Search

Purpose: The randomized phase III coBRIM study (NCT01689519) demonstrated improved progression-free survival (PFS) and overall survival (OS) with addition of cobimetinib to vemurafenib compared with vemurafenib in patients with previously untreated BRAFV600 mutation–positive advanced melanoma. We report long-term follow-up of coBRIM, with at least 5 years since the last patient was randomized. Patients and Methods: Eligible patients were randomized 1:1 to receive either oral cobimetinib (60 mg once daily on days 1–21 in each 28-day cycle) or placebo in combination with oral vemurafenib (960 mg twice daily). Results: 495 patients were randomized to cobimetinib plus vemurafenib (n = 247) or placebo plus vemurafenib (n = 248). Median follow-up was 21.2 months for cobimetinib plus vemurafenib and 16.6 months for placebo plus vemurafenib. Median OS was 22.5 months (95% CI, 20.3–28.8) with cobimetinib plus vemurafenib and 17.4 months (95% CI, 15.0–19.8) with placebo plus vemurafenib; 5-year OS rates were 31% and 26%, respectively. Median PFS was 12.6 months (95% CI, 9.5–14.8) with cobimetinib plus vemurafenib and 7.2 months (95% CI, 5.6–7.5) with placebo plus vemurafenib; 5-year PFS rates were 14% and 10%, respectively. OS and PFS were longest in patients with normal baseline lactate dehydrogenase levels and low tumor burden, and in those achieving complete response. The safety profile remained consistent with previously published reports. Conclusions: Extended follow-up of coBRIM confirms the long-term clinical benefit and safety profile of cobimetinib plus vemurafenib compared with vemurafenib monotherapy in patients with BRAFV600 mutation–positive advanced melanoma.

Semiconductor nanowires are unique materials for studying nanoscale phenomena; the possibility of forming silicon nanowires on bulk silicon-on-insulator substrates in a top-down process ensures complete incorporation of this technology into integrated electronic systems. In addition, the use of ferromagnetic contacts in combination with the high quality of ferromagnetic–semiconductor interfaces open up prospects for the use of such structures in spintronics devices, in particular, spin transistors. A simple approach is proposed to create semiconductor nanowire-based active devices, specifically, bottom-gate Schottky-barrier field-effect transistors with a metal (Fe) source and drain synthesized on a silicon-on-insulator substrate and the transport characteristics of the designed transistors are investigated.

This article studies the problem of social activity of the students of the Sverdlovsk Region in the context of intensified digitalization of the social space caused by the forced self-isolation. The purpose of this paper is to analyze the self-regulation of students’ social activity in the conditions of digitalization of the social space. Following the traditions of the Ural scientific sociological school and the concepts of multidimensionality and self-organization of youth, as well as using community, activity, and other sociological approaches, the authors use a terminological series of concepts of the student educational community and its social activity. They distinguish the terms “classical (traditional) community”, acting in a real socio-cultural space, “virtual community”, the social action of which is transferred to cyberspace, and “digital community”, occupying an intermediate position and acting in a multidimensional real-virtual sociocultural space. The social activity of students is considered as one of the types (along with educational and professional) social activity and a form of actual general cultural socialization. Employing the materials of empirical research in 2020 and with the involvement of data from the secondary analysis of regional monitoring, the authors study the factors, intensity, and dynamics of social activity of students. The results reveal the ideas of students about people serving as role models and multidirectional tendencies in value attitudes — to help society and specific people and to material success and comfort. New data were obtained on the social problems of concern to students, the preferred type of public activity, attitude towards the citizenship of the Russian Federation, the dynamics of participation in electoral processes, collective actions, participation in the work of public organizations and charitable events. The attitude of students to youth public organizations and conditions for active participation of young people in social and political activities is revealed. The authors conclude on the reorientation of a significant part of the student body towards the priority of the personal over the public and about the further stratification and atomization of the student community.

A method for interpreting uncertainty of predictions provided by machine learning survival models is proposed. It is called UncSurvEx and aims to determine which features of an analyzed example lead to uncertain predictions of an explainable black-box survival model. One of the ideas behind the proposed method is to approximate the uncertainty measure of a local black-box survival model prediction by the uncertainty measure of the Cox proportional hazards model at the local area around a test example. The linear relationship between covariates and predictions in the corresponding Cox model allows determining quantitative impacts of covariates on the uncertainty measure. A specific certainty measure of the survival function, taking into account the most uncertain survival function, is introduced to interpret the prediction uncertainty. The $L_{2}$ -norm is used to compute the distance between survival functions. The method leads to an unconstrained non-convex optimization problem which is solved by means of the well-known Broyden–Fletcher–Goldfarb–Shanno algorithm. A lot of numerical experiments demonstrate the uncertainty interpretation method.

The choice of first-line therapy for metastatic skin melanoma has become quite a difficult task in the last few years: a practicing oncologist has got drugs of different classes with different mechanisms of action – immune checkpoint blockers and BRAF/MEK tyrosine kinase inhibitors. In this study, the authors have tried to systematize their own experience and current scientific information about the treatment of patients with metastatic skin melanoma in the first and subsequent lines of therapy. The paper discusses the importance of correct initial assessment of the patient’s condition and prevalence of the process, the role of currently available biomarkers in metastatic skin melanoma, as well as the effectiveness and safety of possible first-line therapy options for this disease. Due to the lack of data for direct comparison of combined immunotherapy, combined targeted therapy, triple combination of BRAFi + MEKi + aPDl1 blockers and aPD1 monotherapy obtained from randomized studies, we have to draw parallels between studies with different diagnoses and different patient populations. Since the practicing oncologist has to do the same thing in his or her daily work, we believe that our review will be very useful for these doctors. The article also discusses the possibility of using drug therapy in the second and subsequent lines and a successful strategy to return to previous treatment in some cases. The clinical observation of a patient with metastatic foot skin melanoma, which has been observed in our department since 2010, is given as an illustration of the success of sequential treatment tactics with different drugs. Clinical observation shows the success of returning to some previously used therapy lines, including after the progression on them.

A new adaptive weighted deep forest algorithm which can be viewed as a modification of the confidence screening mechanism is proposed. The main idea underlying the algorithm is based on adaptive weigting of every training instance at each cascade level of the deep forest. The confidence screening mechanism for the deep forest proposed by Pang et al., strictly removes instances from training and testing processes to simplify the whole algorithm in accordance with the obtained random forest class probability distributions. This strict removal may lead to a very small number of training instances at the next levels of the deep forest cascade. The presented modification is more flexible and assigns weights to instances in order to differentiate their use in building decision trees at every level of the deep forest cascade. It overcomes the main disadvantage of the confidence screening mechanism. The proposed modification is similar to the AdaBoost algorithm to some extent. Numerical experiments illustrate the outperformance of the proposed modification in comparison with the original deep forest. It is also illustrated how the proposed algorithm can be extended for solving the transfer learning and distance metric learning problems.

