Melanoma is a cancer arising from melanocytes, the cells that produce the pigment melanin. Most melanomas develop in the skin, but they can also originate in the eye, mucosal surfaces, and, rarely, other sites. Although melanoma is less common than the major keratinocyte cancers of the skin, its capacity to spread through lymphatic vessels and the bloodstream makes it responsible for a disproportionate share of skin cancer deaths.
Melanoma encompasses biologically distinct diseases rather than a single uniform tumor. Ultraviolet-associated cutaneous melanomas differ genetically and clinically from melanomas of the palms, soles, nail apparatus, mucous membranes, and uveal tract of the eye. Tumor thickness, ulceration, regional spread, and distant metastasis are central determinants of outcome. Surgical treatment can be curative for localized disease, while immune checkpoint inhibition and molecularly targeted drugs have substantially changed the treatment of advanced disease since 2011.
Cellular origin and molecular biology
Melanocytes originate predominantly from the embryonic neural crest and migrate into the skin and several other tissues during development. In the epidermis, they occupy the basal layer and transfer pigment-containing melanosomes to neighboring keratinocytes. Melanin contributes to skin and hair color and helps protect cellular DNA from ultraviolet damage. Melanoma develops when accumulated genetic and regulatory alterations allow a melanocyte or its precursor to proliferate, evade normal growth controls, invade surrounding tissue, and potentially disseminate.
A benign melanocytic nevus, commonly called a mole, is a localized proliferation of melanocytes. Many nevi contain activating alterations in growth-signaling genes, particularly BRAF, yet remain stable because additional mechanisms restrain cell division. The presence of an oncogenic alteration alone therefore does not establish malignancy. Progression can involve disruption of cell-cycle control, maintenance of chromosome ends through telomerase activation, and loss of mechanisms that normally trigger senescence or cell death. Most cutaneous melanomas arise without a recognizable pre-existing nevus, although a minority develop in association with one.
The mitogen-activated protein kinase pathway is a major driver of cutaneous melanoma. Activating BRAF variants occur in approximately 40–50% of cutaneous melanomas, with the V600E substitution being the most frequent. NRAS alterations occur in a smaller proportion, and loss of NF1 function defines another molecular group. These changes activate signaling through the RAS–RAF–MEK–ERK cascade, promoting growth and survival. Alterations involving CDKN2A, PTEN, and the TERT promoter can cooperate with these drivers. Molecular categories overlap imperfectly with the traditional microscopic subtypes, so appearance alone does not reliably predict a tumor’s actionable genetic alterations.
Ultraviolet-associated melanomas often carry numerous somatic mutations, including characteristic substitutions at adjacent pyrimidine bases. This high tumor mutational burden can generate abnormal proteins recognizable by immune cells. Nevertheless, tumors may escape immune destruction by reducing antigen presentation, excluding immune cells from the tumor microenvironment, or engaging inhibitory pathways such as PD-1. This interaction with the immune system provides the biological basis for several effective melanoma treatments.
Other melanoma groups follow different molecular routes. Acral and mucosal melanomas tend to have fewer ultraviolet-associated point mutations and more large-scale chromosome alterations. KIT alterations occur in subsets of these tumors. Uveal melanoma commonly involves GNAQ or GNA11 pathway alterations, while changes involving BAP1, SF3B1, or EIF1AX help distinguish biologically different groups. These differences explain why a drug effective in one melanoma subtype cannot automatically be assumed effective in another.
Epidemiology and population distribution
The worldwide burden of melanoma is unevenly distributed. International estimates for 2022 placed the annual number of newly diagnosed cutaneous melanomas at approximately 330,000, with about 60,000 deaths. These figures describe melanoma of the skin rather than all ocular and mucosal melanomas. Comparisons between countries depend on cancer registration quality, population age structure, diagnostic practices, and the completeness of death certification, as well as on genuine differences in exposure and susceptibility.
