Intraoperative diagnosis is essential for providing safe and effective care during cancer surgery 1. The existing workflow for intraoperative diagnosis based on hematoxylin and eosin-staining of processed tissue is time-, resource-, and labor-intensive 2,3. Moreover, interpretation of intraoperative histologic images is dependent on a contracting, unevenly distributed pathology workforce 4. Here, we report a parallel workflow that combines stimulated Raman histology (SRH) 5-7 , a label-free optical imaging method, and deep convolutional neural networks (CNN) to predict diagnosis at the bedside in near real-time in an automated fashion. Specifically, our CNN, trained on over 2.5 million SRH images, predicts brain tumor diagnosis in the operating room in under 150 seconds, an order of magnitude faster than conventional techniques (e.g., 20-30 minutes) 2. In a multicenter, prospective clinical trial (n = 278) we demonstrated that CNN-based diagnosis of SRH images was non-inferior to pathologist-based interpretation of conventional histologic images (overall accuracy, 94.6% vs. 93.9%). Our CNN learned a hierarchy of recognizable histologic feature representations to classify the major histopathologic classes of brain tumors. Additionally, we implemented a semantic segmentation method to identify tumor infiltrated, diagnostic regions within SRH images. These results demonstrate how intraoperative cancer diagnosis can be streamlined, creating a complimentary pathway for tissue diagnosis that is independent of a traditional pathology laboratory.

Glioblastoma remains a lethal diagnosis with a 5-year survival rate of less than 10%. (NEJM 352:987-96, 2005) Although immunotherapy-based approaches are capable of inducing detectable immune responses against tumor-specific antigens, improvements in clinical outcomes are modest, in no small part due to tumor-induced immunosuppressive mechanisms that promote immune escape and immuno-resistance. Immunotherapeutic strategies aimed at bolstering the immune response while neutralizing immunosuppression will play a critical role in improving treatment outcomes for glioblastoma patients. In vivo murine models of glioma provide an invaluable resource to achieving that end, and their use is an essential part of the preclinical workup for novel therapeutics that need to be tested in animal models prior to testing experimental therapies in patients. In this article, we review five contemporary immunocompetent mouse models, GL261 (C57BL/6), GL26 (C57BL/6) CT-2A (C57BL/6), SMA-560 (VM/Dk), and 4C8 (B6D2F1), each of which offer a suitable platform for testing novel immunotherapeutic approaches.

Glioblastoma multiforme (GBM) is the most common and aggressive malignant adult primary brain tumor. Despite surgical resection followed by radiotherapy and chemotherapy, the median survival rate is approximately 14 months. Although experimental therapies are in clinical trials for GBM, there is an urgent need for a peripheral GBM biomarker for measuring treatment response. As we have previously demonstrated that the long noncoding RNA HOX Transcript Antisense Intergenic RNA, or HOTAIR, is dysregulated in GBM and required for GBM cell proliferation, we hypothesized that HOTAIR expression may be utilized as a peripheral biomarker for GBM. HOTAIR expression was measured in serum from 43 GBM and 40 controls using quantitative real-time PCR (qRT-PCR). The PCR products were subsequently subcloned into pCR™4-TOPO®TA vectors for DNA sequencing. A ROC curve was also generated to examine HOTAIR’s prognostic value. The amount of HOTAIR in serum exosomes and exosome-depleted supernatant was calculated by qRT-PCR. The relative HOTAIR expression was also investigated in 15 pairs of GBM serum and tumors. We detected HOTAIR in serum from GBM patients. HOTAIR levels in serum samples from GBM patients was significantly higher than in the corresponding controls (P < 0.0001). The area under the ROC curve distinguishing GBM patients from controls was 0.913 (95% CI: 0.845–0.982, P < 0.0001), with 86.1% sensitivity and 87.5% specificity at the cut-off value of 10.8. HOTAIR expression was significantly correlated with high grade brain tumors. In addition, Pearson correlation analysis indicated a medium correlation of serum HOTAIR levels and the corresponding tumor HOTAIR levels (r = 0.734, P < 0.01). We confirmed via sequencing that the amplified HOTAIR from serum contained the HOTAIR sequence and maps to the known HOTAIR locus at 12q13. The serum-derived exosomes contain HOTAIR and the purified exosomes were validated by western blot and nanoparticle tracking analysis. Importantly, our results demonstrate that serum HOTAIR can be used as a novel prognostic and diagnostic biomarker for GBM.Electronic supplementary materialThe online version of this article (10.1186/s12943-018-0822-0) contains supplementary material, which is available to authorized users.