Co-stimulatory Markers and Vitamin D Status in Anti-PD1 Treated Melanoma Patients

Background Advanced melanoma still is associated with severe disease and early death. New immunotherapies including the checkpoint inhibitors anti-programmed death 1(PD-1) and anti-cytotoxic T-lymphocyte-associated protein 4 (CTLA4) improve survival compared to chemotherapy. It still remains to be elucidated why the treatment only benefits some patients. The tumour cells expression of programmed death ligand 1 (PD-L1) and other co-stimulatory receptors, as well as the presence of soluble isoforms are speculated to influence treatment success. Possible modulators of the PD-1 pathway like vitamin D might also influence the treatment outcome. Investigating these relations could provide useful markers to predict response to anti-PD-1 therapy in advanced melanoma patients.

PD-1, a transmembranous molecule present on activated T, B and natural killer (NK) cells is essential in maintaining self tolerance. PD-1 has two widely expressed ligands, PD-L1 and programmed death ligand 2 (PD-L2). Both PD-1 and its ligands are present in soluble (s) forms. In autoimmune disease like rheumatoid arthritis sPD-1 concentrations are associated with disease markers and scores.

PD-L1 can be highly expressed by cancers and thereby silence the T cell immune attack. Using anti-PD-1 antibodies, the immune cells are not inhibited by the PD-1 pathway, and proliferation, activation and T cell survival increases, improving tumour cells targeting. Soluble PD-1, sPD-L1 and sPD-L2 may be potential markers in melanoma patients who would benefit from the anti- PD1 treatment.

In autoimmune diseases as rheumatoid arthritis, Crohn's disease and in multiple sclerosis vitamin D deficiency is associated with increased disease activity. Active 1.25-dihydroxyvitamin D3 (1.25-vitD) binds to the vitamin D receptor (VDR) and the 1.25-vitD-VDR complex functions as a transcription factor. Regarding melanoma, VDR variants are associated with melanoma risk. In mice, vitamin D stimulation has been shown to increase the PD-L2 expression in dendritic cells (DCs). In humans vitamin D stimulation increased the DC PD-L1 expression and interleukin 10 (IL-10) production and PD-L1 blockage resulted in increased interferon gamma and decreased IL-10 production.

Our preliminary pilot studies have demonstrated that vitamin D initially decreases PD-1 expression in Crohn's disease T cells but after 26 weeks of vitamin D treatment PD-1 expression increases in CD4+CD25+ T cells (unpublished data). These results might indicate that vitamin D treatment initially increase T cell activation replaced by a decreased T cell activation after long-time vitamin D treatment in inflamed patients. This might influence the anti-PD-1 treatment response in advanced melanoma. Improved knowledge of the interaction between anti-PD-1treatment, vitamin-D and sPD-1, PD-L1 and PD-L2 could provide new beneficial modifications to the treatment regimens currently available -for instance as a new predictor of treatment response.

Methods and participants Advanced melanoma patients 40 patients diagnosed with advanced melanoma will be included in the present observational study designed to identify response-related biomarkers of anti-PD-1 therapy. Patients are included at the Department of Oncology, Aarhus University Hospital (AUH). The protocol is approved by the Central Denmark Regional Committee for Health Research Ethics and the Danish Data Protection Agency. The protocol comprises blood samples at baseline, 3 and 6 weeks after treatment initiation with anti-PD1 therapy. The study period is 6 weeks followed by 3 years of follow-up. Medical history and symptoms and serious immune side affects will be recorded at each visit in both the study period and in follow-up. Standard biochemical markers will be obtained according to guidelines during follow-up and will be independent of the project. Patients are follow-up every third week during the anti-PD-1 treatment. The treatment is stopped because of disease progression or server adverse events. After ended anti-PD-1 treatment patients are followed-up every 3rd month. The follow-up visits are used to detect time to treatment response and time to treatment failure.

The included persons do not receive economic compensation for participating in the project. Aim one to three will be investigated when all patients have ended the study period of 6 weeks. Aim five and six will be investigated when all patients have ended the follow-up period

Healthy volunteers 20 Healthy volunteers (HV) are included as controls, matched by age and gender. Matched HV are obtained from buffy coats, an established collaboration with the Blood Bank, Aarhus University Hospital. It is not possible to trace the identity of a donor to a specific blood sample.

Peripheral Blood Mononuclear Cells From advanced melanoma patients 36 ml of blood are collected at baseline and after 6 weeks of therapy (a total of 72 ml). From HV, blood from the buffy coat is used. Peripheral blood mononuclear cells (PBMCs) are isolated with ficoll, an established protocol. PBMCs are cryopreserved at -140C and stored in the INART bio bank of Biomedicine, Aarhus University. The PBMCs are used to investigate expression of co-stimulatory molecules; among these PD-1, PD-L1, PD-L2, CTLA-4, activation markers, cell proliferation, cell viability, cytokine production as well as vitamin D related cell markers together with mRNA expressions of vitamin D and PD-1 related genes.

Plasma and Serum At baseline and after 3 and 6 weeks of therapy 18 ml of blood are obtained to isolate serum and plasma (a total of 54 ml blood). Plasma and serum from HV are obtained from the buffy coat. Plasma and serum is isolated, cryopreserved at -80C and stored in the INART bio bank of Biomedicine, Aarhus University. In plasma, sPD-1, sPD-L1, sPD-L2 and CTLA-4 concentrations with more are determined.

Study-related biochemical markers At baseline and after 6 weeks of initial treatment a total amount of 55 ml of blood are used for the following blood samples: 25-hydroxyvitamin D2+D3, 1.25-dihydroxyvitamin D3, Parathyroid hormone, free calcium-ion and phosphate. Vitamin D binding protein concentration is measured at baseline. All the listed blood samples despite vitamin D binding protein concentration are measured as a standard operation procedure by the Clinical Biochemistry department at Aarhus University Hospital. Vitamin D binding protein concentrations are measured at Statens Serum Institut a public enterprise under the Danish Ministry of Health.

Melanoma cell line The patient data will be supported by in vitro experiments using a melanoma cell line. PD-1 and vitamin D related surface markers, cytokine production and mRNA expressions will be associated to anti-PD-1 therapy, sPD-1, sPD-L1, sPD-L1 and vitamin D.

Statistics ELISA-made results (sPD-1, sPD-L1, sPD-L2, sCLTA-4) are presented as median with total range and differences are calculated using Wilcoxon non-parametric statistics. Other results are examined for normal distribution. Normal distributed data are compared with student's t-test and non-normal distributed data are compared with Wilcoxon non-parametric statistics. Association between two variables will be estimated using Spearman's rank correlation coefficient (Spearman's rho). A p-value below 0.05 will be considered statistically significant.