The highest recorded incidence rates occur in populations with a high proportion of people of European ancestry living in high-ultraviolet environments, particularly Australia and New Zealand. Rates are also high in parts of northern and western Europe and North America. Melanoma is much less common in many African and Asian populations, although low average incidence does not mean that individuals in those populations cannot develop it. Acral melanoma represents a larger proportion of melanomas in darker-skinned populations chiefly because ultraviolet-associated cutaneous melanoma is less frequent, not because acral melanoma is necessarily more common in absolute terms.
Incidence generally increases with age, but melanoma also contributes appreciably to cancer diagnoses in younger adults. Patterns differ by sex and age: in several predominantly light-skinned populations, women have higher incidence at younger adult ages, while men have higher incidence later in life. Anatomical distributions also differ. The trunk is a frequent site in men, whereas the lower limbs are relatively prominent in women. These patterns reflect a combination of biological factors, clothing and exposure practices, and differences in detection.
Over recent decades, many countries have recorded rising melanoma incidence, particularly among older adults. Increased ultraviolet exposure, changing recreational habits, longer lifespans, more intensive examination, and more frequent biopsy all contribute. An increase in diagnosis does not necessarily represent an equivalent increase in life-threatening disease. Detection of thin lesions and melanoma in situ can raise recorded incidence, and some lesions may never have caused clinical harm during the affected person’s lifetime. This possibility is studied as overdiagnosis, although separating it from beneficial early detection is difficult.
Mortality trends have not always paralleled incidence. In some countries, melanoma death rates have stabilized or declined despite continued growth in diagnoses. Earlier detection, prevention efforts, and effective treatment for advanced disease may all contribute. Population-level interpretation requires attention to age-standardized rates and treatment availability rather than simple counts. Differences in stage at diagnosis, specialist access, and timely treatment also connect melanoma outcomes with social determinants of health and the organization of health services.
Environmental, phenotypic, and inherited susceptibility
Ultraviolet radiation is the principal established environmental cause of most cutaneous melanomas. UVB can directly damage DNA, while UVA contributes through oxidative damage and other mechanisms; both are relevant to carcinogenesis. Intermittent intense exposure and episodes of sunburn, particularly early in life, are associated with increased melanoma risk. Chronic cumulative exposure is especially important for some tumors arising on persistently sun-damaged skin, including lentigo maligna melanoma. The relationship is therefore not adequately described by a single measure of total lifetime sunlight.
Artificial tanning devices are another source of carcinogenic ultraviolet exposure. In 2009, the International Agency for Research on Cancer classified ultraviolet-emitting tanning devices as carcinogenic to humans. Exposure patterns, age at first use, and frequency of use affect observed associations. Ultraviolet exposure is not, however, an established explanation for every melanoma: it has little or no demonstrated causal role in most acral, mucosal, and uveal melanomas.
Constitutional traits strongly influence susceptibility. Fair skin that burns readily, freckling, light eyes, and red or blond hair are associated with higher risk of ultraviolet-related melanoma. Variation in the MC1R gene contributes to pigmentation differences and melanoma susceptibility. A high number of common nevi and the presence of atypical nevi are also established risk markers. These features indicate increased susceptibility but do not imply that each individual mole is likely to transform into cancer.
Family history can reflect shared pigmentation, shared environmental exposure, and inherited predisposition. Germline pathogenic variants in CDKN2A are among the best-established causes of familial melanoma and can also be associated with pancreatic cancer. Less common predisposition genes include CDK4, BAP1, and POT1. BAP1 tumor predisposition syndrome can involve uveal melanoma, cutaneous melanoma, mesothelioma, and other tumors. Large congenital melanocytic nevi are another distinct setting associated with increased melanoma risk, although the magnitude of risk varies substantially with the characteristics of the congenital lesion.
Personal history is also important. A previous melanoma identifies a population with increased risk of an additional primary melanoma. Immunosuppression, including that associated with solid-organ transplantation, is linked to increased melanoma incidence and can affect outcomes. These associations do not make melanoma an infectious disease; rather, they indicate that immune surveillance contributes to control of malignant melanocytic cells.
