HMB45 Melanoma: Clinical Significance and Diagnostic Value

HMB45 is one of the most widely used immunohistochemical stains in the diagnosis of melanoma, prized for its high specificity for melanocytic tumors. It works by binding to a protein called gp100 (also known as Pmel17), which is found in immature melanosomes, the organelles responsible for pigment production in melanocytes. While no single marker is perfect, HMB45 has been a diagnostic workhorse for decades because it rarely flags non-melanocytic cancers as melanoma. That strength, though, comes with trade-offs that matter in real clinical decisions.

What HMB45 Actually Detects

HMB45 is a mouse-derived monoclonal antibody originally raised against melanoma tissue extracts. It recognizes gp100/Pmel17, a glycoprotein concentrated in the pre-melanosomes of pigment-producing cells.¹ Cells that have been transfected with pMel17 cDNA react with HMB45 as well as with related diagnostic antibodies NKI-beteb and HMB-50, confirming that this single protein is the core target shared across several melanoma-diagnostic reagents.² Immunoelectron microscopy has further shown that HMB45 reacts specifically with immature melanosomes, and the antigen appears to be a sialylated glycoconjugate, meaning that enzyme treatment with neuraminidase wipes out HMB45 staining entirely.

A useful quirk of this biology is that HMB45 staining tends to be strongest in actively proliferating or immature melanocytes, which is why melanoma cells typically light up while mature, resting melanocytes in normal skin do not. This distinction between active and quiescent pigment cells is what gives HMB45 much of its diagnostic power.

Sensitivity and Specificity in Melanoma Diagnosis

In one of the landmark early studies examining 200 paraffin-embedded cutaneous tumors, HMB45 labeled about 93% of melanomas while achieving 100% specificity. It did not stain any of the 133 non-melanocytic neoplasms in the series. By contrast, the alternative marker S-100 caught every melanoma (100% sensitivity) but also stained 15 of those non-melanocytic tumors, dragging down its specificity.³ That trade-off captures the core character of HMB45: when it stains positive, you can be quite confident you are looking at a melanocytic lesion, but a negative result does not rule melanoma out.

More recent work paints a less rosy picture of HMB45’s sensitivity when used alone. In a study comparing melanocytic markers across a mix of primary and metastatic melanomas, HMB45 was positive in roughly 78% of cases, compared with 100% for S-100 and 75% for Melan-A.⁴ And when the newer marker PRAME was compared head-to-head with HMB45 in primary melanocytic tumors, diffuse PRAME positivity showed both higher sensitivity (about 74%) and higher specificity (about 97%) than HMB45 alone (roughly 56% sensitivity, 86% specificity).⁵ These numbers vary across studies because tumor populations and staining protocols differ, but the trend is consistent: HMB45 is specific rather than sensitive, and using it as a solo marker risks missing a meaningful proportion of melanomas.

Where HMB45 Falls Short

The biggest blind spot involves spindle cell melanoma, a subtype in which tumor cells are elongated and resemble fibroblasts or smooth-muscle cells. In a series of 56 spindle cell melanoma cases, more than half, 30 of 56, were HMB45-negative.⁶ Because spindle cell melanoma can already be tricky to recognize under the microscope, relying on HMB45 here could lead to a dangerous miss. Pathologists working up a suspected spindle cell lesion will almost always add S-100 or SOX10, both of which perform better in this setting.

Desmoplastic melanoma, another subtype in which abundant fibrous tissue surrounds sparse tumor cells, is also notorious for HMB45 negativity. When a melanoma has these unusual growth patterns, a negative HMB45 result should be interpreted with caution, not as reassurance.

How HMB45 Compares to Other Melanocytic Markers

Pathologists rarely use HMB45 in isolation. A typical melanoma workup involves a panel of markers, each with different strengths. Understanding how they complement each other helps explain why HMB45 remains useful despite its limitations.

S-100 is the most sensitive melanocytic marker available, catching essentially all melanomas, including spindle cell and desmoplastic variants. However, S-100 also stains nerve-sheath tumors, dendritic cells, and some carcinomas, so a positive result does not automatically mean melanoma. SOX10, a nuclear transcription factor, matches S-100 in sensitivity for detecting melanoma metastases in lymph nodes and does a better job at avoiding non-melanocytic tumors.⁷ In benign melanocytic lesions, both SOX10 and Melan-A stain significantly more melanocytes than HMB45, meaning that HMB45 can undercount melanocytes relative to its panel mates.⁸

PRAME is a newer entrant that has generated excitement because of its ability to distinguish malignant from benign melanocytic lesions. Strong diffuse PRAME expression correlates with melanoma, while benign nevi tend to be PRAME-negative. Combining PRAME with HMB45 in a dual stain can allow a pathologist to see both malignancy-associated and melanocyte-associated signals on a single slide.⁹

Telling Nevi from Melanoma Using Staining Patterns

One of the most clinically useful features of HMB45 is not just whether cells stain positive, but how the staining is distributed through the lesion. In ordinary benign moles (compound nevi), HMB45 tends to show a gradient pattern: superficial melanocytes near the skin surface stain strongly, and deeper cells stain weakly or not at all. Melanomas, by contrast, tend to stain diffusely throughout the lesion, including deep dermal components.

