Neurocognitive and endocrine dysfunction are potential complications of cranial irradiation. However, risk factors are poorly understood, impeding accurate prognostication and exploration of potential preventive interventions. The objective of this study was to evaluate the prognostic value of various vascular and genotypic risk factors for the development of radiation-related toxicities. This single-institution retrospective cohort study included patients with metastatic and malignant primary brain tumors who received cranial irradiation as part of their initial tumor-directed therapy. Demographic and treatment characteristics were collected, as well as putative vascular and genotypic risk factors. Univariate, multivariate, and machine-learning analyses were performed using five pre-specified measures of radiation-related toxicity. The primary outcome was change in mini-mental status exam (MMSE). Eighty patients (53 https://afranklin22.shinyapps.io/PRMMSERadiationRiskFactors/ ) Two pre-treatment laboratory values (elevated homocysteine and ApOE genotype) were strongly associated with post-radiation neurocognitive and endocrine dysfunction using a variety of domains. Our predictive algorithm can aid clinicians in stratifying baseline risk and should be validated in prospective trials and with additional metrics of ND.

Glioblastoma (GBM) is the most aggressive primary brain tumor with limited therapeutic options and extremely poor prognosis. Aberrant activation of the receptor tyrosine kinase MET drives tumor progression, therapeutic resistance, and reduced survival, particularly in the mesenchymal GBM subtype. Given its crucial role in GBM recurrence and progression, we investigated the mechanisms of resistance to MET inhibition using patient-derived glioma-initiating cells (GICs) and orthotopic xenograft mouse models. GICs were treated with the MET inhibitor crizotinib to elucidate the mechanism of adaptive resistance. Prolonged MET inhibition induced a senescent-like phenotype in GICs, associated with downregulation of BNIP3, a mitochondrial protein regulating mitophagy. We showed that BNIP3 downregulation led to activation of mTOR signaling, promoting cellular survival and adaptive resistance. Combining crizotinib with the mTOR inhibitor everolimus effectively suppressed mTOR activity, reduced cell viability, and induced mitochondrial alterations, apoptosis, and necroptosis. In orthotopic GBM xenograft models, combined MET and mTOR inhibition significantly prolonged survival compared with single-agent treatments. Notably, sequential treatment—crizotinib followed by everolimus—further enhanced therapeutic efficacy. These effects were achieved without significant weight loss, supporting tolerability of the treatment regimen. Our findings identify the BNIP3-mTOR axis as a critical mediator of resistance to MET inhibition and demonstrate that combined inhibition of MET and mTOR exhibits significant synergy against GBM.

Background:Glioblastoma (GBM) is the most common malignant primary brain tumor and is associated with a poor prognosis. Repurposed posaconazole (PCZ) has demonstrated efficacy in vitro by targeting key metabolic pathways. The aim of this study was to determine the neuropharmacological profile of PCZ in patients with GBM through a Phase 0 first-in-human trial for this indication (NCT04825275). Methods:We conducted an open-label, non-randomized, parallel-arm trial in participants with primary or recurrent GBM. Patients in the treatment arm received oral PCZ for 7-10 days presurgically to achieve steady-state concentrations. Microdialysis catheters were implanted to measure interstitial drug levels. Postoperative tissue and fluid samples were analyzed to evaluate PCZ pharmacokinetics (mass spectrometry) and pharmacodynamics, including vascular density, apoptosis, and metabolite concentrations (lactate and pyruvate) compared to controls. Results:The trial was closed early due to slow accrual, enrolling 2 participants in the PCZ arm and three in the control arm. No drug-related safety issues were observed. Although PCZ was undetectable in microdialysate fluid, it accumulated in tumor tissue, with concentrations showing an association with CD31-measured endothelial content. PCZ-treated tumors exhibited trends toward lower BCL2 expression indicative of increased apoptosis, and significantly lower pyruvate levels in the tumor periphery compared to controls. Plasma lactate levels in treated patients normalized to control levels over 24 h. Conclusions:Despite limited accrual, this trial provides the first clinical evidence that oral PCZ accumulates in human GBM tissue and shows preliminary signals of metabolic modulation. These findings provide biological rationale warranting further clinical investigation, potentially utilizing direct tissue sampling rather than microdialysis.

