Resection and radiotherapy for intracranial ependymoma: a multiinstitutional 50-year experience

OBJECTIVE Maximal safe resection is the standard-of-care treatment for adults with intracranial ependymoma. The value of adjuvant radiotherapy remains unclear as these tumors are rare and current data are limited to a few retrospective cohort studies. In this study, the authors assembled a cohort of patients across multiple international institutions to assess the utility of adjuvant radiotherapy in this patient population. METHODS Adults with intracranial ependymoma managed surgically at the University Health Network in Toronto, Canada, the University of Oklahoma Health Sciences Center in Oklahoma City, Oklahoma, and The Ottawa Hospital in Ottawa, Canada, were included in this study. The primary end points were progression-free survival (PFS) and overall survival (OS). Clinicopathological variables were assessed in univariate and multivariate Cox proportional hazard models for prognostic significance of PFS and OS. RESULTS A total of 122 patients diagnosed between 1968 and 2019 were identified for inclusion. The majority of patients had grade II ependymomas on histopathology (78%) that were infratentorially located (71%), underwent gross-total (GTR) or near-total resection (NTR; 55%), and were treated with adjuvant radiotherapy (67%). A volumetric analysis of the extent of resection in 49 patients with available tumor volume data supported the accuracy of the categorical GTR, NTR, and subtotal resection (STR) groups utilized. Independent statistically significant predictors of poorer PFS in the multivariate analysis included STR or biopsy (vs GTR/NTR; HR 5.4, 95% confidence interval [CI] 2.4–11.0, p < 0.0001) and not receiving adjuvant radiotherapy; cranial (HR 0.5, 95% CI 0.2–1.1) and craniospinal (HR 0.2, 95% CI 0.04–0.5) adjuvant radiotherapy regimens improved PFS (p = 0.0147). Predictors of poorer OS in the multivariate analysis were grade III histopathology (vs grade II: HR 5.7, 95% CI 1.6–20.2, p = 0.0064) and undergoing a biopsy/STR (vs GTR/NTR: HR 9.8, 95% CI 3.2–30.1, p = 0.0001). CONCLUSIONS The results of this 50-year experience in treating adult intracranial ependymomas confirm an important role for maximal safe resection (ideally GTR or NTR) and demonstrate that adjuvant radiotherapy improves PFS. This work will guide future studies as testing for molecular ependymoma alterations become incorporated into routine clinical practice.

P.130 Clinical prognostic factors in adult intracranial ependymoma patients – A fifty year multi-institutional experience

Background: Standard of care treatment for adult intracranial ependymoma patients includes maximal safe surgical resection, while the role for adjuvant radiotherapy remains unclear with existing data from small retrospective series’. Accordingly, we built a multi-institutional cohort to assess the prognostic value of adjuvant radiotherapy and other clinical factors in these patients. Methods: Patients managed for adult intracranial ependymomas from 1968 onwards within the University Health Network in Toronto, The University of Oklahoma Health Sciences Center, and The Ottawa Hospital were identified. Multivariate models assessing the prognostic value of clinical factors were built using variables with known prognostic value and/or significance in univariate Cox models. Results: Of 122 ependymomas, 71% were infratentorial, 78% grade two, 55% gross/near-totally resected (GTR/NTR), and 65% treated with adjuvant radiotherapy. Multivariate analyses identified GTR/NTR (vs. subtotal resection; HR=0.2, 95%CI=0.1–0.4, p<0.0001) and cranial (HR=0.5, 95%CI=0.2–1.1) or craniospinal (HR=0.2, 95%CI=0.04–0.5) radiotherapy receipt (p=0.01) as independent statistically significant predictors of more favourable PFS. Grade II pathology (vs. grade III; HR=0.2, 95%CI=0.05–0.6, p=0.006) and GTR/NTR (vs. subtotal resection; HR=0.1, 95%CI=0.03–0.3, p=0.0001) were independent statistically significant predictors of better OS. Conclusions: This work confirms the importance of maximal safe resection for adult intracranial ependymomas and establishes that adjuvant radiotherapy improves progression-free survival in these patients.

