Welcome to this November twenty twenty-six review from Dermatologic Surgery. We've got four pieces on the docket this month — a SEER registry study on amelanotic melanoma, a commentary on a novel wound matrix for post-Mohs defects, a cross-sectional look at the quality of Mohs content on TikTok, and a case report on a nasty bleeding complication in a patient on a JAK inhibitor. Let's get into it. First up is an original research article: "Amelanotic Melanoma is Associated With Higher Rates of Metastasis at Presentation and Disease-specific Death: A National Cancer Registry Study." The clinical problem here is one you already know in your gut — amelanotic melanoma is a diagnostic trap, it doesn't look like melanoma, and it gets biopsied late. The last population-level data on whether that translates into worse outcomes is over a decade old, so these authors went back to SEER to ask, in the modern era, is amelanotic melanoma still showing up with more metastatic disease and dying of it more often than pigmented melanoma. Methodologically, this is a registry-based retrospective cohort, pulling from SEER-17 using histology codes, from the year two thousand through twenty twenty-two. That gave them a lopsided but very large comparison: about four hundred forty-five thousand melanotic melanomas against only about seventeen hundred amelanotic cases — which itself tells you something about how rare and probably underrecognized this subtype is. For the key analyses, they used two different statistical tools, and it's worth pausing on why they picked each one. For the question of whether amelanotic melanoma predicts metastasis already present at diagnosis, they used multivariable logistic regression — that's your standard tool whenever the outcome is binary, metastatic or not, and you want to isolate the effect of one variable, amelanotic versus melanotic, while holding age, sex, income, and anatomic site constant. For survival, though, they shifted to something called Fine-Gray competing risk regression rather than a standard Cox model. Here's why that distinction matters: in a population of mostly older melanoma patients, a lot of people are going to die of heart disease or something unrelated before melanoma ever kills them. A standard Cox model, or a simple Kaplan-Meier curve, treats those non-melanoma deaths as if the patient was simply "censored" — lost to follow-up — which can distort your estimate of melanoma-specific mortality risk. Fine-Gray regression instead explicitly models the competing event, so the hazard ratios it produces — called subdistribution hazard ratios — more accurately reflect the actual probability of dying specifically from melanoma in a population where other causes of death are competing for the same patients. That's the right tool when your outcome is disease-specific death in an elderly cohort, and it's a nice one to have in your back pocket for any registry analysis you read going forward. Onto results. Patients with amelanotic melanoma were notably older — about seventy-one percent were over sixty, compared to about sixty percent in the melanotic group — and they had more tumor burden on the head and neck and extremities, with more distant spread at diagnosis. After adjustment, amelanotic histology carried about a two-fold increase in the odds of presenting with nodal or distant metastasis already in place, and that was statistically significant. Younger age and male sex also independently raised the odds of metastatic presentation, for what it's worth. On the mortality side, amelanotic melanoma carried a significant and clinically meaningful increase in disease-specific death risk — roughly forty percent higher risk relative to melanotic melanoma, even after adjusting for stage and other variables. Lower household income, head and neck location, and of course having distant or nodal disease at baseline were all independently tied to higher disease-specific death as well — and that metastatic disease variable carried an enormous hazard, roughly an eight-fold increase, which is no surprise but confirms the model is behaving as expected. Interestingly, radiation and systemic therapy were both associated with higher death hazards too, but the authors are upfront that this is very likely confounding by indication — sicker patients get radiated and get systemic therapy — and they explicitly warn against reading that as cause and effect. The authors' interpretation is that this gap probably reflects a combination of inherently more aggressive tumor biology in amelanotic lesions and ongoing delayed recognition — the lesion simply doesn't trigger the same index of suspicion clinically. They're appropriately honest about the limitations: this is retrospective registry data, SEER's staging detail is coarse, and critically, SEER extracts the word "amelanotic" directly from the free text of pathology reports and only allows one morphology code per tumor, so there's real potential for misclassification in both directions. For your practice, this isn't practice-changing in a procedural sense — there's no new surgical technique here — but it is a legitimate reinforcement of vigilance. The actionable piece is diagnostic: nonpigmented, nonspecific papules or plaques, especially in older patients, deserve a lower threshold for biopsy, and this data says that threshold still isn't low enough across the population, more than a decade after the original studies raised the same alarm. Next is a commentary — not a primary study, but a critical appraisal — on a trial of a microporous annealed