Inclusion criteria 1.

Male or female patient ≥ 18 years of age.

2.

Confirmed diagnosis of malignant melanoma, NSCLC or mesothelioma.

3.

Shared decision by oncologist and patient to proceed with CPI treatment, either with the combination of ipilimumab and nivolumab, or with single-agent nivolumab, pembrolizumab or atezolizumab as standard of care.

4.

Patient is judged as being capable of understanding the information sheet and of giving informed consent according to the Mental Capacity Act 2005.

5.

Written informed consent to participate in the study.

Exclusion criteria 1.

Known pre-existing autoimmune or immune-mediated inflammatory disease requiring immunomodulatory treatment, including (but not limited to) inflammatory bowel disease (Crohn’s disease, ulcerative colitis) autoimmune endocrinopathy or hepatitis, vitiligo and inflammatory arthritis.

2.

Received enteral or parenteral steroids within past month (topical, inhaled or intranasal permitted).

3.

Previous treatment with CPI therapy.

4.

Vaccination within the past 4 weeks, except COVID-19 vaccination permitted.

5.

Known chronic infection.

6.

Current pregnancy, or pregnancy planned within next 6 months.

7.

Inability to provide informed consent and/or undergo any of the procedures mandated by the study.

Screening visit

Before undergoing any study related procedure (including screening procedures), all potential participants will provide full, written informed consent. Participant demographics, full medical history including cancer history, general physical examination, and patient symptom questionnaires are completed at the screening visit. Standard of care blood tests, research blood tests and six skin swabs (Table 2) are taken, and a pregnancy test is taken for women of childbearing potential. Sites undergoing skin swab were: forehead, upper chest, upper back, dorsum of the hand and forearm; vitiligo-like depigmenting rashes were also swabbed if present. The full schedule of events is shown below in Table 2. Participants at consent can decide to contribute optional additional samples including: regular stool samples for microbiome/mycobiome analysis, accompanied by optional completion of a nutritional questionnaire, and optional gastrointestinal or dermatological tissue if they develop a SirAE affecting these sites that trigger the need for biopsy as part of routine care.

Table 2 Schedule of eventsFollow-up: post-cycle 1,2,3,4, and 9 months

Patients undergo formal study clinical review post-cycles 1, 2, 3, and 4 months of CPI treatment and at 9–10 months depending on the CPI regimen as detailed in Table 2 to coincide with routine care visits to hospital. If patients change CPI therapy prior to completing the scheduled 6 study visits, then they can continue within the study from cycle 1 of their new treatment until they have completed 6 scheduled study visits in total. Adverse events and concurrent medications are recorded at each visit in addition to clinical review. Clinical review will include collation of a detailed symptom directed questionnaire from patients, general physical examination, collection of skin swabs and routine blood tests. Participants who have consented to optional stool samples will also provide these following each visit, they will furthermore be asked to submit a completed dietary questionnaire.

Adverse events and principle clinical end point

The reporting of adverse events will be undertaken at each visit, with these graded as mild, moderate, or severe, and allocation of a numerical grade according to the CTCAE version 52. AEs are classified in terms of their relatedness to CPI therapy as follows: unrelated, unlikely, possible, probable, definitely and not assessable.

Significant IrAEs for purposes of the MEDALLION study (SirAEs) are defined according to strict criteria and recorded at scheduled/unscheduled visits, being those considered “definitely” or “probably” related to administration of the CPI therapy, and meeting organ system-specific criteria as outlined in Table 1. SirAEs constitute the principle clinical endpoint of the study.

Where a definite IrAE is determined to have occurred by the investigator but the above SirAE criteria are not fulfilled this will be recorded as a “non-significant IrAE” and the patient will remain in follow-up.

IrAE ‘ad hoc’ visit

Participants can directly request an unscheduled visit if they suspect they are developing an IrAE, or they may be identified at routine clinical review. Research specific biological sampling will take place at the time of incident IrAE.

Withdrawal criteria

Participants can decide to withdraw their consent at any point during the study. The clinical investigator may withdraw a participant from the trial at any time if this is considered necessary, and for any reason including:

i.

Symptomatic deterioration.

ii.

Participant withdrawal of consent or inability (through incapacity or otherwise) to provide consent for study-specific procedures to proceed.

iii.

Significant protocol deviation or non-compliance, including failure to attend for > 2 consecutive visits.

iv.

An adverse event such that continuation of CPI therapy is no longer appropriate, even if SirAE criteria are not fulfilled.

v.

Termination of the clinical trial by the sponsor.

vi.

Investigator’s discretion that it is in the best interest of the participant to withdraw.

vii.

The patient has fulfilled the SirAE criteria and completed their IrAE ad hoc visit.

