing data from patient medical records and, therefore patients will not be required to participate in any risky procedures, treatments or hospital visits. The study will therefore not require explicit informed consent from eligible participants. A potential ethical issue could arise around explicit consent of patients for collection and publication of their data. The investigators aim to circumvent this by only using data that has already been recorded from direct patient care. The investigators will pseudo-anonymise personal data and mitigate risk of identification through: 1. Direct health care providers screening for eligible patients from clinic records based on clear inclusion and exclusion criteria 2. Allocation of de-identified serial numbers for patients on the database used to collect and record relevant data 3. Exportation and storage of de-identified data from all sites on a common trusted research environment (TRE) 'BRIDGE' for blinded analysis by the Research team 4. Reporting of anonymised/de-identified data only, for publication In addition, the investigators aim to reduce selection bias by eliminating the need for explicit consent as unwell patients with greater clinical needs may be unable to consent and would not be included in the study, thereby only selecting for well patients and potentially underrepresenting a vital group of patients, compromising the scientific validity of the study. As this is a multicentre study, de-identified, anonymised data from all centres will be exported and stored in a single secure password protected TRE for analysis. The main centre in charge of maintaining and analysing the database, with appropriate data sharing agreements with individual sites, will be The Royal Marsden team. The investigators do not anticipate any legal issues arising from this study.