Welcome to this January twenty twenty-six run-through of the Journal of the American Academy of Dermatology. Four brief reports caught our attention this month, and they cluster nicely around things you deal with every single week in clinic — margin control in a rare adnexal cancer, the sunscreen safety noise your patients ask you about, perioperative risk in patients on biologic immunosuppression, and how you're prescribing pain control after your own procedures. Let's get into it.

First up is a retrospective registry study looking at clinical factors and surgical treatment in sebaceous carcinoma, with a specific eye toward whether Muir-Torre syndrome status changes the picture. This is genuine original research, so we'll walk it through in full.

The background here is straightforward: sebaceous carcinoma is an adnexal neoplasm that sometimes travels with Muir-Torre syndrome, the allelic variant of Lynch syndrome, and surgical resection with margin control remains the gold standard. What hadn't been well characterized is how survival outcomes differ by Muir-Torre status once you control for the usual prognostic variables, and specifically whether the surgical modality you choose — Mohs versus wide local excision — actually moves the needle on survival rather than just recurrence.

Methodologically, the authors pulled from the SEER-17 registry, which is the standard move when you need population-scale numbers for a tumor this rare — you simply cannot power a survival analysis on sebaceous carcinoma from a single center. Because SEER doesn't capture genetic testing results, Muir-Torre status couldn't be confirmed directly; instead they inferred it using the Mayo risk-scoring system, the same approach used in prior work in this space, which is worth flagging honestly as a real limitation baked into the design rather than something to gloss over. They ran univariate screening followed by multivariable Cox models for overall survival and a competing-risk model for sebaceous-carcinoma-specific survival, which is the right choice when patients are dying of other causes at meaningful rates, as they do in an elderly cancer population — it keeps you from crediting a treatment with a survival benefit that's really just other-cause mortality diluting the denominator.

Across four thousand four hundred fifty patients, the independent poor prognostic factors were about what you'd expect and are worth having at your fingertips: advancing age, tumor size at or above three centimeters, and regional or distant disease at diagnosis all significantly worsened both overall and cancer-specific survival, with distant disease carrying roughly a seven to eight-fold increase in cancer-specific mortality risk — clinically the single biggest driver in the model. Periocular location was actually associated with better overall survival, plausibly a lead-time or detection-bias effect given how early periocular lesions tend to present. Now the part that matters most for you: any surgical treatment was associated with meaningfully improved overall and cancer-specific survival, and in a head-to-head sub-analysis, Mohs surgery outperformed wide local excision on both fronts — a meaningful reduction in cancer-specific death, not just a statistically significant but trivial difference. The authors are explicit about why they think this happens: wide local excision only ever samples a small fraction of the surgical margin, whereas Mohs affords complete margin assessment, and in a tumor prone to subclinical extension and skip areas, that's plausibly what's driving the survival difference rather than any inherent tumor biology. Sentinel lymph node biopsy was linked to better overall survival but not cancer-specific survival, which the authors appropriately caveat given how few patients actually underwent the procedure. And Muir-Torre status itself showed a trend toward worse overall survival — consistent with the known excess of visceral malignancy in these patients — but no significant effect on sebaceous-carcinoma-specific survival, meaning the syndrome doesn't appear to make the skin cancer itself more aggressive, it just comes with other cancer risks riding along.

Limitations are the ones inherent to any SEER-based retrospective study: tumor size was missing in over half of patients, which weakens that variable's precision; Muir-Torre status is inferred rather than confirmed; and the small number of sentinel node procedures makes that finding hypothesis-generating at best.

Practical takeaway — this one is genuinely practice-relevant. It reinforces that complete circumferential margin assessment, meaning Mohs, should be your default for sebaceous carcinoma regardless of Muir-Torre status, and it gives you an actual survival-based argument, not just a local-recurrence argument, to bring to tumor boards and to referring physicians when justifying Mohs over wide excision for this tumor type.

