Welcome back to the journal club. This is our walkthrough of Dermatologic Surgery for May twenty twenty-six, and we've got four pieces on the docket this time: a combination "how we do it" technique for acne scarring, a reliability study on MART-1 margin assessment in Mohs for melanoma, a systematic review of the JAK-STAT pathway in keloids, and a systematic review looking at Mohs for benign adnexal tumors. Let's get into it.

First up is a "How We Do It" piece — so this is technique-article territory, not a hypothesis-driven trial, and it reads that way: steps, rationale, ideal patient, then a modest retrospective case series tacked on to support it. The group, led by Turki with Tseng and Murrell, describes a four-step combination protocol for acne scarring that layers chemical reconstruction of skin scars — CROSS — using eighty-nine percent carbolic acid, with cannula subcision, fractionated ablative erbium:YAG resurfacing, and volumetric filler where needed. The logic is straightforward and matches how most of us already think about acne scar subtyping: ice pick and boxcar scars get the carbolic acid, painted in a ring for boxcar morphology and dropped directly into the base for ice pick scars; rolling scars get tumescent subcision through a single sideburn entry point with a fanning technique; then two passes of fractionated erbium:YAG at a fairly conservative hundred-micron depth, with wet gauze debridement between passes; and finally filler for residual volume loss. The stated rationale for combining these rather than reaching for something like a full-face phenol-croton peel is recovery time — they're quoting six to eight days downtime versus a much longer recovery with deep peeling — plus a lower dyschromia risk in darker skin types, which matters because this cohort is genuinely diverse: roughly a quarter of patients were Fitzpatrick type four or above.

Mechanistically, the authors offer an interesting synergy argument worth flagging as their proposed explanation rather than a proven mechanism — that the carbolic acid draws water into the wound edges, and that this chromophore-rich, edematous tissue then interacts favorably with the erbium laser's water-target absorption, theoretically narrowing the zone of thermal injury and lowering post-inflammatory hyperpigmentation risk. It's plausible, but it's mechanistic reasoning layered onto a retrospective series, not something they measured directly.

On methods, this was a retrospective chart review of forty-nine patients, mean age right around thirty-one, majority female. Outcome assessment combined three things: a patient self-rated improvement survey about four months out, a validated psychosocial questionnaire administered before and after, and blinded reviewer grading of standardized photographs using the Goodman and Baron qualitative scale. Using blinded photo grading alongside patient-reported outcomes is a sensible way to triangulate efficacy when you don't have a comparator arm — it doesn't eliminate the lack of a control group, but it does guard against pure placebo-driven patient satisfaction bias.

Results: every single patient achieved at least a one-point improvement on the Goodman and Baron scale, and on self-grading, about a third called their improvement moderate and closer to four in ten called it significant — so the large majority landed somewhere in the meaningfully-improved category. The psychosocial data were the more compelling piece for me — statistically significant improvements across the board in feeling self-conscious, feeling attractive, worrying the scars wouldn't resolve, and social avoidance, with fairly large absolute swings, for instance patients endorsing feeling self-conscious dropped from about eight in ten before treatment to roughly one in seven after. Complications were all short-term and self-resolving: some erythema, a small rate of post-inflammatory hyperpigmentation, filler nodules in about a quarter of patients who received filler, and a handful of transiently enlarged scars.

The honest limitations here are the ones inherent to any retrospective, uncontrolled case series with patient-reported outcomes at a single follow-up point — no control arm, no comparison against any single-modality treatment, relatively short follow-up, and an outcome measure that blends subjective self-report with a qualitative rather than quantitative photographic scale. Practically, I'd call this interesting and additive rather than practice-changing on its own — it's a reasonable combination protocol with a sensible mechanistic rationale and a good safety profile in skin of color, but it's not powered or designed to tell you it outperforms the individual modalities you're probably already combining in some fashion. If you're already doing CROSS and subcision and resurfacing sequentially, this gives you a specific, replicable protocol and reassurance about the safety profile in higher Fitzpatrick types, which is the genuinely useful takeaway.

Next, an original study that's squarely in our procedural wheelhouse — the interrater and intrarater reliability of Mohs surgeons interpreting MART-1 immunostained frozen section margins for melanoma. This addresses a real gap: as Mohs for melanoma expands, we've had good outcome data — local recurrence rates comparable to permanent section evaluation — but nobody had actually formally tested how consistently different surgeons read the same MART-1 slide, or how consistent a single surgeon is with themselves over time.

