Neurotoxicity and management of primary and secondary central nervous system lymphoma after adoptive immunotherapy with CD19-directed chimeric antigen receptor T-cells

BACKGROUND

Chimeric antigen receptor (CAR) T-cells targeting CD19 have been established as a leading engineered T-cell therapy for B-cell lymphomas; however, data for patients with central nervous system (CNS) involvement are limited.

METHODS

We retrospectively report on CNS-specific toxicities, management, and CNS response of 45 consecutive CAR T-cell transfusions for patients with active CNS lymphoma at the Massachusetts General Hospital over a 5-year period.

RESULTS

Our cohort includes 17 patients with primary CNS lymphoma (PCNSL; 1 patient with 2 CAR T-cell transfusions) and 27 patients with secondary CNS lymphoma (SCNSL). Mild ICANS (grade 1-2) was observed after 19/45 transfusions (42.2%) and severe immune effector cell-associated neurotoxicity syndrome (ICANS) (grade 3-4) after 7/45 transfusions (15.6%). A larger increase in C-reactive protein (CRP) levels and higher rates of ICANS were detected in SCNSL. Early fever and baseline C-reactive protein levels were associated with ICANS occurrence. CNS response was seen in 31 cases (68.9%), including a complete response of CNS disease in 18 cases (40.0%) which lasted for a median of 11.4 ± 4.5 months. Dexamethasone dose at time of lymphodepletion (but not at or after CAR T-cell transfusion) was associated with an increased risk for CNS progression (hazard ratios [HR] per mg/d: 1.16, P = .031). If bridging therapy was warranted, the use of ibrutinib translated into favorable CNS-progression-free survival (5 vs. 1 month, HR 0.28, CI 0.1-0.7; P = .010).

CONCLUSIONS

CAR T-cells exhibit promising antitumor effects and a favorable safety profile in CNS lymphoma. Further evaluation of the role of bridging regimens and corticosteroids is warranted.

NIMG-59. EVALUATION OF THE RESPONSE ASSESSMENT CRITERIA IN NEWLY DIAGNOSED AND RECURRENT GLIOBLASTOMA

BACKGROUND

We sought to compare the Response Assessment in Neuro-Oncology (RANO), modified RANO (mRANO), and immunotherapy RANO (iRANO) in a large population of patients with newly diagnosed (nGBM) and recurrent (rGBM) glioblastoma.

METHODS

Bidimensional measurements of enhancing disease and FLAIR sequence evaluation were performed by two independent readers on brain MRIs of consecutive patients with IDH-wildtype nGBM and rGBM treated at a single institution. Discrepancies were evaluated by a third reader. Dates of disease progression (PD) were identified using RANO, mRANO, iRANO, and other response assessment criteria variations. Spearman’s correlations between PFS and OS were calculated using iterative multiple imputations for censored observations.

RESULTS

526 nGBM and 580 rGBM cases were included. Spearman’s correlations were not significantly different between RANO and mRANO in nGBM (0.69 [95% CI 0.62 to 0.75] vs. 0.67 [0.60, 0.73]) and rGBM (0.48 [0.40, 0.55] vs. 0.50 [0.42, 0.57]). Evaluation of FLAIR did not improve the correlation in patients who received antiangiogenic therapy. Acquisition of confirmation scans was associated with increased correlation only when PD was identified within 12 weeks of completion of radiation in nGBM. The use of the post-radiation MRI as a baseline was associated with increased correlation compared to use of the pre-radiation MRI in nGBM (0.67 [0.60, 0.73] vs. 0.53 [0.42, 0.62]). The correlation with iRANO was similar to RANO and mRANO among 98 patients with nGBM and 175 patients with rGBM who received immunotherapy.

CONCLUSIONS

RANO and mRANO demonstrated similar correlations between PFS and OS. The evaluation of FLAIR can be omitted, while confirmation scans were only beneficial in nGBM in the first 12 weeks after completion of radiation. There was a trend in favor of the post-radiation MRI as the baseline scan in nGBM. The use of iRANO criteria did not add a significant benefit in patients who received immunotherapy.

NCOG-09. RADIATION INDUCED CEREBRAL VOLUME LOSS AND CEREBROVASCULAR DISEASE RISK IN SURVIVORS OF GRADE 2 AND 3 GLIOMAS

INTRODUCTION

Cranial irradiation with protons (PRT) as compared with photons (PHT) reduces radiation exposure to healthy brain tissue and may mitigate early- and late-delayed complications. Prior work in our group has demonstrated that PRT results in less brain atrophy over 2 years post-treatment as compared with PHT. It is unknown whether individual factors, such as cerebrovascular disease risk (CVDR; known to be important in post-RT brain changes), interacts with treatment type to affect brain volume change.

METHODS

Patients treated with PRT for grade 2 or 3 gliomas were meticulously matched to patients treated with PHT using an eleven-tiered criterion (i.e., age, sex, tumor type, tumor location, laterality, IDH mutation status, 1p19q deletion status, concurrent chemotherapy, adjuvant chemotherapy, total Gy dose, and number of fractions). Brain volume changes were evaluated by measuring changes in ventricular size in the contralesional hemisphere (to reduce impact of tumor-related changes on volumetric measurements). A CVDR score was created, incorporating history of hypertension, hyperlipidemia, smoking, and diabetes. Correlation and multiple regression analyses examined the relationship between CVDR and brain volume loss over 2 years.

RESULTS

Mann-Whitney tests of independence showed no significant differences in CDVR between PHT and PRT treatment groups. Cerebral volume loss correlated significantly with CVDR (r =.404, p =0.02) and treatment type (r=-.366, p =0.036). Age and gender were unrelated to volume loss. A multiple regression analysis that included treatment type (PHT, PRT) and CVDR accounted for 23% of the variance in volume loss, F(2, 32) = 4.383, p < .021.

CONCLUSION

The current study demonstrates that neurotoxicity of radiation therapy is related to both CVDR and radiation treatment type. Longitudinal data, including a larger sample size, are warranted to confirm these preliminary findings. Inclusion of routine neuropsychological outcome data will be critical to understand the functional significance of these results.

CTNI-50. A RANDOMIZED PHASE 2 TRIAL OF CEDIRANIB/OLAPARIB VERSUS BEVACIZUMAB IN PATIENTS WITH RECURRENT GLIOBLASTOMA: UPDATED RESULTS

PURPOSE

Hypoxia from inhibition of angiogenesis reduces DNA repair capacity. Targeting homologous recombination pathway also has anti-angiogenic effects. To examine the synergistic effects of disrupting DNA repair pathways and angiogenesis in glioblastoma, we combine cediranib (ced), a pan vascular endothelial growth factor (VEGF) receptor inhibitor with olaparib (ola), a poly-ADP ribose polymerase (PARP) inhibitor compared with bevacizumab (bev), a VEGF-A inhibitor alone. To identify potential biomarkers predicting response to treatments, we also perform whole exome sequencing on archival tissues.

