Melanoma is rapidly increasing in incidence throughout the world. Early stages are curable with surgical approaches with excellent prognosis. However, a substantial proportion of patients progress to metastatic disease with survival rates of less than 5% making melanoma the culprit for over 65% of all skin-cancer related deaths. Novel agents targeting the immune system and the signaling pathways of melanoma are generating new promise, but chemotherapy remains an important therapeutic alternative, despite low response rates. The resistance of melanoma to chemotherapy is in part due to DNA repair mechanisms that allow cells to survive alkylation damage. Several novel agents targeting the abrogation of DNA repair pathways alone and in combination with cytotoxic agents have been developed with varying measures of success. In this dissertation, we first identified the epigenetic silencing of the DNA mismatch repair (MMR) gene MLH1 as a determinant of response and survival for melanoma patients treated with alkylator-based chemotherapy (dacarbazine/ temozolomide). We then determined the safe dosage of the epigenetic agent decitabine that can be administered in combination with temozolomide. The safety, tolerability and efficacy of the combination of decitabine and temozolomide were evaluated in a Phase II population. We finally determined the pharmacokinetic and pharmacodynamic effects of treatment with the combination of decitabine and temozolomide in the blood and tumor tissues of metastatic melanoma patients.

Immunotherapy utilizing cytokines or immune regulatory check point blockade has consistently demonstrated superior clinical efficacy in melanoma when compared to tumor peptide immunization strategies reported to date. In this project, I conducted 2 model studies representing alternative immunotherapeutic approaches (non-antigen specific combination of interferon-á2b and an anti-CTLA4 monoclonal antibody, IFN-Treme compared to a tumor antigen specific multi-epitope vaccine given in adjuvant with the potent combination of a TLR-9 agonist and GM-CSF) designed to overcome tumor immune evasion and conducted separately in a similar patient population. In addition to evaluating safety and clinical efficacy, I tested the following hypotheses: (1) Clinical benefits are likely to be associated with markers of reversal of immune tolerance (autoimmunity). (2) Clinical benefits may be predicted by baseline peripheral biomarkers of immune tolerance/suppression (C-reactive protein, CRP and absolute lymphocyte count, ALC). (3) Superior antitumor efficacy is likely to be associated with more effective downregulation of the host suppressor immune response (circulating T regulatory cells, T-reg and myeloid derived suppressor cells, MDSC). My findings supported superior clinical efficacy that was associated with more significant modulation of immune tolerance by the combination of IFN-Treme. Autoimmunity correlated with improved clinical outcome among the recipients of IFN-Treme (but not the vaccine) and suggested more significant reversal of immune tolerance. Baseline CRP and ALC were significantly predictive of therapeutic benefit with the IFN-Treme combination and may serve as variables for stratification of future trials, as these are validated in larger studies. Finally, my findings supported more significant downregulation of the host suppressor immune response by the nonspecific IFN-á/Treme regimen as compared to the vaccine-TLR agonist/GM-CSF combination. There was apparent increase in CD4+CD25hi+ CD39+ Treg but this was associated with an increase in the overall CD4+ T cell population suggesting that direct inhibition of CTLA4 suppressive effects on T effector cells leading to their expansion and prolonged activation is likely more important than the regimens effect on T-reg. In addition, I saw parallel downregulation in several populations of MDSC following treatment with IFN-Treme which may have had a role in the reduction of immune suppression and superior clinical outcome observed.