Skip to content

Melanoma red flags and safety concerns

TL;DR — Three categories of harm dominate. First, lesions that defeat the standard visual rules: amelanotic melanoma, where dermoscopy sensitivity falls to 61% (95% CI 37–81) and reflectance confocal microscopy to 67% (51–81) against far higher figures for pigmented lesions (Lan 2020, PMID 31747045); nodular melanoma, which grows at a median 0.49 mm/month against 0.12 mm/month for superficial spreading melanoma (Liu 2006, PMID 17178980); and nail-unit and acral lesions, where the commonest mimic — subungual haemorrhage — shares dermoscopic features with melanoma (Mun 2013, PMID 23302009). Second, immunotherapy emergencies: roughly 40% of checkpoint-treated patients experience an immune-related adverse event, pulmonary toxicity carries a threefold increased risk of requiring intensive care, and endocrine toxicities occur in 25–50% of recipients, ranging from asymptomatic subclinical hypothyroidism to fatal adrenal crisis or diabetic ketoacidosis (Li 2025, PMID 40145913; Wright 2023, PMID 36481794). Third, delayed and permanent toxicity: 43.2% of adjuvant anti-PD-1 recipients had an event persisting beyond 12 weeks after cessation, mostly grade 1–2 and mostly unresolved at last follow-up (Patrinely 2021, PMID 33764387). This is a research knowledge base, not medical advice.

Lesions that break the rules

Presentation Why it is missed Evidence
Amelanotic / hypomelanotic melanoma Lacks the pigment every visual algorithm depends on Pooled dermoscopy sensitivity 61% (37–81) and specificity 90% (74–97); RCM sensitivity 67% (51–81), specificity 89% (86–92) across 7 studies and 1,111 lesions (Lan 2020, PMID 31747045). Presentations classified as erythematous macule/patch on sun-exposed skin, dermal plaque or nodule without epidermal change, and exophytic nodule (Gualandri 2009, PMID 19207640)
Nodular melanoma Fast-growing, often clinically unremarkable, frequently lacks the asymmetry and colour variation the ABCD rule detects Median growth 0.49 mm/month vs 0.12 for superficial spreading; one third of all melanomas grow ≥0.5 mm/month (Liu 2006, PMID 17178980). Thin nodular melanoma's clinical presentation is often unremarkable and dermoscopic features are more suggestive of malignancy than clinical ones (Kalkhoran 2010, PMID 20231503). Against T1 superficial spreading melanoma, T1 nodular melanoma is less likely to show regression (OR 0.46, 0.29–0.72) and more likely to have mitoses (OR 1.97, 1.33–2.93) and regional metastasis (OR 1.77, 1.02–3.05) (Dessinioti 2019, PMID 30863861)
Nail-unit melanoma About two-thirds present as longitudinal melanonychia, which has a broad benign differential; adult dermoscopic parameters are unvalidated in children Conway 2023, PMID 36980308
Subungual haemorrhage as the mimic Overlapping features: across 90 lesions in 64 patients, 84% showed more than one colour, 92% a homogeneous pattern, 42% globular patterns, 39% streaks, 54% peripheral fading and 16% nail-plate destruction Mun 2013, PMID 23302009
Acral melanoma Amelanotic acral melanoma is diagnosed later than pigmented melanoma; acral lentiginous histotype carries independently worse disease-free survival after adjustment (adjusted HR 1.25, 1.02–1.52) Wu 2024, PMID 37690705; Mandalà 2022, PMID 35421840
Scalp and neck primaries Concealed by hair; carry independently worse melanoma-specific survival (adjusted HR 1.28, 1.16–1.41) and higher odds of lung (aOR 2.39) and multisite (aOR 2.23) metastasis than other head-and-neck sites Rashid 2026, PMID 41803580
Melanoma in skin of colour Later stage at diagnosis, different anatomical distribution and dermatoscopic patterns; acral and mucosal subtypes over-represented Brunsgaard 2023, PMID 35533771; Mangione 2023, PMID 37071089

Fast-growing melanomas occur disproportionately in people the risk models would not flag: rapid growth was more frequent in men (geometric mean ratio 1.7), those ≥70 (2.8), and those with fewer nevi (<50, GMR 2.0) and fewer freckles (2.5) (PMID 17178980).