Abstract Background: Primary DLBCL of CNS (PCNSL) that relapses after or is refractory to high-dose methotrexate (HD-MTX) has poor long-term survival of 20%. PCNSL biology includes chronic active BCR signaling targeted by ibrutinib. Ibrutinib with temozolomide, etoposide, liposomal doxorubicin, dexamethasone, and rituximab (TEDDI-R) induces durable remissions in relapsed or refractory PCNSL but carries Aspergillus risk (Lionakis et al. Cancer Cell 2017). We added isavuconazole prophylaxis and report outcomes with escalating ibrutinib doses. Methods: Study conduct has been described (Roschewski et al. ASH 2024). Adults with relapsed or refractory PCNSL were enrolled in a phase 1 study with expansion. Prior BTKi, HIV+, and EBV+ were excluded. Isavuconazole started 3d before ibrutinib and continued throughout therapy. In phase 1, three dose levels of ibrutinib (280/420/560mg) were tested with TEDD-R. Two expansion cohorts received ibrutinib either continuously or on a fixed schedule (d1-10/cycle). Pts received up to 6 cycles without consolidation or maintenance. Tumors were molecularly classified by LymphGen. The primary objective was to determine the highest ibrutinib dose safely given with isavuconazole. Secondary objectives included overall response rate (ORR), PFS, and OS. Results: Thirty pts enrolled, including 10 in phase 1 and 10 each in the expansion cohorts. Median age was 63y (range 40-78), with 6 (20%) ≥70y. All had prior HD-MTX and 5 (17%) had prior stem cell transplant. Nineteen (63%) pts were refractory to HD-MTX. No DLTs were observed; ibrutinib 560mg was used in expansion. Neutropenia occurred in 6% (G3) and 40% (G4) of cycles, with febrile neutropenia in 11%. Thrombocytopenia occurred in 13% (G3) and 14% (G4) of cycles. ≥G3 infections occurred in 53% of pts, most common being UTI (20%), none fungal or opportunistic. Palmar plantar erythrodysesthesia occurred in 17 (57%) pts, managed with liposomal doxorubicin dose reduction. Notable ≥G3 non-hematologic toxicities included venous thromboembolism in 20%, and fatigue, syncope, mucositis, and hypokalemia in 13% pts each. G2 supraventricular arrhythmias occurred in 2 (7%) pts. Twenty-two pts died: 16 (73%) from progression, 2 (9%) from COVID, 1 (5%) from HBV reactivation, and 3 (14%) unknown. ORR was 90% (95% CI, 74-97), including complete response (CR) in 60% (95% CI, 42-75). CRs occurred across LymphGen subtypes: MCD 5 (71%), A53 2 (100%), BN2 1 (100%), and Other 2 (40%). No pt received consolidation. With a median f/u of 5.3y, the 2y PFS was 29% (95% CI, 14-46) and the 2y OS 47% (95% CI, 28-63). Ibrutinib schedule and HD-MTX refractoriness did not affect 2y PFS (P=0.94 and 0.68, respectively). Conclusions: Ibrutinib 560mg is safe across ages in TEDDI-R, and concurrent isavuconazole substantially reduces Aspergillus risk. TEDDI-R achieves high CR rate in relapsed or refractory PCNSL, including HD-MTX refractory disease. Durable remissions occur without consolidation. Cancer Therapy Evaluation Program sponsored this trial [NCT02203526] along with NCI’s Intramural Research Program. Citation Format: Rahul Lakhotia, Christopher Melani, Tatyana Gavrilova, Jagan R. Muppidi, James D. Phelan, Michail S. Lionakis, Kieron Dunleavy, Lode J. Swinnen, Matthias Holdhoff, Catherine Lai, Sami Ibrahimi, Michael Glantz, Jan Drappatz, John A. Butman, Stefania Pittaluga, Kim Johnson, Atekelt Tadese, Hyoyoung Choo-Wosoba, John D. Heiss, William D. Figg, Elaine S. Jaffe, S. Percy Ivy, Richard F . Little, Louis M. Staudt, Mark Roschewski, Wyndham H. Wilson. Final analysis of a Phase 1 study of ibrutinib dose-escalation in TEDDI-R with isavuconazole for relapsed or refractory primary DLBCL of the CNS [abstract]. In: Proceedings of the Fifth AACR International Meeting on Advances in Malignant Lymphoma: From Discovery to Clinical Impact; 2026 Jun 24-27; Philadelphia, PA. Philadelphia (PA): AACR; Blood Cancer Discov 2026;7(3_Suppl):Abstract nr A047.

