Welcome to this October 2026 edition of our journal review, covering the Journal of the American Academy of Dermatology. We've got four pieces for you today — a letter and its response forming a methodologic back-and-forth over acral melanoma surgery, a National Cancer Database brief report on sex disparities in dermatofibrosarcoma protuberans care, and a Medicare claims analysis on radiation trends for keratinocyte carcinoma. Let's get into it.

First up is a paired exchange — a letter to the editor and the original authors' response — both centered on the Frankel et al retrospective cohort study comparing Mohs micrographic surgery to conventional excision for hand and foot melanomas. You'll recall that original study found Mohs associated with fewer reconstructions performed on positive margins, fewer repeat surgeries, and shorter time to definitive reconstruction. This exchange is really a masterclass in how to critique — and defend — a retrospective comparative design, so it's worth walking through carefully even though neither piece is itself a new study.

The letter, from Xu and colleagues, raises two core methodologic concerns. First, baseline imbalance: the conventional excision cohort had deeper, more advanced tumors than the Mohs cohort, and the letter writers argue that this disparity in tumor burden could confound comparisons of secondary outcomes like repeat excision rates, even though the primary positive-margin outcome didn't differ by Breslow depth subgroup within the conventional excision arm itself. Their proposed fix is a textbook one — propensity score matching on Breslow depth and tumor stage to balance the two groups before comparing outcomes, which would strengthen causal inference beyond what an unadjusted cohort comparison can offer.

Second, and probably the sharper point, they flag a definitional asymmetry: positive margins for conventional excision were defined pathologically, as melanoma within a millimeter of ink or requiring re-excision, while for Mohs a positive margin was simply defined as discontinuing the procedure without clearance. Since no Mohs case could, by definition, proceed to reconstruction with a known positive margin, the letter writers argue the real contrast being measured is delayed versus immediate margin assessment, not a head-to-head comparison of final pathologic clearance.

The response, from Kapoor, Miller, and the original author group, pushes back thoughtfully rather than dismissively. On the confounding point, they bring data: among invasive melanomas with positive margins after conventional excision, the vast majority — about seven in ten — were positive at the peripheral margin only, meaning the positive margins were overwhelmingly driven by subclinical in-situ spread rather than the invasive tumor component itself. Their argument is that this supports prior literature showing subclinical spread tracks with anatomic location, not tumor stage, which blunts the confounding concern even without formal matching. On the definitional point, they essentially concede the conceptual distinction — yes, Mohs allows immediate margin confirmation before reconstruction — but reframe it as a feature, not a flaw: regardless of whether a surgeon delays conventional excision reconstruction to wait for pathology or proceeds and then has to repeat surgery, the patient experiences the same inconvenience and delay. And they add a data point that really lands: among patients needing repeat excision after conventional excision, about one in six still had persistently positive margins on the second attempt, and one patient needed three surgeries before achieving clearance. Both sides agree a prospective, propensity-matched, multicenter study would be the ideal next step, but neither pretends the current retrospective data settles the question definitively.

The practical takeaway here is less about new evidence and more about how to read comparative retrospective surgical outcomes studies critically. For your own practice, the underlying clinical signal — that immediate margin assessment before reconstruction avoids the real-world problem of operating on bad margins — remains sound and clinically intuitive regardless of which side of this exchange you find more persuasive. But if you're citing this literature to colleagues or in multidisciplinary tumor board discussions, it's worth knowing the nuance: the headline positive-margin comparison is partly an artifact of differing definitions, and the degree of residual confounding from tumor stage differences is a legitimate open question pending matched or prospective data.

Next, a brief report using the National Cancer Database to look at sex-based disparities in Mohs micrographic surgery utilization and margin outcomes for dermatofibrosarcoma protuberans, or DFSP. The background here is straightforward: DFSP is locally aggressive, complete excision is the backbone of cure, and margin status drives recurrence risk — but nobody had really looked at whether sex influences how that surgical care gets delivered.

This is a retrospective cohort study pulling adult DFSP patients from the National Cancer Database between 2004 and 2023, identified by histology and site codes, with three outcomes of interest: Mohs utilization, positive margin rates, and overall survival. Methodologically, this is a nice opportunity to talk about why they reached for the tools they did. For the two binary outcomes — getting Mohs or not, having a positive margin or not — they used multivariable logistic regression, which models the odds of a yes-or-no outcome as a function of several predictors simultaneously, letting you isolate the independent association of sex while holding age, insurance status, facility type, tumor size, and so on constant. That's the right tool here because you can't just compare raw margin-positivity rates between men and women without accounting for the fact that, as the data show, men in this cohort also had larger tumors and more head-and-neck primaries — factors that independently affect margin positivity regardless of sex.

For survival, they used the combination you'd expect: Kaplan-Meier curves with a log-rank test to visualize and compare raw survival between sexes, followed by multivariable Cox proportional hazards regression to see whether sex remained an independent predictor of mortality after adjusting for the same covariates. The Cox model is doing something subtly different from logistic regression — instead of asking about the odds of an event by some fixed endpoint, it compares the instantaneous risk of death at every moment someone remains under observation, which naturally handles the fact that patients in a database spanning twenty years have wildly different follow-up times. That's a better fit than logistic regression would be for a time-to-event outcome like survival.

