Purpose: BAY1436032, an inhibitor of mutant isocitrate dehydrogenase 1 (mIDH1), was active against multiple IDH1-R132X solid tumors in preclinical models. This first-in-human study was designed to determine the safety and pharmacokinetics of BAY1436032, and to evaluate its potential pharmacodynamics and antitumor effects. Patients and Methods: The study comprised of dose escalation and dose expansion cohorts. BAY1436032 tablets were orally administered twice daily on a continuous basis in subjects with mIDH1 solid tumors. Results: In dose escalation, 29 subjects with various tumor types were administered BAY1436032 across five doses (150–1,500 mg twice daily). BAY1432032 exhibited a relatively short half-life. Most evaluable subjects experienced target inhibition as indicated by a median maximal reduction of plasma R-2-hydroxyglutarate levels of 76%. BAY1436032 was well tolerated and an MTD was not identified. A dose of 1,500 mg twice daily was selected for dose expansion, where 52 subjects were treated in cohorts representing four different tumor types [lower grade glioma (LGG), glioblastoma, intrahepatic cholangiocarcinoma, and a basket cohort of other tumor types]. The best clinical outcomes were in subjects with LGG (n = 35), with an objective response rate of 11% (one complete response and three partial responses) and stable disease in 43%. As of August 2020, four of these subjects were in treatment for >2 years and still ongoing. Objective responses were observed only in LGG. Conclusions: BAY1436032 was well tolerated and showed evidence of target inhibition and durable objective responses in a small subset of subjects with LGG.

Purpose: The prospective, randomized ERGO2 trial investigated the effect of calorie-restricted ketogenic diet and intermittent fasting (KD-IF) on re-irradiation for recurrent brain tumors. The study did not meet its primary endpoint of improved progression-free survival in comparison to standard diet (SD). We here report the results of the quality of life/neurocognition and a detailed analysis of the diet diaries. Methods: 50 patients were randomized 1:1 to re-irradiation combined with either SD or KD-IF. The KD-IF schedule included 3 days of ketogenic diet (KD: 21–23 kcal/kg/d, carbohydrate intake limited to 50 g/d), followed by 3 days of fasting and again 3 days of KD. Follow-up included examination of cognition, quality of life and serum samples. Results: The 20 patients who completed KD-IF met the prespecified goals for calorie and carbohydrate restriction. Substantial decreases in leptin and insulin and an increase in uric acid were observed. The SD group, of note, had a lower calorie intake than expected (21 kcal/kg/d instead of 30 kcal/kg/d). Neither quality of life nor cognition were affected by the diet. Low glucose emerged as a significant prognostic parameter in a best responder analysis. Conclusion: The strict caloric goals of the ERGO2 trial were tolerated well by patients with recurrent brain cancer. The short diet schedule led to significant metabolic changes with low glucose emerging as a candidate marker of better prognosis. The unexpected lower calorie intake of the control group complicates the interpretation of the results. Clinicaltrials.gov number: NCT01754350; Registration: 21.12.2012.

Background: The ERGO2 (Ernaehrungsumstellung bei Patienten mit Rezidiv eines Glioblastoms) MR-spectroscopic imaging (MRSI) subtrial investigated metabolism in patients randomized to calorically restricted ketogenic diet/intermittent fasting (crKD-IF) versus standard diet (SD) in addition to re-irradiation (RT) for recurrent malignant glioma. Intracerebral concentrations of ketone bodies (KB), intracellular pH (pH_i), and adenosine triphosphate (ATP) were non-invasively determined. Methods: 50 patients were randomized (1:1): Group A keeping a crKD-IF for nine days, and Group B a SD. RT was performed on day 4–8. Twenty-three patients received an extended MRSI-protocol (¹H decoupled ³¹P MRSI with 3D chemical shift imaging (CSI) and 2D ¹H point-resolved spectroscopy (PRESS)) at a 3T scanner at baseline and on day 6. Voxels were selected from the area of recurrent tumor and contralateral hemisphere. Spectra were analyzed with LCModel, adding simulated signals of 3-hydroxybutyrate (βOHB), acetone (Acn) and acetoacetate (AcAc) to the standard basis set. Results: Acn was the only reliably MRSI-detectable KB within tumor tissue and/or normal appearing white matter (NAWM). It was detected in 4/11 patients in Group A and in 0/8 patients in Group B. MRSI results showed no significant depletion of ATP in tumor tissue of patients at day 6 during crKD-IF, even though there were a significant difference in ketone serum levels between Group A and B at day 6 and a decline in fasting glucose in Group A from baseline to day 6. The tumor specific alkaline pH_i was maintained. Conclusions: Our metabolic findings suggest that tumor cells maintain energy homeostasis even with reduced serum glucose levels and may generate additional ATP through other sources.