The effects of itraconazole, a strong CYP3A4 inhibitor, on the steady-state pharmacokinetics of vemurafenib were evaluated in a phase 1, multicenter, open-label, fixed-sequence study. Patients with BRAFV600 mutation-positive metastatic malignancies received oral vemurafenib 960 mg twice daily on days 1 to 20 (period A) and oral vemurafenib 960 mg twice daily with oral itraconazole 200 mg once daily on days 21 to 40 (period B). A mixed-effects analysis of variance model was used to compare log-transformed area under the concentration-time curve during the dosing interval and maximum plasma concentration values for vemurafenib in 8 patients between period B (vemurafenib plus itraconazole) and period A (vemurafenib alone). Multiple doses of itraconazole increased steady-state exposure of vemurafenib by approximately 40%, with geometric least squares mean ratios (period B/period A) of 140% (90% confidence interval, 121-161) for both maximum plasma concentration and area under the concentration-time curve during the dosing interval. There was no apparent increase in incidence or severity of adverse events during coadministration of vemurafenib with itraconazole. In conclusion, coadministration of itraconazole with vemurafenib resulted in a modest increase in exposure of vemurafenib at steady state and was generally well tolerated.

Summary Background IMspire150 aimed to evaluate first-line combination treatment with BRAF plus MEK inhibitors and immune checkpoint therapy in BRAFV600 mutation-positive advanced or metastatic melanoma. Methods IMspire150 was a randomised, double-blind, placebo-controlled phase 3 study done at 112 institutes in 20 countries. Patients with unresectable stage IIIc–IV, BRAFV600 mutation-positive melanoma were randomly assigned 1:1 to 28-day cycles of atezolizumab, vemurafenib, and cobimetinib (atezolizumab group) or atezolizumab placebo, vemurafenib, and cobimetinib (control group). In cycle 1, all patients received vemurafenib and cobimetinib only; atezolizumab placebo was added from cycle 2 onward. Randomisation was stratified by lactate dehydrogenase concentration and geographical region. Blinding for atezolizumab was achieved by means of an identical intravenous placebo, and blinding for vemurafenib was achieved by means of a placebo tablet. The primary outcome was investigator-assessed progression-free survival. This trial ( ClinicalTrials.gov , NCT02908672 ) is ongoing but no longer recruiting patients. Findings Between Jan 13, 2017, and April 26, 2018, 777 patients were screened and 514 were enrolled and randomly assigned to the atezolizumab group (n=256) or control group (n=258). At a median follow-up of 18·9 months (IQR 10·4–23·8), progression-free survival as assessed by the study investigator was significantly prolonged with atezolizumab versus control (15·1 vs 10·6 months; hazard ratio [HR] 0·78; 95% CI 0·63–0·97; p=0·025). Common treatment-related adverse events (>30%) in the atezolizumab and control groups were blood creatinine phosphokinase increased (51·3% vs 44·8%), diarrhoea (42·2% vs 46·6%), rash (40·9%, both groups), arthralgia (39·1% vs 28·1%), pyrexia (38·7% vs 26·0%), alanine aminotransferase increased (33·9% vs 22·8%), and lipase increased (32·2% vs 27·4%); 13% of patients in the atezolizumab group and 16% in the control group stopped all treatment because of adverse events. Interpretation The addition of atezolizumab to targeted therapy with vemurafenib and cobimetinib was safe and tolerable and significantly increased progression-free survival in patients with BRAFV600 mutation-positive advanced melanoma. Funding F Hoffmann–La Roche and Genentech.

This paper addresses the Ekthesis Chronica (Ἔκθεσις χρονική), a Greek chronicle compiled by an anonymous cleric of the Patriarchate of Constantinople in the first half of the sixteenthcentury, which encompassed the events of the Late Byzantine and Early Ottoman history. Its distinctive feature is a recurrent alternation of seemingly mutually excluding points of view. Its neighboring chapters comply with the demands of different genres, accepting the set of values associated with them. The imaginary world of the chapters dealing with the events prior to 1453 reminds the reader of the heroic world of chivalric romances. The chapters describing the fall of Constantinople are may be read as a prosaic lamentation of the loss of the city which embodied the Byzantine civilization as a whole. In the post-Byzantine section, there appeared three approaches to the Ottoman rule over the Greeks. Whenever the chronicle-writer switches to the apocalyptic mode, the sultan becomes an infidel murderer of Christians. If, by contrast, he adopts the aretalogic (hagiographic) mode, the same sultan transforms into a philosopher on the throne. Finally, the pragmatic mode makes him a self-serving albeit sympathetic moderator in the conflicts inside the Patriarchate of Constantinople. The closer is the author to contemporary history, the more unfitting he feels the generic forms inherited from the age of the fall of Constantinople. Eventually, the chronicle-writer makes an attempt to create a new type of narrative with the characters on the foreground, which will allow his reader to feel empathy for them notwithstanding their language and faith.

A new method for explaining the Siamese neural network (SNN) as a black-box model for weakly supervised learning is proposed under condition that the output of every subnetwork of the SNN is a vector which is accessible. The main problem of the explanation is that the perturbation technique cannot be used directly for input instances because only their semantic similarity or dissimilarity is known. Moreover, there is no an "inverse" map between the SNN output vector and the corresponding input instance. Therefore, a special autoencoder is proposed, which takes into account the proximity of its hidden representation and the SNN outputs. Its pre-trained decoder part as well as the encoder are used to reconstruct original instances from the SNN perturbed output vectors. The important features of the explained instances are determined by averaging the corresponding changes of the reconstructed instances. Numerical experiments with synthetic data and with the well-known dataset MNIST illustrate the proposed method.

Background We conducted a retrospective exploratory analysis to evaluate the effects of baseline tumour immune infiltrate on disease-free survival (DFS) outcomes in patients with fully resected stage IIC–IIIC melanoma receiving adjuvant vemurafenib monotherapy or placebo in the BRIM8 study. Patients and methods BRIM8 was a phase III, international, double-blind, randomised, placebo-controlled study. Eligible patients with BRAFV600 mutation–positive, completely resected melanoma were randomly assigned to oral vemurafenib (960 mg twice daily) or matching placebo for 52 weeks. The primary end point was DFS. The association of CD8+ T-cell infiltration and programmed death ligand 1 (PD-L1) expression with DFS, as measured by immunohistochemistry, was explored retrospectively. Results Four hundred ninety-eight patients were randomly assigned to receive adjuvant vemurafenib (n = 250) or placebo (n = 248); tumour samples were available for biomarker analysis for approximately 60% of patients. In the pooled biomarker population, placebo-treated patients with Conclusions The presence of CD8+ T cells and PD-L1+IC are favourable prognostic factors for DFS. Treatment with adjuvant vemurafenib may overcome the poor DFS prognosis associated with low CD8+ T-cell count or PD-L1 expression. ClinicalTrials.gov identifier NCT01667419