Methodologically, this is a well-constructed agreement study, and it's worth walking through why they built it this way. Twenty ACMS fellowship-trained Mohs surgeons were recruited and asked to independently review twenty digitized en-face frozen section slides — a deliberately curated mix of eight originally called positive, eight negative, and four indeterminate at the time of the actual case — with no clinical history provided, to isolate pure histologic interpretation from clinical context. They read the same slide set twice, six weeks apart, which is the design feature that lets you separate interrater reliability, agreement between different surgeons, from intrarater reliability, a single surgeon's consistency with themselves over time. Six weeks is long enough that surgeons shouldn't simply remember their prior call on a given slide. Statistically they used Fleiss kappa for the multi-rater interrater comparisons and Cohen's kappa averaged across individual raters for intrarater agreement — the standard, appropriate choice for categorical agreement data of this kind, and they followed formal reporting guidelines for reliability studies, which adds credibility to the methodology.

The results were reassuring. Interrater agreement was substantial both times — about eighty-four percent raw agreement on the first read and eighty-eight percent on the second, with kappa values around zero-point-seven, which by standard benchmarks is solidly in the "substantial agreement" range, trending toward "almost perfect" on the second read. Intrarater agreement was even stronger — about ninety-five percent raw agreement with a kappa around zero-point-nine, essentially near-perfect consistency for individual surgeons rereading their own prior interpretations. When they built a consensus "true" margin status for each slide and checked individual rater accuracy against it, average accuracy came out around ninety-three percent, and interestingly this did not track with years in practice — more experienced surgeons weren't meaningfully more accurate than less experienced ones, which the authors note as unexpected but don't over-interpret. The two slides that generated real disagreement were both melanoma in situ cases showing diffuse basal melanocyte staining without the more decisive features — nesting, pagetoid scatter, cytologic atypia — that characterized the high-agreement positive slides, and interestingly oblique transection of adnexal structures created a false appearance of stacking and pagetosis in those problem slides, which is a nice piece of practical pattern-recognition teaching buried in the discussion.

The context they offer is genuinely useful for framing how good this is: agreement between radiologists reading chest films for an infiltrate, or between radiologists reading mammograms, both land in the fair-to-moderate range with kappas well below what was seen here. So Mohs surgeons reading MART-1 en-face margins are doing considerably better than comparable binary diagnostic tasks in other specialties. Limitations worth naming honestly: this is a still-image, decontextualized reading exercise rather than live intraoperative interpretation with real tissue, real depth of field, and real clinical correlation; the case mix was curated by the investigators rather than consecutive; and four indeterminate-at-the-time cases were deliberately built into the sample, which may not mirror true prevalence of ambiguous margins in practice. Practically, though, I'd call this practice-affirming and, frankly, reassuring rather than practice-changing — it doesn't ask you to alter technique, but it does provide solid evidentiary backing for the reliability of MART-1 en-face margin assessment that you can cite when defending Mohs for melanoma to colleagues or payers, and the specific pitfall they highlight — diffuse basal staining without nesting or atypia, especially with oblique adnexal transection — is worth keeping in mind at the scope next time you're on the fence with an in situ case.

Third, a systematic review on the JAK-STAT pathway in keloid pathogenesis and treatment. This is pure evidence synthesis, so no methods-rationale beat here beyond the standard PRISMA search process — they searched five databases through October twenty twenty-four and found a hundred nine articles, whittled down to twenty-three meeting inclusion criteria: one clinical trial and the rest basic science, in vitro and animal work.

The throughline is mechanistic: JAK-STAT signaling, particularly through JAK1, JAK2, and STAT3, appears meaningfully upregulated in keloid fibroblasts compared with normal fibroblasts, driving the fibroblast proliferation and collagen overproduction that defines the phenotype. STAT3 gets the most attention across the preclinical literature — elevated in keloid tissue and fibroblasts across multiple studies, and knocking it down, whether with siRNA or with agents like niclosamide or cucurbitacin, consistently impaired fibroblast proliferation and collagen synthesis in vitro. JAK2 inhibition with agents like ruxolitinib blocked STAT3-driven effects in keloid fibroblasts, and there's a nice mechanistic thread around IL-6 and thrombospondin-1 upstream of JAK2-STAT3 activation. STAT1 is also implicated, and interestingly, the review notes that corticosteroids — our actual current first-line intralesional therapy — appear to work partly by directly inhibiting JAK-STAT signaling, which is a nice bit of mechanistic continuity tying old therapy to new pathway biology. Other STAT family members, STAT2 and the less-studied JAK3, TYK2, and STAT5 isoforms, have much thinner or essentially absent keloid-specific data.

