Consistency of convolutional neural networks in dermoscopic melanoma recognition: A prospective real-world study about the pitfalls of augmented intelligence.

BACKGROUND: Deep-learning convolutional neural networks (CNNs) have outperformed even experienced dermatologists in dermoscopic melanoma detection under controlled conditions. It remains unexplored how real-world dermoscopic image transformations affect CNN robustness. OBJECTIVES: To investigate the consistency of melanoma risk assessment by two commercially available CNNs to help formulate recommendations for current clinical use. METHODS: A comparative cohort study was conducted from January to July 2022 at the Department of Dermatology, University Hospital Basel. Five dermoscopic images of 116 different lesions on the torso of 66 patients were captured consecutively by the same operator without deliberate rotation. Classification was performed by two CNNs (CNN-1/CNN-2). Lesions were divided into four subgroups based on their initial risk scoring and clinical dignity assessment. Reliability was assessed by variation and intraclass correlation coefficients. Excisions were performed for melanoma suspicion or two consecutively elevated CNN risk scores, and benign lesions were confirmed by expert consensus (n = 3). RESULTS: 117 repeated image series of 116 melanocytic lesions (2 melanomas, 16 dysplastic naevi, 29 naevi, 1 solar lentigo, 1 suspicious and 67 benign) were classified. CNN-1 demonstrated superior measurement repeatability for clinically benign lesions with an initial malignant risk score (mean variation coefficient (mvc): CNN-1: 49.5(±34.3)%; CNN-2: 71.4(±22.5)%; p = 0.03), while CNN-2 outperformed for clinically benign lesions with benign scoring (mvc: CNN-1: 49.7(±22.7)%; CNN-2: 23.8(±29.3)%; p = 0.002). Both systems exhibited lowest score consistency for lesions with an initial malignant risk score and benign assessment. In this context, averaging three initial risk scores achieved highest sensitivity of dignity assessment (CNN-1: 94%; CNN-2: 89%). Intraclass correlation coefficients indicated 'moderate'-to-'good' reliability for both systems (CNN-1: 0.80, 95% CI:0.71-0.87, p < 0.001; CNN-2: 0.67, 95% CI:0.55-0.77, p < 0.001). CONCLUSIONS: Potential user-induced image changes can significantly influence CNN classification. For clinical application, we recommend using the average of three initial risk scores. Furthermore, we advocate for CNN robustness optimization by cross-validation with repeated image sets. TRIAL REGISTRATION: ClinicalTrials.gov (NCT04605822).

Improved diagnosis by automated macro- and micro-anatomical region mapping of skin photographs.

BACKGROUND: The exact location of skin lesions is key in clinical dermatology. On one hand, it supports differential diagnosis (DD) since most skin conditions have specific predilection sites. On the other hand, location matters for dermatosurgical interventions. In practice, lesion evaluation is not well standardized and anatomical descriptions vary or lack altogether. Automated determination of anatomical location could benefit both situations. OBJECTIVE: Establish an automated method to determine anatomical regions in clinical patient pictures and evaluate the gain in DD performance of a deep learning model (DLM) when trained with lesion locations and images. METHODS: Retrospective study based on three datasets: macro-anatomy for the main body regions with 6000 patient pictures partially labelled by a student, micro-anatomy for the ear region with 182 pictures labelled by a student and DD with 3347 pictures of 16 diseases determined by dermatologists in clinical settings. For each dataset, a DLM was trained and evaluated on an independent test set. The primary outcome measures were the precision and sensitivity with 95% CI. For DD, we compared the performance of a DLM trained with lesion pictures only with a DLM trained with both pictures and locations. RESULTS: The average precision and sensitivity were 85% (CI 84-86), 84% (CI 83-85) for macro-anatomy, 81% (CI 80-83), 80% (CI 77-83) for micro-anatomy and 82% (CI 78-85), 81% (CI 77-84) for DD. We observed an improvement in DD performance of 6% (McNemar test P-value 0.0009) for both average precision and sensitivity when training with both lesion pictures and locations. CONCLUSION: Including location can be beneficial for DD DLM performance. The proposed method can generate body region maps from patient pictures and even reach surgery relevant anatomical precision, e.g. the ear region. Our method enables automated search of large clinical databases and make targeted anatomical image retrieval possible.

[Clinical, pathology-associated and molecular biomarkers of hidradenitis suppurativa/acne inversa]. / Klinische, pathologische und molekulare Biomarker der Hidradenitis suppurativa/Acne inversa.