Statistical considerations

The primary objective of the study is to generate pilot data from a substantive cohort to support hypotheses that may be tested and/or validated in future investigations. Statistical considerations are applied to address secondary (hypothesis testing) and exploratory (hypothesis generating) objectives.

Sample size justification

Observations made in a separate study, BIOFLARE [26], on pSTAT3 expression in a subset of circulating CD4 + T cells of RA patients in remission who develop disease flare following treatment cessation versus those who do not, were used as a basis for sample size calculation. Assuming CPI recipients with and without SirAEs will display a difference of similar magnitude in CD4 + T cell pSTAT3, enrolment of 66 participants with complete datasets will afford MEDALLION 80% power to detect a difference of ~ 24% between groups at alpha 0.05 (pilot data used currently in preparation for publication; personal communication, JDI and F Rayner). However, the longitudinal nature of our study is an important aspect, as it will allow taking account of the exact time the IrAEs occur. There are very few relevant similar datasets but one recent report found in a Swiss population three immune-related predictors of IrAEs in melanoma patients treated with CPIs: CXCL10, IL-10 and regulatory T cell (Treg) levels with hazards ratios of 12.6, 4.0 and 3.4, respecitvely [27]. Detecting effect sizes broadly similar to these with power 80% at significance level 0.05 requires a minimum of 21 events. Based on the anticipated 32% participants experiencing SirAEs, recruitment and following up of the aforementioned 66 patients will be sufficient for such analyses.

Analysis plan

A broad range of measurable immune parameters will be evaluated in MEDALLION, in addition to the microbiome, but the aim of the main statistical analysis (secondary study objective) is to addresses the hypothesis that pSTAT3 expression in circulating CD4 + T cells pSTAT3 predicts SirAE development. The study design ensures that, in addition to testing a specific hypothesis, our unique cohort will form a substrate for a wide range of exploratory analyses in relation to cellular immunity and the microbiome during the development of SirAEs. Descriptive statistics will be used to describe recruitment rates, reasons for refusal to participate, refusal rates for optional procedures and missing data. They will also be applied to compare baseline circulating CD4 + T cell pSTAT3 between patients developing a significant SirAE versus those that do not (e.g. Mann-Whitney U test), with similar comparisons made up to the time of incident SirAE or matched time point. Exploratory analyses in relation to other biological parameters will then be undertaken in a similar manner. Markers’ association with time-to-SirAE will furthermore be evaluated using Cox proportional hazards regression. Significantly predictive markers will be graphically displayed by Kaplan-Meier curves after dichotomisation. Mixed effect logistic regression, a form of general linear mixed model (GLMM), will be applied to repeated measurements to test association between SirAE occurrence and marker longitudinal trends. Statistically significant trends will be depicted graphically. GLMM framework is flexible and will allow to exploratively adjust key clinical covariates although their number will be limited by available sample size.

Data handling

The number of patients approached, interested in taking part and screened will be collected via a log completed by staff conducting screening. Data for an individual patient will be collected by the study PI or their delegated person and recorded in the secure, password-protected electronic case report form (eCRF) for the trial. Patient identification on the eCRF will be through a unique trial identifier number. A record linking the patient’s name to the unique identifier number will be held securely at the trial site, and is the responsibility of the PI. As such, patients cannot be identified from eCRFs.

The PI or delegated person will monitor completeness and quality of data recording in eCRFs and will correspond regularly with investigators to avoid missing data where possible and ensuring continuous high quality of data. Patients will complete the paper assessment tools as required. The tools will also only be identified using the unique patient identifier number. Overall responsibility for data collection lies with the PI. Data collected on paper assessment tools will be entered onto a secure validated clinical data management system. A study identifier number will be used to identify participants on all paper data collection forms throughout the duration of the trial. Data will be handled, computerised, and stored in accordance with the Data Protection Act 2018. No participant identifiable data will leave the study site. The quality and retention of study data will be the responsibility of the PI. All study data will be retained in accordance with the latest Directive on Good Clinical Practice (2005/28/EC) and local policy.

Staff involved in the conduct of the trial, including the PI and study staff involved in screening and intervention will have access to the site files for patients at the hospital study site. Clinical information shall not be released without the written permission of the participant, except as necessary for monitoring and auditing by the Sponsor or regulatory authorities. Secure pseudonymised electronic data may however be released to named members of the study team for analysis purposes. The PI and trial site staff involved with this trial may not disclose or use for any purpose other than performance of the trial, any data, record, or other unpublished, confidential information disclosed to those individuals for the purpose of the trial.

All trial data will be stored securely in accordance with Good Clinical Practice (GCP) and the Sponsor guidelines (Newcastle JRO standard operating procedures; SOPs). Any personal identifiable information will be stored at the study site or Newcastle upon Tyne Hospital Foundation Trust (NuTH) archiving facilities, for up to 5 years before secure disposal.