The one clinical data point is a small trial of tofacitinib, a JAK1/3 inhibitor, in seven keloid patients over twelve weeks at five milligrams twice daily. Patients showed significant improvement on validated scar assessment scales — softer, thinner, less rough keloids — along with significant improvement in itch and pain. One patient withdrew due to mild liver function abnormality. That's obviously a very small, uncontrolled dataset — seven patients, no comparator arm, short follow-up — so it's hypothesis-generating rather than practice-changing.

The authors are appropriately honest that grading the quality of evidence for JAK-STAT inhibitors as keloid therapy wasn't even attempted given how sparse the clinical data are, and they explicitly flag systemic side effect risk as an ongoing concern with any systemic JAK inhibitor approach, echoing the black-box safety conversation that's already familiar to us from JAK inhibitor use in other dermatologic contexts. For a surgeon's practical takeaway: this is squarely in the "interesting, not yet actionable" category. There's a coherent, biologically plausible pathway here that helps explain why keloids behave the way they do and why corticosteroids and existing therapies partially work, and it's reasonable to keep an eye on topical or intralesional JAK-STAT-targeted formulations as they mature, since systemic dosing is unlikely to be how this eventually gets used for a scar-limited condition. But there is nothing here to change how you counsel or manage a keloid patient today.

Last, a systematic review addressing a genuinely practical gray-zone question — when, if ever, is Mohs micrographic surgery justified for benign adnexal tumors of the skin, given that these lesions sit outside the current appropriate use criteria. The authors searched PubMed, EMBASE, and Web of Science through March twenty twenty-five specifically for named adnexal tumor types paired with Mohs terminology, screened down from four hundred four records to thirty included studies, all of which were case reports or small case series — meaning by CEBM and GRADE standards this entire literature sits at the lowest formal evidence tier, level four, low quality. That's an important framing point the authors are upfront about before you even get to the findings.

Across those thirty studies, they extracted eighty-five individual tumors treated with Mohs, spanning trichoblastoma, desmoplastic trichoepithelioma, hidradenoma, chondroid syringoma, spiradenoma, cylindroma, desmoplastic trichilemmoma, papillary eccrine adenoma, pilomatrixoma, proliferating trichilemmal tumor, and tubular apocrine adenoma. The pattern across nearly all of them is remarkably consistent: about nine in ten were on the head and neck, clustering on the cheek, nose, and lip; roughly a quarter presented as recurrences after a prior excision attempt; median tumor duration before Mohs was about two years; and the median number of stages required to clear was two, though the range went as high as ten stages for some lesions. Follow-up averaged well over three years, and across the entire pooled cohort, there was only a single reported recurrence after Mohs clearance — which, even acknowledging the low quality of the underlying evidence, is a notably low number.

The stated rationale for choosing Mohs over standard excision, when authors gave one, clustered around four recurring themes: cosmetically or functionally sensitive location, large tumor size, clinically ill-defined margins, and concern for locally aggressive behavior or prior recurrence. Desmoplastic trichoepithelioma is a nice illustrative case within the review — twenty-nine cases pooled from three studies, all head and neck, chosen for Mohs specifically because of poorly defined clinical margins and sensitive location, mostly cleared in one to two stages, with zero recurrences reported. Trichoblastoma and hidradenoma showed similar patterns — sensitive location, large size, or recurrent disease driving the decision, with excellent clearance and minimal recurrence.

The obvious limitation, which the authors state plainly, is that this entire evidence base is built from published case reports and small series — meaning there's inherent publication bias toward successful outcomes, no denominator of benign adnexal tumors that were treated with standard excision for comparison, and no controlled data to actually prove Mohs outperforms wide local excision for these entities. It's also worth noting selection bias in the reverse direction — these are, by definition, the adnexal tumors that were unusual enough, large enough, or recurrent enough that someone thought to write them up, so this isn't a representative sample of adnexal tumors generally.

Practically, I'd frame this as evidence-supportive rather than practice-changing, because most of us are already applying exactly this kind of clinical judgment — reaching for Mohs on the recurrent, ill-defined, or cosmetically high-stakes adnexal tumor rather than routinely using it for a straightforward, well-circumscribed lesion. What this review adds is a citable, organized evidence base you can point to when justifying that choice to a payer or tumor board, along with reassurance that the reported recurrence rate after Mohs for these tumors is very low when the indication is chosen thoughtfully. It's a useful reference document, not a new mandate.

That wraps this month's four articles — a practical combination technique for acne scarring in skin of color, reassuring reliability data behind MART-1 margin reading in melanoma Mohs, an early mechanistic look at JAK-STAT targeting in keloids, and a supportive but low-quality evidence base for selective use of Mohs in benign adnexal tumors. Thanks for listening, and I'll see you next month.