Diagnostic-process failures

  • Positive predictive value of a clinically suspicious lesion is 6.4% (95% CI 5.5–7.4) — 16 concerning lesions biopsied per melanoma found in one academic centre's year of practice (Soltani-Arabshahi 2015, PMID 25582536). This cuts both ways: most excised lesions are benign, and a low threshold is what makes the yield low.
  • Pathology in the diagnostic middle is unreliable. Accuracy against expert consensus was 25% for moderate atypia, 40% for severe atypia/in situ and 43% for pT1a melanoma; at population level 8.0% of melanocytic biopsies are overinterpreted and 9.2% underinterpreted (Elmore 2017, PMID 28659278). Second opinions reduce misclassification from 52.8% to 36.7% at best, and in situ and thin invasive melanoma remain intractable (Piepkorn 2019, PMID 31603483).
  • Partial biopsy can understage. Deep margins were positive in 33% of shave and 23% of punch biopsies with residual melanoma at excision in 40.6% and 60.4% respectively; margin or sentinel-node recommendations changed in 6–9% of cases (Jones 2023, PMID 36739830).
  • Referral pathways have a modest yield. The rate of positive skin-cancer diagnosis and its stage in two-week-wait referrals in England has been analysed by age (Keith 2017, PMID 28044351).
  • Delay is not the main determinant of thickness. In 3,772 Queenslanders, there was no significant association between melanoma thickness and reported time to diagnosis except for post-presentation delay in physician-detected nodular melanoma (Baade 2006, PMID 17116832) — the corollary being that thickness is driven mainly by growth rate, so a fast-growing tumour can present thick without any delay.

Immunotherapy emergencies

Syndrome Recognition points
Pneumonitis Approximately 40% of checkpoint-treated patients experience an immune-related event; pulmonary toxicity can be rapidly progressive and potentially fatal, and confers a threefold increased risk of requiring intensive-care-level care. Onset generally 6–12 weeks after initiation but can develop within days of the first dose; clinical manifestations are highly variable (Li 2025, PMID 40145913). Anti-PD-(L)1 fatalities are more often pneumonitis than colitis (Wang 2018, PMID 30242316)
Colitis The commonest any-grade event in randomised-trial meta-analysis (14.5%), with combination therapy raising the risk (RR 3.56, 1.56–8.12) (Da 2019, PMID 32082164). Anti-CTLA-4 fatalities were usually colitis — 135 of 193 (70%) (PMID 30242316). Real-world management shows gaps in guideline adherence (PMID 41902969)
Endocrine emergencies Endocrine toxicities occur in 25–50% of recipients depending on regimen, spanning asymptomatic subclinical hypothyroidism to fatal adrenal crisis, thyroid dysfunction or diabetic ketoacidosis (Wright 2023, PMID 36481794). Hypophysitis with secondary adrenal insufficiency can present as hyponatraemia and be mistaken for SIADH (PMID 42359175; PMID 42528939). Myxoedema crisis has been reported (PMID 28239466). The ESE guideline requires baseline endocrine testing before each treatment cycle (Husebye 2022, PMID 36149449)
Myocarditis and neuromuscular overlap Incidence of checkpoint-inhibitor myotoxicity at 6 months was 0.7–0.9% in a French national cohort of 172,363 treated adults (Salem 2026, PMID 40884033); driven by α-myosin-specific T cells (Axelrod 2022, PMID 36385524). Myocarditis–myositis–myasthenia gravis overlap is a recognised and frequently fatal syndrome (PMID 39202282; PMID 42384108)
Emergency-department recognition generally Dedicated reviews and clinical guidance for emergency physicians exist precisely because these presentations are non-specific and lead to inappropriate discharge and revisits (Hryniewicki 2018, PMID 30120013; Long 2020, PMID 32421502)

Two features make these emergencies distinctive. They can occur after treatment has stopped, and they are treated with immunosuppression rather than with drug withdrawal alone. ASCO's guidance is that therapy generally continues with close monitoring for grade 1 toxicity, except for some neurological, haematological and cardiac events (Schneider 2021, PMID 34724392) — and those exceptions are exactly the fatal ones.

