Psoriasis and eczema are chronic inflammatory skin diseases with overlapping histopathological features, which often lead to diagnostic uncertainty even among experienced dermatopathologists. To address this challenge, we developed a computer-assisted diagnostic framework that combines the Virchow foundation model, pretrained on 1.5 million whole-slide images, with multi-instance learning (MIL) to classify psoriasis and eczema from digitized histopathology slides. Using an internal dataset (n = 40) and an external validation cohort (n = 40), equally balanced between both conditions and annotated by board-certified dermatopathologists, our best-performing configuration (Virchow + CLAM) achieved 85% accuracy, a macro-averaged F1 score of 0.80, and an AUC of 0.81 on the external cohort. This substantially outperformed baseline convolutional neural networks, which reached 61% accuracy, and models relying solely on pretrained feature extractors without MIL, which achieved an average accuracy of 68.8%. In a reader study on the same external cohort, individual dermatopathologist accuracies ranged from 47.5 to 70.0%, with a majority-vote consensus accuracy of 62.5%; our method outperformed both the average individual reader and the consensus under histology-only conditions. Furthermore, the model generates attention heatmaps that provide supportive visual context by highlighting regions associated with model predictions. Importantly, this study is designed as a methodological proof-of-concept conducted under controlled, histology-only conditions and is not intended for direct clinical deployment. Rather than demonstrating clinical readiness, it illustrates the potential of domain-specific foundation models combined with MIL for addressing diagnostically challenging inflammatory dermatoses.

Introduction:Anti-PD-1 immune checkpoint inhibitors have substantially changed melanoma management in both the adjuvant and advanced settings. This study aimed to estimate the percentage of stage III and IV melanoma patients exposed to different therapies and describe treatment patterns among anti-PD-1-treated patients. Methods:We performed a retrospective chart review across 22 centers in Australia, France, Germany, South Korea, the United Kingdom, and the United States. Two patient populations were analyzed: (1) an aggregate cohort with stage III/IV melanoma (2018-2021), and (2) patients who received anti-PD-1 therapy in adjuvant or advanced settings from January 2018-12 months before data collection. Treatment patterns were evaluated. Results:In the aggregate cohort (N = 16,260), observation/watchful waiting predominated in stage III patients (56.8%) in the adjuvant setting, whereas systemic therapy was most common for stage IV disease in the advanced setting. In the anti-PD-1 cohort (N = 738), anti-PD-1 monotherapy predominated in the adjuvant setting (97.5%), while both anti-PD-1 monotherapy (63.1%) and anti-PD-1 plus anti-CTLA-4 (36.9%) were commonly used in the advanced setting. Across successive lines of therapy, anti-PD-1-based regimens remained frequently used despite substantial attrition between treatment lines. Disease progression was the leading cause of treatment discontinuation in advanced disease. Discussion:This real-world study highlights the central role of anti-PD-1 therapy across disease stages. Anti-PD-1 retreatment and reuse of anti-PD-1-based regimens were frequently observed across treatment lines, highlighting the limited availability of established post-anti-PD-1 treatment options. Novel strategies are urgently needed for patients progressing after anti-PD-1 therapy.

Cutaneous sarcoidosis presents with diverse morphologies, yet comparative data on systemic involvement, disease course and treatment response across subtypes remain limited. We conducted a 2-centre cross-sectional study including patients with histologically confirmed sarcoidosis identified through ICD codes, stratified into patients with and without cutaneous involvement and further by cutaneous subtype. Demographics, extracutaneous organ involvement, treatment response, clinical course and the organ leading to initial diagnosis were analysed. Among 125 patients (median age 53 years; 57% female), 99 had cutaneous sarcoidosis and 26 had no skin involvement. All major subtypes of cutaneous sarcoidosis were represented: plaque (39), papular (19), scar (15), subcutaneous (13) and lupus pernio (4); erythema nodosum (9) represented acute disease. Papular sarcoidosis (19) showed the mildest systemic involvement. Lymph nodes (64%) and lungs (48%) were most frequently affected with subtype specific variation. Skin lesions led to diagnosis in 78% of cases. Remission with topical therapy was rare (11%), whereas systemic glucocorticoids (21%) and other immunosuppressants (37%) showed higher response rates; sirolimus (11%), hydroxychloroquine (10%) and TNF inhibitors (8%) were most effective. Relapse occurred in 40% and 46% required long-term systemic therapy. Cutaneous sarcoidosis subtypes differ markedly in systemic involvement, treatment response and disease course, supporting biological heterogeneity and the need for individualized management.

