Welcome back to the journal review. This is your October twenty twenty-six run-through of JAAD, and we've got four pieces on deck — a brief report on a novel analgesic approach in nail surgery, two JAAD Game Changer commentaries worth knowing about even in their short form, and a meatier retrospective cohort study on long-term low-dose capecitabine for field cancerization. Let's get into it.

First up is a brief report, essentially a pilot retrospective cohort study, looking at liposomal bupivacaine in nail procedures. You already know liposomal bupivacaine from Mohs reconstruction — it's that extended-release formulation, branded Exparel, that's FDA approved not as your primary block but as a postsurgical analgesic add-on, releasing drug over something like three days. The gap here is straightforward: nail procedures — matrixectomies, nail bed and matrix biopsies, avulsions — are notoriously painful postoperatively, and nobody had looked at whether this extended-release formulation helps in that specific anatomic context, since prior dermatology data on liposomal bupivacaine is basically confined to Mohs defects.

Methodologically, this is about as modest as retrospective pilot work gets, and the authors are upfront about that. Twenty-four patients, twelve getting a liposomal bupivacaine and bupivacaine admixture via wing block, twelve getting ropivacaine, across ten different physicians at University of Miami. Outcomes — pain, sensation, numbness, and motor function, each on a zero-to-ten scale — were captured by phone call on postop days one through three. The statistical workhorse here is a linear mixed model, one fit separately for each of those four outcomes, with postoperative day, age, sex, and anesthetic type as fixed effects. This is the right tool for a reason worth dwelling on: each patient contributes multiple pain scores over time — day one, two, three — and those repeated measurements on the same person are correlated with each other in a way that a simple t-test or ANOVA would ignore, artificially inflating your confidence. A mixed model handles this by allowing each patient their own baseline trajectory while still estimating an overall population-level effect of anesthetic type, essentially borrowing strength across patients while respecting that their repeated measures aren't independent data points. Significance was assessed with a likelihood ratio test, which compares how much better the model fits once you add the anesthetic variable versus leaving it out.

The results: liposomal bupivacaine dropped pain by about a point and a half relative to ropivacaine, which did not cross the conventional threshold for statistical significance, but — and the authors flag this explicitly — a point-and-a-half reduction actually meets the published minimal clinically important difference for pain on this kind of scale, which runs around one-and-a-third to two points. So you have a classic underpowered-but-promising signal: not statistically significant, but potentially clinically meaningful if it held up in a larger sample. Motor function moved in the same direction, about a two-point improvement, again short of significance. Sensation showed no difference between groups, and interestingly numbness trended slightly higher with liposomal bupivacaine, though that didn't reach significance either. No adverse events in either arm.

Limitations are exactly what you'd expect from twenty-four patients, retrospective design, heterogeneous procedures, ten different surgeons with presumably variable technique, no blinding, and no placebo arm. The authors call it a pilot and ask for a properly powered prospective trial, which is the right instinct.

Practical takeaway: this is interesting, not yet practice-changing. The trend is encouraging enough that if you're already comfortable using liposomal bupivacaine off-label and have access to it, there's a reasonable biologic and now preliminary clinical rationale for trying it in nail surgery, particularly matrixectomies where postop pain is a known patient complaint. But don't treat a point-five-ish p-value in a twelve-patient arm as practice-defining evidence — this is a hypothesis-generating pilot, full stop.

Next, two JAAD Game Changer pieces — these are short retrospective commentaries from Dr. Jane Grant-Kels reflecting on how a prior JAAD paper changed her practice, not full methods write-ups, so I'll give you the clinical substance without inventing structure that isn't there.