Background MGMT promoter methylation status has been utilized as a predictor of response to temozolomide in patients with IDH-wildtype glioblastoma (GBM). Traditional methods of methylation status identification include methylation-specific polymerase chain reaction and pyrosequencing (PyroSeq). Though widely used, each method has disadvantages with respect to determining methylation cut-off values, tumor content required for evaluation, prognostic accuracy, and financial expense.Methods We have developed a method of MGMT classification using artificial intelligence and a large clinicogenomic database of 5841 GBM patients. We evaluate the performance of this novel classification strategy for predicting temozolomide treatment efficacy in comparison to PyroSeq techniques.Results MGMTai reliably predicted MGMT methylation status in GBM patients with available PyroSeq data. Comparative bucketing of methylation status based on MGMTai and PyroSeq yielded high sensitivity and positive predictive value concordance, though new MGMT promoter methylation (MGMTmet) status percentages were drawn according to PyroSeq methylation percent values. Overall survival with temozolomide (TMZ) treatment was comparable between PyroSeq and MGMTai; however, MGMTai by decile and MGMTai stratified into 3 scoring buckets yielded more distinct and predictive survival patterns with increasing MGMTai score compared to PyroSeq.Conclusion Implementation of an AI-based molecular classification system, MGMTai, can better describe MGMT methylation status in comparison to the traditional PyroSeq method. Patient MGMTmet status by MGMTai scoring drew more distinct survival curves and predicted TMZ-treated GBM patient survival more efficiently, less expensively, and with better precision and reproducibility than PyroSeq.

Background:Glioblastoma (GBM) lacks effective therapies for recurrent disease. Unlike cancers with successful fusion-targeted treatments (eg BCR-ABL1 in CML), the incidence and therapeutic potential of gene fusions in GBM remain unclear. We analyzed a large genomic database to define fusion frequency and molecular associations. Methods:4800 IDH-wildtype GBM samples (WHO 2021) underwent NextGen DNA sequencing (592-gene panel/whole exome) and Whole Transcriptome Sequencing for fusions at Caris Life Sciences. Fisher-Exact/Chi-Square tests, adjusted by Benjamini-Hochberg (q < 0.05), assessed significance. Results:Pathogenic fusions occurred in 428 (8.9%) samples, primarily FGFR3 (37%, n = 159; FGFR3: TACC3, n = 134), MET (21%, n = 92), and EGFR (20%, n = 87). Pathogenic or likely pathogenic fusions included NTRK2 (n = 27), PDGFRA (n = 23), ROS1 (n = 14), and BRAF (n = 10). Fusion-positive tumors had higher MET (7.5% vs. 0.7%), FGFR3 (5% vs. 0.2%), CDK4 (17% vs. 11%), and MDM2 (12% vs. 7.5%) amplifications, but lower EGFR mutations (6.1% vs. 18%), amplifications (6.1% vs. 18%), and EGFRvIII (11.9% vs. 22.5%) (all q < 0.05). Median survival was 16.6 months (fusion-positive) vs. 15.5 months (fusion-negative) (P = 0.043). Tyrosine kinase inhibitor (TKI)-treated fusion-positive patients (n = 37) showed no significant survival benefit (18.4 vs. 16.5 months, P = .971). Conclusions:Approximately 9% of GBMs harbor targetable fusions, with five genes (FGFR3, MET, EGFR, NTRK2, PDGFRA) comprising 8%. These findings support multi-arm clinical trials to evaluate targeted therapies, potentially improving outcomes for molecularly defined GBM subgroups.

