Effective Targeting of Melanoma Cells by Combination of Mcl-1 and Bcl-2/Bcl-xL/Bcl-w Inhibitors.

Recent advances in melanoma therapy have significantly improved the prognosis of metastasized melanoma. However, large therapeutic gaps remain that need to be closed by new strategies. Antiapoptotic Bcl-2 proteins critically contribute to apoptosis deficiency and therapy resistance. They can be targeted by BH3 mimetics, small molecule antagonists that mimic the Bcl-2 homology domain 3 (BH3) of proapoptotic BH3-only proteins. By applying in vitro experiments, we aimed to obtain an overview of the possible suitability of BH3 mimetics for future melanoma therapy. Thus, we investigated the effects of ABT-737 and ABT-263, which target Bcl-2, Bcl-xL and Bcl-w as well as the Bcl-2-selective ABT-199 and the Mcl-1-selective S63845, in a panel of four BRAF-mutated and BRAF-WT melanoma cell lines. None of the inhibitors showed significant effectiveness when used alone; however, combination of S63845 with each one of the three ABTs almost completely abolished melanoma cell survival and induced apoptosis in up to 50-90% of the cells. Special emphasis was placed here on the understanding of the downstream pathways involved, which may allow improved applications of these strategies. Thus, cell death induction was correlated with caspase activation, loss of mitochondrial membrane potential, phosphorylation of histone H2AX, and ROS production. Caspase dependency was demonstrated by a caspase inhibitor, which blocked all effects. Upregulation of Mcl-1, induced by S63845 itself, as reported previously, was blocked by the combinations. Indeed, Mcl-1, as well as XIAP (X-linked inhibitor of apoptosis), were strongly downregulated by combination treatments. These findings demonstrate that melanoma cells can be efficiently targeted by BH3 mimetics, but the right combinations have to be selected. The observed pronounced activation of apoptosis pathways demonstrates the decisive role of apoptosis in the loss of cell viability by BH3 mimetics.

Federated Learning for Decentralized Artificial Intelligence in Melanoma Diagnostics.

Importance: The development of artificial intelligence (AI)-based melanoma classifiers typically calls for large, centralized datasets, requiring hospitals to give away their patient data, which raises serious privacy concerns. To address this concern, decentralized federated learning has been proposed, where classifier development is distributed across hospitals. Objective: To investigate whether a more privacy-preserving federated learning approach can achieve comparable diagnostic performance to a classical centralized (ie, single-model) and ensemble learning approach for AI-based melanoma diagnostics. Design, Setting, and Participants: This multicentric, single-arm diagnostic study developed a federated model for melanoma-nevus classification using histopathological whole-slide images prospectively acquired at 6 German university hospitals between April 2021 and February 2023 and benchmarked it using both a holdout and an external test dataset. Data analysis was performed from February to April 2023. Exposures: All whole-slide images were retrospectively analyzed by an AI-based classifier without influencing routine clinical care. Main Outcomes and Measures: The area under the receiver operating characteristic curve (AUROC) served as the primary end point for evaluating the diagnostic performance. Secondary end points included balanced accuracy, sensitivity, and specificity. Results: The study included 1025 whole-slide images of clinically melanoma-suspicious skin lesions from 923 patients, consisting of 388 histopathologically confirmed invasive melanomas and 637 nevi. The median (range) age at diagnosis was 58 (18-95) years for the training set, 57 (18-93) years for the holdout test dataset, and 61 (18-95) years for the external test dataset; the median (range) Breslow thickness was 0.70 (0.10-34.00) mm, 0.70 (0.20-14.40) mm, and 0.80 (0.30-20.00) mm, respectively. The federated approach (0.8579; 95% CI, 0.7693-0.9299) performed significantly worse than the classical centralized approach (0.9024; 95% CI, 0.8379-0.9565) in terms of AUROC on a holdout test dataset (pairwise Wilcoxon signed-rank, P < .001) but performed significantly better (0.9126; 95% CI, 0.8810-0.9412) than the classical centralized approach (0.9045; 95% CI, 0.8701-0.9331) on an external test dataset (pairwise Wilcoxon signed-rank, P < .001). Notably, the federated approach performed significantly worse than the ensemble approach on both the holdout (0.8867; 95% CI, 0.8103-0.9481) and external test dataset (0.9227; 95% CI, 0.8941-0.9479). Conclusions and Relevance: The findings of this diagnostic study suggest that federated learning is a viable approach for the binary classification of invasive melanomas and nevi on a clinically representative distributed dataset. Federated learning can improve privacy protection in AI-based melanoma diagnostics while simultaneously promoting collaboration across institutions and countries. Moreover, it may have the potential to be extended to other image classification tasks in digital cancer histopathology and beyond.