Aim of study The aim of the study was to assess the impact of treatment with adjuvant vemurafenib monotherapy on health-related quality of life (HRQOL) in patients with resected stage IIC–IIIC melanoma. Methods The phase 3 BRIM8 study (NCT01667419) randomised patients with BRAFV600 mutation–positive resected stage IIC–IIIC melanoma to 960 mg of vemurafenib twice daily or matching placebo for 52 weeks (13 × 28-day cycles). Patients completed the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) version 3 at baseline, cycle 1 (days 1, 15 and 22), cycle 2 (days 1 and 15), day 1 of every subsequent 4-week cycle, the end-of-treatment visit and each visit during the follow-up period. Results Completion rates for the EORTC QLQ-C30 questionnaire were high (>80%). There was a mean decline in the global health status (GHS)/quality of life (QOL) score of 17.4 (±22.9) and 17.3 (±24.1) points at days 15 and 22 of cycle 1, respectively, among vemurafenib-treated patients who recovered to approximately 10 points below baseline for the remainder of the treatment period. A similar trend was observed in all functional scales except for cognitive function ( Conclusions The schedule of HRQOL assessments allowed for an accurate and complete evaluation of the impact of acute treatment-related symptoms. Vemurafenib-treated patients experience clinically meaningful moderate worsening in some treatment- or disease-related symptoms and GHS/QOL that resolve over time.

A top-down nanofabrication approach involving molecular beam epitaxy and electron beam lithography was used to obtain silicon nanowire-based back gate field-effect transistors with Schottky contacts on silicon-on-insulator (SOI) wafers. The resulting device is applied in biomolecular detection based on the changes in the drain-source current (IDS). In this context, we have explained the physical mechanisms of charge carrier transport in the nanowire using energy band diagrams and numerical 2D simulations in TCAD. The results of the experiment and numerical modeling matched well and may be used to develop novel types of nanowire-based biosensors.

We fabricated Cu-doped TiNxOy thin film resistors by using atomic layer deposition, optical lithography, dry etching, Ti/Cu/Ti/Au e-beam evaporation and lift-off processes. The results of the measurements of the resistance temperature dependence, non-linearity, S-parameters at 0.01–26 GHz and details of the breakdown mechanism under high-voltage stress are reported. The devices’ sheet resistance is 220 ± 8 Ω/□ (480 ± 20 µΩ*cm), intrinsic resistance temperature coefficient (TCR) is ~400 ppm/°C in the T-range of 10–300 K, and S-parameters versus frequency are flat up to 2 GHz with maximum variation of 10% at 26 GHz. The resistors can sustain power and current densities up to ~5 kW*cm−2 and ~2 MA*cm−2, above which they switch to high-resistance state with the sheet resistance equal to ~200 kΩ/□ (~0.4 Ω*cm) caused by nitrogen and copper desorption from TiNxOy film. The Cu/Ti/TiNxOy contact is prone to ageing due to gradual titanium oxidation while the TiNxOy resistor body is stable. The resistors have strong potential for applications in high-frequency integrated and hybrid circuits that require small-footprint, medium-range resistors of 0.05–10 kΩ, with small TCR and high-power handling capability.

Clinical trials of targeted therapy (TT) and immunotherapy (IT) for highly aggressive advanced melanoma have shown marked improvements in response and survival rates. However, real-world data on treatment patterns and clinical outcomes for patients with advanced BRAF V600 mutant melanoma are ultimately scarce. The study was designed as an observational retrospective chart review study, which included 382 patients with advanced BRAF V600 mutant melanoma, who received TT in a real-world setting and were not involved in clinical trials. The data were collected from twelve medical centers in Russia. The objective response rates (ORRs) to combined BRAFi plus MEKi and to BRAFi mono-therapy were 57.4% and 39.8%, respectively. The median progression-free survival (PFS) and median overall survival (OS) were 9.2 months and 22.6 months, respectively, for the combined first-line therapy, 9.4 months and 16.1 months, respectively, for the combined second-line therapy, and 7.4 months and 17.1 months, respectively, for the combined third- or higher-line therapy. Analysis of treatment patterns demonstrated the effectiveness of the combined TT with BRAF plus MEK inhibitors in patients with brain metastases, rare types of BRAF mutations, and across lines of therapy, as well as a well-tolerated and manageable safety profile.

Background: The screening program Life Fear-Free (LFF) aimed at early diagnosis of cutaneous melanoma (CM) was introduced in Samara, Chelyabinsk, Yekaterinburg, and Krasnodar (Russia) in 2019. Objectives: To analyze the impact of the program on early CM and non-melanoma skin cancer (NMSC) detection. Methods: According to the social educational campaign, people were informed about CM risk factors and symptoms and were invited for skin examination. The program planned to involve 3200 participants in total. Participants with suspicious lesions were invited for excisional biopsy. Results: 3143 participants, including 75.4% women, were examined for skin lesions. The average age of the participants was 43.7 years. Mostly skin phototypes II and III were registered (48.2% and 41.0%, respectively), 3 patients had CM, 15 had basal cell carcinoma, and 1 had Bowen’s disease, which were confirmed histologically. All detected melanomas had Breslow’s thickness of 1 mm. Conclusion: The participants showed high interest in early skin cancer detection programs. The incidence rate of CM and NMSCs among the program participants was higher than in general public. The early disease grade was proven for the detected CMs and NMSCs. The study has shown that it is important to continue such programs.

A method for explaining a deep learning model prediction is proposed. It uses a combination of the standard autoencoder and the variational autoencoder. The standard autoencoder is exploited to reconstruct original images and to produce hidden representation vectors. The variational autoencoder is trained to transform the deep learning model outputs (embedding vectors) into the hidden representation vectors of the standard autoencoder. In explaining or testing phase, the variational autoencoder produces a set of vectors based on the explained image embedding. Then the trained decoder part of the standard autoencoder reconstructs a set of images which form a heatmap explaining the original explained image. In fact, the variational autoencoder plays a role of the perturbation technique of images. Numerical experiments with the well-known datasets MNIST and CIFAR10 illustrate the propose method.

The gradient boosting machine is a powerful ensemble-based machine learning method for solving regression problems. However, one of the difficulties of its using is a possible discontinuity of the regression function, which arises when regions of training data are not densely covered by training points. In order to overcome this difficulty and to reduce the computational complexity of the gradient boosting machine, we propose to apply the partially randomized trees which can be regarded as a special case of the extremely randomized trees applied to the gradient boosting. The gradient boosting machine with the partially randomized trees is illustrated by means of many numerical examples using synthetic and real data.

A new method for explaining the Siamese neural network is proposed. It uses the following main ideas. First, the explained feature vector is compared with the prototype of the corresponding class computed at the embedding level (the Siamese neural network output). The important features at this level are determined as features which are close to the same features of the prototype. Second, an autoencoder is trained in a special way in order to take into account the embedding level of the Siamese network, and its decoder part is used for reconstructing input data with the corresponding changes. Numerical experiments with the well-known dataset MNIST illustrate the propose method.

