Therapy — from Wide Excision to Lifileucel

1. The Therapeutic Landscape Today

Melanoma treatment is stratified by stage, mutational status, and bulk of disease:

The DREAMseq trial (Atkins et al., JCO 2023) randomised BRAF V600-mutant IV melanoma patients to anti-PD-1+anti-CTLA-4 first vs. BRAF+MEK first; the immunotherapy-first sequence had ~20-percentage-point higher 2-year OS (72% vs 52%). For BRAF V600-mutant stage IV, immunotherapy first is now standard outside specific clinical scenarios (rapidly progressive symptomatic visceral disease).

2. Surgical Excision & Margins

Wide local excision (WLE) remains curative for nearly all stage 0–II melanoma. Margins are tied to Breslow thickness:

The historical 5 cm margin (Handley 1907) has been steadily walked back. The MelMarT-2 trial (Jakub et al., due ~2025) is randomising 1 cm vs 2 cm margins for T2–T3 melanoma, anticipating non-inferiority of the narrower margin and substantially less morbidity.

3. Sentinel Lymph Node & MSLT-II Reset

The MSLT-II trial (Faries et al., NEJM 2017) was a paradigm-shifter: 1939 patients with positive SLN were randomised to immediate completion lymph-node dissection (CLND) vs. observation (with nodal-basin ultrasound surveillance). The result: melanoma-specific survival was equivalent at 3 years (86% vs 86%), but CLND produced ~25% rate of clinically significant lymphedema.

CLND has therefore been largely abandoned in favour of active nodal-basin ultrasound surveillance plus effective adjuvant systemic therapy. This is a textbook case of less-aggressive surgery enabled by more-effective systemic therapy — SLN biopsy continues, but its role is staging, not the gateway to a morbid completion dissection.

4. The DTIC Era — Why Cytotoxic Chemotherapy Failed

Until 2011 the standard of care for stage IV melanoma was dacarbazine (DTIC) — an imidazole-carboxamide alkylator approved in 1975. Response rates were ~10%, durable response <5%, and median overall survival ~6–9 months. Equivalent cytotoxic combinations (CVD, Dartmouth regimen) added toxicity without survival benefit.

The reasons melanoma is intrinsically chemotherapy-resistant include high baseline expression of P-glycoprotein, robust DNA-damage repair (paradoxical given the UV burden — melanoma cells are repair-competent for chemotherapy adducts despite not having repaired the original UV damage), and a high apoptotic threshold mediated by overexpressed Bcl-2 family. Effective treatment had to bypass chemotherapy altogether.

5. BRAF + MEK Inhibitor Combinations

The discovery of BRAF V600E in ~50% of melanomas (Davies et al., Nature 2002) opened the first targeted-therapy era. Vemurafenib (PLX4032), the first selective BRAF V600 inhibitor:

• BRIM-3 trial (Chapman et al., NEJM 2011): 675 untreated BRAF V600E+ stage IV melanoma randomised to vemurafenib vs DTIC. ORR 48% vs 5%; 6-month OS 84% vs 64%; HR for death 0.37.
• Median PFS as monotherapy ~6–7 months — resistance through MAPK reactivation is the rule.
• Class toxicity: cutaneous SCC and keratoacanthoma (paradoxical RAS activation in BRAF-WT keratinocytes), photosensitivity, arthralgia, fatigue.

Adding a MEK inhibitor downstream of BRAF blunts pathway reactivation and substantially extends PFS while reducing cutaneous side effects. Three approved combinations:

Resistance mechanisms to BRAF+MEK include NRAS Q61 acquisition, MEK1 P124 mutations, BRAF amplification or splice variants (e.g. p61BRAF), and MITF amplification. Median PFS ~12–15 months — striking activity, but still rarely curative as monotherapy. Hence the modern emphasis on combining or sequencing with checkpoint blockade.

Triplet therapy — BRAFi+MEKi+ anti-PD-1 (e.g. atezolizumab/spartalizumab in IMspire150 / COMBI-i): the trials showed a modest PFS gain but did not improve OS in unselected populations, and triplet therapy is not standard.