TIGIT+ NK cells in combination with specific gut microbiota features predict response to checkpoint inhibitor therapy in melanoma patients.

BACKGROUND: Composition of the intestinal microbiota has been correlated to therapeutic efficacy of immune checkpoint inhibitors (ICI) in various cancer entities including melanoma. Prediction of the outcome of such therapy, however, is still unavailable. This prospective, non-interventional study was conducted in order to achieve an integrated assessment of the connection between a specific intestinal microbiota profile and antitumor immune response to immune checkpoint inhibitor therapy (anti-PD-1 and/or anti-CTLA-4) in melanoma patients. METHODS: We assessed blood and stool samples of 29 cutaneous melanoma patients who received immune checkpoint inhibitor therapy. For functional and phenotypical immune analysis, 12-color flow cytometry and FluoroSpot assays were conducted. Gut microbiome was analyzed with shotgun metagenomics sequencing. To combine clinical, microbiome and immune variables, we applied the Random Forest algorithm. RESULTS: A total of 29 patients was analyzed in this study, among whom 51.7% (n = 15) reached a durable clinical benefit. The Immune receptor TIGIT is significantly upregulated in T cells (p = 0.0139) and CD56high NK cells (p = 0.0037) of responders. Several bacterial taxa were associated with response (e.g. Ruminococcus torques) or failure (e.g. Barnesiella intestinihominis) to immune therapy. A combination of two microbiome features (Barnesiella intestinihominis and the Enterobacteriaceae family) and one immune feature (TIGIT+ CD56high NK cells) was able to predict response to ICI already at baseline (AUC = 0.85; 95% CI: 0.841-0.853). CONCLUSIONS: Our results reconfirm a link between intestinal microbiota and response to ICI therapy in melanoma patients and furthermore point to TIGIT as a promising target for future immunotherapies.

Three-Year Overall Survival with Tebentafusp in Metastatic Uveal Melanoma.

BACKGROUND: Tebentafusp, a T-cell receptor-bispecific molecule that targets glycoprotein 100 and CD3, is approved for adult patients who are positive for HLA-A*02:01 and have unresectable or metastatic uveal melanoma. The primary analysis in the present phase 3 trial supported a long-term survival benefit associated with the drug. METHODS: We report the 3-year efficacy and safety results from our open-label, phase 3 trial in which HLA-A*02:01-positive patients with previously untreated metastatic uveal melanoma were randomly assigned in a 2:1 ratio to receive tebentafusp (tebentafusp group) or the investigator's choice of therapy with pembrolizumab, ipilimumab, or dacarbazine (control group), with randomization stratified according to the lactate dehydrogenase level. The primary end point was overall survival. RESULTS: At a minimum follow-up of 36 months, median overall survival was 21.6 months in the tebentafusp group and 16.9 months in the control group (hazard ratio for death, 0.68; 95% confidence interval, 0.54 to 0.87). The estimated percentage of patients surviving at 3 years was 27% in the tebentafusp group and 18% in the control group. The most common treatment-related adverse events of any grade in the tebentafusp group were rash (83%), pyrexia (76%), pruritus (70%), and hypotension (38%). Most tebentafusp-related adverse events occurred early during treatment, and no new adverse events were observed with long-term administration. The percentage of patients who discontinued treatment because of adverse events continued to be low in both treatment groups (2% in the tebentafusp group and 5% in the control group). No treatment-related deaths occurred. CONCLUSIONS: This 3-year analysis supported a continued long-term benefit of tebentafusp for overall survival among adult HLA-A*02:01-positive patients with previously untreated metastatic uveal melanoma. (Funded by Immunocore; IMCgp100-202 ClinicalTrials.gov number, NCT03070392; EudraCT number, 2015-003153-18.).