Delayed and permanent harm

  • Chronic immune-related events persisted beyond 12 weeks after adjuvant anti-PD-1 cessation in 43.2% of 387 patients, 96.4% grade 1–2, most unresolved at last follow-up; one fatal myocarditis and one fatal neurotoxicity occurred in the same cohort (Patrinely 2021, PMID 33764387). Extended follow-up to a median beyond 18 months confirms the pattern (Goodman 2023, PMID 37535354).
  • Endocrine replacement is usually lifelong, and high-dose glucocorticoids do not prevent it except possibly in severe thyroid eye disease and hypophysitis affecting vision (PMID 36149449).
  • High corticosteroid peak dose for treatment-related events was associated with worse progression-free survival across six trials' anti-PD-1 + anti-CTLA-4 arms (Verheijden 2024, PMID 39110922) — so treating toxicity aggressively is not without cost.
  • Lymphoedema occurred in 24.1% of patients randomised to completion lymph-node dissection in MSLT-II (Faries 2017, PMID 28591523).

Populations needing extra caution

Group Concern
Solid organ transplant recipients Checkpoint inhibitors carry 36.2% one-year acute rejection and 18.4% graft loss against 31.6% response (Saleem 2025, PMID 40545616) — see special populations
Pregnancy Melanoma is the commonest cancer to metastasise to placenta and fetus: 27 of 87 reported placental/fetal metastases, with the fetus affected in 6 of 27 and 5 of 6 infants dying (Alexander 2003, PMID 12775744). Placental examination after delivery in metastatic disease is a defined action
Children Sentinel-node positivity 46% with 7% fatality; a positive node in an atypical spitzoid tumour predicted nothing in prospective follow-up (Hawryluk 2024, PMID 38040338; Gassenmaier 2022, PMID 35104769)
Older adults with comorbidity Age was not associated with grade ≥3 events but comorbidity count was (multivariable OR 1.83, 0.99–3.40) (Özkan 2024, PMID 39368226)
Heritable retinoblastoma survivors 50-year cumulative melanoma incidence 4.5% vs 0.7% in non-heritable survivors, with skin cancers arising ~20 years earlier (Kleinerman 2021, PMID 34153328)
Symptomatic brain metastases Intracranial clinical benefit 16.7% vs 57.4% in asymptomatic patients; 36-month overall survival 36.6% vs 71.9% (Tawbi 2021, PMID 34774225)

Prevention-side cautions

  • Indoor tanning raises melanoma risk (RR 1.27, 1.16–1.39; 1.75, 1.14–2.69 for early-onset disease), with higher risk for first exposure at ≤20 years and for ≥10 sessions per year (An 2021, PMID 34885049). IARC classified artificial tanning devices as carcinogenic in 2009 (Gandini 2019, PMID 30811691).
  • Behavioural counselling changes behaviour but has not consistently reduced sunburn, and melanoma outcomes are essentially unmeasured (Henrikson 2018, PMID 29558557).
  • Overdiagnosis is a harm, not only a statistical artefact: an estimated 59–60% of US melanoma diagnoses in White patients in 2014 (Adamson 2022, PMID 35293957), with consequences argued rather than measured (Kutzner 2020, PMID 32841508) — see screening and overdiagnosis.

Interpretation rules for this page

  • Pigment-dependent tools are less sensitive in amelanotic disease; pooled sensitivity was 61% for dermoscopy and 67% for reflectance confocal microscopy, despite specificity of 90% and 89%, respectively (PMID 31747045).
  • Growth rate, not delay, drives thickness at presentation in population data (PMID 17178980; PMID 17116832).
  • A negative dermoscopic impression does not exclude nodular melanoma (PMID 20231503; PMID 30863861).
  • Immunotherapy toxicity can begin days after the first dose or months after the last (PMID 40145913; PMID 33764387).
  • The grade of an immune-related event does not predict its permanence (PMID 33764387).
  • Treat the exceptions as the rule in cardiac, neurological and haematological toxicity — those are where "continue with monitoring" does not apply (PMID 34724392).
  • Nothing here is clinical guidance; it is a summary of published safety literature.

