Welcome back to the podcast, where we walk through the latest issue of the Journal of the American Academy of Dermatology — this is the May twenty twenty-six issue, and we've got four pieces worth your time today, ranging from a diagnostic imaging study to a technique note, a medicolegal case series, and a short statistical letter from the editor that's genuinely worth sitting with. Let's get into it.

First up is an original study: a blinded retrospective analysis looking at line-field confocal optical coherence tomography, or LC-OCT, for evaluating atypical nevi, out of a private pigmented lesion practice in New York. The background here is one you already live with — equivocal melanocytic lesions are a diagnostic gray zone, and every noninvasive tool we have comes with tradeoffs. Reflectance confocal microscopy, the Vivascope platforms, gives great cellular resolution but loses fidelity below about a hundred to a hundred fifty microns of depth, and the smaller handheld probe sacrifices field of view. Traditional OCT penetrates deeper but its resolution is too coarse to actually characterize melanocytes. LC-OCT, introduced back in twenty eighteen, is pitched as the bridge — it gives you OCT-like depth with RCM-like cellular resolution, plus three-dimensional stacks. The gap this paper addresses is that the diagnostic criteria for LC-OCT in melanocytic lesions are still relatively new, built mostly on European data, and the authors wanted to refine and validate those criteria in their own patient population, integrated into a real-world workflow that also uses short-term mole monitoring, RCM, and the DermTech genomic melanoma test.

Methodologically, this was a one-center, one-visit pilot study with follow-up as needed, enrolling seventy-five lesions from patients with dermoscopically atypical nevi. The design logic is worth noting explicitly: rather than a randomized comparison, they used a pragmatic, risk-stratified triage protocol — small or highly suspicious lesions went straight to RCM or biopsy, while larger or less suspicious lesions got RCM, biopsy, or DermTech depending on insurance and location. That's not something a controlled trial would tolerate, but it's exactly how these lesions actually get worked up in a specialty pigmented-lesion clinic, so the authors are essentially validating imaging criteria against the ground truth their practice already generates. Every lesion got an LC-OCT scan regardless of pathway, and one blinded-adjacent structure was built in: a single physician evaluator interpreted the LC-OCT criteria, while an unblinded coauthor tracked the actual clinical and histologic outcomes separately, which is the closest they could get to blinding without a fully independent adjudication panel. Statistically, they used chi-squared or Fisher's exact testing criterion by criterion, comparing benign, atypical, and malignant-or-conservatively-re-excised categories.

On results: of the seventy-five lesions, forty-four ended up benign, sixteen were atypical without need for re-excision, and fifteen showed atypical intraepidermal proliferation, early evolving melanoma, or moderate-to-severe dysplasia requiring conservative re-excision. When they tried to separate benign from atypical as distinct groups, the differences in LC-OCT criteria were weak and not particularly useful — those two categories looked similar on imaging, which itself is a clinically honest and useful negative finding. But when they collapsed benign and atypical together and compared that group against the malignant-or-excision-recommended group, real signal emerged. Using just the subset with confirmed pathology, three criteria stood out as statistically significant discriminators: irregular honeycomb pattern, atypical dendritic cells, and moderate pleomorphism. When they expanded to the full seventy-five-lesion dataset, six criteria reached significance, adding atypical epidermal cells, moderate-to-severe pleomorphism, and clefting to the list. One more notable negative finding — there was no association between any LC-OCT criterion and DermTech result, suggesting the imaging features and the genomic assay are picking up different, non-overlapping information rather than redundant signal.

The authors' own limitations are refreshingly candid — small sample, single center, and by their own analysis, LC-OCT doesn't reliably separate benign from mildly atypical nevi, only the more meaningfully atypical or malignant end of the spectrum from everything else. There's also an inherent selection bias, since every lesion entering the study was already dermoscopically flagged as atypical, so this doesn't tell you how LC-OCT performs on a general unselected population.

