Welcome to this December twenty twenty-five run-through of the Journal of the American Academy of Dermatology. Four pieces on the docket today — a surgical pearl for Mohs processing, a prospective cohort study on body mass index and melanoma outcomes, a large multi-institutional cohort study on satellitosis and in-transit metastasis in cutaneous squamous cell carcinoma, and a brief survey report on sun protection at golf courses. Let's get into it.

First up is a surgical pearl — a "how we do it" piece out of Curitiba and Dallas, addressing a problem every one of us has run into: the large Mohs specimen that simply won't fit on a single histology slide. The traditional workaround has been to divide the specimen into multiple pieces, each sized to fit its own slide, which means multiple blocks. The authors point out the real cost of that approach — every time you divide tissue, you increase the risk of floaters, and you increase the risk that residual microscopic tumor or unwiped curettage debris drops out of the central cut surface, which can generate a false-positive deep margin. There's an existing alternative of mounting a single section across two slides held together, but the authors note that's technically fiddly.

Their solution, which they call the alternating single-section method, is elegantly simple. Rather than dividing the specimen into pieces during processing, you process and section the entire block as one continuous section. You fit as much of that section onto the first slide as will go, and when a portion doesn't fit, the contralateral edge of that same section gets fit onto a second slide. Sequential cuts continue in this alternating fashion — left portion on one slide, right portion on the next — so that when you mentally superimpose the two slides, you're looking at one hundred percent of the surgical margin, without ever having physically divided the specimen. Each slide is read on its own, but together they reconstruct the whole margin.

The advantage here is mechanical: less tissue manipulation means fewer opportunities for floaters and for the false-negative or false-positive margin calls that come with manipulating a fragmented specimen. The tradeoff the authors are upfront about is a longer freezing time for that larger single block — but they argue that's offset by the lab efficiency of sectioning, staining, and reading fewer blocks overall. This is a genuinely practical, low-barrier technique — nothing here requires new equipment, just a change in how you approach the embedding and cutting of oversized specimens. For anyone running their own lab or training technicians, this is directly actionable tomorrow morning.

Second, a brief report out of the Queensland Melanoma Project, prospectively examining whether body mass index predicts tumor stage at diagnosis and recurrence in high-risk primary melanoma. The background here is the obesity-melanoma literature, which has generally suggested that obesity — through chronic inflammation — correlates with worse prognosis and thicker tumors at diagnosis. This group set out to test that directly in a prospective cohort.

Methodologically, this is a secondary analysis nested within a previously published prospective cohort of seven hundred adults with newly diagnosed T1b through T4b melanoma, recruited across Queensland health facilities between twenty ten and twenty fourteen. Body mass index, comorbidities, socioeconomic status, and skin-check frequency were all captured by baseline questionnaire, while tumor stage came from pathology and recurrence came from follow-up or cancer registry linkage. They used logistic regression for the tumor-stage-at-diagnosis question and Cox proportional hazards for the seven-year recurrence question, adjusting for the usual suspects — age, sex, smoking, statin use, skin-check frequency, socioeconomic status. A prospective cohort with registry-linked outcomes is a sensible design choice here because it avoids the recall and selection biases that would plague a retrospective look at recurrence, and because tumor stage and comorbidity data were collected before outcomes were known.

The results actually run counter to the obesity-thick-tumor hypothesis. Obese patients were significantly more likely to present with an earlier tumor stage, not a later one, compared with healthy-weight patients — roughly a two-fold increase in odds, and this held up after adjustment. When they stratified by presence of serious comorbidities, that association got even stronger in the comorbid subgroup, essentially a three-fold increase in odds of earlier-stage diagnosis. Critically, when they looked at seven-year recurrence, there was no association with body mass index at all — the hazard ratios sat right around one regardless of weight category, and this was true whether or not comorbidities were adjusted for.

The authors' own interpretation is worth relaying directly: they don't think this reflects any biologically protective effect of obesity on melanoma behavior. Instead, they suspect it's a surveillance artifact — obese patients, particularly those with comorbidities, likely have more frequent contact with the healthcare system for management of those comorbidities, creating more opportunities for incidental or opportunistic lesion detection at an earlier stage. The obvious limitation is that most of the covariates here, aside from height and weight, were self-reported, and this remains an association study that can't fully disentangle detection bias from any true biological effect. For practice, this isn't practice-changing — it doesn't change how you counsel a patient on prognosis based on their weight — but it is a useful mental corrective if you've been assuming, based on older literature, that obesity should raise your index of suspicion for a thicker, worse-prognosis tumor. Here it did the opposite, most plausibly through a detection pathway rather than a biological one.

Third, and the most substantial paper this issue: a multi-institutional cohort study on satellitosis or in-transit metastasis in cutaneous squamous cell carcinoma — what the authors abbreviate as S-ITM — looking at risk factors and prognostic significance. This is squarely in our wheelhouse. The clinical problem is that S-ITM, defined as satellite tumor deposits or metastases in the dermis or subcutaneous tissue that are physically separate from the primary tumor and its margins, occurring before any nodal involvement, is well recognized as a marker of poor outcomes in the existing literature — one prior study put its prognosis on par with nodal disease. And yet, despite that evidence, S-ITM is not incorporated into either the American Joint Committee on Cancer eighth edition or the Brigham and Women's Hospital staging systems. That's the gap this paper is built to address.