METHODS

Bevacizumab-naïve adult patients with first or second recurrence of glioblastoma after radiation and temozolomide were randomly assigned to cediranib (30mg PO daily)/olaparib (200mg PO twice daily) or bevacizumab (10mg/kg IV every 2 weeks). The primary end point was progression-free survival (PFS) at 6 months. The secondary end points included safety and overall survival (OS). Whole exome sequencing of formalin-fixed paraffin embedded archival tissue from 23 patients was performed.

RESULTS

We are presenting the final data from this trial. Between December 2017 and November 2018, a total 70 adult patients with recurrent glioblastoma were randomly assigned to receive ced/ola (n = 35) or bev (n = 35). With a data cut off on 5/30/2022, median PFS was 118 days and 92 days in ced/ola and bev groups, respectively (hazard ratio, 1.099, 95% CI 0.6-2, p = 0.76). Median overall survival was 269.5 days and 192 days in ced/ola and bev groups, respectively (hazard ratio, 0.6892, 95% CI 0.39-1.2, p = 0.2). Whole exome sequencing was performed in total 23 patients (24 samples), 14 patients in the ced/ola group and 9 patients in the bev group.

CONCLUSION

No significant survival benefit was observed in patients with recurrent glioblastoma treated with ced/ola compared to patients treated with bev monotherapy. Potential biomarkers predicting response to treatment identified from the whole exome sequencing on archival tissues will be presented.

NCOG-20. BRAIN ATROPHY IN GLIOBLASTOMA PATIENTS FOLLOWING TREATMENT WITH CHEMORADIATION OR CHEMORADIATION WITH ANTI-ANGIOGENIC THERAPY IN NRG/RTOG 0825 PARTICIPANTS

BACKGROUND

The clinical trial NRG/RTOG 0825 sought to determine if the addition of bevacizumab to temozolomide and radiation improves survival in patients with newly diagnosed glioblastomas. Tertiary objectives included measuring the effect of the addition of bevacizumab to standard chemoradiation on neurocognitive function and quality of life. In this study, we describe brain atrophy changes as measured by ventricular volume expansion.

METHODS

We analyzed longitudinal MRI brain studies obtained from NRG/RTOG-0825. Volume changes in the contralesional (non-tumor) lateral ventricle were measured. Patients were included if they had either a scan at post operative (week 0) or post radiation baseline (week 10). Patients were also required to have at least one follow-up MRI brain scan 6 months or more from their baseline scans (at Week 0 or Week 10). Volumes were delineated using tissue segmentation in Slicer software.

RESULTS

177 patients were identified with eligible baseline scans at Week 0 and 162 patients at Week 10. For participants analyzed at 6 months from the Week 0 scan, mean ventricular volume increased by 54.70% (SEM: 3.21%, t = 6.41, p < 0.001, N = 135). For patients analyzed at 6 months from the Week 10 scan, mean ventricular volume increased by 31.89% (SEM: 2.52%, t = 3.96, p < 0.001, Nf117).

CONCLUSIONS

This study presents evidence of progressive brain volume loss in patients with glioblastoma treated with standard chemoradiation with or without anti-VEGF therapy. This is one of the largest sample sizes of volumetric analysis in real world patients with glioblastoma. These volume changes begin early in the disease course and may precede treatment. Next directions include correlating these volumetric changes with neurocognitive score changes, quality of life scores and analyzing these changes by treatment arm.

Microenvironmental Landscape of Human Melanoma Brain Metastases in Response to Immune Checkpoint Inhibition

Melanoma-derived brain metastases (MBM) represent an unmet clinical need because central nervous system progression is frequently an end stage of the disease. Immune checkpoint inhibitors (ICI) provide a clinical opportunity against MBM; however, the MBM tumor microenvironment (TME) has not been fully elucidated in the context of ICI. To dissect unique elements of the MBM TME and correlates of MBM response to ICI, we collected 32 fresh MBM and performed single-cell RNA sequencing of the MBM TME and T-cell receptor clonotyping on T cells from MBM and matched blood and extracranial lesions. We observed myeloid phenotypic heterogeneity in the MBM TME, most notably multiple distinct neutrophil states, including an IL8-expressing population that correlated with malignant cell epithelial-to-mesenchymal transition. In addition, we observed significant relationships between intracranial T-cell phenotypes and the distribution of T-cell clonotypes intracranially and peripherally. We found that the phenotype, clonotype, and overall number of MBM-infiltrating T cells were associated with response to ICI, suggesting that ICI-responsive MBMs interact with peripheral blood in a manner similar to extracranial lesions. These data identify unique features of the MBM TME that may represent potential targets to improve clinical outcomes for patients with MBM.

Prospective biomarker study in newly diagnosed glioblastoma: Cyto-C clinical trial

Background

Glioblastoma (GBM) has a 5-year survival rate of 3%-5%. GBM treatment includes maximal resection followed by radiotherapy with concomitant and adjuvant temozolomide (TMZ). Cytochrome C oxidase (CcO) is a mitochondrial enzyme involved in the mechanism of resistance to TMZ. In a prior retrospective trial, CcO activity in GBMs inversely correlated with clinical outcome. The current Cyto-C study was designed to prospectively evaluate and validate the prognostic value of tumor CcO activity in patients with newly diagnosed primary GBM, and compared to the known prognostic value of MGMT promoter methylation status.

Methods

This multi-institutional, blinded, prospective biomarker study enrolled 152 patients with newly diagnosed GBM who were to undergo surgical resection and would be candidates for standard of care. The primary end point was overall survival (OS) time, and the secondary end point was progression-free survival (PFS) time. Tumor CcO activity and MGMT promoter methylation status were assayed in a centralized laboratory.

Results

OS and PFS did not differ by high or low tumor CcO activity, and the prognostic validity of MGMT promoter methylation was confirmed. Notably, a planned exploratory analysis suggested that the combination of low CcO activity and MGMT promoter methylation in tumors may be predictive of long-term survival.

Conclusions

Tumor CcO activity alone was not confirmed as a prognostic marker in GBM patients. However, the combination of low CcO activity and methylated MGMT promoter may reveal a subgroup of GBM patients with improved long-term survival that warrants further evaluation. Our work also demonstrates the importance of performing large, multi-institutional, prospective studies to validate biomarkers. We also discuss lessons learned in assembling such studies.

CTIM-18. LUMINOS-101: INITIAL SAFETY AND TOLERABILITY OF PVSRIPO AND PEMBROLIZUMAB COMBINATION THERAPY IN RECURRENT GLIOBLASTOMA

BACKGROUND

Recurrent glioblastoma (rGBM) is rapidly fatal with current therapies. PVSRIPO is an intratumoral immunotherapy targeting CD155 on antigen-presenting and malignant cells of solid tumors. Preclinically, PVSRIPO treatment leads to systemic, tumor antigen-specific, polyfunctional T-cell–mediated anti-tumor response, predominately driven by type I/III interferons. This inflammatory signature generates anti-tumor immunity and upregulates the programmed death (PD)-1 immune checkpoint in the tumor microenvironment. Preclinical models (including GBM) have shown that PVSRIPO+anti-PD-1/L1 therapy was more efficacious than either agent alone, warranting further investigation.

