Assessment of Short Immunotherapy After Radical Surgery of High-risk Malignant Melanoma

A prospective randomized international multicenter study to compare Short versus Long Adjuvant immunotherapy after radical surgery of stage IIb-c, III and IV cutaneous malignant Melanoma, Acronym: Grand SLAM (Study Long Adjuvant Melanoma) Purpose and aims Between 2013 and 2024 the incidence of cutaneous malignant melanoma (CMM) increased from 3400 to over 5800 cases/year in Sweden (Swedish Melanoma Registry, SweMR) and the number of cases is expected to multiply during the next decades (www.cancerfonden.se).

Since 2017, post-operative treatment with programmed death 1 (PD-1) inhibitors for 12 months is routine in the Western world for stage III melanoma patients. However, a rational for having chosen a 12-month treatment period in the registration studies is lacking.

As in other malignancies, adjuvant treatment for CMM is most often given in vain as many patients would not relapse even without adjuvant treatment and a large group of patients relapse despite having received adjuvant treatment. Noteworthy, no randomized study has investigated whether a shorter adjuvant treatment is as effective as 12 months and there is no ongoing study on this subject (www.clinicaltrials.gov).

One major disadvantage with a long adjuvant treatment is the increased risk for severe toxicity. In the adjuvant studies the observed grade 3-4 toxicity is 10-15 % (1, 2). Another drawback with adjuvant treatment is the resources required, including costs. The cost for checkpoint inhibitors has been estimated to raise from astonishing 24 billion USD in 2021 to 46 billion USD in 2026 (https://www.researchandmarkets.com/report/checkpoint-inhibitors). A large unknown part of this sum could be referred to adjuvant treatments.

In conclusion, since adjuvant studies on malignant melanoma patients usually have been conducted with pharmaceutical companies as sponsors, the important question whether a shorter adjuvant treatment period is as effective as the current one-year schedule has so far not been addressed. A shorter treatment period would clearly be advantageous for patients and lead to a substantial reduction in drug costs and health care resources.

The aim is to conduct a prospective randomized international multicenter non-inferiority study with systemic immunotherapy comparing treatment for 6 months (experimental arm) versus treatment for 12 months (standard arm) in patients having undergone radical surgery for stage IIb-c, III and IV CMM. The main research question is whether 6 months of immunotherapy in this patient group with high risk of relapse is as effective as the current 12-month scheme. The primary clinically relevant outcome variables are distant metastatic free survival (DMFS) and relapse free survival (RFS). Substudies, investigating the risk with food supplements (Melanoma Kost, MelKo) and biomarkers, are also planned.

Survey of the field The introduction of adjuvant immunotherapy was based on two phase III studies, one for nivolumab (Opdivo®) and the other for pembrolizumab (Keytruda®). Two years ago, results were presented from another randomized phase III study assessing immunotherapy in patients operated for thick CMM without lymph node involvement (stage IIb-c). Treatment with PD-1 inhibitor significantly prolonged RFS compared to placebo. In addition, results from a randomized phase II study in patients with macroscopic stage III and operable stage IV CMM were published last year showing that the event-free survival was significantly better for patients receiving neoadjuvant PD-1 inhibitor compared to standard adjuvant treatment. The ongoing phase III NADINA study aims to confirm the increased benefit of neoadjuvant over adjuvant immunotherapy (NCT04949113).

There are ongoing studies assessing new drugs in the adjuvant setting for high risk CMM patients. Two phase III studies investigate the potential benefit of adding a LAG 3 inhibitor to anti PD-1 treatment: In RELATIVITY-098, relatlimab is given together with nivolumab (Opdualag®) (NCT05002569) while fianlimab is tested together with cemiplimab (Libtayo®) in the other study (NCT05608291). Furthermore, whether pembrolizumab in combination with the TIGIT-inhibitor, vibostolimab is more effective than pembrolizumab alone is addressed in another large study (NCT05665595). Yet another potential registration study is KEYNOTE-942 where patients who have undergone surgery for different stages of high risk CMM, receive pembrolizumab combined with injections with a neoantigen mRNA-based vaccine which has shown promising results in a phase 2 study (NCT03897881).

Study design Research questions PICO: High risk CMM/Treatment at standard doses with checkpoint inhibitor for 6 months/Standard treatment with checkpoint inhibitor for 12 months/DMFS and RFS.

Variables and measures DMFS and RFS at 2 years measured in months from start of treatment.

The patient must be randomized and start systemic treatment within 12 weeks after final surgery, i.e. excision + sentinel node biopsy, lymph-node dissection or metastasectomy.

Study procedures Randomization procedure Randomization is performed after stratification according to tumor stage. Study treatment Adjuvant treatment with current immunotherapy standard drugs (at present nivolumab or pembrolizumab) will be given in both study groups. The standard drugs might change during the study, based on upcoming results from ongoing adjuvant studies (see above). Patients who have undergone neoadjuvant immunotherapy are also eligible for the study providing that a complete pathological response has not been achieved which usually is the case in 20% of the cases (4). Since the neoadjuvant treatment duration is two months, these patients will in the standard arm receive 10 months post-operative treatment and in the experimental arm, four months only. Follow-up visits, blood tests and management of side effects will follow current practice based on national guidelines.