BACKGROUND AND OBJECTIVES: Intraventricular chemotherapy administered through an Ommaya reservoir (OmR) constitutes an integral part of therapy for patients with leptomeningeal metastasis. Unfortunately, OmR infections remain a frequent, costly, morbid, and occasionally fatal complication, limiting the benefit of this approach. We evaluate the efficacy, cost savings, and toxicity of vancomycin coadministered with intraventricular chemotherapy for the prevention of OmR-associated bacterial meningitis. METHODS: This was a cohort study comparing a treatment group treated from May 1, 2021, to April 30, 2022, and a retrospective control group treated from 2016 to 2021. Patients were included if they had a diagnosis of leptomeningeal metastasis and subsequent placement of an OmR, followed by delivery of planned intraventricular chemotherapy. Patients in the treatment group received a 10-mg dose of intraventricular vancomycin in addition to their standard chemotherapy regimen at each treatment. We compared this group to a retrospective cohort from the preceding 5 years who did not receive intraventricular vancomycin, evaluating the rate of infection, adverse events, and associated treatment costs between groups. RESULTS: The infection rate was 0% (95% CI 0%-6.75%) among the 63 patients and 0% (0%-0.76%) in the 501 consecutive treatments administered over the 12-month study period, compared with 10.25% (7.39%-14.0%) among the 322 patients and 1.71% (1.22%-2.39%) in 1932 treatments over the preceding 5 years. The absolute risk reduction was 10.3% (3.83-14.04), P = .0028. The number needed to treat was 10 (7-26). Cost per vancomycin dose was $10.00 ($5010 over 12 months). The cost of nonsurgical treatment of 1 OmR-associated infection is $88 372, which translates into $618 604 for the estimated 7 patients over 12 months who developed an OmR-associated infection. No treatment-associated toxicity was observed. CONCLUSION: Prophylactic intraventricular vancomycin eliminated OmR-associated infections without added toxicity and with dramatic cost savings.

Background Central nervous system lymphoma (CNSL), is a rare subtype of non-Hodgkin lymphoma, primarily affecting the brain and spinal cord. Most therapeutic systemic agents have limited penetration of the blood-brain and blood-cerebrospinal fluid (CSF) barrier, with the latter potentially promoting a treatment "sanctuary" for cancer cells. Evaluation of occult disease, particularly in the CSF, is challenging. In limited clinical experience, the addition of multiagent intraventricular chemotherapy (MAIVC), delivered through intracranially implanted CSF reservoirs, to systemic therapy has demonstrated encouraging outcomes, enhancing both progression-free survival and overall survival. However, given the potential morbidity associated with MAIVC, identification of minimally invasive biomarkers for guiding patient selection and management is necessary. Leveraging the longitudinal, large volume of CSF, the objective of this study was to identify CSF-based proteomic biomarkers that can serve as reliable indicators of CSF clearance in response to MAIVC and CNSL treatment outcome.Methods One hundred fifteen CSF samples from 59 CNSL patients receiving MAIVC were profiled using a high-throughput protocol coupled with mass-spectrometry that only requires 30 mu L of CSF.Results More than 2000 unique proteins were detected using shotgun proteomics. Cerebrospinal fluid proteomics revealed key proteins (SGCE, LCP1, AGRN, OLFML3, and HRSP12) distinguishing early from never responders to MAIVC, with area under the receiver operating characteristic (AUROC) 0.86 (95% CI: 0.696-1). By integrating tumor volume from brain MRI scans with proteomic data, we identified potential intraventricular tumor burden markers for CNSL management, in particular LCP1.Conclusions The study identified CSF-based proteomic biomarkers, particularly LCP1, that can classify MAIVC response and indicate tumor burden in CNSL patients. Central nervous system lymphoma (CNSL) is a rare cancer that affects the brain and spinal cord. Treatment of CNSL is difficult because most chemotherapies are not able to fully reach the fluid that surrounds the brain and spinal cord. This study investigated whether specific proteins in cerebrospinal fluid could predict patient response to multiagent intraventricular chemotherapy (MAIVC), a treatment delivered directly into the brain. By analyzing cerebrospinal fluid samples from 59 CNSL patients, researchers identified several key proteins that could distinguish patients who would respond early to treatment from those who would not respond. In particular, a protein called LCP1 was found to be a potential marker of treatment response and tumor burden. These findings may help doctors better determine which patients will benefit most from this specialized treatment approach.