Human Milk Oligosaccharides Are Stable Over One-Week of Lactation and Over Six-Hours Following a Standardized Meal

Objectives Our prior studies revealed that human milk oligosaccharides (HMOs) have numerous health benefits for nursing infants, and can also change dynamically over 24 months of lactation. However, the extent to which HMOs vary over days of time and also in response to acute factors such as maternal diet has not been established. The aims of this study were to determine the stability of HMO concentrations over 7-days, and in response to a standardized meal and sugar-sweetened beverage (SSB) over 6-hours. Methods Lactating mothers were enrolled at 6 weeks. Participants received in-person instructions and materials to complete procedures at home. In the first experiment (N = 11), mothers pumped a fasted sample of human milk at 7:00 AM for 7-consecutive days. In the second experiment (N = 33), mothers pumped a fasted sample of human milk at 6:00 AM. Mothers then consumed a standardized meal plus SSB provided by the study team. Thereafter, mothers pumped a non-fasted sample every hour for 6-consecutive hours. Samples were analyzed for nineteen HMOs. One-factor repeated-measures ANOVA was used to test changes in HMOs. Results There were no significant changes in any of the HMOs over 7-consecutive days, including 2’-fucosyllactose (2’FL) (interaction, P = 0.99), lacto-N-fucopentaose I (LNFPI) (interaction, P = 0.36), disialyl-lacto-N-tetraose (DSLNT) (interaction, P = 0.99), and lacto-N-neotetraose (LNnT) (interaction, P = 0.90). Further, there were no significant changes in any of the HMOs following a standardized meal plus SSB, as 2’FL (interaction, P = 0.12), LNFPI (interaction, P = 0.95), DSLNT (interaction, P = 0.66), and LNnT (interaction, P = 0.67) did not change in response to the diet intervention over 6-consecutive hours. Conclusions Our findings reveal that HMOs are stable over one-week during early lactation, and are not acutely affected by maternal food and SSB consumption. Funding Sources National Institutes of Health (HD098288, HD080444, and DK056341); Harold Hamm Diabetes Center at the University of Oklahoma Health Sciences Center.

622 PD-L1 is a potential predictive biomarker for response to RM-1929 treatment in recurrent head and neck squamous cell carcinoma patients

BackgroundRM-1929 is an antibody-dye conjugate comprised of cetuximab covalently linked to the photoactivable dye, IRDye® 700DX (IR700). After systemic infusion of RM-1929, illumination of the tumor with 690 nm non-thermal red light activates the drug and results in targeted and rapid tumor necrosis. Previous preclinical data have shown that RM-1929 treatment triggers immunogenic cell death and activates the innate and adaptive immune response. A retrospective analysis of PD-L1 expression from the phase I/IIa clinical trial in patients with recurrent head and neck squamous cell carcinoma (rHNSCC) (NCT02422979) was conducted. The analysis explored correlations of PD-L1 expression, including combined proportion score (CPS) and tumor proportion score (TPS), with clinical outcomes such as response rate and overall survival.MethodsPD-L1 expression prior to RM-1929 treatment was assessed by immunohistochemistry in 18 out of 30 patients enrolled in Part II of the trial, based on sample availability. PD-L1 expression was evaluated using TPS and CPS. Responders were defined as patients that achieved complete response or partial response, and non-responders had either stable disease or progressive disease. Overall survival (OS) was analyzed using the Kaplan-Meier method.ResultsResponders (n=10) had a TPS of 4.3±2.4 (mean±SEM), which was substantially lower than in non-responders (n=8) with a TPS of 39.4±11.8. Similarly, CPS was lower in responders (8.6±3.6) compared to non-responders (50.0±13.5). The best target response rate for all patients included in this analysis was 56%. Patients with CPS=40 had a response rate of 76.9% (n=13) compared to 0% in patients with CPS>40 (n=5). This suggests that a CPS cut-off of =40 led to enrichment of the best target response rate. The median OS of patients with CPS=40 (13.0±0.8 months) was also higher than in patients with CPS>40 (3.1±0.8 months) and in all patients (12.0±2.9 months).ConclusionsThese results suggest that rHNSCC patients with lower PD-L1 expression levels may be more responsive to RM-1929 treatment and CPS/TPS could potentially be predictive biomarkers in identifying patients with a higher probability of benefiting from this treatment. Given the limited number of patients in this analysis, additional clinical trials will be needed to validate PD-L1 expression as an effective predictive biomarker for RM-1929 treatment.AcknowledgementsThe authors would like to thank all patients and their families for their participation in this trial. The authors would also like to thank the following investigators for the contribution of samples included in this trial analysis: Dr. David Cognetti (Thomas Jefferson University Hospital), Dr. Ann M Gillenwater (University of Texas MD Anderson Cancer Center), Dr. Mary Jo Fidler (Rush University Medical Center), Dr. Samith T. Kochuparambil (Virginia Piper Cancer Institute ), Dr. John Campana (University of Colorado Head and Neck Specialists), and Dr. Nilesh R. Vasan (University of Oklahoma Health Sciences Center).Trial RegistrationNCT02422979Ethics ApprovalThe trial was approved by the following Instution Ethics Boards and IRB# as listed: UCSF Institutional Review Board (#17-21904), Thomas Jefferson University, IRB (#16C.328), University of Oklahoma Health Sciences Center Institutional Review Board (#5723), University of Texas MD Anderson Cancer Center - Institutional Review Board (#IRB 2 IRB00002203), Quorum Review IRB (#30458/1), Rush University Medical Center Institutional Review Board (#15030601-IRB01), and Catholic Health Initiatives Institute for Research and Innovation (CIRI) Institutional Review Board (CHIRB) (# IRB00009715).ConsentN/A

