Search

Supplementary Table S2

Suppl Data incl Figure S1 and Table S3

Most patients whose large bowel cancer has spread to other organs do not respond to immune therapy. We detected a rare gene mutation, termed 9p24.1 copy-number gain (CNG), in an otherwise incurable colorectal cancer that provoked an immune therapy response. We identified this gene mutation by gene-panel sequencing of DNA from a liver metastasis biopsy from a patient who had disease refractory to standard therapies. Following immune checkpoint blockade (ICB) with pembrolizumab (anti–PD-1), the patient experienced conversion of the tumor phenotype from one with epithelial features to that of an inflamed microenvironment, detected by high-resolution RNA sequencing. Circulating tumor DNA disappeared over the first weeks of therapy. As assessed by standard radiographic measurement, the patient had a partial response that was durable. This patient's response may support the use of histology-agnostic ICB in solid tumors that carry the rare 9p24.1 CNG.

Checkpoint blockade can reverse T-cell exhaustion and promote antitumor responses. Although blocking the PD-1 pathway has been successful in Hodgkin lymphoma, response rates have been modest in B-cell non-Hodgkin lymphoma (NHL). Coblockade of checkpoint receptors may therefore be necessary to optimize antitumor T-cell responses. Here, characterization of coinhibitory receptor expression in intratumoral T cells from different NHL types identified TIGIT and PD-1 as frequently expressed coinhibitory receptors. Tumors from NHL patients were enriched in CD8+ and CD4+ T effector memory cells that displayed high coexpression of TIGIT and PD-1, and coexpression of these checkpoint receptors identified T cells with reduced production of IFNγ, TNFα, and IL2. The suppressed cytokine production could be improved upon in vitro culture in the absence of ligands. Whereas PD-L1 was expressed by macrophages, the TIGIT ligands CD155 and CD112 were expressed by lymphoma cells in 39% and 50% of DLBCL cases and in some mantle cell lymphoma cases, as well as by endothelium and follicular dendritic cells in all NHLs investigated. Collectively, our results show that TIGIT and PD-1 mark dysfunctional T cells and suggest that TIGIT and PD-1 coblockade should be further explored to elicit potent antitumor responses in patients with NHL.

Supplementary figures S1-S8.

Supplementary Table S1-S2.

PDF file - 1.29MB

Table S2. Expression of TIGIT ligands in B cells and T cells.

Supplementary Figure 1. IL-21 produc􀆟on is reduced in CD4 FL T cells; Supplementary Figure 2. TCR-induced signaling effectors have different kine􀆟c; Supplementary Figure 3. Iden􀆟fica􀆟on of T-cell subsets; Supplementary Figure 4. Distribu􀆟on of T-cell subsets in FL LN; Supplementary Figure 5. Expression pa􀆩erns of co-inhibitory receptors in CD8 and CD4 T-cell subsets; Supplementary Figure 6. Contribu􀆟on of Tregs among CD4+TIGIT+ T cells; Supplementary Figure 7. TIGIT is a highly expressed co-inhibitory receptor in FL; Supplementary Figure 8. TIGIT expression is stable over 􀆟me; Supplementary Figure 9. Recovery of TCR-induced signaling in TIGIT+ CD8+ FL T cells is robust.

Purpose: Activating mutations of the anaplastic lymphoma kinase (ALK) were recently described in neuroblastoma. We carried out a meta-analysis of 709 neuroblastoma tumors to determine their frequency and mutation spectrum in relation to genomic and clinical parameters, and studied the prognostic significance of ALK copy number and expression.Experimental Design: The frequency and type of ALK mutations, copy number gain, and expression were analyzed in a new series of 254 neuroblastoma tumors. Data from 455 published cases were used for further in-depth analysis.Results: ALK mutations were present in 6.9% of 709 investigated tumors, and mutations were found in similar frequencies in favorable [International Neuroblastoma Staging System (INSS) 1, 2, and 4S; 5.7%] and unfavorable (INSS 3 and 4; 7.5%) neuroblastomas (P = 0.087). Two hotspot mutations, at positions R1275 and F1174, were observed (49% and 34.7% of the mutated cases, respectively). Interestingly, the F1174 mutations occurred in a high proportion of MYCN-amplified cases (P = 0.001), and this combined occurrence was associated with a particular poor outcome, suggesting a positive cooperative effect between both aberrations. Furthermore, the F1174L mutant was characterized by a higher degree of autophosphorylation and a more potent transforming capacity as compared with the R1275Q mutant. Chromosome 2p gains, including the ALK locus (91.8%), were associated with a significantly increased ALK expression, which was also correlated with poor survival.Conclusions: ALK mutations occur in equal frequencies across all genomic subtypes, but F1174L mutants are observed in a higher frequency of MYCN-amplified tumors and show increased transforming capacity as compared with the R1275Q mutants. Clin Cancer Res; 16(17); 4353–62. ©2010 AACR.

