Merkel cell carcinoma

Meet the patient

A 78-year-old fair-skinned farmer has a firm, shiny, red-violet nodule on his right cheek that has tripled in size over six weeks. It does not hurt, it does not bleed, and he is in your room only because his daughter insisted. He is certain it is "just a cyst".[2]

Before you reach for the curette, hold the recognition question every MCC case turns on: is this one of the AEIOU lesions — painless, expanding, in an older, UV-exposed or immunosuppressed patient — that I must biopsy deeply and then stain for CK20 and TTF-1? A cyst does not treble its size in six weeks on the face of a 78-year-old, and a cyst is not your working diagnosis here.[2][6]

What is MCC — and why does it kill more often than melanoma?

MCC is a high-grade primary cutaneous neuroendocrine carcinoma — rare, fast, and disproportionately lethal. Derived from epidermal Merkel cells (specialised light-touch mechanoreceptors), it is built from small round blue cells with neuroendocrine differentiation that spread early through lymphatics and blood. The line examiners want is its per-lesion mortality: MCC kills more often per lesion than melanoma, which is precisely why a painless red nodule on a sun-damaged forehead is never a cyst until proven otherwise.[2][6]

In the 2022 WHO classification of neuroendocrine neoplasms, MCC sits as a high-grade poorly differentiated neuroendocrine carcinoma of cutaneous origin. That matters because it is by definition high-grade and aggressive — there is no well-differentiated, indolent cousin the way there is in gut or lung. The WHO label also draws the boundary against other small round blue cell tumours and against extrapulmonary neuroendocrine carcinomas that metastasise to skin.[1]

The old names are viva gold even if they no longer drive management: trabecular carcinoma (the original 1972 Toker description), primary neuroendocrine carcinoma of the skin, and the obsolete APUDoma of skin. The histological growth patterns — trabecular, intermediate or diffuse (the commonest), and small cell — are descriptive only; all three behave aggressively, so the pattern does not change prognosis or staging.[6]

Etymology for the viva: Merkel cells are named for Friedrich Sigmund Merkel, who described them in 1875 as Tastzellen — "touch cells" — long before anyone guessed their malignant counterpart would be a virus-driven skin killer. The cell outlived its discoverer; the tumour is still catching up to its reputation.[6]

How common, who, and why — the AEIOU patient

MCC is rare but rising, and it selects its patients with eerie consistency: elderly, fair-skinned, sun-damaged, male, and often immunosuppressed. In the United States SEER-18 registry the age-adjusted incidence is approximately 0.7 per 100,000 person-years, climbing 6 to 8 percent annually. Queensland, Australia reports the world's highest incidence at 1.6 to 2.0 per 100,000 — high ambient UV, a fair-skinned population, and aggressive surveillance all conspiring.[2][6]

The demographic fingerprint is almost exam-scriptable: median age around 75 years, fewer than 5 percent present before 50, and a roughly 2:1 male predominance. MCC is uncommon in deeply pigmented populations, but when it does arise there it tends to present later and run a worse course — delayed recognition, not different biology.[4][6]

Three risk factors do almost all the work: chronic ultraviolet exposure, immunosuppression, and Merkel cell polyomavirus carriage. UV is the strongest attributable risk overall. Immunosuppression is where the relative risks go vertical: solid organ transplantation raises risk 10 to 24-fold, and chronic lymphocytic leukaemia (CLL) raises it 16 to 30-fold — a number examiners love. HIV with a low CD4 count, prior PUVA, ciclosporin, arsenic, and chronic autoimmune immunosuppression round out the list.[3][4]

Everyone forgets: CLL is not just a B-cell problem — it produces a profound T-cell and NK-cell defect, and many patients have then been fludarabine- or BTK-inhibitor-treated on top. That layered immune failure is exactly what lets MCPyV-transformed cells escape surveillance and expand, which is why the CLL cohort carries the single highest MCC relative risk in the books.[4]

[1]

The AEIOU criteria are a recognition tool, not a scoring system. They are highly sensitive: a painless, rapidly growing, red-violet nodule on sun-exposed skin in an older or immunosuppressed patient is MCC until proven otherwise. The trap is the other direction — do not use a "missing letter" to exclude MCC, because not every lesion hits all five. AEIOU is a prompt to biopsy, not a rule-out.[2][6]