BACKGROUND AND PURPOSE: Preclinical evidence points toward a metabolic reprogramming in isocitrate dehydrogenase (IDH) mutated tumor cells with down-regulation of the expression of genes that encode for glycolytic metabolism. We noninvasively investigated lactate and Cr concentrations, as well as intracellular pH using ¹H/phosphorus 31 (³¹P) MR spectroscopy in a cohort of patients with gliomas. MATERIALS AND METHODS: Thirty prospectively enrolled, mostly untreated patients with gliomas met the spectral quality criteria (World Health Organization II [n = 7], III [n = 16], IV [n = 7]; IDH-mutant [n = 23]; IDH wild-type [n = 7]; 1p/19q codeletion [n = 9]). MR imaging protocol included 3D ³¹P chemical shift imaging and ¹H single-voxel spectroscopy (point-resolved spectroscopy sequence at TE = 30 ms and TE = 97 ms with optimized echo spacing for detection of 2-hydroxyglutarate) from the tumor area. Values for absolute metabolite concentrations were calculated (phantom replacement method). Intracellular pH was determined from ³¹P chemical shift imaging. RESULTS: At TE = 97 ms, lactate peaks can be fitted with little impact of lipid/macromolecule contamination. We found a significant difference in lactate concentrations, lactate/Cr ratios, and intracellular pH when comparing tumor voxels of patients with IDH-mutant with those of patients with IDH wild-type gliomas, with reduced lactate levels and near-normal intracellular pH in patients with IDH-mutant gliomas. We additionally found evidence for codependent effects of 1p/19q codeletion and IDH mutations with regard to lactate concentrations for World Health Organization tumor grades II and III, with lower lactate levels in patients exhibiting the codeletion. There was no statistical significance when comparing lactate concentrations between IDH-mutant World Health Organization II and III gliomas. CONCLUSIONS: We found indirect evidence for metabolic reprogramming in IDH-mutant tumors with significantly lower lactate concentrations compared with IDH wild-type tumors and a near-normal intracellular pH.

Non-small-cell lung cancer (NSCLC) shows a high incidence of brain metastases (BM). Early detection is crucial to improve clinical prospects. We trained and validated classifier models to identify patients with a high risk of developing BM, as they could potentially benefit from surveillance brain MRI. Consecutive patients with an initial diagnosis of NSCLC from January 2011 to April 2019 and an in-house chest-CT scan (staging) were retrospectively recruited at a German lung cancer center. Brain imaging was performed at initial diagnosis and in case of neurological symptoms (follow-up). Subjects lost to follow-up or still alive without BM at the data cut-off point (12/2020) were excluded. Covariates included clinical and/or 3D-radiomics-features of the primary tumor from staging chest-CT. Four machine learning models for prediction (80/20 training) were compared. Gini Importance and SHAP were used as measures of importance; sensitivity, specificity, area under the precision-recall curve, and Matthew's Correlation Coefficient as evaluation metrics. Three hundred and ninety-five patients compromised the clinical cohort. Predictive models based on clinical features offered the best performance (tuned to maximize recall: sensitivity∼70%, specificity∼60%). Radiomics features failed to provide sufficient information, likely due to the heterogeneity of imaging data. Adenocarcinoma histology, lymph node invasion, and histological tumor grade were positively correlated with the prediction of BM, age, and squamous cell carcinoma histology were negatively correlated. A subgroup discovery analysis identified 2 candidate patient subpopulations appearing to present a higher risk of BM (female patients + adenocarcinoma histology, adenocarcinoma patients + no other distant metastases). Analysis of the importance of input features suggests that the models are learning the relevant relationships between clinical features/development of BM. A higher number of samples is to be prioritized to improve performance. Employed prospectively at initial diagnosis, such models can help select high-risk subgroups for surveillance brain MRI.