Liver-directed treatment is associated with improved survival and increased response to immune checkpoint blockade in metastatic uveal melanoma: results from a retrospective multicenter trial.

Metastases of uveal melanoma (UM) spread predominantly to the liver. Due to low response rates to systemic therapies, liver-directed therapies (LDT) are commonly used for tumor control. The impact of LDT on the response to systemic treatment is unknown. A total of 182 patients with metastatic UM treated with immune checkpoint blockade (ICB) were included in this analysis. Patients were recruited from prospective skin cancer centers and the German national skin cancer registry (ADOReg) of the German Dermatologic Cooperative Oncology Group (DeCOG). Two cohorts were compared: patients with LDT (cohort A, n = 78) versus those without LDT (cohort B, n = 104). Data were analyzed for response to treatment, progression-free survival (PFS), and overall survival (OS). The median OS was significantly longer in cohort A than in cohort B (20.1 vs. 13.8 months; P = 0.0016) and a trend towards improved PFS was observed for cohort A (3.0 vs. 2.5 months; P = 0.054). The objective response rate to any ICB (16.7% vs. 3.8%, P = 0.0073) and combined ICB (14.1% vs. 4.5%, P = 0.017) was more favorable in cohort A. Our data suggest that the combination of LDT with ICB may be associated with a survival benefit and higher treatment response to ICB in patients with metastatic UM.

Patterns of pseudoprogression across different cancer entities treated with immune checkpoint inhibitors.

BACKGROUND: Pseudoprogression (PsPD) is a rare response pattern to immune checkpoint inhibitor (ICI) therapy in oncology. This study aims to reveal imaging features of PsPD, and their association to other relevant findings. METHODS: Patients with PsPD who had at least three consecutive cross-sectional imaging studies at our comprehensive cancer center were retrospectively analyzed. Treatment response was assessed according to immune Response Evaluation Criteria in Solid Tumors (iRECIST). PsPD was defined as the occurrence of immune unconfirmed progressive disease (iUPD) without follow-up confirmation. Target lesions (TL), non-target lesions (NTL), new lesions (NL) were analyzed over time. Tumor markers and immune-related adverse events (irAE) were correlated. RESULTS: Thirty-two patients were included (mean age: 66.7 ± 13.6 years, 21.9% female) with mean baseline STL of 69.7 mm ± 55.6 mm. PsPD was observed in twenty-six patients (81.3%) at FU1, and no cases occurred after FU4. Patients with iUPD exhibited the following: TL increase in twelve patients, (37.5%), NTL increase in seven patients (21.9%), NL appearance in six patients (18.8%), and combinations thereof in four patients (12.5%). The mean and maximum increase for first iUPD in sum of TL was 19.8 and 96.8 mm (+ 700.8%). The mean and maximum decrease in sum of TL between iUPD and consecutive follow-up was - 19.1 mm and - 114.8 mm (-60.9%) respectively. The mean and maximum sum of new TL at first iUPD timepoint were 7.6 and 82.0 mm respectively. In two patients (10.5%), tumor-specific serologic markers were elevated at first iUPD, while the rest were stable or decreased among the other PsPD cases (89.5%). In fourteen patients (43.8%), irAE were observed. CONCLUSIONS: PsPD occurred most frequently at FU1 after initiation of ICI treatment. The two most prevalent reasons for PsPD were TL und NTL progression, with an increase in TL diameter commonly below + 100%. In few cases, PsPD was observed even if tumor markers were rising compared to baseline. Our findings also suggest a correlation between PsPD and irAE. These findings may guide decision-making of ICI continuation in suspected PsPD.

Enhanced Apoptosis and Loss of Cell Viability in Melanoma Cells by Combined Inhibition of ERK and Mcl-1 Is Related to Loss of Mitochondrial Membrane Potential, Caspase Activation and Upregulation of Proapoptotic Bcl-2 Proteins.

