Welcome to this February twenty twenty-six rundown from the Journal of the American Academy of Dermatology. Four brief reports on the docket this time — a SEER analysis on immunotherapy and Merkel cell carcinoma, a meta-analysis on acral melanoma and mechanical stress, an international consensus statement redefining the Hutchinson sign, and a large multicenter cohort on how cutaneous squamous cell carcinoma actually kills people. Let's get into it.

First up is a population-based cohort study looking at known-primary versus unknown-primary Merkel cell carcinoma in the era of immune checkpoint inhibitors. The background here is a genuinely interesting immunologic puzzle. Unknown-primary Merkel cell — metastatic disease with no detectable cutaneous primary — has long been thought to represent a kind of spontaneous immune clearance of the primary tumor, and prior work has tied that phenotype to a higher tumor mutation burden and better baseline survival. Checkpoint inhibitors, meanwhile, clearly improve survival in metastatic Merkel cell overall. The open question this group tackled: does the checkpoint inhibitor benefit land equally on known-primary and unknown-primary patients, or does the underlying immunobiology of each group change how much they gain?

Methodologically, this is a SEER registry study, so it's retrospective and observational by nature — you take what the registry gives you. They split the cohort into a pre-checkpoint-inhibitor era, twenty twelve through twenty sixteen, and a checkpoint-inhibitor era, twenty seventeen through twenty twenty-one, and compared cancer-specific survival with log-rank tests and Cox regression, adjusting for age, sex, race, radiotherapy, and chemotherapy. The authors don't spell out why they chose an era-based split rather than tracking actual checkpoint inhibitor exposure, but the likely reason — and this is my inference, not theirs — is that SEER doesn't reliably capture individual drug administration, so era serves as a pragmatic proxy for exposure, with the acknowledged tradeoff that you're not actually confirming who got a checkpoint inhibitor.

Two hundred ninety-nine stage four patients met inclusion criteria — about three-quarters known-primary, one-quarter unknown-primary. Known-primary patients had roughly twice the rate of bone metastases and were less likely to receive radiotherapy. Here's the key finding: before the checkpoint inhibitor era, unknown-primary patients had a clear survival advantage over known-primary patients, consistent with that immune-clearance hypothesis. But once you move into the checkpoint inhibitor era, that gap disappears — there's no significant difference between the two groups anymore. Digging into why, the story is that known-primary patients saw a significant survival improvement in the checkpoint inhibitor era, while unknown-primary patients showed essentially no significant change. In other words, checkpoint inhibitors seem to have closed the gap by lifting the known-primary group up, not by leaving unknown-primary patients behind — but unknown-primary patients simply didn't show the same additional gain.

The authors' interpretation is that unknown-primary tumors may already be operating near a ceiling of immune-mediated control — higher mutation burden, more neoantigen presentation, already better immunogenicity — so there's less additional benefit left for a checkpoint inhibitor to unlock. They also raise the alternative that unknown-primary patients might have unmeasured barriers to actually receiving checkpoint inhibitors, like transplant history or a higher risk of immune-related adverse events. The honest limitations are real here: small sample size, no data on which patients actually received a checkpoint inhibitor, and an era-based design that can't fully separate the drug effect from other secular trends in Merkel cell management over that decade. The authors themselves say this is too preliminary to change practice. For you, the takeaway is interesting-not-actionable — it's a hypothesis-generating signal that the immunologic subtype of metastatic Merkel cell may modulate checkpoint inhibitor benefit, worth watching for a prospective or biomarker-driven follow-up, but not something that should change how you counsel a patient with unknown-primary disease about immunotherapy today.

Second article: a systematic review and meta-analysis asking why acral lentiginous melanoma clusters where it clusters. The clinical gap is straightforward — we have solid risk factor data for UV-driven melanoma, but acral lentiginous melanoma doesn't follow that biology, and the field has been debating whether mechanical stress from gait and grip is actually driving site selection.

The design is a meta-analysis pooling thirty-two studies across multiple countries, registered prospectively, searching for melanoma location relative to what they defined as stress-bearing areas — the high-pressure zones of the plantar foot during gait — versus non-stress-bearing areas like the arch. This is the right design choice for this question because no single institution accumulates enough acral melanomas to answer a site-distribution question with precision — you need pooled data across populations, and a random-effects model appropriately assumes true variability between study populations rather than forcing everything into one fixed effect.

