Melanoma: A New Frontier in Personalized Medicine

The standard of care for patients with advanced-stage melanoma has shifted from empiric treatment with chemotherapy that has low response rates to targeted therapy with response rates that are substantially higher, and immunotherapy, which has shown the most durable disease control. Chemotherapy is not intrinsically able to differentiate between tumor and normal host tissues and thus results in toxicities such as nausea, vomiting, cardiac and bone marrow effects with myelosuppression, and secondary neoplasm risk, to name a few. On the other hand, targeted agents are aimed at molecular features that are more restricted to cancer cells, and immunotherapeutic agents modulate the immune response and reverse tumor-mediated immunosuppression, with risks of collateral damage through immunomodulation. Thus, targeted therapy and immunotherapy may yield greater efficacy with less or differing patterns of attendant toxicity.

Significant strides have been made toward understanding the mechanisms that drive tumor development and growth, and these discoveries have directly driven the development of new targeted therapies. Tumor targeted and immunomodulatory agents have yielded unprecedented response rates in previously refractory solid tumors. In recent years, multiple new agents have received FDA approval for the treatment of melanoma. These include ipilimumab, peginterferon alfa-2b, and vemurafenib in 2011; dabrafenib and trametinib in 2013; the combination of these 2 agents in 2014; pembrolizu­mab and nivolumab in 2014; the combinations of ipilimumab + nivolumab and vemurafenib + cobimetinib in 2015; and talimogene laherparepvec in 2015.¹

Although responses to these new agents are unprecedented, the fact remains that most patients still do not have enduring responses to immunotherapy. On the other hand, whereas a large proportion of patients demonstrate some benefit from BRAF or BRAF/MEK inhibition, most patients eventually develop resistance.² A fundamental question in the field is how to predict which patients will respond to specific therapies. Defining melanoma populations—or patients—using biomarkers permits a more personalized approach to treatment.

A cancer biomarker is a measurable substance that helps predict disease activity or aid with prognosis. Melanoma is particularly interesting because the American Joint Committee on Cancer (AJCC) TNM staging system takes into account the thickness of the primary tumor, the presence or absence of ulceration, and the presence or absence of nodal micrometastases or macrometastases.³ Even with an initial biopsy of the primary lesion, several features that are routinely described in pathology reports aid with prognosis, such as presence or absence of ulceration, tumor-infiltrating lymphocytes (TILs), and mitoses.³ At the time of sentinel lymph node sampling (for patients with thicker cutaneous melanomas), the presence or absence of melanoma in the sentinel lymph node is also a strong prognostic marker.³

Prognostic factors describe characteristics of the patient or the tumor that may have an impact on the natural course of the disease if left untreated, as well as the patient’s outcome.⁴ It is important to note that in the context of oncology, prognostic markers allow us to identify patients who are at increased risk of recurrence or metastases. Thus, patients who are identified as high risk may benefit from adjuvant systemic therapy. On the other hand, predictive biomarkers can be used to identify which patients are most likely to respond to a certain treatment. Thus, predictive biomarkers allow us to choose a treatment that has the highest likelihood of efficacy in an individual patient (such as hormone receptor or HER2 receptor status in breast cancer patients, or KRAS status in colon cancer).⁴

Genomic Biomarkers

The Cancer Genome Atlas (TCGA) Network recently published a genomic classification of cutaneous melanoma.⁵ This exhaustive analysis included global molecular analysis on a total of 333 cutaneous melanoma samples. Whole exome sequencing was performed on 318 samples. Not surprisingly, the mean mutation rate was the highest that has been reported for any cancer type analyzed by TCGA thus far (16.8 mutations/Mb). Melanomas were then classified based on the identification of significantly mutated genes. Melanomas were thus classified as harboring mutations in BRAF (52%), RAS (28%), NF-1 (14%), or triple wild-type (14%). No significant differences in postaccession survival were found among the 4 groups.⁵