Targeting of MAP kinase pathways by BRAF inhibitors has evolved as a key therapy for BRAF-mutated melanoma. However, it cannot be applied for BRAF-WT melanoma, and also, in BRAF-mutated melanoma, tumor relapse often follows after an initial phase of tumor regression. Inhibition of MAP kinase pathways downstream at ERK1/2, or inhibitors of antiapoptotic Bcl-2 proteins, such as Mcl-1, may serve as alternative strategies. As shown here, the BRAF inhibitor vemurafenib and the ERK inhibitor SCH772984 showed only limited efficacy in melanoma cell lines, when applied alone. However, in combination with the Mcl-1 inhibitor S63845, the effects of vemurafenib were strongly enhanced in BRAF-mutated cell lines, and the effects of SCH772984 were enhanced in both BRAF-mutated and BRAF-WT cells. This resulted in up to 90% loss of cell viability and cell proliferation, as well as in induction of apoptosis in up to 60% of cells. The combination of SCH772984/S63845 resulted in caspase activation, processing of poly (ADP-ribose) polymerase (PARP), phosphorylation of histone H2AX, loss of mitochondrial membrane potential, and cytochrome c release. Proving the critical role of caspases, a pan-caspase inhibitor suppressed apoptosis induction, as well as loss of cell viability. As concerning Bcl-2 family proteins, SCH772984 enhanced expression of the proapoptotic Bim and Puma, as well as decreased phosphorylation of Bad. The combination finally resulted in downregulation of antiapoptotic Bcl-2 and enhanced expression of the proapoptotic Noxa. In conclusion, combined inhibition of ERK and Mcl-1 revealed an impressive efficacy both in BRAF-mutated and WT melanoma cells, and may thus represent a new strategy for overcoming drug resistance.

Stereotactic radiosurgery and combined immune checkpoint therapy with ipilimumab and nivolumab in patients with melanoma brain metastases: A retrospective monocentric toxicity analysis.

Purpose and objective: Adding stereotactic radiosurgery (SRS) to combined immune checkpoint therapy with ipilimumab and nivolumab (IPI + NIVO) has led to promising results for patients with melanoma brain metastases (MBM). This study retrospectively analyzes the toxicity profile depending on the timing of SRS with regard to IPI + NIVO. Materials and methods: For this study, the clinical database was searched for all patients with MBM who were treated with SRS and IPI + NIVO. The patients were separated into three groups: group A completed IPI + NIVO (usually up to four cycles) >14 days before SRS, in group B IPI + NIVO was initiated>14 days after SRS, and group C received SRS concurrently to IPI + NIVO. Treatment related toxicity was obtained from clinical and neuroradiological records. Analyses were performed using the Fisher-Yates-test. Results: 31 patients were assessed including six (19.4 %), seven (22.6 %) and 18 (58.1 %) patients, in groups A, B and C, respectively. Baseline prognostic markers between groups were balanced. In total, five (16.1 %) patients experienced neurological grade 3 toxicities related to SRS. All of these five patients were in group C, which was near-significantly correlated with a risk for grade 3 toxicities (p = 0.058). Post-hoc analyses showed that a maximum time period of seven days between SRS and IPI + NIVO was significantly correlated with grade 3 toxicity (p = 0.048). Conclusion: Application of SRS to IPI + NIVO within a seven-day span was related to higher toxicity rates in this retrospective analysis. After previous studies focused on immune checkpoint monotherapies with SRS and declared it as safe, this study indicates that concomitant application of IPI + NIVO and SRS might increase side effects. Prospective validation is warranted to corroborate these findings.

[Clinical and diagnostic features of skin cancer on darker skin]. / Klinische und diagnostische Besonderheiten bei Hautkrebs auf dunklerer Haut.

Although the rates of melanoma and nonmelanoma skin cancer are generally lower in people with darker skin tones compared to people with fairer skin tones, there is a higher rate of misdiagnosis and delay in diagnosis in people of color, which in turn increases the risk of a worse treatment outcome for these patients.

Oligometastatic disease and visceral resections in advanced malignant melanoma: a propensity-matched analysis.

