← OncoDesk·MedDesk·web edition — reference only, not medical advice
all sections · full handbook

13Melanoma (skin)

Snapshot. Early management is driven by ONE NUMBER — BRESLOW THICKNESS, which sets margin, sentinel-node indication and stage. The fork: RESECTABLE (surgery ± PERIOPERATIVE immunotherapy) vs UNRESECTABLE/METASTATIC (immunotherapy FIRST, BRAF/MEK held in reserve). Drivers: Breslow, ulceration, nodal burden, BRAF V600 status, LDH, subtypeacral, mucosal and uveal are biologically SEPARATE DISEASES, not variants.
Workup
EXCISIONAL biopsy, full-thickness (⚠ never shave/punch through the thickest part — destroys the Breslow measurement) · histology for Breslow (to 0.1 mm) + ulceration (the T-category; mitotic rate dropped from AJCC-8 T1) + subtype (superficial spreading/nodular/lentigo maligna/acral/mucosal/desmoplastic) · full skin + nodal exam · sentinel node biopsy — discuss/offer from Breslow >0.8 mm · BRAF V600 (stage III–IV) · LDH + CT/PET + brain MRI (stage IV — melanoma is the most brain-tropic solid tumour) · HLA-A*02:01 (uveal only) · KIT sequencing (acral/mucosal — testing BRAF alone misses the actionable target here).
Treatment by setting
SettingTreatment
IN SITU / STAGE IA
lentigo maligna, or T1a <0.8 mm non-ulcerated
WLE 0.5–1 cm wide local excision
No SLNB, no systemic therapy — surveillance only <5% sentinel-node positivity
STAGE IB–IIA
T1b–T3a, 0.8–2 mm
WLE 1–2 cm by depth
SLNB — discuss / offer at T1b, indicated from T2a
No approved adjuvant therapy at stage IIA — surveillance
STAGE IIB/IIC, node-negative
T3b–T4b, >2 mm, ulcerated or thick
WLE 2 cm + SLNB
Node-negative → adjuvant pembrolizumab ×1 year (KEYNOTE-716) RFS HR 0.65 — first anti-PD-1 to significantly improve RFS and distant-metastasis-free survival at stage II · adjuvant interferon is historical — do not offer
🔴 STAGE III, MICROSCOPIC
sentinel-node positive
⚠⚠ NO reflexive completion node dissection (MSLT-II) MSLT-II / DeCOG-SLT: no melanoma-specific survival benefit, real lymphoedema cost → nodal ultrasound surveillance instead
Adjuvant systemic ×12 months → nivolumab (CheckMate-238) or pembrolizumab (KEYNOTE-054)
BRAF V600E/K → dabrafenib + trametinib (COMBI-AD)
🔴 STAGE III, MACROSCOPIC, resectable
palpable nodal disease
NEOADJUVANT IMMUNOTHERAPY, not upfront surgery — the standard-of-care shift
Ipilimumab + nivolumab ×2 → response-adapted surgery ± adjuvant (NADINA) 12/24-mo EFS 85.2%/77.3% vs 61.7%/55.7%, HR 0.40; ~60% needed no further therapy after 2 cycles · cost: grade ≥3 AEs 29.7% vs 14.7%
Alternative → neoadjuvant pembrolizumab ×3 → surgery → adjuvant (SWOG S1801) 2-yr EFS 72% vs 49%
UNRESECTABLE III / STAGE IV
any BRAF status, first line
Immunotherapy first
Ipilimumab + nivolumab (CheckMate-067) 10-yr OS 43% vs 37% (nivo alone) vs 19% (ipi alone) — longest checkpoint-inhibitor follow-up in any tumour
Nivolumab + relatlimab — a less toxic doublet (RELATIVITY-047) mPFS 10.2 vs 4.6 mo; mature OS benefit at 3 yr (median 51.0 vs 34.1 mo, 3-yr OS 54.6% vs 48.0%) · reasonable frailer-patient alternative
🔴 STAGE IV, BRAF-mutant — sequencing
Immunotherapy FIRST, even in BRAF-mutant disease (DREAMseq) ipi+nivo → BRAF/MEK at progression gave 2-yr OS 72% vs 52% vs the reverse sequence
Reserve BRAF/MEK for AFTER immunotherapy except when rapid cytoreduction is needed for symptomatic / high-burden / high-LDH disease
BRAF/MEK regimens — encorafenib + binimetinib (COLUMBUS) · dabrafenib + trametinib (COMBI-v) · vemurafenib + cobimetinib (coBRIM) cross-trial comparison only, not head-to-head
🔴 BRAIN METASTASES (M1d)
Asymptomatic, steroid-free → systemic IO upfront (CheckMate-204) ipi+nivo in active brain mets — intracranial response 55%, sparing / deferring RT
Symptomatic → radiation oncology early; SRS standard IO benefit falls to 22% — for limited / symptomatic disease
ACRAL / MUCOSAL
a different disease
⚠ KIT is the actionable target here, not BRAF KIT mutations in up to 36% acral / 39% mucosal (BRAF/NRAS only ~10–15%)
Exon 11/13 KIT mutations → imatinib exon 17 or amplification do NOT respond
BRAF/MEK less effective in mucosal (NF1-loss resistance); anti-PD-1 activity lower in both; worse prognosis
UVEAL
HLA-A*02:01-positive, unresectable/metastatic
⚠ No BRAF, poor checkpoint response, HEPATOTROPIC — do NOT treat like cutaneous melanoma
Tebentafusp gp100×CD3 T-cell-receptor bispecific — median OS 21.6 vs 16.9 mo, HR 0.68 · HLA typing gatekeeps eligibility
Watch
⚠ THE FOUR TRAPS: shave-biopsying a suspicious lesion · reflex node dissection for a positive sentinel node · starting BRAF/MEK first in a fit BRAF-mutant patient (large OS penalty per DREAMseq) · sending macroscopic stage III straight to surgery without a neoadjuvant discussion.
Ipilimumab+nivolumab → grade ≥3 irAEs in ~a third: colitis, hepatitis, hypophysitis, thyroiditis, pneumonitis, myocarditis — needs a steroid/infliximab pathway. BRAF/MEK toxicity differs by combo — PYREXIA with dabrafenib/trametinib; PHOTOSENSITIVITY with vemurafenib/cobimetinib; also cutaneous SCC (paradoxical MAPK activation), reduced LVEF. Tebentafusp → cytokine release syndrome — step-up inpatient dosing.
Annual skin exam for life at every stage — second-primary risk is lifelong; stage IIB+ adds periodic CT/PET + brain MRI surveillance.
← OvarianHead & Neck →