Supplementary Figure 1. IL-21 produc􀆟on is reduced in CD4 FL T cells; Supplementary Figure 2. TCR-induced signaling effectors have different kine􀆟c; Supplementary Figure 3. Iden􀆟fica􀆟on of T-cell subsets; Supplementary Figure 4. Distribu􀆟on of T-cell subsets in FL LN; Supplementary Figure 5. Expression pa􀆩erns of co-inhibitory receptors in CD8 and CD4 T-cell subsets; Supplementary Figure 6. Contribu􀆟on of Tregs among CD4+TIGIT+ T cells; Supplementary Figure 7. TIGIT is a highly expressed co-inhibitory receptor in FL; Supplementary Figure 8. TIGIT expression is stable over 􀆟me; Supplementary Figure 9. Recovery of TCR-induced signaling in TIGIT+ CD8+ FL T cells is robust.

Purpose: T cells infiltrating follicular lymphoma (FL) tumors are considered dysfunctional, yet the optimal target for immune checkpoint blockade is unknown. Characterizing coinhibitory receptor expression patterns and signaling responses in FL T-cell subsets might reveal new therapeutic targets.Experimental Design: Surface expression of 9 coinhibitory receptors governing T-cell function was characterized in T-cell subsets from FL lymph node tumors and from healthy donor tonsils and peripheral blood samples, using high-dimensional flow cytometry. The results were integrated with T-cell receptor (TCR)-induced signaling and cytokine production. Expression of T-cell immunoglobulin and ITIM domain (TIGIT) ligands was detected by immunohistochemistry.Results: TIGIT was a frequently expressed coinhibitory receptor in FL, expressed by the majority of CD8 T effector memory cells, which commonly coexpressed exhaustion markers such as PD-1 and CD244. CD8 FL T cells demonstrated highly reduced TCR-induced phosphorylation (p) of ERK and reduced production of IFNγ, while TCR proximal signaling (p-CD3ζ, p-SLP76) was not affected. The TIGIT ligands CD112 and CD155 were expressed by follicular dendritic cells in the tumor microenvironment. Dysfunctional TCR signaling correlated with TIGIT expression in FL CD8 T cells and could be fully restored upon in vitro culture. The costimulatory receptor CD226 was downregulated in TIGIT+ compared with TIGIT− CD8 FL T cells, further skewing the balance toward immunosuppression.Conclusions: TIGIT blockade is a relevant strategy for improved immunotherapy in FL. A deeper understanding of the interplay between coinhibitory receptors and key T-cell signaling events can further assist in engineering immunotherapeutic regimens to improve clinical outcomes of cancer patients. Clin Cancer Res; 24(4); 870–81. ©2017 AACR.

The tumor microenvironment (TME) and limited immune surveillance play important roles in lymphoma pathogenesis. Here we aimed to characterize immunological profiles of diffuse large B-cell lymphoma (DLBCL) and predict the outcome in response to immunochemotherapy. We profiled the expression of 730 immune-related genes in tumor tissues of 81 patients with DLBCL utilizing the Nanostring platform, and used multiplex immunohistochemistry to characterize T-cell phenotypes, including cytotoxic T cells (CD8, Granzyme B, OX40, Ki67), T-cell immune checkpoint (CD3, CD4, CD8, PD1, TIM3, LAG3), as well as regulatory T-cells and Th1 effector cells (CD3, CD4, FOXP3, TBET) in 188 patients. We observed a high degree of heterogeneity at the transcriptome level. Correlation matrix analysis identified gene expression signatures with highly correlating genes, the main cluster containing genes for cytolytic factors, immune checkpoint molecules, T cells and macrophages, together named a TME immune cell signature. Immunophenotyping of the distinct cell subsets revealed that a high proportion of immune checkpoint positive T cells translated to unfavorable survival. Together, our results demonstrate that the immunological profile of DLBCL TME is heterogeneous and clinically meaningful. This highlights the potential impact of T-cell immune checkpoint in regulating survival and resistance to immunochemotherapy. (Registered at clinicaltrials.gov identifiers: NCT01502982 and NCT01325194.)