DNA methylation provides a stable record of cellular identity, capturing epigenetic programs that distinguish specialized cell states despite a shared genome. Because malignant transformation and tumour progression are accompanied by extensive epigenetic remodeling, we hypothesized that the methylome of melanocytic lesions contains biologically and clinically relevant information for both diagnosis and disease progression. In a cohort of 1,001 tissue samples prospectively collected across eight German university hospitals profiled using Illumina Infinium MethylationEPIC arrays, we compared machine-learning models based on selected Cytosine phosphate Guanine (CpG) methylation sites with models incorporating biology-guided features, including epigenetic age acceleration, cell type composition and copy-number variation burden. In an external test set, the best diagnostic classifier was CpG-based and distinguished melanocytic nevi, noninvasive melanoma and invasive melanoma with a macro-averaged area under the receiver operating characteristic curve of 0.919 (95

Abstract Background: In-transit metastasis (ITM) in melanoma is associated with poor prognosis, yet patients show widely variable clinical outcomes from rapid progression to durable responses. To investigate the mechanisms underlying these variations, we performed multi-omics profiling of sequential tumor biopsies from an ITM melanoma patient who progressed to stage IV over 4 years. Methods: Whole-exome sequencing (WES) from 14 tumors was analyzed using ABSOLUTE for purity-adjusted variant and CNA calling. Pyclone and PhylogicNDT defined mutational clusters based on cancer cell fraction (CCF) and inferred tumor lineages. The phylogenetic map was then integrated with clinical metadata. Gene set enrichment analysis (GSEA) on bulk RNA-seq assessed lineage- and time-associated melanoma signatures and transcriptional programs. Results: Phylogenetic reconstruction captured 3 major lineages (L1: ITM; L2: ITM + distant subcutaneous metastases; L3: distant skin metastases) from a common ancestral clone. Combination of anti-PD1 and intralesional T-VEC therapy reshaped the genomic landscape of tumor clones, selectively favoring L3, which expanded into clonality while L1 became undetectable, consistent with therapy-associated selection. L2, branched off L1, also exhibited persistent survival and immune resistance. RNA-seq showed evidence for a coordinated shift from a highly differentiated state towards an AXL-high state following clinical immune intervention. Transcriptional tumor state heterogeneity increased over time, independent of lineages. Both ITM (L2) and distant (L3) lineages initially displayed more invasive, mesenchymal-like cell state profiles immediately following the clone formation, but transitioned towards more differentiated tumor states, benefiting cell proliferation and clonal expansion. By the end of the clinical course, all distant lesions across anatomical sites converged on an MITF-driven, melanocytic-like proliferative state. Conclusion: This study reconstructed a phylogenetic map for tumor evolution through multi-omics, longitudinal data from an ITM patient to reveal key pathways and tumor state changes associated with metastatic events and heterogeneity in clinical responses. In the next step, the individual phylogenetic trees will be compared with other patients in the cohort to explore intra-tumoral, inter-tumoral, and inter-patient heterogeneity in tumor progression, metastasis sites, and resistance to therapy. Citation Format: Yourong Bao, Anne Zaremba, Giuseppe Tarantino, Tuulia Vallius, Mark Woodnorth, Mariana Lopez Leon, Yingxiao Shi, Zoltan Maliga, Samira Makhzami, Tyler Aprati, Bojan Karlas, Valerie Glutsch, Bastian Schilling, Jessica Cecile Hassel, Carola Berking, Jochen Utikal, Friedegund Meier, Frank Meiss, Lucie Heinzerling, Katharina Kähler, Jiajia Chen, Lisa Zimmer, Antje Sucker, Elisabeth Livingstone, Eva Hadaschik, Christine G. Lian, George F. Murphy, Yevgeniy R. Semenov, Genevieve Boland, Peter Karl Sorger, Florian Rambow, David Liu, Dirk Schadendorf. Multi-omics analysis of longitudinal melanoma samples reveals evolutionary transitions and therapy-associated cell-state switching [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3535.