Two roads to one tumour — MCPyV versus UV

MCC has two mutually exclusive origins, and both converge on the same switch: loss of the retinoblastoma (RB1) tumour suppressor. That convergence is the single most elegant fact in MCC biology, and it is also why a virus-driven and a UV-driven tumour answer to the same immunotherapy.[3][4]

The MCPyV-driven pathway (approximately 80 percent). MCPyV is a small double-stranded DNA polyomavirus with a 5.4 kb circular genome, discovered in 2008 by Feng and colleagues through digital transcriptome subtraction of MCC tissue. In positive tumours the viral genome is clonally integrated — present in every tumour cell, proving integration is an early transforming event, not a passenger. Viral large T antigen binds and functionally inactivates RB1, releasing E2F and driving uncontrolled G1-to-S progression; small T antigen stabilises large T, inhibits protein phosphatase 2A, and activates cap-dependent translation. The virus does the heavy lifting; UV damage is largely absent.[3][4]

The UV-driven, virus-negative pathway (approximately 20 percent). These tumours carry no MCPyV DNA and instead show a heavy burden of UV-signature mutations — C to T transitions at dipyrimidine sites — with recurrent hits in TP53, RB1, NOTCH1, KMT2D and other chromatin modifiers. Molecularly they look more like aggressive keratinocyte carcinomas, arise on chronically sun-damaged skin, and may carry a worse prognosis in some cohorts. Critically, the UV mutation load generates abundant neoantigens, so they are immunogenic too.[3][4]

Why both subtypes answer to checkpoint blockade: MCPyV-positive tumours display viral large T and small T antigens to the immune system; virus-negative tumours display UV-induced neoantigens. Either way the tumour is broadcasting foreign peptide, which is why anti-PD-1 and anti-PD-L1 antibodies achieve response rates in MCC that few solid tumours can match.[4]

MCC spreads first through lymphatics to the draining nodal basin — sentinel lymph node biopsy finds occult nodal disease in roughly 30 percent of clinically node-negative patients. Haematogenous dissemination follows later to lung, liver, bone, and brain. In-transit and satellite nodules between the primary and the basin are a regional red flag and usually herald nodal involvement.[5][6]

The lesion you will miss — red-violet, painless, growing

The classic MCC lesion is a firm, dome-shaped, shiny, red-violet to flesh-coloured, painless papule or nodule, 0.5 to 2 cm at presentation — and it is almost always called something benign first. Cyst, lipoma, arthropod bite, vascular lesion, basal cell carcinoma: these are the labels that delay the diagnosis. The feature that should stop you is the combination MCC is famous for — painless AND expanding rapidly. The overlying skin is often thin and telangiectatic; ulceration is less common than in squamous cell carcinoma but appears in advanced lesions.[2][6]

The classic trap: the painless, firm, mobile, red-violet nodule gets dismissed as a benign cyst or lipoma and either watched or shallowly curetted. A rapidly growing subcutaneous nodule on sun-exposed skin in an elderly or immunosuppressed patient is not a cyst — it is a biopsy. A few weeks of delay is enough for nodal metastasis and an upstage.[2][5]

About half of MCCs arise on the head and neck, roughly 40 percent on the extremities (forearm and hand especially), and about 10 percent on the trunk. Rare sites include the buttock, genitalia, periocular skin, lip, and vulva; mucosal MCC is uncommon and behaves more aggressively. Because the head and neck dominate, the lesion is often first spotted by a hairdresser or a family member rather than the patient — a detail worth asking about, because patient-reported delay is the rule.[2]

The presentation bends with the host. In the elderly and frail, MCC arrives late as a bleeding or ulcerated nodule with satellites, written off as a bruise or "age change". In CLL, expect a rapidly growing blue-red nodule on the head, neck, or upper trunk with an ugly course. In transplant recipients, lesions may be atypical, multiple, and aggressive under ongoing immunosuppression. In HIV, risk climbs as the CD4 count falls below 200 cells per microlitre and patients present younger.[4]

Occasionally MCC presents as an unknown primary — a nodal or visceral metastasis with a regressed cutaneous source — and rarely in previously irradiated or chronically inflamed skin, the Marjolin analogue for neuroendocrine disease. In-transit and satellite nodules between primary and basin strongly imply nodal spread and upstage the patient on the spot.[5]

Why does the lesion look red-violet and shiny?

