Welcome back to the journal review. This is our June twenty twenty-six run through the Journal of the American Academy of Dermatology, and we've got four pieces to get through today — a letter challenging a SEER-based vulvar cancer study, a Veterans Affairs cohort study linking Parkinson's disease to melanoma, a TriNetX cohort study testing whether blood pressure medications affect keratinocyte carcinoma risk, and a letter reporting an association between syphilis and cutaneous squamous cell carcinoma. Let's get into it.

First up is a letter to the editor — not a new study, but a structured critique of a paper JAAD published on vulvar squamous cell carcinoma, the one using Surveillance, Epidemiology, and End Results, or SEER, registry data to look at staging, treatment timing, and survival disparities. The correspondents, a group of oncology clinicians, raise five methodological flags that are worth internalizing any time you're reading a large registry study like this. First, the cohort spans more than two decades, over which the FIGO staging system itself changed — the twenty twenty-one revision refined nodal and depth-of-invasion criteria significantly enough to cause real stage migration. Pool data across eras without harmonizing staging, and your survival curves may just be reflecting definitional drift rather than true biology. Second, SEER simply doesn't capture the pathologic granularity that actually drives adjuvant treatment decisions — margin width, lymphovascular invasion, nodal metastasis size and number, extranodal extension. Without those, any survival difference attributed to a treatment modality could just as easily be confounded by unmeasured disease aggressiveness or surgical quality. Third, they push back on the paper's conclusion that surgery alone gives the best survival for regional disease, pointing out that the field has moved toward sentinel lymph node biopsy following the GROINSS-V trials, and that era-specific variation in sentinel node adoption and radiation technique needs to be accounted for before drawing that kind of conclusion. Fourth, HPV and p16 status, along with p53 patterns, are known to stratify prognosis and radiosensitivity in vulvar SCC, and SEER doesn't capture any of that biology — so the discussion of demographic disparities should be tempered accordingly. And finally, because this is typically an elderly, comorbid population, overall survival endpoints may be muddied by non-cancer mortality, and the letter argues for competing-risk models and cancer-specific survival instead. There's no new data here — it's a methodologic gut-check. But it's a genuinely useful checklist for how to read any SEER-based oncology paper that crosses your desk, vulvar or otherwise.

Second, an original retrospective cohort study out of the Veterans Affairs system looking at Parkinson's disease and melanoma risk. The background here is that prior meta-analyses have already suggested an epidemiologic link between Parkinson's disease and melanoma, with proposed shared biology around immune dysregulation, oxidative stress, and alpha-synuclein aggregation, but the mechanism remains speculative and this hadn't been tested in a large national veteran cohort. Methodologically, this is about as large as retrospective cohorts get — over seven million veterans age fifty and up, drawn from the Veterans Affairs Cancer Registry and the Corporate Data Warehouse, followed from their first eligible primary care visit between two thousand nine and twenty twenty-two. Parkinson's disease was modeled as a time-varying exposure requiring two separate diagnostic codes at least thirty days apart, which is a sensible design choice to avoid misclassifying a single miscoded encounter as true disease, and melanoma was captured through the cancer registry rather than diagnosis codes alone, which gives more confidence in outcome ascertainment. They used Cox proportional hazards models with age as the time scale — a standard approach when age is your dominant driver of both exposure and outcome risk — and adjusted for sex, race, ethnicity, Agent Orange exposure, and comorbidity burden. The results: about two percent of this massive cohort had Parkinson's disease, and those patients were nearly twice as likely to develop melanoma after adjustment, a hazard ratio of one point eight, and that held up as statistically significant with a tight confidence interval. Veterans with Parkinson's were also older at melanoma diagnosis and had a higher proportion of head and neck melanomas compared to those without Parkinson's disease. Agent Orange exposure was also independently associated with increased melanoma hazard, but importantly, the Parkinson's association survived adjustment for that exposure, arguing against Agent Orange as the sole explanation. The authors frame this nearly two-fold hazard as comparable in magnitude to other established melanoma risk factors we already use for risk-stratified screening — things like a personal history of blistering sunburn or immunosuppression. That's a meaningful clinical anchor: this isn't a statistically significant but trivial effect, this is in the range of things that already change how we think about surveillance intervals. Limitations are the usual for administrative data — surveillance bias, since patients with Parkinson's may simply have more health system contact and therefore more opportunity for melanoma to be caught, and incomplete capture of care obtained outside the VA system. The practical takeaway: this is not yet practice-changing enough to mandate a new formal surveillance pathway, but it's a legitimate signal that patients with Parkinson's disease — particularly given the head and neck predominance suggesting a possible sun-exposure interaction — deserve a lower threshold for full skin exams and probably belong in your mental bucket of higher-risk patients, alongside transplant recipients and those with major photodamage history.