Practically, this is interesting and hypothesis-generating rather than practice-changing. If you're working in a practice with access to LC-OCT, this gives you a slightly more refined shortlist of features — irregular honeycomb, atypical dendritic morphology, moderate-to-severe pleomorphism, and clefting — that appear to correlate with the higher end of the atypia spectrum. But it does not yet let you use LC-OCT to avoid biopsy on a lesion you're worried about, and it doesn't help you separate benign from mildly atypical, which is often the exact clinical decision point you're wrestling with at the bedside.

Next, a brief report and technique piece — this one describes a tissue processing method the authors call combined peripheral and deep en face margin assessment, or PDEMA, paired with bread-loafing, aimed specifically at research settings that need to confirm complete clearance of a treated skin cancer. The clinical problem is one you'll recognize from any trial of a nonsurgical basal cell carcinoma treatment — these tumors grow slowly, follow-up windows have historically run five years, topical or intralesional treatments can leave skip areas or incomplete penetration, and conventional posttreatment histology, whether vertical bread-loaf sections or horizontal en face sections, always leaves some fraction of the tissue unexamined. Standard vertical sectioning after margin clearance typically samples seven to ten micron sections at two-to-four millimeter intervals, which by definition skips large stretches of debulked tissue.

The technique itself: after the experimentally treated area is excised, first the peripheral and deep margins are inked, mapped, frozen, and sectioned horizontally exactly as in Mohs, checking for residual tumor at the margin, with additional stages taken if positive. Once margins are clear, those same frozen horizontal sections are then re-processed vertically, at ten micron thickness — but critically, at two-hundred-to-four-hundred micron intervals rather than the conventional two-to-four millimeters, which is roughly a tenfold increase in sampling density of the debulked tissue. Formalin-fixed paraffin-embedded tissue can be substituted for speed if frozen processing isn't necessary. The authors also note this pairs well with established German techniques — the Tübingen Torte and Tübingen Muffin margin methods — which can be layered in with the same vertical follow-up step.

There's no results section here in the traditional sense, since this is a methods description rather than a study, but the authors are direct about limitations: it requires additional histotechnician training, is labor- and time-intensive, and is meaningfully more costly, which is why they position it explicitly as a research tool for clinical trials evaluating novel nonsurgical modalities — not something to bring into daily Mohs practice. The takeaway for you is mostly conceptual rather than actionable at the bedside: if you're ever asked to consult on or design a trial evaluating a topical, laser, or energy-based treatment for superficial basal cell carcinoma, this gives you a rigorous, near-exhaustive histologic clearance standard to reference or request.

Sitting right alongside that piece in the brief reports section is a short, unrelated note worth a quick mention since it's directly relevant to your daily practice — a brief report flagging the dramatic shift toward prepectoral breast implant placement, which rose from a small minority of reconstructions in twenty eighteen to the large majority by twenty twenty-four. The practical point for anyone doing biopsies, excisions, or Mohs on the chest wall is that prepectoral implants sit immediately beneath skin and subcutaneous fat with no muscular buffer, raising the risk of implant compromise during cutaneous procedures, particularly on previously radiated, thinned skin. The authors offer a simple bedside trick — have the patient tense the pectoralis by pressing hands together or against the hips; skin wrinkling suggests a subpectoral implant, while a smooth, taut appearance suggests prepectoral placement — worth filing away before your next chest wall case.

Third, a retrospective descriptive study, essentially a legal-database case series, cataloguing cryotherapy-related medicolegal claims across the US and Canada from nineteen eighty-eight through twenty twenty-five. The rationale is straightforward — cryotherapy is perceived as about as low-risk and low-liability as dermatologic procedures get, but no one had actually looked systematically at malpractice exposure tied to it. The methodology was a structured search of LexisNexis for physician-performed cryotherapy cases alleging misdiagnosis, wrong indication, consent issues, supervision failures, or procedural injury, extracting specialty, setting, lesion type, allegation, outcome, and damages — this is inherently a descriptive legal-literature review rather than a clinical study, so there's no control group or statistical testing, just tabulation of what turned up.

