Welcome back to this month's rundown from the Journal of the American Academy of Dermatology, April twenty twenty-six issue. We've got four pieces today spanning a rare genodermatosis, germline testing strategy in melanoma, a review of the fast-moving adjuvant and neoadjuvant landscape in resectable melanoma, and a SEER-based look at how outcomes vary across Asian American subgroups with cutaneous melanoma. Let's get into it.

First up is an observational study looking at cutaneous leiomyoma severity in hereditary leiomyomatosis and renal cell cancer syndrome — HLRCC for short. As you know, this is the autosomal dominant condition driven by pathogenic variants in fumarate hydratase, presenting with cutaneous leiomyomas, early-onset uterine leiomyomas, and renal cell carcinoma. The cutaneous leiomyomas show up in the vast majority of patients by mid-adulthood and often precede the other manifestations, so they're a genuine diagnostic sentinel. What hadn't been well characterized, though, was what predicts lesion burden, pain, and quality-of-life impact — which matters enormously for counseling and triage in these patients.

This was a cross-sectional study out of the McGill University Health Centre's oncogenodermatology clinic, enrolling seventy-eight genetically confirmed HLRCC patients over about two and a half years. Every patient got a standardized total-body skin exam by a single board-certified dermatologist, and the team captured leiomyoma count and distribution, a zero-to-ten pain score, and Dermatology Life Quality Index results. They then built three separate regression models — a negative binomial model for total lesion count, an ordinal logistic model for pain severity, and a binary logistic model for any quality-of-life impairment — and applied a false-discovery-rate correction since they were running multiple comparisons. This single-clinic design makes sense to me, though the authors don't spell it out explicitly, simply because HLRCC is rare enough that you need a specialized referral population to get any meaningful sample size at all.

On the results: ninety-two percent of patients had at least one cutaneous leiomyoma, with a median of fifteen lesions, ranging as high as over a hundred, concentrated on the trunk and upper limbs. Total lesion count rose modestly but significantly with age — roughly a seventeen percent increase per decade — but wasn't tied to sex or variant type. Pain was reported by about six in ten patients, with touch as the leading trigger, and a quarter had moderate to severe pain. Two things stood out as significant predictors of pain: head and neck location, which conferred about a four-fold increase in odds, and higher overall lesion count, though that per-lesion effect was small on its own. Quality-of-life impairment was reported by roughly half the cohort, mostly mild, with under one in ten reaching moderate impact — and the single strongest predictor of that quality-of-life burden was simply the presence of pain, carrying nearly an eight-fold increase in odds, independent of demographics, genotype, or lesion count.

The authors frame this as one of the largest phenotype-focused looks at leiomyoma morbidity in HLRCC, and it adds real specificity to a literature that previously just said pain was common — now we know roughly how common, where it clusters anatomically, and how directly it translates into life-quality burden. Limitations are the ones you'd expect from a rare-disease cohort: modest sample size that limited more granular modeling, a cohort enriched for frameshift variants and skewed heavily white, and a cross-sectional design that can't establish causality or trajectory over time. Practically, this isn't practice-changing in the sense of giving you a new intervention — notably, ninety percent of these patients had never received any treatment for their lesions — but it is a solid argument for building routine, systematic skin exams and directed pain assessment into HLRCC follow-up, with a particular eye toward head and neck lesions as a marker of higher symptom burden. The therapeutic side — botulinum toxin, capsaicin, gabapentinoids — remains aspirational, flagged by the authors as needing prospective study.

Second, a multicenter prospective cohort study asking a genuinely provocative question: what do you find if you offer germline cancer susceptibility testing to melanoma patients regardless of whether they meet guideline criteria? Current National Comprehensive Cancer Network guidance restricts CDKN2A testing to patients with a personal or family history hitting specific thresholds — multiple invasive melanomas, or a mix of melanoma with pancreatic cancer or astrocytoma. The obvious gap: how many clinically relevant variants are guidelines missing by only testing the people who already look high-risk?

This came out of Mayo Clinic's INTERCEPT program, which prospectively recruited adult patients with any new or active pathologically confirmed cancer diagnosis — no selection by cancer type, personal or family history, stage, or age — and ran germline sequencing on everyone. Two hundred forty-five of those patients had melanoma. The rationale for this unselected, universal-testing design is really the whole point of the study: you cannot test whether guideline-restricted testing under-captures actionable variants if your cohort is already guideline-selected, so they deliberately removed that filter.

What they found: about seven percent of these unselected melanoma patients carried a pathogenic germline variant, and nearly half carried a variant of unknown significance — a notably high rate that's a real practical headache with any broad-panel approach. Of the eighteen patients with a pathogenic variant, two-thirds had a variant with potential clinical utility, about a quarter had trial eligibility implications, and two-thirds had an established management pathway. But here's the more nuanced point: only five of those eighteen variants were actually in genes with established melanoma biology — CDKN2A and MITF. Of those five, three would have been missed entirely by personal or family history criteria — so two out of five true melanoma-risk variants, forty percent, fell outside current guidelines. Overall, about five percent of the whole unselected cohort derived some incremental benefit beyond what guideline-based testing would have caught, and the large majority of all pathogenic-variant carriers overall didn't meet testing criteria — though it's worth noting most of those were incidental findings in genes unrelated to melanoma itself, picked up simply because everyone was tested for everything.

