10.2 Melanoma, Mesenchymal, and Neuroendocrine Markers

10.2 Melanoma, Mesenchymal, and Neuroendocrine Markers

Quick Answer: Melanoma screens: S100 (nuclear and cytoplasmic), SOX10 (nuclear), Melan-A/MART-1 and HMB-45 (cytoplasmic), PRAME (nuclear, practical melanoma-versus-nevus use). Muscle: cytoplasmic desmin and SMA; nuclear myogenin for skeletal-muscle differentiation. Beta-catenin: membranous versus nuclear as a Wnt-pathway pattern example. Neuroendocrine: synaptophysin, chromogranin, nuclear INSM1, membrane CD56. A keratin-poor epithelioid tumor is not automatically a true sarcoma. Chapter 9 already taught CK7/CK20 maps—do not reuse them here.

This OpenExamPrep section is independent teaching on melanoma, mesenchymal, and neuroendocrine staining patterns covering published QIHC staining topic IV.B.2.a. It is not an ASCP publication and does not claim Board, CAP, or WHO-classification approval. Clones named below are examples. These panels are lineage screens, not diagnoses.

Melanocytic markers at practical QIHC level

Melanoma can mimic carcinoma, lymphoma, and sarcoma. The first job is still the H&E (pigment versus chromogen, nesting versus sheets, nucleoli). Then use melanocytic antibodies in their expected compartments.

S100 must show nuclear staining to count as a real S100 result; cytoplasmic-only S100 is not the textbook melanocytic pattern. Because S100 is promiscuous (nerve, cartilage, Langerhans cells, some carcinomas, some histiocytic lesions), a positive S100 starts a panel; it does not finish one.

SOX10 is the nuclear transcription-factor screen many laboratories now lead with for suspected melanoma and nerve-sheath tumors. Nuclear localization is required. SOX10-positive spindle cells in a scar-like dermis raise desmoplastic melanoma and scar/nerve-sheath differentials together—Melan-A and HMB-45 often will not rescue you there.

Melan-A (MART-1) and HMB-45 are cytoplasmic. They are useful in epithelioid melanoma and in confirming melanocytic differentiation when S100/SOX10 are positive. Desmoplastic melanoma is the classic HMB-45-negative / Melan-A-negative (or only focal) pitfall: S100 and SOX10 remain the workhorse stains. Do not call a SOX10-positive desmoplastic tumor “not melanoma” because HMB-45 is negative.

PRAME (preferentially expressed antigen in melanoma) is nuclear. At a practical screening level, many melanomas are PRAME-positive and many banal nevi are PRAME-negative. Exceptions exist: some melanomas are PRAME-negative; some nevi (including some dysplastic or traumatized lesions) can show labeling. Teach PRAME as supportive nuclear context, not as a yes/no melanoma assay and not as a replacement for a dermatopathologist’s threshold.

Pigment versus chromogen: coarse brown-black melanin is already on the H&E. Use a red chromogen or a bleach protocol when brown DAB and melanin collide. Section 9.1 already taught pigment traps; the same rule applies here.

Mesenchymal muscle markers: desmin, SMA, myogenin

When the H&E looks like muscle or a spindle-cell sarcoma, the usual IHC trio is desmin, smooth-muscle actin (SMA), and myogenin—read in the correct compartment.

Desmin is a muscle intermediate filament, cytoplasmic. It marks skeletal, cardiac, and many smooth-muscle cells, and it is the expected stain in rhabdomyosarcoma and leiomyosarcoma. Desmin also labels some myofibroblasts and some non-muscle tumors; it is a muscle-class stain, not a sarcoma stamp.

SMA is cytoplasmic (often membrane-accentuated). It marks smooth muscle and myofibroblasts. A healing biopsy site full of SMA-positive myofibroblasts is not leiomyosarcoma. Vascular smooth muscle in vessel walls is the usual internal control; do not score the artery and call the tumor positive.

Myogenin is nuclear. It indicates skeletal-muscle differentiation and is the marker that points toward rhabdomyosarcoma in the small-round-cell and spindle-cell differentials. Cytoplasmic myogenin is not a rhabdomyosarcoma call. Leiomyosarcoma is typically myogenin-negative (desmin/SMA-positive). MyoD1 is a related nuclear skeletal-muscle factor used in some panels; if you see it, still score nuclei.

Beta-catenin as a nuclear-versus-cytoplasmic pattern example

Beta-catenin is not a melanoma stain. It is taught here because QIHC staining-pattern content includes reading compartment, and beta-catenin is a clean teaching example of membranous versus nuclear signal with different meaning.