PURPOSE: Malignant melanoma is among the tumours with the highest increase in incidence of solid tumours in Germany. While most patients are diagnosed at an early stage and show a good prognosis, advanced stages of malignant melanoma are accompanied with a poor prognosis and limited treatment options. Comparable to other tumour entities, the resection of visceral metastases could lead to a better prognosis. Supplementary, the subgroup of oligometastatic patients might benefit from surgical therapy to a greater extent. METHODS: This retrospective study analysed 351 patients treated between 2006 and 2017 at the University Hospital of Cologne. A total of 121 patients showed visceral metastases, with which we compared patients with a diffuse tumour spread to patients in an oligometastatic state. Furthermore, we evaluated the effect of visceral resection of oligometastatic, malignant melanoma. RESULTS: Our analysis showed that patients with an oligometastatic malignant melanoma had a significantly better prognosis than patients with a diffuse pattern of metastases, if they showed visceral metastases. Furthermore, the resection of visceral metastases leads to a significant gain in median overall survival time (13.6 vs. 34.2 months) and in progression-free survival (9.6 vs. 3.8 months). CONCLUSION: The resection of visceral metastases is a rational treatment option in advanced malignant melanoma. Although our study is limited by a small cohort of patients (n = 18), we believe that the resection of visceral metastases will be fundamental in the treatment of malignant melanoma. In particular, patients in an oligometastatic stage could be an eligible group for surgical treatment.

Immune checkpoint inhibition and targeted therapy for melanoma: A patient-oriented cross-sectional comparative multicentre study.

BACKGROUND: Choosing the adequate systemic treatment for melanoma is driven by clinical parameters and personal preferences. OBJECTIVE: Evaluation of the impact of disease and treatment on the daily life of patients receiving systemic therapy for melanoma. METHODS: A German-wide, cross-sectional comparative study was conducted at 13 specialized skin cancer centres from 08/2020 to 03/2021. A questionnaire was distributed to assess patients' perception of disease and symptoms, the impact of their current treatment on quality of life (QOL) and activities, adverse events (AEs), therapeutic visits, as well as believe in and satisfaction with their current systemic melanoma treatment. Patient-reported outcomes (PROs) were rated on a continuous numerical rating scale or selected from a given list. RESULTS: Four hundred and fourteen patients with systemic melanoma therapy were included. 359 (87%) received immune checkpoint inhibition (ICI) and 55 (13%) targeted therapy (TT). About 1/3 of patients were adjuvantly treated, the remaining because of unresectable/metastatic melanoma. In subgroup analyses, only in the adjuvant setting, TT patients reported a significant decrease in their treatment associated QOL compared to patients with ICI (p = 0.02). Patients with TT were 1.9 times more likely to report AEs than patients with ICI, a difference being significant just for the adjuvant setting (p = 0.01). ICI treatment intervals differed significantly between adjuvant and unresectable/metastatic setting (p = 0.04), though all patients, regardless of their specific ICI drug, evaluated their treatment frequency as adequate. TT patients with dabrafenib/trametinib (n = 37) or encorafenib/binimetinib (n = 15) did not differ regarding the strain of daily pill intake. Patients older than 63 years rated various PROs better than younger patients. CONCLUSIONS: Patients evaluated their treatment mainly positively. ICI might be preferred over TT regarding QOL and patient-reported AEs in the adjuvant setting. Older melanoma patients appeared to be less impacted by their disease and more satisfied with their treatment.

Real-world outcomes using PD-1 antibodies and BRAF + MEK inhibitors for adjuvant melanoma treatment from 39 skin cancer centers in Germany, Austria and Switzerland.