METHODS

Adults with histologically confirmed rGBM (1-2 prior progressions), Karnofsky performance status (KPS) ≥70, and an active, supratentorial, contrast-enhancing lesion (1-5.5 cm), received PVSRIPO (5x107 TCID50) intratumorally via convection-enhanced delivery (Day 1), followed by 200 mg pembrolizumab IV at week 2, given every 3 weeks for up to 24 months, to evaluate the safety/efficacy of the combination. A safety lead-in period (n=3-6) with a minimum 21–28-day delay before treatment of subsequent patients was planned, with a data safety monitoring board (DSMB) evaluating safety/tolerability prior to expansion (up to N=30).

RESULTS

The first 3 patients enrolled (ages 55-60, KPS 90-100) all received PVSRIPO followed by pembrolizumab (1-5 cycles), as planned. At cutoff (26-106 days of follow-up), there were no dose-limiting toxicities, treatment-emergent (TE) serious adverse events (SAE), or TEAEs necessitating a delay in initial/subsequent pembrolizumab treatments. All patients experienced a related TEAE, all grade 1 or 2 in severity. One patient experienced an AE of special interest (peritumoral edema, resulting in headache and hemiparesis), successfully managed with low-dose bevacizumab and corticosteroids. The DSMB unanimously recommended the study proceed without modification.

CONCLUSIONS

Intratumoral PVSRIPO+pembrolizumab was reasonably well tolerated, warranting continued investigation of the safety and efficacy of this combination in patients with rGBM.

TAMI-29. MR SPECTROSCOPY MEASURES OF LAC/NAA AND NAA/CHO DIFFERENTIATE SURVIVORSHIP IN PATIENTS WITH RECURRENT GLIOBLASTOMA TREATED WITH ANTI-ANGIOGENIC THERAPY

Patients with recurrent glioblastoma (rGBM) are often started on anti-angiogenic therapy such as bevacizumab. However, determining treatment failure using conventional MRI methods remains challenging. We prospectively collected longitudinal MR spectroscopy data in 33 patients with rGBM and quantified various metabolites including N-acetylaspartate (NAA), Choline (Cho), and Lactate (Lac). After stratifying patients by 9 month survival, we found that longer-term survivors had decreased Lac/NAA and increased NAA/Cho compared to shorter-term survivors. ROC analyses illustrated that intratumoral changes in NAA/Cho were predictive of survival at 1 day (AUC 0.92), 2 weeks (AUC 0.75), 8 weeks (AUC 0.71) and 16 weeks (AUC 0.85) but not 4 weeks (AUC 0.60). Intratumoral changes in Lac/NAA were predictive of survival at all time points tested (AUCs > 0.76 for all time points). At 8 weeks, 90% of patients with increased Lac/NAA from baseline and 88% of patients with decreased NAA/Cho did not survive 9 months. Changes in NAA/Cho and Lac/NAA may serve as early biomarkers of anti-angiogenic treatment failure.

CTIM-01. PHASE II TRIAL OF PEMBROLIZUMAB AND LENVATINIB FOR LEPTOMENINGEAL METASTASES

Leptomeningeal metastasis (LMD) is a late complication of cancer with poor prognosis and median survival of approximately 4-6 weeks without treatment. Whole brain radiation remains the mainstay of treatment, however it can cause significant neurocognitive sequelae and has not been shown to prolong overall survival. Thus, new treatment strategies are urgently needed to improve outcomes in patients with LMD. Results from recent Phase 2 studies of immune checkpoint inhibitors in LMD shows promising improvement in overall survival. Combining anti-VEGF therapy and immunotherapy may control symptoms due to inflammation and tumor-induced irritation, minimize steroid use, and promote improved efficacy of immunotherapy through modulation of the tumor immune microenvironment. We designed a multi-institutional, single-arm Phase 2 study of pembrolizumab in combination with lenvatinib in patients with LMD from any solid tumor. The primary objective is to estimate the overall survival rate at 6 months (OS6). A Simon two-stage design with a total sample size of 19 evaluable patients will be used to compare a null hypothesis of OS6 of 25% against an alternative hypothesis of 55%. Secondary objectives include assessing safety of pembrolizumab and lenvatinib in this patient population, systemic response rate, intracranial/intraspinal response rate, and progression-free survival. We will also explore clinician-reported neurologic outcomes and patient-reported quality of life and symptom burden. Blood, cerebrospinal fluid, and tissue biomarkers will be analyzed to determine predictors of response. Patients must be on minimal doses of steroids prior to study enrollment and cannot have received prior immune checkpoint inhibitor or anti-VEGF therapy. Response to treatment will be determined using RANO-BM for intracranial disease and RECIST 1.1 for systemic disease. Study accrual is anticipated over 12-24 months with anticipated total study duration of 30 months.

CTIM-02. PHASE II STUDY OF IPILIMUMAB AND NIVOLUMAB IN LEPTOMENINGEAL CARCINOMATOSIS

Leptomeningeal disease (LMD) is an increasingly common complication from solid tumor malignancies with a poor prognosis and limited treatment options. We conducted a single arm Phase II study of combined ipilimumab and nivolumab in patients with LMD from solid tumor malignancies (NCT02939300). Patients received manufacturer-specific dosing regimens of combined ipilimumab and nivolumab based on primary-tumor histology until definitive progression or unacceptable toxicity. The primary end point was rate of overall survival at 3 months (OS3). A Simon two-stage design was used to compare a null hypothesis OS3 of 18% against an alternative of 44%. Eighteen patients with diverse primary tumor histologies were enrolled and all received at least one dose of combined ipilimumab and nivolumab. Median follow up based on patients still alive was 8.0 months (range: 0.5 to 15.9 months). The study met its primary endpoint as 8 of 18 (OS3 0.44; 90% CI: 0.24 to 0.66) patients were alive at three months after enrollment. One third of patients experienced one (or more) grade-3 or higher adverse events possibly related to treatment. Two patients discontinued protocol treatment due to unacceptable toxicity (hepatitis and colitis, respectively). The most frequent adverse events overall included fatigue (N=7), nausea (N=6), fever (N=6), anorexia (N=6) and rash (N=6). Combined ipilimumab and nivolumab has an acceptable safety profile and demonstrates promising activity in LMD patients; this therapeutic approach should be studied in larger, multicenter clinical trials to validate these results as well as better identify patients who will benefit.