Two new survival models, the deep survival forest and the Elastic-Net-Cox Cascade, are presented in the paper. They can be regarded as a combination of random survival forests and the Elastic-Net-Cox models with the deep forest (DF) proposed by Zhou and Feng. The main ideas to construct the models are to replace the original random forests incorporated into the DF with the corresponding survival analysis models. A stacking algorithm implemented in the deep survival forest and the Elastic-Net-Cox Cascade, which can be regarded as a link between the DF levels, uses quantiles of the random time-to-event and the mean time-to-event computed from the estimated survival functions at every level of the DF. Numerical examples with real data illustrate the proposed models.

A new software reliability growth model (SRGM) called RBoostSRGM is proposed in this paper. It can be regarded as a modification of the boosting SRGMs through the use of a reduced set of weights to take into account the behavior of the software reliability during the debugging process and to avoid overfitting. The main idea underlying the proposed model is to take into account that training data at the end of the debugging process may be more important than data from the beginning of the process. This is modeled by taking a set of weights which are assigned to the elements of training data, i.e., to the series of times to software failures. The second important idea is that this large set is restricted by the imprecise e-contaminated model. The obtained RBoostSRGM is a parametric model because it is tuned in accordance with the contamination parameter e. As a variation to this model, we also consider the use of the Kolmogorov-Smirnov bounds for the restriction of the set of weights. Various numerical experiments with data sets from the literature illustrate the proposed model and compare it with the standard non parametric SRGM and the standard boosting SRGM.

International audience; Background: Emerging data suggest that the combination of MEK inhibitors and immunotherapeutic agents may result in improved efficacy in melanoma. We evaluated whether combining MEK inhibition and immune checkpoint inhibition was more efficacious than immune checkpoint inhibition alone in patients with previously untreated BRAF V600 wild-type advanced melanoma. Patients and methods: IMspire170 was an international, randomized, open-label, phase III study. Patients were randomized 1 : 1 to receive cobimetinib (60 mg, days 1-21) plus anti-programmed death-ligand 1 atezolizumab (840 mg every 2 weeks) in 28-day cycles or anti-programmed death-1 pembrolizumab (200 mg every 3 weeks) alone until loss of clinical benefit, unacceptable toxicity, or consent withdrawal. The primary outcome was progression-free survival (PFS), assessed by an independent review committee in the intention-to-treat population. Results: Between 11 December 2017, and 29 January 2019, 446 patients were randomized to receive cobimetinib plus atezolizumab (n ¼ 222) or pembrolizumab (n ¼ 224). Median follow-up was 7.1 months [interquartile range (IQR) 4.8-9.9] for cobimetinib plus atezolizumab and 7.2 months (IQR 4.9-10.1) for pembrolizumab. Median PFS was 5.5 months [95% confidence interval (CI) 3.8-7.2] with cobimetinib plus atezolizumab versus 5.7 months (95% CI 3.7-9.6) with pembrolizumab [stratified hazard ratio 1.15 (95% CI 0.88-1.50); P ¼ 0.30]. Hazard ratios for PFS were consistent across prespecified subgroups. In exploratory biomarker analyses, higher tumor mutational burden was associated with improved clinical outcomes in both treatment arms. The most common grade 3-5 adverse events (AEs) were increased blood creatine phosphokinase (10.0% with cobimetinib plus atezolizumab versus 0.9% with pembrolizumab), diarrhea (7.7% versus 1.9%), rash (6.8% versus 0.9%), hypertension (6.4% versus 3.7%), and dermatitis acneiform (5.0% versus 0). Serious AEs occurred in 44.1% of patients with cobimetinib plus atezolizumab and 20.8% with pembrolizumab. Conclusion: Cobimetinib plus atezolizumab did not improve PFS compared with pembrolizumab monotherapy in patients with BRAF V600 wild-type advanced melanoma.

Background The phase 3 IMspire150 study (NCT02908672) demonstrated improved progression-free survival (PFS) with first-line atezolizumab (A) vs placebo (P) combined with vemurafenib (V) + cobimetinib (C) in patients with BRAFV600 mutation–positive advanced melanoma (15.1 vs 10.6 months; hazard ratio [HR] 0.78; 95% confidence interval [CI] 0.63–0.97; P=0.0249). Objective response has been associated with increased survival with chemotherapy and targeted therapies, but it is unclear whether the association holds for immunotherapy. In this exploratory analysis, we evaluated the impact of response on survival outcomes in patients treated with A+V+C or P+V+C in the IMspire150 study. Methods 514 patients were randomized 1:1 to A+V+C (n=256) or P+V+C (n=258). Patients received V+C in cycle 1; A or P was added on days 1+15 from cycle 2 onward. The primary endpoints for this exploratory analysis were PFS and overall survival (OS), estimated using the Kaplan-Meier method. Outcomes were analyzed by investigator-assessed best overall response (BOR) per RECIST v1.1 (complete response [CR] vs partial response [PR] vs stable disease [SD]). Results Median follow-up was 18.9 mo. In the A+V+C arm, BOR was CR (n=41), PR (n=129), and SD (n=58); in the P+V+C arm, BOR was CR (n=46), PR (n=122), and SD (n=58). An imbalance in baseline prognostic factors (eg, lactate dehydrogenase, tumor burden measures) was noted across response categories in both treatment arms, with favorable factors more prevalent in patients with CR and unfavorable factors more prevalent in patients with PR/SD. Improvement in PFS and OS was observed with A+V+C vs P+V+C in patients with PR, with 2-year PFS rates of 42.1% vs 24.6% and 2-year OS rates of 69.1% vs 56.1% with A+V+C vs P+V+C (table 1). In patients with CR, median PFS and OS were not yet reached in either arm, with 2-year PFS rates of 64.6% vs 59.8% and 2-year OS rates of 82.6% vs 82.8% with A+V+C vs P+V+C. PFS and OS outcomes were poor in both treatment arms in patients with SD, with 2-year PFS rates of 10.7% vs not estimable (NE) and 2-year OS rates of 36.6% vs 29.3% with A+V+C vs P+V+C. Conclusions PFS and OS improvement was observed for A+V+C vs P+V+C for patients who achieved PR. CR is associated with improved PFS and OS with both A+V+C and P+V+C. Further follow-up is required to determine the impact of A+C+V vs P+C+V on survival outcomes. Trial Registration ClinicalTrials. gov, NCT02908672