BACKGROUND: Programmed death-1 (PD-1) antibodies and BRAF + MEK inhibitors are widely used for adjuvant therapy of fully resected high-risk melanoma. Little is known about treatment efficacy outside of phase III trials. This real-world study reports on clinical outcomes of modern adjuvant melanoma treatment in specialized skin cancer centers in Germany, Austria and Switzerland. METHODS: Multicenter, retrospective study investigating stage III-IV melanoma patients receiving adjuvant nivolumab (NIV), pembrolizumab (PEM) or dabrafenib + trametinib (D + T) between 1/2017 and 10/2021. The primary endpoint was 12-month recurrence-free survival (RFS). Further analyses included descriptive and correlative statistics, and a multivariate linear-regression machine learning model to assess the risk of early melanoma recurrence. RESULTS: In total, 1198 patients from 39 skin cancer centers from Germany, Austria and Switzerland were analysed. The vast majority received anti PD-1 therapies (n = 1003). Twelve-month RFS for anti PD-1 and BRAF + MEK inhibitor-treated patients were 78.1% and 86.5%, respectively (hazard ratio [HR] 1.998 [95% CI 1.335-2.991]; p = 0.001). There was no statistically significant difference in overall survival (OS) in anti PD-1 (95.8%) and BRAF + MEK inhibitor (96.9%) treated patients (p > 0.05) during the median follow-up of 17 months. Data indicates that anti PD-1 treated patients who develop immune-related adverse events (irAEs) have lower recurrence rates compared to patients with no irAEs (HR 0.578 [95% CI 0.443-0.754], p = 0.001). BRAF mutation status did not affect overall efficacy of anti PD-1 treatment (p > 0.05). In both, anti PD-1 and BRAF + MEK inhibitor treated cohorts, data did not show any difference in 12-month RFS and 12-month OS comparing patients receiving total lymph node dissection (TLND) versus sentinel lymph node biopsy only (p > 0.05). The recurrence prediction model reached high specificity but only low sensitivity with an AUC = 0.65. No new safety signals were detected. Overall, recorded numbers and severity of adverse events were lower than reported in pivotal phase III trials. CONCLUSIONS: Despite recent advances in adjuvant melanoma treatment, early recurrence remains a significant clinical challenge. This study shows that TLND does not reduce the risk of early melanoma recurrence and should only be considered in selected patients. Data further highlight that variables collected during clinical routine are unlikely to allow for a clinically relevant prediction of individual recurrence risk.

Dataset of a retrospective multicenter cohort study on characteristics of immune checkpoint inhibitor-induced encephalitis and comparison with HSV-1 and anti-LGI1 encephalitis.

Over the past decade, cancer immunotherapy with immune checkpoint inhibitors (ICIs) has significantly improved the outcome of many malignancies. However, with the broad use of ICIs, neurological immune related adverse events (irAE) are increasingly recognized. ICI-induced encephalitis (ICI-iE) is a particularly severe irAE, often leading to treatment termination, long-term sequalae or death. Despite its high morbidity and mortality, data on clinical features and diagnostic criteria are limited. We aimed to define clinical, radiologic and laboratory characteristics of ICI-iE and identify factors that discriminate it from anti-leucine-rich glioma-inactivated (anti-LGI)-1 encephalitis and herpes simplex virus (HSV)-1 encephalitis - two alternative causes of encephalitis - to increase the awareness of ICI-iE and improve its diagnosis and management. To that end, we retrospectively collected 30 cases of ICI-iE that were reported to the Side Effect Registry Immuno-Oncology (SERIO) and 46 cases of anti-LGI1 encephalitis or herpes simplex virus (HSV)-1 encephalitis that presented to a large German neurological referral center (Charité Universitätsmedizin Berlin) between January 2015 and September 2021. Signs and symptoms, imaging and electroencephalogram features, laboratory findings and outcome measures were assessed using standardized case report forms as well as patients' medical records and compared between the groups. The data reported here represents the largest primary cohort of patients with ICI-iE to date and the first comparison with other types of encephalitis. As all three disorders - ICI-iE, HSV-1 encephalitis and anti-LGI1 encephalitis - are rare neurological entities, this dataset can be used as a reference in future clinical studies on ICI-induced neurotoxicity, neurological autoimmune disorders, and central nervous system infections.

Characteristics of immune checkpoint inhibitor-induced encephalitis and comparison with HSV-1 and anti-LGI1 encephalitis: A retrospective multicentre cohort study.