They found twenty eligible cases over that thirty-seven-year window — a small number in absolute terms, which itself is informative. Dermatologists were the most frequently named specialty, involved in roughly half of cases, followed by family medicine, plastic surgery, and podiatry. Looking at the twenty-two lesions treated, the largest single category — over a quarter — were melanocytic nevi that had been misdiagnosed as warts or seborrheic keratoses before cryotherapy was applied, followed by warts, squamous cell carcinomas, and actinic keratoses. Of the twenty cases, just over half alleged an inappropriate indication for cryotherapy — essentially, treating something without adequate biopsy confirmation that turned out to be malignant — and forty percent alleged direct procedural injury: burns, scarring, infection, blistering, even one partial nasal amputation. Only one case involved inadequate supervision of a trainee.

Of the thirteen cases that reached final resolution, the large majority — over three-quarters — favored the defendant physician, with only two resulting in plaintiff verdicts. But those two are worth knowing in detail because they illustrate the tail risk. One involved a child with a recurrent scalp lesion repeatedly called benign and treated with liquid nitrogen over time, which turned out to be metastatic melanoma resulting in the child's death — noneconomic damages were originally set at four and a quarter million dollars, though statutory caps reduced that to two hundred fifty thousand. The other involved cryotherapy for an anal wart where excessive liquid nitrogen caused a third-degree burn with permanent scarring and chronic post-traumatic stress; the court upheld combined damages of roughly seven hundred twenty thousand dollars for past and future pain and suffering.

The authors' stated limitation is important: this only captures published, litigated cases and excludes the presumably much larger universe of out-of-court settlements, so the true liability burden from cryotherapy is certainly undercounted. Practically, this isn't practice-changing in a technical sense — nothing here suggests cryotherapy itself is riskier than assumed — but it is a useful reminder that the liability that does materialize almost always traces back to using cryotherapy as a substitute for biopsy on a lesion with any atypical features, rather than to the physical complications of the freeze itself. For a Mohs surgeon who may see referred or recurrent lesions with a cryotherapy history in the chart, it's a nice reinforcing data point for your own threshold to biopsy before treating.

Finally, a short letter from the editor, Dirk Elston, on a purely statistical theme — standardized mean difference versus the P value — and it's less a study than a teaching point, so there's no methods or limitations to walk through, just the argument itself. The core message is one you've probably internalized already but rarely see stated this bluntly in print: a P value only tells you whether an observed difference is unlikely to be due to chance, and it is highly sensitive to sample size — with a big enough cohort, almost any real effect, no matter how trivially small, will become statistically significant. What the P value cannot tell you is how big that effect actually is, and that's where the standardized mean difference comes in, expressing the size of a difference in standard deviation units, independent of how many subjects you enrolled. Elston uses the classic Physicians' Health Study aspirin example to make the point concrete — over twenty-two thousand subjects, and the reduction in myocardial infarction was significant at a vanishingly small P value, yet the actual absolute effect size was under one percent, which is why the sweeping public recommendation for aspirin prophylaxis was later walked back once smaller confirmatory studies came in. He also flags the standard confounding caveat — correlation isn't causation, illustrated with his rooster-and-sunrise line — and closes by noting that the journal is now explicitly encouraging authors to report effect size, likely confounders, and variability in outcomes alongside significance testing, not instead of it.

There's no clinical takeaway to extract in the usual sense here, but there is a durable habit worth reinforcing every time you read a paper in this journal or any other: when you see a striking P value attached to a large database or registry study — which describes an increasing share of what's published in dermatologic surgery these days — immediately ask what the actual effect size was before you let it change how you counsel or treat a patient.

That wraps this month's rundown — a promising but still-early imaging refinement for atypical nevi, a rigorous but research-only tissue-sectioning technique, a sobering reminder about documentation and biopsy thresholds before reaching for the cryo canister, and a clean statistical gut-check from the editor's desk. Thanks for listening, and I'll see you next month.