Alliance A071401: Phase II trial of FAK inhibition in meningiomas with somatic NF2 mutations.

2502 Background: Patients with progressive or recurrent meningiomas have limited treatment options. Clinical trials of systemic therapies for meningiomas have failed to demonstrate benefit. FAK inhibition has a synthetic lethal relationship with NF2 loss. Given the predominance of NF2 mutations in meningiomas, we evaluated the efficacy of GSK2256098, a FAK inhibitor, as part of the first genomically-driven phase II study in recurrent or progressive grade I-III meningiomas. Methods: Eligible patients (pts) whose tumors screened positively for NF2 mutations were treated with GSK2256098 750mg po bid until progressive disease in 2 separate cohorts: grade I or II/III meningiomas. Two co-primary endpoints were used: progression-free survival at 6 months (PFS6) and response rate (RR) by Macdonald criteria; per study design, the trial would be declared positive if either endpoint was met. RR was evaluated across the overall cohort; PFS6 was evaluated within each subgroup. Historical benchmark data was obtained from Kaley et al. Neuro Oncol 2014. In the grade I group, 12 evaluable pts provided >79% power to detect a PFS6 rate >65% (vs. null hypothesis of 25%; alpha=0.014). In the grade II/III group, 24 evaluable pts provided >85% power to detect a PFS6 >41.5% (vs. null 15%; alpha=0.02). The threshold for promising results for PFS6 was: 7+/12(grade I) and 8+/24(grade II/III) pts. For RR, 36 evaluable pts provided >94% power to detect RR >20% (vs. null 2.5%; alpha= 0.012). Results: Of 322 pts screened for all mutation cohorts of the study, 36 eligible and evaluable pts with NF2 mutations were enrolled. Across all grades, one pt had a partial response and 24 had stable disease as best response to treatment. In Grade I pts, the observed PFS6 rate was 83% (10/12 pts; 95% CI: 52-98%). In Grade II/III pts, the observed PFS6 rate was 33% (8/24 pts; 95% CI: 16-55%). The study met PFS6 efficacy endpoint both for the Grade I and the Grade II/III cohorts. Treatment was well tolerated. Only 7 patients had a maximum grade-3 adverse event that was at least possibly related to treatment; toxicities across these pts included: proteinuria (2), rash (1), pain (1), ALT (1), AST (1), cholecystitis (1), hypertriglyceridemia (1), apraxia (1), and lymphopenia (1) with no grade 4 or 5 events. Conclusions: GSK2256098 had excellent tolerability andresulted in an improved PFS6 rate in pts with recurrent or progressive NF2-mutated meningiomas. Trial endpoint was met. FAK inhibition warrants further evaluation in this patient population. Support: U10CA180821, U10CA180882; https://acknowledgments.alliancefound.org Clinical trial information: NCT02523014 .

Imaging Criteria in Neuro-oncology

The identification of more effective therapies for brain tumors has been limited in part by the lack of reliable criteria for determining response and progression. Since its introduction in 1990, the MacDonald criteria have been used in neuro-oncology clinical trials to determine response, but they fail to address issues such as pseudoprogression, pseudoresponse, and nonenhancing tumor progression that have arisen with more recent therapies. The Response Assessment in Neuro-Oncology (RANO) working group, a multidisciplinary international group consisting of neuro-oncologists, medical oncologists, neuroradiologists, neurosurgeons, radiation oncologists, and neuropsychologists, was formed to improve response assessment and clinical trial endpoints in neuro-oncology. Although it was initially focused on response assessment for gliomas, the scope of the RANO group has been broadened to include brain metastases, leptomeningeal metastases, spine tumors, pediatric brain tumors, and meningiomas. In addition, subgroups have focused on response assessment during immunotherapy and use of positron emission tomography, as well as determination of neurologic function, clinical outcomes assessment, and seizures. The RANO criteria are currently a collective work in progress, and refinements will be needed in the future based on data from clinical trials and improved imaging techniques.