Effective T-cell activation requires both antigen recognition and co-stimulatory signaling. However, tumor cells typically lack the necessary co-stimulatory molecules, rendering T-cell receptor engagement with tumor antigens insufficient to induce potent anti-tumor immune responses. To overcome this limitation, we converted tumor cells into professional antigen-presenting cell (APC)-like cells by equipping them with the co-stimulatory molecule CD80. Electroporation of melanoma and lung cancer cell lines with CD80-encoding mRNA significantly increased their ability to stimulate tumor-antigen-specific T cells, as indicated by increased activation markers and cytokine secretion. Most importantly, CD80 expression converted melanoma cells into APC-like cells capable of priming and expanding naïve T cells that recognize endogenous tumor antigens. To enable clinical application, we subsequently used poly-(beta aminoester) (pBAE) polymers, which efficiently encapsulate mRNA into nanometric polyplexes. As previously reported, pBAE nanoparticles (NPs) efficiently penetrate plasma membranes and escape endosomal degradation both in vitro and in vivo. Here, we evaluated pBAE NPs encapsulating GFP-encoding mRNA in human cell lines derived from melanoma, lung cancer, leukemia, adenocarcinoma, uveal melanoma, and Merkel cell carcinoma. The pBAE NPs effectively transfected adherent tumor cells, and tumor spheroids in vitro and in ovo. When delivering CD80 mRNA, NPs achieved expression rates of more than 50% in most cell lines, which lasted for at least 72 h. Here as well, nanoparticle-mediated CD80 expression enabled melanoma cells to prime and expand naïve T cells recognizing endogenous tumor antigens. Importantly, T cells primed by CD80-expressing tumor cells were also capable of killing tumor cells that had not been transfected. In line with these findings, T cells stimulated with untransfected tumor cells showed high levels of checkpoint receptors CTLA-4 and PD-1, whereas the expression was significantly lower when using CD80-transfected tumor cells. This proof-of-concept study indicates that conversion of tumor cells into professional APC-like cells with CD80 mRNA is feasible and results in potent anti-tumor CD8+ T-cell responses. With the use of mRNA-loaded pBAEs, the prerequisites for a future application in vivo are given. Therefore, the application of pBAE NPs encapsulating CD80 mRNA to specifically transfect tumor cells presents a new promising strategy for the in vivo treatment of various cancers.

BACKGROUND:Erythrodermic psoriasis (EP) is a rare but severe condition. Because of its low prevalence, there are no standardized treatment recommendations for EP. Specific EP guidelines are outdated, prioritizing conventional disease-modifying antirheumatic drugs (cDMARDs) and tumor necrosis factor-alpha (TNF-α) inhibitors. METHODS:We conducted a multicenter retrospective chart analysis in five academic centers in Bavaria, Germany (Augsburg, Erlangen, LMU Munich, TU Munich, Regensburg). Patients diagnosed with EP between 2019 and 2024 who received systemic treatment were included in the study. RESULTS:A total of 29 patients were included. cDMARDs were initiated in 8 patients (27.6%). Biologics were used in 21 patients (72.4%). Psoriasis Area and Severity Index (PASI) decreased from 31.9 to 10.8 across all therapies (p < 0.001). PASI 75 was achieved with methotrexate (1), cyclosporine (1), fumarates (1), infliximab (1), ustekinumab (2), ixekizumab (1), secukinumab (2), risankizumab (4), and guselkumab (1). PASI 100 was achieved with infliximab (1), ustekinumab (1), and risankizumab (2). Adverse events occurred most frequently in the cDMARDs group. CONCLUSION:There is a wide variety of treatment approaches. Standardized guidelines are needed. Biologic therapies, especially interleukin (IL)17 and IL23-inhibitors, showed favorable outcomes in this cohort and warrant prospective evaluation.