6. The Structural Basis of BRAFi Selectivity

Vemurafenib’s ~100-fold preference for BRAF V600E over wild-type BRAF rests on the V600E-stabilised conformation of the kinase domain. The PDB structure 3OG7 (Bollag et al., Nature 2010) shows the inhibitor lodged in the ATP- binding cleft of the active monomer:

7. Anti-CTLA-4 (Ipilimumab) — the First Checkpoint

CTLA-4 is an inhibitory receptor on activated T cells (and constitutively on regulatory T cells) that competes with CD28 for B7 ligands on antigen-presenting cells, raising the threshold for T-cell activation. James Allison’s group showed in 1996 (Leach et al., Science) that anti-CTLA-4 antibody could induce tumour rejection in mice; ipilimumab was the human therapeutic incarnation.

• Hodi et al., NEJM 2010. 676 stage IV melanoma patients randomised to ipilimumab vs gp100 peptide vaccine vs both: ipilimumab arm had median OS 10.1 vs 6.4 months — the first ever OS benefit in advanced melanoma.
• Long-term follow-up: ~22% of ipilimumab patients are alive at 10+ years — the “tail of the curve” that defines durable immunotherapy response.
• Toxicity is the cost: immune-related adverse events (irAEs) include colitis (~10%), hypophysitis (~5%), hepatitis, pneumonitis, dermatitis. ~10–15% of patients require systemic glucocorticoids.

Ipilimumab monotherapy (3 mg/kg q3w × 4 doses) is now largely supplanted by anti-PD-1, but the combination of ipi + nivo retains a key role.

8. Anti-PD-1 — Pembrolizumab & Nivolumab

PD-1 (programmed death-1) is an inhibitory receptor expressed on chronically antigen-stimulated T cells; it binds PD-L1 / PD-L2 on tumour cells and myeloid cells, recruits SHP-2 phosphatase, and dampens TCR signalling. Tasuku Honjo’s group identified PD-1 in 1992 and elucidated its role in peripheral tolerance and tumour escape.

Two anti-PD-1 antibodies dominate melanoma practice:

• Pembrolizumab (humanised IgG4 anti-PD-1). KEYNOTE-006 (Robert et al., NEJM 2015): pembro vs ipilimumab, 5-year OS 39% vs 31%, HR 0.73. KEYNOTE-054 (Eggermont et al., NEJM 2018): adjuvant pembro vs placebo in stage III, 1-yr RFS 75% vs 61%.
• Nivolumab (fully human IgG4 anti-PD-1). CheckMate-066 (Robert et al., NEJM 2015): nivo vs DTIC, 1-yr OS 73% vs 42%. CheckMate-238 (Weber et al., NEJM 2017): adjuvant nivo vs ipilimumab in resected stage IIIB-IV, 12-mo RFS 71% vs 61%.

Anti-PD-1 monotherapy has ORR ~40–45% in stage IV melanoma; ~50% of responders maintain response at 5 years. Toxicity is more favourable than ipilimumab: any-grade irAE ~70%, but grade 3+ in only ~10–15%.

9. PD-1 / PD-L1 Structural Biology

The PD-1:PD-L1 interaction surface is a shallow protein-protein interface, ~1700 Ų of buried surface, lacking deep pockets — classically “undruggable” for small molecules. Antibody therapeutics succeed because they can present a competing surface large enough to outcompete PD-L1 for the PD-1 binding face. The crystal structure 5IUS shows PD-1 in complex with PD-L1, illuminating the epitope that pembrolizumab and nivolumab block.

10. Ipilimumab + Nivolumab — Dual Checkpoint Blockade

The combination of anti-CTLA-4 + anti-PD-1 acts on non-redundant checkpoints: CTLA-4 controls priming and Treg suppression in lymph nodes; PD-1 controls effector-phase exhaustion in the tumour. Combination produces synergy — with synergistic toxicity to match.

• CheckMate-067 (Larkin et al., NEJM 2015; updated NEJM 2019, NEJM 2022 10-yr update): 945 untreated stage IV patients randomised 1:1:1 to nivo+ipi vs nivo vs ipi.
• 10-year OS: 43% (nivo+ipi) vs 37% (nivo) vs 19% (ipi).
• Median OS: not reached for nivo+ipi vs ~36 mo nivo vs ~19 mo ipi.
• Grade 3+ irAEs: nivo+ipi ~60%; nivo ~25%; ipi ~30%.
• ~35–40% of nivo+ipi patients are alive and treatment-free at 10 years — the most striking durable-response curve in modern oncology.