Short Course Radiotherapy Concomitant with Temozolomide in GBM Patients: A Phase II Study

Purpose Despite recent advances, the prognosis of glioblastoma (GBM) remains poor. The aim of this study was to assess the efficacy and tolerability of multiple daily fraction radiotherapy performed with multiple temozolomide (TMZ) administrations in newly diagnosed patients with GBM. Methods This trial was a prospective, open-label, monocentric, nonrandomized, single arm, phase II study. The primary endpoint was the proportion of progression-free patients at 12 months, and the secondary endpoints were overall survival (OS) and toxicity. Thirty-five patients underwent two radiotherapy courses concomitant with TMZ after surgery. At each course, radiation was delivered 3 times daily, 2 Gy/fraction, for 5 consecutive days, and the total dose was 60 Gy; concurrent TMZ was administered in a total dose of 150-200 mg/m2/day. Results The primary endpoint failed to be applied; Macdonald criteria could be used in 16 (46%) patients with local or intracerebral recurrence (group A). In 12 patients, due to suspicion of radiation necrosis vs recurrence, Macdonald criteria were not applied (group B). The OS was 22 months, and OS probabilities at 12, 18, and 24 months were 82%, 59%, and 44%, respectively. Hematologic toxicities generally did not exceed grade 2. The quality of life and cognitive functioning did not significantly change between baseline and the first follow-up. In the multivariate analysis, necrosis and pseudoprogression were significant prognostic factors of OS. Conclusions To improve local control and OS, a more aggressive treatment schedule should be explored. The related higher necrosis risk and its implications regarding local control deserve further investigation.

Response Assessment in Neuro-Oncology (RANO): more than imaging criteria for malignant glioma

The introduction of antiangiogenic therapies for the treatment of malignant glioma and the effect of these agents on standard imaging studies were the stimuli for forming a small group of investigators to critically evaluate the limitations of the Macdonald criteria in assessing response to treatment. The initial goal of this group was to highlight the challenges in accurately determining the efficacy of therapeutic interventions for malignant glioma and to develop new criteria that could be implemented in clinical care as well as in the design and conduct of clinical trials. This initial Response Assessment in Neuro-Oncology (RANO) effort started in 2008 and over the last 7 years, it has expanded to include a critical review of response assessment across several tumor types as well as endpoint selection and trial design to improve outcome criteria for neuro-oncological trials. In this paper, we review the overarching principles of the RANO initiative and the efforts to date. We also highlight the diverse and expanding efforts of the multidisciplinary groups of investigators who have volunteered their time as part of this endeavor.

A novel, reproducible, and objective method for volumetric magnetic resonance imaging assessment of enhancing glioblastoma

Object Robust methodology that allows objective, automated, and observer-independent measurements of brain tumor volume, especially after resection, is lacking. Thus, determination of tumor response and progression in neurooncology is unreliable. The objective of this study was to determine if a semi-automated volumetric method for quantifying enhancing tissue would perform with high reproducibility and low interobserver variability. Methods Fifty-seven MR images from 13 patients with glioblastoma were assessed using our method, by 2 neuroradiologists, 1 neurosurgeon, 1 neurosurgical resident, 1 nurse practitioner, and 1 medical student. The 2 neuroradiologists also performed traditional 1-dimensional (1D) and 2-dimensional (2D) measurements. Intraclass correlation coefficients (ICCs) assessed interobserver variability between measurements. Radiological response was determined using Response Evaluation Criteria In Solid Tumors (RECIST) guidelines and Macdonald criteria. Kappa statistics described interobserver variability of volumetric radiological response determinations. Results There was strong agreement for 1D (RECIST) and 2D (Macdonald) measurements between neuroradiologists (ICC = 0.42 and 0.61, respectively), but the agreement using the authors' novel automated approach was significantly stronger (ICC = 0.97). The volumetric method had the strongest agreement with regard to radiological response (κ = 0.96) when compared with 2D (κ = 0.54) or 1D (κ = 0.46) methods. Despite diverse levels of experience of the users of the volumetric method, measurements using the volumetric program remained remarkably consistent in all users (0.94). Conclusions Interobserver variability using this new semi-automated method is less than the variability with traditional methods of tumor measurement. This new method is objective, quick, and highly reproducible among operators with varying levels of expertise. This approach should be further evaluated as a potential standard for response assessment based on contrast enhancement in brain tumors.