374 A first-in-human Phase 1/2 open label trial evaluating the safety, pharmacology, and preliminary efficacy of VT1021 in subjects with advanced solid tumors

BackgroundVT1021, a cyclic pentapeptide, reprograms myeloid-derived suppressor cells (MDSCs) and induces the production of thrombospondin-1 (Tsp-1) in the tumor microenvironment (TME). Tsp-1, via binding to CD36 and CD47, induces apoptosis in tumor and endothelial cells, blocks the ‘do-not-eat-me’ signal, increases the M1:M2 macrophage ratio and activates cytotoxic T lymphocytes (CTLs). Preclinical studies showed robust anti-tumor activities of VT1021 in multiple animal models.MethodsThis is a first-in-human, Ph 1/2, open-label, multicenter dose escalation and expansion study in advanced solid tumors. The primary objectives are to determine the recommended Phase 2 dose (RP2D) and characterize the safety and tolerability of VT1021. Secondary objectives are to characterize the adverse event (AE) profile, evaluate pharmacokinetics (PK), and describe preliminary efficacy. Exploratory objectives include evaluation of pharmacodynamic effects of VT1021 in tumor, TME, and peripheral blood. The expansion phase focuses on ovarian, pancreatic, triple negative breast cancer, glioblastoma, and a basket cohort with high CD36-expressing tumors.ResultsIn the escalation phase, 46 subjects received between 0.5–15.6 mg/kg of VT1021 by IV infusion twice weekly. VT1021 has been well tolerated through all doses tested. One patient dosed at 1.0 mg/kg developed a grade 3 infusion reaction and 3 patients dosed at 1.0, 6.6, and 8.8 mg/kg respectively developed grade 2 infusion reactions. Other drug related AEs included grade 1–2 fatigue (n=7), nausea (n=4), constipation (n=2), increased aspartate aminotransferase (n=2) and blood bilirubin (n=2), hypomagnesaemia (n=2), and dizziness (n=2). Dose proportionality was observed in PK analysis. Among 28 evaluable subjects, one partial response (thymoma, 372+ days on treatment) and 11 stable disease (SD) in 9 different solid tumors have been observed for a disease control rate of 43%. Seven of eleven SDs had high CD36 AND high CD47 expression with an average duration of 162 days on study. VT1021 induced Tsp-1 production in peripheral blood cells at most dose levels. In addition, on-study biopsies exhibited increased Tsp-1 expression in the TME by activation of p53 in MDSCs, increased CTL infiltration, increased M1:M2 macrophage ratio, and reduced regulatory T cells in the TME. The RP2D was declared to be 11.8 mg/kg and enrollment in tumor-specific expansions is on-going.ConclusionsThrough all doses tested, VT1021 was safe and well tolerated, with dose proportional PK properties. In addition, VT1021 has demonstrated activities in reprogramming the TME which resulted in a high disease control rate in subjects with tumors expressing both high CD36 and high CD47.Trial RegistrationNCT03364400Ethics ApprovalThe study was approved by Northwestern University Medical School institutional review board (IRB), approval number 00000418, Horizon Oncology Center IRB, approval number 00001313, South Texas Accelerated Research Therapeutic IRB, approval number 00003657, University of Oklahoma Health Sciences Center IRB, approval number 00006075, Cleveland Clinic IRB, approval number 00000536, Florida Cancer Specialists IRB, approval number 00006075, Case Western IRB, approval number 00000536, Beth Israel Deaconess Hospital and Dana Farber Cancer Institute IRB, approval number 00000753 and MD Anderson IRB, approval number 00006023.