Background/aim The chromosome translocation t(14;21)(q11;q22) was reported in four pediatric T-cell lymphoblastic leukemias and was shown to activate the OLIG2 gene. Materials and methods A pediatric T-cell lymphoblastic lymphoma was investigated using G-banding chromosome analysis, fluorescence in situ hybridization (FISH), and immunocytochemistry. Results The malignant cells carried a t(14;21)(q11;q22) aberration. The translocation moves the enhancer elements of TRA/TRD from band 14q11 to 21q22, a few thousands kbp downstream of OLIG1 and OLIG2, resulting in the production of both OLIG1 and OLIG2 proteins. Conclusion The translocation t(14;21)(q11;q22) occurs in some pediatric T-cell lymphoblastic malignancies. Activation of both OLIG1 and OLIG2 by t(14;21)(q11;q22) in T-lymphoblasts and the ensuing deregulation of thousands of genes could explain the highly malignant disease and resistance to treatment that has characterized this small group of patients.

In neuroblastoma, MYCN amplification and 11q-deletion are important, although incomplete, markers of high-risk disease. It is therefore relevant to characterize additional alterations that can function as prognostic and/or predictive markers. Using SNP-microarrays, a group of neuroblastoma patients showing amplification of one or multiple 12q loci was identified. Two loci containing CDK4 and MDM2 were commonly co-amplified, although amplification of either locus in the absence of the other was observed. Pharmacological inhibition of CDK4/6 with ribociclib or abemaciclib decreased proliferation in a broad set of neuroblastoma cell lines, including CDK4/MDM2-amplified, whereas MDM2 inhibition by Nutlin-3a was only effective in p53wild-type cells. Combined CDK4/MDM2 targeting had an additive effect in p53wild-type cell lines, while no or negative additive effect was observed in p53mutated cells. Most 12q-amplified primary tumors were of abdominal origin, including those of intrarenal origin initially suspected of being Wilms’ tumor. An atypical metastatic pattern was also observed with low degree of bone marrow involvement, favoring other sites such as the lungs. Here we present detailed biological data of an aggressive neuroblastoma subgroup hallmarked by 12q amplification and atypical clinical presentation for which our in vitro studies indicate that CDK4 and/or MDM2 inhibition also could be beneficial.

In neuroblastoma (NB), the ALK receptor tyrosine kinase can be constitutively activated through activating point mutations or genomic amplification. We studiedDiagnostic tumor samples were available from 1,092 HR-NBL1/SIOPEN patients to determineGenomicGenetic alterations of

T cells modified to express chimeric antigen receptor (CAR) targeting CD19 (CD19CAR) have produced remarkable clinical responses in patients with relapsed/refractory B-cell acute lymphoblastic leukemia. CD19CAR T-cell therapy has also demonstrated prominent effects in B-cell non-Hodgkin lymphoma (B-NHL) patients. However, a subset of patients who relapse after CD19CAR T-cell therapy have outgrowth of CD19− tumor cells. Hence, development of alternative CARs targeting other B-cell markers represents an unmet medical need for B-cell acute lymphoblastic leukemia and B-NHL. Here, we confirmed previous data by showing that, overall, B-NHL has high expression of CD37. A second-generation CD37CAR was designed, and its efficacy in T cells was compared with that of CD19CAR. In vitro assessment of cytotoxicity and T-cell function upon coculture of the CAR T cells with different target B-cell lymphoma cell lines demonstrated comparable efficacy between the 2 CARs. In an aggressive B-cell lymphoma xenograft model, CD37CAR T cells were as potent as CD19CAR T cells in controlling tumor growth. In a second xenograft model, using U2932 lymphoma cells containing a CD19− subpopulation, CD37CAR T cells efficiently controlled tumor growth and prolonged survival, whereas CD19CAR T cells had limited effect. We further show that, unlike CD19CAR, CD37CAR was not sensitive to antigen masking. Finally, CD37CAR reactivity was restricted to B-lineage cells. Collectively, our results demonstrated that CD37CAR T cells also can effectively eradicate B-cell lymphoma tumors when CD19 antigen expression is lost and support further clinical testing for patients with relapsed/refractory B-NHL.