particle wound matrix for post-Mohs defects, often shortened to MAP. The commentary authors are walking us through and contextualizing a separate first-in-human randomized trial by Hau and colleagues, so think of this as expert guided interpretation rather than new data. The underlying trial randomized forty patients, two to one, to a single application of this flowable MAP matrix versus a conventional hydrocolloid dressing, across three centers, for clean full-thickness post-Mohs defects suited to healing by secondary intention. It was well designed for an early-phase study — an active comparator rather than a placebo, blinded photographic review by five independent Mohs surgeons, and scar assessment around six months using a validated tool. The primary safety endpoint was clean, with no serious adverse events. On efficacy, there was no significant difference in time to seventy-five percent granulation or complete wound closure. There was a numerical trend toward faster wound-bed readiness with MAP — about fifteen days versus about twenty-nine days with the hydrocolloid — but that didn't reach statistical significance, so it's a hypothesis, not a finding. Where MAP did pull ahead with statistical significance was in blinded observer-rated scar quality at roughly six months, though patient-reported scar satisfaction showed no difference between groups. Pain was low and comparable in both arms. The commentary's real value is in the critique. They point out that this is a genuinely rare thing in the skin-substitute world — a randomized, actively controlled trial — because most of the evidence supporting the dozens of products on the market, collagen matrices, placental products, fish-derived xenografts, human allografts, is extrapolated from chronic lower-extremity wounds or comes from retrospective series with no head-to-head comparisons. They flag that this evidence gap isn't academic — Medicare Part B spending on skin substitutes exploded from roughly two hundred fifty million dollars in twenty nineteen to over ten billion dollars in twenty twenty-four, which is part of why CMS moved to a flat per-square-centimeter payment rate starting this year. Against that backdrop, they highlight something practically important: MAP was applied once, while the hydrocolloid control needed dressing changes every few days for two weeks and then weekly after — so even without a formal cost-effectiveness analysis, there's a plausible burden-reduction story here worth testing properly. But they're equally clear about the limits: small sample, unequal two-to-one randomization, open-label design, an exploratory post hoc endpoint for wound-bed readiness, and a population restricted to small, clean, non-irradiated, non-immunosuppressed wounds — so none of this generalizes to your harder cases. The takeaway: interesting, not yet actionable. This is a methodological stepping stone showing that comparative, randomized wound-care research is feasible in dermatologic surgery, which is valuable in itself, but it's not grounds to change what you reach for in your own secondary-intention defects based on this trial alone. Third is an original cross-sectional study: "Cutting Through TikTok: Assessing Quality and Usefulness of Mohs Videos." The gap they identify is straightforward — YouTube content on Mohs surgery has been evaluated before, but nobody had looked at TikTok, despite it now being a dominant source of health information, particularly for younger patients who increasingly use it as a search engine outright. Methodologically, they did a cross-sectional content analysis. They searched TikTok with Mohs-related keywords, used a power calculation to justify a sample of one hundred twelve videos, and had two independent raters score each video two ways: the validated fifteen-item DISCERN instrument, which grades health information quality on trustworthiness, balance, and reliability, and a custom twenty-item "usefulness" checklist the authors built directly from American College of Mohs Surgery and American Academy of Dermatology patient materials, where points were only awarded if information was both present and accurate. For the statistics, they used multivariable linear regression to see whether continuous video features — length, views, likes, comments — predicted the quality and usefulness scores, and they used ANOVA to compare mean scores across categorical groups, like video type or uploader identity. It's worth a quick word on why ANOVA is the right tool there: when you're comparing means across three or more groups — say, educational versus promotional versus personal-testimony videos — running a string of separate t-tests between each pair inflates your false-positive rate. ANOVA instead tests all the group means simultaneously by partitioning the total variance into between-group and within-group components, giving you one overall significance test before you drill down into which specific groups differ. That's exactly what they did here to then isolate, for instance, that educational videos significantly outscored personal testimonies. The results are not flattering. Across one hundred twelve videos representing nearly twelve million cumulative views, most were uploaded by patients or the general public rather than clinicians, and the overwhelming majority — about eighty percent — were personal testimonies rather than educational content; only about one in nine was genuinely educational. Quality was poor across the board: the mean DISCERN score corresponded to roughly two and a half out of five, and not a single