The red-violet colour comes from a densely vascular dermal tumour showing through a thin, overlying epidermis; the shine is the dome of the nodule stretching that thinned epidermis taut. Painlessness plus rapid growth is the combination that should pull you away from a benign cyst or vascular malformation and toward MCC.

[1]

The mimics — and the one discriminator that saves you

The differential of a rapidly growing red-violet nodule on sun-exposed elderly skin is broad; the histology narrows it, and two stains settle it. MCC must be separated from common benign and malignant skin lesions and, most dangerously, from a cutaneous metastasis of small cell lung carcinoma.[5][6]

The discriminating line, learn it cold: small round blue cells that are CK20 perinuclear punctate positive and TTF-1 negative are MCC; the reverse — CK20 negative and TTF-1 positive — is metastatic small cell lung carcinoma. Neuroendocrine markers (chromogranin, synaptophysin, CD56, NSE) are positive in both and therefore do not discriminate. MCPyV large T antigen staining helps because it is positive in MCC and negative in SCLC.[6]

The other mimics sort out by their own lineages: basal cell carcinoma is BerEP4 and BCL2 positive, CK20 negative, slow and pearly; squamous cell carcinoma is CK5/6 and p40 positive, often keratotic or tender; amelanotic melanoma is S100, SOX10, HMB45, and melan-A positive; cutaneous lymphoma (especially B-cell) is CD45 positive with lineage markers; atypical fibroxanthoma is CD10, CD68, and vimentin positive in sun-damaged skin. Each has a stain that MCC does not.[6]

Serum antibody to MCPyV oncoproteins (capsid and large T) has been studied as a biomarker: a rising titre tracks tumour burden and may herald recurrence. It is not yet a standalone diagnostic test, but in antibody-positive patients it earns a place in surveillance.[4]

Bedside round — find the nodes, not just the lesion

Examination in suspected MCC rarely makes the diagnosis; its job is to find nodal spread and to lower your biopsy threshold. Inspect the lesion for colour, surface, ulceration, and telangiectasia, and palpate for firmness and depth. Then examine the draining basin with the same care you gave the primary: cervical and supraclavicular nodes for head and neck lesions, axillary and epitrochlear for upper limb, inguinal and popliteal for lower limb.[5]

Everyone forgets: the epitrochlear and popliteal nodes. A palpable node of even 1 cm in the basin draining an MCC is not normal — it warrants urgent fine-needle aspiration, not watchful waiting. These two stations are missed on hurried exams more than any other.[5]

Survey the skin around the primary for satellite and in-transit metastases — small skin-coloured or pink nodules between the primary and the basin. Perform a full skin examination: patients may carry a second primary MCC, another skin cancer, or significant photodamage. Photograph and measure the lesion for the multidisciplinary record. MCC is typically non-tender, so true tenderness should redirect you toward an inflamed cyst, abscess, or glomus tumour.[5]

An ABCDE-style assessment adapted for non-pigmented lesions keeps you honest: Asymmetry, irregular Border, Colour (red-violet or erythematous), Diameter (the 6 mm melanoma rule does not apply to nodules — any rapidly growing bump earns a biopsy), and Evolution (growth over weeks). Immune suppression, advanced age, or a UV-exposed site each lower the biopsy threshold on their own.[6]

In immunosuppressed patients the picture distorts: CLL brings small, firm, rapidly growing red-violet nodules; HIV brings atypical distributions and younger age; transplant recipients bring aggressive ulceration and multiple primaries. Refer all of them urgently to the multidisciplinary team.[4]

Biopsy it properly, then stage it

Diagnosis needs a full-thickness biopsy with dermal tissue; a superficial shave is the recurring error that misses MCC. For small lesions, an excisional biopsy with 1 to 2 mm margins is ideal. For larger or anatomically awkward lesions, a 4 mm or deeper punch through the centre is acceptable. A superficial shave that fails to sample the dermis can read as nondiagnostic or, worse, as a benign cyst — go deep.[5][6]

The diagnostic immunohistochemical panel, in order of how much it earns you:[1]

