Transcriptomic Analysis Reveals Prognostic Molecular Signatures of Stage I Melanoma

Abstract

Purpose: Previously identified transcriptomic signatures have been based on primary and metastatic melanomas with relatively few American Joint Committee on Cancer (AJCC) stage I tumors, given difficulties in sampling small tumors. The advent of adjuvant therapies has highlighted the need for better prognostic and predictive biomarkers, especially for AJCC stage I and stage II disease.Experimental Design: A total of 687 primary melanoma transcriptomes were generated from the Leeds Melanoma Cohort (LMC). The prognostic value of existing signatures across all the AJCC stages was tested. Unsupervised clustering was performed, and the prognostic value of the resultant signature was compared with that of sentinel node biopsy (SNB) and tested as a biomarker in three published immunotherapy datasets.Results: Previous Lund and The Cancer Genome Atlas signatures predicted outcome in the LMC dataset (P ¼ 10 À8 to 10 À4 ) but showed a significant interaction with AJCC stage (P ¼ 0.04) and did not predict outcome in stage I tumors (P ¼ 0.3-0.7). Consensus-based classification of the LMC dataset identified six classes that predicted outcome, notably in stage I disease. LMC class was a similar indicator of prognosis when compared with SNB, and it added prognostic value to the genes reported by Gerami and colleagues. One particular LMC class consistently predicted poor outcome in patients receiving immunotherapy in two of three tested datasets. Biological characterization of this class revealed high JUN and AXL expression and evidence of epithelial-to-mesenchymal transition.Conclusions: A transcriptomic signature of primary melanoma was identified with prognostic value, including in stage I melanoma and in patients undergoing immunotherapy.

Tumour gene expression signature in primary melanoma predicts long-term outcomes: A prospective multicentre study

“…The published signatures from Thakur et al (16) (LMC_150; n=150 genes) as well as those from Gerami et al (12) (Decision-Dx Melanoma™; n=27 genes) were not found to be significantly associated with survival in multivariate models in our dataset, which may in part be reflective a higher proportion of stage III patients in these data than in these earlier studies (Supplementary figure S7).…”

“…In particular, we found that adding the signature to the clinical covariates correctly predicted an additional 42 cases (Figure 1E). We then tested the performance of published prognostic signatures from Gerami et al (12) (Decision-Dx Melanoma™; n=27 genes) and Thakur et al (16) (LMC_150; n=150 genes) using the same machine learning models in our dataset and were not able to deduce that these signatures perform better than the baseline clinical covariates at the 5% significance level (Supplementary figure S4 & Supplementary table S3).…”

“…A primary melanoma transcriptomic dataset from the Leeds melanoma cohort study (LMC DASL array) was used as independent replication. This represents a population-controlled cohort study, as previously described (16). This study recorded data on melanoma-specific survival in 677 patients, calculated from the time of diagnosis to the time of last follow-up or time of death from melanoma, whichever occurred first.…”

“…The test measures the expression of individual genes from formalin-fixed paraffin-embedded (FFPE) primary melanomas to provide a binary classification of low (class 1) or high (class 2) risk for developing metastases within 5 years of diagnosis (with A and B subclasses to further stratify risk) (13-15). A further recent unsupervised clustering analyses based on 677 primary melanoma transcriptomes embedded within a population-controlled cohort study from the Leeds Melanoma Cohort (LMC) identified a six-class 150-gene prognostic signature (herein reference to as LMC_150) (16). The signature uniquely demonstrated prognostic relevance in patients with stage I primary melanoma and further predicted poor outcomes in patients undergoing immunotherapy (16).…”

“…A further recent unsupervised clustering analyses based on 677 primary melanoma transcriptomes embedded within a population-controlled cohort study from the Leeds Melanoma Cohort (LMC) identified a six-class 150-gene prognostic signature (herein reference to as LMC_150) (16). The signature uniquely demonstrated prognostic relevance in patients with stage I primary melanoma and further predicted poor outcomes in patients undergoing immunotherapy (16). However, owing at least in-part to a lack of prospective reproducible data, proving the clinical utility of such prognostic molecular tools remains a considerable challenge and there are currently no established prognostic biomarkers able to accurately identify truly high-risk patients (17).…”