Hidradenitis suppurativa/acne inversa is a scarring disease of the intertrigines that is now intensively researched. Improved pathogenetic understanding has led to the introduction of tumor necrosis factor alpha (TNF­α) inhibition, which represents a major advance over traditional broad immunosuppression and antibiotic administration. In addition, a wide range of newer and promising treatments is or is about to be clinically evaluated. These include various specific antibodies against cytokines and the complement system and small molecules. Successful use of the individual drugs depends on the stratification of suitable patient groups with the help of clinically relevant biomarkers. While molecular investigations have shown a number of possible biomarkers and/or therapeutic target molecules, the detection of robust predictive biomarkers is still in its initial phase. In summary, the therapeutic options for hidradenitis suppurativa/acne inversa are improving through the introduction of new drugs, possibly in combination with surgical interventions, whereby the possibilities for predictive therapeutic decisions through the discovery of biomarkers would revolutionize the chances of therapeutic success.

Genomic profiling of late-onset basal cell carcinomas from two brothers with nevoid basal cell carcinoma syndrome.

BACKGROUND: Nevoid basal cell carcinoma syndrome (NBCCS) is an autosomal dominant genetic disorder. It is commonly caused by mutations in PTCH1 and chiefly characterized by multiple basal cell carcinomas (BCCs) developing prior to the age of 30 years. In rare cases, NBCCS presents with a late onset of BCC development. OBJECTIVE: To investigate BCC tumorigenesis in two brothers, who showed characteristic features of NBCCS but developed their first BCCs only after the age of 40 years. Two other siblings did not show signs of NBCCS. RESULTS: We obtained blood samples from four siblings and nine BCCs from the two brothers with NBCCS. Whole exome sequencing and RNA sequencing revealed loss of heterozygosity (LOH) of PTCH1 in eight out of nine tumours that consistently involved the same haplotype on chromosome 9. This haplotype contained a germinal splice site mutation in PTCH1 (NM_001083605:exon9:c.763-6C>A). Analysis of germline DNA confirmed segregation of this mutation with the disease. All BCCs harboured additional somatic loss-of-function (LoF) mutations in the remaining PTCH1 allele which are not typically seen in other cases of NBCCS. This suggests a hypomorphic nature of the germinal PTCH1 mutation in this family. Furthermore, all BCCs had a similar tumour mutational burden compared to BCCs of unrelated NBCCS patients while harbouring a higher number of damaging PTCH1 mutations. CONCLUSIONS: Our data suggest that a sequence of three genetic hits leads to the late development of BCCs in two brothers with NBCCS: a hypomorphic germline mutation, followed by somatic LOH and additional mutations that complete PTCH1 inactivation. These genetic events are in line with the late occurrence of the first BCC and with the higher number of damaging PTCH1 mutations compared to usual cases of NBCCS.

Acne and hidradenitis suppurativa.

Acne and hidradenitis suppurativa (HS) both centre on hair follicles. They often occur together as part of the acne tetrad, but are found in distinct localizations. Acne is primarily defined by the presence of comedones and inflammatory lesions. However, in HS the intertriginous localization and chronicity play equally important roles for the diagnosis to the inflammatory lesions. Genetics, bacteria, environmental factors and innate inflammation have all been found to play a role in acne and/or HS. Surprisingly, there is little overlap between the findings so far. The genetics of acne and HS are distinct, bacteria have not been shown convincingly to play a role in HS, and the important risk factors obesity and smoking in HS cannot be easily translated to acne. The one driving factor central to both diseases is innate inflammation, most strikingly involving interleukin-1. Hence the interleukin-1 family, as already shown in autoinflammatory conditions associated with acne, could represent attractive treatment targets.

Paradoxical ulcerative colitis during adalimumab treatment of psoriasis resolved by switch to ustekinumab.

Here we report the case of a patient with psoriasis who developed ulcerative colitis most likely caused by adalimumab. After cessation of adalimumab, colitis improved significantly. However, as psoriasis worsened, the patient was switched to ustekinumab, which resulted in complete cessation of colitis. During the 2-year follow-up under ustekinumab therapy, no further gastrointestinal complaints occurred. Paradoxical psoriasis manifestations in inflammatory bowel disease (IBD) under tumour necrosis factor (TNF)-inhibitor therapy have been reported and paradoxical IBD occurred rarely (mostly Crohn disease) in patients with rheumatological conditions treated with infliximab or etanercept. Due to the highly probable association of adalimumab with the onset of colitis in this case, we would like to suggest the term 'paradoxical ulcerative colitis' (PUC) for this as yet extremely rarely reported phenomenon. To the best of our knowledge this is the first description of PUC in a patient with psoriasis and in adalimumab treatment. Our observation suggests that ustekinumab is an effective treatment option in patients with paradoxical anti-TNF-driven inflammatory reactions like psoriasis or IBD.