In most epithelia, beta-catenin sits at the adherens junction: expected stain is membranous (often with a little cytoplasm). When the Wnt destruction complex is disrupted (CTNNB1 mutation or equivalent), beta-catenin accumulates in the nucleus. Desmoid-type fibromatosis is the mesenchymal example: look for nuclear beta-catenin in the spindle cells (often not every nucleus). Many colorectal carcinomas and some other tumors also show nuclear beta-catenin—this section is not a GI atlas.

Screening rules:

• Membranous beta-catenin in epithelium is the default physiologic pattern, not a “positive special stain.”
• Nuclear labeling (usually with cytoplasm) is the abnormal Wnt-type pattern the pathologist wants documented.
• Cytoplasmic-only blush without membrane or nucleus is not a desmoid call and is often noise or a titer problem.
• Endothelium and some normal cells can show weak nuclear staining—compare with the tumor spindles, do not over-call a vessel.

If a stem describes nuclear beta-catenin in a deep fibromatosis-like proliferation, name nuclear localization. If it describes crisp membranes in benign colonic epithelium, that is the normal pattern, not disease.

Neuroendocrine markers: synaptophysin, chromogranin, INSM1, CD56

Neuroendocrine (NE) IHC supports NE differentiation; it does not by itself prove a primary organ. Poorly differentiated NE carcinomas can be keratin-pale or keratin-dot-like (Chapter 9 taught keratin localization; this section does not reprint coordinate CK tables).

Synaptophysin is the usual sensitive cytoplasmic screen. Chromogranin is the more specific granule marker and may be negative when the tumor is poorly granulated. INSM1 is scored in nuclei; it has become the preferred NE transcription-factor stain in many laboratories because it is less promiscuous than CD56. CD56 is a membrane stain that will light up true NE tumors and many non-NE mimics—do not diagnose neuroendocrine carcinoma on CD56 alone.

Normal islets, pulmonary NE cells, or adrenal medulla (depending on the control tissue) are internal or external controls. A completely negative NE panel with negative internal NE cells is a failed stain, not proof the tumor is “non-NE.”

Keratin-poor tumors versus true sarcoma

Chapter 9 covered pan-keratins and CK7/CK20 coordinate maps. Do not reprint those tables. The screening idea that still belongs here:

Absence of keratin does not equal sarcoma. Adrenal cortical carcinoma, some hepatocellular carcinomas, some sarcomatoid carcinomas, and some melanoma and hematolymphoid tumors are keratin-poor or keratin-negative. Presence of keratin does not equal carcinoma in every spindle-cell tumor: synovial sarcoma and epithelioid sarcoma are classic keratin-positive mesenchymal tumors.

A practical first-line frame after H&E:

1. Broad keratin (already discussed in Chapter 9) plus CD45 plus SOX10/S100—carcinoma versus hematopoietic versus melanoma screens.
2. If those are negative and the tumor looks mesenchymal, add muscle markers (desmin/SMA/myogenin) and other directed stains the pathologist orders (not a fishing expedition of every sarcoma antibody).
3. If the tumor looks neuroendocrine, add synaptophysin/chromogranin/INSM1 rather than calling it sarcoma because keratin is faint.

True sarcoma is a morphology-plus-panel diagnosis. The QIHC technologist’s job is to produce interpretable lineage stains in the right compartment with working controls, not to issue “undifferentiated sarcoma” from a single negative keratin.

Worked example

A dermal spindle-cell tumor in sun-damaged skin is HMB-45-negative and Melan-A-negative. S100 and SOX10 are strongly nuclear (S100 also cytoplasmic) in the spindles. PRAME shows nuclear labeling in a subset. SMA highlights only vessel walls; desmin and myogenin are negative; keratin is negative. That cluster is a desmoplastic-melanoma-type melanocytic pattern, not “not melanoma because HMB-45 is negative,” and not a myofibroblastic sarcoma because SMA is present in vessels. A second case, a deep abdominal wall mass, shows SMA and desmin in cytoplasm of long fascicles, myogenin-negative nuclei, and nuclear beta-catenin in a fraction of spindles: document nuclear beta-catenin as a desmoid-type Wnt pattern for the pathologist—do not call it melanoma because S100 was not even indicated. A third case, a lung mass with salt-and-pepper chromatin, is CD56-positive but INSM1-negative, synaptophysin-negative, chromogranin-negative, with CD56 also staining lymphocytes: do not sign out neuroendocrine carcinoma on CD56 alone.

Lineage markers are pattern tools under topic IV.B.2.a. They feed the pathologist’s diagnosis; they are not the diagnosis.