Open questions

  • Can any tool raise amelanotic-melanoma sensitivity above the ~60–67% ceiling (PMID 31747045)?
  • Should nodular melanoma have a separate detection pathway given that its growth rate makes interval detection structurally difficult (PMID 17178980)?
  • What proportion of chronic grade 1–2 immune events ever resolve beyond five years (PMID 37535354)?
  • Does aggressive corticosteroid treatment of immune-related events causally reduce antitumour efficacy (PMID 39110922)?
  • How should melanoma surveillance pathways identify heritable retinoblastoma survivors, who are not on standard germline panels (PMID 34153328; PMID 42201696)?
  • What are the measurable harms of melanoma overdiagnosis — insurance, employment, psychological — which are currently argued rather than quantified (PMID 32841508; PMID 42279200)?

References

  1. Lan J, et al. The diagnostic accuracy of dermoscopy and reflectance confocal microscopy for amelanotic/hypomelanotic melanoma: a systematic review and meta-analysis. The British journal of dermatology. 2020;183:210-219. PMID 31747045
  2. Liu W, et al. Rate of growth in melanomas: characteristics and associations of rapidly growing melanomas. Archives of dermatology. 2006;142:1551-8. PMID 17178980
  3. Mun JH, et al. Dermoscopy of subungual haemorrhage: its usefulness in differential diagnosis from nail-unit melanoma. The British journal of dermatology. 2013;168:1224-9. PMID 23302009
  4. Li C, et al. Immune Checkpoint Inhibitor-associated Pneumonitis: A Narrative Review. The western journal of emergency medicine. 2025;26:210-218. PMID 40145913
  5. Wright JJ, et al. Approach to the Patient With Immune Checkpoint Inhibitor-Associated Endocrine Dysfunction. The Journal of clinical endocrinology and metabolism. 2023;108:1514-1525. PMID 36481794
  6. Patrinely JR Jr, et al. Chronic Immune-Related Adverse Events Following Adjuvant Anti-PD-1 Therapy for High-risk Resected Melanoma. JAMA oncology. 2021;7:744-748. PMID 33764387
  7. Gualandri L, et al. Clinical features of 36 cases of amelanotic melanomas and considerations about the relationship between histologic subtypes and diagnostic delay. Journal of the European Academy of Dermatology and Venereology : JEADV. 2009;23:283-7. PMID 19207640
  8. Kalkhoran S, et al. Historical, clinical, and dermoscopic characteristics of thin nodular melanoma. Archives of dermatology. 2010;146:311-8. PMID 20231503
  9. Dessinioti C, et al. Distinct Clinicopathological and Prognostic Features of Thin Nodular Primary Melanomas: An International Study from 17 Centers. Journal of the National Cancer Institute. 2019;111:1314-1322. PMID 30863861
  10. Conway J, et al. Adult and Pediatric Nail Unit Melanoma: Epidemiology, Diagnosis, and Treatment. Cells. 2023;12. PMID 36980308
  11. Wu Q, et al. Clinicopathologic features, delayed diagnosis, and survival in amelanotic acral melanoma: A comparative study with pigmented melanoma. Journal of the American Academy of Dermatology. 2024;90:369-372. PMID 37690705
  12. Mandalà M, et al. Acral lentiginous melanoma histotype predicts outcome in clinical stage I-II melanoma patients: an International multicenter study. ESMO open. 2022;7:100469. PMID 35421840
  13. Rashid S, et al. Prognostic Implications of Primary Site in Cutaneous Head and Neck Melanoma After the Implementation of Sentinel Node Biopsy: A SEER-Based Analysis (2011-2020). Annals of surgical oncology. 2026;33:5002-5008. PMID 41803580
  14. Brunsgaard EK, et al. Melanoma in skin of color: Part I. Epidemiology and clinical presentation. Journal of the American Academy of Dermatology. 2023;89:445-456. PMID 35533771
  15. Mangione CM, et al. Screening for Skin Cancer: US Preventive Services Task Force Recommendation Statement. JAMA. 2023;329:1290-1295. PMID 37071089