When somatic copy number alterations were assessed across subtypes, focal amplifications of BRAF and programmed death-1 (PD-1) and its ligand 1 (PD-L1) were observed at significant frequencies in the BRAF-mutant subtype. This is especially interesting when one considers the potential combination of BRAF and PD-1 inhibition, or BRAF/MEK and PD-1 inhibition. Furthermore, PTEN mutations and deletions were also more frequent in BRAF-mutant melanomas, suggesting PI3Kβ inhibitors in clinical development may be of benefit in this population, either with a BRAF inhibitor or following progressive disease. Interestingly, loss of PTEN was recently shown to promote resistance to T-cell–mediated immunotherapy. Specifically, loss of PTEN resulted in decreased T-cell tumor infiltration, and treatment with a PI3Kβ inhibitor resulted in improved efficacy of treatment with anti–PD-1 or anti–cytotoxic T-lymphocyte antigen 4 (CTLA-4) antibodies in murine models.⁶

In the TCGA analysis, amplification and mRNA overexpression of AKT3 were significantly increased in the RAS, NF-1, and triple wild-type subgroups compared with the BRAF subgroup. Patients harboring these mutations may therefore benefit from AKT inhibitors more than patients who are BRAF mutant. TERT promoter mutations were observed in a high proportion of BRAF (75%), RAS (71.9%), and NF-1 (83.3%) subtypes, but only in 6.7% of triple wild-type.

Samples were also analyzed on the basis of gene expression. On the basis of the 1500 genes with the most variant expression levels, 329 samples were divided into 3 stable clusters. Based on mRNA transcripts, these were named “immune” (51%), “keratin” (31%), and “MITF low” (18%). In patients with regionally metastatic tumors, postaccession survival was significantly different among these subtypes (P <.001), indicating that these transcriptomically delineated subclasses may be biologically distinct. Specifically, patients in the immune subclass had better survival than patients in the other 2 subclasses (P = .003), which is not surprising considering the immunogenicity and importance of the immune response in melanoma.⁷ To assess the clinical importance of this observation, TCGA developed a semiquantitative measure of the number of lymphocytes in a given sample (LScore). The LScore was significantly associated with prolonged postaccession survival, as would be expected from prior studies that have shown tumor-associated lymphocytes to be a favorable prognostic factor in melanoma.^8,9 Furthermore, a high LScore very strongly correlated with samples that were in the immune subclass (P <1e-12). Levels of 2 lymphocyte signaling–associated proteins (LCK and SYK) were measured to assess whether transcriptome features defining the immune subtype are seen at the protein level. High LCK protein expression correlated with favorable postaccession survival, and these also tended to be in samples with a high LScore. A combination of the 3 overlapping characteristics was strongly predictive of melanoma outcome (P = 8e-6). Given that transcriptome analysis is not practical in the clinical setting, the predictive power of LScore and high LCK expression were tested. Interestingly, samples with high LScore and high LCK expression were associated with significantly higher postaccession survival (P = 7.9e-5).⁵ Thus, combining LCK protein expression and TIL score may be a stronger prognostic tool than either marker alone. A recent study used the TCGA data set to categorize patients by immune phenotype. A median of 123 mutations having a high immunogenicity were found in the T-cell–inflamed cohort versus a median of 176 mutations in the non–T-cell–inflamed cohort. Furthermore, both cohorts showed comparable expression of cancer testis, differentiation, and mutational antigens. Thus, lack of a spontaneous immune infiltrate is unlikely to be due to a lack of antigens, but rather lack of antigen recognition.¹⁰

The TNM classification identifies patients with stage III melanoma as being at high risk of recurrence, and the current standard is to treat those patients with adjuvant high-dose interferon-alpha¹¹ or ipilimumab.¹² Although a positive sentinel lymph node biopsy (SLNB) is the most accurate prognostic marker for patients with melanoma,^13-15 the majority of patients with intermediate thickness primaries who develop metastasis had a negative SLNB.¹⁶ Thus, patients with high-risk stage I/II disease are not captured by AJCC staging. A 28-gene signature was recently developed to identify which patients with early-stage (I/II) melanoma are more likely to develop metastatic disease.¹⁷ Primary tumor samples were evaluated using reverse transcription-polymerase chain reaction, and radial basis machine modeling allowed stratification of tumors into low risk (Class 1) and high risk (Class 2). Five-year disease-free survival (DFS) was 97% for the predicted Class 1 and 31% for the predicted Class 2 (P <.0001). The negative predictive value was 93%, whereas the positive predictive value was 72%. Furthermore, the gene expression profile signature was found to be an independent predictor of metastatic risk.¹⁷ In a separate study, the prognostic accuracy of this gene expression profile was compared with SLNB.¹⁸ The gene expression profile was found to be a better predictor for DFS, distant metastasis-free survival, and overall survival (OS) in both univariate and multivariate analysis (P <.001 for all).¹⁸ However, these results should be interpreted with some caution, as the risk of metastatic disease in the patients with a negative SLNB was 30%, which is higher than would be expected in the general population.

