Welcome back to this June 2026 rundown from the Journal of the American Academy of Dermatology. Four items on the docket this time — two research letters, one letter-to-the-editor commentary, and one cross-sectional survey study. Let's get into it.

First up is a research letter out of Shanghai Dermatology Hospital looking at high-frequency ultrasound, or HFUS, for diagnosing subungual glomus tumors. This is really a single-center retrospective case series framed as a letter to the editor, prompted by a prior paper from Lee and colleagues that had shown plain X-ray is essentially useless for these tumors, and that MRI, while sensitive and specific, misses lesions under three millimeters and is expensive. So the gap here is whether HFUS can thread that needle — good enough resolution, but cheaper and more accessible than MRI.

The authors pulled sixty-five suspected cases over a four-year window. Twenty-nine were pathologically confirmed via excision, and thirty-six were managed conservatively based on typical HFUS findings alone, with at least six months of follow-up to confirm nothing was progressing unexpectedly. That's actually the clever part of the design — instead of only looking at surgically confirmed cases, they used the nonsurgical cohort as a real-world validation set, essentially asking whether HFUS characteristics could substitute for tissue diagnosis in guiding management. The authors' working hypothesis was that symptom severity, not imaging appearance, was what actually drove the decision to operate — and the data support that: the surgical group had significantly longer symptom duration, more pinpoint tenderness, more intermittent pain, and more cold sensitivity, while the imaging characteristics themselves — tumor size, thickness, Doppler flow pattern, stalk sign, bone remodeling — didn't differ significantly between the surgical and nonsurgical groups. In other words, HFUS appearance was consistent regardless of whether symptoms were severe enough to warrant excision.

On the imaging itself, lesions averaged about six millimeters in long diameter and just over three millimeters in thickness — meaningfully smaller than what MRI can reliably pick up, which is really the headline message. Nearly all showed detectable blood flow, about half showed intense Doppler signal, roughly a third showed the stalk sign — that's a defined feeding vessel entering the tumor — and cup-shaped bone remodeling of the underlying phalanx was seen in the large majority, around nine in ten pathologically confirmed cases, without cortical breakthrough or periosteal reaction, which helps distinguish this from a more aggressive bony process. Interestingly, only about six percent of patients presented with the classic pain-tenderness-cold sensitivity triad you were taught to expect, which the authors rightly flag as evidence that relying on classic symptoms alone will make you miss most of these.

Limitations are the ones you'd expect: retrospective, single center, a fairly small pathologically confirmed sample of twenty-nine, which limits statistical power, and no comparison against energy Doppler or other modalities within the same cohort. The authors also note a cost comparison specific to their practice setting in China — HFUS running about sixteen dollars per case versus sixty-five for MRI — which won't translate directly to US costs but underscores the directional argument.

Practical takeaway: this isn't practice-changing in the sense of introducing a new technique — most of you already know HFUS is useful for nail unit tumors — but it's a nice piece of confirmatory evidence that HFUS can detect subcentimeter glomus tumors that MRI would plausibly miss, and that its imaging phenotype is stable enough to support a genuine watch-and-wait strategy in patients with mild or absent symptoms, rather than reflexively pursuing excision or advanced imaging on every suspected case. Worth keeping in your back pocket when a patient has classic erythronychia and tenderness but wants to avoid surgery.

Next, a retrospective multicenter cohort study characterizing dermatologic adverse events from loncastuximab tesirine, the CD19-targeted antibody-drug conjugate approved for relapsed or refractory diffuse large B-cell lymphoma. The clinical gap here is straightforward — this drug's cutaneous toxicity profile has really only been described through clinical trial data and scattered case reports, so this group set out to characterize it properly in a real-world, multi-institution setting across three tertiary cancer centers.