• CK20 — perinuclear punctate (dot-like) positivity is pathognomonic for MCC. Diffuse cytoplasmic CK20 is not the pattern and should not be called positive.
• Neuroendocrine markers — chromogranin A, synaptophysin, CD56 (NCAM), and neuron-specific enolase are usually positive; they confirm neuroendocrine lineage but do not separate MCC from SCLC.
• MCPyV large T antigen — positive in approximately 80 percent of cases (CM2B4 antibody); confirms the viral pathway and helps exclude SCLC.
• TTF-1 — negative in MCC; positive in most SCLCs. This is the discriminator.
• CDX2, CEA, CD45/LCA — usually negative in MCC, excluding intestinal metastasis, adenocarcinoma, and lymphoma respectively.[6]

[1]

Staging uses the AJCC 8th edition (2017) TNM system for MCC:[1]

• T stage: T1 is 2 cm or less; T2 is more than 2 cm up to 5 cm; T3 is more than 5 cm; T4 invades deep extradermal structures (fascia, muscle, cartilage, bone).
• N stage: N0 no regional nodes; N1a microscopic nodal metastasis on SLNB only; N1b macroscopic nodal disease clinically or radiologically; N2 in-transit metastasis without nodal metastasis; N3 in-transit or satellite metastasis with nodal metastasis.
• M stage: M0 none; M1a distant skin, soft tissue, or distant nodes; M1b lung; M1c other viscera.[5]

[5]

Sentinel lymph node biopsy (SLNB) is mandatory, not optional, for clinically node-negative T1 or greater disease. It upstages roughly 30 percent of patients who looked node-negative and changes management on the spot. Perform it at the same sitting as definitive excision, using dual mapping — lymphoscintigraphy plus blue dye, or radiocolloid plus fluorescent dye.[5][6]

Imaging at baseline is contrast-enhanced CT of chest, abdomen, and pelvis, because lung and liver head the distant-metastasis list. 18F-FDG PET-CT adds value for occult nodal or distant disease when CT is equivocal or the primary is high-risk. Brain MRI is reserved for neurological symptoms. Serum MCPyV antibody can serve as a baseline and surveillance marker where available, with a rising titre flagging recurrence.[5]

The NCCN 2024 risk stratification sorts patients into low risk (primary 1 cm or less, node negative, no lymphovascular invasion), intermediate (1 to 2 cm, node negative), and high risk (any node positive, more than 2 cm, lymphovascular invasion, or immunosuppression), with adjuvant therapy tailored to the band. ESMO-EURACAN 2024 aligns: systemic staging with CT or PET-CT, mandatory SLNB for T1 or greater, and multidisciplinary referral at every stage.[5]

Acute and urgent presentations

True MCC emergencies are uncommon, but when they arrive the move is local control first, multidisciplinary referral immediately. A bleeding or ulcerated primary gets direct pressure, an alginate or haemostatic dressing, analgesia and tetanus cover as needed, urgent biopsy if not already done, and a photograph for the MDT. Surgical haemostasis in the emergency department is rarely required.[5]

A fungating MCC in a transplant recipient is biopsied urgently, but immunosuppression is not reduced unilaterally — abrupt withdrawal can reject the graft. The transplant team, dermatologist, medical oncologist, and radiation oncologist decide jointly, and conversion to an mTOR inhibitor such as sirolimus or everolimus may both lower MCC risk and preserve graft function.[5]

Suspected spinal cord compression from vertebral metastasis is the oncological emergency. Manage by the book: high-dose dexamethasone, urgent whole-spine MRI, and urgent radiotherapy; surgical decompression for progressive motor deficit, spinal instability, or a single compressive lesion in a fit patient. MCC is highly radiosensitive, so radiotherapy often palliates quickly.[5]

Tumour lysis syndrome is rare in MCC because the burden is usually smaller than in bulky lymphoma or leukaemia, but it can surface in disseminated disease or after a brisk immunotherapy response. Prophylaxis is aggressive hydration, allopurinol, and rasburicase for high-risk patients, with close electrolyte monitoring.[5]

Surgery, nodes, and radiotherapy — the localised disease spine

For localised MCC the spine is fixed: adequate-margin surgery plus SLNB, with adjuvant radiotherapy layered on for high-risk features. Metastatic disease breaks off into immunotherapy; chemotherapy is the last resort, not the first.[5][6]