In a recent study, Snyder et al conducted whole exome sequencing on 64 patients with metastatic melanoma who had been treated with ipilimumab.¹⁹ While mutational load correlated with the degree of clinical benefit (P = .01), it was insufficient to predict benefit. However, the investigators discovered a neoantigen signature that was present in tumors of patients who had a strong response to CTLA-4 blockade.¹⁹ Thus, exome sequencing may predict which patients are likely to benefit from treatment with ipilimumab.

Molecular Biomarkers

BRAF is a proto-oncogene that phosphorylates regulatory serine residues on MEK1 and MEK2. Approximately 50% of patients with cutaneous melanoma are BRAF V600 mutant. Of these, 80% to 90% are V600E mutations (glutamic acid replacing valine) while ~5% are V600K mutations (lysine replacing valine).²⁰ BRAF mutations result in activation of the RAS/RAF/MEK/ERK pathway, leading to cellular proliferation and antiapoptotic activity, which ultimately aid in tumor progression. Advanced melanomas carrying BRAF mutations are associated with an earlier age at onset, truncal primaries, and may lack the findings of chronic UV skin damage; clinically, melanoma that is BRAF mutation–positive has a more aggressive clinical course, and patients who are not treated with BRAF inhibitors have shorter OS.²¹ Interestingly, BRAF V600E mutation frequency has been found to be significantly higher (P <.0024) in metastatic melanoma compared to primary melanomas,²² although these results should be interpreted with caution; most of the primary and metastatic samples were not from the same patients, and BRAF mutation may be an acquired mutation that leads to metastasis. BRAF-mutated melanoma has been reported to have a higher risk of ulceration and to have a more advanced stage at initial diagnosis compared with patients with wild-type tumors.²³ Thus, BRAF is both a prognostic and a predictive marker, and 2 BRAF inhibitors have been shown to have significant antitumor activity in phase 3 trials in patients with advanced melanoma.

Vemurafenib, a highly specific BRAF inhibitor, transformed the targeted therapy paradigm. In the phase 3 registration trial, 675 patients with V600E unresectable or metastatic melanoma were randomized to treatment with vemurafenib or dacarbazine.²⁴ Patients in the vemurafenib arm had a 63% decrease in hazard of death (P <.0001) and a 74% decrease in hazard of tumor progression (P <.0001), with a confirmed response rate of 48% compared with 5% with dacarbazine. Benefit was noted in all subgroups, including the high-risk cohorts, with an objective response rate (ORR) of 48%. In the phase 3 dabrafenib trial, 250 patients with V600E unresectable or metastatic melanoma were randomized to receive either dabrafenib or dacarbazine.²⁵ Median progression-free survival (PFS) was 5.1 months for dabrafenib and 2.7 months for dacarbazine, with a hazard ratio (HR) of 0.30 (P <.0001). Although virtually all patients treated with a BRAF inhibitor eventually develop resistance, usually at 7 months, treatment with BRAF inhibition is often continued if the disease is slowly progressive.²⁶ In a retrospective, single-institution analysis of 95 patients who had progressive disease while being treated with BRAF inhibition, continued treatment beyond initial progression was associated with prolonged OS even after adjusting for potential prognostic factors at the time of progression.²⁷

Trametinib, a MEK inhibitor, was initially approved for patients whose disease had progressed on BRAF inhibition based on prolongation of OS.²⁸ It was approved for combination therapy with dabrafenib based on the results of the phase 3 COMBI-d trial, which revealed significant prolongation in PFS and OS with the combination compared with dabrafenib alone (HR, 0.67 and HR, 0.71, respectively).^29,30 ORR was also significantly increased with the combination. Similarly, cobimetinib was approved for use with vemurafenib based on the results of a phase 3 study that revealed significantly increased PFS with the combination compared with vemurafenib plus placebo (HR, 0.58).³¹ ORR was also increased, and there was a trend toward improved OS, although follow-up data have not yet matured.³²

