Integrating imaging and genomic data for the discovery of distinct glioblastoma subtypes: a joint learning approach

Scientific Reports · 2024 · 23 citations

• Computer Science
• Artificial Intelligence
• Computational biology

Glioblastoma is a highly heterogeneous disease, with variations observed at both phenotypical and molecular levels. Personalized therapies would be facilitated by non-invasive in vivo approaches for characterizing this heterogeneity. In this study, we developed unsupervised joint machine learning between radiomic and genomic data, thereby identifying distinct glioblastoma subtypes. A retrospective cohort of 571 IDH-wildtype glioblastoma patients were included in the study, and pre-operative multi-parametric MRI scans and targeted next-generation sequencing (NGS) data were collected. L21-norm minimization was used to select a subset of 12 radiomic features from the MRI scans, and 13 key driver genes from the five main signal pathways most affected in glioblastoma were selected from the genomic data. Subtypes were identified using a joint learning approach called Anchor-based Partial Multi-modal Clustering on both radiomic and genomic modalities. Kaplan-Meier analysis identified three distinct glioblastoma subtypes: high-risk, medium-risk, and low-risk, based on overall survival outcome (p < 0.05, log-rank test; Hazard Ratio = 1.64, 95% CI 1.17-2.31, Cox proportional hazard model on high-risk and low-risk subtypes). The three subtypes displayed different phenotypical and molecular characteristics in terms of imaging histogram, co-occurrence of genes, and correlation between the two modalities. Our findings demonstrate the synergistic value of integrated radiomic signatures and molecular characteristics for glioblastoma subtyping. Joint learning on both modalities can aid in better understanding the molecular basis of phenotypical signatures of glioblastoma, and provide insights into the biological underpinnings of tumor formation and progression.

Impact of primary kidney disease on the effects of empagliflozin in patients with chronic kidney disease: secondary analyses of the EMPA-KIDNEY trial

The Lancet Diabetes & Endocrinology · 2023 · 102 citations

• Medicine
• Internal medicine
• Intensive care medicine

BACKGROUND: The EMPA-KIDNEY trial showed that empagliflozin reduced the risk of the primary composite outcome of kidney disease progression or cardiovascular death in patients with chronic kidney disease mainly through slowing progression. We aimed to assess how effects of empagliflozin might differ by primary kidney disease across its broad population. METHODS: , or death from kidney failure) were assessed using prespecified Cox models, and eGFR slope analyses used shared parameter models. Subgroup comparisons were performed by including relevant interaction terms in models. EMPA-KIDNEY is registered with ClinicalTrials.gov, NCT03594110. FINDINGS: per year (95% CI 1·16-1·59), representing a 50% (42-58) reduction in the rate of chronic eGFR decline. This relative effect of empagliflozin on chronic eGFR slope was similar in analyses by different primary kidney diseases, including in explorations by type of glomerular disease and diabetes (p values for heterogeneity all >0·1). INTERPRETATION: In a broad range of patients with chronic kidney disease at risk of progression, including a wide range of non-diabetic causes of chronic kidney disease, empagliflozin reduced risk of kidney disease progression. Relative effect sizes were broadly similar irrespective of the cause of primary kidney disease, suggesting that SGLT2 inhibitors should be part of a standard of care to minimise risk of kidney failure in chronic kidney disease. FUNDING: Boehringer Ingelheim, Eli Lilly, and UK Medical Research Council.

Clinical measures, radiomics, and genomics offer synergistic value in AI-based prediction of overall survival in patients with glioblastoma

Scientific Reports · 2022 · 86 citations

• Computer Science
• Artificial Intelligence
• Machine Learning

Multi-omic data, i.e., clinical measures, radiomic, and genetic data, capture multi-faceted tumor characteristics, contributing to a comprehensive patient risk assessment. Here, we investigate the additive value and independent reproducibility of integrated diagnostics in prediction of overall survival (OS) in isocitrate dehydrogenase (IDH)-wildtype GBM patients, by combining conventional and deep learning methods. Conventional radiomics and deep learning features were extracted from pre-operative multi-parametric MRI of 516 GBM patients. Support vector machine (SVM) classifiers were trained on the radiomic features in the discovery cohort (n = 404) to categorize patient groups of high-risk (OS < 6 months) vs all, and low-risk (OS ≥ 18 months) vs all. The trained radiomic model was independently tested in the replication cohort (n = 112) and a patient-wise survival prediction index was produced. Multivariate Cox-PH models were generated for the replication cohort, first based on clinical measures solely, and then by layering on radiomics and molecular information. Evaluation of the high-risk and low-risk classifiers in the discovery/replication cohorts revealed area under the ROC curves (AUCs) of 0.78 (95% CI 0.70-0.85)/0.75 (95% CI 0.64-0.79) and 0.75 (95% CI 0.65-0.84)/0.63 (95% CI 0.52-0.71), respectively. Cox-PH modeling showed a concordance index of 0.65 (95% CI 0.6-0.7) for clinical data improving to 0.75 (95% CI 0.72-0.79) for the combination of all omics. This study signifies the value of integrated diagnostics for improved prediction of OS in GBM.

A Probability-Based Investigation on the Setup Robustness of Pencil-beam Proton Radiation Therapy for Skull-Base Meningioma

International Journal of Particle Therapy · 2021 · 1 citations

• Computer Science
• Nuclear medicine
• Computer Science

INTRODUCTION: The intracranial skull-base meningioma is in proximity to multiple critical organs and heterogeneous tissues. Steep dose gradients often result from avoiding critical organs in proton treatment plans. Dose uncertainties arising from setup errors under image-guided radiation therapy are worthy of evaluation. PATIENTS AND METHODS: Fourteen patients with skull-base meningioma were retrospectively identified and planned with proton pencil beam scanning (PBS) single-field uniform dose (SFUD) and multifield optimization (MFO) techniques. The setup uncertainties were assigned a probability model on the basis of prior published data. The impact on the dose distribution from nominal 1-mm and large, less probable setup errors, as well as the cumulative effect, was analyzed. The robustness of SFUD and MFO planning techniques in these scenarios was discussed. RESULTS: The target coverage was reduced and the plan dose hot spot increased by all setup uncertainty scenarios regardless of the planning techniques. For 1 mm nominal shifts, the deviations in clinical target volume (CTV) coverage D99% was -11 ± 52 cGy and -45 ± 147 cGy for SFUD and MFO plans. The setup uncertainties affected the organ at risk (OAR) dose both positively and negatively. The statistical average of the setup uncertainties had <100 cGy impact on the plan qualities for all patients. The cumulative deviations in CTV D95% were 1 ± 34 cGy and -7 ± 18 cGy for SFUD and MFO plans. CONCLUSION: It is important to understand the impact of setup uncertainties on skull-base meningioma, as the tumor target has complex shape and is in proximity to multiple critical organs. Our work evaluated the setup uncertainty based on its probability distribution and evaluated the dosimetric consequences. In general, the SFUD plans demonstrated more robustness than the MFO plans in target coverages and brainstem dose. The probability-weighted overall effect on the dose distribution is small compared to the dosimetric shift during single fraction.