Third, a TriNetX-based retrospective cohort study asking a very practical, biologically plausible question: do renin-angiotensin-aldosterone system inhibitors — ACE inhibitors and angiotensin receptor blockers — actually reduce keratinocyte carcinoma risk, as some smaller prior studies have suggested. The rationale for even asking is that angiotensin two signaling in keratinocytes has been proposed to promote proliferation and angiogenesis, so blocking that pathway is mechanistically plausible as chemoprevention, but the existing literature has been genuinely contradictory. Methodologically, this is a nice example of a new-user, active-comparator design, which is exactly the right approach here: rather than comparing RAAS inhibitor users to non-users, which invites confounding by indication since people started on antihypertensives for different reasons look different in ways that affect cancer risk, they compared new RAAS inhibitor initiators against new calcium channel blocker initiators, a comparator with no established link to keratinocyte carcinoma. That's a much cleaner counterfactual. They ran three separate cohorts — patients with no prior keratinocyte carcinoma looking at incident risk, patients with prior keratinocyte carcinoma looking at recurrence, and transplant recipients as a high-risk subgroup — and used propensity score matching within each, achieving excellent balance, with standardized mean differences well under the threshold for concern. The results were about as clean a null as you'll see: five-year cumulative incidence of keratinocyte carcinoma was essentially identical between RAAS inhibitor and calcium channel blocker initiators, around two and a half percent in both groups, with a hazard ratio right at one. Recurrence risk among those with prior keratinocyte carcinoma was likewise unchanged, as was risk among transplant recipients, and basal cell and squamous cell subtypes analyzed separately showed the same flat, non-significant pattern across the board. This directly contradicts earlier smaller studies — one from the VATTC chemoprevention trial cohort and one from a renal transplant cohort — that had reported thirty to over fifty percent reductions in keratinocyte carcinoma with ACE inhibitor or ARB use. The authors reasonably attribute the discrepancy to those prior studies being smaller, lacking an active comparator, and studying more intensively monitored populations, rather than to any flaw in the biological hypothesis itself. Limitations here are the standard TriNetX caveats — diagnosis codes without pathology confirmation, no dosing data, and residual confounding that propensity matching can't fully eliminate. The practical takeaway is genuinely practice-relevant in a negative sense: there is no basis here for recommending RAAS inhibitors as chemoprevention, and if you've had patients or referring physicians asking about switching antihypertensives for skin cancer prevention purposes, this large, well-designed study says don't bother — treat blood pressure medication choice as blood pressure medication choice, full stop.

Fourth and last, a letter reporting a retrospective cohort study on syphilis and cutaneous squamous cell carcinoma risk, again built on TriNetX data. The premise is that syphilis produces chronic mucocutaneous inflammation, ulceration, scarring, and toll-like receptor activation, all of which are plausible carcinogenic co-factors, but this link had never actually been tested for nonmelanoma skin cancer. They compared patients with early syphilis, late syphilis, and syphilis overall against matched controls, looking at five-year risk of invasive squamous cell carcinoma — explicitly excluding in-situ disease — and basal cell carcinoma, with matching on age, sex, race and ethnicity, immunosuppressant use, and radiation history, and excluding anyone with prior phototherapy exposure to avoid that confounder. The findings: early syphilis was associated with roughly a doubling of overall squamous cell carcinoma risk, and that was statistically significant. The site-specific pattern is the more interesting part — the risk was highest on the trunk, over a four-fold increase, and also significantly elevated on the head and neck, but not on the extremities. Late syphilis showed a different pattern, with elevated truncal and extremity risk but no significant head-and-neck or overall effect. Basal cell carcinoma risk was not significantly affected in either group, which the authors reasonably interpret as consistent with distinct pathogenesis — basal cell carcinoma arising from deeper follicular structures less exposed to superficial mucocutaneous inflammation. The truncal predominance is biologically coherent too, since that's exactly where secondary syphilitic eruptions tend to concentrate, supporting a localized inflammation-driven mechanism rather than a purely systemic one. On limitations, the authors are appropriately cautious: ICD-10 coding doesn't capture disease stage, treatment status, or duration, so we can't tell whether treating the infection mitigates the risk; syphilis is heterogeneous and patients without cutaneous involvement may not share this risk profile; and there's likely behavioral confounding, since populations at higher risk for syphilis may also have higher tanning bed use or sun exposure patterns that weren't captured. The authors themselves note that the magnitude of risk here is modest compared to genuinely high-risk groups like transplant recipients, and they frame this explicitly as preliminary. Practically, this isn't practice-changing — you're not adding syphilis serology to your skin cancer risk calculus tomorrow — but it's a reasonable argument for maintaining a lower threshold for biopsying persistent or atypical lesions on the trunk in patients with a syphilis history, given that secondary syphilitic eruptions and early squamous cell carcinoma can genuinely look alike clinically, and this study is a decent reminder that clinical mimicry cuts both ways diagnostically.

That wraps up this month's four articles — a methodologic critique worth keeping in your back pocket for reading any SEER-based paper, a real signal on Parkinson's disease and melanoma risk that nudges surveillance thinking without yet mandating a new protocol, a clean and practically useful null result on RAAS inhibitors and keratinocyte carcinoma, and a preliminary but biologically coherent link between syphilis and truncal squamous cell carcinoma. Thanks for listening, and we'll see you next month.