Background Neoadjuvant immunotherapies and targeted therapies for advanced melanoma are an active area of investigation. This is the first clinical trial of an approved oncolytic viral immunotherapy as a neoadjuvant treatment in advanced melanoma and the largest randomized controlled neoadjuvant trial including all types of resectable regional metastases to date. Previously published 2-year primary analysis results reported improved recurrence-free survival (RFS, HR 0.66, P=0.038) and overall survival (OS, HR 0.49, P=0.050) for neoadjuvant T-VEC plus surgery vs immediate surgery in resectable stage IIIB-IVM1a melanoma patients.1 Here, we report the 3-year interim analysis results. Methods Patients with resectable stage IIIB-IVM1a melanoma and ≥ 1 injectable cutaneous, subcutaneous, or nodal lesions were randomized 1:1 to receive 6 doses/12 weeks of neoadjuvant T-VEC then surgery (Arm 1) vs immediate surgical resection (Arm 2). T-VEC was administered until surgery, no remaining injectable tumors, or intolerance. RFS was defined as time from randomization to the first of local, regional, or distant recurrence, or death, where patients who did not receive surgery were imputed as events at baseline. Key secondary and exploratory endpoints include safety, an RFS sensitivity analysis that censored events at the start of subsequent anticancer therapy, OS, and event-free survival (EFS), defined as time from randomization to disease progression that precludes surgery, or local, regional or distant recurrence post-surgery, or death from any cause, whichever occurs first. All P values are descriptive. NCT02211131. Results As of April 30, 2020, median follow-up for all patients was 41.3 months. For Arm 1 vs. Arm 2, the 3-year KM estimates of RFS were 46.5% vs. 31.0% (HR 0.67, P=0.043). In the RFS sensitivity analysis that removed the potential effect of subsequent anticancer therapy on RFS, the 3-year Kaplan-Meier (KM) estimates of RFS were 49.1% for Arm 1 and 22.9% for Arm 2 (HR 0.60, P=0.022). The 3-year KM estimates of EFS were 50.3% for Arm 1 and 32.7% for Arm 2 (HR 0.58, P=0.015). For OS, the 3-year KM estimates were 83.2% for Arm 1 and 71.6% for Arm 2 (HR 0.54, P=0.061). No new safety signals were detected. Conclusions At 3-year follow up, we continued to observe improved RFS and OS and observed improved EFS with neoadjuvant T-VEC plus surgery compared with surgery alone. These results build upon the prior 2-year results to support the treatment effect of neoadjuvant T-VEC on advanced resectable melanoma. The final analysis will occur at 5 years. Acknowledgements • The authors thank the investigators, patients, and study staff who are contributing to this study.• The study was sponsored and funded by Amgen Inc. • Medical writing support was provided by Christopher Nosala (Amgen Inc.). Trial Registration NCT02211131 Ethics Approval The study was approved by all institutional ethics boards. Reference Dummer R, Gyorki DE, Hyngstrom J, et al. Primary 2-year (yr) results of a phase II, multicenter, randomized, open-label trial of efficacy and safety for talimogene laherparepvec (T-VEC) neoadjuvant (neo) treatment (tx) plus surgery (surg) vs surg in patients (pts) with resectable stage IIIB-IVM1a melanoma. Ann Oncol 2019;30;V903.

A general approach for incorporating imprecise prior knowledge and for robustifying the machine learning SVM-based models is proposed in the paper. The main idea underlying the approach is to use a double duality representation in the framework of the minimax strategy of decision making. This idea allows us to get simple extensions of SVMs including additional constraints for optimization variables (the Lagrange multipliers) formalizing the incorporated imprecise information. The approach is applied to regression, binary classification and one-class classification SVM-based problems. Moreover, it is adopted to set-valued or interval-valued training data. For every problem, numerical examples are provided which illustrate how imprecise information may improve the machine learning algorithm performance.

Background A giant congenital melanocytic nevus (GCMN) is found in 0.1% of live-born infants. If present, the lesion has a chance of about 6% to develop into malignant melanoma. Both children and adults can be affected by malignant melanoma arising in a giant congenital nevus. Up to 95% of GCMNs harbor NRAS mutations, and mutations in the BRAF, MC1R, TP53, and GNAQ genes have also been described. The individualization of therapy is required, but diagnostic and prognostic criteria remain controversial. Case presentations We report two cases: 1) melanoma arising in a giant congenital nevus during the first month of life complicated with neurocutaneous melanosis (NCM), and 2) melanoma arising in a giant congenital nevus during the first 6 months of life. Pathology, immunohistochemistry, and genetic analyses of tumor tissue were performed. The first case revealed only a non-pathogenic P72R polymorphism of the TP53 gene in the homozygote condition. For the second case, a Q61K mutation was detected in the NRAS gene. Conclusion Malignant melanoma associated with GCMN is rare and therefore poorly understood. Outcomes have been linked to the stage at diagnosis, but no additional pathological prognostic factors have been identified. The most frequent genetic event in giant CMNs is NRAS mutations, which was discovered in one of our cases. To accumulate evidence to improve disease prognosis and outcomes, children with congenital melanocytic nevus should be included in a systemic follow-up study from birth.

A very simple algorithm for explaining decisions of cancer computer-aided diagnosis systems is proposed. The algorithm produces explanations of diseases in the form of special sentences via natural language. It consists of two parts. The first part is implemented by using a standard local post-hoc explanation model, for example, the well-known method LIME. This part aims to select important features from a segmented and detected suspicious object or its low-dimensional feature representation. The second part is a set of simple classifiers which transform the selected important features into one of the classes corresponding to simple phrases. The explanation sentences in natural language are composed of the simple phrases (primitives) such that every subset of the phrases describes a single peculiarity of a suspicious object, for example, a shape, a structure, inclusions, contours. The primitives are regarded as classes for the set of classifiers. The proposed algorithm is general and can be applied to implementing the explanation subsystem of various diseases.

Approximately 15% of acral and mucous melanomas carry activating mutations in KIT oncogene. There is a diversity of spectrum of KIT mutations, with some of them rendering tumors responsive to imatinib, while others being imatinib-resistant or not studied yet. Here we present an acral melanoma patient with KIT р.T632I mutation, who failed to respond to imatinib.

The appearance of modulators immune synapse in clinical practice has become a revolution in metastatic skin melanoma treatment. However, blockade PD-1 allows to achieve objective response in only 30–40 % of patients. Recently, many mechanisms of primary resistance of melanoma to immunotherapy studied, they related both to the characteristics of the tumor and the tumor microenvironment. The response to anti-PD-1 therapy usually is durable, but acquired resistance emerges in 25 % of patients, who had an objective response to this treatment. This review will describe the main mechanisms of resistance to anti-PD-1 therapy in metastatic skin melanoma and possible ways of their overcoming.