The numbers are striking. Across roughly forty-two hundred patients and over five thousand melanomas, with acral lentiginous melanoma making up about half of those, ninety-one percent of foot acral lentiginous melanomas sat in stress-bearing areas — the heel alone accounted for almost half, the forefoot another quarter or so — compared to only about eight percent in the arch, a non-stress-bearing zone. That's a meaningfully lopsided distribution, especially when you consider the arch and the stress-bearing zones aren't even close to equal in surface area — stress-bearing areas only make up sixty to seventy percent of the plantar surface, yet they're capturing the overwhelming majority of tumors. And critically, when they ran the same comparison for ordinary melanocytic nevi of the foot, there was no significant preference for stress-bearing areas at all — nevi were roughly split. That contrast is the strongest piece of evidence here: something about melanoma specifically, not just pigmented lesion formation in general, is drawn to these mechanically stressed zones. The same pattern held on the hands and feet more broadly — the thumbnail carried a disproportionate share of nail-unit acral lentiginous melanoma at roughly six in ten cases, and the great toe accounted for about three-quarters of toe cases — both high-load, high-friction digits. No left-right laterality, no sex difference in location. One more clinically relevant wrinkle: tumors arising in stress-bearing areas were more than twice as likely to present with a Breslow depth over one millimeter, though they weren't more likely to ulcerate, spread to nodes, or present at a higher stage.

The discussion is appropriately cautious about a prior negative study, a twenty seventeen US paper that found no weight-bearing predilection — the authors note they excluded it for missing location-level event counts and speculate the discrepancy might reflect differences in ancestry or lifestyle across study populations, which is a reasonable but unproven explanation. The stated limitation is that they couldn't fully control for the unequal surface area between stress-bearing and non-stress-bearing zones or between individual digits, which tempers how much you can say about relative risk per unit area rather than just raw proportion of cases. Practically, this doesn't change your biopsy threshold — you're already suspicious of pigmented lesions on the heel, forefoot, thumb, and great toe — but it does reinforce that these are true biological hot spots rather than noticing bias, and it strengthens the counseling point to patients about self-exam of high-pressure sites, plus it adds mechanistic plausibility to the finding that stress-bearing tumors trend toward greater Breslow depth at diagnosis, which is worth keeping in mind when you're gauging pretest suspicion on a deep, symptomatic heel lesion.

Third is not an original study but an expert consensus statement — a formal position paper from the European Nail Society addressing decades of terminology confusion around the Hutchinson sign. The background is almost historical: the eponym traces to Jonathan Hutchinson's eighteen eighty-six description of "melanotic whitlow," and periungual pigmentation has been treated as a red flag for subungual melanoma ever since, but the term has been applied inconsistently — specifically, two unresolved questions have circulated for years: should the label be reserved only for pigmentation around a confirmed nail melanoma, and can pigmentation around a non-melanoma pigmented nail lesion be called pseudo-Hutchinson sign?

Twenty-three experts went through a preconsensus questionnaire followed by a formal consensus meeting, and they landed on ten agreed points, which functionally redefine how you should use this terminology going forward. The headline redefinition is that the Hutchinson sign is not diagnostic of melanoma by itself — it's melanin pigmentation anywhere on the perionychium, cuticle, proximal and lateral nail folds, hyponychium, or surrounding digital pad, associated with a melanocytic nail lesion of any kind, benign or malignant. So the sign should trigger a melanoma workup, but its presence does not mean the patient has melanoma. They explicitly carved out exogenous pigmentation — silver nitrate, potassium permanganate staining — as not qualifying as a true Hutchinson sign at all. They also formalized two related terms you'll want in your working vocabulary: pseudo-Hutchinson sign, which is pigment simply shining through the thin free margin of the cuticle and proximal nail fold, caused by a range of benign and malignant lesions and importantly not something you can call "benign" reflexively — and micro-Hutchinson sign, faint or small periungual hyperpigmentation that only becomes visible under magnification or dermoscopy, which again can arise from either melanoma or benign melanocytic lesions.

Two practical points worth flagging directly for your workflow. First, the panel does not recommend biopsying the periungual skin itself when a Hutchinson sign is present, because that approach can miss the actual melanoma diagnosis — the lesion of interest is the nail matrix, and matrix histopathology is what's needed to confirm or exclude melanoma, not skin adjacent to it. Second, they endorsed a preferred immunohistochemical panel for working up these lesions — Melan-A slash MART-1, PRAME, SOX10, and MITF — as the melanocyte markers of choice. Interestingly, the group also debated abandoning the eponym altogether in favor of purely descriptive language, but deliberately chose to keep "Hutchinson sign" in circulation, reasoning that the named sign is more likely to prompt clinicians to actually pursue further workup than a generic phrase like "periungual hyperpigmentation" would. This is a terminology and diagnostic-framework paper rather than a data study, so there's no results or limitations section to speak of — the actionable takeaway is definitional but genuinely practice-relevant: any periungual pigmentation associated with a melanocytic nail lesion in an adult should be called a Hutchinson sign, should prompt nail matrix biopsy rather than periungual skin biopsy, should be interpreted with the caveat that it does not itself confirm melanoma, and micro-Hutchinson sign under dermoscopy carries the same diagnostic weight and workup obligation as the classic macroscopic finding.