Abstract Background Central nervous system lymphoma (CNSL), is a rare subtype of non-Hodgkin lymphoma, primarily affecting the brain and spinal cord. Most therapeutic systemic agents have limited penetration of the blood-brain and blood-cerebrospinal fluid barrier, with the latter potentially promoting a treatment “sanctuary” for cancer cells. Evaluation of occult disease, particularly in the cerebrospinal fluid (CSF), is challenging. In limited clinical experience the addition of multi-agent intraventricular chemotherapy (MAIVC), delivered through intracranially-implanted CSF reservoirs, to systemic therapy has demonstrated encouraging outcomes, enhancing both progression-free survival and overall survival. However, given the potential morbidity associated with MAIVC, identification of minimally-invasive biomarkers for guiding patient selection and management are necessary. Leveraging the longitudinal, large volume of CSF, the objective of this study was to identify CSF-based proteomic biomarkers that can serve as reliable indicators of CSF clearance in response to MAIVC and CNSL treatment outcome. Methods 115 CSF samples from 59 CNSL patients receiving MAIVC were profiled using a high-throughput protocol coupled with mass-spectrometry that only requires 30 μL of CSF. Results More than 2000 unique proteins were detected using shotgun proteomics. CSF proteomics revealed key proteins (SGCE, LCP1, AGRN, OLFML3, HRSP12) distinguishing early from never responders to MAIVC, with AUROC 0.86 (95% CI: 0.696-1). By integrating tumor volume from brain MRI scans with proteomic data, we identified potential intraventricular tumor burden markers for CNSL management, in particular LCP1. Conclusions The study identified CSF-based proteomic biomarkers, particularly LCP1, that can classify MAIVC response and indicate tumor burden in CNSL patients.

Patients with radiographically detectable lesions in their brain or other symptoms compatible with brain tumors pose challenges for diagnosis. The only definitive way to diagnose such patients is through brain biopsy, an invasive and dangerous procedure. In this study, we present a new workflow termed "CSF-BAM" that simultaneously identifies B-cell or T-cell receptor sequences, aneuploidy, and mutations using amplification of both strands of the DNA from cerebrospinal fluid (CSF) samples. We applied CSF-BAM to a validation set of 209 samples from patients with brain cancers. Among the 129 samples from patients with the most common aggressive cancer types, the sensitivity of detection was 81%. None of 30 CSF-BAM assays were positive in CSF samples from patients without brain cancers (100% specificity). CSF-BAM provides an integrated approach to identify neoplasia in the central nervous system, provides information about the genetics and immune environment, and has the potential to inform patient management. SIGNIFICANCE:There is a paucity of technologies beyond surgical biopsy that can accurately diagnose central nervous system neoplasms. We developed a novel, sensitive, and highly specific assay that can detect brain cancers by comprehensively identifying somatic mutations, chromosomal copy-number changes, and adaptive immunoreceptor repertoires from samples of CSF. See related commentary by Weiss, p. 1976.

Glioblastoma (GBM) is an aggressive primary brain malignancy with poor prognosis due to rapid progression, extensive invasiveness, and intrinsic resistance to standard therapies. Aberrant activation of receptor tyrosine kinases (RTKs), particularly MET, drives tumor proliferation, invasion, and therapy resistance. Here, we show that MET inhibition with crizotinib induces senescence and mitochondrial dysfunction in glioma-initiating cells (GICs), in part via downregulation of the mitochondrial protein BNIP3. However, BNIP3 downregulation activates mTOR signaling, enabling adaptive resistance. Targeting mTOR with everolimus in combination with crizotinib synergistically enhances anti-tumor effects, inducing apoptosis, senescence, and necroptosis, and significantly reducing cell viability and sphere-forming capacity. In orthotopic GBM xenograft models, this combination, particularly in a sequential regimen, markedly prolongs survival without overt toxicity. Our findings identify a BNIP3-mTOR signaling axis as a critical mediator of resistance to MET inhibition and provide a mechanistic rationale for combined MET and mTOR targeting as a promising therapeutic strategy in GBM.