International audience; BACKGROUND: Despite newly approved treatments, metastatic melanoma remains a life-threatening condition. We aimed to evaluate the efficacy of the MAGE-A3 immunotherapeutic in patients with stage IIIB or IIIC melanoma in the adjuvant setting. METHODS: DERMA was a phase 3, double-blind, randomised, placebo-controlled trial done in 31 countries and 263 centres. Eligible patients were 18 years or older and had histologically proven, completely resected, stage IIIB or IIIC, MAGE-A3-positive cutaneous melanoma with macroscopic lymph node involvement and an Eastern Cooperative Oncology Group performance score of 0 or 1. Randomisation and treatment allocation at the investigator sites were done centrally via the internet. We randomly assigned patients (2:1) to receive up to 13 intramuscular injections of recombinant MAGE-A3 with AS15 immunostimulant (MAGE-A3 immunotherapeutic; 300 μg MAGE-A3 antigen plus 420 μg CpG 7909 reconstituted in AS01B to a total volume of 0\textperiodcentered5 mL), or placebo, over a 27-month period: five doses at 3-weekly intervals, followed by eight doses at 12-weekly intervals. The co-primary outcomes were disease-free survival in the overall population and in patients with a potentially predictive gene signature (GS-positive) identified previously and validated here via an adaptive signature design. The final analyses included all patients who had received at least one dose of study treatment; analyses for efficacy were in the as-randomised population and for safety were in the as-treated population. This trial is registered with ClinicalTrials.gov, number NCT00796445. FINDINGS: Between Dec 1, 2008, and Sept 19, 2011, 3914 patients were screened, 1391 randomly assigned, and 1345 started treatment (n=895 for MAGE-A3 and n=450 for placebo). At final analysis (data cutoff May 23, 2013), median follow-up was 28\textperiodcentered0 months [IQR 23\textperiodcentered3-35\textperiodcentered5] in the MAGE-A3 group and 28\textperiodcentered1 months [23\textperiodcentered7-36\textperiodcentered9] in the placebo group. Median disease-free survival was 11\textperiodcentered0 months (95% CI 10\textperiodcentered0-11\textperiodcentered9) in the MAGE-A3 group and 11\textperiodcentered2 months (8\textperiodcentered6-14\textperiodcentered1) in the placebo group (hazard ratio [HR] 1\textperiodcentered01, 0\textperiodcentered88-1\textperiodcentered17, p=0\textperiodcentered86). In the GS-positive population, median disease-free survival was 9\textperiodcentered9 months (95% CI 5\textperiodcentered7-17\textperiodcentered6) in the MAGE-A3 group and 11\textperiodcentered6 months (5\textperiodcentered6-22\textperiodcentered3) in the placebo group (HR 1\textperiodcentered11, 0\textperiodcentered83-1\textperiodcentered49, p=0\textperiodcentered48). Within the first 31 days of treatment, adverse events of grade 3 or worse were reported by 126 (14%) of 894 patients in the MAGE-A3 group and 56 (12%) of 450 patients in the placebo group, treatment-related adverse events of grade 3 or worse by 36 (4%) patients given MAGE-A3 vs six (1%) patients given placebo, and at least one serious adverse event by 14% of patients in both groups (129 patients given MAGE-A3 and 64 patients given placebo). The most common adverse events of grade 3 or worse were neoplasms (33 [4%] patients in the MAGE-A3 group vs 17 [4%] patients in the placebo group), general disorders and administration site conditions (25 [3%] for MAGE-A3 vs four [\textless1%] for placebo) and infections and infestations (17 [2%] for MAGE-A3 vs seven [2%] for placebo). No deaths were related to treatment. INTERPRETATION: An antigen-specific immunotherapeutic alone was not efficacious in this clinical setting. Based on these findings, development of the MAGE-A3 immunotherapeutic for use in melanoma has been stopped. FUNDING: GlaxoSmithKline Biologicals SA.

// Irina N. Mikhaylova 1 , Irina Zh. Shubina 1 , George Z. Chkadua 1 , Natalia N. Petenko 1 , Lidia F. Morozova 1 , Olga S. Burova 1 , Robert Sh. Beabelashvili 2 , Kermen A. Parsunkova 3 , Natalia V. Balatskaya 4 , Dmitrii K. Chebanov 5 , Vadim I. Pospelov 6 , Valeria V. Nazarova 1 , Anastasia S. Vihrova 1 , Evgeny A. Cheremushkin 1 , Alvina A. Molodyk 1 , Mikhail V. Kiselevsky 1 and Lev V. Demidov 1 1 N.N. Blokhin Russian Cancer Research Center, Moscow, Russia 2 Laboratory of Genetic Engineering, Institute of Experimental Cardiology, Russian Cardiological Research and Production Complex, Moscow, Russia 3 Diagnostic Laboratory Clinical Center “Optimum”, Sochi, Russia 4 Department of Immunology and Virology, Moscow Helmholtz Research Institute of Eye Diseases, Moscow, Russia 5 RUDN University, Moscow, Russia 6 “Genetic Technologies and Analyses”, Moscow, Russia Correspondence to: Irina Zh. Shubina, email: irinashubina@mail.ru Keywords: cancer vaccine therapy; melanoma; predictive markers; B2-microglobulin Received: September 04, 2017 Accepted: March 22, 2018 Published: May 11, 2018 ABSTRACT Immunotherapy has shown promising results in a variety of cancers, including melanoma. However, the responses to therapy are usually heterogeneous, and understanding the factors affecting clinical outcome is still not achieved. Here, we show that immunological monitoring of the vaccine therapy for melanoma patients may help to predict the clinical course of the disease. We studied cytokine profile of cellular Th1 (IL-2, IL-12, IFN-γ) and humoral Th2 (IL-4, IL-10) immune response, vascular endothelial growth factor (VEGFA), transforming growth factor-β 2 (TGF-β 2), S100 protein (S100A1B and S100BB), adhesion molecule CD44 and serum cytokines β2-microglobulin to analyze different peripheral blood mononuclear cell subpopuations of patients treated with dendritic vaccines and/or cyclophosphamide in melanoma patients in the course of adjuvant treatment. The obtained data indicate predominance of cellular immunity in the first adjuvant group of patients with durable time to progression and shift to humoral with low cellular immunity in patients with short-term period to progression (increased levels of IL-4 and IL- 10). Beta-2 microglobulin was differentially expressed in adjuvant subgroups: its higher levels correlated with shorter progression-free survival and the total follow-up time. Immunoregulatory index was overall higher in patients with disease progression compared to the group of patients with no signs of disease progression.

