ABSTRACT
Background: As our molecular understanding of pediatric central nervous system (CNS) tumors evolves, so too do diagnostic criteria, prognostic biomarkers, and clinical management decision making algorithms. Here, we explore the clinical utility of wide-breadth assays, including whole-exome sequencing (WES), RNA sequencing (RNA-seq), and methylation array profiling as an addition to more conventional diagnostic tools for pediatric CNS tumors. Methods: This study comprises an observational, prospective cohort followed at a single academic medical center over 3 years. Paired tumor and normal control specimens from 53 enrolled pediatric patients with CNS tumors underwent WES. A subset of cases also underwent RNA-seq (n = 28) and/or methylation array analysis (n = 27). Results: RNA-seq identified the driver and/or targetable fusions in 7/28 cases, including potentially targetable NTRK fusions, and uncovered possible rationalized treatment options based on outlier gene expression in 23/28 cases. Methylation profiling added diagnostic confidence (8/27 cases) or diagnostic subclassification endorsed by the WHO (10/27 cases). WES detected clinically pertinent tier 1 or tier 2 variants in 36/53 patients. Of these, 16/17 SNVs/INDELs and 10/19 copy number alterations would have been detected by current in-house conventional tests including targeted sequencing panels. Conclusions: Over a heterogeneous set of pediatric tumors, RNA-seq and methylation profiling frequently yielded clinically relevant information orthogonal to conventional methods while WES demonstrated clinically relevant added value primarily via copy number assessment. Longitudinal cohorts comparing targeted molecular pathology workup vs broader genomic approaches including therapeutic selection based on RNA expression data will be necessary to further evaluate the clinical benefits of these modalities in practice.
ABSTRACT
Covid-19 is affecting ethnic minorities more severely, but we will never understand why if we don't collect the right data, says Alisha Dua.
ABSTRACT
Hyperspectral imaging (HSI) systems have the potential to retrieve in vivo hemodynamic and metabolic signals from the exposed cerebral cortex. The use of multiple narrow wavelength bands in the near infrared (NIR) range theoretically allows not only to image brain tissue oxygenation and hemodynamics via mapping of hemoglobin concentration changes, but also to directly quantify cerebral metabolism via measurement of the redox states of mitochondrial cytochrome-c-oxidase (CCO). The aim of this study is to assess the possibility of performing hyperspectral imaging of in vivo cerebral oxyhemoglobin (HbO2), deoxyhemoglobin (HHb) and oxidized CCO (oxCCO) using commercially available HSI devices. For this reason, a hyperspectral snapshot solution based on Cubert GmbH technology (S185 FireflEYE camera) has been tested on the exposed cortex of mice during normoxic, hypoxic and hyperoxic conditions. The system allows simultaneous acquisition of 138 wavelength bands between 450 and 998 nm, with spectral sampling and resolution of ~4 to 8 nm. From the hyperspectral data, relative changes in concentration of hemoglobin and oxCCO are estimated and hemodynamic and metabolic maps of the imaged cortex are calculated for two different NIR spectral ranges. Spectroscopic analysis at particular regions of interest is also performed, showing typical oxygen-dependent hemodynamic responses. The results highlight some of the potentials of the technology, but also the limitations of the tested commercial solution for such specific application, in particular regarding spatial resolution.