Checkpoint blockade can reverse T-cell exhaustion and promote antitumor responses. Although blocking the PD-1 pathway has been successful in Hodgkin lymphoma, response rates have been modest in B-cell non-Hodgkin lymphoma (NHL). Coblockade of checkpoint receptors may therefore be necessary to optimize antitumor T-cell responses. Here, characterization of coinhibitory receptor expression in intratumoral T cells from different NHL types identified TIGIT and PD-1 as frequently expressed coinhibitory receptors. Tumors from NHL patients were enriched in CD8+ and CD4+ T effector memory cells that displayed high coexpression of TIGIT and PD-1, and coexpression of these checkpoint receptors identified T cells with reduced production of IFNγ, TNFα, and IL2. The suppressed cytokine production could be improved upon in vitro culture in the absence of ligands. Whereas PD-L1 was expressed by macrophages, the TIGIT ligands CD155 and CD112 were expressed by lymphoma cells in 39% and 50% of DLBCL cases and in some mantle cell lymphoma cases, as well as by endothelium and follicular dendritic cells in all NHLs investigated. Collectively, our results show that TIGIT and PD-1 mark dysfunctional T cells and suggest that TIGIT and PD-1 coblockade should be further explored to elicit potent antitumor responses in patients with NHL.

Background: To compare the Nordic and the Memorial Sloan-Kettering Cancer Center (MSKCC) chemotherapy protocols for Overall Survival (OS) and Progression-Free Survival (PFS) for intracranial primary CNS lymphoma (PCNSL). Methods: A prospective database at Oslo University Hospital of PCNSL was reviewed over a 12-year period (2003–2014). Results: Overall, 79 patients with PCNSL were identified, of whom 57 received chemotherapy. MSKCC with Rituximab (RTX) was used in 18 patients (32%) who had median OS of 46.3 months [9.8–131.9] and median PFS of 34.6 months [6.4–131.9]. The Nordic protocol was used in 14 patients (25%) who had median OS of 30.9 months [2.7–106.3] and PFS of 14.3 months [0.0–106.3]. The MSKCC was used without RTX in 25 patients (44%) who had OS of 15.2 months [0.7–136.5] and PFS of 12.0 months [0.0–117.0]. MSKCC with RTX had a significantly longer median OS (p

Separating and labeling each nuclear instance (instance-aware segmentation) is the key challenge in nuclear image segmentation. Deep Convolutional Neural Networks have been demonstrated to solve nuclear image segmentation tasks across different imaging modalities, but a systematic comparison on complex immunofluorescence images has not been performed. Deep learning based segmentation requires annotated datasets for training, but annotated fluorescence nuclear image datasets are rare and of limited size and complexity. In this work, we evaluate and compare the segmentation effectiveness of multiple deep learning architectures (U-Net, U-Net ResNet, Cellpose, Mask R-CNN, KG instance segmentation) and two conventional algorithms (Iterative h-min based watershed, Attributed relational graphs) on complex fluorescence nuclear images of various types. We propose and evaluate a novel strategy to create artificial images to extend the training set. Results show that instance-aware segmentation architectures and Cellpose outperform the U-Net architectures and conventional methods on complex images in terms of F1 scores, while the U-Net architectures achieve overall higher mean Dice scores. Training with additional artificially generated images improves recall and F1 scores for complex images, thereby leading to top F1 scores for three out of five sample preparation types. Mask R-CNN trained on artificial images achieves the overall highest F1 score on complex images of similar conditions to the training set images while Cellpose achieves the overall highest F1 score on complex images of new imaging conditions. We provide quantitative results demonstrating that images annotated by under-graduates are sufficient for training instance-aware segmentation architectures to efficiently segment complex fluorescence nuclear images.