Swiss S1 guideline for the treatment of rosacea.

Rosacea (in German sometimes called 'Kupferfinne', in French 'Couperose' and in Italian 'Copparosa') is a chronic and frequently relapsing inflammatory skin disease primarily affecting the central areas of the face. Its geographic prevalence varies from 1% to 22%. The differential diagnosis is wide, and the treatment is sometimes difficult and varies by stage of rosacea. For erythematous lesions and telangiectasia, intense pulsed light (IPL) therapy and lasers are popular treatment option. In addition, a vasoconstrictor agent, brimonidine, has recently been developed. For papulopustular rosacea, topical antibiotics, topical and systemic retinoids, as well as systemic antibiotics are used. A topical acaricidal agent, ivermectin, has undergone clinical development and is now on the market. In the later stages, hyperplasia of the sebaceous glands develops, resulting in phymatous growths such as the frequently observed bulbous nose or rhinophyma. Ablative laser treatments have largely replaced classical abrasive tools. Here, we reviewed the current evidence on the treatment of rosacea, provide a guideline (S1 level) and discuss the differential diagnosis of rosacea.

Genetic architecture of acne vulgaris.

Acne vulgaris is a ubiquitary skin disease characterized by chronic inflammation of the pilosebaceous unit resulting from bacterial colonization of hair follicles by Propionibacterium acnes, androgen-induced increased sebum production, altered keratinization and inflammation. Here, we review our current understanding of the genetic architecture of this intriguing disease. We analysed genomewide association studies (GWAS) and candidate genes studies for acne vulgaris. Moreover, we included GWAS studies for the associated disease polycystic ovary syndrome (PCOS). Overall, the available data revealed sixteen genetic loci flagged by single nucleotide polymorphisms (SNPs), none of which has been confirmed yet by independent studies. Moreover, a GWAS for PCOS identified 21 susceptible loci. The genetic architecture is complex which has been revealed by GWAS. Further and larger studies in different populations are required to confirm or disprove results from candidate gene studies as well to identify signals that may overlap between different populations. Finally, studies on rare genetic variants in acne and associated diseases like PCOS may deepen our understanding of its pathogenesis.

Impact of UVA on pruritus during UVA/B phototherapy of inflammatory skin diseases: a randomized double-blind study.

BACKGROUND: Narrowband (TL-01) UVB phototherapy (UVB nb) is effective in treating inflammatory skin disease. The addition of UVA is traditionally advocated to reduce pruritus, but lacks evidence for this recommendation. OBJECTIVES: The aim of this study was to assess the effect of UVB nb and UVA phototherapy in combination compared against UVB nb monotherapy on pruritus, disease activity and quality of life. METHODS: In this double-blind randomized clinical trial, 53 patients suffering from inflammatory skin diseases with pronounced itching (Visual Analogue Scale (VAS) for pruritus ≥5) were randomized into two treatment groups. One group received UVB nb (311 nm) phototherapy alone and another group received a combination of UVB nb and UVA (320-400 nm) phototherapy. UV therapy was performed three times per week over 16 weeks. Pruritus (VAS and 5-D itch score), disease activity and quality of life (Dermatology Life Quality Index, DLQI) were assessed at baseline and weeks 4, 8, 12 and 16. RESULTS: In both treatment groups, there was a reduction in pruritus scores, disease activity and DLQI. No difference in pruritus score, disease activity and quality of life could be detected between the group receiving UVB nb alone and those receiving UVB nb combined with UVA. CONCLUSIONS: Phototherapy with UVB nb alone, and UVB nb combined with UVA are equally effective in treating inflammatory skin disease and indifferent in reducing disease-associated pruritus. Given this non-inferiority for UVB nb monotherapy, the recommendation of adding UVA to UVB nb phototherapy for pruritic inflammatory skin disease should be abandoned.

The continuing evolution of targeted therapy for inflammatory skin disease.