Background Anti-programmed cell death protein 1 (PD-1) immunotherapy has revolutionized the treatment of stage III and IV melanoma. Real-world data on its resistance is needed to facilitate the development of combinatorial approaches to overcome anti-PD-1 resistance.Objectives To characterize anti-PD-1 resistance and assess whether progressive disease assigned by clinicians is concordant with scan data assessed by independent central reviewers (ICR).Methods A retrospective chart review was conducted in adult patients with stage III/IV melanoma who initiated anti-PD-1 therapy from January 2018 until 12 months before the start of data collection at 22 sites across six countries. Primary resistance and late relapse in the adjuvant setting, and primary, secondary resistance, and late progression in the advanced setting were assigned using Society for Immunotherapy of Cancer definitions. Demographic and clinical characteristics by type of resistance were compared with appropriate univariate tests. Time to resistance (TTR) and overall survival were analyzed using Kaplan-Meier. To compare the concordance of progression assigned by clinicians and ICR, the positive predictive value (PPV) was calculated in a subset of patients.Results Of 981 eligible patients, 738 were included. In the adjuvant setting (n=240), 53 (22.1%) patients developed primary resistance and 60 (25.0%) experienced late relapse. In the advanced setting (n=498), 222 (44.6%), 50 (10.0%), and 64 (12.9%) patients developed primary, secondary resistance, and late progression. Type of resistance significantly differed by country, race, type of BRAF mutation, and PD-L1 expression in both settings; and by sex, disease stage and tumor thickness in the adjuvant setting only (p<0.05). Mean (SD) TTR was 47.7 (1.3) and 24.2 (1.0) months in the adjuvant and advanced setting, respectively. Patients with primary resistance had the poorest overall survival. The PPV of progression assigned by clinicians was 87.2% (95% CI 72.6% to 95.7%).Conclusions This study showed that a substantial proportion of patients with melanoma receiving anti-PD-1 therapy in the adjuvant (47.1%) and advanced (67.5%) settings developed resistance or late relapse/progression, highlighting an unmet medical need. Real-world clinical practice provided a reliable assessment of progression. Factors associated with different types of resistance were identified. Further study is warranted to evaluate their impact on patient risk stratification. (Graphical abstract)