The authors are appropriately measured here. They note their pathogenic-variant and variant-of-unknown-significance rates track with prior unselected-testing literature, and that CDKN2A status remains the strongest known risk factor for a subsequent melanoma — so identifying it, even outside current criteria, could meaningfully shift surveillance intensity for that patient and their family. They also cite prior data that learning this information doesn't seem to increase patient distress. But the limitations are real: no benchmark exists for expected variant-of-unknown-significance rates in unselected melanoma patients specifically, so they can't say whether their signal is inflated; there's no extended follow-up; the cohort is overwhelmingly white; and disease stage skewed toward more advanced presentations, which limits how far you can generalize. Critically, the authors explicitly say they are not advocating for universal germline testing in melanoma — they're flagging the financial and psychological costs directly — and instead call this a signal that alternative, smarter targeting strategies deserve investigation. So take this as genuinely interesting and hypothesis-generating, underscoring that current criteria have real gaps, but not something to translate into ordering panels on every melanoma patient tomorrow.

Third, a clinical review synthesizing where adjuvant and neoadjuvant therapy stand for resectable cutaneous melanoma — this is a review, not original data, so it's walking us through the trial landscape rather than generating new results. The framing is that immune checkpoint inhibitors and MAPK-pathway-targeted therapy have transformed outcomes over the past fifteen years, but a meaningful subset of patients still recur, and the field has been actively testing whether giving immunotherapy before surgery — neoadjuvant — beats the old standard of surgery followed by adjuvant therapy. The biological rationale the authors lay out is straightforward: treating before resection lets checkpoint blockade prime a T-cell response against a tumor that still contains its neoantigens and tumor-infiltrating lymphocytes, rather than removing that antigen source first.

On the adjuvant side for stage three and four disease, the review runs through the foundational trials. ECOG 1609 established that ipilimumab at three milligrams per kilogram beat high-dose interferon with a real survival benefit, while the ten-milligram dose added toxicity without added efficacy. KEYNOTE-054 showed adjuvant pembrolizumab after resection nearly doubled five-year recurrence-free survival relative to placebo — about fifty-five percent versus thirty-eight percent — a genuinely large and clinically meaningful gap, with a real but manageable increase in serious toxicity. CheckMate 238 then pitted nivolumab against ipilimumab head-to-head and found nivolumab cut the relative risk of recurrence by roughly a quarter, with dramatically less severe toxicity — under fifteen percent versus nearly half the patients on ipilimumab — which is why nivolumab, not ipilimumab, became the adjuvant PD-1 backbone.

The neoadjuvant story is where things get more interesting for surgical planning. SWOG 1801 randomized stage three-B through four-C patients to neoadjuvant-plus-adjuvant pembrolizumab versus adjuvant-only, and found a striking two-year event-free survival advantage — about seventy-two percent versus forty-nine percent — with comparable toxicity between arms. That trial alone shifted guidelines toward recommending neoadjuvant therapy for clinical stage three disease. The OpACIN and OpACIN-Neo trials then worked out dual-checkpoint dosing: OpACIN was a small feasibility study showing neoadjuvant combination therapy drove more robust tumor-specific T-cell expansion than adjuvant dosing, and OpACIN-Neo tested three dosing schedules, finding that lower-dose ipilimumab combined with standard-dose nivolumab delivered pathologic response rates in the high seventies percent range with a much better safety profile — only about one in five patients with severe toxicity, compared to forty to fifty percent in the higher-ipilimumab arms, one of which even had a treatment-related death from encephalitis. That safety data is why the lower-ipilimumab, standard-nivolumab combination became the regimen carried forward.

That regimen was then tested definitively in NADINA, a phase three trial in macroscopic resectable stage three disease, comparing two cycles of neoadjuvant ipilimumab-nivolumab followed by surgery, against surgery followed by a full year of adjuvant nivolumab — with a clever response-adapted design where patients achieving a pathologic complete or near-complete response after neoadjuvant therapy skipped further adjuvant treatment altogether. Twelve-month event-free survival was about eighty-four percent in the neoadjuvant arm versus fifty-seven percent in the adjuvant-only arm — a large, clinically meaningful difference. And critically, pathologic response after neoadjuvant therapy stratified outcomes beautifully: recurrence-free survival was ninety-five percent in major pathologic responders, dropping to seventy-six percent with partial response, and fifty-seven percent with no response — meaning the surgical specimen itself becomes a powerful prognostic readout that could guide who really needs further systemic therapy. Toxicity was higher upfront in the neoadjuvant arm, but surgical complication rates were essentially identical between groups, which is a reassuring point for anyone worried that pre-treating would complicate the operation itself.

