P14.06 Radiomics for the non-invasive determination of the BRAF mutational status in patients with melanoma brain metastases

BACKGROUND

The BRAF V600E mutation is present in approximately 50% of patients with melanoma and is an important prerequisite for a response to targeted therapies such as BRAF inhibitors. In the majority of patients, the BRAF mutational status is based on the analysis of tissue samples from the extracranial primary tumor only. Since the extracranial and intracranial BRAF mutational status may be discrepant, the additional information on the BRAF mutational status of melanoma brain metastases would be of clinical value, e.g., for the prediction of response to targeted therapies. Here, we evaluated the potential of structural MRI radiomics for the determination of the intracranial BRAF mutational status in patients with melanoma brain metastases.

MATERIAL AND METHODS

Fifty-nine patients with melanoma brain metastases from two university hospitals (group 1, 45 patients; group 2, 14 patients) underwent surgery with subsequent genetic analysis of the brain metastases tissue to determine the BRAF mutational status. All patients underwent structural MRI preoperatively. Areas of contrast enhancement were manually segmented and analyzed. Group 1 was used for model training and validation, group 2 for model testing. After image preprocessing, 1,316 radiomics features were extracted using the open-source PyRadiomics package. A test-retest analysis was performed to identify robust features prior to feature selection. Finally, the best performing radiomics model was applied to the test data (group 2). Diagnostic performances were evaluated using receiver operating characteristic (ROC) analyses.

RESULTS

Twenty-two patients (49%) in group 1, and 6 patients (43%) in group 2 had an intrametastatic BRAF V600E mutation. Using a six parameter radiomics signature, a linear support vector machine classifier yielded an average area under the ROC curve (AUC) of 0.87 (accuracy, 85%; sensitivity, 78%; specificity, 91%) for prediction of the BRAF mutational status in the training data (group 1). Finally, the classifier achieved an AUC of 0.85 (accuracy, 86%; sensitivity, 83%; specificity, 88%) in the test data (group 2).

CONCLUSION

The developed radiomics classifier allows a non-invasive prediction of the intracranial BRAF V600E mutational status in patients with melanoma brain metastases and may be of value for treatment decisions.

OS08.4.A Retrospective analysis of in vivo 1H-magnetic resonance spectroscopy based on a machine learning approach enables reliable prediction of IDH mutation in patients with glioma

BACKGROUND

Mutation of isocitrate dehydrogenase (IDH) is not only an important landmark in the development of low-grade gliomas, but also has prognostic significance and is a potential therapeutic target. There is a high need to determinate IDH mutation status at diagnosis and during the course of therapy in a non-invasive and reliable manner. We established a machine learning approach based on a support vector machine to detect IDH mutation status in in vivo standard 1H-magnetic resonance spectroscopy (1H-MRS) at 3T with an accuracy of 88.2%, a sensitivity of 95.5% (95% CI, 77.2–99.9%), and a specificity of 75% (95% CI, 42.85–94.5%) in a prospective monocentric clinical trial. Here, the same method is applied in a retrospective cohort at 1.5T and tested for transferability.

MATERIAL AND METHODS

Validation cohort. The validation cohort comprised 100 patients with glioma for which standard in vivo 1H-MRS spectra had been acquired between 2002 and 2007. Standard single voxel spectroscopy had been measured at 1.5T using a PRESS sequence with a TR of 1500ms and a TE of 30ms. One sample had to be excluded due to non-malignant histology and for 15 samples the IDH mutation status was not available. Therefore, the validation cohort comprised 84 samples, of which 35 were bearing an IDH mutation in immunohistochemistry (sequencing for confirmation is outstanding). Machine learning. To transfer our method to an independent validation cohort our previously established machine learning approach was first trained on all samples of the 3T group. The trained algorithm was then applied to the data of the validation cohort. Here, among other factors the different field strengths, with which the spectra were acquired (3T vs. 1.5T) had to be considered.

RESULTS

27 samples of the validation cohort had to be excluded due to poor spectra quality. Our approach correctly detected IDH mutation status in 47 of 62 patients (75.8%), although the technical conditions were significantly different from our published prospective cohort. 17 of 30 patients bearing an IDH mutation were correctly identified, while 30 of 32 wild type patients were determined successfully.

CONCLUSION

Our approach to detect IDH mutation status has promising application in an unselected retrospective cohort, demonstrating transferability across different technical conditions. Further investigations to improve our technique and an advanced neuropathological processing of the samples are planned.

The macro-metastasis/organ parenchyma interface (MMPI) - A hitherto unnoticed area

The macro-metastasis/organ parenchyma interface (MMPI) was previously considered an inert anatomical border which sharply separates the affected organ parenchyma from the macro-metastatic tissue. Recently, infiltrative growth of macro-metastases from various primary tumors was described in the brain, liver and lung, with significant impact on survival. Strikingly, the MMPI patterns differed between entities, so that at least nine different patterns were described. The MMPI patterns could be further classified into three major groups: displacing, epithelial and diffuse infiltrating. Additionally, macro-metastases are a source of further tumor cell dissemination in the affected organ; and these intra-organ metastatic dissemination tracks starting from the MMPI also vary depending on the anatomical structures of the colonized organ and influence disease outcome. In spite of their relevance, MMPIs and organ-specific dissemination tracks are still largely overlooked by many clinicians, pathologists and/or researchers. In this review, we aim to address this important issue and enhance our current understanding of the different MMPI patterns and dissemination tracks in the brain, liver and lung.

Use of the pCONus as an adjunct to coil embolization of acutely ruptured aneurysms

Introduction

Coil embolization of ruptured aneurysms has become the standard treatment in many situations. However, certain aneurysm morphologies pose technical difficulties and may require the use of adjunctive devices.

Objective

To present our experience with the pCONus, a new neck bridging device, as an adjunct to coil embolization for acutely ruptured aneurysms and discuss the technical success, angiographic and clinical outcomes.

Methods

We conducted a retrospective review of our database of prospectively collected data to identify all patients who presented with acute subarachnoid hemorrhage that required adjunctive treatment with the pCONus in the acute stage. We searched the database between April 2011 and April 2016.

Results

21 patients were identified (13 male, 8 female) with an average age of 54.6 years (range 31–73). 8 aneurysms were located at the basilar artery tip, 7 at the anterior communicating artery, 4 at the middle cerebral artery bifurcation, 1 pericallosal, and 1 basilar fenestration. 61.8% patients achieved modified Raymond–Roy classification I or II at immediate angiography, with 75% of patients having completely occluded aneurysms or stable appearance at initial follow-up. There were no repeat aneurysmal ruptures and two device-related complications (no permanent morbidity). Four patients in our cohort died.

Conclusions

Use of the pCONus is safe and effective in patients with acutely ruptured aneurysms and carries a high rate of technical success.