video reached the "good" or "high quality" tier. For comparison, the American College of Mohs Surgery's own benchmark patient-education video scored about four and a half out of five — a stark gap between professional and user-generated content. Usefulness followed the same pattern, landing in the "moderate" range overall. The only variable that reliably predicted better scores on either measure was video length — longer videos scored modestly higher on both quality and usefulness — while views, likes, and comments had no significant relationship to quality at all, which is itself a useful finding: popularity on this platform tells you nothing about accuracy. Educational videos, unsurprisingly, scored significantly higher than personal testimonies and promotional content on both measures. The authors note this mirrors prior YouTube-based studies using the same DISCERN tool, suggesting the deficit in Mohs-related content quality is a platform-agnostic problem, not something specific to TikTok's format or audience. They're appropriately cautious that this is a snapshot in time from a single search date and assesses only spoken and visual video content, not captions, so some nuance may be missed. Practically, this isn't going to change how you perform surgery, but it is directly relevant to your patient counseling and any outreach your practice does. The clear signal is that patients are absorbing mostly anecdote-driven content with real information gaps, and that when clinicians or institutions do produce educational content, it measurably outperforms the baseline. If your practice has any interest in patient-facing content, this is reasonably strong evidence that it's a vacuum worth filling. Last is a case report: "Postoperative Hemorrhage After Mohs Micrographic Surgery in a Patient on Ruxolitinib." No statistics to teach here, just a cautionary clinical vignette worth filing away. An eighty-year-old man with a history of melanoma and keratinocyte carcinoma presented with a squamous cell carcinoma of the left cheek. He was on ruxolitinib, ten milligrams twice daily, for a JAK2-positive myeloproliferative disorder, plus low-dose aspirin for coronary disease. His preoperative counts already showed mild anemia and thrombocytopenia, with a platelet count around eighty-five thousand. He underwent a two-stage Mohs excision, final defect about four by two centimeters, closed with an island pedicle advancement flap, and the surgery itself was uneventful with no excessive intraoperative bleeding. The trouble started on postoperative day one, with active oozing at the flap margins that responded to pressure. Then on day six it recurred more seriously — platelets had dropped further to the high fifty-thousands — with multiple small hematomas requiring evacuation. Over the following weeks he had repeated hematoma reformation despite gel foam, pressure dressings, and holding his aspirin, eventually requiring platelet transfusion. He declined oral tranexamic acid. The bleeding only resolved once ruxolitinib itself was held, after which platelets climbed to the low one hundred twenty-thousands and the bleeding stopped within about a week. The wound ultimately healed by fourteen weeks. The teaching point here is that ruxolitinib's bleeding risk doesn't track cleanly with platelet count alone. This patient's counts, while reduced, stayed above the conventional fifty-thousand threshold typically cited as safe for minor cutaneous surgery, yet he still bled persistently and needed the drug held to stop it. The authors cite a prior published case with a similar pattern, where platelet function testing confirmed actual drug-induced platelet dysfunction despite adequate counts — suggesting the bleeding risk from ruxolitinib is at least partly independent of the platelet number itself, likely related to its effect on growth factor and cytokine signaling needed for normal hematopoiesis and platelet function. They also flag a dosing nuance worth remembering: per the drug's own label, a platelet count in the fifty-to-one-hundred-thousand range should correspond to a lower dose, five milligrams twice daily, but this patient was actually on ten milligrams twice daily at the time of surgery — a mismatch that may have compounded his risk. Since there's no formal guideline for perioperative JAK inhibitor management, the authors' practical recommendations are to loop in the prescribing oncology team before surgery, obtain a current complete blood count and consider platelet function testing, consider dose reduction or temporary interruption for significant cytopenias, plan for lower-risk and simpler reconstructive options where feasible, have a low threshold to use tranexamic acid, and commit to close postoperative monitoring given how delayed and recurrent this bleeding pattern can be. The practical takeaway here is genuinely actionable: don't let a platelet count above fifty thousand give you false reassurance in a patient on ruxolitinib or similar JAK inhibitors. Treat it as a distinct bleeding-risk category warranting its own perioperative conversation with the patient's hematology or oncology team, regardless of the numbers on the CBC. That wraps this November roundup — a registry study reinforcing vigilance for amelanotic melanoma, a sober look at where skin-substitute evidence still falls short, a reality check on the patient education content actually reaching your patients online, and a reminder that JAK inhibitors can bleed independent of the numbers in front of you. Thanks for listening, and we'll see you next month.