Fourth and last, a multicenter retrospective cohort characterizing exactly how patients die from cutaneous squamous cell carcinoma. The clinical gap is that disease-specific death from cutaneous squamous cell carcinoma is a known but understudied outcome — prior single-center series were small, twenty-one and forty patients respectively, and squamous cell carcinoma deaths are excluded from many national cancer registries entirely, so we've been leaning on limited data despite mortality that reportedly rivals melanoma in aggregate burden.

This was a retrospective chart review across twelve centers — nine US academic sites, one US private practice, one Spanish academic center, and one Brazilian public-private center — spanning tumors diagnosed from nineteen ninety-one through twenty twenty-three. The rationale for a large multicenter retrospective design here is fairly self-evident and the authors imply it directly: disease-specific death is a rare enough event in squamous cell carcinoma that no single center, even a busy academic one, accumulates enough cases to characterize patterns of death with any granularity, so pooling across twelve sites over three decades was necessary just to get a workable sample.

Out of nearly fifteen thousand patients with over twenty-three thousand tumors, about two percent developed disease-specific death, and two hundred seventy-eight of those patients had complete data for analysis. The cohort was mostly male, about three in ten were immunosuppressed, and the vast majority of fatal tumors were on the head or neck. Initial treatment had been Mohs surgery in about a third, standard excision in more than half, and a small fraction had no surgery at all, with about a quarter receiving adjuvant radiation. Now the core finding, which echoes and extends those earlier single-center reports: local recurrence developed in almost half of tumors, and the great majority of patients — about eighty-six percent — eventually developed some form of metastasis, whether in-transit, nodal, or distant, with nodal metastasis being by far the most common route at about two-thirds. Roughly three in ten patients already had metastasis at or within three months of initial presentation, meaning the disease was essentially aggressive from the start, while just over half developed metastasis later in the disease course. The headline mortality pattern: about two-thirds of patients died with locoregional disease rather than distant metastasis, and roughly one in twenty died from persistent local disease without ever metastasizing at all — notably, every one of those persistent-disease deaths was a head and neck tumor. When distant metastasis did occur, lung was the leading site at about four in ten, followed by bone at about one in five. Prognostically, this is grim, fast-moving disease once it turns fatal — ninety percent of these patients died within five years of primary treatment, median survival was about twenty-one months from diagnosis to death, and only about nine months from the point of metastasis to death.

The authors' discussion is appropriately restrained: this confirms, at a much larger scale, what the earlier single-institution studies suggested — that locoregional disease control, not distant spread, is the dominant driver of squamous cell carcinoma mortality. They flag one significant limitation themselves, and it's an important one for interpreting relevance to your current practice: most of these patients were treated before twenty eighteen, before immunotherapy was approved for advanced cutaneous squamous cell carcinoma, so this cohort reflects a pre-immunotherapy natural history, and the authors explicitly call for modern cohorts to see whether these death patterns hold up in the current treatment era. There's also the inherent retrospective, multicenter selection bias to keep in mind — these are referral-heavy academic and specialty practice populations, which likely enriches for more aggressive or higher-risk tumors than community-wide squamous cell carcinoma overall. Practically, though, this is squarely relevant to how you frame surveillance and adjuvant treatment discussions: it reinforces that aggressive locoregional control — surgical margin adequacy, nodal staging, and consideration of adjuvant radiation in high-risk cases — is where the mortality battle is actually being fought for squamous cell carcinoma, since distant metastasis, while it happens, accounts for only about a third of fatal cases. The one-in-twenty patients dying from uncontrolled local disease without ever metastasizing, concentrated entirely in the head and neck, is a pointed reminder that for high-risk head and neck squamous cell carcinoma, local control alone can be the life-or-death variable, independent of systemic spread.

That wraps this February roundup — a hint that checkpoint inhibitors may unevenly benefit known versus unknown-primary Merkel cell carcinoma, strong pooled evidence tying acral lentiginous melanoma to biomechanically stressed sites, a much-needed terminology reset on the Hutchinson sign that should change how you document and biopsy periungual pigmentation, and a sobering multicenter confirmation that locoregional, not distant, disease is what kills most patients with fatal cutaneous squamous cell carcinoma. Thanks for listening.