Risk factors operate probabilistically rather than deterministically. Many people with extensive sun exposure or numerous nevi never develop melanoma, while some melanomas occur without conspicuous risk factors. Likewise, the absence of dark pigment does not exclude melanocytic malignancy, and darker skin does not eliminate risk. Etiological distinctions are particularly important when considering tumors on the soles, beneath nails, or at mucosal sites.
Cutaneous forms and clinical manifestations
The traditional clinical and microscopic classification of cutaneous melanoma includes superficial spreading melanoma, nodular melanoma, lentigo maligna melanoma, and acral lentiginous melanoma. These categories remain useful descriptions, although modern classification increasingly incorporates the degree of solar damage, anatomical location, and molecular alterations. A subtype describes a pattern of tumor development; it does not replace measurement of thickness or assessment of spread.
Superficial spreading melanoma is the most frequent subtype in many predominantly European-ancestry populations. It commonly begins as an irregularly pigmented patch or slightly raised lesion with asymmetry, uneven borders, and variation in color. A lateral or “radial” growth component may precede substantial invasion. Nodular melanoma, by contrast, often presents as a raised, rapidly enlarging papule or nodule and may become deeply invasive without a conspicuous broad superficial component. It can be black, blue, brown, pink, or skin-colored, and may ulcerate or bleed.
Lentigo maligna is melanoma in situ arising on chronically sun-damaged skin, most often on the head and neck of older adults. When an invasive component develops, the lesion is termed lentigo maligna melanoma. Its often indistinct peripheral extent and extension along hair follicles can complicate assessment of surgical margins. Acral lentiginous melanoma occurs on the palms, soles, and nail apparatus. Nail-unit melanoma may produce an enlarging longitudinal pigmented band, irregular pigmentation, nail disruption, or a mass. Pigment extending onto adjacent nail-fold skin, termed Hutchinson sign, is an important clinical finding but is not independently diagnostic.
The ABCDE mnemonic describes asymmetry, border irregularity, color variation, diameter, and evolution. Diameter greater than about 6 millimeters is a conventional warning feature, but smaller melanomas occur. Evolution—change in size, shape, color, surface, or symptoms—is particularly important. The “ugly duckling” concept refers to a lesion that differs from an individual’s other pigmented lesions. Neither approach detects every melanoma, especially nodular, amelanotic, or unusual-site tumors.
Amelanotic melanoma contains little visible pigment and can resemble an inflamed papule, a vascular lesion, or another skin cancer. Desmoplastic melanoma is an uncommon fibrous form, often occurring on sun-damaged head and neck skin; it may resemble a scar and can show growth along nerves. Symptoms such as itching, tenderness, bleeding, or ulceration are nonspecific, and early melanoma may be entirely asymptomatic.
Regional spread can produce nearby skin or subcutaneous deposits, termed satellites or in-transit metastases according to their relationship to the primary site and regional nodes. Enlarged lymph nodes may indicate nodal involvement, although microscopic nodal disease causes no palpable enlargement. Distant manifestations depend on the affected organ and can include respiratory, neurological, skeletal, or abdominal symptoms rather than changes in the original skin lesion.
Diagnostic assessment and pathological examination
Melanoma diagnosis combines clinical assessment with tissue examination. Evaluation typically documents lesion evolution, previous skin cancers, family history, ultraviolet exposure, and the appearance of the wider skin surface. Dermoscopy uses magnification and controlled illumination to reveal structures not readily visible to the unaided eye. Irregular pigment networks, atypical dots and globules, blue-white structures, and abnormal vascular patterns can increase diagnostic suspicion, but no single feature is universally present.