Several mRNA biomarkers have been extensively investigated. Here, we focus on those that have been shown to have some prognostic implication. GalNac-T is a key enzyme involved in synthesis of gangliosides GM2 and GD2, which are markers for aggressive melanoma.^33,34 In blood from 126 melanoma patients, GalNac-T mRNA was significantly increased in patients with AJCC stage II, III, and IV versus patients with stage I.^35,36 All normal donor blood samples (n = 37) were negative for β1→4GalNac-T mRNA expression. Patients were prospectively followed, and 71% of those with progression were GalNac-T positive.³⁵

FABP7 has been shown to be involved in cell proliferation and invasion in vitro.³⁷ Levels of mRNA were tested in cutaneous melanomas and were found to be significantly higher in primary melanomas than in metastatic samples (P <.0001).^36,38 This was also true in 37 paired primary and metastatic samples (P <.001). Furthermore, detection of FABP7 in metastatic tissue inversely correlated with relapse-free survival (RFS) and OS (P <.0001 for both). Interestingly, when FABP7 was tested in patient’s blood as a surrogate biomarker for circulating tumor cells, levels decreased with disease progression. Thus, FABP7 may serve as a prognostic biomarker of early-stage melanoma developing to metastatic disease.³⁸ Similarly, lower expression of survivin, an antiapoptotic molecular inhibitor, has been shown to correlate with improved OS in stage IV patients receiving adjuvant immunotherapy following cytoreductive surgery.³⁹

Cellular Biomarkers

PD-1 is expressed on activated T cells, and when binding to PD-L1, expressed on tumor cells and antigen-presenting cells, downregulates T-cell receptor (TCR) signaling.⁴⁰ This induces T-cell anergy and apoptosis, and, subsequently, immune suppression. Anti–PD-1 agents have yielded unmatched rates of durable clinical responses in patients who respond to treatment, but the fact remains that most patients do not respond. In an elegant study, tumor samples were analyzed from 46 patients with metastatic melanoma before and after treatment with pembrolizumab.⁴¹ Using immunohistochemistry, CD8+ expression was analyzed before and after treatment in the invasive tumor margin and in the tumor parenchyma. CD8+ density in the tumor margin pretreatment was higher in responders compared with nonresponders. A greater increase in CD8+ density from baseline to posttreatment correlated with decreased tumor size (P = .0002), and cells that were double positive for CD8+ and Ki67 in the tumor parenchyma were increased in responders. Granzyme B expression was also increased in the response group (P <.0001) after treatment, as was p-STAT1 before (P = .002) and during (P <.0001) treatment. Furthermore, the response group had increased CD8+, PD-1, and PD-L1 at the invasive margin and the tumor parenchyma compared with the group that progressed (P <.0001, P = .0002, and P = .006, respectively). Interestingly, previous treatment with ipi­limumab did not alter these findings. Quantitative multiplexed immunofluorescence revealed a significant correlation between the proximity of the receptor and its ligand and response to therapy (P = .005). The high number of CD8+ T cells at the margin suggested an antigen-specific immune response, and this was confirmed by the finding that a more restricted TCR beta chain correlated with clinical response to treatment (P = .004). Indeed, samples from responders had over 10 times as many clones expand after treatment. This predictive model was tested in 15 patients whose treatment outcome was blinded to the investigators. They were able to accurately predict the outcome in 4 of 5 patients with progression and 9 of 9 patients with response.⁴¹ Thus, CD8+ density and expression of PD-1/PD-L1 pretreatment may be used in clinical practice as predictors of patients who may benefit from anti–PD-1 therapy.

CTLA-4 is expressed on T cells, and after binding to CD80 or CD86 on antigen-presenting cells acts as a negative regulator of T-cell activation. In a phase 2 neoadjuvant study of ipilimumab (an anti–CTLA-4 antibody) in patients with stage III melanoma, circulating T regulatory cells (Tregs) were significantly increased at week 6 compared with baseline (P = .02), and increased Tregs correlated with improved PFS (P = .034).^42,43 Although Tregs are typically thought to relate with worse outcomes,⁴⁴ the increase noted in this study may be due to early activation of Tregs in the antitumor response.⁴⁵ Circulating myeloid-derived suppressor cells (MDSCs) were decreased, and monocytic MDSCs, a key subset,⁴⁶ were notably reduced (P <.0001). Interestingly, decreased circulating monocytic MDSCs correlated with improved PFS (P = .03).