The first revisits a paper on baseline serum neutrophil-to-lymphocyte ratio, or NLR, in acral melanoma versus nonacral melanoma. The finding driving this one: in advanced-stage disease, baseline NLR was significantly higher in acral lentiginous melanoma than in nonacral cutaneous melanoma, and a high NLR in the acral group was independently associated with worse progression-free survival, holding up even after adjusting for ulceration, Breslow depth at the two-millimeter cutoff, and nodal involvement. The practical lesson, as the commentary puts it, is that a simple, cheap, already-available lab value — a ratio you can pull off any CBC with differential — carries independent prognostic weight in acral lentiginous melanoma specifically. For those of you managing these patients alongside medical oncology, this is a data point worth having in your back pocket when counseling on expected trajectory, particularly in a histologic subtype that already behaves more aggressively than its nonacral counterparts stage for stage.

The second Game Changer looks back at a single-center ten-year retrospective review of Mohs micrographic surgery for melanoma in situ on the head and neck — eight hundred forty-six cases, the vast majority primary, a minority recurrent. The number that matters for your consent conversations: only about six in ten of these lesions cleared with a five-millimeter margin, and you needed to push out to fifteen millimeters to reach a ninety-seven percent clearance rate. Cheek location and larger preoperative lesion size were the two factors that predicted needing more than five millimeters to clear. The stated clinical translation — and this is useful even outside the Mohs context — is that when Mohs isn't available or appropriate and you're relying on standard wide local excision for head and neck melanoma in situ, margins of ten millimeters or more are often going to be necessary, not the traditional five. For those of us doing this under Mohs technique, the actionable piece is less about margin width, since we're already clearing histologically regardless of distance, and more about defect-size counseling — you can now tell patients up front, especially for cheek lesions or larger in-situ disease, to expect a bigger defect and plan their reconstruction conversation accordingly, rather than anchoring them to a five-millimeter mental model that this data says is frequently inadequate.

Now the main course: a retrospective cohort study, with an embedded survey component, on low-dose long-term capecitabine for refractory field cancerization, out of Brigham and Women's and Dana-Farber.

The background problem is one you deal with constantly — patients with extensive, topical-refractory field cancerization, often transplant recipients or otherwise immunosuppressed, who are cycling through surgery after surgery with real cumulative morbidity, what the authors nicely term surgical fatigue. Capecitabine, the oral prodrug of five-fluorouracil, has some prior signal in solid organ transplant populations — a systematic review cited here found roughly a two-thirds reduction in squamous cell carcinoma incidence at twelve months — but tolerability looked rough, with historical discontinuation rates reported as high as twenty to forty percent. Critically, nobody had data on long-term use with planned treatment holidays, and nobody had looked at non-transplant immunosuppressed or immunocompetent patients. That's the gap this study targets.

Design-wise, this is a single-institution retrospective cohort, twenty patients identified through institutional oncology data systems, reviewed for demographics, dosing, treatment holidays, concurrent chemoprevention, toxicity, and pre- versus post-treatment counts of keratinocyte carcinoma surgeries, biopsies, and topical treatments — paired with a separate survey study administered to surviving patients using the Skin Cancer Index, a validated quality-of-life instrument, plus questions on side effects and treatment preference. A retrospective design here makes sense for the reason it usually does in rare, long-duration exposures like this — you're not going to randomize twenty patients to a multi-year capecitabine protocol with treatment holidays; you're capturing real-world practice as it already happened, and pairing it with patient-reported outcomes gives you the qualitative dimension a chart review alone can't.

The statistical core is the Wilcoxon matched-pairs signed-rank test, comparing each patient's own annual keratinocyte carcinoma surgery count before treatment against their count after treatment, at the one, two, and three-year marks. This is worth a moment because it's a nice contrast to the mixed model we just discussed. Where a mixed model handles repeated measures with varying numbers of observations per patient, the Wilcoxon signed-rank test is built for a simpler paired-before-and-after question in a small sample where you can't assume the differences are normally distributed — instead of comparing means, it ranks the magnitude of each patient's within-person change and tests whether those ranks are symmetrically distributed around zero, which is what you'd expect if treatment did nothing. It's the natural nonparametric choice here precisely because with sample sizes dropping to eleven and then five patients by years two and three, you have neither the data volume nor any particular reason to trust a normality assumption, and each patient fundamentally serves as their own control, which is statistically efficient when numbers are this small.