Digital photography and sequential dermoscopic imaging can document change in selected populations, particularly people with numerous atypical nevi. Reflectance confocal microscopy provides cellular-level images of superficial skin and is used in some specialist settings, including assessment of equivocal lesions on cosmetically sensitive sites. These methods are adjuncts. A definitive diagnosis ordinarily depends on a biopsy interpreted by a pathologist rather than on an image or clinical scoring rule alone.
For a suspected cutaneous melanoma, complete removal of the visible lesion with a narrow clinical margin is a common diagnostic approach because it permits assessment of architecture and maximum depth. Partial sampling may be used for very large lesions or anatomically difficult sites, but it can miss the most invasive region. The effect of the biopsy technique depends on whether adequate representative tissue is obtained; a superficial specimen that transects the tumor can prevent accurate thickness measurement.
Microscopic assessment considers asymmetry, poor circumscription, abnormal distribution of melanocytes, upward spread within the epidermis, cellular atypia, mitotic activity, and invasion into the dermis. No single microscopic feature establishes every case. Immunohistochemical markers such as SOX10, S100, Melan-A, and HMB45 help establish melanocytic differentiation. PRAME staining and selected molecular tests may assist in difficult cases, but they are interpreted alongside morphology rather than as independent verdicts.
The histopathology report records features relevant to staging and treatment. Breslow thickness measures the vertical depth of an invasive melanoma in millimeters, usually from the top of the epidermal granular layer, or the base of an ulcer, to the deepest invasive tumor cell. Ulceration means pathological loss of the epidermis overlying the tumor and has prognostic significance. Mitotic rate, microsatellites, lymphovascular invasion, neurotropism, and margin status provide additional information. Traumatic surface damage must be distinguished from genuine tumor-associated ulceration.
The differential diagnosis includes benign and atypical nevi, seborrheic keratosis, pigmented basal cell carcinoma, vascular lesions, dermatofibroma, and other malignancies. Some melanocytic tumors remain diagnostically challenging even after specialist review. Molecular profiling has a separate role when identifying treatment-relevant alterations, particularly BRAF V600 variants in advanced or selected high-risk disease. Such testing characterizes therapeutic targets and should not be confused with inherited genetic testing.
Staging, routes of spread, and prognosis
Cutaneous melanoma is commonly classified using the eighth edition of the American Joint Committee on Cancer TNM system. In this cancer staging framework, T describes the primary tumor, N describes regional nodal and certain regional non-nodal metastases, and M describes distant metastasis. Melanoma in situ is stage 0; invasive disease confined to the primary site is stage I or II; regional metastatic disease is stage III; and distant metastatic disease is stage IV. The detailed stage groups combine several pathological and clinical features rather than tumor diameter alone.
Breslow thickness is central to the T category. Major thickness boundaries are 1, 2, and 4 millimeters, with ulceration further separating categories. For thin tumors, the 0.8-millimeter threshold helps distinguish T1 subgroups: T1a comprises tumors below 0.8 millimeters without ulceration, while T1b includes tumors below 0.8 millimeters with ulceration and tumors from 0.8 through 1.0 millimeter with or without ulceration. Clark level, an older description of anatomical invasion, is no longer a principal determinant of modern T staging.
Regional staging may involve sentinel lymph node biopsy. A tracer identifies the first draining node or nodes, which are removed and examined for microscopic tumor. This procedure provides prognostic information and can identify stage III disease that is not apparent on examination or imaging. Its use depends on primary tumor characteristics and clinical context. A negative result reduces the likelihood of regional nodal involvement but cannot guarantee that future metastasis will not occur.
Melanoma can spread through the lymphatic system to regional nodes or through the bloodstream to distant organs. Common distant sites include skin and soft tissue, lung, liver, bone, and brain. In stage IV classification, the anatomical site of metastasis and serum lactate dehydrogenase contribute to risk assessment; central nervous system involvement has a distinct category. Computed tomography, positron emission tomography, and magnetic resonance imaging are used selectively to evaluate spread. Their yield is much lower in asymptomatic, low-stage disease than in advanced disease.