  16. Soltani-Arabshahi R, et al. Predictive value of biopsy specimens suspicious for melanoma: support for 6-mm criterion in the ABCD rule. Journal of the American Academy of Dermatology. 2015;72:412-8. PMID 25582536
  17. Elmore JG, et al. Pathologists' diagnosis of invasive melanoma and melanocytic proliferations: observer accuracy and reproducibility study. BMJ (Clinical research ed.). 2017;357:j2813. PMID 28659278
  18. Piepkorn MW, et al. Assessment of Second-Opinion Strategies for Diagnoses of Cutaneous Melanocytic Lesions. JAMA network open. 2019;2:e1912597. PMID 31603483
  19. Jones S, et al. Clinical Impact and Accuracy of Shave Biopsy for Initial Diagnosis of Cutaneous Melanoma. The Journal of surgical research. 2023;286:35-40. PMID 36739830
  20. Keith DJ, et al. Rate of positive diagnosis of skin cancer and its stage in two-week wait referrals in England according to age. Clinical and experimental dermatology. 2017;42:145-152. PMID 28044351
  21. Baade PD, et al. The relationship between melanoma thickness and time to diagnosis in a large population-based study. Archives of dermatology. 2006;142:1422-7. PMID 17116832
  22. Wang DY, et al. Fatal Toxic Effects Associated With Immune Checkpoint Inhibitors: A Systematic Review and Meta-analysis. JAMA oncology. 2018;4:1721-1728. PMID 30242316
  23. Da L, et al. Organ-Specific Immune-Related Adverse Events Associated With Immune Checkpoint Inhibitor Monotherapy Versus Combination Therapy in Cancer: A Meta-Analysis of Randomized Controlled Trials. Frontiers in pharmacology. 2019;10:1671. PMID 32082164
  24. Alekhina N, et al. Real-world management and outcomes of immune-mediated diarrhea and colitis: gaps in guideline adherence and opportunities for implementation. Supportive care in cancer : official journal of the Multinational Association of Supportive Care in Cancer. 2026;34. PMID 41902969
  25. Al Zein S. Immune Checkpoint Inhibitor-Induced Hypophysitis Presenting as Severe Hyponatremia: A Nephrology Perspective on Avoiding the Syndrome of Inappropriate Antidiuretic Hormone (SIADH) Pitfall. Cureus. 2026;18:e109633. PMID 42359175
  26. Thu AN, et al. Life-Threatening Secondary Adrenal Crisis Triggered by Pembrolizumab-Induced Hypophysitis: A Case Report. Cureus. 2026;18:e111692. PMID 42528939
  27. Khan U, et al. Nivolumab induced myxedema crisis. Journal for immunotherapy of cancer. 2017;5:13. PMID 28239466
  28. Husebye ES, et al. Endocrine-related adverse conditions in patients receiving immune checkpoint inhibition: an ESE clinical practice guideline. European journal of endocrinology. 2022;187:G1-G21. PMID 36149449
  29. Salem JE, et al. Incidence and risk factors of immune checkpoint inhibitor myocardial and muscle toxicity: a French nationwide study. European heart journal. 2026;47:1014-1030. PMID 40884033
  30. Axelrod ML, et al. T cells specific for α-myosin drive immunotherapy-related myocarditis. Nature. 2022;611:818-826. PMID 36385524
  31. Lipe DN, et al. Myocarditis, Myositis, and Myasthenia Gravis Overlap Syndrome Associated with Immune Checkpoint Inhibitors: A Systematic Review. Diagnostics (Basel, Switzerland). 2024;14. PMID 39202282
  32. Allen WM, et al. Immune checkpoint inhibitor-induced myasthenia gravis and myocarditis: a fatal immune-related adverse event. Immunologic research. 2026;74. PMID 42384108
  33. Hryniewicki AT, et al. Management of Immune Checkpoint Inhibitor Toxicities: A Review and Clinical Guideline for Emergency Physicians. The Journal of emergency medicine. 2018;55:489-502. PMID 30120013
  34. Long B, et al. Oncologic Emergencies: Immune-Based Cancer Therapies and Complications. The western journal of emergency medicine. 2020;21:566-580. PMID 32421502
