Abstract Recent studies have highlighted the importance of using molecular biomarkers (IDH, 1p19q, TERT) to group gliomas that have similar clinical behavior, response to therapy, and outcome. An emerging hypothesis is that glioma specific genetic and/or molecular alterations manifest as specific observable changes in MR anatomical imaging. Morphologic and texture features, originating from anatomical MRI, have been investigated as imaging biomarkers to predict MGMT and glioma group status. These texture or morphologic based approaches pose several challenges including requirements for several preprocessing steps such as intensity standardization, skull stripping, and tumor segmentation. Deep learning is an important evolving technology in many different fields, including anatomic imaging, and can be used to empirically identify important features in a variety of modalities, including MRI. Importantly deep learning precludes the need for extensive pre-processing. We describe a convolutional neural network, evaluating resnet50, vgg16, inception and xception neural network architectures, that can predict IDH, 1p19q and TERT status utilizing conventional T2 weighted MRI imaging with intensity normalization and nonuniform intensity normalization (N4) bias corrections. The dataset consisted of 432 images (340 for training and 92 for validation) from patients published by Eckel-Passow et al New England Journal of Medicine (2015). The system achieved a weighted f1 score of 0.901, 0.937 and 0.924 for IDH, 1p19q and TERT prediction on the test dataset, respectively. IDH status was misclassified in 9 out of 92 patients, while 1p19q and TERT status was misclassified in 6 and 7 patients respectively. Our results demonstrate the potential of deep learning architectures applied to conventional MRI to predict molecular glioma groups.

Abstract

BACKGROUND

Heat shock protein vaccines can be used to stimulate anti-tumor immune responses. An autologous vaccine (HSPPC-96) generated from patient resected tumors has been previously studied in single arm phase I/II trials for newly diagnosed and recurrent GBM. Based on promising survival in these studies, a randomized, multi-centered phase II trial of HSPPC-96 for recurrent GBM compared to bevacizumab was undertaken (NCT01814813).

METHODS

Patients with 1st/2nd recurrence of resectable GBM were enrolled and underwent surgery with generation of HSPPC-96 from autologous tumors. Post-operative eligibility included a volumetric extent of resection ≥90% (by central review) and sufficient tumor to generate a minimum of 4 vaccine doses. Patients were randomized (1:1:1) to receive HSPPC-96 vaccine followed by bevacizumab at subsequent progression vs. concurrent HSPPC-96 vaccine and bevacizumab vs. bevacizumab alone. The primary endpoint was overall survival (OS) aiming to detect a 36% reduction in HR with 85% power and alpha=0.1 for pooled HSPPC-96 groups vs. bevacizumab alone. Planned enrollment was 165 patients (n=55/arm). An interim analysis was pre-planned at 50% of events (65 deaths). Secondary endpoints included progression-free survival and adverse events.

RESULTS

At final analysis, 90 patients were enrolled with a distribution of 29:30:31. The study was terminated for futility after the interim analysis. In the intention to treat population, OS for the HSPPC-96 treated groups was 7.5 vs. 10.7 months for bevacizumab alone (HR=2.06 [95% CI 1.18–3.60], p=0.008). Among patients treated per protocol (n=73), OS for HSPPC-96 patients was 8.6 vs. 12.3 months (HR=1.99 [95% CI 1.03–3.81], p=0.03). Trends were similar for progression-free survival between groups. HSPPC-96 toxicity was well tolerated with no attributable serious adverse events.

CONCLUSIONS

The study failed to demonstrate a survival benefit for patients treated with HSPPC-96 alone or in combination with bevacizumab compared to bevacizumab alone. Correlative analyses are ongoing. SUPPORT: U10CA180882, U10CA180821, U10CA180868, U24CA196171.

Background

Whole brain radiotherapy (WBRT) is the standard of care to improve intracranial control following resection of brain metastasis. However, stereotactic radiosurgery (SRS) to the surgical cavity is widely used in an attempt to reduce cognitive toxicity, despite the absence of high-level comparative data substantiating efficacy in the postoperative setting. We aimed to establish the effect of SRS on survival and cognitive outcomes compared with WBRT in patients with resected brain metastasis.