Treatment of inflammatory skin disease has evolved from non-specific suppression of immune cells to increasingly precise targeting and modulation of immune mechanisms at all levels. This has led to dramatic treatment successes and deepened understanding of the pathophysiology. The cycle of in vitro studies, animal models, clinical trials, and case series of non-primary indications is a feedback loop that informs and guides the design of ever better disease models and therapeutic targets. Not only are we constantly discovering new molecules driving skin inflammation, we have also found that psoriasis and other autoimmune conditions are driven by distinct mediators occurring in early and late phases, which could be an opportunity for phase-specific or multipronged interventions. The deeper our mechanistic understanding, the more likely we will be able to discover subtle strategies to reprogram each patients' immune cells without having to dampen or eliminate their protective effects against pathogens and tumors. Lastly, ongoing genomic studies might soon confirm interesting genetic markers for predictive personalized medicine, the earliest currently being evaluated in psoriasis such as HLA-Cw6 and TNFAIP3. Taken together, the continued evolution of immune therapies in skin will potentially allow an unprecedented form of medicine that is not bent on silencing the pathogenic mechanism, but rather aims at using subtle interventions to shepherd the immune cell swarm back on the correct path.

Unexplained recurrent fever: when is autoinflammation the explanation?

Recurrent fever can be the sole or leading manifestation of a variety of diseases including malignancies, autoimmune diseases and infections. Because the differential diagnoses are manifold, no formal guidelines for the approach of patients with recurrent fever exists. The newly recognized group of autoinflammatory diseases are often accompanied by repetitive fever attacks. As these episodes are frequently associated by a variety of divergent presentations, the differentiation of other causes for febrile illnesses can be difficult. In this article, we first review disease entities, which frequently present with the symptom of recurrent fever. In a next step, we summarize their characteristic pattern of disease presentation. Finally, we analyse key features of autoinflammatory diseases, which are helpful to distinguish this group of diseases from the other causes of recurrent fever. Recognizing these symptom patterns can provide the crucial clues and, thus, lead to the initiation of targeted specific diagnostic tests and therapies.

How not to miss autoinflammatory diseases masquerading as urticaria.

Urticarial skin reactions are one of the most frequent problems seen by allergists and clinical immunologists in daily practice. The most common reason for recurrent wheals is spontaneous urticaria. There are, however, several less common diseases that present with urticarial rash, such as urticarial vasculitis and autoinflammatory disorders. The latter include cryopyrin-associated periodic syndrome and Schnitzler's syndrome, both rare and disabling conditions mediated by increased interleukin-1 secretion. Apart from the urticarial rash, patients are suffering from a variety of systemic symptoms including recurrent fever attacks, arthralgia or arthritis and fatigue. Autoinflammatory diseases are often associated with a diagnostic delay of many years and do not respond to antihistamines and other treatments of urticaria. Also, the chronic inflammation may lead to long-term complications such as amyloidosis. It is therefore important not to miss these diseases when diagnosing and treating patients with chronic recurrent urticarial rash. Here, we present clinical clues and tips that can help to identify autoinflammatory disorders in patients presenting with chronic urticarial rash and discuss their clinical picture and management.

Alitretinoin abrogates innate inflammation in palmoplantar pustular psoriasis.

BACKGROUND: Palmoplantar pustular psoriasis is often recalcitrant to therapy. Here we evaluated the therapeutic effect of alitretinoin in patients with recalcitrant palmoplantar pustular psoriasis and investigated subsequent immunopathological alterations. METHODS: Seven patients with palmoplantar pustular psoriasis were treated with oral alitretinoin 30 mg once daily for 12 weeks. Efficacy was assessed by palmoplantar pustular psoriasis area and severity index (PPPASI), visual analogue scales (VAS) on intensity of pain and pruritus and an overall patient assessment. Immunohistochemical staining for neutrophil elastase, CD3, CD4, CD8, CD1a CD11c, CD303,CD68, CD69, CD208 and HLA-DR was on lesional skin biopsies obtained before and after 12 weeks of treatment. RESULTS: PPPASI and VAS for pruritus and pain decreased significantly after 12 weeks of treatment with alitretinoin. The overall patient assessment ranged from 60% to 90% clinical improvement. In correlation with clinical improvement a significant reduction, particularly of neutrophils, macrophages and dendritic cells, was also observed in the skin sections. Alitretinoin was well tolerated except for headache during the first month of treatment in two patients. Limitations of the study are a missing control group and the concomitant usage of topical therapy. DISCUSSION: Our findings suggest that alitretinoin may represent a new and promising therapy for recalcitrant palmo-plantar psoriasis and warrants further controlled studies to confirm efficacy and safety of alitretinoin in this disease.