Prognosis varies greatly within the term melanoma. A thin localized tumor and a tumor involving multiple organs have very different expected courses. Population registries generally report five-year relative survival above 99% for localized cutaneous melanoma, substantially lower survival for regional disease, and lower still for distant disease. Such broad categories are not interchangeable with detailed AJCC stage groups, and figures vary by country, calendar period, and access to treatment.
Historical survival estimates can underestimate outcomes achieved with newer systemic therapies. Conversely, clinical-trial results may not apply directly to every patient because trial eligibility, organ function, disease burden, and prior treatment differ. Recurrence may occur years after initial treatment, including occasionally after a long disease-free interval, reflecting the capacity of disseminated melanoma cells to persist before becoming clinically detectable.
Treatment of localized and regionally advanced disease
Surgery is the principal treatment for localized cutaneous melanoma. After diagnostic biopsy, wide local excision removes the primary site and a surrounding margin of clinically normal skin. Commonly used margins for invasive tumors range from 1 to 2 centimeters according to Breslow thickness, while melanoma in situ is generally excised with smaller margins. Actual microscopic extension, anatomical constraints, and histological subtype affect how clearance is achieved. Randomized trials of excision margins helped establish that routinely removing very large amounts of normal skin is unnecessary.
Lentigo maligna and other lesions with poorly defined borders can require specialized margin assessment. Staged excision or selected micrographic approaches are used in some centers, particularly on the head and neck. Reconstruction may involve direct closure, a skin graft, or a flap. Surgical planning balances adequate tumor removal with preservation of function; the extent of reconstruction does not itself indicate the biological aggressiveness of the melanoma.
Management of regional nodes changed substantially following randomized evidence. The MSLT-II trial, reported in 2017, found that immediate completion lymph-node dissection after a positive sentinel node did not improve melanoma-specific survival compared with nodal observation using ultrasound, although it improved regional disease control. Complete dissection also caused more lymphedema. Consequently, a positive sentinel node no longer automatically leads to removal of all nodes in that basin. Clinically apparent nodal disease is a different situation and may require surgery, systemic treatment, or a combined approach.
Adjuvant therapy is treatment delivered after complete resection to reduce recurrence risk. Anti-PD-1 drugs have demonstrated benefit in resected stage III melanoma and, subsequently, in selected stage IIB and IIC disease. For resected BRAF V600-mutant stage III melanoma, combined BRAF and MEK inhibition is another established option. The relevant trial outcomes often include recurrence-free and distant metastasis-free survival; these endpoints are related to, but not identical with, overall survival.
Neoadjuvant therapy places systemic treatment before planned surgery. Trials in resectable, clinically detectable stage III melanoma have shown that preoperative immunotherapy can produce substantial pathological responses and improve event-free outcomes in defined settings. This approach can expose the intact tumor and its antigens to an activated immune response before resection. Its implementation requires coordination among surgical, medical, radiological, and pathological specialists, and it does not apply uniformly to every localized melanoma.
Radiation therapy has a selective role. It can improve local control in circumstances such as certain desmoplastic tumors, unresectable lesions, or high-risk regional disease, but it is not a routine substitute for excision of a typical operable primary melanoma. In-transit metastases may be treated with combinations of surgery, systemic therapy, intralesional therapy, or regional limb-directed treatment, depending on their distribution and whether disease exists elsewhere.
Systemic treatment of unresectable and metastatic melanoma
Modern systemic treatment rests principally on cancer immunotherapy and molecularly targeted cancer therapy. Before these approaches became available, metastatic melanoma was commonly treated with dacarbazine or related chemotherapy, and durable disease control was uncommon. High-dose interleukin-2 produced long-lasting responses in a small minority but required intensive monitoring because of substantial toxicity. The introduction of ipilimumab and vemurafenib in 2011 marked a major change in the treatment landscape.