Methods

In this randomised, controlled, phase 3 trial, adult patients (aged 18 years or older) from 48 institutions in the USA and Canada with one resected brain metastasis and a resection cavity less than 5·0 cm in maximal extent were randomly assigned (1:1) to either postoperative SRS (12-20 Gy single fraction with dose determined by surgical cavity volume) or WBRT (30 Gy in ten daily fractions or 37·5 Gy in 15 daily fractions of 2·5 Gy; fractionation schedule predetermined for all patients at treating centre). We randomised patients using a dynamic allocation strategy with stratification factors of age, duration of extracranial disease control, number of brain metastases, histology, maximal resection cavity diameter, and treatment centre. Patients and investigators were not masked to treatment allocation. The co-primary endpoints were cognitive-deterioration-free survival and overall survival, and analyses were done by intention to treat. We report the final analysis. This trial is registered with ClinicalTrials.gov, number NCT01372774.

Findings

Between Nov 10, 2011, and Nov 16, 2015, 194 patients were enrolled and randomly assigned to SRS (98 patients) or WBRT (96 patients). Median follow-up was 11·1 months (IQR 5·1-18·0). Cognitive-deterioration-free survival was longer in patients assigned to SRS (median 3·7 months [95% CI 3·45-5·06], 93 events) than in patients assigned to WBRT (median 3·0 months [2·86-3·25], 93 events; hazard ratio [HR] 0·47 [95% CI 0·35-0·63]; p<0·0001), and cognitive deterioration at 6 months was less frequent in patients who received SRS than those who received WBRT (28 [52%] of 54 evaluable patients assigned to SRS vs 41 [85%] of 48 evaluable patients assigned to WBRT; difference -33·6% [95% CI -45·3 to -21·8], p<0·00031). Median overall survival was 12·2 months (95% CI 9·7-16·0, 69 deaths) for SRS and 11·6 months (9·9-18·0, 67 deaths) for WBRT (HR 1·07 [95% CI 0·76-1·50]; p=0·70). The most common grade 3 or 4 adverse events reported with a relative frequency greater than 4% were hearing impairment (three [3%] of 93 patients in the SRS group vs eight [9%] of 92 patients in the WBRT group) and cognitive disturbance (three [3%] vs five [5%]). There were no treatment-related deaths.

Interpretation

Decline in cognitive function was more frequent with WBRT than with SRS and there was no difference in overall survival between the treatment groups. After resection of a brain metastasis, SRS radiosurgery should be considered one of the standards of care as a less toxic alternative to WBRT for this patient population.

Funding

National Cancer Institute.

BACKGROUND: A recent phase III trial (NCT01372774) comparing use of stereotactic radiosurgery [SRS] versus whole-brain radiation therapy [WBRT] after surgical resection of a single brain metastasis revealed that declines in cognitive function were more common with WBRT than with SRS. A secondary endpoint in that trial, and the primary objective in this secondary analysis, was to identify baseline biomarkers associated with cognitive impairment after either form of radiotherapy for brain metastasis. Here we report our findings on APOE genotype and serum levels of associated proteins and their association with radiation-induced neurocognitive decline. METHODS: In this retrospective analysis of prospectively collected samples from a completed randomized clinical trial, patients provided blood samples every 3 months that were tested by genotyping and enzyme-linked immunosorbent assay, and results were analyzed in association with cognitive impairment. RESULTS: The APOE genotype was not associated with neurocognitive impairment at 3 months. However, low serum levels of ApoJ, ApoE, or ApoA protein (all P < .01) and higher amyloid beta (Aβ 1-42) levels (P = .048) at baseline indicated a greater likelihood of neurocognitive decline at 3 months after SRS, whereas lower ApoJ levels were associated with decline after WBRT (P = .014). CONCLUSIONS: Patients with these pretreatment serum markers should be counseled about radiation-related neurocognitive decline.