BIOM-04. SENSITIVE DETECTION OF LEPTOMENINGEAL DISEASE USING CELL-FREE DNA FROM CEREBROSPINAL FLUID

Leptomeningeal disease is a devastating complication of cancer that is frequently underdiagnosed due to the low sensitivity of cerebrospinal fluid cytology, the current gold-standard diagnostic method. We performed genomic sequencing on cerebrospinal fluid specimens obtained from patients with suspected or confirmed leptomeningeal disease to identify tumor-derived cell-free DNA. From the same fluid draw, cerebrospinal fluid cytology was assayed for comparison. 30 patients underwent cytology and cell-free DNA analysis. This study consisted of two patient populations: 22 patients with cytology-confirmed leptomeningeal disease without parenchymal tumors abutting their cerebrospinal fluid and 8 patients with parenchymal brain metastases with no evidence of leptomeningeal disease. The primary outcome was the diagnostic accuracy of cell-free DNA, defined as the number of correct diagnoses out of the total number of tests assayed. A total of 30 patients, 23 female and 7 male, with a median age of 51 participated in this study. Participants mostly presented with metastatic solid malignancies. In patients previously diagnosed with leptomeningeal disease via cytology with no parenchymal tumor abutting cerebrospinal fluid, cell-free DNA was accurate in diagnosis of leptomeningeal disease in 45 of 48 follow-up samples (94%; 95% CI, 83%-99%). Cytology was accurate in 36 of 48 follow-up samples (75%; 95% CI, 60%-86%). Cell-free DNA was significantly more accurate (P=.02) and sensitive (P=.02) than cytology in patients without parenchymal tumors abutting the cerebrospinal fluid. In three patients with parenchymal brain metastases abutting the cerebrospinal fluid and no suspicion for leptomeningeal disease, cytology was negative in all three patients; whereas, cell-free DNA was positive in all three. This study demonstrates the improved sensitivity and accuracy of cell-free DNA in diagnosing leptomeningeal disease with the exception of parenchymal tumors abutting cerebrospinal fluid. Overall, these results will lead to improved diagnosis of leptomeningeal disease and potentially earlier intervention and clinical trial enrollment.

BIOM-09. MYO-INOSITOL LEVELS ON MR SPECTROSCOPY CAN PREDICT FAILURE OF ANTI-ANGIOGENIC TREATMENT IN RECURRENT GLIOBLASTOMA

BACKGROUND

Recurrent glioblastoma (rGBM) patients are often treated with anti-angiogenic agents such as bevacizumab (BEV). Despite therapeutic promise, conventional MR methods fail to determine which patients may not benefit. PURPOSE: The purpose of this study was to utilize magnetic resonance spectroscopic imaging (MRSI) with intermediate and short echo time to generate corrected Myo-inositol normalized by contralateral creatine (mI/c-Cr) in patients with rGBM treated with BEV and investigate whether it can predict survivorship prior to BEV initiation (baseline) and at 1-day, 4-weeks, and 8-weeks thereafter.

METHODS

We conducted a prospective, longitudinal study and evaluated spectroscopic data of myo-inositol (mI), a glial marker and osmoregulator within the brain, normalized to contralateral-creatine (mI/c-Cr) in the intratumoral, contralateral normal appearing white matter, and peritumoral volumes of rGBM patients. Area under the ROC curve (AUC) was calculated for all volumes at baseline, 1-day, 4-weeks, and 8-weeks after treatment to determine mI/c-Cr’s ability to predict survivorship.

RESULTS

21 participants (62 ± 12 years, 15 men) were evaluated. Lower mI/c-Cr in the tumor prior to and during BEV treatment predicted poor survivorship, with ROC analyses illustrating an AUC of 0.75 at baseline, 0.87 at 1-day, and 1 at 8 weeks. Lower levels of mI/c-Cr were also observed in the contralateral and the peritumoral volumes for shorter-term survivors. In the contralateral volume, lower mI/Cr was predictive of shorter-term survival at baseline and all other timepoints. Within the peritumoral volume, lower mI/c-Cr was predictive of shorter-term survival at baseline (AUC=0.80), 1-day (AUC=0.93), and 4-weeks (AUC=0.68).

CONCLUSIONS

Lower levels of mI/c-Cr within intratumoral, contralateral, and peritumoral volumes were predictive of poor survivorship and anti-angiogenic treatment failure as early as one month before BEV treatment. Acquiring MRSI alongside conventional MR imaging modalities can convey critical information regarding tumor microenvironment that informs management of patients with rGBM.

CTNI-53. RADIATION TREATMENT VOLUMES BEFORE AND AFTER BRAF/MEK THERAPY IN NEWLY DIAGNOSED PAPILLARY CRANIOPHARYNGIOMAS: A CORRELATIVE ANALYSIS OF THE ALLIANCE A071601 PHASE II TRIAL

PURPOSE

Standard of care for craniopharyngiomas is surgery with or without radiotherapy (RT). Cohort A of Alliance A071601 evaluated the efficacy of BRAF/MEK inhibition with vemurafenib/cobimetinib in patients with previously untreated papillary craniopharyngiomas (PCP), which carry the BRAF V600E mutation. Cohort B is currently enrolling patients with recurrence after RT. In a correlative analysis, we examined changes in RT volumes after BRAF/MEK therapy in Cohort A.

METHODS

Previously unirradiated patients with BRAF-mutated PCP were treated with vemurafenib/cobimetinib. Sixteen patients had scans available before starting vemurafenib/cobimetinib (“pre-therapy”) and after completing therapy (“post-therapy”). Two patients went off study treatment after 8 and 9 days due to side-effects and were excluded for this analysis. Gross target volumes (GTV) were contoured on pre-therapy and post-therapy scans. On post-therapy scans, an additional target comprising gross disease and at-risk regions for microscopic residual disease (GTV-micro) was defined and considered the treatment volume. Clinical target volume (CTV) was a 5-mm uniform expansion on pre-therapy GTV and post-therapy GTV-micro. Volumes were independently reviewed by two radiation oncologists. Changes in volumes from pre- versus post-therapy were compared using the Wilcoxon signed rank test.

RESULTS

In 14 patients evaluated, 57% were female and median age at enrollment was 49.5 years (range 33-83). Median time on treatment was 8.9 months (range 4.0-18.0). Median GTV pre-therapy was 3.8 mL (range 0.2-23.4) versus 0.3 mL (range 0.0-3.2) post-therapy (p=0.0001) and 1.7 mL (range 0.1-8.0) post-therapy GTV-micro (p=0.0001). Median CTV pre-therapy was 13.7 mL (range 2.8-51.8) versus 9.1 mL (range 2.2-27.5) post-therapy (p=0.0001). All tumors abutted the optic chiasm pre-therapy, only 6 did post-therapy.

CONCLUSIONS

Vemurafenib/cobimetinib resulted in smaller RT volumes. BRAF/MEK inhibitors could reduce RT volumes and spare dose to surrounding normal structures. Enrollment to Cohort B of Alliance A071601 should be considered for patients with recurrent tumors after RT.