2066 Background: The oncometabolite 2-hydroxyglutarate (2HG) produced by IDH1/2 mt in HGG has profound effects on numerous pathways including DNA damage repair (DDR). We investigated the prognostic effect of DDR mt in IDH mutant vs. wild type (wt) tumors in a large cohort using a real-world database. Methods: A total of 4894 HGG tumors tested at Caris Life Sciences (Phoenix, AZ) with NextGen sequencing of DNA (592-gene panel or whole exome sequencing) were included in the study. DDR alteration was defined as a pathogenic mutation in one of > 20 DDR genes. Patient survival was obtained by insurance claims data and calculated from the initiation of tissue collection (rwOS). Cox proportional hazards model was used to calculate hazard ratios (HR) and log-rank tests to calculate p values, which were adjusted for multiple comparisons. Significance was set at p<0.05. Results: In the 1121 HGG carrying either IDH1 or 2 mutations, 100 carried a DDR mutation (8.9%) . When comparing DDR mutant (mt) vs. wild type (wt), no difference was seen in patient age (median 39 vs. 38 yrs; p = 0.8); gender (female 45% vs. 42%, p = 0.9), race or ethnicity (p > 0.1). The most frequent mutations were seen in MSH6 (24% of the DDR mt), ATM (18%), MLH1 (15%), MSH2 (13%), MSH3 (10%) and BRCA2 (10%). When comparing the rwOS of DDR mt vs. wt, a significantly shorter survival was seen (24m vs. 51m, HR = 1.87, 95% CI [1.41-2.48], p < 0.001); the effect persisted in the subset of tumors collected prior to temozolomide treatment (26m vs. 64m, HR = 1.92 [1.35-2.74], p < 0.001). In contrast, in IDH wt tumors, patients with (N = 223) or without DDR mutation (N = 3550) showed similar survival (17.5m vs. 20.6m, p = 0.1). In the IDH mutant cohort, DDR mt was associated with an increased tumor mutational burden (TMB) compared to DDR wt tumors (median = 6 vs. 4 mutations/mb, by Wilcoxon). Multivariate analysis within the IDH mutant tumors indicated that both TMB and DDR status were independently associated with poorer rwOS, with TMB showing an adjusted HR of 1.01 per unit increase (p = 0.005) and DDR status with an adjusted HR of 1.59 (p = 0.028). Conclusions: In a large real-world database, we demonstrate IDH mt HGG with a DDR mutation exhibit significantly poorer survival compared to DDR wt. This is not seen in IDH wt, where survivals of the two groups are similar. These results stand in sharp contrast to reported prognostic effect of DDR mutation in many other solid tumors. The data suggest that DDR mutations in the context of 2HG accumulation in IDH mt HGG may be an indicator of profound genomic instability that confers severe negative impact on patient survival. Clinicians managing high-grade gliomas should consider the presence of DDR mutations in IDH mutant patients as a poor prognostic category in this overall favorable prognostic group and consider therapeutic approaches accordingly.

Supplementary Table S1: Cohorts for CSF-BAM components. Supplementary Table S2A: SafeBSeqS primer sequences. Supplementary Table S2B: SafeTseqS primer sequences. Supplementary Table S2C: Aneuploidy analysis primer sequences. Supplementary Table S2D: Mutation analysis primer sequences. Genomic coordinates refer to hg19. Supplementary Table S3: Summary of SafeBSeqS analysis in WBC control samples. Supplementary Table S4: Summary of CSF-BAM SafeBSeqS analysis in CSF. Supplementary Table S5: Summary of SafeTSeqS analysis in WBC control samples. Supplementary Table S6: Summary of CSF-BAM SafeTSeqS analysis in CSF. Supplementary Table S7: SafeBSeqS and SafeTSeqS analysis for CSF validation set 2. Supplementary Table S8: Summary of aneuploidy analysis in control samples. Supplementary Table S9: Summary of CSF-BAM aneuploidy analysis in CSF samples. Supplementary Table S10: Reproducibility of aneuploidy analysis. Supplementary Table S11: Comparison of aneuploidy analysis with CSF-BAM and Real-CSF. Supplementary Table S12: CSF-BAM mutation training set in non-cancer CSF samples. Supplementary Table S13: CSF-BAM mutation analysis in CSF samples. Supplementary Table S14: Demographic characteristics by sample. Supplementary Table S15: Diagnostic categories by sample. Supplementary Table S16: CSF-BAM results summary. Supplementary Table S17: CSF-BAM detection based on CSF reservoir abutment by sample. All samples with available clinical data are included. P-values were calculated by Fisher’s exact test. Supplementary Table S18A: TCR Clusters obtained via CSF-BAM. Supplementary Table S18B: TCR specificity annotation. Supplementary Table S19: CSF samples from patients with multiple sclerosis analyzed with SafeTSeqS and SafeBSeqS.