AIM: Immune checkpoint inhibitor-induced encephalitis (ICI-iE) is a rare but life-threatening toxicity of immune checkpoint inhibitor treatment. We aim to identify the characteristics of ICI-iE and describe factors that discriminate it from herpes simplex virus (HSV)-1 encephalitis and anti-leucine-rich glioma-inactivated 1 (anti-LGI1) encephalitis, as two alternative entities of encephalitis. METHODS: In this retrospective multicentre cohort study, we collected patients with ICI-iE reported to the Side Effect Registry Immuno-Oncology from January 2015 to September 2021 and compared their clinical features and outcome with 46 consecutive patients with HSV-1 or anti-LGI1 encephalitis who were treated at a German neurological referral centre. RESULTS: Thirty cases of ICI-iE, 25 cases of HSV-1 encephalitis and 21 cases of anti-LGI1 encephalitis were included. Clinical presentation of ICI-iE was highly variable and resembled that of HSV-1 encephalitis, while impairment of consciousness (66% vs. 5%, p = .007), confusion (83% vs. 43%; p = .02), disorientation (83% vs. 29%; p = .007) and aphasia (43% vs. 0%; p = .007) were more common in ICI-iE than in anti-LGI1 encephalitis. Antineuronal antibodies (17/18, 94%) and MRI (18/30, 60%) were mostly negative in ICI-iE, but cerebrospinal fluid (CSF) showed pleocytosis and/or elevated protein levels in almost all patients (28/29, 97%). Three patients (10%) died of ICI-iE. Early immunosuppressive treatment was associated with better outcome (r = 0.43). CONCLUSIONS: ICI-iE is a heterogeneous entity without specific clinical features. CSF analysis has the highest diagnostic value, as it reveals inflammatory changes in most patients and enables the exclusion of infection. Early treatment of ICI-iE is essential to prevent sequelae and death.

Model soups improve performance of dermoscopic skin cancer classifiers.

BACKGROUND: Image-based cancer classifiers suffer from a variety of problems which negatively affect their performance. For example, variation in image brightness or different cameras can already suffice to diminish performance. Ensemble solutions, where multiple model predictions are combined into one, can improve these problems. However, ensembles are computationally intensive and less transparent to practitioners than single model solutions. Constructing model soups, by averaging the weights of multiple models into a single model, could circumvent these limitations while still improving performance. OBJECTIVE: To investigate the performance of model soups for a dermoscopic melanoma-nevus skin cancer classification task with respect to (1) generalisation to images from other clinics, (2) robustness against small image changes and (3) calibration such that the confidences correspond closely to the actual predictive uncertainties. METHODS: We construct model soups by fine-tuning pre-trained models on seven different image resolutions and subsequently averaging their weights. Performance is evaluated on a multi-source dataset including holdout and external components. RESULTS: We find that model soups improve generalisation and calibration on the external component while maintaining performance on the holdout component. For robustness, we observe performance improvements for pertubated test images, while the performance on corrupted test images remains on par. CONCLUSIONS: Overall, souping for skin cancer classifiers has a positive effect on generalisation, robustness and calibration. It is easy for practitioners to implement and by combining multiple models into a single model, complexity is reduced. This could be an important factor in achieving clinical applicability, as less complexity generally means more transparency.

Development of Lymphopenia during Therapy with Immune Checkpoint Inhibitors Is Associated with Poor Outcome in Metastatic Cutaneous Melanoma.

Predictive markers for immune checkpoint inhibitor (ICI) therapy are needed. Thus, baseline blood counts have been investigated as biomarkers, showing that lymphopenia at the start of therapy with (ICI) is associated with a worse outcome in metastatic melanoma. We investigated the relationship between the occurrence of lymphopenia under ICI and disease outcome. Patients with metastatic melanoma who had undergone therapy with ICI were identified in our database. Only patients with a normal lymphocyte count at baseline were included in this retrospective study. Progression-free survival (PFS) and overall survival (OS) were compared between patients in which lymphopenia occurred during ICI therapy and those who did not develop lymphopenia. In total, 116 patients were analyzed. Lymphopenia occurred in 42.2% of patients, with a mean onset after 17 weeks (range 1-180 weeks). The occurrence of lymphopenia during immunotherapy was significantly associated with a shorter PFS and OS. Patients who developed lymphopenia (n = 49) had a mean PFS of 13.3 months (range 1-67 months) compared to 16.9 months (range 1-73 months) for patients who did not develop lymphopenia (n = 67; p = 0.025). Similarly, patients with lymphopenia had a significantly shorter OS of 28.1 months (range 2-70 months) compared with 36.8 months (range 4-106 months) in patients who did not develop lymphopenia (p = 0.01). Patients with metastatic melanoma who develop lymphopenia during ICI therapy have a worse prognosis with significantly shorter PFS and OS compared with patients who do not develop lymphopenia.