On to the numbers, and these are worth stating plainly since they'd actually inform a treatment conversation. Mean annual keratinocyte carcinoma surgeries dropped from about three point nine before treatment to two point nine in year one — a twenty-four percent reduction that did not reach statistical significance, likely underpowered at eighteen patients. By year two, with eleven patients followed, surgeries dropped from about three point five to well under one, a seventy-three percent reduction that was statistically significant. By year three, with only five patients remaining, the reduction was back down to twenty-seven percent and not significant — almost certainly a sample-size artifact rather than a true waning of effect, given how few patients survived to that time point with adequate follow-up. Topical treatments and biopsies for concerning lesions also trended down across all three years, though none of those secondary endpoints reached significance. Put simply: about half of patients saw fewer keratinocyte carcinomas at one year, over eighty percent did at two years, and two of five did at three years — again, that last number reflects a very thin remaining cohort.

Toxicity is the other half of the story and it's notably better than prior transplant-cohort data would predict. Three-quarters of patients had at least one adverse effect, but most were handled with simple watchful waiting or a temporary treatment holiday rather than discontinuation. Only two patients, ten percent, stopped therapy permanently, both for cutaneous toxicity — a meaningfully lower discontinuation rate than the twenty-to-forty-percent range reported in earlier transplant-focused literature, which the authors attribute plausibly to the built-in treatment holiday structure: roughly six to ten months of cycling followed by a two-to-four-month break, repeated as needed, rather than continuous dosing. The most common adverse effects were gastrointestinal symptoms, cutaneous reactions, fatigue, and palmoplantar erythrodysesthesia, each affecting roughly a third to just under half of patients. On the patient-reported side, three-quarters felt capecitabine improved their skin lesions at least moderately, and a striking five in six said they preferred it to alternative treatments they'd tried.

The authors' own limitations section is honest and brief: retrospective design and small sample size, with attrition across the three yearly analyses meaning the year-two and year-three numbers rest on progressively thinner cohorts. I'd add that this is also a single-institution, entirely white, non-Hispanic population, split evenly between immunosuppressed and immunocompetent patients, so generalizability to more diverse populations and to community practice settings without dedicated DPD genotyping and lab monitoring infrastructure remains an open question. The year-three loss of significance is a statistical power issue, not necessarily a loss of efficacy, but you can't fully disentangle those with the data presented.

Practical takeaway, and I'd call this genuinely practice-relevant for anyone managing heavily field-cancerized patients who've failed topical therapy: low-dose capecitabine with a cycled treatment-holiday structure — roughly six to ten months on, two to four months off — appears to meaningfully reduce surgical burden, particularly by year two, with a tolerability profile considerably better than older transplant-cohort data suggested, and with strong patient-reported preference over alternatives. This is a reasonable option to have in your chemoprevention armamentarium alongside nicotinamide and acitretin for patients who've exhausted topical options, especially transplant recipients and other high-burden patients, provided you build in dihydropyrimidine dehydrogenase screening and routine lab monitoring as this cohort did. What's not yet established is optimal dosing, ideal holiday timing, and durability beyond three years, so this is promising real-world evidence supporting an existing off-label practice rather than a new standard to adopt wholesale — but it's considerably more actionable than the nail block pilot we opened with.

That wraps this month's review. To summarize the signal-to-noise: the capecitabine cohort is the one with real, if preliminary, practice relevance; the two Game Changer pieces hand you concrete numbers worth keeping for patient conversations on acral melanoma prognosis and head-and-neck melanoma in situ margins; and the liposomal bupivacaine nail study is a hypothesis worth watching, not yet acting on. Thanks for listening, and I'll see you next month.