Surgery — margins by size, Mohs by site

Wide local excision (WLE) is the standard primary treatment, and the margin is dictated by tumour size, not by surgeon preference. NCCN recommends 1 cm clinical margins for tumours 1 cm or less and 2 cm clinical margins for tumours larger than 1 cm, taken down to deep fascia. On the eyelid, lip, nose, ear, or cosmetically critical head-and-neck sites those margins are often unattainable — there Mohs micrographic surgery gives complete circumferential and deep margin control while sparing tissue. With Mohs, a narrow peripheral margin is acceptable provided the deep margin is clear.[5][6]

Sentinel lymph node biopsy — mandatory, not discretionary

SLNB is mandatory for clinically node-negative patients with T1 or greater disease, performed at the same time as WLE. Dual mapping identifies the sentinel node(s), which are then examined histologically with immunohistochemistry. A positive SLNB upstages the patient and redirects the nodal plan.[5]

A positive sentinel node — dissection, radiotherapy, or both

For a positive SLNB, current practice offers completion therapeutic lymph node dissection (TLND), adjuvant radiotherapy to the nodal basin, or both. NCCN 2024 favours nodal basin radiotherapy with or without completion dissection; ESMO-EURACAN 2024 similarly recommends nodal basin radiotherapy for a positive SLNB. The choice turns on patient factors, node number, extracapsular extension, and morbidity concerns such as lymphoedema.[5]

Radiotherapy — MCC is highly radiosensitive

MCC is one of the most radiosensitive solid tumours, and radiotherapy earns its place in nearly every high-risk localised case. Adjuvant radiotherapy to the primary is recommended for close or positive margins, tumours larger than 1 cm, lymphovascular invasion, head and neck site, or immunosuppression — typically 50 to 66 Gy in 2 Gy fractions. The nodal basin receives 50 to 66 Gy if SLNB is positive or TLND is not performed. For the non-surgical candidate, definitive radiotherapy alone at 60 to 66 Gy in 2 to 2.5 Gy fractions over 5 to 6 weeks can deliver excellent local control.[5][6]

Metastatic MCC — immunotherapy first, chemotherapy last

Immune checkpoint inhibitors are first-line for metastatic MCC because they produce durable responses that chemotherapy simply cannot match. The principal agents, with the trials that licensed them:[4][5]

[7] [8]

• Avelumab (anti-PD-L1), 800 mg intravenously every 2 weeks — FDA and EMA approved for metastatic MCC on the JAVELIN Merkel 200 data; first-line objective response rate approximately 62 percent, durable at 6 months or longer.[7]
• Pembrolizumab (anti-PD-1), 200 mg intravenously every 3 weeks — approved on KEYNOTE-017, with a first-line objective response rate of approximately 56 percent across both MCPyV-positive and virus-negative tumours.[8]
• Nivolumab (anti-PD-1), 240 mg every 2 weeks or 480 mg every 4 weeks — also active in metastatic MCC and evaluated in the neoadjuvant setting.[4]

After progression on a single checkpoint inhibitor, options include nivolumab plus ipilimumab and investigational combinations such as cabozantinib plus nivolumab and ipilimumab. Dual immunotherapy brings lower response rates and higher toxicity, so patient selection matters; the optimal treatment duration and the role of maintenance remain under study.[5]

Chemotherapy has been demoted — and that demotion is itself an exam answer. Reserved for patients refractory to or ineligible for immunotherapy, or for urgent palliation, the standard regimen is etoposide 100 mg per square metre on days 1 to 3 plus cisplatin or carboplatin every 3 weeks for 4 to 6 cycles (topotecan is the alternative). Response rates are modest at 20 to 30 percent, median progression-free survival is only 3 to 4 months, and responses are not durable. International guidelines moved chemotherapy off first-line precisely because checkpoint blockade is both more effective and more durable.[5][6]

Before and after surgery — adjuvant and neoadjuvant immunotherapy

Adjuvant and neoadjuvant checkpoint blockade are the active frontier in resectable MCC, and both have randomised evidence behind them.[9][10]

[9] [10]