Design-wise, this is a chart review of ninety-eight patients treated over roughly a three-to-four year window, capturing demographics, adverse event details, management, and any resulting changes to cancer therapy, with severity graded using the standard Common Terminology Criteria for Adverse Events framework. A retrospective multicenter design makes sense here simply because of the rarity and specificity of the exposure — there's no way to prospectively power a study around a fairly uncommon drug's skin toxicity, so pooling records across centers is really the only feasible way to get an adequately sized descriptive cohort.

The key finding: one in five patients developed a dermatologic adverse event. Of those, the overwhelming majority — nineteen out of twenty — had a photodistributed eruption, hitting sun-exposed sites like the face, neck, hands, forearms, and chest, with onset typically around a month after starting treatment. Severity skewed mild to moderate, roughly half grade one and half grade two, with just one grade three bullous reaction. Histology showed an interface dermatitis pattern with keratinocyte dysplasia, consistent with a cytotoxic drug reaction. Importantly, neither prior radiation exposure nor use of other photosensitizing medications was statistically associated with developing the rash — and even in patients with a radiation history, the eruptions showed up at sites distinct from the old radiation field, arguing against a radiation recall phenomenon. A smaller subset developed telangiectatic eruptions on the extremities, appearing later and in two cases persisting even after the drug was stopped — a previously described but still poorly understood reaction with this class of tesirine-based conjugates. One patient had a morbilliform eruption. Across the entire cohort, nobody required a dose hold or discontinuation specifically because of skin toxicity — everything was managed with topical steroids, antihistamines, and emollients, with systemic steroids needed only rarely.

The authors' discussion ties the photosensitivity to the tesirine payload itself — a pyrrolobenzodiazepine dimer — hypothesizing off-target cutaneous DNA damage that heightens ultraviolet sensitivity, independent of concurrent photosensitizing drugs. The obvious limitation is the retrospective design, which limits causality and standardized photo-documentation, and the authors suggest future work should look at whether these skin reactions track with tumor response, and whether proactive sun protection counseling can blunt the effect.

Practically, this is useful, actionable information even though it's not changing surgical practice — if you're the dermatologist consulted on one of these lymphoma patients, the message is reassuring: photodistributed eruptions are common, expected, almost always manageable conservatively, and shouldn't automatically trigger a conversation about holding an oncologically important drug. Proactive sun protection counseling at treatment initiation is probably the one genuinely actionable practice point here.

Third, a letter to the editor — a response piece, not original data — reacting to the recent phase three randomized trial by Hong and colleagues comparing radiotherapy against imiquimod for complex lentigo maligna. The letter writers, a group out of Tallahassee with a strong background in Grenz ray therapy, aren't presenting new data; they're contextualizing the trial's radiotherapy arm against older, more superficial radiation modalities that they argue got left out of the conversation.

Their core point is historical and technical: lentigo maligna is by definition a purely epidermal process, essentially zero depth, and the traditional treatment for that using Grenz ray — an ultrasoft radiation just above the ultraviolet range — was designed around exactly that superficiality, penetrating only half a millimeter to one and a half millimeters, delivered in just five to seven fractions, with minimal setup burden and preservation of hair follicles. They contrast that with the modalities actually used in the Hong trial — superficial energy, electron beam, intensity-modulated and dimensional-conformal radiotherapy — which penetrate far deeper, require immobilization, and are delivered daily over five to six weeks. They point out the trial's twenty-two percent alopecia rate is likely a direct consequence of using deeper-penetrating modalities than necessary for a lesion that's purely epidermal, and they suggest Grenz ray, or reviving it, might reduce that toxicity burden while maintaining efficacy, since it has one of the highest chromosomal-damage coefficients across the light spectrum for treating superficial disease. They do flag a real practical obstacle to any revival, though — the loss of national calibration standards for Grenz ray units after the reference source at Oak Ridge was decommissioned.

