In a study involving patients with non–small-cell lung cancer with a
MET
exon 14 skipping mutation, the use of tepotinib (a selective MET inhibitor) was associated with a partial response in ...approximately half the patients. The main adverse event was peripheral edema.
Abstract
Computational models for drug sensitivity prediction have the potential to significantly improve personalized cancer medicine. Drug sensitivity assays, combined with profiling of cancer cell ...lines and drugs become increasingly available for training such models. Multiple methods were proposed for predicting drug sensitivity from cancer cell line features, some in a multi-task fashion. So far, no such model leveraged drug inhibition profiles. Importantly, multi-task models require a tailored approach to model interpretability. In this work, we develop DEERS, a neural network recommender system for kinase inhibitor sensitivity prediction. The model utilizes molecular features of the cancer cell lines and kinase inhibition profiles of the drugs. DEERS incorporates two autoencoders to project cell line and drug features into 10-dimensional hidden representations and a feed-forward neural network to combine them into response prediction. We propose a novel interpretability approach, which in addition to the set of modeled features considers also the genes and processes outside of this set. Our approach outperforms simpler matrix factorization models, achieving R
$$=$$
=
0.82 correlation between true and predicted response for the unseen cell lines. The interpretability analysis identifies 67 biological processes that drive the cell line sensitivity to particular compounds. Detailed case studies are shown for PHA-793887, XMD14-99 and Dabrafenib.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
Drug sensitivity prediction constitutes one of the main challenges in personalized medicine. Critically, the sensitivity of cancer cells to treatment depends on an unknown subset of a large number of ...biological features. Here, we compare standard, data-driven feature selection approaches to feature selection driven by prior knowledge of drug targets, target pathways, and gene expression signatures. We asses these methodologies on Genomics of Drug Sensitivity in Cancer (GDSC) dataset, evaluating 2484 unique models. For 23 drugs, better predictive performance is achieved when the features are selected according to prior knowledge of drug targets and pathways. The best correlation of observed and predicted response using the test set is achieved for Linifanib (r = 0.75). Extending the drug-dependent features with gene expression signatures yields the most predictive models for 60 drugs, with the best performing example of Dabrafenib. For many compounds, even a very small subset of drug-related features is highly predictive of drug sensitivity. Small feature sets selected using prior knowledge are more predictive for drugs targeting specific genes and pathways, while models with wider feature sets perform better for drugs affecting general cellular mechanisms. Appropriate feature selection strategies facilitate the development of interpretable models that are indicative for therapy design.
Full text
Available for:
IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
Copy number alterations constitute important phenomena in tumor evolution. Whole genome single-cell sequencing gives insight into copy number profiles of individual cells, but is highly noisy. Here, ...we propose CONET, a probabilistic model for joint inference of the evolutionary tree on copy number events and copy number calling. CONET employs an efficient, regularized MCMC procedure to search the space of possible model structures and parameters. We introduce a range of model priors and penalties for efficient regularization. CONET reveals copy number evolution in two breast cancer samples, and outperforms other methods in tree reconstruction, breakpoint identification and copy number calling.
The microRNA (miRNA) landscape changes during the progression of cancer. We defined a metastasis-associated miRNA landscape using a systematic approach. We profiled and validated miRNA and mRNA ...expression in a unique series of human colorectal metastasis tissues together with their matched primary tumors and corresponding normal tissues. We identified an exclusive miRNA signature that is differentially expressed in metastases. Three of these miRNAs were identified as key drivers of an EMT-regulating network acting though a number of novel targets. These targets include SIAH1, SETD2, ZEB2, and especially FOXN3, which we demonstrated for the first time as a direct transcriptional suppressor of N-cadherin. The modulation of N-cadherin expression had significant impact on migration, invasion, and metastasis in two different in vivo models. The significant deregulation of the miRNAs defining the network was confirmed in an independent patient set as well as in a database of diverse malignancies derived from more than 6,000 patients. Our data define a novel metastasis-orchestrating network based on systematic hypothesis generation from metastasis tissues.
To optimize neuroblastoma treatment stratification, we aimed at developing a novel risk estimation system by integrating gene expression-based classification and established prognostic markers.
Gene ...expression profiles were generated from 709 neuroblastoma specimens using customized 4 × 44 K microarrays. Classification models were built using 75 tumors with contrasting courses of disease. Validation was performed in an independent test set (n = 634) by Kaplan-Meier estimates and Cox regression analyses.
The best-performing classifier predicted patient outcome with an accuracy of 0.95 (sensitivity, 0.93; specificity, 0.97) in the validation cohort. The highest potential clinical value of this predictor was observed for current low-risk patients 5-year event-free survival (EFS), 0.84 ± 0.02 vs. 0.29 ± 0.10; 5-year overall survival (OS), 0.99 ± 0.01 vs. 0.76 ± 0.11; both P < 0.001 and intermediate-risk patients (5-year EFS, 0.88 ± 0.06 vs. 0.41 ± 0.10; 5-year OS, 1.0 vs. 0.70 ± 0.09; both P < 0.001). In multivariate Cox regression models for low-risk/intermediate-risk patients, the classifier outperformed risk assessment of the current German trial NB2004 EFS: hazard ratio (HR), 5.07; 95% confidence interval (CI), 3.20-8.02; OS: HR, 25.54; 95% CI, 8.40-77.66; both P < 0.001. On the basis of these findings, we propose to integrate the classifier into a revised risk stratification system for low-risk/intermediate-risk patients. According to this system, we identified novel subgroups with poor outcome (5-year EFS, 0.19 ± 0.08; 5-year OS, 0.59 ± 0.1), for whom we propose intensified treatment, and with beneficial outcome (5-year EFS, 0.87 ± 0.05; 5-year OS, 1.0), who may benefit from treatment de-escalation.