In the adjuvant setting, ADMEC-O randomised completely resected patients to nivolumab 480 mg every 4 weeks for 1 year or observation; disease-free survival was higher with nivolumab at 12 and 24 months, though overall survival data are immature. The phase 3 ADAM (adjuvant avelumab) and STAMP (adjuvant pembrolizumab) trials have completed accrual and results are awaited.[9]

In the neoadjuvant setting, CheckMate 358 gave two doses of nivolumab before surgery in resectable MCC and produced a pathological complete response in approximately 47 percent, with 54 percent showing radiographic reduction of 30 percent or more. Recurrence-free survival tracked pathological response, making neoadjuvant immunotherapy a genuine option in high-risk resectable disease.[10]

Follow-up — intensive, because recurrence is common

MCC recurs in up to 40 percent of patients, most within the first 2 years, so surveillance is front-loaded and aggressive. Clinical examination every 3 to 6 months for 2 years, then every 6 to 12 months until 5 years. Teach self-examination and urge patients to report any new lump promptly. In stage III or IV disease, add surveillance CT or PET-CT every 6 to 12 months for the first 3 years. In patients whose tumours were MCPyV-antibody positive at baseline, trend the serum titre — a rise may precede imaging evidence of recurrence.[5]

The special patients who break the rules

MCPyV-positive versus virus-negative. Viral-positive MCC (roughly 80 percent in cooler climates) carries a lower UV mutational burden and generally responds robustly to checkpoint blockade. Virus-negative MCC is more common in high-UV regions, carries a heavier mutation load, looks molecularly closer to keratinocyte carcinoma, and may have a worse prognosis in some cohorts. Both are treated identically by stage, though trials increasingly stratify by viral status.[3][4]

Transplant recipients carry the worst prognosis of any MCC cohort. They develop multiple primaries, and checkpoint inhibitors carry a real risk of graft rejection — highest in kidney, heart, and lung recipients. Management is shared with the transplant team: reduce or convert immunosuppression to an mTOR inhibitor, and consider immunotherapy case by case, with the risk-benefit discussion documented.[4]

HIV. Risk rises as the CD4 count falls below 200 cells per microlitre; patients are younger and disease more aggressive. Checkpoint inhibitors are generally feasible alongside effective antiretroviral therapy, with opportunistic-infection prophylaxis guided by the CD4 count. Optimise ART before systemic therapy.[4]

CLL — the single highest relative risk. Patients with CLL carry a 16 to 30-fold increased MCC risk: the B-cell immune defect plus prior chemo-immunotherapy compounds the surveillance failure, and prognosis is worse than in CLL-free patients. The CLL regimen itself may need to change to permit effective immunotherapy.[4]

Primary nodal MCC (unknown primary). Isolated nodal or visceral disease with no cutaneous primary behaves like cutaneous MCC and is managed with nodal dissection plus adjuvant radiotherapy and systemic staging; prognosis is poorer than for localised cutaneous disease.[5]

Children and adolescents. Fewer than 1 percent of cases occur here, and the diagnosis must be made cautiously — other small round blue cell tumours, congenital neurocristopathies, and lymphoma all mimic MCC. Small series suggest better outcomes than in adults, but long-term data are thin.[5]

Mucosal MCC (lip, eyelid, vulva). Uncommon and more aggressive. Anatomical constraints force narrow margins, lowering the threshold for adjuvant radiotherapy.[5]

What goes wrong — complications and pitfalls

Complications track the disease and its treatment in three rings: local, regional, and distant. Locally: bleeding, ulceration, superinfection, recurrence, and in-transit or satellite metastases — roughly 40 percent of recurrences are local or in-transit. Regionally: nodal recurrence, nodal rupture, and lymphoedema after dissection. Distally: liver, lung, and bone metastases (the last with pathological fracture), brain metastases with mass effect or seizures, and rare leptomeningeal disease with a dismal prognosis.[5]

Paraneoplastic phenomena are uncommon but exam-worthy: ectopic ACTH causing Cushing syndrome, PTHrP causing hypercalcaemia, and a rare Lambert-Eaton myasthenic syndrome from neuronal autoimmunity. Hunt for these when the symptoms do not fit the lesion.[5]

Treatment complications deserve the same respect. Radiotherapy brings skin atrophy, telangiectasia, and lymphoedema. Immune checkpoint inhibitors bring immune-related adverse events (irAEs) — colitis, hepatitis, pneumonitis, thyroiditis, hypophysitis, adrenal insufficiency, and the rare but potentially lethal myocarditis. Recognition plus early corticosteroids and withholding the offending agent are non-negotiable.[4][5]

[1]

Prognosis — staged, brutal, beatable early

MCC per-lesion mortality exceeds melanoma, and survival collapses with stage — which is why early, deep biopsy is the single most consequential decision in the whole disease.[2][5]

Five-year overall survival by AJCC 8th edition stage is approximately:[1]

• Stage I (T1 N0 M0): 55 to 79 percent depending on cohort.
• Stage IIA-IIB (T2-4 N0 M0): 30 to 60 percent.
• Stage III (node-positive): 30 to 50 percent.
• Stage IV (distant metastasis): 14 to 25 percent.[1]

Roughly half of patients with localised disease survive 5 years, dropping sharply with regional and distant spread. The overall recurrence rate is about 40 percent, most within the first 2 years. Independent adverse prognostic factors include head, neck, or trunk primary site, tumour size greater than 2 cm, lymphovascular invasion, immunosuppression, MCPyV-negative status in some cohorts, a positive SLNB, and depth greater than 6 mm or invasion beyond the subcutis.[5]

Follow-up is intensive and effectively lifelong: clinical examination every 3 to 6 months for 2 years, then every 6 to 12 months until 5 years, with imaging reserved for high-risk or symptomatic disease and serum MCPyV antibody titres trended in antibody-positive tumours. Patient education is load-bearing — any new lesion, lump, or neurological symptom prompts urgent review. Refer palliative care early for symptomatic metastatic disease, declining performance status despite immunotherapy, leptomeningeal disease, uncontrolled pain, or bleeding; early palliative involvement improves symptom control even while disease-modifying therapy continues.[5]

Evidence, guidelines, and regional deltas

Four trials rewrote MCC management in under a decade, and every fellow should name them. JAVELIN Merkel 200 made avelumab first-line and later-line for metastatic disease. KEYNOTE-017 did the same for pembrolizumab. CheckMate 358 proved neoadjuvant nivolumab feasible and effective. ADMEC-O delivered the first randomised evidence for adjuvant nivolumab in resected MCC. Phase 3 adjuvant trials ADAM (avelumab) and STAMP (pembrolizumab) are awaited.[7][8][9][10]

Regional guidelines differ in emphasis rather than principle. NCCN 2024 (United States) backs broad adjuvant radiotherapy for high-risk primaries and nodal basins, with completion lymph node dissection or observation after a positive SLNB. ESMO-EURACAN 2024 reserves adjuvant radiotherapy for positive margins, positive nodes, and immunosuppression, noting that omission in low-risk disease may not compromise overall survival. Both recommend avelumab, pembrolizumab, or nivolumab as first-line systemic therapy for metastatic disease. NICE UK (NG12) frames MCC under the umbrella of rapidly growing skin lesions requiring urgent recognition and referral. Cancer Council Australia aligns with ESMO-EURACAN, reflecting the high local UV burden.[5]

Live controversies: whether to omit completion lymph node dissection after a positive SLNB (NCCN 2024 allows observation plus nodal radiotherapy); the role and timing of neoadjuvant immunotherapy; the optimal duration of adjuvant immunotherapy; and how to manage immunotherapy-refractory disease. Chemotherapy's demotion from first-line — short responses, inferior durability versus checkpoint blockade — is now consensus, not debate.[5]

Serum MCPyV antibody titres are an emerging biomarker: rising titres track tumour burden and recurrence and may enter routine surveillance, though the assay is not yet standardised worldwide and must be read alongside imaging and clinical findings.[4]

The consultant's mantra

Painless, growing, red-violet, sun-exposed, elderly — biopsy it; CK20 dot positive, TTF-1 negative.[1]

Hold that single line and the rest of MCC falls into place: the lesion you must not call a cyst, the stain that separates it from small cell lung carcinoma, and the patient for whom immunotherapy — not chemotherapy — is first-line when disease has spread.[6][5]

Exam pearls

[1]

Red flags

[1]

Ward-round test

A 79-year-old fair-skinned man has a painless red-violet nodule on his cheek that doubled in six weeks. Recognition cue and your first step?