Distinguishing infiltrative basal cell carcinoma (BCC) from poorly differentiated cutaneous squamous cell carcinoma (cSCC) remains a significant histopathological challenge. Automated deep learning approaches hold promise for improving diagnostic reliability, yet robust external validation is essential. In this study, we developed a weakly supervised deep learning model to classify these diagnostically challenging subtypes and evaluated its generalizability across internal and external cohorts, as well as in comparison to a dermatopathology foundation model (HistoGPT). The model employed a multiple-instance learning framework (CLAM) using the histopathology-specific transformer Phikon for feature extraction from whole-slide images. Slide-level ground-truth diagnoses from the collected images (n = 335, University Hospital Erlangen) were derived from routine clinical practice and re-evaluated by two board-certified dermatopathologists. Performance was assessed on an internal test set of 84 whole-slide images (27 cSCC and 57 BCC) and two external datasets: Queensland cohort (n = 10, curated in-distribution cases) and the COBRA cohort (n = 200, broad, partly out-of-distribution cases). Model discrimination was quantified using ROC curves, while accuracy, sensitivity, and specificity were reported alongside 95% Wilson confidence intervals (CIs). On the internal test set, the model achieved perfect classification [area under the receiver operating characteristic (AUC) = 1.0; 100% accuracy, sensitivity, and specificity]. Similarly, strong performance was observed in the Queensland cohort (AUC = 1.0), although limited by sample size. In the more heterogeneous COBRA cohort, discrimination remained high (AUC = 0.923, 95% CI 0.885-0.961), requiring threshold adjustment to correct for marked calibration shift (balanced accuracy 86.5% at Youden's J). Attention heatmaps highlighted histologically meaningful regions. In zero-shot evaluation on the internal test set, HistoGPT achieved an overall accuracy of 77%, with high class-wise sensitivity for BCC (98%, 95% CI 91-100) but markedly reduced sensitivity for cSCC (33%, 95% CI 19-52). Fine-tuning a task-specific classifier on the HistoGPT backbone substantially improved performance, achieving near-perfect discrimination and 98% balanced accuracy. These findings demonstrate that weakly supervised deep learning enables highly accurate classification of diagnostically challenging BCC and cutaneous squamous cell carcinoma subtypes. However, reliable deployment across institutions necessitates careful calibration and domain adaptation, and even powerful foundation models such as HistoGPT benefit from targeted fine-tuning to ensure robust performance in dermatopathology.

Systemic sclerosis (SSc) is an autoimmune disease characterized by vasculopathy and fibrotic remodeling of the skin and internal organs. Fibrotic tissue changes are considered hardly reversible with current therapies, suggesting that new strategies are required to modulate the disease-associated molecular and cellular phenotype to enable regeneration of affected tissues. Here, analyzing skin biopsy samples from patients with SSc who had received CD19-CAR T cell therapy as part of the CASTLE study or named patient use, we demonstrate structural regeneration of SSc skin structure, as evidenced by recovery of skin papillae. Consistent with these histological changes, cyclic in situ hybridization and imaging mass cytometry analyses suggested that fibroblast populations shifted towards a physiological state, both in terms of composition and function. Moreover, we describe signs of vascular repair and changes in epidermal cell function. These results suggest that B cell depletion using CD19-CAR T cell therapy may lead to skin tissue remodeling in SSc and highlight its potential for tissue regeneration in fibrotic diseases.

BACKGROUND:Melanoma of unknown primary (MUP) accounts for up to 10% of all cases of metastatic melanoma. The origin of MUP is unclear, and previous studies on the clinical course in comparison with cutaneous melanoma have reported varying results. OBJECTIVES:The study was conducted to assess the oncogenic pattern of MUP and to compare the clinical course with that of melanoma of known cutaneous primary (MKP). METHODS:Patients with MUP diagnosed between 1999 and 2022 were included in this cohort study from the prospective multicentre real-world DeCOG (German Dermatologic Cooperative Oncology Group) registry ADOREG and from the tumour database of the University Hospital Essen. For comparison, a cohort of consecutive patients with MKP with stage III or IV disease from the skin cancer centre Essen between 1999 and 2022 was used. The median follow-up was 3.5 years. Available tumour samples were genetically analysed using next-generation sequencing. Survival outcome adjusted for age, sex and tumour stage was compared between patients with MUP and those with MKP, using the Kaplan-Meier method. RESULTS:In total, 727 patients with MUP and 587 with MKP were identified. Median age at first diagnosis was 62 years [interquartile range (IQR) 53-73] for MUP and 63 years (IQR 52-75) for MKP. Most patients were male, 448 (61.6%) with MUP and 330 (56.2%) with MKP. Of the 727 patients with MUP, 337 (46.4%) had stage III disease at first diagnosis. Overall survival (OS) at 48 months was 65% [95% confidence interval (CI) 61-69] for all patients with MUP and 68% (95% CI 64-72) for all patients with MKP. Survival analyses adjusted for age, sex and tumour stage showed comparable OS rates for patients with MUP and MKP. Genetic analyses of tumours from 110 patients with MUP detected C>T and CC→TT mutations as the most common nucleotide variations. Median tumour mutational burden was 5 mutations per Mb (95% CI 4-5). Activating BRAF V600E (c.620T>A) mutations were detected in 43 patients (39.1%), activating NRAS mutations in 37 (33.6%), RAC1 P29S mutations in 7 (6.3%) and TERT promoter mutations in 73 (66.4%) of the analysed tumour samples. No activating KIT mutations were detected. CONCLUSIONS:The oncogenic pattern of MUP showed an ultraviolet mutation signature, suggesting an origin in sun-exposed areas. OS of MUP is comparable with that of MKP.

