Predicting melanoma outcome using clinical and biological indicators

Abstract

Globally, we are seeing an increasing rate of melanoma while melanoma mortality is reducing. This is significant as more patients with cancers including melanoma are now surviving longer, which increases their risk of being diagnosed with subsequent melanomas. In chapter 2, we examined patients with multiple invasive melanomas, looking at their pattern of progression towards sentinel node metastasis. 651 patients with primary invasive cutaneous melanoma who were referred to the Princess Alexandra Hospital melanoma clinic and underwent sentinel lymph node biopsy between 1994 and 2011 were included in this study. Information on their index melanoma as well as any melanomas before and after were collected and analysed. Logistic regression model was used to determine which factors predicted sentinel node metastasis. Additional stage II melanoma was an independent and significant predictor of sentinel lymph node metastasis. When we evaluated melanoma-specific survival in this cohort, patients with additional stage I melanoma had better melanoma-specific survival, whereas additional stage II and stage II melanoma prior to index melanoma had poorer melanoma-specific survival as expected. Patients with additional invasive melanomas fare poorer compared to those with single invasive melanoma, especially if the additional melanoma is of a later stage. This highlights the importance of patient education in the melanoma survivor population, as well as closer surveillance to diagnose any subsequent melanomas at an early stage.Next, we explored the role of various biomarkers using immunohistochemistry, including P-STAT5 (phosphorylated signal transducer and activator of transcription 5) and a panel comprising SOX18 (SRY-related HMG-box 18), Ki67, CD31, D2-40, in predicting melanoma outcome in a cohort of clinical stage I and II melanoma patients. Identifying novel prognostic biomarkers in locally invasive melanomas to predict clinical outcomes is crucial with the new possibilities in adjuvant therapy. It enables further stratification of patients to reserve adjuvant therapy to high risk patients, provide closer follow up to improve survival, while preventing unnecessary exposure of low risk patients to toxicity from adjuvant therapy.We studied cytokine receptor signalling via P-STAT5 to reflect tumour immune function as a potential prognostic marker of melanoma. We studied a cohort of stage Ib and II melanoma patients, with 189 melanoma tissue samples being analysed for active P-STAT5 level by immunohistochemistry and correlated this with patient and tumour characteristics and survival outcomes after an average of 10 years of follow-up. High P-STAT5 level was associated with poor prognostic factors of ulceration and increased tumour thickness. However, it was found to be independently and inversely associated with melanoma-specific death to 10 years (HR=0.25 [0.07-0.87] p=0.029). P-STAT5 allows to identify a subset of patients that have better melanoma outcomes despite having poor clinicopathological characteristics at diagnosis possibly through immune system activation. This finding could allow major improvements in stratification of the risk of progression.Melanoma progression and metastasis via the vascular and lymphatic channels results in extremely poor prognosis. We evaluated the expression of biomarkers representing melanoma vascularisation and assessed their association with melanoma outcomes including SLN metastasis and melanoma death in a prospective cohort. The presence of Ki67+ nuclei in D2-40+ vessels was an independent predictor of SLN metastasis and melanoma death, which is a novel finding as a combination biomarker in a cohort of melanoma patients. Positive SOX18 was associated with lower SLN metastasis and better melanoma-specific survival.