Chlormethin-Gel zur Behandlung der Mycosis fungoides: Ein Expertenkonsens aus Deutschland, Österreich und der Schweiz (DACH-Region) zum Therapiemanagement.

HINTERGRUND: Chlormethin-Gel ist in Europa zur Therapie von Patienten mit Mycosis fungoides in allen Krankheitsstadien zugelassen. Die optimalen Behandlungsregime hinsichtlich Frequenz, Dosierung, Kombinations- oder Erhaltungstherapien sind noch nicht vollständig etabliert. METHODIK: Zehn in der Erforschung und Behandlung kutaner T-Zell-Lymphome erfahrene Experten aus Deutschland, Österreich und der Schweiz (DACH-Region) wurden schriftlich zu Indikation, Anwendungsfrequenz, Beurteilung des Therapieerfolgs, Begleittherapie, Nebenwirkungen, Kombinationstherapien in späteren Krankheitsstadien, Erhaltungstherapie und Adhärenz im Rahmen der Therapie der Mycosis fungoides mit Chlormethin-Gel befragt. Die strukturiert aufbereiteten Ergebnisse der Umfrage wurden in einer Konsensuskonferenz diskutiert und Empfehlungen zum Management der Therapie mit Chlormethin-Gel entwickelt. ERGEBNISSE: Wesentlich für die Therapie mit Chlormethin-Gel ist ein individuelles, symptomorientiertes Therapiemanagement. Systemische Nebenwirkungen des Wirkstoffs sind wegen der fehlenden systemischen Verfügbarkeit bei topischer Anwendung unwahrscheinlich. Die häufig auftretende allergische oder irritativ-toxische Kontaktdermatitis kann durch eine Anpassung des Therapieregimes, Therapiepausen sowie nebenwirkungsspezifische und unterstützende Maßnahmen häufig beherrscht werden. Ein einschleichender Therapiebeginn mit Anwendung von Chlormethin-Gel jeden zweiten Tag kann die Tolerabilität wesentlich verbessern, insbesondere wenn die Therapie alternierend mit topischen Kortikosteroiden erfolgt. SCHLUSSFOLGERUNGEN: Die Anwendung von Chlormethin-Gel bei Mycosis fungoides wird durch die begleitende Kontaktdermatitis häufig eingeschränkt. Mit einem geeigneten Therapie- und Nebenwirkungsmanagement können vermeidbare Therapieabbrüche verhindert werden und mehr Patienten von der Therapie profitieren.

Explainable artificial intelligence in skin cancer recognition: A systematic review.

BACKGROUND: Due to their ability to solve complex problems, deep neural networks (DNNs) are becoming increasingly popular in medical applications. However, decision-making by such algorithms is essentially a black-box process that renders it difficult for physicians to judge whether the decisions are reliable. The use of explainable artificial intelligence (XAI) is often suggested as a solution to this problem. We investigate how XAI is used for skin cancer detection: how is it used during the development of new DNNs? What kinds of visualisations are commonly used? Are there systematic evaluations of XAI with dermatologists or dermatopathologists? METHODS: Google Scholar, PubMed, IEEE Explore, Science Direct and Scopus were searched for peer-reviewed studies published between January 2017 and October 2021 applying XAI to dermatological images: the search terms histopathological image, whole-slide image, clinical image, dermoscopic image, skin, dermatology, explainable, interpretable and XAI were used in various combinations. Only studies concerned with skin cancer were included. RESULTS: 37 publications fulfilled our inclusion criteria. Most studies (19/37) simply applied existing XAI methods to their classifier to interpret its decision-making. Some studies (4/37) proposed new XAI methods or improved upon existing techniques. 14/37 studies addressed specific questions such as bias detection and impact of XAI on man-machine-interactions. However, only three of them evaluated the performance and confidence of humans using CAD systems with XAI. CONCLUSION: XAI is commonly applied during the development of DNNs for skin cancer detection. However, a systematic and rigorous evaluation of its usefulness in this scenario is lacking.