Mutant isocitrate dehydrogenase 1 (IDH1) is common in gliomas, and produces D-2-hydroxyglutarate (D-2-HG). The full effects of IDH1 mutations on glioma biology and tumor microenvironment are unknown. We analyzed a discovery cohort of 169 World Health Organization (WHO) grade II-IV gliomas, followed by a validation cohort of 148 cases, for IDH1 mutations, intratumoral microthrombi, and venous thromboemboli (VTE). 430 gliomas from The Cancer Genome Atlas were analyzed for mRNAs associated with coagulation, and 95 gliomas in a tissue microarray were assessed for Tissue Factor (TF) protein. In vitro and in vivo assays evaluated platelet aggregation and clotting time in the presence of mutant IDH1 or D-2-HG. VTE occurred in 26–30% of patients with wild-type IDH1 gliomas, but not in patients with mutant IDH1 gliomas (0%). IDH1 mutation status was the most powerful predictive marker for VTE, independent of variables such as GBM diagnosis and prolonged hospital stay. Microthrombi were far less common within mutant IDH1 gliomas regardless of WHO grade (85–90% in wild-type versus 2–6% in mutant), and were an independent predictor of IDH1 wild-type status. Among all 35 coagulation-associated genes, F3 mRNA, encoding TF, showed the strongest inverse relationship with IDH1 mutations. Mutant IDH1 gliomas had F3 gene promoter hypermethylation, with lower TF protein expression. D-2-HG rapidly inhibited platelet aggregation and blood clotting via a novel calcium-dependent, methylation-independent mechanism. Mutant IDH1 glioma engraftment in mice significantly prolonged bleeding time. Our data suggest that mutant IDH1 has potent antithrombotic activity within gliomas and throughout the peripheral circulation. These findings have implications for the pathologic evaluation of gliomas, the effect of altered isocitrate metabolism on tumor microenvironment, and risk assessment of glioma patients for VTE.

No abstract

Objective We aimed to determine the prevalence of hydronephrosis (HN) in patients who underwent renal sonography for new onset acute kidney injury (AKI) and to identify clinical factors predictive of HN. In patients with HN, we sought to investigate how routine renal US affects patient management, including performance of interventional procedures. Methods A retrospective chart review identified 274 adults with AKI who underwent renal ultrasound (US) at an urban teaching hospital from January 2011 through July 2011. The prevalence of HN was determined. Electronic medical records were reviewed for co-morbidities, including risk factors for HN such as pelvic mass, prior renal or pelvic surgery and neurogenic bladder, and for subsequent interventions and outcomes. Results US demonstrated HN in 28 patients (10%); 5 (18%) had subsequent interventions. In a multivariable logistic regression model with the outcome being HN, all considered risk factors, pelvic mass, prior renal or pelvic surgery and neurogenic bladder, were significantly associated with HN (odds ratio 6.4 (95% confidence interval 2.7, 15.4) and p < 0.001) when adjusting for age and diabetes mellitus. Diabetes had a negative predictive value for HN. No diabetic patients younger than 85 years and without clinical risk factors had HN. Conclusions HN is infrequently seen on sonograms in hospitalized patients with AKI who lack risk factors for urinary tract obstruction. Deferral of the US pending a trial of medical management is safe and will reduce medical costs. Adoption of clinical guidelines to assess patients’ risk levels for HN is critical to avoid unnecessary imaging.

No abstract