The design is a retrospective cohort pooling almost nine thousand cutaneous squamous cell carcinoma patients across twelve institutions in the United States, Brazil, and Spain, spanning twenty-five years of data collection. Given that S-ITM is a rare event, a retrospective multi-institutional pooling strategy is really the only feasible way to generate adequate statistical power — you simply couldn't accrue enough events at a single center in a reasonable timeframe, and the authors' explicit rationale is the rarity of the outcome. They used two separate analytic strategies matched to two separate questions. To find risk factors for developing S-ITM, they ran multivariable logistic regression comparing the seventy-seven patients who developed S-ITM against the roughly eighty-eight hundred who didn't. Then, to isolate whether S-ITM itself independently predicts poor outcomes — rather than simply being a marker of already-bad tumor biology — they built a propensity-matched control group, matched on things like immunosuppression, location, diameter, depth, differentiation, perineural invasion, lymphovascular invasion, treatment modality, and margin status, and then used a Fine-Gray subdistribution hazard model to compare cumulative incidence of local recurrence, nodal metastasis, distant metastasis, and disease-specific death. The Fine-Gray approach is the right tool here because it properly accounts for competing risks — for instance, death from another cause competing with the chance of observing a recurrence — which a standard Cox model would mishandle.

On risk factors, four things came out as independent predictors of developing S-ITM: increasing patient age, a history of immunosuppression — which roughly quadrupled the odds — and, unsurprisingly, higher Brigham and Women's Hospital tumor stage, with a dramatic, near thirty-fold increase in odds for T3 tumors compared with T1. Lymphovascular invasion also independently raised risk, again by roughly four to five-fold. Notably, primary tumor treatment modality, margin positivity, and adjuvant treatment did not reach significance as predictors of who goes on to develop S-ITM.

On outcomes, this is where the paper earns its conclusion. Compared with propensity-matched controls, patients with S-ITM had roughly double the risk of local recurrence, meaningfully elevated risk of nodal metastasis, over a four-fold increase in distant metastasis, and — the number that matters most clinically — well over a four-fold increase in disease-specific death. In absolute terms, disease-specific death occurred in roughly half of the S-ITM group compared to under one in ten of the matched controls, and all-cause mortality was three-fold higher. Because this was a matched comparison controlling for tumor size, depth, differentiation, perineural invasion, lymphovascular invasion, and treatment, the authors argue S-ITM is conveying independent prognostic information beyond what's already captured by existing staging variables.

Limitations are exactly what you'd expect from a retrospective multi-institutional design — the rarity of the event limits statistical power despite the large denominator, follow-up time was shorter in the S-ITM group, which could bias event detection, and pooling data across twelve sites over twenty-five years introduces heterogeneity in documentation and treatment practice patterns that the authors can't fully control for. Still, this is about as good as the evidence base for a rare event like this is likely to get without a dedicated prospective registry.

The practical takeaway is genuinely significant for our field. This adds substantially to a growing body of literature arguing that satellitosis or in-transit metastasis should be incorporated into cutaneous squamous cell carcinoma staging, since it appears to carry independent prognostic weight comparable in magnitude to markers we already stage on, like large-caliber perineural invasion. It's not yet practice-changing in the sense of altering the staging systems themselves — that requires the staging committees to act — but clinically, if you identify satellite lesions or in-transit deposits on a high-stage, immunosuppressed, or lymphovascular-invasion-positive tumor, this data says you should treat that patient as high risk regardless of what box the current staging system puts them in, with a low threshold for multidisciplinary discussion around adjuvant radiation, systemic therapy, and closer surveillance.

Last, a brief report with a public health flavor — a survey study on sun protection measures at Michigan golf courses. The premise is straightforward: golfing involves prolonged outdoor exposure often during peak ultraviolet hours, and there's been little empirical assessment of what golf courses actually offer in terms of sun protection infrastructure and education. The methodology was a phone survey of just over a hundred golf courses near Detroit, with seventy-five responding, asking about sunscreen availability, sun protection factor levels, protective apparel, staff policies, and educational materials — a simple, low-cost design appropriate for a descriptive feasibility question like this one, though of course it relies on self-report from whoever answered the phone rather than any independent verification.

The findings: about seven in ten courses provided sunscreen for patrons, though only about one in five offered it free of charge, and there was a clear public-private divide — private courses were roughly three times as likely to offer free sunscreen as public courses, a statistically significant and practically meaningful gap. Most courses that did provide sunscreen offered at least SPF thirty, with lotion the dominant form. Broad-brimmed hats were available for purchase at about three-quarters of courses, again more common at private clubs. Sun-protective shirts were available at roughly half. But the standout finding, and the one worth remembering, is that only two of the seventy-five courses — under three percent — offered any educational materials on sun-safe practices, and only about one in ten had any staff sun protection policy or training program, again with no meaningful public-private difference on that particular point.

There's no results-versus-discussion tension to unpack here beyond what's on its face — this is a descriptive snapshot, not a hypothesis test, and the authors' conclusion follows directly: physical sun protection products are reasonably available, especially at private courses, but the educational infrastructure to actually get golfers and staff to use them consistently is essentially absent. For us as dermatologists, this isn't a clinical practice-changer, but it is a useful piece of advocacy ammunition — it's a concrete, quotable gap if you're involved in public health outreach, tumor board community education initiatives, or partnering with local recreational facilities on skin cancer prevention messaging.

That wraps our four articles this month — a practical processing pearl you can implement at your next large Mohs case, a reassuring but non-actionable finding on body mass index and melanoma stage, a substantial cohort study that strengthens the case for staging satellitosis and in-transit metastasis in cutaneous squamous cell carcinoma, and a reminder that sun protection infrastructure without education is only half the job. Thanks for listening, and I'll see you next issue.