They also raise a fair clinical caveat about the original trial itself: a recurrence rate of about twenty-one percent within two years is fairly high, and given that lentigo maligna is known to recur well beyond the three-year mark, they'd want to see five-year data before drawing firm conclusions about radiotherapy's durability. For their own practice, they say they rarely use radiotherapy at all, favoring permanent-section slow Mohs followed by topical imiquimod, and they note — anecdotally, not as trial data — that nearly every aggressive facial melanoma they've encountered has arisen within a lentigo maligna, which is their way of pushing back against complacency toward this entity even though it behaves indolently.

There's no new methodology or results to weigh here — this is entirely a discussion-and-context piece. The practical takeaway is modest but relevant: if you're counseling poor surgical candidates about radiotherapy for lentigo maligna, it's worth knowing that not all radiotherapy is equivalent — superficial Grenz ray techniques exist as a theoretically better-matched, lower-toxicity option relative to the deeper modalities used in the pivotal trial, even though practical access to calibrated Grenz ray units today is essentially the limiting factor. Interesting historical and technical context, not something that changes today's referral patterns.

Last, a cross-sectional brief report using the National Health Interview Survey to examine skin cancer risk factors and protective behaviors in US veterans compared with nonveterans. The background problem is well established — veterans carry a higher skin cancer burden than nonveterans — but the drivers of that gap, particularly around behavior, haven't been well characterized. This study pooled nearly two decades of survey data, from 2000 through 2018, giving them a huge sample — over fifty-eight thousand veterans and more than half a million nonveterans — which is really the whole rationale for using this dataset: a nationally representative survey lets you detect and adjust for confounders across a population that's otherwise hard to study prospectively, though as always with cross-sectional design, you get association, not causation.

The headline numbers: veterans had roughly three times the odds of reporting any skin cancer history compared to nonveterans, and this held up separately for melanoma, nonmelanoma skin cancer, and unspecified skin cancer. After adjusting for the obvious confounders — age, sex, race, region, education, income, insurance, smoking, alcohol, obesity, survey year, and phototype — that association shrank considerably but remained statistically significant and clinically real, landing around a one-third to one-half increase in odds across the different skin cancer categories. That drop from a three-fold unadjusted difference down to roughly one-and-a-third to one-and-a-half after adjustment tells you a good chunk of the raw gap is explained by demographics like age and sex, but not all of it — something intrinsic to veteran status, plausibly cumulative sun exposure during service, is still doing work here.

The behavioral findings are the more actionable and somewhat counterintuitive part. Veterans were more likely to engage in physical sun protection — long sleeves, ankle-length clothing, hats, and especially caps — and more likely to have had a full-body skin exam in the past year, findings that persisted after adjustment. But veterans were significantly less likely to use sunscreen frequently, and this gap didn't fully explain away with adjustment either. They also reported less indoor tanning and fewer sunburns in the past year, though those associations became nonsignificant after adjustment, suggesting they were largely explained by demographic differences like the older age and predominantly male composition of the veteran group.

The authors' interpretation is that the elevated skin cancer burden likely traces back to intense, intermittent ultraviolet exposure during active deployment, especially in high-UV regions, and that the behavioral split — more protective clothing and screening exams, but less sunscreen — probably reflects military culture and access patterns, where physical protection and mandated exams are built into service norms, but sunscreen access and education during deployment may be more limited. The obvious limitation, which the authors state plainly, is that cross-sectional design can't establish causality, and self-reported skin cancer history and behaviors always carry recall and reporting bias.

Practically, this doesn't change how you manage an individual patient in front of you, but it's a useful population-health data point: if you have a veteran in your practice, the survey data supports being proactive about sunscreen counseling specifically, since that's the one protective behavior where they're falling behind despite otherwise good sun-protective habits and higher screening engagement. It's a reasonable evidence base for tailoring public health messaging or in-clinic counseling toward this population, even if it isn't reshaping individual surgical or diagnostic decision-making.

That wraps our four articles for this June 2026 issue — a nice mix of niche diagnostic imaging, real-world oncologic drug toxicity, a spirited historical rebuttal on radiotherapy technique, and a population-level look at who's actually getting skin cancer and why. Thanks for listening, and I'll see you next issue.