SUPPORT

https://acknowledgments.alliancefound.org

CTIM-30. PHASE II TRIAL OF PEMBROLIZUMAB IN RECURRENT AND RESIDUAL HIGH-GRADE MENINGIOMAS

INTRODUCTION

High-grade meningiomas are associated with significant neuro-cognitive morbidity and a poor prognosis. Systemic therapies, to date, have demonstrated minimal efficacy. We recently found that high-grade meningiomas harbor an immunosuppressive tumor microenvironment and that programmed cell death-ligand 1 (PD-L1) expression may contribute to the aggressive phenotype of these tumors. Therefore, we conducted a single-arm, open-label phase II trial evaluating efficacy of pembrolizumab, a PD-1 inhibitor, in a cohort of 24 patients with recurrent and progressive grade II and III meningiomas.

METHODS

The primary endpoint was the rate of progression-free survival at 6 months. The trial distinguished between 6-month PFS (PFS-6) rates of 26% vs. 52%. If at least 10 patients demonstrated a 6-month PFS, among the 24 patients, the agent would be considered worthy of further study. This design has at least 88% power using an exact binomial test with a one-sided significance level of 0.1.

RESULTS

Between November 2017 to December 2019, twenty-four patients were enrolled. The majority of the patients in our cohort were heavily pre-treated; prior to enrolling to the study, twenty patients underwent more than one surgical resection and twelve patients had received more than one round of radiotherapy. Our study met its primary endpoint and achieved a 6-month progression-free survival rate of 0.50 (90% exact CI: 0.32-0.68) and a median PFS of 8.3 months (90% CI: 4.1-12.9 months). For the twelve patients who achieved the PFS-6 primary endpoint, median PFS from the start of treatment was 17.3 months (90% CI: 9.7 – 24.3 months). Four patients had grade-3 or higher adverse events that were at least possibly treatment-related, including colitis, skin infection, encephalopathy and transaminitis.

CONCLUSION

Our study achieved its primary endpoint. These results suggest that pembrolizumab exerts promising activity on these tumors and results in prolonged PFS compared to historical controls.

CTNI-54. A SINGLE ARM PHASE II STUDY OF THE DUAL MTORC1/MTORC2 INHIBITOR VISTUSERTIB PROVIDED FOR SPORADIC PATIENTS WITH GRADE II-III MENINGIOMAS THAT RECUR OR PROGRESS AFTER SURGERY AND RADIATION

Grade II/III meningiomas have increased rates of recurrence with no approved medical therapies. The historical progression-free survival at 6 months (PFS-6) is 25% with rates >35% declared of interest for drug development. NF2 gene inactivation occurs in about half of meningiomas. Based on our studies showing mTORC1 and mTORC2/SGK1 pathway activation in NF2-deficient meningiomas and the paradoxical activation of the mTORC2/AKT pathway, we hypothesized that mTORC1/mTORC2 inhibitors would be active in meningiomas. We studied the effect of vistusertib in patients with progressive/recurrent grade II/III meningiomas (NCT03071874). Vistusertib was administered orally at 125mg twice daily on two consecutive days each week. MRIs were obtained every 56 days. Tumor size was defined as the largest cross-sectional area. Progression was defined as ≥ 25% increase in the sum of products of all measurable lesions over smallest sum observed. The primary endpoint was PFS-6. Secondary endpoints included toxicity, radiographic response, and correlative studies including immunohistochemistry for mTORC1/2 pathway activation and genetic biomarkers. Twenty-eight patients (13 female, median age 58 years, median KPS 80%) were enrolled. Median tumor size was 4.4cm; 71% were grade II and 50% harbored pathogenic NF2 variants. Four patients discontinued treatment voluntarily and 1 each withdrew for intercurrent illness and non-compliance. PFS-6 is 47% (CI, 26%-65%) and OS-12 is 72% (95%CI, 48%-86%). PFS but not OS was shorter for patients with grade 3 meningiomas; there was no difference in PFS/OS between genetic groups. Adverse events at least possibly related to vistusertib with frequency >10% include nausea, fatigue, hypophosphatemia, diarrhea, anorexia, dry mouth, hypertriglyceridemia, hypertension, vomiting, increased ALT, constipation, and weight loss. Vistusertib treatment was associated with a PFS-6 rate exceeding the target of 35% for recurrent high-grade meningioma. Adverse events were tolerable in this patient population. These data support the continued development of mTORC1/2 inhibitors in this setting.

IMMU-08. PHASE II TRIAL OF PEMBROLIZUMAB AND LENVATINIB FOR LEPTOMENINGEAL METASTASES

Leptomeningeal metastasis (LMD) is a late complication of cancer with poor prognosis and median survival of approximately 4-6 weeks without treatment. Whole brain radiation remains the mainstay of treatment, however it can cause significant neurocognitive sequelae and has not been shown to prolong overall survival. Thus, new treatment strategies are urgently needed to improve outcomes in patients with LMD. Results from recent Phase 2 studies of immune checkpoint inhibitors in LMD shows promising improvement in overall survival. Combining anti-VEGF therapy and immunotherapy may control symptoms due to inflammation and tumor-induced irritation, minimize steroid use, and promote improved efficacy of immunotherapy through modulation of the tumor immune microenvironment. We designed a multi-institutional, single-arm Phase 2 study of pembrolizumab in combination with lenvatinib in patients with LMD from any solid tumor. The primary objective is to estimate the overall survival rate at 6 months (OS6). A Simon two-stage design with a total sample size of 19 evaluable patients will be used to compare a null hypothesis of OS6 of 25% against an alternative hypothesis of 55%. Secondary objectives include assessing safety of pembrolizumab and lenvatinib in this patient population, systemic response rate, intracranial/intraspinal response rate, and progression-free survival. We will also explore clinician-reported neurologic outcomes and patient-reported quality of life and symptom burden. Blood, cerebrospinal fluid, and tissue biomarkers will be analyzed to determine predictors of response. Patients must be on minimal doses of steroids prior to study enrollment and cannot have received prior immune checkpoint inhibitor or anti-VEGF therapy. Response to treatment will be determined using RANO-BM for intracranial disease and RECIST 1.1 for systemic disease. Study accrual is anticipated over 12-24 months with anticipated total study duration of 30 months.

Detection of Leptomeningeal Disease Using Cell-Free DNA From Cerebrospinal Fluid

Importance

Leptomeningeal disease (LMD) is a devastating complication of cancer that is frequently underdiagnosed owing to the low sensitivity of cerebrospinal fluid (CSF) cytologic assessment, the current benchmark diagnostic method. Improving diagnostic sensitivity may lead to improved treatment decisions.

Objective

To assess whether cell-free DNA (cfDNA) analysis of CSF may be used to diagnose LMD more accurately than cytologic analysis.