Similarly, cellular profiling of the tumor microenvironment has been used to predict survival. In a recent study, 62 stage II/III melanomas (32 ulcerated and 30 nonulcerated) were evaluated for the presence of TILs.⁴⁷ TILs were associated with improved OS (P = .034) and RFS (P = .002) in ulcerated but not in nonulcerated melanoma. In addition, CD2, a marker that has previously been shown to correlate with RFS and OS,⁴⁸ correlated with improved OS (P = .021) and RFS (P = .001) in ulcerated melanoma but not in nonulcerated melanoma. BRAF status did not correlate with TILs or CD2+ expression; however, this was done on a small number (41) of tumors.⁴⁷ These findings suggest that TILs may be useful as a prognostic marker in patients with early-stage melanoma and may help determine which patients with stage II melanoma could benefit from trials for adjuvant therapy.

KEY POINTS

• Immune checkpoint inhibitors and targeted therapies have ushered in a new era in the treatment of advanced or metastatic melanoma
• Although the paradigm has shifted, the fact remains that most patients do not respond to immune therapy
• Many respond to targeted therapy but will inevitably develop resistance
• A key question is how to determine which patients are at risk of recurrence, and which are likely to benefit from specific therapies

Conclusion

The use of immune checkpoint inhibitors and targeted therapies has ushered in a new era in the treatment of patients with advanced or metastatic melanoma. Although the paradigm has shifted, with an unprecedented number of durable responses for those patients who respond to these therapies, the fact remains that most patients do not respond to immune therapy. Furthermore, while many respond to targeted therapy, they inevitably develop resistance. As combinatorial and novel strategies continue to develop, a key question is how to determine which patients are at risk of recurrence, and which are likely to benefit from specific therapies. Thus, we must continue to develop prognostic and predictive biomarkers and validate their use in clinical practice. Doing so will allow us to better select which patients to treat with certain therapies to maximize therapeutic benefit while minimizing exposure and toxicity for patients who are unlikely to benefit from these agents and may be better served on clinical trials.

Financial Disclosures

Dr Najjar: No financial disclosures to declare.
Dr Kirkwood: Consultant to Bristol-Myers Squibb, GlaxoSmithKline, Merck; trial support from Prometheus.