This is the AEIOU lesion — Asymptomatic, Expanding, in an Older, UV-exposed patient — so MCC is the working diagnosis, not a cyst. The first step is a deep punch (4 mm or more) or excisional biopsy through the centre of the lesion with a full immunohistochemistry panel: CK20, neuroendocrine markers (chromogranin, synaptophysin, CD56), MCPyV large T, and TTF-1. Do NOT curette or shave superficially — you will miss the dermal component. Stage the draining nodal basin clinically and plan sentinel lymph node biopsy at definitive excision. [1][5]

A small round blue cell tumour on haematoxylin and eosin staining. CK20 perinuclear punctate positive, TTF-1 negative. Diagnosis, and what does the reverse pattern mean?

Merkel cell carcinoma. CK20 perinuclear punctate (dot) positivity plus TTF-1 negativity is the pathognomonic MCC pattern, and MCPyV large T will be positive in roughly 80 percent. The reverse — CK20 negative and TTF-1 positive — is metastatic small cell lung carcinoma, which demands chest staging and a completely different systemic pathway. Neuroendocrine markers are positive in both and do not discriminate. [6]

Stage IV MCC, treatment-naive. What is first-line, and what is the trap?

First-line is anti-PD-1 or anti-PD-L1 immunotherapy — avelumab 800 mg every 2 weeks, pembrolizumab 200 mg every 3 weeks, or nivolumab. JAVELIN Merkel 200 showed an objective response rate around 62 percent with avelumab and KEYNOTE-017 around 56 percent with pembrolizumab, with durable responses in both viral-positive and virus-negative tumours. The trap is reaching for chemotherapy first — etoposide-platinum is now reserved for refractory, ineligible, or urgently palliative disease because its responses are modest (20 to 30 percent) and short-lived (median progression-free survival 3 to 4 months). [7][8][5]

References

1. [1]Rindi G, Mete O, Uccella S, et al. Overview of the 2022 WHO Classification of Neuroendocrine Neoplasms Endocr Pathol, 2022.PMID 35294740
2. [2]Strong J, Hallaert P, Brownell I. Merkel Cell Carcinoma Hematol Oncol Clin North Am, 2024.PMID 39060119
3. [3]Hernandez LE, Mohsin N, Yaghi M, et al. Merkel cell carcinoma: An updated review of pathogenesis, diagnosis, and treatment options Dermatol Ther, 2022.PMID 34967084
4. [4]Becker JC, Stang A, Schrama D, et al. Merkel Cell Carcinoma: Integrating Epidemiology, Immunology, and Therapeutic Updates Am J Clin Dermatol, 2024.PMID 38649621
5. [5]Lugowska I, Becker JC, Ascierto PA, et al. Merkel-cell carcinoma: ESMO-EURACAN Clinical Practice Guideline for diagnosis, treatment and follow-up ESMO Open, 2024.PMID 38796285
6. [6]Becker JC, Stang A, DeCaprio JA, et al. Merkel cell carcinoma Nat Rev Dis Primers, 2017.PMID 29072302
7. [7]D'Angelo SP, Russell J, Lebbé C, et al. Efficacy and Safety of First-line Avelumab Treatment in Patients With Stage IV Metastatic Merkel Cell Carcinoma: A Preplanned Interim Analysis of a Clinical Trial JAMA Oncol, 2018.PMID 29566106
8. [8]Nghiem PT, Bhatia S, Lipson EJ, et al. PD-1 Blockade with Pembrolizumab in Advanced Merkel-Cell Carcinoma N Engl J Med, 2016.PMID 27093365
9. [9]Becker JC, Ugurel S, Leiter U, et al. Adjuvant immunotherapy with nivolumab versus observation in completely resected Merkel cell carcinoma (ADMEC-O): disease-free survival results from a randomised, open-label, phase 2 trial Lancet, 2023.PMID 37451295
10. [10]Topalian SL, Bhatia S, Amin A, et al. Neoadjuvant Nivolumab for Patients With Resectable Merkel Cell Carcinoma in the CheckMate 358 Trial J Clin Oncol, 2020.PMID 32324435