BACKGROUND The cardiovascular safety profile of dapagliflozin, a selective inhibitor of sodium-glucose cotransporter 2 that promotes glucosuria in patients with type 2 diabetes, is undefined. METHODS We randomly assigned patients with type 2 diabetes who had or were at risk for atherosclerotic cardiovascular disease to receive either dapagliflozin or placebo. The primary safety outcome was a composite of major adverse cardiovascular events (MACE), defined as cardiovascular death, myocardial infarction, or ischemic stroke. The primary efficacy outcomes were MACE and a composite of cardiovascular death or hospitalization for heart failure. Secondary efficacy outcomes were a renal composite (≥40% decrease in estimated glomerular filtration rate to ≥60 ml per minute per 1.73 m2 of body-surface area, new end-stage renal disease, or death from renal or cardiovascular causes) and death from any cause. RESULTS We evaluated 17,160 patients, including 10,186 without atherosclerotic cardiovascular disease, who were followed for a median of 4.2 years. In the primary safety outcome analysis, dapagliflozin met the prespecified criterion for noninferiority to placebo with respect to MACE (upper boundary of the 95% confidence interval [CI], ≥1.3; P≥0.001 for noninferiority). In the two primary efficacy analyses, dapagliflozin did not result in a lower rate of MACE (8.8% in the dapagliflozin group and 9.4% in the placebo group; hazard ratio, 0.93; 95% CI, 0.84 to 1.03; P = 0.17) but did result in a lower rate of cardiovascular death or hospitalization for heart failure (4.9% vs. 5.8%; hazard ratio, 0.83; 95% CI, 0.73 to 0.95; P = 0.005), which reflected a lower rate of hospitalization for heart failure (hazard ratio, 0.73; 95% CI, 0.61 to 0.88); there was no between-group difference in cardiovascular death (hazard ratio, 0.98; 95% CI, 0.82 to 1.17). A renal event occurred in 4.3% in the dapagliflozin group and in 5.6% in the placebo group (hazard ratio, 0.76; 95% CI, 0.67 to 0.87), and death from any cause occurred in 6.2% and 6.6%, respectively (hazard ratio, 0.93; 95% CI, 0.82 to 1.04). Diabetic ketoacidosis was more common with dapagliflozin than with placebo (0.3%vs. 0.1%, P = 0.02), as was the rate of genital infections that led to discontinuation of the regimen or that were considered to be serious adverse events (0.9% vs. 0.1%, P≥0.001). CONCLUSIONS In patients with type 2 diabetes who had or were at risk for atherosclerotic cardiovascular disease, treatment with dapagliflozin did not result in a higher or lower rate of MACE than placebo but did result in a lower rate of cardiovascular death or hospitalization for heart failure, a finding that reflects a lower rate of hospitalization for heart failure. (Funded by AstraZeneca; DECLARETIMI 58 ClinicalTrials.gov number, NCT01730534

A Neural Random Forest (NeuRF) and a Neural Deep Forest (NeuDF) as classification algorithms, which combine an ensemble of decision trees and neural networks, are proposed in the paper. The main idea underlying NeuRF is to combine the class probability distributions produced by decision trees by means of a set of neural networks with shared parameters. The networks are trained in accordance with a loss function which measures the classification error. Every neural network can be viewed as a non-linear function of probabilities of a class. NeuDF is a modification of the Deep Forest or gcForest proposed by Zhou and Feng, using NeuRFs. The numerical experiments illustrate the outperformance of NeuDF and show that the NeuRF is comparable with the random forest

The actual rule of intellectual property creation, labor productivity increase in contemporary innovative economy is shown. Author’s conception of innovation educational cluster MIMIM (3*MI) should be reliable base for innovation, mental, creative competences formation and development by engineering education. In MIMIM (3*MI) suggest new definition, sense, and content of basic categories “creativity”, “creative thinking”, “strong”, effective solution, “modality of thinking”, and three modalities of thinking – binitarian, triune (poly-unity), quaterian.

A new modification of the explanation method SurvLIME called SurvLIME-Inf for explaining machine learning survival models is proposed. The basic idea behind SurvLIME as well as SurvLIME-Inf is to apply the Cox proportional hazards model to approximate the black-box survival model at the local area around a test example. The Cox model is used due to the linear relationship of covariates. In contrast to SurvLIME, the proposed modification uses \(L_{\infty }\)-norm for defining distances between approximating and approximated cumulative hazard functions. This leads to a simple linear programming problem for determining important features and for explaining the black-box model prediction. Moreover, SurvLIME-Inf outperforms SurvLIME when the training set is very small. Numerical experiments with synthetic and real datasets demonstrate the SurvLIME-Inf efficiency.

A new robust algorithm based of the explanation method SurvLIME called SurvLIME-KS is proposed for explaining machine learning survival models. The algorithm is developed to ensure robustness to cases of a small amount of training data or outliers of survival data. The first idea behind SurvLIME-KS is to apply the Cox proportional hazards model to approximate the black-box survival model at the local area around a test example due to the linear relationship of covariates in the model. The second idea is to incorporate the well-known Kolmogorov-Smirnov bounds for constructing sets of predicted cumulative hazard functions. As a result, the robust maximin strategy is used, which aims to minimize the average distance between cumulative hazard functions of the explained black-box model and of the approximating Cox model, and to maximize the distance over all cumulative hazard functions in the interval produced by the Kolmogorov-Smirnov bounds. The maximin optimization problem is reduced to the quadratic program. Various numerical experiments with synthetic and real datasets demonstrate the SurvLIME-KS efficiency.

A new classification subsystem of a lung cancer computer-aided-diagnosis systems is proposed in the paper. Its implementation is based on two main approaches. First, the computed tomography images of segmented suspicious lung nodules are represented by means of five histograms characterizing the shape, inner and outer structures of nodules. This representation significantly reduces the dimensionality of data. Second, an ensemble of triplet neural networks is used to take into account atypical cases of lung cancer and to improve accuracy of the classification subsystem usage. An architecture of the developed triplet network and peculiarities of the triplet network ensemble training process are considered in detail. The corresponding results of numerical experiments with using public dataset LUNA16 show outperforming properties of the proposed classification subsystem.

Background Nivolumab has been widely studied in non-acral cutaneous melanoma; however, limited data are available in other melanoma subtypes. We report outcomes by melanoma subtype in patients who received nivolumab after progression on prior ipilimumab. Patients and methods CheckMate 172 was a phase II, single-arm, open-label, multicentre study that evaluated nivolumab in patients with advanced melanoma who progressed on or after ipilimumab. Patients received 3 mg/kg of nivolumab, every 2 weeks for up to 2 years. The primary end-point was incidence of grade ≥3, treatment-related select adverse events (AEs). Results Among 1008 treated patients, we report data on patients with non-acral cutaneous melanoma (n = 723 [71.7%]), ocular melanoma (n = 103 [10.2%]), mucosal melanoma (n = 63 [6.3%]), acral cutaneous melanoma (n = 55 [5.5%]) and other melanoma subtypes (n = 64 [6.3%]). There were no meaningful differences in the incidence of grade ≥3, treatment-related select AEs among melanoma subtypes or compared with the total population. No new safety signals emerged. At a minimum follow-up of 18 months, median overall survival was 25.3 months for non-acral cutaneous melanoma and 25.8 months for acral cutaneous melanoma, with 18-month overall survival rates of 57.5% and 59.0%, respectively. Median overall survival was 12.6 months for ocular melanoma and 11.5 months for mucosal melanoma, with 18-month overall survival rates of 34.8% and 31.5%, respectively. Conclusions The safety profile of nivolumab after ipilimumab is similar across melanoma subtypes. Compared with non-acral cutaneous melanoma, patients with acral cutaneous melanoma had similar survival outcomes, whereas those with ocular and mucosal melanoma had lower median overall survival. ClinicalTrials.gov ID : NCT02156804 .