Design, Setting, and Participants

This diagnostic study conducted in a neuro-oncology clinic at 2 large, tertiary medical centers assessed the use of genomic sequencing of CSF samples obtained from 30 patients with suspected or confirmed LMD from 2015 through 2018 to identify tumor-derived cfDNA. From the same CSF samples, cytologic analyses were conducted, and the results of the 2 tests were compared. This study consisted of 2 patient populations: 22 patients with cytologically confirmed LMD without parenchymal tumors abutting their CSF and 8 patients with parenchymal brain metastases with no evidence of LMD. Patients were considered positive for the presence of LMD if previous CSF cytologic analysis was positive for malignant cells. The analysis was conducted from 2015 to 2018.

Main Outcomes and Measures

The primary outcome was the diagnostic accuracy of cfDNA analysis, defined as the number of tests that resulted in correct diagnoses out of the total number of tests assayed. Hypotheses were formed before data collection.

Results

In total, 30 patients (23 women [77%]; median age, 51 years [range, 28-81 years]), primarily presenting with metastatic solid malignant neoplasms, participated in this study. For 48 follow-up samples from patients previously diagnosed via cytologic analysis as having LMD with no parenchymal tumor abutting CSF, cfDNA findings were accurate in the assessment of LMD in 45 samples (94%; 95% CI, 83%-99%), whereas cytologic analysis was accurate in 36 samples (75%; 95% CI, 60%-86%), a significant difference (P = .02). Of 43 LMD-positive samples, CSF cfDNA analysis was sensitive to LMD in 40 samples (93%; 95% CI, 81%-99%), and cytologic analysis was sensitive to LMD in 31 samples (72%; 95% CI, 56%-85%), a significant difference (P = .02). For 3 patients with parenchymal brain metastases abutting the CSF and no suspicion of LMD, cytologic findings were negative for LMD in all 3 patients, whereas cfDNA findings were positive in all 3 patients.

Conclusions and Relevance

This diagnostic study found improved sensitivity and accuracy of cfDNA CSF testing vs cytologic assessment for diagnosing LMD with the exception of parenchymal tumors abutting CSF, suggesting improved ability to diagnosis LMD. Consideration of incorporating CSF cfDNA analysis into clinical care is warranted.

Consensus recommendations for a standardized brain tumor imaging protocol for clinical trials in brain metastases

A recent meeting was held on March 22, 2019, among the FDA, clinical scientists, pharmaceutical and biotech companies, clinical trials cooperative groups, and patient advocacy groups to discuss challenges and potential solutions for increasing development of therapeutics for central nervous system metastases. A key issue identified at this meeting was the need for consistent tumor measurement for reliable tumor response assessment, including the first step of standardized image acquisition with an MRI protocol that could be implemented in multicenter studies aimed at testing new therapeutics. This document builds upon previous consensus recommendations for a standardized brain tumor imaging protocol (BTIP) in high-grade gliomas and defines a protocol for brain metastases (BTIP-BM) that addresses unique challenges associated with assessment of CNS metastases. The "minimum standard" recommended pulse sequences include: (i) parameter matched pre- and post-contrast inversion recovery (IR)-prepared, isotropic 3D T1-weighted gradient echo (IR-GRE); (ii) axial 2D T2-weighted turbo spin echo acquired after injection of gadolinium-based contrast agent and before post-contrast 3D T1-weighted images; (iii) axial 2D or 3D T2-weighted fluid attenuated inversion recovery; (iv) axial 2D, 3-directional diffusion-weighted images; and (v) post-contrast 2D T1-weighted spin echo images for increased lesion conspicuity. Recommended sequence parameters are provided for both 1.5T and 3T MR systems. An "ideal" protocol is also provided, which replaces IR-GRE with 3D TSE T1-weighted imaging pre- and post-gadolinium, and is best performed at 3T, for which dynamic susceptibility contrast perfusion is included. Recommended perfusion parameters are given.

PATH-27. CLINICAL, RADIOLOGIC AND PROGNOSTIC PROFILE OF IDH WILD TYPE DIFFUSE ASTROCYTOMA WITH MOLECULAR FEATURES OF GLIOBLASTOMA

BACKGROUND

cIMPACT-NOW-3 recommends that IDH-wildtype diffuse astrocytic glioma Grade II or III with EGFR amplification, or combined whole chromosome 7 gain and whole chromosome 10 loss, or TERT promoter mutation should receive an integrated classification: Diffuse astrocytic glioma, IDH-wildtype with molecular features of glioblastoma, WHO grade IV. The aims of this study were: Outline the features of a cohort of patients with molecular glioblastoma according to the above criteria; Assess clinical and molecular factors that may inform prognosis; Determine if cIMPACT-NOW-3 recommendation changed clinical practice and clinical trial enrolment.

METHODS

61 patients diagnosed with IDH-wildtype diffuse astrocytic glioma Grade II or III and EGFR amp or mTERT or chromosome (+7/-10) between 2011 and 2019 at MGH were included in this single center retrospective cohort study. Data collected: sex, age, extent of surgery, functional status, histological grade, molecular diagnostics and treatment. Progression was defined using RANO criteria, progression was quantified in terms of months from the initial surgery. Survival was defined in terms of months from initial surgery to date of death or last known visit.

RESULTS

mOS was 16 months, mPFS was 9 months. 14 patients (23%) survived > 24 months, 7 ≥ 36 months (mOS 32 months; 9 deceased). The probability of survival in patients with markedly enhancing tumors was 0.5 at 3 months versus 0.5 at 11 months in non-enhancing tumors. The probability of survival in patients who underwent biopsy only was 0.5 at 5 months compared to 0.5 at 12 months in patients with maximally resected tumors. 1 patient was enrolled in a clinical trial at diagnosis. 6 were enrolled at time of recurrence.

CONCLUSIONS

mOS and pFS in the deceased patients was comparable to historical data on survival in IDH wild-type glioblastomas. Inclusion criteria in clinical trials have not reflected the cIMPACT-NOW-3 recommendation so far.

NIMG-05. ADVANCED IMAGING TO ASSESS LONGITUDINAL VASCULAR CHANGES IN BRAIN METASTASES TREATED WITH CHECKPOINT INHIBITION

Immune checkpoint inhibitors (ICI) have recently been shown to be effective for brain metastases (BM) in melanoma and lung cancer. However, accurately assessing intracranial response in patients undergoing ICI is a challenge, as current measures cannot reliably distinguish pseudoprogression from true tumor progression. To identify potential biomarkers of response, we analyzed standard post-contrast and dynamic susceptibility contrast MRI to identify characteristic vascular signatures as part of an ongoing Phase 2 study of pembrolizumab for patients with untreated or progressive, previously treated BM from any histology. Tumor volume measurements were calculated by summating all enhancing voxels. A volumetric increase of >40% was categorized as progressive disease (PD), a decrease of >60% as partial response (PR), and stable disease (SD) as between -60% and +40%. 78 patients have been enrolled, of whom 44 have received at least baseline advanced MR imaging. Histologies include 21 with breast cancer, 5 with non-small cell lung cancer, 4 with melanoma, and 13 with other cancers. At baseline, the total number of BM was 1-50+ per patient. Based on summing the entire enhancing intracranial disease burden, best volumetric responses for the 33 evaluable patients include 4 PR, 10 SD, and 19 PD. On preliminary analysis, there was a correlation between increased tumor cerebral blood volume/flow with tumor progression. Correlation of additional vascular physiologic parameters (e.g. vessel caliber, tissue oxygenation) to volumetric response, patient outcome, and standardized response criteria (iRANO) are ongoing. Our data provides potential evidence that effective ICI is associated with a decrease in perfusion. Ongoing analyses to uncover additional vascular changes – specifically longitudinal metrics reflecting vascular structure and function - within BM to ICI are pending. These findings have potential to explore mechanisms of ICI response and resistance, as well as biomarkers of response.

