One of the first directed therapeutic strategies to be tested in clinical trials for the treatment of metastatic melanoma targeted Bcl-2, an antiapoptotic protein that is overexpressed, thereby favoring cell survival. This overexpression occurs in many neoplasms and in 80% of melanomas. One drug, oblimersen can be used as Bcl-2 antisense therapy. The first phase I-II clinical trial in which oblimersen and dacarbazine were combined obtained acceptable results, encouraging design of another trial. In that phase III trial patients were randomized to a group given only dacarbazine or a group given dacarbazine and oblimersen. In total, 771 patients were included. The group that received dacarbazine and oblimersen had a higher objective response rate (13.5% vs 7.5% in the dacarbazine-only group; P=.007) and a longer disease-free interval (median, 2.6 months vs 1.6 months; P<.001). However, there was only a numerical trend towards longer overall survival (median, 9.0 months vs 7.8 months; P=.077). Although a retrospective analysis indicated that the subgroup of patients without elevated serum lactate dehydrogenase hormone (LDH) did benefit in terms of overall survival (median, 11.4 vs 9.7 months; P=.02),16 this finding was not confirmed in subsequent trials. Oblimersen has been combined with other cytostatic agents such as paclitaxel and temozolomide, but the survival outcomes have not improved and so this line of investigation has been abandoned as a possible treatment for melanoma.8,17 Currently, treatments such as ABT-737 and ABT-263 (BH3 mimetics) and YM155 (survivin inhibitor) targeting other antiapoptotic proteins are being tested in clinical trials.17

RAF is a family of proteins (ARAF, BRAF, and CRAF) that intervene in intracellular signaling of MAP kinases, thereby regulating cell proliferation, differentiation, and survival. BRAF mutations are present in 30% to 70% of melanomas and, as mentioned above, particularly in those associated with chronic sun-induced damage,12,13 the most common type of melanoma. In more than 90% of cases, the BRAF mutation is always the same, a point mutation at exon 15 (thymine replaced by adenine). The result is that valine is replaced by glutamic acid at position 600 of the BRAF protein (V600E mutation). In addition, 15% to 30% of melanomas have NRAS mutations. As NRAS encodes a protein located early in the MAP kinase pathways (Fig. 1), it also acts as a BRAF activator. Interestingly, BRAF mutations have been detected in 20% and NRAS mutations in 80% of common acquired and congenital melanocytic nevi, although the biological significance of this finding is not clear.10 The above findings suggest that BRAF is a very important protein for proliferation of melanocytic cells.18–20

Studies have attempted to find a link between the presence of BRAF and NRAS mutations and the clinical and histopathologic characteristics of melanomas with these mutations.21–23 Although further studies are necessary, in summary, we can now say that melanomas with BRAF mutations are usually more pigmented lesions and are located on the trunk and limbs of middle-aged adults (<50 years) who have a history of sun exposure in childhood, few freckles, and no actinic keratosis. From the histologic point of view, they are found on skin with little actinic elastosis and are more frequently associated with the superficial-spreading variant of melanoma; neoplastic cells tend to invade the upper strata of the epidermis, where they form intraepidermal nests (pagetoid growth pattern); the epidermis is usually thicker and the lesion sharply defined with respect to surrounding skin; the cells are usually rounded, large, and clearly pigmented; and the Breslow depth and mitotic rate are usually lower. Clinically, metastasis tends to be to regional lymph nodes and survival is longer.

c-kit has been considered a potential therapeutic target in melanoma for a long time. In fact, c-kit is a protein that acts as a fundamental growth factor receptor in epidermal melanocytes and has an essential role in the differentiation and migration of melanocytic cells during embryonic development.37 Consistent with this, many melanomas express c-kit. However, in recent years, certain clinical and nonclinical findings have suggested that c-kit is not a valid target. On the one hand, lower c-kit expression as many melanomas progress has been demonstrated,38,39 and we know that in certain cell lines, loss of this growth factor is, paradoxically, related to increased metastatic potential.40,41 On the other, the first clinical trials in patients with metastatic melanoma using c-kit inhibitors, such as imatinib, had discouraging results.42,43

Although there is experimental and clinical evidence of isolated responses to imatinib and of rare mutations in exon 11 of c-kit in some melanoma cells,44–48 this gene was not considered a possible therapeutic target until 2006, when Curtin et al,13 who investigated certain types of uncommon melanomas, established that 39% of mucosal melanomas, 36% of acral melanomas, and 28% of melanomas had c-kit abberations.13 Such aberrations have not been found in uveal melanoma. The low proportion of these types of melanomas in the aforementioned clinical studies of metastatic melanoma would explain the lack of response observed.49 In a clinical trial, designed before these findings were known, but published in 2008, the only patient with response to imatinib had acral lentiginous melanoma.50Genetic aberrations in c-kit include mutations and amplifications of this gene. This point is important because the 2 types of aberration do not necessarily have to have the same biological significance. Of the 39% to 28% of all genetic aberrations of c-kit, only 15% to 38% of mucosal melanomas, 8% to 23% of acral melanomas, and 0 to 17% of melanomas on skin with chronic sun-induced damage are related to mutations. Most mutations, as in gastrointestinal stromal tumors (GIST) in which c-kit mutations are typical, are located on exon 11. However, higher proportions of tumors are associated with mutations located on exons 13, 17, and 18 in melanoma than in GIST and these mutations are in fact the ones associated with resistance to imatinib. Moreover, the coexistence of mutations and amplifications of c-kit has been reported in melanoma but is uncommon in GIST. Studies performed to date seem to indicate that there is not always a relationship between positive and negative immunohistochemical findings for c-kit gene products and the presence or absence of genetic aberrations. Thus, molecular studies seem to be essential to establish whether aberrations of this gene are present.14,51–54