Combination of gene expression-based classification and established prognostic markers improves risk estimation of patients with low-risk/intermediate-risk neuroblastoma. We propose to implement our revised treatment stratification system in a prospective clinical trial.
BACKGROUND: Current risk stratification systems for neuroblastoma patients consider clinical, histopathological, and genetic variables, and additional prognostic markers have been proposed in recent ...years. We here sought to select highly informative covariates in a multistep strategy based on consecutive Cox regression models, resulting in a risk score that integrates hazard ratios of prognostic variables. METHODS: A cohort of 695 neuroblastoma patients was divided into a discovery set (n=75) for multigene predictor generation, a training set (n=411) for risk score development, and a validation set (n=209). Relevant prognostic variables were identified by stepwise multivariable L1-penalized least absolute shrinkage and selection operator (LASSO) Cox regression, followed by backward selection in multivariable Cox regression, and then integrated into a novel risk score. RESULTS: The variables stage, age, MYCN status, and two multigene predictors, NB-th24 and NB-th44, were selected as independent prognostic markers by LASSO Cox regression analysis. Following backward selection, only the multigene predictors were retained in the final model. Integration of these classifiers in a risk scoring system distinguished three patient subgroups that differed substantially in their outcome. The scoring system discriminated patients with diverging outcome in the validation cohort (5-year event-free survival, 84.9±3.4 vs 63.6±14.5 vs 31.0±5.4; P<.001), and its prognostic value was validated by multivariable analysis. CONCLUSION: We here propose a translational strategy for developing risk assessment systems based on hazard ratios of relevant prognostic variables. Our final neuroblastoma risk score comprised two multigene predictors only, supporting the notion that molecular properties of the tumor cells strongly impact clinical courses of neuroblastoma patients.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
To evaluate the impact of a predefined gene expression-based classifier for clinical risk estimation and cytotoxic treatment decision making in neuroblastoma patients.
Gene expression profiles of 440 ...internationally collected neuroblastoma specimens were investigated by microarray analysis, 125 of which were examined prospectively. Patients were classified as either favorable or unfavorable by a 144-gene prediction analysis for microarrays (PAM) classifier established previously on a separate set of 77 patients. PAM classification results were compared with those of current prognostic markers and risk estimation strategies.
The PAM classifier reliably distinguished patients with contrasting clinical courses (favorable n = 249 and unfavorable n = 191; 5-year event free survival EFS 0.84 +/- 0.03 v 0.38 +/- 0.04; 5-year overall survival OS 0.98 +/- 0.01 v 0.56 +/- 0.05, respectively; both P < .001). Moreover, patients with divergent outcome were robustly discriminated in both German and international cohorts and in prospectively analyzed samples (P <or= .001 for both EFS and OS for each). In subgroups with clinical low-, intermediate-, and high-risk of death from disease, the PAM predictor significantly separated patients with divergent outcome (low-risk 5-year OS: 1.0 v 0.75 +/- 0.10, P < .001; intermediate-risk: 1.0 v 0.82 +/- 0.08, P = .042; and high-risk: 0.81 +/- 0.08 v 0.43 +/- 0.05, P = .001). In multivariate Cox regression models based on both EFS and OS, PAM was a significant independent prognostic marker (EFS: hazard ratio HR, 3.375; 95% CI, 2.075 to 5.492; P < .001; OS: HR, 11.119, 95% CI, 2.487 to 49.701; P < .001). The highest potential clinical impact of the classifier was observed in patients currently considered as non-high-risk (n = 289; 5-year EFS: 0.87 +/- 0.02 v 0.44 +/- 0.07; 5-year OS: 1.0 v 0.80 +/- 0.06; both P < .001).
Gene expression-based classification using the 144-gene PAM predictor can contribute to improved treatment stratification of neuroblastoma patients.
The present study investigated the genetic contribution to alcohol dependence (AD) using genome-wide association data from three German samples. These comprised patients with: (i) AD; (ii) chronic ...alcoholic pancreatitis (ACP); and (iii) alcohol-related liver cirrhosis (ALC). Single marker, gene-based, and pathway analyses were conducted. A significant association was detected for the
locus in a gene-based approach (
= 1.2 × 10
;
= 0.020). This was driven by the AD subsample. No association with
was found in the combined ACP + ALC sample. On first inspection, this seems surprising, since
is a robustly replicated risk gene for AD and may therefore be expected to be associated also with subgroups of AD patients. The negative finding in the ACP + ALC sample, however, may reflect genetic stratification as well as random fluctuation of allele frequencies in the cases and controls, demonstrating the importance of large samples in which the phenotype is well assessed.
Cell competition is an emerging principle underlying selection for cellular fitness during development and disease. Competition may be relevant for cancer, but an experimental link between defects in ...competition and tumorigenesis is elusive. In the thymus, T lymphocytes develop from precursors that are constantly replaced by bone-marrow-derived progenitors. Here we show that in mice this turnover is regulated by natural cell competition between 'young' bone-marrow-derived and 'old' thymus-resident progenitors that, although genetically identical, execute differential gene expression programs. Disruption of cell competition leads to progenitor self-renewal, upregulation of Hmga1, transformation, and T-cell acute lymphoblastic leukaemia (T-ALL) resembling the human disease in pathology, genomic lesions, leukaemia-associated transcripts, and activating mutations in Notch1. Hence, cell competition is a tumour suppressor mechanism in the thymus. Failure to select fit progenitors through cell competition may explain leukaemia in X-linked severe combined immune deficiency patients who showed thymus-autonomous T-cell development after therapy with gene-corrected autologous progenitors.
Full text
Available for:
DOBA, IJS, IZUM, KILJ, KISLJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK