Welcome back to the journal review. This month we're working through the December twenty twenty-five issue of the Journal of the American Academy of Dermatology, and we've got four pieces on deck — two brief reports, a full original study, and a letter-and-response exchange that's worth hearing as a pair even though we'll take them in the order they appeared. Let's get into it.

First up is a brief report titled "Adolescent tanning perceptions and resistance to prevention: insights from a skin cancer education program." This is a survey-based brief report, not a full trial, and it's tackling a problem you've probably sensed anecdotally in clinic — that no matter how much sun-safety messaging we throw at teenagers, a subset just doesn't budge. The authors, a group out of Baylor, wanted to know why, and specifically whether the belief that "a tan makes you more attractive" is driving resistance to prevention messaging.

The methods here are straightforward and worth noting for what they are: this piggybacked on an existing public skin cancer education outreach program for grade-school students. After the educational session, the students completed a survey asking whether they agreed that having a tan makes them more attractive, along with questions about their actual outdoor and indoor tanning behavior in the past six months, and whether the session itself changed their intentions around sun protection. About a thousand students completed it, a strong eight-in-ten response rate. This is a cross-sectional, self-report design — the authors didn't state it explicitly, but the rationale is fairly obvious: you're capturing attitudes and behavior at a single time point right after an intervention, which is a pragmatic and low-cost way to screen for effect-modifying beliefs before investing in a more expensive longitudinal design.

The results are the headline here. About one quarter of students agreed that tanning improves attractiveness. And that belief mattered enormously. Compared to those who disagreed, the "tan equals attractive" group had roughly a twelve-fold increased odds of outdoor sunbathing and about a four-fold increased odds of indoor tanning in the last six months, and both held up after adjusting for grade level, parental education, and Fitzpatrick type. That first number is not a subtle effect — a twelve-fold odds ratio is about as large as you'll ever see in a behavioral study. The same group also trended toward roughly double the odds of saying the educational session would not change their sunburn-prevention or sun-avoidance behavior, though that particular comparison landed in the marginally-significant range rather than clearly significant.

The authors' discussion leans into a biopsychosocial framing — biological, in that ultraviolet exposure triggers endorphin release with real addictive potential; social, in that tanned skin remains a beauty ideal; and psychological, in that tanning may function as a coping mechanism tied to self-esteem. They explicitly raise the idea of screening for tanning as a behavioral addiction using tools like the Behavioral Addiction Indoor Tanning Screener, and they point to sunless tanning promotion — dihydroxyacetone-based products — as an evidence-based alternative, citing a beach-based randomized trial where women offered sunless tanning products were significantly less likely to sunbathe at one-year follow-up.

Limitations are inherent to the design — this is self-reported behavior and self-reported future intention, not verified outcomes, and it's a single intervention program, so generalizability beyond that population and that curriculum is uncertain. But for a brief report, the effect size on outdoor tanning is hard to ignore.

Practically, I wouldn't call this practice-changing in the sense of altering how you manage a patient in the chair, but it is directly actionable for anyone involved in prevention or public health messaging, and probably worth internalizing for how you counsel adolescent patients and their parents. If a teenager tells you they think tanned skin looks better, generic "wear sunscreen" messaging is very likely to bounce off. The more useful move, per this data, is to reframe the conversation around sunless alternatives rather than pure restriction, and possibly to screen for compulsive tanning behavior the way you'd screen for any other behavioral pattern with an addiction component.

Next, a letter to the editor — "Comments on Fakult et al's Epidemiology of Merkel cell carcinoma in the United States." This is a correspondence piece responding to a SEER-based epidemiologic study on Merkel cell carcinoma trends from two thousand to twenty twenty-one, so there's no independent methods or results section here — it's a critique, and I'll walk through the substance of the pushback.

The letter writers raise three main points. First, the original study found no positive correlation between solar irradiance and Merkel cell carcinoma incidence at the population level, which on its face seems to challenge ultraviolet radiation as a causal driver. The letter writers call this out as a likely ecological fallacy — SEER data is aggregated at the population level and doesn't capture individual UV exposure histories, so a null correlation at that level doesn't actually contradict the substantial body of evidence linking UV to Merkel cell carcinoma in light-skinned patients and in sun-exposed tumor sites. Second, they push back on the original authors' suggestion that recent survival improvements might be attributable to immunotherapies like avelumab and pembrolizumab, pointing out — correctly — that SEER has no treatment-level granularity, so that attribution is speculative. Third, they urge caution around the claim that incidence and mortality have "stabilized" in recent years, raising the possibility that reporting lag, diagnostic saturation, and pandemic-related healthcare disruption could produce an artifactual plateau rather than a true biological one. They also flag that Merkel cell carcinoma is molecularly heterogeneous — polyomavirus-positive versus UV-driven tumors — and that lumping these together at a population level could obscure genuinely divergent trends.

This is a methodologic critique piece through and through, and it's a good one for calibrating how much weight to put on SEER-based epidemiologic conclusions generally — a useful reminder any time you see population registry data used to infer either causality or treatment effect.

Now, rather than jumping to the response immediately, let's go to the substantive original study in this issue, and we'll circle back to the reply at the end, since it directly continues this conversation.

That original study is titled "Genomic and clinical characterization of anogenital basal cell carcinoma: a retrospective and prospective cohort study," out of a multi-center French collaborative. This is a full original study, and it's addressing a genuinely under-characterized entity — basal cell carcinoma arising in sun-protected anogenital skin, which the authors note occurs at an incidence roughly a thousand-fold lower than basal cell carcinoma in sun-exposed sites. The core gap: cumulative UV exposure is the dominant established driver of basal cell carcinoma pathogenesis, so what's actually driving tumorigenesis in a site that never sees the sun, and does that biology have implications for how these tumors should be managed?

Methodologically, this combined retrospective and prospective cohort collection across seven French academic centers from two thousand six through twenty twenty-four — fifty patients total. The authors don't spell out why they chose this hybrid retrospective-prospective design, but the reasoning is fairly intuitive for a rare-disease study: anogenital basal cell carcinoma is uncommon enough that you can't accrue a meaningful cohort prospectively in any reasonable timeframe, so retrospective case capture extends your denominator, while the prospective arm going forward allows for consistent, protocol-driven sample collection — critical here because a subset of tumors underwent whole-genome or whole-exome sequencing, which requires fresh tissue collected before treatment. That's not something you can retrieve retrospectively from archived material with the same fidelity. Twelve of the fifty cases had adequate fresh tissue and consent to undergo sequencing.

On the clinical side, the cohort was mostly women — forty of fifty — with vulvar involvement being the dominant site in women and perianal involvement dominant in men. Average age at diagnosis was seventy-eight. Three patients had a prior history of pelvic radiotherapy in the exact area where the basal cell carcinoma later developed — small numbers, but worth flagging as a possible radiation-associated pathway distinct from UV carcinogenesis. Nodular and infiltrative subtypes together accounted for the large majority of cases, and about one in five vulvar basal cell carcinomas arose in a background of lichen sclerosus — that's a meaningful minority and supports a chronic-inflammation-associated pathway analogous to what we already accept for vulvar squamous cell carcinoma.

Treatment was overwhelmingly conventional wide local excision — no Mohs surgery was performed in this cohort at all, which is itself notable given how much anogenital anatomy typically demands tissue conservation. A small number of patients received neoadjuvant hedgehog inhibitor therapy before surgery, and in those cases, tumor size shrank by at least half prior to excision — a meaningful debulking effect in a location where morbidity from wide excision is a real concern. Two elderly patients were managed with observation alone after a geriatric multidisciplinary discussion, reflecting individualized goals of care rather than protocol.

The recurrence numbers are the most clinically actionable data point here: among the forty-eight treated patients, recurrence occurred in fifteen percent over a median follow-up of seven years. Six of those seven recurrences had been treated with conventional excision with clear margins — not Mohs, plain wide local excision — and recurred anyway despite histologically negative margins. That's a real signal that standard margin assessment in this anatomic site may be under-detecting subclinical extension, which is exactly the kind of finding that should make a Mohs surgeon's ears perk up, even though the numbers here are too small to draw firm treatment-algorithm conclusions.