Disfiguring annular sarcoidosis improved by adalimumab.

Depending on the location, dermatoses can produce blemishes that severely impair quality of life and require highly effective treatment that is otherwise used for extensive skin involvement. We report the case of a 39-year-old, otherwise healthy male disfigured by an 8 × 7-cm hypopigmented and centrally atrophic annular plaque with erythematous indurated borders in an area of scar tissue on his forehead. Skin biopsies revealed non-caseating granulomas, and hilar involvement was identified, leading to the diagnosis of systemic sarcoidosis stage II with cutaneous involvement. The lesions proved resistant to multiple therapies, but responded within 4 months to adalimumab with regression of the lesion and inflammatory infiltrate. The visual analogue scale of disease activity decreased from 7/10 to 3.5/10, and the Dermatology Life Quality Index from 16/30 to 3/30 points. In conclusion, TNF-α inhibition can control inflammation and disfigurement by cutaneous sarcoidosis and restore quality of life.

Arsenic trioxide down-regulates antiapoptotic genes and induces cell death in mycosis fungoides tumors in a mouse model.

BACKGROUND: Mycosis fungoides (MF) is the most frequent cutaneous T-cell lymphoma (CTCL). Arsenic trioxide (As(2)O(3)) has recently been shown to be effective against leukemias, so we studied whether As(2)O(3) induces apoptosis of CTCL cells in vitro. We further investigated if As(2)O(3) is effective in a MF mouse model. MATERIAL AND METHODS: Annexin V/7-amino-actinomycin-D stainings were carried out to investigate if As(2)O(3) induced apoptosis of CTCL cell lines. To study the underlying mechanisms, the effects of As(2)O(3) on various transcription factors and apoptosis regulating proteins were analyzed by western blots, electrophoretic mobility shift assays and transcription factor enzyme-linked immunosorbent assays. The ability of As(2)O(3) to induce tumor regression was investigated in a MF mouse model. RESULTS: As(2)O(3)-induced apoptosis was paralleled by a reduction of the DNA-binding activities of transcription factors of the NFkB and signal transducer and activator of transcription gene families and reduced expression of the antiapoptotic proteins bcl-1, bcl-xL and mcl-1. Local injections of 200 muM As(2)O(3) into tumors caused complete remissions in five of six mice and one partial remission. CONCLUSIONS: As(2)O(3) induced apoptosis of CTCL cells by the down-regulation of transcription factors that stimulate the expression of antiapoptotic genes. Local injection of As(2)O(3) into MF tumor-bearing mice resulted in tumor regression.

Cholestasis protects the liver from ischaemic injury and post-ischaemic inflammation in the mouse.

BACKGROUND AND HYPOTHESIS: Cholestasis is associated with high morbidity and mortality in patients undergoing major liver surgery, but the mechanisms responsible remain elusive. Increased ischaemic liver injury and inflammation may contribute to the poor outcome. METHODS: Common bile duct ligation (biliary obstruction with hyperbilirubinaemia) or selective ligation of the left hepatic duct (biliary obstruction without hyperbilirubinaemia) was performed in C57BL/6 mice before 1 h of hepatic ischaemia and 1, 4 or 24 h of reperfusion. Infection with the intracellular hepatic pathogen Listeria monocytogenes for 12 and 48 h was used to study ischaemia-independent hepatic inflammation. RESULTS: Cholestatic mice showed considerable protection from ischaemic liver injury as determined by transaminase release, histological liver injury and neutrophil infiltration. In cholestatic mice, reduced injury correlated with a failure to activate nuclear factor kappaB (NFkappaB) and tumour necrosis factor alpha (TNFalpha) mRNA synthesis, two key mediators of post-ischaemic liver inflammation. After selective bile duct ligation, both the ligated and the non-ligated lobes showed blocked activation of NFkappaB as well as reduced induction of TNFalpha mRNA synthesis and neutrophil infiltration. By contrast, infection with L monocytogenes showed comparable activation of NFkappaB and hepatic recruitment of neutrophils 12 h after infection. CONCLUSION: Cholestasis does not increase but rather dramatically protects the liver from ischaemic injury and inflammation. This effect is mediated by a systemic factor, but not bilirubin, and is associated with a preserved capacity to trigger an inflammatory response to other stimuli such as a bacterial pathogen.