References

1. Zhu Z, Liu W, Gotlieb V. The rapidly evolving therapies for advanced melanoma—towards immunotherapy, molecular targeted therapy, and beyond. Crit Rev Oncol Hematol. 2016;99:91-99.
2. Moriceau G, Hugo W, Hong A, et al. Tunable-combinatorial mechanisms of acquired resistance limit the efficacy of BRAF/MEK cotargeting but result in melanoma drug addiction. Cancer Cell. 2015;27:240-256.
3. Balch CM, Gershenwald JE, Soong SJ, et al. Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol. 2009;27:6199-6206.
4. Italiano A. Prognostic or predictive? It’s time to get back to definitions! J Clin Oncol. 2011;29:4718.
5. Cancer Genome Atlas Network. Genomic classification of cutaneous melanoma. Cell. 2015;161:1681-1696.
6. Peng W, Chen JQ, Liu C, et al. Loss of PTEN promotes resistance to T cell-mediated immunotherapy. Cancer Discov. 2016;6:202-216.
7. Ramirez-Montagut T, Turk MJ, Wolchok JD, et al. Immunity to melanoma: unraveling the relation of tumor immunity and autoimmunity. Oncogene. 2003;22:3180-3187.
8. Azimi F, Scolyer RA, Rumcheva P, et al. Tumor-infiltrating lymphocyte grade is an independent predictor of sentinel lymph node status and survival in patients with cutaneous melanoma. J Clin Oncol. 2012;30:2678-2683.
9. Bogunovic D, O’Neill DW, Belitskaya-Levy I, et al. Immune profile and mitotic index of metastatic melanoma lesions enhance clinical staging in predicting patient survival. Proc Natl Acad Sci U S A. 2009;106:20429-20434.
10. Gajewski TF, Zha Y, Hernandez K, et al. Density of immunogenic antigens and presence or absence of the T cell-inflamed tumor microenvironment in metastatic melanoma. J Clin Oncol. 2015;33(suppl). Abstract 3002.
11. Kirkwood JM, Ibrahim JG, Sosman JA, et al. High-dose interferon alfa-2b significantly prolongs relapse-free and overall survival compared with the GM2-KLH/QS-21 vaccine in patients with resected stage IIB-III melanoma: results of intergroup trial E1694/S9512/C509801. J Clin Oncol. 2001;19:2370-2380.
12. Eggermont AM, Chiarion-Sileni V, Grob JJ, et al. Adjuvant ipilimumab versus placebo after complete resection of high-risk stage III melanoma (EORTC 18071): a randomised, double-blind, phase 3 trial. Lancet Oncol. 2015;16:522-530.
13. Balch CM, Gershenwald JE, Soong SJ, et al. Multivariate analysis of prognostic factors among 2,313 patients with stage III melanoma: comparison of nodal micrometastases versus macrometastases. J Clin Oncol. 2010;28:2452-2459.
14. Morton DL, Thompson JF, Cochran AJ, et al. Final trial report of sentinel-node biopsy versus nodal observation in melanoma. N Engl J Med. 2014;370:599-609.
15. Wong SL, Balch CM, Hurley P, et al. Sentinel lymph node biopsy for melanoma: American Society of Clinical Oncology and Society of Surgical Oncology joint clinical practice guideline. J Clin Oncol. 2012;30:2912-2918.
16. Morton DL, Thompson JF, Cochran AJ, et al. Sentinel-node biopsy or nodal observation in melanoma. N Engl J Med. 2006;355:1307-1317.
17. Gerami P, Cook RW, Wilkinson J, et al. Development of a prognostic genetic signature to predict the metastatic risk associated with cutaneous melanoma. Clin Cancer Res. 2015;21:175-183.
18. Gerami P, Cook RW, Russell MC, et al. Gene expression profiling for molecular staging of cutaneous melanoma in patients undergoing sentinel lymph node biopsy.J Am Acad Dermatol. 2015;72:780-785.
19. Snyder A, Makarov V, Merghoub T, et al. Genetic basis for clinical response to CTLA-4 blockade in melanoma. N Engl J Med. 2014;371:2189-2199.
20. Davies H, Bignell GR, Cox C, et al. Mutations of the BRAF gene in human cancer. Nature. 2002;417:949-954.
21. Long GV, Menzies AM, Nagrial AM, et al. Prognostic and clinicopathologic associations of oncogenic BRAF in metastatic melanoma. J Clin Oncol. 2011;29:1239-1246.
22. Shinozaki M, Fujimoto A, Morton DL, et al. Incidence of BRAF oncogene mutation and clinical relevance for primary cutaneous melanomas. Clin Cancer Res. 2004;10:1753-1757.
23. Ellerhorst JA, Greene VR, Ekmekcioglu S, et al. Clinical correlates of NRAS and BRAF mutations in primary human melanoma. Clin Cancer Res. 2011;17:229-235.
24. Chapman PB, Hauschild A, Robert C, et al. Improved survival with vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med. 2011;364:2507-2516.