Purpose: In neuroblastoma (NB), the ALK receptor tyrosine kinase can be constitutively activated through activating point mutations or genomic amplification. We studied ALK genetic alterations in high-risk (HR) patients on the HR-NBL1/SIOPEN trial to determine their frequency, correlation with clinical parameters, and prognostic impact. Materials and methods: Diagnostic tumor samples were available from 1,092 HR-NBL1/SIOPEN patients to determine ALK amplification status (n = 330), ALK mutational profile (n = 191), or both (n = 571). Results: Genomic ALK amplification (ALKa) was detected in 4.5% of cases (41 out of 901), all except one with MYCN amplification (MNA). ALKa was associated with a significantly poorer overall survival (OS) (5-year OS: ALKa [n = 41] 28% [95% CI, 15 to 42]; no-ALKa [n = 860] 51% [95% CI, 47 to 54], [P < .001]), particularly in cases with metastatic disease. ALK mutations (ALKm) were detected at a clonal level (> 20% mutated allele fraction) in 10% of cases (76 out of 762) and at a subclonal level (mutated allele fraction 0.1%-20%) in 3.9% of patients (30 out of 762), with a strong correlation between the presence of ALKm and MNA (P < .001). Among 571 cases with known ALKa and ALKm status, a statistically significant difference in OS was observed between cases with ALKa or clonal ALKm versus subclonal ALKm or no ALK alterations (5-year OS: ALKa [n = 19], 26% [95% CI, 10 to 47], clonal ALKm [n = 65] 33% [95% CI, 21 to 44], subclonal ALKm (n = 22) 48% [95% CI, 26 to 67], and no alteration [n = 465], 51% [95% CI, 46 to 55], respectively; P = .001). Importantly, in a multivariate model, involvement of more than one metastatic compartment (hazard ratio [HR], 2.87; P < .001), ALKa (HR, 2.38; P = .004), and clonal ALKm (HR, 1.77; P = .001) were independent predictors of poor outcome. Conclusion: Genetic alterations of ALK (clonal mutations and amplifications) in HR-NB are independent predictors of poorer survival. These data provide a rationale for integration of ALK inhibitors in upfront treatment of HR-NB with ALK alterations. Key Objective: High risk neuroblastoma (HR-NB) is one of the most difficult childhood cancers to cure. This study examined whether the presence of an ALK alteration (amplification or mutation) was associated with a poor prognosis in a large patient series treated on the prospective European high-risk neuroblastoma trial (HR-NBL1). Knowledge Generated: We found that ALK amplification or clonal mutation was associated with inferior prognosis in patients with HR-NB and both are independent prognostic variables on multivariate analysis. To our knowledge, this is the first study to report the highly prognostic significance of ALK amplification in HR-NB. Relevance: As ALK can be targeted therapeutically, this study convincingly argues for the introduction of ALK inhibitors for upfront management of patients with HR-NB with ALK aberrations. Importantly, the prognostic significance of ALK alterations included a subgroup of trial patients treated with the current standard of care for HR-NB including anti-GD2 immunotherapy. info:eu-repo/semantics/publishedVersion

Survival of patients with high-risk diffuse large B-cell lymphoma (DLBCL) is suboptimal, and the risk of central nervous system (CNS) progression is relatively high. We conducted a phase 2 trial in 139 patients aged 18 to 64 years who had primary DLBCL with an age-adjusted International Prognostic Index (aaIPI) score of 2 to 3 or site-specific risk factors for CNS recurrence. The goal was to assess whether a dose-dense immunochemotherapy with early systemic CNS prophylaxis improves the outcome and reduces the incidence of CNS events. Treatment consisted of 2 courses of high-dose methotrexate in combination with biweekly rituximab (R), cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP-14), followed by 4 courses of R-CHOP-14 with etoposide (R-CHOEP) and 1 course of high-dose cytarabine with R. In addition, liposomal cytarabine was administered intrathecally at courses 1, 3, and 5. Coprimary endpoints were failure-free survival and CNS progression rates. Thirty-six (26%) patients experienced treatment failure. Progression occurred in 23 (16%) patients, including three (2.2%) CNS events. At 5 years of median follow-up, failure-free survival, overall survival, and CNS progression rates were 74%, 83%, and 2.3%, respectively. Treatment reduced the risk of progression compared with our previous trial, in which systemic CNS prophylaxis was given after 6 courses of biweekly R-CHOEP (hazard ratio, 0.49; 95% CI, 0.31-0.77; P = .002) and overcame the adverse impact of an aaIPI score of 3 on survival. In addition, outcome of the patients with BCL2/MYC double-hit lymphomas was comparable to the patients without the rearrangements. The results are encouraging, with a low toxic death rate, low number of CNS events, and favorable survival rates. This trial was registered at www.clinicaltrials.gov as #NCT01325194.