Isocitrate dehydrogenase mutant (IDHm) low-grade gliomas have a slow growing phase, but eventually become aggressive tumors. To delay long-term toxic effects of chemoradiation, low-risk patients are monitored after surgery. Traditionally, patients >40 were considered high-risk, however, this remains controversial. Recent promising results with the IDHm-inhibitor vorasidenib sparking debate, challenging age-based risk stratification. We evaluated survival relative to age and molecular data. 598 IDHm astrocytoma grades 2-3 and 288 IDHm oligodendroglioma grades 2-3 were analyzed by NGS and WTS at Caris Life Sciences (Phoenix, AZ). Samples were stratified by age at diagnosis (12-26y, 27-40y, 41-60y, and >60y). Overall survival was obtained from insurance claims data and analyzed using Kaplan-Meier and Cox proportional hazards models. Comparisons in survival were made between 27-40y and 41-60y given larger sample size. Multivariate regression analysis included radiotherapy, temozolomide, and mutation status as covariates. IDHm astrocytomas age distribution was 12-26y, n=74 (12.4%); 27-40y, n=271 (45.3%); 41-60y, n=205 (34.3%); and >60y, n=48 (8.0%). Univariate analysis showed that 27-40y patients had shorter survival (HR=1.63, 95% CI:1.07–2.50, p=0.022). However, after multivariate analysis, age was not associated with survival (HR =1.02, 95% CI:0.74-1.4, p=0.912). In contrast, TP53 (HR=4.0, 95%CI:1.43-11.24, p=0.008– mutation rate=95.4%) and TERT-promoter (HR=10.36, 95% CI:4.05-26.45, p<0.0001– mutation rate=9.0%) mutations were independently associated with poorer survival. IDHm oligodendrogliomas age distribution was 12-26y, n=18 (5.5%); 27-40y, n=76 (23.2%); 41-60y, n=137 (41.8%); and >60y, n=57 (17.4%). Univariate and multi-variate analysis did not show any association between age and survival (HR =0.94, 95% CI:0.48-1.83, p=0.912, and HR =1.33, 95% CI:0.81-2.19, p=0.248, respectively). However, KRAS mutations were independently associated with poorer survival (HR=4.36, 95% CI:1.12-16.92, p=0.033- mutation rate=3%). While age (27-40y vs. 41-60y) was not associated with survival, genetic alterations such as KRAS (oligodendroglioma), TP53 and TERT mutations (astrocytoma) were independently associated with poorer survival.

Supplementary Figure S1. SafeBSeqS amplification. Supplementary Figure S2. SafeTSeqS amplification. Supplementary Figure S3. Aneuploidy reproducibility from two independent aliquots and libraries. Supplementary Figure S4. Metrics for targeted mutation panel. Supplementary Figure S5. Correlation of predicted aneuploidy neoplastic content to mutation neoplastic content. Supplementary Figure S6. A, TCR UIDs and B, BCR UIDS recovered for each CSF sample. Supplementary Figure S7. A, TCR clonality and B, BCR clonality for evaluable samples with total UIDs ≥20. Supplementary Figure S8. BCR clonality ROC and BCR IGHV4-34 gene segment usage. Supplementary Figure S9. Case reports demonstrating CSF-BAM clinical applicability.

Despite 50 years of intensive clinical and laboratory research, multiple single-arm and retrospective studies, and 8 randomized controlled trials (RCTs), the prognosis for patients with leptomeningeal metastases (LM) remains appalling. More recently, potentially paradigm-changing insights into the cellular and molecular anatomy of the leptomeningeal space, and the genomic underpinnings of tumors metastasizing to the cerebrospinal fluid (CSF), coupled with novel therapeutic strategies have resulted in a renaissance of interest in this disease. To maximize the impact of further research, we reviewed the patient-level data from all 8 RCTs. Using a PRISMA-compliant search with multiple databases, 8 qualifying RCTs were identified. Study-specific information regarding patient demographics, enrollment, treatment, and response outcomes were recorded using structured data extraction tools. De-identified patient-level data was acquired from study sponsors or was extracted from Kaplan Meier curves. Summary statistics were calculated using a random effects model and inverse variance technique. Eight randomized phase III studies conducted between 1987 and 2019 were identified. These studies accrued 429 patients (mean 53.5/study) from 161 institutions. Average study duration was 4.8 years; MTX, Ara-C, Thiotepa, DepoCyt, and systemic therapy were investigated. Five of 8 studies included Depocyt (no longer available). Median overall survival was 31.5 weeks. Only 4 studies provided class I evidence for OS and only 2 for PFS. All other studies and endpoints provided class III or IV evidence. Challenges to accrual, reasons for loss to follow-up, specific patient characteristics, multiple outcomes, and outcomes for specific patient subsets were all analyzed. The design and conduct of RCTs for leptomeningeal metastases, accrual, retention, treatment selection, and outcomes have been disappointing. The nature and reasons for these failures should help improve study efficiency and reliability for future trials.

