New insights in melanoma biology: Running fast towards precision medicine

Abstract

Cutaneous melanoma is one of the most aggressive and the deadliest form of skin cancers characterized by a high degree of intra-tumor heterogeneity harboring mutations in genes involved in different signaling pathways. The majority of melanomas are diagnosed in the early stage and are treatable with surgical resection with a five-year survival rate ranging 82% to 98% for stages I-II. On the other hand, the prognosis of highly aggressive, late stage, melanoma is still very poor. For later stage and metastatic melanoma the alkylating agents, Dacarbazine or Temozolomide were the earliest treatment option; however, patients easily become resistant to these drugs and they also have serious side effects with low overall response and survival rates. 'Targeted therapies', such as Vemurafenib and Dabrafenib (BRAF V600E inhibitors) , were developed in order to specifically target mutated proteins, with a response rate around 50%; however, most patients relapse within few months [1]. The combination of BRAF inhibitors with MEK inhibitors, such as Cobimetinib and Trametinib, achieved response rates as high as 70%; however, most patients still face progressive disease even if treated with combination therapy, with less than 25% of patients being progression-free at 3 years. Along with the development of 'targeted therapies', the immune-checkpoint inhibitors, such as Ipilimumab (a CTLA-4 inhibitor), Nivolumab and Pembrolizumab (PD-1 inhibitors), were developed to improve the endogenous antitumor immune response. However, they are effective in only a subset of patients and may be associated with serious side effects and the development of acquired resistance in initially responders [2, 3]. Thus,

Artificial intelligence in melanoma: A systematic review

“…Artificial intelligence has multiple main branches, such as machine learning, artificial neural networks, deep learning, and computer vision 69–73 . In the era of precision medicine and the exploding of Internet‐based technology, we have observed the application of all the above algorithms in melanoma 74,75 . Artificial intelligence systems have high‐level performance in many tasks, better predicting patients' prognosis, and assisting in diagnosis 53,63 .…”

“…[69][70][71][72][73] In the era of precision medicine and the exploding of Internet-based technology, we have observed the application of all the above algorithms in melanoma. 74,75 Artificial intelligence systems have high-level performance in many tasks, better predicting patients' prognosis, and assisting in diagnosis. 53,63 As artificial intelligence models continue to advance, there is a growing trend that the roles of clinicians might be revolutionized and redefined in medical practice.…”