Risk Factors Associated with 6-Month Recurrent VTE Among Patients in Two Large Population-Based Surveillance Systems

Introduction The incidence of recurrent venous thromboembolism (VTE), encompassing deep vein thrombosis (DVT) and pulmonary embolism (PE), is dependent on multiple patient demographic factors and medical co-morbidities, such as cancer and surgery. We sought to determine risk factors associated with 6-month cumulative incidence of recurrent VTE. Methods A detailed description of the population-based surveillance systems have been previously published (Wendelboe 2015 and Ortel 2019 pending publication). In brief, the Centers for Disease Control and Prevention collaborated with University of Oklahoma Health Science Center and Duke University to establish surveillance systems that utilized active and passive methods to obtain data on all VTE events that occurred within Oklahoma and Durham counties, respectively between April 2012 and March 2014. This is the first report combining data from the two surveillance systems. Eligibility for the current analysis included 1) patients 18 years or older at the time of index VTE 2) no reports of patient death between index VTE and recurrent VTE or within 6 months of index VTE 3) patient data from a hospital system or out-patient clinics associated with a hospital (i.e. individuals treated at non-hospital-based out-patient clinics were excluded due to missing treatment data) 4) and index VTE occurred at least 6 months prior to the end of the surveillance period or a 6-month follow-up data abstraction was performed by the site. Recurrent VTE was defined as 1) VTE (either DVT or PE) in a different location and diagnosed after the index VTE or 2) VTE in the same location and diagnosed greater than 90 days after index VTE. Results During the surveillance period 4,205 patients were diagnosed with an index VTE, of these 2,883 (68.6%) were eligible for analysis. The 6-month incidence of VTE recurrence was 5.8% (n=166). Recurrent VTE events were diagnosed within 3 to 179 days post index VTE. Compared to patients without recurrent VTE, patients with recurrent VTE were more likely to be younger (≤ 60 years of age) and black (Table 1). They were also more likely to have a DVT only as their index VTE, to have a provoked VTE, and to have had a VTE prior to their index VTE; they were less likely to present with symptoms at the time their index VTE was diagnosed. The majority of index DVTs were located in the veins of the lower extremities, however among patients with recurrent DVT there was an increased proportion of index DVTs diagnosed in the upper extremities, other locations, and in more than one location. These differences remained after multivariable adjustment (Table 2). Clinical characteristics of recurrent VTE events are summarized in Table 3. Recurrent VTEs were primarily DVT only (73.5%), located in the lower extremity (52.7%), symptomatic at presentation (73.4%), and were associated with transient provoking factors (50.6%) including hospitalization (41.6%). Several patients (n=25) were hospitalized from the date of their index VTE to the date of VTE recurrence. Use of pharmacologic prophylaxis was high at time of recurrence (33.1%) compared to use at index VTE (12.3%). Discussion Currently, there is no U.S. national VTE surveillance system. Our VTE surveillance results show significant differences in the risk of VTE recurrence according to both patient demographic factors and clinical features of the index VTE. Factors associated with higher recurrence risk include 60 years or younger, black race, and index VTE that was DVT only, asymptomatic, and associated with persistent or transient provoking factors. Fatal recurrent VTEs may not have been identified at time of death, potentially underestimating the VTE recurrence incidence and influence of risk factors. The proportions of recurrent VTE events reported here is similar to the proportions reported in previous cohort studies, indicating that these population-based surveillance systems captured most recurrent VTE events among patients seen within hospital systems for their index VTE. Disclosures Ortel: Instrumentation Laboratories: Consultancy. Raskob:Bayer Healthcare: Consultancy, Honoraria; Daiichi Sankyo: Consultancy, Honoraria; Pfizer: Consultancy, Honoraria; Tetherex: Consultancy; Novartis: Consultancy; Anthos: Consultancy; Janssen R&D, LLC: Consultancy, Honoraria; Boehringer Ingelheim: Consultancy; BMS: Consultancy, Honoraria; Portola: Consultancy; Eli Lilly: Consultancy.