The CheckMate-067 10-year OS for stage IV melanoma is a numerical and conceptual landmark: in 2010, before ipi, fewer than ~10% of stage IV patients survived 5 years; by 2022, ~43% were surviving 10 years on dual-checkpoint blockade. This is a transformation almost without precedent in solid-tumour oncology.

RELATIVITY-047 (Tawbi et al., NEJM 2022) added a third combination: nivolumab + relatlimab (anti-LAG-3) vs nivo alone. PFS 10.1 vs 4.6 months; lower-toxicity alternative to ipi+nivo. Nivo+relatlimab is now an option for patients who are PD-L1-low or for whom ipi+nivo toxicity is prohibitive.

11. TIL Therapy & T-VEC — Beyond Checkpoint Blockade

For patients who progress on anti-PD-1 (and ipi+nivo where used), three options with FDA approval:

• Lifileucel (Iovance) — the first FDA-approved TIL therapy for any solid tumour, approved February 2024. Tumour-infiltrating lymphocytes are harvested from a resected metastasis, expanded ex vivo with IL-2 to ~50 billion cells, and infused after non-myeloablative lymphodepletion. ORR ~31%, median DoR ~10 months in heavily pretreated patients (Chesney et al., JITC 2022; LD-A pivotal cohort).
• Talimogene laherparepvec (T-VEC, Imlygic) — an oncolytic herpes simplex virus expressing GM-CSF, injected intralesionally for unresectable cutaneous/subcutaneous/nodal melanoma. Approved 2015; ORR ~26%, with abscopal-like responses at distant uninjected sites. Modest activity as monotherapy; combination with anti-PD-1 trials ongoing (MASTERKEY-265 negative for OS).
• Tebentafusp — bispecific TCR-anti-CD3 fusion (gp100-specific) approved 2022 for HLA-A*02:01+ uveal melanoma. The first systemic therapy with OS benefit in uveal melanoma (HR 0.51 vs investigator's choice).

Lifileucel’s 2024 approval marks two firsts: first solid-tumour TIL therapy and first one-time T-cell product approved outside CD19-CAR. The SPECTRUM-D / TILVANCE trial is now testing earlier-line lifileucel + pembrolizumab vs pembro alone, anticipating a likely move into first or second line in the late 2020s.

12. Adjuvant & Neoadjuvant Therapy

Adjuvant (post-resection) systemic therapy is now standard for stage IIB–IV (resected) melanoma:

• Anti-PD-1 (pembrolizumab or nivolumab) — KEYNOTE-054, CheckMate-238 in stage III; KEYNOTE-716 in stage IIB/IIC: ~25% reduction in recurrence at 3 years.
• Dabrafenib + trametinib in BRAF V600 — COMBI-AD (Long et al., NEJM 2017): 5-yr RFS 52% vs 36% placebo.
• Choice between adjuvant anti-PD-1 vs BRAF+MEK in BRAF V600 stage III is unsettled; SWOG-S1801 favoured pembrolizumab for the long-term response curve.

Neoadjuvant immunotherapy — giving anti-PD-1 (or ipi+nivo) before resection — is the latest development:

• SWOG-S1801 (Patel et al., NEJM 2023): randomised stage III/IV resectable melanoma to neoadjuvant pembrolizumab (3 doses) + adjuvant pembrolizumab vs adjuvant pembrolizumab alone. Event-free survival at 2 yrs: 72% vs 49%.
• NADINA (Blank et al., NEJM 2024): neoadjuvant ipi+nivo × 2 cycles in stage III: pathologic CR ~60%; 2-yr EFS 84% vs 57% with adjuvant nivo only.

Neoadjuvant immunotherapy is increasingly the preferred approach for resectable stage III melanoma — both because pathologic response is a strong surrogate for long-term outcome and because the tumour-bearing patient mounts a more effective immune response than after surgical removal.