Part VI
Staging — AJCC 8th Edition
The staging system that drives treatment decisions: Breslow thickness, ulceration, mitotic rate, sentinel-node status, distant metastasis site, and the prognostic gradient from stage IA to stage IV M1d.
1. Why Stage Melanoma?
Few cancers benefit as much from accurate staging as cutaneous melanoma. Staging dictates:
- Surgical margins for wide local excision (1 cm vs 2 cm).
- Whether to perform sentinel lymph node biopsy.
- Whether to offer adjuvant immunotherapy (anti-PD-1) or BRAF/MEK in BRAF-mutants — with major implications for relapse risk.
- Frequency of clinical follow-up and imaging surveillance.
- Eligibility for clinical trials.
The AJCC (American Joint Committee on Cancer) 8th edition (Gershenwald et al., CA Cancer J Clin 2017; implementation 2018) is the global standard. It refines the 7th edition primarily by adjusting T1 sub-stratification and dropping mitotic rate from the staging equation while retaining its prognostic role.
2. Breslow Thickness — the Single Strongest Prognostic Variable
Breslow thickness is the vertical distance in millimetres from the granular layer of the epidermis (or, in ulcerated tumours, from the base of the ulcer) to the deepest invasive tumour cell, measured by the pathologist with a calibrated ocular micrometer. Introduced by Alexander Breslow (Cancer 1970), it has dominated melanoma prognostication for fifty years.
The relationship between Breslow thickness and 10-year mortality is roughly monotonic and approximately log-linear:
Below 1.0 mm, 10-year mortality is <5%; from 1–2 mm ~10–15%; from 2–4 mm ~25–30%; above 4 mm ~40–50%. Each ~0.1 mm matters at the AJCC cut-points (0.8, 1.0, 2.0, 4.0 mm).
3. Ulceration & Mitotic Rate
Ulceration — histological demonstration of full-thickness loss of overlying epidermis with neutrophilic reactive change — is the second-strongest independent prognostic variable after thickness. It approximately upstages a tumour by one T-substage: an ulcerated 1.5 mm melanoma behaves like an unulcerated 3 mm melanoma in registry data.
Mitotic rate — mitoses per square millimetre — was a T1 substage modifier in AJCC 7 (T1a vs T1b) but was removed from AJCC 8 because thickness cut-points already capture most of its prognostic information. Mitotic rate ≥1/mm² remains a strong adverse marker within stage I disease and is reported on every histopathology report.
Other histopathological features routinely reported but not in the formal staging equation: Clark level (now largely supplanted), regression, microsatellitosis, perineural invasion, lymphovascular invasion, and tumour-infiltrating lymphocytes (TILs) — the last with potential favourable prognostic significance.
4. AJCC 8 T Categories
| T category | Thickness | Ulceration |
|---|---|---|
| Tis | In situ | — |
| T1a | <0.8 mm | No ulceration |
| T1b | <0.8 mm with ulceration, OR 0.8–1.0 mm (with or without) | — |
| T2a | 1.0–2.0 mm | No |
| T2b | 1.0–2.0 mm | Yes |
| T3a | 2.0–4.0 mm | No |
| T3b | 2.0–4.0 mm | Yes |
| T4a | >4.0 mm | No |
| T4b | >4.0 mm | Yes |
The AJCC 8 update introduced the 0.8 mm cut-point within T1, recognising that T1a (<0.8 mm, non-ulcerated) has >99% 5-year MSS and rarely benefits from SLN biopsy, while T1b (≥0.8 mm or ulcerated) has a meaningfully higher SLN- positive yield (~5–12%) and is now routinely offered SLN biopsy.
5. Sentinel Lymph Node Biopsy
The sentinel lymph node (SLN) is the first lymph node draining a primary tumour, identified intraoperatively by technetium-99m radio-colloid ± isosulfan blue dye. The technique was introduced for melanoma by Donald Morton in 1992 and has become the standard staging procedure for clinically node-negative T1b and above.
- Probability of a positive SLN scales with Breslow thickness: ~5% at 0.8–1.0 mm; ~12% at 1–2 mm; ~20–25% at 2–4 mm; ~30–40% above 4 mm.
- Performed at the time of (or just before) wide local excision; pathology examines multiple step sections, often supplemented by S100/MART-1 immunohistochemistry.
- Even very small “submicro” tumour deposits are reported, but burden categories matter prognostically (tumour deposit <0.1 mm, 0.1–1.0 mm, >1.0 mm).
The MSLT-I trial (Morton et al., NEJM 2014) showed SLN biopsy provides accurate staging without clear OS benefit but with melanoma-specific survival benefit in T2–T3 patients. The MSLT-II trial (Faries et al., NEJM 2017) then asked whether completion lymph node dissection (CLND) after a positive SLN improves survival. Answer: no. CLND added only morbidity (lymphoedema in ~25%) without melanoma-specific survival benefit, so CLND has been largely abandoned in favour of nodal-basin ultrasound surveillance plus effective adjuvant systemic therapy.