  35. Schneider BJ, et al. Management of Immune-Related Adverse Events in Patients Treated With Immune Checkpoint Inhibitor Therapy: ASCO Guideline Update. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2021;39:4073-4126. PMID 34724392
  36. Goodman RS, et al. Extended Follow-Up of Chronic Immune-Related Adverse Events Following Adjuvant Anti-PD-1 Therapy for High-Risk Resected Melanoma. JAMA network open. 2023;6:e2327145. PMID 37535354
  37. Verheijden RJ, et al. Corticosteroids for Immune-Related Adverse Events and Checkpoint Inhibitor Efficacy: Analysis of Six Clinical Trials. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2024;42:3713-3724. PMID 39110922
  38. Faries MB, et al. Completion Dissection or Observation for Sentinel-Node Metastasis in Melanoma. The New England journal of medicine. 2017;376:2211-2222. PMID 28591523
  39. Saleem N, et al. Outcomes of Solid Organ Transplant Recipients With Advanced Cancers Receiving Immune Checkpoint Inhibitors: A Systematic Review and Individual Participant Data Meta-Analysis. JAMA oncology. 2025;11:1150-1159. PMID 40545616
  40. Alexander A, et al. Metastatic melanoma in pregnancy: risk of transplacental metastases in the infant. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2003;21:2179-86. PMID 12775744
  41. Hawryluk EB, et al. Risk factors and outcomes of melanoma in children and adolescents: A retrospective multicenter study. Journal of the American Academy of Dermatology. 2024;90:716-726. PMID 38040338
  42. Gassenmaier M, et al. Diagnostic and prognostic classification of atypical spitzoid tumours based on histology and genomic aberrations: A prospective cohort study with long-term follow-up. European journal of cancer (Oxford, England : 1990). 2022;163:200-210. PMID 35104769
  43. Özkan A, et al. Adjuvant immunotherapy in older patients with stage III and resected stage IV melanoma: Toxicity and recurrence-free survival outcomes from the Dutch melanoma treatment registry. European journal of cancer (Oxford, England : 1990). 2024;212:115056. PMID 39368226
  44. Kleinerman RA, et al. Increased Risk of Skin Cancer in 1,851 Long-Term Retinoblastoma Survivors. The Journal of investigative dermatology. 2021;141:2849-2857.e3. PMID 34153328
  45. Tawbi HA, et al. Long-term outcomes of patients with active melanoma brain metastases treated with combination nivolumab plus ipilimumab (CheckMate 204): final results of an open-label, multicentre, phase 2 study. The Lancet. Oncology. 2021;22:1692-1704. PMID 34774225
  46. An S, et al. Indoor Tanning and the Risk of Overall and Early-Onset Melanoma and Non-Melanoma Skin Cancer: Systematic Review and Meta-Analysis. Cancers. 2021;13. PMID 34885049
  47. Gandini S, et al. Epidemiological evidence of carcinogenicity of sunbed use and of efficacy of preventive measures. Journal of the European Academy of Dermatology and Venereology : JEADV. 2019;33 Suppl 2:57-62. PMID 30811691
  48. Henrikson NB, et al. Behavioral Counseling for Skin Cancer Prevention: Evidence Report and Systematic Review for the US Preventive Services Task Force. JAMA. 2018;319:1143-1157. PMID 29558557
  49. Adamson AS, et al. Estimating Overdiagnosis of Melanoma Using Trends Among Black and White Patients in the US. JAMA dermatology. 2022;158:426-431. PMID 35293957
  50. Kutzner H, et al. Overdiagnosis of melanoma - causes, consequences and solutions. Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG. 2020;18:1236-1243. PMID 32841508
  51. Goldstein AM, et al. Prevalence of Familial Melanoma Genes and Cancer Risk Among Genomically Ascertained Individuals. JAMA dermatology. 2026;162:692-700. PMID 42201696
  52. Greco ME, et al. Overdiagnosis of Melanoma In Situ. Journal of clinical medicine. 2026;15. PMID 42279200