Multiparametric MR-PET Imaging Predicts Pharmacokinetics and Clinical Response to GDC-0084 in Patients with Recurrent High-Grade Glioma

PURPOSE

GDC-0084 is an oral, brain-penetrant small-molecule inhibitor of PI3K and mTOR. Because these two targets alter tumor vascularity and metabolism, respectively, we hypothesized multiparametric MR-PET could be used to quantify the response, estimate pharmacokinetic (PK) parameters, and predict progression-free survival (PFS) in patients with recurrent malignant gliomas.

PATIENTS AND METHODS

Multiparametric advanced MR-PET imaging was performed to evaluate physiologic response in a first-in-man, multicenter, phase I, dose-escalation study of GDC-0084 (NCT01547546) in 47 patients with recurrent malignant glioma.

RESULTS

Measured maximum concentration (C _max) was associated with a decrease in enhancing tumor volume (P = 0.0287) and an increase in fractional anisotropy (FA; P = 0.0418). Posttreatment tumor volume, ¹⁸F-FDG uptake, K^trans, and relative cerebral blood volume (rCBV) were all correlated with C _max. A linear combination of change in ¹⁸F-FDG PET uptake, apparent diffusion coefficient (ADC), FA, K^trans, v_p, and rCBV was able to estimate both C _max (R² = 0.4113; P < 0.0001) and drug exposure (AUC; R² = 0.3481; P < 0.0001). Using this composite multiparametric MR-PET imaging response biomarker to predict PK, patients with an estimated C _max > 0.1 μmol/L and AUC > 1.25 μmol/L*hour demonstrated significantly longer PFS compared with patients with a lower estimated concentration and exposure (P = 0.0039 and P = 0.0296, respectively).

CONCLUSIONS

Results from this study suggest composite biomarkers created from multiparametric MR-PET imaging targeting metabolic and/or physiologic processes specific to the drug mechanism of action may be useful for subsequent evaluation of treatment efficacy for larger phase II-III studies.

First-in-Human Phase I Study to Evaluate the Brain-Penetrant PI3K/mTOR Inhibitor GDC-0084 in Patients with Progressive or Recurrent High-Grade Glioma

PURPOSE

GDC-0084 is an oral, brain-penetrant small-molecule inhibitor of PI3K and mTOR. A first-in-human, phase I study was conducted in patients with recurrent high-grade glioma.

PATIENTS AND METHODS

GDC-0084 was administered orally, once daily, to evaluate safety, pharmacokinetics (PK), and activity. Fluorodeoxyglucose-PET (FDG-PET) was performed to measure metabolic responses.

RESULTS

Forty-seven heavily pretreated patients enrolled in eight cohorts (2-65 mg). Dose-limiting toxicities included 1 case of grade 2 bradycardia and grade 3 myocardial ischemia (15 mg), grade 3 stomatitis (45 mg), and 2 cases of grade 3 mucosal inflammation (65 mg); the MTD was 45 mg/day. GDC-0084 demonstrated linear and dose-proportional PK, with a half-life (∼19 hours) supportive of once-daily dosing. At 45 mg/day, steady-state concentrations exceeded preclinical target concentrations producing antitumor activity in xenograft models. FDG-PET in 7 of 27 patients (26%) showed metabolic partial response. At doses ≥45 mg/day, a trend toward decreased median standardized uptake value in normal brain was observed, suggesting central nervous system penetration of drug. In two resection specimens, GDC-0084 was detected at similar levels in tumor and brain tissue, with a brain tissue/tumor-to-plasma ratio of >1 and >0.5 for total and free drug, respectively. Best overall response was stable disease in 19 patients (40%) and progressive disease in 26 patients (55%); 2 patients (4%) were nonevaluable.

CONCLUSIONS

GDC-0084 demonstrated classic PI3K/mTOR-inhibitor related toxicities. FDG-PET and concentration data from brain tumor tissue suggest that GDC-0084 crossed the blood-brain barrier.

NIMG-34. MULTI-PARAMETRIC MR-PET IMAGING PREDICTS PHARMACOKINETICS AND CLINICAL RESPONSE TO GDC-0084 IN HUMAN RECURRENT HIGH-GRADE GLIOMA

Alterations in the PI3K pathway are found in the majority of malignant gliomas, but lack of efficacy has caused investigators to question the viability of this target, particularly due to lack of brain penetration. GDC-0084 was specifically optimized to penetrate the brain, targeting both PI3K and mTOR. Since these two targets alter tumor vascularity and metabolism, respectively, we hypothesized that multi-parametric MR-PET could be used to quantify the response, estimate pharmacokinetic (PK) parameters, and predict progression-free survival (PFS) in patients with recurrent malignant gliomas. In this first-in-man, multicenter, phase I, dose-escalation study (NCT01547546), we show in 47 patients that the measured maximum concentration (Cmax) of GDC0084 was associated with a decrease in enhancing tumor volume (P=0.0287) and an increase in fractional anisotropy (FA) (P=0.0418). Post-treatment tumor volume, 18F-FDG uptake, Ktrans, and relative cerebral blood volume (rCBV) were all correlated with Cmax. A linear combination of change in 18F-FDG PET uptake, apparent diffusion coefficient (ADC), FA, Ktrans, vp, and rCBV were able to estimate both Cmax (R2=0.4113, P< 0.0001) and drug exposure (AUC) (R2=0.3481, P< 0.0001). Using this composite multi-parametric MR-PET imaging response biomarker to predict PK, patients with an estimated Cmax >0.1 uM and AUC > 1.25 uM*hr demonstrated significantly longer PFS compared with patients with a lower estimated concentration and exposure (P=0.0039 and P=0.0296, respectively). Results from the current study suggest composite biomarkers created from multi-parametric MR-PET imaging targeting metabolic and/or physiologic processes specific to the drug mechanism of action may be useful for subsequent evaluation of treatment efficacy for larger phase II-III studies.