So what did they find, in round numbers. Out of nearly nine thousand patients, just under half were male. Men presented with somewhat larger tumors and more head-and-neck primaries, which is clinically relevant real estate for tissue-sparing surgery. Yet men were less likely to receive Mohs — about one in nine versus about one in seven for women — and after adjustment, male sex was independently associated with roughly a fifth lower odds of undergoing Mohs, a statistically significant and clinically meaningful finding given that this is exactly the population where margin-controlled surgery matters most. Men also had meaningfully higher odds of positive margins, on the order of a quarter higher, again statistically significant. And on survival, male sex was independently associated with worse overall survival, a statistically significant roughly quarter increase in hazard, even after adjusting for the measured confounders. Interestingly, when they stratified by facility type, these sex disparities in Mohs use and margin positivity disappeared in academic centers but persisted in non-academic settings — suggesting this isn't about DFSP biology differing by sex, but about where and how care gets delivered.

The authors are appropriately cautious in their discussion — this is a retrospective database study with the usual limitations: no detail on local recurrence or disease-specific survival, and residual confounding they can't fully exclude. They frame their conclusion carefully, as an association between sex and disparities in access to optimal surgical management rather than a causal claim.

Practically, this isn't practice-changing in the sense of altering your surgical technique, but it is a genuine access-and-equity signal worth carrying into how you think about referral patterns. If you practice in a non-academic or community Mohs setting, this is a nudge to examine your own referral funnel — are male patients with DFSP, particularly head-and-neck lesions, reaching you for margin-controlled surgery at the same rate as female patients? It's the kind of finding that should prompt a look at local practice patterns rather than a change at the bench.

Last on the docket is a Medicare claims analysis looking at trends in radiation therapy, including image-guided delivery, among dermatologist-performed procedures for keratinocyte carcinoma. Background: surgery remains primary treatment for basal cell and squamous cell carcinoma, with radiation relegated to a secondary option for select non-surgical candidates per AAD, NCCN, and ASTRO guidelines — and NCCN specifically says that radiation should be delivered by radiation oncologists. The rise of portable in-office superficial radiation therapy devices has driven explosive growth in dermatologist radiation billing, so the authors wanted to characterize exactly what's being treated and how.

This is a retrospective cohort study using a large deidentified random sample of Medicare beneficiaries, about five and a half million of them, aged sixty-five and up with continuous Parts A, B, and D coverage, spanning 2009 to 2021. They captured every keratinocyte carcinoma treated by a physician dermatologist using CPT codes for destruction, excision, Mohs, photodynamic therapy, and radiation, paired with a same-day keratinocyte carcinoma diagnosis code. Methodologically this is a fairly simple descriptive and trend analysis rather than a risk-adjusted model — they used chi-square tests to compare procedure-type distributions between general dermatologists and dermatologists who also bill for Mohs, and Kendall's Tau-B to test for trend over time. That's a reasonable choice: Kendall's Tau-B is a rank-based correlation measure well suited to ordinal trend data like "proportion of procedures by year," and it's a logical alternative to simple linear regression when you just want to know whether a trend is monotonically increasing or decreasing over time without assuming a strictly linear relationship.

The results: out of nearly two and a seven-tenths million treated keratinocyte carcinomas in this massive cohort, radiation therapy accounted for a small slice, about half a percent, split roughly evenly between basal cell and squamous cell carcinoma. But that proportion increased significantly over the study period. Within radiation cases, the overwhelming majority — about nine in ten — were superficial radiation therapy, with brachytherapy making up the remainder. Dermatologists who also billed for Mohs performed the majority of this radiation, close to three in five cases, though the overall proportion of their keratinocyte carcinoma caseload treated with radiation was similar to general dermatologists. The number the authors really want you to sit with is this: image guidance, almost entirely ultrasound, was used in about one in five radiation treatments across the study period, and that proportion increased significantly from 2015 onward.

Here's the tension the authors highlight in their discussion — ASTRO guidelines and a 2023 position statement explicitly state that image guidance is neither required nor appropriate for superficial radiation therapy of keratinocyte carcinoma, since these are clinically visible tumors with no accepted imaging endpoint for treatment response. Yet image-guided delivery, mostly ultrasound, is rising. The implication, though the authors don't say this outright, is a possible billing or reimbursement driver behind that trend rather than a clinical one. Limitations are notable: the data end in 2021, right as superficial radiation therapy utilization may have been accelerating further; Medicare data limits generalizability to older, predominantly white populations; they relied on diagnostic codes and assumed all dermatologist-billed radiation was definitive treatment; and they couldn't further characterize which dermatologists were fellowship-trained Mohs surgeons using radiation adjunctively versus general dermatologists using superficial radiation therapy as a primary modality.

This one is squarely in the "interesting but not yet practice-changing for your own hands" category, since as a Mohs surgeon you're unlikely to be the one billing superficial radiation. But it's relevant intelligence for how you counsel patients and for multidisciplinary conversations — if you're seeing patients referred for superficial radiation, or fielding questions about it as an alternative to surgery, it's worth knowing that its use is growing, that the image-guidance trend may be outpacing its evidence base and guideline support, and that there's currently no consensus guideline specifically governing radiation delivery for keratinocyte carcinoma, which leaves this an unsettled area worth watching rather than acting on.

That wraps up this October 2026 roundup — a reminder that even well-designed retrospective cohort studies benefit from rigorous letter-and-response scrutiny, that access disparities in margin-controlled surgery for rare tumors like DFSP deserve attention at the practice-pattern level, and that claims-based trend data can surface emerging practice patterns, like image-guided superficial radiation, before the guidelines have caught up. Thanks for listening, and we'll see you next month.