BACKGROUND:Patients with BRAF wildtype metastatic melanoma who exhibit primary resistance to immune checkpoint inhibitors (ICIs) have a poor prognosis. Chemotherapy has been shown to induce genetic mutations, modify the tumour microenvironment and microbiome, and influence immune system activity. OBJECTIVES:To investigate in a prospective multicentre phase II trial whether two applications of an alkylating agent (dacarbazine) can sensitize patients with metastatic melanoma who are nonresponsive to ICIs to the same checkpoint inhibitor regime. METHODS:The PROMIT trial (NCT04225390) enrolled patients with histologically confirmed BRAF wildtype metastatic melanoma who exhibited primary resistance to ICI therapy. Following radiological evidence of primary resistance to ICI (ipilimumab + nivolumab or pembrolizumab) upon the first staging after initiation, patients received two doses of dacarbazine at 850 mg m-2 intravenously on days 1 and 21. Subsequently, 1 week after application of the second dose of dacarbazine, patients were rechallenged with the same ICI therapy to which they had previously shown progressive disease. RESULTS:In total, 53 patients were enrolled across four German skin cancer centres. Of these patients, 38 were evaluable for efficacy, having received at least one dose of ICI re-exposure. The overall objective response rate was 18% (95% confidence interval 0.08-0.34), with 7 of 38 patients achieving a partial response. The disease control rate was 37%. Therapy was well tolerated, with treatment-related ≥ grade 3 CTCAE adverse events occurring in 10% of patients. No new safety signals were observed. CONCLUSIONS:The study indicates that short-term chemotherapy followed by ICI rechallenge can overcome primary ICI resistance in patients with melanoma, supporting its potential as a new therapeutic option in clinical practice.

Pyoderma gangrenosum (PG) is an immune-mediated skin condition that is commonly associated with inflammatory bowel diseases (IBD). Its diagnosis and management are often challenging due to the heterogeneity of the clinical presentation and the need for an individualized therapeutic approach. We present the case of a 19-year-old Caucasian male with Crohn’s disease, who received therapy with a biological agent and developed severe PG in the retroauricular region following mastoidectomy for mastoiditis, which was complicated by cerebral venous sinus thrombosis. The diagnosis of PG was made with the help of the PARACELSUS-score based on lesion characteristics, histology, the patient’s IBD history, and lack of response to antibiotics. Treatment with upadacitinib, a selective Janus kinase 1 inhibitor, was initiated after resolution of the previous thrombotic event. The patient responded rapidly to therapy, with a notable reduction of the local inflammatory activity and wound size, and ultimately complete healing. Anticoagulation therapy did not need to be continued during treatment. The patient has been on upadacitinib for 25 months to date without adverse events or recurrence of thrombosis. This case highlights upadacitinib as a promising and safe treatment option for PG, even in patients with Crohn’s disease and a history of thrombotic events after careful benefit–risk assessment.