The genomic findings are the other major contribution. Tumor mutational burden was low, averaging about three and a half mutations per megabase, and there was no evidence of UV mutational signatures — confirming at the molecular level what the anatomic site already suggested. PTCH1 was mutated in the large majority of sequenced tumors, and loss of heterozygosity at the PTCH1 locus on chromosome 9q was present in literally every sample, consistent with biallelic inactivation and sonic hedgehog pathway activation as the core oncogenic driver — the same final pathway as sun-exposed basal cell carcinoma, just reaching it without UV-induced mutagenesis. Interestingly, TP53 mutations, which show up in roughly seven in ten UV-driven basal cell carcinomas, were seen in only two of twelve anogenital tumors here — a statistically significant difference, though the authors are appropriately cautious given how small that comparison is, and they explicitly call for validation in larger cohorts. HPV testing was performed in a handful of p16-positive cases and came back negative in all of them, arguing against papillomavirus as a driver in this specific cohort.

Limitations are exactly what you'd expect and the authors state them plainly — small sample size, especially for the genomic arm, and a retrospective design for a substantial portion of the cohort, which limits standardization of margin reporting and follow-up intervals.

Practically, here's how I'd frame the takeaway. The recurrence signal — fifteen percent over seven years, concentrated in patients treated with conventional excision and clear margins — is the piece worth actually sitting with, because it raises a legitimate question about whether standard bread-loafed margin assessment is adequate in this site, given how deceptively infiltrative anogenital basal cell carcinoma can be. It's not yet practice-changing in the sense of "you must do Mohs on every anogenital basal cell carcinoma" — the numbers are too small and no Mohs cases were even included for comparison — but it's a reasonable rationale for lower threshold toward complete circumferential margin assessment or closer long-term follow-up in these patients, particularly those with lichen sclerosus or prior pelvic radiation, both of which the authors flag explicitly as populations warranting long-term surveillance. The genomic story — hedgehog pathway activation without UV signature — is scientifically satisfying and supports hedgehog inhibitor use as a legitimate neoadjuvant option when surgery is anatomically difficult, which the authors' own data support with meaningful debulking response. That part is already actionable today for surgeons managing large or anatomically challenging anogenital basal cell carcinoma.

Now let's close the loop on that Merkel cell carcinoma exchange. The final piece is the original authors' formal response, titled "Response to Sheu et al." This is a reply letter, so again, no new data — just point-by-point clarification.

On the UV correlation question, the original authors push back gently: they note their paper actually did report findings consistent with UV causality — Merkel cell carcinoma arising predominantly in sun-exposed sites, increasing with age, occurring more often in lighter-skinned patients, and showing an age-related increase in sun-exposed sites specifically in White patients. They agree the population-level null correlation likely reflects aggregation bias rather than a true absence of UV effect — essentially conceding the ecological fallacy point while defending their other lines of evidence. On the immunotherapy attribution, they concede the critique is fair, note they'd already flagged alternative explanations like earlier detection and improved surgical technique in their original discussion, and add a useful methodologic point worth remembering generally: any real survival benefit from a therapy approved in twenty seventeen may take five to ten years to become visible in population-level data that only runs through twenty twenty-one, and population data isn't the right tool to establish a treatment effect in the first place — that requires randomized trials. On the stabilization question, they agree further research is needed and acknowledge the three-year reporting lag inherent to SEER limits real-time trend assessment, especially around pandemic-era disruption. And they agree that the racial incidence disparity and the biological heterogeneity between polyomavirus-positive and UV-driven Merkel cell carcinoma deserve dedicated future study, explicitly recommending that SEER incorporate polyomavirus status as a captured variable going forward.

There's no new clinical action item from this exchange, but as a pair, this letter and response are a genuinely nice worked example of how to read a SEER-based epidemiology paper critically — treating incidence trends, survival trends, and treatment attribution as three separate claims that each carry their own distinct evidentiary weaknesses.

That wraps our four pieces this month. The through-line, if there is one, is a reminder to keep two habits sharp — read population-level epidemiologic claims with an eye for aggregation bias and unmeasured confounding, the way this Merkel cell carcinoma exchange models so well, and stay alert to genuinely novel biology showing up in unusual anatomic sites, the way anogenital basal cell carcinoma's UV-independent, hedgehog-driven, and possibly under-margined behavior did here. Thanks for listening, and we'll see you next month.