ACTR-61. A RANDOMIZED PHASE 2 TRIAL OF CEDIRANIB IN COMBINATION WITH OLAPARIB VERSUS BEVACIZUMAB IN PATIENTS WITH RECURRENT GLIOBLASTOMA

BACKGROUND

Like most proliferating tumors, GBM relies heavily on accurate DNA repair for maintenance of genome stability. Dysfunction in repair of both single and double strand DNA breaks by PARP inhibition and impairment of homologous recombination, respectively, would be synthetically lethal. In this study we combined the PARP inhibitor olaparib with cediranib, a pan VEGF receptor inhibitor. Cediranib may mediate disruption in the homologous recombination pathway through its antiangiogenic properties.

METHODS

Through the Experimental Therapeutics Clinical Trials Network, we performed an open-label randomized phase II study of bevacizumab (BEV)- naive adult patients with first or second recurrence of glioblastoma after radiation and temozolomide. Patients were randomized 1:1 to receive either olaparib 200 mg by mouth twice daily with cediranib 30 mg by mouth daily or BEV 10 mg/kg IV every 2 weeks. The primary endpoint was progression-free survival at 6 months (PFS6). Secondary endpoints included safety and overall survival. Exploratory objectives included blood, tissue and imaging-based biomarkers of response to treatment.

RESULTS

Seventy patients were enrolled. Median age was 60.5 years (range: 19–79), 39% females, median KPS was 90 (range: 60–100). Baseline characteristics were well balanced. With a data cut-off of 5/2/2019, PFS6 was 14% [95% CI 4–30%] in the cediranib/olaparib arm vs 30.9% [95% CI 12.7–51.2%] in the BEV arm. Median OS was 247 days in the cediranib/olaparib arm vs 201 days in the BEV arm, HR 0.816, 95% CI (0.431, 1.546). Related grade 3, 4 or 5 toxicity was experienced in 29% vs 12% of patients for the cediranib/olaparib vs BEV arm.

CONCLUSION

Treatment with cediranib/olaparib failed to increase PFS and OS in patients with recurrent GBM. Blood, tissue and imaging correlates will be presented to help understand why this treatment combination was unsuccessful.

CMET-33. PHASE II STUDY OF PALBOCICLIB IN BRAIN METASTASES HARBORING CDK PATHWAY ALTERATIONS

BACKGROUND

Up to 25% of all cancer patients will develop brain metastases and prognosis remains poor. In preclinical work, we discovered that more than 50% brain metastases genomically diverge from primary tumors and harbor alterations in genes involved with cell cycle regulation. We thus initiated a phase II study to evaluate the efficacy and safety of the cyclin dependent kinase (CDK) 4/6 inhibitor, palbociclib, in patients with recurrent brain metastases with alterations in the CDK pathway (NCT02896335).

METHODS

The primary endpoint of the trial is the rate of intracranial benefit (defined as CR, PR, or SD, per RANO) at 8 weeks after the start of palbociclib. A Simon two-stage design was used to compare the rate of intracranial benefit for a null rate of 10% against an alternative of 30%. Fifteen patients were to be enrolled in the first stage. If fewer than 2 responders were observed, then the study would stop for insufficient evidence of efficacy. If 6 or more responders were observed among the total of 30, this treatment regimen would be considered worthy of further study. CSF, blood samples and tumor samples were collected to elucidate the genomic determinants of response to CDK inhibitors in the brain.

RESULTS

A total of 14 patients have been accrued (5 with breast cancer, 4 with melanoma, 3 esophageal and 2 with non-small cell lung cancer) thus far. One or more grade-3 or higher adverse events at least possibly related to treatment were seen in six (42%) patients, the majority being hematologic toxicities. At the time of the data analysis, eight (57%) patients had achieved intracranial benefit. Therefore, the study met primary endpoint.

CONCLUSIONS

In this unique, genomically-guided brain metastasis trial, we demonstrate that the CDK 4/6 inhibitor, palbociclib, is well-tolerated and has activity in patients with brain metastases.

NIMG-43. LONGITUDINAL TRACKING AND GROWTH RATE CHARACTERIZATION OF BRAIN METASTASES ON MAGNETIC RESONANCE IMAGING

PURPOSE

Measuring treatment response is vital for assessing efficacy of treatment regimen for patients with brain metastases (BM). Unfortunately, manual delineation of all lesions on MRI across time-points is prohibitively time-consuming, making it infeasible to track individual lesion growth/shrinkage rates as part of the clinical workflow. To overcome this challenge, we propose a deep learning approach to segment all BM, and furthermore, show that certain brain regions are more prone to high-growth rate lesions.

METHODS

163 longitudinal MRIs from 77 patients with MPRAGE-post contrast imaging protocol were prospectively obtained from Massachusetts General Hospital (MGH). An expert neuro-oncologist provided ground truth segmentations for all patients. A 3D U-Net architecture was trained to automatically segment BM; training was stopped when validation set Dice score plateaued to prevent overfitting. To enable lesion tracking, all time-points per patient were affinely registered to each other. Every lesion was subsequently classified based on its growth rate (responder: overall lesion shrinkage; inconclusive: 0% to 40% lesion growth; non-responder: more than 40% lesion growth). Characterization of global lesion growth rate patterns was accomplished by affinely registering all time-points to the MNI brain atlas. Segmented lesions were projected onto the atlas, which was qualitatively analyzed to identify spatial regions composed primarily of one class of lesion.

RESULTS

For automatic segmentation, we report a mean dice score of 0.778, 0.737, and 0.704 on training, validation, and testing sets respectively. Furthermore, we find that the largest BM with the highest average growth rate (non-responders) tend to be located in the posterior frontal/parietal lobes, while smaller, lower growth rate lesions (responders) tend to be localized in the frontal lobes. The posterior fossa was found to be heterogeneous in lesion size and growth rate.

CONCLUSION

We developed automatic metastatic lesion tracking over time-points and identified brain regions associated with differing growth rate lesions.

BSCI-10. NEUROLOGICAL DYSFUNCTION CAUSED BY BRAIN TUMOR-GENERATED SOLID STRESS IS REVERSED BY LITHIUM

The compression of brain tissue by a tumor mass is believed to be a major cause of the clinical symptoms seen in patients. However, the biological consequences of these physical stresses on the brain tissue are unknown. Using clinical imaging and preclinical studies, we discovered that a subgroup of primary and metastatic brain tumors, classified as nodular based on the growth pattern, exert compressive solid stress on the surrounding brain tissue, leading to a decrease in local vascular perfusion, as well as neuronal death and impaired function. We demonstrated a causal link between solid stress and neurological dysfunction, by applying and removing cerebral compression, mimicking the mechanics of tumor growth and surgical resection respectively. Finally, we showed that treatment with lithium reduced solid stress-induced neuronal death and improved motor coordination in mice. Our results indicate that brain tumor-generated solid stress impairs neurological function in patients and show lithium as a potential therapeutic intervention to counter these effects.