PURPOSE: Whole brain radiation therapy (WBRT) remains a commonly used cancer treatment, although controversy exists regarding the optimal dose/fractionation to optimize intracranial tumor control and minimize resultant cognitive deficits. METHODS AND MATERIALS: NCCTG N107C [Alliance]/CEC.3 randomized 194 patients with brain metastases to either stereotactic radiosurgery alone or WBRT after surgical resection. Among the 92 patients receiving WBRT, sites predetermined the dose/fractionation that would be used for all patients treated at that site (either 30 Gy in 10 fractions or 37.5 Gy in 15 fractions). Analyses were performed using Kaplan-Meier estimates, log rank tests, and Fishers exact tests. RESULTS: Among 92 patients treated with surgical resection and adjuvant WBRT, 49 were treated with 30 Gy in 10 fractions (53%), and 43 were treated with 37.5 Gy in 15 fractions (47%). Baseline characteristics, including cognitive testing, were well balanced between groups with the exception of primary tumor type (lung cancer histology was more frequent with protracted WBRT: 72% vs 45%, P = .01), and 93% of patients completed the full course of WBRT. A more protracted WBRT dose regimen (37.5 Gy in 15 fractions) did not significantly affect time to cognitive failure (hazard ratio [HR], 0.9; 95% confidence interval [CI], 0.6-1.39; P = .66), surgical bed control (HR, 0.52 [95% CI, 0.22-1.25], P = .14), intracranial tumor control (HR, 0.56 [95% CI, 0.28-1.12], P = .09), or overall survival (HR, 0.72 [95% CI, 0.45-1.16], P = .18). Although there was no reported radionecrosis, there is a statistically significant increase in the risk of at least 1 grade ≥3 adverse event with 37.5 Gy in 15 fractions versus 30 Gy in 10 fractions (54% vs 31%, respectively, P = .03). CONCLUSIONS: This post hoc analysis does not demonstrate that protracted WBRT courses reduce the risk of cognitive deficit, improve tumor control in the hypoxic surgical cavity, or otherwise improve the therapeutic ratio. Adverse events were significantly higher with the lengthened course of WBRT. For patients with brain metastases where WBRT is recommended, shorter course hypofractionated regimens remain the current standard of care.

A recent meeting was held on March 22, 2019, among the FDA, clinical scientists, pharmaceutical and biotech companies, clinical trials cooperative groups, and patient advocacy groups to discuss challenges and potential solutions for increasing development of therapeutics for central nervous system metastases. A key issue identified at this meeting was the need for consistent tumor measurement for reliable tumor response assessment, including the first step of standardized image acquisition with an MRI protocol that could be implemented in multicenter studies aimed at testing new therapeutics. This document builds upon previous consensus recommendations for a standardized brain tumor imaging protocol (BTIP) in high-grade gliomas and defines a protocol for brain metastases (BTIP-BM) that addresses unique challenges associated with assessment of CNS metastases. The "minimum standard" recommended pulse sequences include: (i) parameter matched pre- and post-contrast inversion recovery (IR)-prepared, isotropic 3D T1-weighted gradient echo (IR-GRE); (ii) axial 2D T2-weighted turbo spin echo acquired after injection of gadolinium-based contrast agent and before post-contrast 3D T1-weighted images; (iii) axial 2D or 3D T2-weighted fluid attenuated inversion recovery; (iv) axial 2D, 3-directional diffusion-weighted images; and (v) post-contrast 2D T1-weighted spin echo images for increased lesion conspicuity. Recommended sequence parameters are provided for both 1.5T and 3T MR systems. An "ideal" protocol is also provided, which replaces IR-GRE with 3D TSE T1-weighted imaging pre- and post-gadolinium, and is best performed at 3T, for which dynamic susceptibility contrast perfusion is included. Recommended perfusion parameters are given.