This gradient pattern has been studied specifically in Spitz nevi, a type of benign mole that can closely mimic melanoma under the microscope. The majority of compound Spitz nevi show a stratified HMB45 staining pattern resembling ordinary nevi rather than melanoma, and this pattern can serve as a helpful extra clue when deciding whether a worrisome Spitzoid lesion is truly malignant.¹⁰ Statistically significant differences in HMB45 staining pattern between Spitz nevi and melanoma have been confirmed alongside other criteria such as maturation, pagetoid spread, and p16 expression.¹¹

There is a wrinkle, though. Halo nevi, the benign moles surrounded by a ring of depigmented skin, can show aberrant HMB45 staining. Half of the halo nevi in one study showed unexpected deep or diffuse HMB45 positivity, a pattern that might normally raise suspicion for melanoma. Because PRAME was negative in all of those cases, adding PRAME to the panel can prevent a pathologist from over-calling a halo nevus as malignant.¹²

Using Multi-Marker Scoring Systems

Because no single stain is perfect, researchers have developed scoring systems that combine several markers to boost accuracy. One such system uses p16, Ki-67, and HMB45 together. A total score from these three stains was able to separate nevi from primary melanomas with sensitivity and specificity above 97% in the initial series, and achieved perfect classification in a validation set of 62 tumors.¹³ Similarly, immunohistochemical double stains pairing Ki67 with MART-1 and HMB45 with MITF have shown promise. When the HMB45 and Ki-67 indices were combined in one study, none of the melanomas were misclassified, though about 13% of nevi were.¹⁴ These panel-based approaches are where HMB45 shines: it contributes high specificity to a combination that also includes more sensitive markers.

Sentinel Lymph Node Assessment

When melanoma is diagnosed, one of the most important staging steps is the sentinel lymph node biopsy. The idea is to find and examine the first lymph node that drains the tumor site, looking for tiny clusters of melanoma cells (micrometastases) that might not be visible on standard staining. HMB45 has proven especially useful here.

In a study of 66 consecutive melanoma patients, routine microscopy found metastases in 18% of patients. Of the remaining patients whose nodes looked clean under standard staining, 7% were found to harbor scattered HMB45-positive cells on immunohistochemistry. Importantly, none of the 244 control lymph nodes from patients without melanoma contained a single HMB45-positive cell, reinforcing the marker’s specificity in this setting.¹⁵ One practical advantage here is that benign nevus cell aggregates, which occasionally sit in lymph nodes and can confuse pathologists, do not stain with HMB45, while melanoma deposits do.

That said, HMB45 is not infallible even in this context. A separate study examining sentinel nodes that had been falsely called negative on initial pathology found that three of 10 micrometastases did not stain with HMB45, though all stained with S-100 and MART-1.¹⁶ The lesson is the same as with primary tumors: HMB45 is excellent for confirming melanoma when it stains positive, but it should not be the sole stain used to rule melanoma out.

Non-Melanoma Tumors That Stain Positive

Although HMB45 is often described as “melanoma-specific,” a handful of non-melanocytic tumors also react with it, and recognizing them is critical to avoid misdiagnosis. The most important group is the PEComa family, a set of rare tumors defined by cells that express both muscle-related and melanocytic markers. PEComas include angiomyolipomas (most commonly found in the kidneys) and lymphangioleiomyomatosis (LAM), a rare lung disease.¹⁷ These tumors consistently express HMB45 alongside smooth-muscle actin, a combination that would be bizarre for melanoma and essentially serves as their diagnostic fingerprint.¹⁸

Clear cell sarcoma is another tumor that can mimic melanoma both under the microscope and on immunohistochemistry. Clear cell sarcoma shares expression of S-100, HMB45, and Melan-A with melanoma, which makes staining alone insufficient to tell them apart. The key differentiator is molecular: clear cell sarcoma carries a characteristic EWSR1-ATF1 fusion gene. PRAME expression can also help, as melanoma tends to show stronger and more frequent PRAME staining than clear cell sarcoma does.