Immune checkpoint inhibitors (ICIs) have substantially improved outcomes in advanced melanoma but are frequently linked to immune-related adverse events (irAEs). Vitiligo is a common cutaneous irAE and has been consistently associated with improved patient outcome, including prolonged progression-free and overall survival. It also represents significant visual stigma, particularly when the face is involved. Traditional treatment comprises topical steroids, calcineurin inhibitors, laser, and phototherapy which often have insufficient effects. Since 2023, the first approved drug for non-segmental vitiligo (NSV) with facial involvement, the topical Janus kinase inhibitor ruxolitinib, has been available. However, experience with its use in ICI-induced vitiligo remains limited. In this exploratory analysis, three patients who developed facial vitiligo following ICI therapy applied 1.5% ruxolitinib cream to affected facial areas twice daily. After six (two patients), and twelve months (one patient), extensive repigmentation was observed, quantified at 95.7%, 78.9%, and 99.1% using a novel semi-automatic tool. Quality-of-life questionnaires showed mean reductions of 57.6% (Vitiligo DLQI) and 68.2% (Vitiligo-specific Quality of Life) in disease burden. Treatment was associated with substantial repigmentation without observed side effects. Further evaluation in larger, prospective cohorts is warranted to better define treatment effects, clinical applicability, and long-term safety.

BACKGROUND AND OBJECTIVES:Complete lymph node dissection (CLND) is the standard of care in patients with regional nodal melanoma macrometastasis. However, evidence on surgical procedures in the era of adjuvant systemic therapies is lacking. PATIENTS AND METHODS:This retrospective multi-center study included stage IIIB-D melanoma patients with nodal macrometastasis undergoing CLND or selective lymph node extirpation (LNE) prior to adjuvant therapy. CLND and LNE were compared regarding recurrence-free survival (RFS), nodal metastasis-free survival (NFS) and overall survival (OS). RESULTS:320 melanoma patients were included (median age 62; 55.5% male). Patients received PD-1 monotherapy (77.8%), targeted therapy (21.3%) or both sequentially (0.9%), as well as adjuvant radiotherapy in 40.9%. RFS and OS did not differ significantly between patients receiving CLND or LNE, while NFS was significantly prolonged following CLND (HR 0.3917; p = 0.005). After adjustment for risk factors by multivariate Cox regression, a prolonged RFS for CLND vs. LNE was found (HR 0.676; p = 0.04), but no benefit for OS. The rate of complications was significantly higher in the CLND group. CONCLUSIONS:CLND showed no OS benefit compared to LNE, while local control was improved. CLND can be recommended in the context of adjuvant therapies, however, the increased rate of complications should be considered.

BACKGROUND:Metastatic uveal melanoma (UM) primarily affects the liver, and due to low response rates to systemic therapies, liver-directed therapies (LDT) are commonly used for hepatic tumor control. Direct head-to-head trials comparing these LDTs are still lacking. METHODS:A retrospective multicenter explorative analysis was conducted to evaluate the clinical outcomes of transarterial chemoembolization (TACE), selective internal radiotherapy (SIRT), and percutaneous hepatic perfusion (PHP, chemosaturation). RESULTS:The present analysis was conducted on a cohort of 121 patients who had been treated for metastatic UM with LDT at five different skin cancer centers. A total of n = 15 patients received TACE (12.4%), n = 51 SIRT (42.1%), and n = 55 PHP (45.5%). The estimated median overall survival (OS) for SIRT was 25.1 months (95% confidence interval [CI], 15.3-34.8), for TACE, 11.4 months (95% CI, 7.8-14.9), and for PHP, 24.9 months (95% CI, 13.7-36.0). In the multivariate analysis that included LDH and additional systemic therapies, there was no significant difference in OS between SIRT and PHP (HR 1.14, 95% CI 0.67-1.93, P = .63), whereas TACE was found to be associated with a substantially reduced in OS compared to PHP (HR 3.25, 95% CI 1.31-8.04, P = .01). Adverse events occurred in 41.6% of patients undergoing SIRT, 64.3% of patients receiving TACE, and 84.6% of patients receiving PHP. CONCLUSION:The findings of this study indicate that SIRT and PHP demonstrate equivalent survival rates in a real-world setting.