25. Hauschild A, Grob JJ, Demidov LV, et al. Dabrafenib in BRAF-mutated metastatic melanoma: a multicentre, open-label, phase 3 randomised controlled trial. Lancet. 2012;380:358-365.
26. Carlino MS, Gowrishankar K, Saunders CA, et al. Antiproliferative effects of continued mitogen-activated protein kinase pathway inhibition following acquired resistance to BRAF and/or MEK inhibition in melanoma. Mol Cancer Ther. 2013;12:1332-1342.
27. Chan MM, Haydu LE, Menzies AM, et al. The nature and management of metastatic melanoma after progression on BRAF inhibitors: effects of extended BRAF inhibition. Cancer. 2014;120:3142-3153.
28. Flaherty KT, Robert C, Hersey P, et al. Improved survival with MEK inhibition in BRAF-mutated melanoma. N Engl J Med. 2012;367:107-114.
29. Long GV, Stroyakovskiy D, Gogas H, et al. Combined BRAF and MEK inhibition versus BRAF inhibition alone in melanoma. N Engl J Med. 2014;371:1877-1888.
30. Long GV, Stroyakovskiy D, Gogas H, et al. Dabrafenib and trametinib versus dabrafenib and placebo for Val600 BRAF-mutant melanoma: a multicentre, double-blind, phase 3 randomised controlled trial. Lancet. 2015;386:444-451.
31. Larkin J, Ascierto PA, Dréno B, et al. Combined vemurafenib and cobimetinib in BRAF-mutated melanoma. N Engl J Med. 2014;371:1867-1876.
32. Larkin JMG, Yan Y, McArthur GA, et al. Update of progression-free survival (PFS) and correlative biomarker analysis from coBRIM: phase III study of cobimetinib (cobi) plus vemurafenib (vem) in advanced BRAF-mutated melanoma. J Clin Oncol. 2015;33(suppl). Abstract 9006.
33. Tsuchida T, Saxton RE, Irie RF. Gangliosides of human melanoma: GM2 and tumorigenicity. J Natl Cancer Inst. 1987;78:55-60.
34. Tsuchida T, Saxton RE, Morton DL, et al. Gangliosides of human melanoma. J Natl Cancer Inst. 1987;78:45-54.
35. Kuo CT, Bostick PJ, Irie RF, et al. Assessment of messenger RNA of β1→4-N-acetylgalactosaminyl-transferase as a molecular marker for metastatic melanoma. Clin Cancer Res. 1998;4:411-418.
36. Tanaka R, Koyanagi K, Narita N, et al. Prognostic molecular biomarkers for cutaneous malignant melanoma. J Surg Oncol. 2011;104:438-446.
37. Goto Y, Matsuzaki Y, Kurihara S, et al. A new melanoma antigen fatty acid-binding protein 7, involved in proliferation and invasion, is a potential target for immunotherapy and molecular target therapy. Cancer Res. 2006;66:4443-4449.
38. Goto Y, Koyanagi K, Narita N, et al. Aberrant fatty acid-binding protein-7 gene expression in cutaneous malignant melanoma. J Invest Dermatol. 2010;130:221-229.
39. Takeuchi H, Morton DL, Elashoff D, et al. Survivin expression by metastatic melanoma predicts poor disease outcome in patients receiving adjuvant polyvalent vaccine. Int J Cancer. 2005;117:1032-1038.
40. Pardoll DM. The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer. 2012;12:252-264.
41. Tumeh PC, Harview CL, Yearley JH, et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature. 2014;515:568-571.
42. Tarhini AA, Zahoor H, Lin Y, et al. Baseline circulating IL-17 predicts toxicity while TGF-β1 and IL-10 are prognostic of relapse in ipilimumab neoadjuvant therapy of melanoma. J Immunother Cancer. 2015;3:39.
43. Tarhini AA, Edington H, Butterfield LH, et al. Immune monitoring of the circulation and the tumor microenvironment in patients with regionally advanced melanoma receiving neoadjuvant ipilimumab. PLoS One. 2014;9:e87705.
44. Mougiakakos D, Choudhury A, Lladser A, et al. Regulatory T cells in cancer. Adv Cancer Res. 2010;107:57-117.
45. Delgoffe GM, Woo SR, Turnis ME, et al. Stability and function of regulatory T cells is maintained by a neuropilin-1-semaphorin-4a axis. Nature. 2013;501:252-256.
46. Walter S, Weinschenk T, Stenzl A, et al. Multipeptide immune response to cancer vaccine IMA901 after single-dose cyclophosphamide associates with longer patient survival. Nat Med. 2012;18:1254-1261.
47. de Moll EH, Fu Y, Qian Y, et al. Immune biomarkers are more accurate in prediction of survival in ulcerated than in non-ulcerated primary melanomas. Cancer Immunol Immunother. 2015;64:1193-1203.
48. Harcharik S, Bernardo S, Moskalenko M, et al. Defining the role of CD2 in disease progression and overall survival among patients with completely resected stage-II to -III cutaneous melanoma. J Am Acad Dermatol. 2014;70:1036-1044.