Immune checkpoint inhibitors release inhibitory controls on antitumor immune responses. Ipilimumab targets CTLA-4, while nivolumab and pembrolizumab target PD-1. Nivolumab combined with ipilimumab can produce durable responses but causes more severe immune-related toxicity than anti-PD-1 treatment alone. Nivolumab combined with the LAG-3 inhibitor relatlimab is another checkpoint regimen. Long-term follow-up of major trials demonstrates that a proportion of patients with advanced melanoma remain alive for many years, a result that was uncommon with historical chemotherapy.
Checkpoint inhibition can cause immune-mediated inflammation in normal organs. Important toxicities include colitis, hepatitis, pneumonitis, skin inflammation, and endocrine disorders involving the thyroid, pituitary, or adrenal glands. Some endocrine effects are permanent. Treatment selection therefore involves both expected antitumor activity and the possibility of serious toxicity. Tumor PD-L1 staining does not serve as a universal gatekeeper for checkpoint treatment in melanoma, and no single biomarker reliably predicts benefit or harm in every patient.
For tumors with an actionable BRAF V600 alteration, combined BRAF and MEK inhibition blocks successive components of a growth-signaling pathway. Established pairs include dabrafenib–trametinib, encorafenib–binimetinib, and vemurafenib–cobimetinib. These combinations often shrink tumors rapidly and are more effective than BRAF inhibition alone, but acquired resistance remains common. Adverse effects vary by regimen and can include fever, skin reactions, ocular effects, and cardiac dysfunction. Treatment sequencing depends on clinical circumstances; a mutation result alone does not settle every therapeutic decision.
Additional approaches address particular disease patterns or treatment-resistant disease. Talimogene laherparepvec, an oncolytic virus therapy, is injected into accessible lesions and is used for selected unresectable cutaneous, subcutaneous, or nodal disease. Tumor-infiltrating lymphocyte therapy expands a patient’s tumor-reactive immune cells outside the body before reinfusion. In 2024, the United States granted accelerated approval to lifileucel for a defined previously treated advanced-melanoma population. Its administration includes lymphodepleting treatment and interleukin-2 support and requires specialized infrastructure.
Brain metastases may be treated with systemic therapy, stereotactic radiation, surgery, or combinations of these methods. Symptoms, lesion size and number, corticosteroid requirements, and extracranial disease influence the approach. Palliative care addresses symptoms and quality of life alongside antitumor treatment. Drug resistance, organ-specific progression, and unequal access to specialist treatments remain important limits on the gains achieved in clinical trials.
Mucosal, uveal, and other uncommon melanomas
Mucosal melanoma arises from melanocytes in mucous membranes, most frequently in the head and neck, anorectal region, and female genital tract. It accounts for a small minority of melanomas and is not strongly associated with ultraviolet exposure. Because the primary sites may be difficult to inspect and early symptoms are often nonspecific, diagnosis frequently occurs after local advancement or spread. Presentation can include nasal obstruction or bleeding, a mass, anorectal bleeding, or genital symptoms, depending on the site.
Its molecular profile differs from that of ordinary sun-exposed cutaneous melanoma. Mucosal tumors generally have a lower burden of ultraviolet-associated point mutations and a greater contribution from structural chromosome changes. Activating KIT alterations occur in a subset and can identify tumors potentially responsive to KIT-directed drugs; the specific alteration matters. BRAF V600 alterations are less common than in cutaneous melanoma. Checkpoint inhibitors have activity, but average response rates in mucosal melanoma have generally been lower than in cutaneous melanoma in available studies.
Localized mucosal disease is often managed with surgery, sometimes combined with radiotherapy for local control. Anatomical constraints can make complete resection difficult, and radical surgery may carry substantial functional consequences. Different mucosal sites have different staging systems and treatment considerations. The cutaneous Breslow-based staging framework cannot simply be transferred to every mucosal tumor.
Uveal melanoma arises in the choroid, ciliary body, or iris. Choroidal melanoma is the most frequent form. Diagnosis commonly relies on specialist ophthalmic examination and ocular imaging rather than the skin-biopsy pathway. Treatment of the primary tumor may include plaque brachytherapy, proton-beam radiation, or removal of the eye, depending on tumor dimensions and location. Control of the ocular primary does not eliminate the possibility of later distant metastasis.