Background: In precision cancer medicine, the challenge is to prioritize DNA driver events, account for resistance markers, and procure sufficient information for treatment that maintains patient safety. The MetAction project, exploring how tumor molecular vulnerabilities predict therapy response, first established the required workflow for DNA sequencing and data interpretation (2014–2015). Here, we employed it to identify molecularly matched therapy and recorded outcome in end-stage cancer (2016–2019). Material and methods: Metastatic tissue from 26 patients (16 colorectal cancer cases) was sequenced by the Oncomine assay. The study tumor boards interpreted called variants with respect to sensitivity or resistance to matched therapy and recommended single-agent or combination treatment if considered tolerable. The primary endpoint was the rate of progression-free survival 1.3-fold longer than for the most recent systemic therapy. The objective response rate and overall survival were secondary endpoints. Results: Both common and rare actionable alterations were identified. Thirteen patients were found eligible for therapy following review of tumor sensitivity and resistance variants and patient tolerability. The interventions were inhibitors of ALK/ROS1-, BRAF-, EGFR-, FGFR-, mTOR-, PARP-, or PD-1-mediated signaling for 2–3 cases each. Among 10 patients who received treatment until radiologic evaluation, 6 (46% of the eligible cases) met the primary endpoint. Four colorectal cancer patients (15% of the total study cohort) had objective response. The only serious adverse event was a transient colitis, which appeared in 1 of the 2 patients given PD-1 inhibitor with complete response. Apart from those two, overall survival was similar for patients who did and did not receive study treatment. Conclusions: The systematic MetAction approach may point forward to a refined framework for how to interpret the complexity of sensitivity versus resistance and patient safety that resides in tumor sequence data, for the possibly improved outcome of precision cancer medicine in future studies.

Key Points Diagnostic and relapse diffuse large B-cell lymphoma (DLBCL) biopsies reveal increased mutational burden/loss of heterozygosity in HLA-A. Serially sampled tumor biopsies provide insight into therapeutic targets and evolutionary divergence in relapsed/refractory DLBCL.

Patients with mantle cell lymphoma (MCL) generally have a dismal prognosis. Intensified induction treatment with rituximab and high dose cytarabine (R_HDAC), and consolidation with high-dose therapy with autologous stem cell support has resulted in 10-year overall survival (OS) higher than 60%. However, the clinical course varies. Diagnostic tools capable of stratifying patients include the MCL International Prognostic Index (MIPI), gene expression-based proliferation signature, Ki-67 proliferation index or tumour cell morphology. Here, we tested the performance of a newly developed Nanostring-based RNA expression-based proliferation assay (MCL35) on formalin-fixed paraffin-embedded tumour tissue from younger patients recruited in or treated according to Nordic MCL protocols compared to the prognosticators listed above. Seventy-four patients were included and the assay performed well in all cases except four, which had inadequate RNA quality. The patients were evenly distributed in the MCL35 low-, intermediate- and high-risk categories. MCL35 low- and intermediate- risk groups had overlapping progression-free survival (PFS), while patients in the high-risk category had significantly inferior PFS. Combining MCL35 with MIPI or the MIPI-C (MIPI with the addition of binary Ki67 score +/-30%) showed a better discrimination than either assessment alone. In conclusion, the MCL35 assay alone or combined with MIPI or MIPI-C scores can identify patients who still have a dismal outcome despite intensified treatment.

The purposes of this study are to study the impact of deep brain involvement on overall survival (OS) and progression-free survival (PFS) in intracranial primary CNS lymphoma (PCNSL), and to explore possible mechanisms for this impact using advanced MRI techniques. Seventy-nine patients with histologically verified PCNSL were identified from a prospective clinical database of patients treated at Oslo University Hospital between 2003 and 2014. Patients were treated per standard chemotherapeutic regimens (N = 57) or no chemotherapy (N = 22). Anatomical MRIs were available in all patients to assess tumor load and location based on contrast agent enhancement visible on T1-weighted images. Diffusion MRIs were available in 33 (42%) patients and perfusion MRI in 13 (16%) patients that received active treatment. Across all patients, OS and PFS were 16.4 and 9.8 months, respectively. In multivariate analysis, MRI-based deep brain involvement (80%) was the only negative significant factor of OS (OR = 14.2; P