6. AJCC 8 N Categories
| N | Definition |
|---|---|
| N0 | No regional node disease |
| N1a | 1 clinically occult (SLN+) node, no in-transit/microsat |
| N1b | 1 clinically detected node |
| N1c | In-transit/satellite/microsatellite, no node |
| N2a | 2–3 clinically occult nodes |
| N2b | 2–3 with at least 1 clinically detected |
| N2c | 1 node + in-transit/satellite/microsat |
| N3a | ≥4 clinically occult |
| N3b | ≥4 with ≥1 clinically detected, or matted nodes |
| N3c | ≥2 nodes + in-transit/satellite/microsat, or matted with in-transit |
In-transit metastasis is a satellite tumour deposit between the primary site and the regional nodal basin, reflecting haematogenous-lymphatic dissemination through the dermal lymphatics. It is counted in the N categorisation rather than M and indicates a more aggressive biology than SLN positivity alone.
7. AJCC 8 M Categories & LDH
Distant metastasis is M1, sub-stratified by anatomical site and serum LDH:
| M sub-category | Site | LDH |
|---|---|---|
| M1a(0) | Distant skin / subcutaneous / nodal | Normal |
| M1a(1) | Distant skin / subcutaneous / nodal | Elevated |
| M1b(0/1) | Lung ± M1a sites | Normal / Elevated |
| M1c(0/1) | Other visceral (liver, bowel, adrenal) ± M1a/b | Normal / Elevated |
| M1d(0/1) | Central nervous system (brain, leptomeningeal) | Normal / Elevated |
The introduction of M1d for CNS metastasis was an AJCC 8 innovation: prior systems lumped CNS into M1c, but CNS-only metastasis is a distinct prognostic group with worse outcome and different therapeutic considerations (radiotherapy, neurosurgery, stereotactic radiosurgery, dual checkpoint blockade).
Serum LDH is a non-specific marker of tumour burden and lactate metabolism. Elevation worsens M-substage from (0) to (1) and is independently prognostic in stage IV. It is not yet replaced by ctDNA-based metrics in routine clinical use, but ctDNA monitoring is emerging.
8. Stage Groupings & the Prognostic Gradient
Pulled together, the AJCC 8 stage groupings produce a steep prognostic gradient spanning >60-fold mortality:
| Stage | TNM | 5-yr MSS |
|---|---|---|
| 0 | Tis N0 M0 | ~99–100% |
| IA | T1a N0 M0 | ~99% |
| IB | T1b–T2a N0 M0 | ~97% |
| IIA | T2b–T3a N0 M0 | ~94% |
| IIB | T3b–T4a N0 M0 | ~87% |
| IIC | T4b N0 M0 | ~82% |
| IIIA | T1a/b–T2a, N1a/N2a, M0 | ~93% |
| IIIB | Various T+N combos | ~83% |
| IIIC | Various T+N combos | ~69% |
| IIID | T4b N3a/b/c M0 | ~32% |
| IV | Any T, Any N, M1 | ~30%* (immunotherapy era) |
*Pre-immunotherapy 5-year stage IV survival was ~5–10% (DTIC era); modern long-term follow-up of CheckMate-067 shows ~43% OS at 10 years for ipi+nivo combination.
The most striking feature of this gradient is the >60-fold mortality difference between IA and IIID — both with localised disease at diagnosis. Behind this gradient lie three biological variables: thickness, ulceration, and SLN status. Catching melanoma at IA preserves nearly normal life expectancy; missing it into IIID changes outcome more than almost any other staging shift in oncology.
9. Molecular Prognostic Signatures & ctDNA
Several gene-expression signatures are now FDA-cleared as prognostic adjuncts in early-stage melanoma:
- DecisionDx-Melanoma (31-GEP). Stratifies stage I/II tumours into class 1A (low risk) vs class 2B (high risk). 5-year recurrence-free survival ~95% (1A) vs ~50% (2B). Used to select SLN candidates among T1b lesions and to escalate adjuvant decisions.
- MelaGenix. An 8-gene RT-PCR signature with similar intent.
- CP-GEP. A ClinicoPathologic + 8-GEP nomogram for SLN-positivity risk.
Circulating tumour DNA (ctDNA) — detected by digital-droplet PCR or next-generation sequencing of plasma — shows promise in melanoma surveillance:
- Detectable in ~80–90% of stage IV patients; sensitivity rises with disease burden.
- Pre-treatment ctDNA load is independently prognostic; on-treatment dynamics predict immunotherapy response within weeks.
- Persistent post-resection ctDNA in stage III is a strong indicator of impending recurrence; potential basis for ctDNA-adapted adjuvant strategies (now in trials).
- BRAF V600 ctDNA is a reliable and quantitative pharmacodynamic marker for BRAF/MEK therapy.
ctDNA is not yet incorporated into AJCC staging; AJCC 9 (anticipated mid-2020s) may incorporate molecular markers more formally.