Second article, a brief report, but a nice piece of public-health epidemiology relevant to every conversation you have at the front desk about sunscreen. This one asks a simple question: after the twenty twenty-one benzene contamination scare and the ongoing reef-toxicity bans on oxybenzone and octinoxate, did patients actually change their sun-protective behavior?

The design is a repeated cross-sectional analysis using NHANES, the National Health and Nutrition Examination Survey, across six cycles from 2009 through 2023. NHANES is the right tool here because it's nationally representative and has asked the same self-reported sun-protection questions across cycles, letting you track a real behavioral trend rather than a convenience sample. They categorized people as consistent users — always or most of the time — versus inconsistent, for three behaviors: sunscreen, shade-seeking, and long-sleeve clothing, and they specifically compared the twenty seventeen to twenty twenty period against twenty twenty-one to twenty twenty-three, since that's the window bracketing the controversies.

The result is almost counterintuitive and worth remembering the next time a patient cites a scary headline as a reason to skip sunscreen: consistent sunscreen use had been fairly flat for most of the study period, dipped slightly to about one in five respondents in twenty seventeen to twenty twenty, and then rose significantly to roughly one in three respondents in twenty twenty-one to twenty twenty-three — nearly a two-fold increase in the odds of consistent use, and this was clearly clinically meaningful, not a statistical artifact. Shade-seeking stayed flat around one in three throughout, with no significant shift, and long-sleeve clothing use, already low at roughly one in eight to one in seven people, showed no significant change either.

The authors' interpretation is that despite the media noise, public trust in sunscreen didn't erode — if anything, awareness and possibly reformulation, marketing, and a preference for products perceived as safer likely pushed usage up rather than down. They also point out that other data have shown no real link between sunscreen use and elevated blood benzene levels, with things like traffic pollution and cigarette smoke being far more plausible contributors to background benzene exposure. The obvious limitation is that this is all self-reported survey data, so recall and social-desirability bias are baked in — people may simply be more comfortable saying they wear sunscreen now than they were reporting shade-seeking honestly.

Practically, this is reassuring but not something that changes your counseling — it's good ammunition to reassure anxious patients that the population-level trend is toward more protection, not less, but it doesn't tell you anything new about what to actually recommend.

Third, a retrospective cohort study on the safety of rituximab in the perioperative dermatologic surgery setting — directly relevant to any of you doing Mohs on patients maintained on B-cell depleting therapy.

The clinical gap here is timing. Rheumatology and orthopedic surgery already have a consensus recommendation — elective joint arthroplasty should happen four to seven months after the last rituximab infusion to lower infection risk — but there's been no equivalent guidance for dermatologic surgery, despite it also being elective, nonemergent surgery in a population that's often immunosuppressed for hematologic malignancy or autoimmune disease.

Methodologically, this is a single academic center study that used TriNetX as a cohort-discovery tool to flag potential patients by procedure codes, then had every chart manually reviewed for actual inclusion and outcome data — the authors don't spell this reasoning out explicitly, but combining an automated database query with manual chart adjudication is a sensible way to get accurate, granular outcome and timing data that a pure claims-based query would miss, at the cost of a much smaller final sample. Of a hundred sixty-one flagged patients, ninety-four met criteria after screening.

The cohort was mostly on rituximab for hematologic malignancy, about a third for autoimmune disease, nearly half were undergoing Mohs surgery specifically, and infections of some kind — mostly respiratory tract infections — were common over the one-year follow-up, alongside a fairly high rate of hospitalization and a notable death rate, though those are almost certainly reflecting the burden of the underlying hematologic or autoimmune disease rather than the surgery itself. The number that actually matters for your consent conversation is the surgical site infection rate: only two patients, about two percent, developed a documented surgical site infection, which the authors point out is essentially in line with the baseline surgical site infection rate in dermatologic surgery generally, quoted as under two percent in prior work. Those two infections occurred in patients operated on 127 and 273 days after their last infusion — so no clear signal that earlier surgery after rituximab was riskier.