Beyond checkpoint blockade, the review touches on lymphocyte-activation-gene-3 inhibition — relatlimab paired with nivolumab, already approved in the metastatic setting — tested in a small thirty-patient neoadjuvant trial with a pathologic complete response rate over fifty percent and no severe immune-related toxicity in the neoadjuvant phase, though this is early and small. Oncolytic virus therapy also gets a mention: talimogene laherparepvec given neoadjuvantly before surgery nearly doubled two-year recurrence-free survival compared with surgery alone — about thirty percent versus seventeen percent — a real but modest absolute effect, and a newer oncolytic agent, vusolimogene oderparepvec, is in an ongoing trial aiming to prevent sentinel node metastases, with no results yet.

Finally, the review covers the extension of this whole paradigm earlier, into stage IIb and IIc disease — which the authors point out actually accounts for a disproportionate share of melanoma deaths given how much more common it is than later-stage disease. KEYNOTE-716 tested adjuvant pembrolizumab against placebo in resected high-risk stage two disease and found a significant, clinically meaningful improvement in distant-metastasis-free survival at three years — roughly eighty-four percent versus seventy-five percent — supporting that the adjuvant checkpoint-blockade benefit isn't confined to later-stage disease.

For you as a Mohs surgeon and dermatologic oncologist, the practical upshot here is genuinely practice-relevant, even if the surgery itself typically sits with surgical oncology: neoadjuvant ipilimumab-nivolumab ahead of resection is now a data-backed, guideline-endorsed pathway for macroscopic stage three disease, pathologic response on the resected specimen is becoming a real decision point for whether adjuvant therapy continues, and the field is actively moving checkpoint blockade earlier into stage II. That matters for how you counsel patients, how you coordinate referral timing with medical oncology, and how you interpret pathology reports coming back from neoadjuvant-treated specimens. The LAG-3 and oncolytic virus data remain investigational and shouldn't yet change your day-to-day counseling.

Last, a retrospective SEER-database study looking at variations in outcomes for cutaneous melanoma among Asian American patients. The setup here is that Asian Americans are known to have lower melanoma incidence but disproportionately worse outcomes, and prior work has already shown some subgroup-level differences — but the authors point out that heterogeneity within Asian American subgroups specifically hasn't been well characterized, and "Asian American" as a single category may be masking very different risk profiles.

Using the SEER seventeen registries from two thousand to twenty twenty-two, they identified first primary cutaneous melanomas in non-Hispanic Asian American patients by histology and site codes, then ran Kaplan-Meier and stage-stratified multivariable Cox models for overall and cancer-specific survival. A registry-based design is really the only way to get here — this is a rare cancer within an already low-incidence population, disaggregated into seven ethnic subgroups, so you need a large population-based dataset just to have enough events per subgroup to say anything statistically meaningful, though that's my read on why they chose it rather than something the authors state outright.

They landed on just over seventeen hundred patients. The subgroup differences in presentation were notable: Korean and Filipino patients were most often diagnosed at sixty-five or older, while Chinese patients had the highest share diagnosed under thirty-five. Filipino patients had the highest rate of distant-stage disease at diagnosis, at sixteen percent. Vietnamese patients stood out with the thickest, most frequently ulcerated tumors — over a third had Breslow depth beyond four millimeters, and over four in ten had ulceration — versus Japanese patients at the low end of both. Chinese patients skewed toward higher household income. By the end of follow-up, about eighteen percent of the whole cohort had died from melanoma and sixteen percent from other causes.

The survival findings are the clinically important part. In stage-stratified adjusted models, Filipino patients had significantly worse overall survival than Chinese patients — roughly a third higher hazard — while Vietnamese patients had significantly worse cancer-specific survival, with about a sixty-six percent higher hazard relative to Chinese patients. Both effects held up after adjusting for stage, so this isn't purely explained by presenting later. Male sex carried an independent, roughly thirty percent higher hazard for both overall and cancer-specific mortality, and diagnosis in the more recent decade, twenty-eleven to twenty twenty-two, was associated with meaningfully better cancer-specific survival, consistent with the immunotherapy era.

The authors' interpretation is that the Vietnamese pattern — thicker, more ulcerated tumors driving worse cancer-specific survival — likely reflects delayed detection or barriers to dermatologic care, while the Filipino pattern, where overall survival is worse but the cancer-specific effect was not statistically significant after adjustment, may point more toward competing mortality risks than melanoma biology itself. The obvious limitation, which the authors acknowledge directly, is that SEER doesn't capture ultraviolet exposure, genetic susceptibility, or skin phototype — so none of the actual mechanistic drivers of these disparities can be tested here, only described. For practice, this is useful population-level context rather than a management-changing finding: it reinforces disaggregating Asian American patients rather than treating them as one risk category, and argues for a lower threshold for biopsy and closer surveillance in Filipino and Vietnamese patients specifically, alongside culturally tailored outreach — genuinely useful epidemiologic signal, but hypothesis-generating rather than something that rewrites a guideline.

That's the April lineup — a rare-disease phenotype study reinforcing systematic skin exams in HLRCC, a germline-testing study that pokes a real hole in current melanoma screening criteria without yet justifying universal testing, a review mapping how neoadjuvant checkpoint blockade is reshaping resectable melanoma management, and a SEER analysis reminding us that Asian American melanoma outcomes are anything but uniform. Thanks for listening, and we'll see you next month.