In non-small cell lung cancer, HMB45 positivity is extremely rare, occurring in less than 1% of cases in one large series, which means an unexpected HMB45-positive lung mass strongly favors a melanocytic origin.¹⁹

Prognostic Clues from HMB45 Staining

Beyond diagnosis, HMB45 expression may carry some prognostic information. In a study of cutaneous melanoma biomarkers, HMB45 staining intensity showed a negative correlation with Breslow depth, the single most important pathological predictor of melanoma outcome. Thinner, earlier-stage tumors tended to have stronger HMB45 expression, while thicker, more advanced tumors showed weaker staining. In univariate analysis, higher HMB45 expression was identified as a positive prognostic factor.²⁰ This makes biological sense: as melanomas progress and dedifferentiate, they may downregulate the melanocytic machinery that HMB45 detects, becoming less “melanocyte-like” in their protein expression.

That inverse relationship also explains why metastatic melanomas sometimes lose HMB45 reactivity. In patients undergoing immunotherapy, researchers have looked at whether treatment changes the expression of melanoma antigens over time. One study using serial fine-needle aspirates found no statistically significant change in HMB45 expression at two months compared to baseline, though there was a trend toward declining staining intensity.²¹ This suggests that antigen loss is something pathologists should keep in mind when interpreting biopsies from patients who have already started treatment.

Technical Challenges with Heavily Pigmented Tumors

Melanoma presents a unique laboratory problem: the very pigment that defines the tumor can obscure immunohistochemical staining. Melanin is brown, and the most common chromogen used in immunostaining (DAB) also produces a brown signal. In heavily pigmented specimens, it becomes nearly impossible to tell melanin deposits from a genuine positive stain. This is a practical headache that directly affects how useful HMB45 and other markers are in pigment-rich tumors.

The standard workaround is to bleach melanin out of the tissue before staining. Various concentrations of potassium permanganate can remove pigment effectively, but the bleaching process itself can damage the very proteins the antibodies need to bind. One study found that cytoplasmic antigens like HMB45, MART-1, and BRAF V600E suffered noticeable intensity loss after bleaching, while nuclear antigens like Ki-67 and SOX10 held up better.²² Another group tested multiple detection methods after bleaching and found that multiplex immunofluorescence preserved the best positivity rates for HMB45 and other markers, while certain chromogens like AEC performed poorly.²³

One clever solution is to use a red chromogen instead of a brown one so that positive immunostaining is visually distinct from melanin. Dual-stain protocols that combine HMB45 with PRAME, for example, can use DAB for nuclear PRAME and a red chromogen for cytoplasmic HMB45, making it possible to read both signals even in pigmented tissue.⁹ These technical choices may sound like laboratory minutiae, but they directly affect whether a pathologist can confidently read the slide and give the patient an accurate diagnosis.

HMB45 and Nuclear Versus Cytoplasmic Localization

The gp100/Pmel17 protein that HMB45 detects is normally a cytoplasmic protein found in melanosomes. However, research has shown that in some melanocytic lesions, the protein shifts to a nuclear location. A study comparing nuclear versus cytoplasmic HMB45 staining found that this relocalization could help discriminate between benign and malignant melanocytic lesions.¹ In routine practice, pathologists focus primarily on the presence and distribution of cytoplasmic staining, but nuclear localization represents an additional dimension that may gain more clinical attention as staining interpretation becomes more nuanced.

Where HMB45 Fits in a Modern Melanoma Workup

The diagnostic landscape has shifted considerably since HMB45 was introduced in the 1980s. Newer markers like PRAME and SOX10 have carved out important roles, and molecular testing for BRAF mutations and gene-expression profiling now contribute to treatment decisions in ways that immunohistochemistry alone cannot. Despite this evolution, HMB45 has not been retired. It continues to occupy a specific niche: a highly specific melanocytic stain that works well in panels, excels at sentinel lymph node screening, and provides pattern-based clues (the gradient) that no single alternative fully replaces.

For patients, the practical takeaway is that a pathology report showing HMB45 positivity is strong evidence that a lesion is melanocytic in origin, but the diagnosis of melanoma itself rests on the full picture, including morphology, staining patterns across multiple markers, and sometimes molecular testing. A negative HMB45 result, meanwhile, should never be taken as proof that melanoma has been excluded, especially in unusual tumor subtypes or metastatic deposits that have lost their melanocytic protein expression. When reading your own pathology report, expect to see HMB45 listed alongside at least one or two companion stains, each covering for the others’ weaknesses.