The authors' conclusion is appropriately measured: patients on rituximab appear to carry a comparable surgical site infection risk to the general dermatologic surgery population, and they explicitly call for larger, multi-institutional cohorts to determine whether timing relative to infusion actually matters. The limitations are the obvious ones for a single-center series of ninety-four patients — underpowered to detect a modest effect of timing, and not generalizable beyond this institution's practice patterns.

Practical takeaway: this is reassuring, real-world safety data that should lower your threshold for proceeding with medically necessary skin cancer surgery in patients on rituximab without insisting on the same rigid four-to-seven-month window used in joint arthroplasty — but it's not yet strong enough evidence to formally rewrite a timing protocol; think of it as permission to use clinical judgment rather than a new mandate.

Fourth and last, a large database study on opioid prescribing patterns after Mohs surgery, using the TriNetX network — this one has real quality-improvement teeth for anyone still writing for oxycodone reflexively after a flap.

Background: dermatologic surgeons already account for the majority of opioid prescriptions written within dermatology, and there's good evidence, including a randomized trial cited in the piece, that acetaminophen-based regimens control post-Mohs pain adequately without opioids. What was missing was a granular, decade-long look at who is still getting opioids and whether certain populations are being over- or under-treated.

The design pulled everyone undergoing Mohs surgery — using the standard first-stage and multistage CPT codes — across TriNetX from 2015 through 2025, tracked opioid prescriptions filled within fourteen days of surgery using RxNorm codes, and ran Cox regression adjusting for known risk factors for prolonged postoperative opioid use, like prior substance use disorders and chronic pain diagnoses. A database of this scale is really the only way to detect modest but real disparities in prescribing across sex and race, since a single-center series would never have the numbers.

Across three hundred fifty-six thousand Mohs patients, thirteen percent received an opioid prescription within two weeks of surgery — and encouragingly, that rate fell from about sixteen and a half percent in the 2015-to-2020 era down to roughly ten percent in 2020-to-2025, a meaningful and clinically real decline, consistent with the broader trend toward opioid-sparing protocols in dermatologic surgery. But the disparities are the headline here: male patients had a modestly but significantly higher likelihood of receiving an opioid, and non-White patients had a meaningfully higher likelihood as well — though the authors are careful to flag that non-White patients made up only about two percent of this cohort, so that estimate should be interpreted with real caution rather than over-read. Patients with a documented history of opioid or alcohol use disorder were actually less likely to receive an opioid prescription, suggesting surgeons are already exercising judgment in that population. The one finding that should genuinely give you pause: patients sixty-five and older were not statistically less likely to receive an opioid than younger adults, despite this being exactly the population where opioids carry the highest risk of falls, delirium, and mortality. Unsurprisingly, comorbidities like osteoarthritis, chronic pain, fibromyalgia, migraine, and pulmonary disease, along with preoperative benzodiazepine or antidepressant use, were all associated with a higher likelihood of receiving an opioid — largely intuitive, but useful to have quantified.

The authors' discussion is direct: despite overall improvement, older adults are not being spared opioid exposure the way they probably should be, and they advocate for more consistent use of acetaminophen and ibuprofen as first-line agents across the board, not just in flagged high-risk groups. Limitations are the usual ones for a claims-based network study — no data on indication or dose adequacy, no ability to confirm whether the prescription was actually filled or taken, and the small non-White subgroup limiting confidence in that specific estimate.

Practical takeaway: the encouraging decade-long decline is nice context, but the actionable piece is the elderly finding — this is a genuine opportunity to formalize an opioid-sparing default in your own practice for your oldest patients specifically, rather than assuming age alone is already driving conservative prescribing, because this data suggests it isn't.

That wraps our four articles for January. The throughline across all of them is really about matching the intervention to the actual risk — complete margin control for a tumor that behaves unpredictably, judicious perioperative timing rather than rigid rules for immunosuppressed patients, and opioid-sparing analgesia especially in the patients least equipped to tolerate opioid side effects. Thanks for listening, and we'll see you next month.