Study Follow-up Scheme in brief A baseline visit 1-11 weeks after the final surgery is scheduled for study information, inclusion and randomization. The national routine schedule for follow-up visits will be applied for all patients in the study according to standard practice which normally includes doctor appointments with physical examinations after each scan. These visits are important for ethical reasons; the results of the scans should always be delivered to the patient in person.

The imaging scheme follows national guidelines but computor tomography (CT) thorax, abdomen and brain, or i.v. contrast enhanced whole body FluoroDeoxyGlucose-Position Emission Tomography (FDG-PET-CT) (including brain) at baseline, and at 6 and 36 months is mandatory (see flow-chart below). The scan at 36 months is not included in Swedish practice but the minimum scheme is clearly less extensive than currently used in most other countries and in previous adjuvant studies (1, 2).

Extra diagnostic evaluation (imaging and laboratory tests) will be allowed for all subjects presenting signs and/or symptoms. Patients in both groups are instructed to contact their study center if they experience any symptoms suspicious of recurrence.

Estimated sample size and power The study design is a non-inferiority trial investigating whether short-term adjuvant immunotherapy, is non-inferior to long-term adjuvant immunotherapy (standard of care) with a non-inferiority margin of 4.0% being clinically appropriate in this study population.

Assumptions for the sample size calculation:

• DMFS at 2 years was 88.1% for stage IIb-c patients (Keynote 716 study) (1)
• DMFS at 2 years was 70% for stage III-IV patients (Checkmate 238 trial) (2)
• DMFS for the sample size will be based on these two measurements with a conservative estimate of 75.0% at 2 years.
• Power = 80% ● Alpha level = 5% ● Non-inferiority Margin = 4%
• Accrual time = 48 months ● Follow up time = 60 months A sample size of 1792 patients (896 in each study arm) will have 80% power with one-sided 95% confidence to declare non-inferiority with a margin of 4.0% based on a 1:1 randomization with a corresponding HR of 1.19. If emerging data demonstrate a higher 3-year DMFS proportion than 75%, the sample size may be either re-estimated or the non-inferiority boundary adjusted or if the number of patients lost to follow-up is of concern.

Statistical methods Anders Berglund, PhD is responsible statistician and the efficacy analyses will include all randomized patients. Kaplan-Meier techniques will be used to plot both outcomes. Comparison of outcomes between the study arms will be based on a Cox regression model. The hazard ratio (HR) associated with the study arms will be derived along with the associated 95% confidence interval (CI). The p value for testing the null hypothesis that the HR between the interventions will be greater than or equal to 1.19 will be derived from this model by comparing the log-likelihood of the fitted model with the log-likelihood of a model where the HR between the groups is set to 1.19 by use of a likelihood ratio test.

Co-primary endpoints

DMFS and RFS at 2 years are co-primary endpoints and to preserve the overall type I error rate, comparisons will be performed according to hierarchical testing:

1. The alternative non-inferiority hypothesis, H11: DMFS short-term treatment is not inferior to long-term treatment (standard of care) for the first co-primary endpoint (DMFS). If H11 is accepted., i.e., non-inferiority is declared in (1), then the following hypothesis will be tested.
2. The alternative non-inferiority hypothesis, H12: DMFS short-term treatment is not inferior to long-term treatment (standard of care) for the second co-primary endpoint (RFS).

5% level of significance will be applied, with comparisons in the sequential testing procedure being conditional on rejection of the null hypothesis of the previous comparison.

Interim analysis The interim analysis will be conducted when two-thirds of the total number of planned subjects are enrolled which is the optimal timing when using the O'Brien-Fleming Boundary (5).

Organisation The PIs in Sweden have been involved in the TRIM study and the PIs in the other countries are also experienced and active researchers at university centers.

The statistician has been responsible for the TRIM study. The study coordinator has been involved in several clinical studies over the years. The monitors from the research unit at Uppsala University Hospital are very experienced.

All principal PIs will be GCP trained. There is support for agreements at the research unit at Uppsala University Hospital and that unit is highly competent in GDPR.

Study subject information will be handled strictly confidentially and only by authorized personnel. When data is processed, name and personal number will be replaced by a code.

Patients are recruited from centers all over Sweden and university clinics in other European countries. Based on data from the SweMR, the expected incidence of stage IIIb-d according to AJCC8 classification at the time of diagnosis, is 250 patients/year. With a recruitment rate of 60%, 150 patients are estimated to be enrolled/year in Sweden only.

In addition, CMM patients diagnosed with stage III at relapse and patients undergoing radical surgery for stage IV CMM are also eligible. We estimate that 50 patients from these subgroups will be included/year, yielding a total recruitment of 200 patients/year.

Time plan

• 2023: Planning, fund-raising, writing the final protocol.
• Spring 2024: The eCRF and randomization procedure were created. Ethical, radiation protection committees and the Swedish Medical Products Agency approvals were gained.
• 2023-2024: Recruitment of centers.
• 2025-2028: Inclusion of patients, around 200/year (Sweden), 450/year in total.
• Interim analysis in 2027
• 2033: Last patient followed-up for 5 years. Risk mitigation As the study is a non-inferiority study it will require a large number of participants. The Swedish Melanoma Study Group, the Nordic Melanoma Group and the European Melanoma Group are in favor of the study. We will take advantage of the organization for the ongoing TRIM study which has recruited around 1300 patients despite that only Swedish centers participate. More university centers will likely join when the study is up and running.

Equipment Need for research infrastructure No equipment is needed for this project. Not applicable for this study.