Uveal melanoma has a marked tendency to metastasize to the liver. Chromosome 3 loss, BAP1 inactivation, and certain gene-expression patterns are associated with higher metastatic risk. Its biology and generally lower sensitivity to conventional checkpoint treatment distinguish it sharply from cutaneous melanoma. Tebentafusp, a bispecific T-cell receptor–based treatment, improved overall survival in a randomized trial and was approved in 2022 for unresectable or metastatic uveal melanoma in adults with the HLA-A*02:01 type. Its eligibility restriction illustrates how treatment can depend on both tumor biology and the patient’s antigen-presentation machinery.
Conjunctival melanoma is separate from uveal melanoma and shares some molecular features with cutaneous disease. Very rare primary melanomas also occur in the central nervous system. In addition, metastatic melanoma is sometimes diagnosed without an identifiable primary site. Such cases may reflect regression of a cutaneous primary or an occult lesion; absence of a visible primary does not mean that metastatic deposits constitute a different non-melanocytic cancer.
Prevention, surveillance, and developing diagnostic approaches
Primary prevention focuses on reducing carcinogenic ultraviolet exposure. Public-health measures include restrictions on commercial tanning, shade provision, occupational exposure controls, protective clothing, and sunscreen education. Sunscreen reduces ultraviolet exposure but is only one component of protection. Follow-up of a randomized sunscreen trial in Queensland, Australia, reported fewer melanomas among participants assigned to regular sunscreen use, although melanoma was not the trial’s original primary endpoint and the number of events was limited.
The evidence for population-wide melanoma screening is more complex than the evidence that ultraviolet radiation causes skin cancer. Detecting a thinner melanoma is associated with better prognosis, but this association alone does not prove that organized screening reduces mortality. Lead-time bias, preferential detection of slower-growing lesions, and overdiagnosis can make screening appear more beneficial than it is. Evidence assessments therefore distinguish routine examination of an asymptomatic general population from targeted surveillance of people with previous melanoma, strong familial susceptibility, or numerous atypical nevi.
After treatment, surveillance addresses two different risks: recurrence of the original melanoma and development of another primary melanoma. Clinical examination can assess the skin, surgical site, and regional lymph nodes. Imaging and laboratory investigations are used according to stage, symptoms, treatment history, and the likelihood that detecting recurrence will alter management. More testing is not automatically more informative: false-positive findings can lead to additional procedures, while repeated imaging carries costs and, for some modalities, radiation exposure.
Follow-up also captures long-term treatment effects. Lymph-node surgery can cause chronic lymphedema, extensive surgery can affect appearance or function, and immune-related endocrine disorders may persist after immunotherapy ends. Fear of recurrence and the burden of repeated examinations are relevant to health-related quality of life. These outcomes are increasingly measured alongside recurrence and survival in treatment studies, rather than being treated as interchangeable with tumor response.
Several technologies are under investigation for risk assessment and monitoring. Artificial intelligence systems can classify skin images, but performance depends on image quality, skin-tone representation, clinical setting, and the spectrum of lesions included in testing. Accuracy in a curated image collection does not establish effectiveness in routine care. Prospective evaluation must also consider missed cancers, unnecessary biopsies, workflow, and the consequences of automated reassurance.
Circulating tumor DNA is being studied as a marker of residual disease and recurrence, particularly after treatment of high-risk melanoma. Commercial gene-expression tests also attempt to refine prognosis, but prognostic association does not by itself establish that test-guided treatment improves outcomes. Investigational personalized cancer vaccines seek to stimulate responses against tumor-specific neoantigens, often in combination with checkpoint inhibition. These approaches are assessed through clinical trials using endpoints such as recurrence-free survival, overall survival, toxicity, and patient-reported functioning.