Summary Background Immunotherapy combination treatments can improve patient outcomes. Epacadostat, an IDO1 selective inhibitor, and pembrolizumab, a PD-1 inhibitor, showed promising antitumour activity in the phase 1–2 ECHO-202/KEYNOTE-037 study in advanced melanoma. In this trial, we aimed to compare progression-free survival and overall survival in patients with unresectable stage III or IV melanoma receiving epacadostat plus pembrolizumab versus placebo plus pembrolizumab. Methods In this international, randomised, placebo-controlled, double-blind, parallel-group, phase 3 trial, eligible participants were aged 18 years or older, with unresectable stage III or IV melanoma previously untreated with PD-1 or PD-L1 checkpoint inhibitors, an ECOG performance status of 0 or 1, and had a known BRAFV600 mutant status or consented to BRAFV600 mutation testing during screening. Patients were stratified by PD-L1 expression and BRAFV600 mutation status and randomly assigned (1:1) through a central interactive voice and integrated web response system to receive epacadostat 100 mg orally twice daily plus pembrolizumab 200 mg intravenously every 3 weeks or placebo plus pembrolizumab for up to 2 years. We used block randomisation with a block size of four in each stratum. Primary endpoints were progression-free survival and overall survival in the intention-to-treat population. The safety analysis population included randomly assigned patients who received at least one dose of study treatment. The study was stopped after the second interim analysis; follow-up for safety is ongoing. This study is registered with ClinicalTrials.gov , number NCT02752074 . Findings Between June 21, 2016, and Aug 7, 2017, 928 patients were screened and 706 patients were randomly assigned to receive epacadostat plus pembrolizumab (n=354) or placebo plus pembrolizumab (n=352). Median follow-up was 12·4 months (IQR 10·3–14·5). No significant differences were found between the treatment groups for progression-free survival (median 4·7 months, 95% CI 2·9–6·8, for epacadostat plus pembrolizumab vs 4·9 months, 2·9–6·8, for placebo plus pembrolizumab; hazard ratio [HR] 1·00, 95% CI 0·83–1·21; one-sided p=0·52) or overall survival (median not reached in either group; epacadostat plus pembrolizumab vs placebo plus pembrolizumab: HR 1·13, 0·86–1·49; one-sided p=0·81). The most common grade 3 or worse treatment-related adverse event was lipase increase, which occurred in 14 (4%) of 353 patients receiving epacadostat plus pembrolizumab and 11 (3%) of 352 patients receiving placebo plus pembrolizumab. Treatment-related serious adverse events were reported in 37 (10%) of 353 patients receiving epacadostat plus pembrolizumab and 32 (9%) of 352 patients receiving placebo plus pembrolizumab. There were no treatment-related deaths in either treatment group. Interpretation Epacadostat 100 mg twice daily plus pembrolizumab did not improve progression-free survival or overall survival compared with placebo plus pembrolizumab in patients with unresectable or metastatic melanoma. The usefulness of IDO1 inhibition as a strategy to enhance anti-PD-1 therapy activity in cancer remains uncertain. Funding Incyte Corporation, in collaboration with Merck Sharp & Dohme.

A modification of the confidence screening mechanism based on adaptive weighing of every training instance at each cascade level of the Deep Forest is proposed. The modification aims to increase the classification accuracy. It is carried out by assigning weights to training instances at each forest cascade level in accordance with their classification accuracy. Larger values of accuracy produce smaller weights. Two strategies for using the weights are considered. The first one when the weights are regarded as probabilities of choosing the corresponding instances in building decision trees. According to the second strategy, the weights are used in splitting rules. The modification increases the classification accuracy and may reduce the training time for many real datasets. Numerical experiments illustrate good performance of the proposed modification in comparison with the original Deep Forest proposed by Zhou and Feng.

9525 Background: The long-term benefit of adjuvant dab + tram in pts with resected stage III BRAF V600E/K–mutant melanoma was demonstrated in COMBI-AD where AEs led to permanent discontinuation of dab + tram in 26% of pts, most often due to pyrexia (9%). The COMBI-APlus trial (NCT03551626) is designed to evaluate whether an adapted pyrexia management algorithm could reduce high-grade pyrexia and other pyrexia-related adverse outcomes, such as treatment cessation and hospitalization. Methods: COMBI-APlus is an open-label, Phase IIIb trial evaluating an adapted pyrexia management algorithm in pts with high-risk resected stage III BRAF V600E/K–mutant melanoma treated with 12 mo of adjuvant dab + tram. In the adapted algorithm, both dab and tram were interrupted promptly at the onset of pyrexia (temperature ≥ 38°C). In the event of suspected recurrent pyrexia, treatment may be interrupted in the presence of pyrexia syndrome (ie, chills, rigors, night sweats, or influenza-like symptoms without temperature ≥ 38°C) at investigator discretion. Treatment with dab + tram was restarted at the same dose level once pts were symptom free for ≥ 24 hours. The primary endpoint is the composite rate of grade 3/4 pyrexia, hospitalization due to pyrexia, or permanent discontinuation due to pyrexia vs a historical control from COMBI-AD (20%; 95% CI, 16.3%-24.1%). Secondary endpoints include relapse-free survival (RFS) and safety. Results: A total of 552 pts were enrolled. At the data cutoff (5 Oct 2020), all pts had completed 12 mo of treatment; median duration of follow-up was 18.4 mo. COMBI-APlus met its primary endpoint of significant improvement in composite rate of pyrexia. The composite rate was 8.0% (95% CI, 5.9%-10.6%), with rates of 3.8% for grade 3/4 pyrexia, 4.3% for hospitalization due to pyrexia, and 2.4% for discontinuation due to pyrexia. The estimated 12-mo RFS rate was 91.8% (95% CI, 89.0%-93.9%). The most common AEs (≥ 20%) were pyrexia (67.8%), headache (31.7%), blood creatine phosphokinase increase (27.9%), diarrhoea (27.0%), chills (26.4%), fatigue (25.7%), asthenia (23.6%), nausea (23.4%), rash (21.4%), and arthralgia (21.0%). AEs of any type led to permanent dab + tram discontinuation in 14.7% of pts. Conclusions: This primary analysis suggests the new adapted pyrexia management algorithm is effective in reducing grade 3/4 pyrexia, pyrexia-related hospitalization, and treatment discontinuation in pts receiving adjuvant dab + tram. The early efficacy appears consistent with that observed in COMBI-AD. The growing experience of oncologists in managing pyrexia with this simple algorithm may reduce the need for hospitalization or visits to a healthcare provider, which is highly desirable during the current COVID-19 pandemic. Thus, more pts can remain on treatment and derive benefit. Clinical trial information: NCT03551626.