Drugs other than imatinib (sunitinib, dasatinib, and nilotinib) that inhibit the tyrosine-kinase activity of the c-kit gene product are already on the market.3 The ability of these drugs to induce the regression of melanomas with c-kit activating mutations has been shown in several isolated clinical trials2,55–58 and in small series of patients.51 Currently, several phase II studies are in progress with different c-kit inhibitors in patients with metastatic melanoma and this genetic aberration. These trials aim to assess the following aspects: 1) whether the different susceptibility and resistance of various c-kit mutations to the different c-kit inhibitors has significant clinical implications, such that the drug to be used should be selected according to the mutation present in each case; 2) which inhibitors act effectively on metastases of the central nervous system; 3) whether clinical response varies according to whether the melanoma has a c-kit activating mutation or only amplifications, or a combination of an activating mutation and an amplification (in a study of imatinib, a response rate of 50% was attained in patients with melanomas with c-kit mutations, but no responses were observed when only amplification was present59); and 4) how resistance to c-kit inhibitors develops. We know that in GIST, this usually occurs through the appearance of additional c-kit mutations, but it is still not known what the mechanism might be in melanoma.8,52,53

The isoforms of the RAS oncogene are KRAS, HRAS, and NRAS. Between 15% and 30% of melanomas have NRAS mutations. RAS activating mutations stimulate the MAP kinase pathway and also the PI3K/AKT pathway (Fig. 1) among others. The first drugs used to try to inhibit the MAP kinase pathway were inhibitors of farnesyl transferase of RAS (tipifarnib or R115777). The only phase II study (single-arm design) was terminated due to lack of response. However, none of the patients were selected for the presence of NRAS mutations. There is some evidence that RAS antagonists can enhance the effect of chemotherapy, but this approach has not been tested in clinical trials. We need to see more work on the synthesis of new more effective RAS inhibitors.8 Another alternative would be a concerted inhibition of the 2 main pathways activated as a result of RAS activation, namely, the MAP kinase pathway and the PI3K/AKT pathway14,60 (see section on Combination Treatments).

MEK is a protein that belongs to the MAP kinase pathway downstream from BRAF. MEK inhibitors synthesized to date are PD0325901, AZD6244, GSK1120212, and E6201. The results of the first phase I and phase II trials suggest that these pharmacologic agents are not effective monotherapies in the treatment of melanoma. However, extensive preclinical data suggest that they would be a good choice to use in combined regimens, both for avoiding resistance when using drugs targeting the BRAF V600E mutation and for treating melanoma with BRAF mutations other than V600E or NRAS mutations, especially if combined with inhibitors of the PI3K/AKT pathway8,14,54,61 (see section on Combination Treatments).

The PI3K/AKT pathway is triggered after RAS activation. In melanoma, the pathway is usually highly activated by genetic aberrations that block the physiological regulatory effect of PTEN (NRAS oncogenic mutations [15% to 30% of melanomas], mutations [10% to 20% of melanomas], or silencing of the PTEN tumor suppressor gene) or by AKT amplifications, and recently an uncommon point mutation of AKT3 E17K that activates AKT was identified (Fig. 1).8,10,14,54,61 Various derivatives of rapamycin (CCI-779 or temsirolimus) have been used as inhibitors of the PI3K/AKT pathway. These agents act on mTOR, a molecule located downstream of AKT/PKB (Fig. 1). There are also dual inhibitors of PI3K and mTOR and of PI3K and AKT.62 Although the clinical outcomes in phase-II trials of these drugs have not been good, several authors have proposed they be used in combination therapy, particularly with inhibitors of the MAP kinase pathway8,10,14,54,61 or even in a combination that simultaneously inhibits 2 points on the PI3K/AKT pathway63 (see section on Combination Treatments). However, the results of a study in a murine model of melanoma suggest that inhibition of the PI3K/AKT pathway could induce immunosuppression of the host, thereby favoring tumor growth.64

Inhibition with drugs that target other molecules in other pathways related to proliferation/survival, such as CDK4 or cyclin D1, which are often amplified in melanoma, has not been adequately achieved.60

The term pleiotropic refers to molecules that act on multiple different cell functions. Within this group, we will discuss the proteasome inhibitors and the histone deacetylase inhibitors. These inhibitors affect functions that are important for tumor development and progression, favoring arrest of the cell cycle, apoptosis, decreased invasive and cell migration capacity, generation of reactive oxygen species, inhibition of angiogenesis, and autophagy.