Background:Intra-cerebrospinal fluid (CSF) chemotherapy delivered through an Ommaya reservoir is a treatment for certain patients with leptomeningeal metastasis (LM). Numerous administration strategies are advocated in the literature and are used in practice. We evaluated differences in drug delivery associated with the three most used techniques. Methods:Consecutive patients with newly diagnosed LM underwent conventional CSF flow studies using indium-111-DTPA (n = 104). Three different administration techniques and two different radionuclide volumes were used. Techniques evaluated were administration of the radionuclide followed by repeated compression of the Ommaya bulb, administration mixed with an equal volume of CSF followed by compression, and barbotage with CSF withdrawn prior to chemotherapy injection. Residual radionuclide counts were measured in the injection syringe, butterfly needle/stopcock apparatus, and reservoir bulb after radionuclide administration. In addition, U.S. neuro-oncologists were surveyed regarding their intra-CSF chemotherapy administration techniques. Results:For the 3 mL injection volume, 53.5%, 28.0%, and 3.14% of the radionuclide never reached the patient using the bulb compression, CSF-drug mixing followed by compression, and barbotage techniques respectively (P < .00001). For the 5 mL injection volume, the corresponding percentages were 23.3%, 18.7%, and 3.0% (P < .0001). The survey of U.S. neuro-oncologists revealed that a majority (56%) do not use a barbotage procedure for chemotherapy administration. Conclusions:A large proportion of radionuclide injected into Ommaya reservoirs never reaches the patient unless a barbotage technique is used. This finding likely applies to intraventricular chemotherapy injections as well and may represent one easily remediable reason for the poor efficacy of intraventricular chemotherapy in patients with LM.

2058 Background: Prognosis for mutant isocitrate dehydrogenase (mIDH) gliomas is influenced by tumor type, size, neurologic deficits, and age. Traditionally, patients over 45 are considered high-risk, prompting consideration of early chemoradiation. Recent promising results with the mIDH inhibitor vorasidenib challenge traditional age-based risk stratification, sparking debate over its role in treatment decisions. We evaluated survival relative to age and molecular data obtained from next-generation sequencing (NGS). Methods: Tumor specimens from 598 mIDH gliomas were analyzed using NGS and WTS at Caris Life Sciences (Phoenix, AZ). Samples were stratified by age at diagnosis into four groups: 12-26y, 27-40y, 41-60y, and > 60y. Real-world overall survival (calculated from initial diagnosis to last contact) was obtained from insurance claims data and analyzed using Kaplan-Meier and Cox proportional hazards models. Covariates in the multivariate regression analysis included radiation treatment, temozolomide treatment, and mutation status of different biomarkers. Results: In mIDH astrocytoma group, age distribution was 12-26y, n = 74 (12.4%); 27-40y, n = 271 (45.3%); 41-60y, n = 205 (34.3%); and > 60y, n = 48 (8.0%). In mIDH oligodendroglioma group, age distribution was 12-26y, n = 18 (5.5%); 27-40y, n = 76 (23.2%); 41-60y, n = 137 (41.8%); and > 60y, n = 57 (17.4%). For each subtype, comparisons in survival were made between patients 27-40y vs. 41-60y given larger sample size, and patients with temozolomide treatment before biopsy were excluded (about 10%). Univariate analysis showed that 27-40y patients had shorter survival in astrocytoma (HR = 1.63, 95% CI: 1.07 – 2.50, p = 0.022). However, after adjusting for confounding factors in multivariate analysis, age was not associated with survival. In contrast, TP53 (HR = 4.0, 95% CI: 1.43-11.24, p = 0.008 – mutation rate = 95.4%) and TERT-promoter (HR = 10.36, 95% CI: 4.05-26.45, p < 0.0001 – mutation rate = 9.0%) mutations were independently associated with poorer survival in astrocytoma patients. Univariate analysis showed that age was not associated with survival in oligodendroglioma (HR = 1.07, 95% CI: 0.79-3.65, p = 0.168). KRAS mutations were independently associated with poorer survival in oligodendroglioma patients (HR = 4.36, 95% CI: 1.12-16.92, p = 0.033 - mutation rate = 3%). Conclusions: In this enriched dataset of mIDH low grade glioma patients, which included NGS, age did not contribute to survival differences when comparing patients between 27-40 years with those aged 41-60 years. Rather, selected genetic alterations such as KRAS for oligodendroglioma and TP53 and TERT mutations for astrocytoma were associated with poorer survival. The results suggest that NGS, rather than age, may drive prognosis for mIDH glioma patients.