Response assessment of novoTTF-100A versus best physician’s choice chemotherapy in recurrent glioblastoma.

2080 Background: The NovoTTF-100A device emits tumor treating electric fields and was tested against Best Physician’s Choice (BPC) chemotherapy in a randomized phase III trial. We analyzed post hoc the characteristics of responders and non-responders in both cohorts. Methods: Macdonald criteria were used to determine tumor response and progression. Kaplan-Meier and Chi-squared statistics were computed for time to response, response duration, progression-free survival (PFS) with and without Simon-Makuch correction, and overall survival (OS). Prognostic factors were compared using the Wilconox rank sum test. Relative hazard rates for responders and non-responders were plotted. Results: The median response duration was 7.3 versus 5.6 months for NovoTTF-100A and BPC chemotherapy respectively (p=0.0009). Five of 14 NovoTTF-100A responders but none of 7 BPC responders had prior low-grade histology. The mean cumulative dexamethasone dose was 35.9 mg for responders versus 485.6 mg for non-responders in the NovoTTF-100A cohort (p<0.0001) as compared to 525.6 mg for responders and 431.0 mg for non-responders in the BPC cohort (p=0.9520). Hazard rate analysis showed delayed tumor progression in responders compared to non-responders. The Simon-Makuch conditional plot, which adjusted for unequal progression-free states, still showed longer PFS in responders than non-responders treated with NovoTTF-100A (χ2=11.5, P=0.0007) or BPC chemotherapy (χ2=5.2, P=0.0222). The median OS was 24.8 months for responders that is longer than 6.2 months for non-responders treated with NovoTTF-100A (χ2=25.7, P<0.0001). In the BPC chemotherapy cohort, the median OS was 20.0 months for responders and 6.8 months for non-responders (χ2=5.1, P=0.0235). There was strong Pearson correlation between response and OS in NovoTTF-100A (P<0.0002) but not in BPC cohort (P=0.2952). Conclusions: Response duration, adjusted Simon-Makuch PFS and OS favor NovoTTF-100A over BPC chemotherapy. Data on prior low-grade histology and dexamethasone dose suggest potential genetic and epigenetic determinants of NovoTTF-100A response. Clinical trial information: NCT00379470.

Bevacizumab in combination with TMZ in patients with recurrent GBM: Final OS and PFS analysis.

2087 Background: BEV provides a consistent clinical benefit in the treatment of relapsing GBM in terms of a delayed progression and increased median overall survival compared to historical controls. The aim of this study is to evaluate the efficacy and toxicity profile of the combination of BEV with dose dense TMZ, reporting the final results of PFS, OS and the toxicity experienced. Methods: A phase II multicenter, national, open-label study in pts diagnosed of recurrent GBM treated with BEV 10 mg/kg day q2w and TMZ 150 mg/m2 days 1-7 and 15-21 q28d d until disease progression or unacceptable toxicity or medical decision, as first line of treatment all pts received radiotherapy and at least 3 cycles of adjuvant TMZ. This study evaluates efficacy by PFS as primary endpoint and as secondary endpoints: OS, RR based on the adapted MacDonald criteria and toxicity profile (NCI CTC v3.0). Results: From June 10 to July 11, 32 evaluable pts were recruited in 8 sites. The baseline characteristics were as follows: 17 males and 15 females, median age 57.5 y (29-74), ECOG PS 0: 25%, PS 1: 50% and PS 2: 25%, 44% patients had gross total resection, 50% had subtotal resection and 6% biopsy only, MGMT promotor was methylated in 12 pts, unmeth in 6 pts and missing in 14 pts. The median number of TMZ or BEV cycles administered across all patients was 4 (TMZ range 1-9 and BEV range 1-25) The median PFS was 4.4 m [IC 95% (3.7 – 5.6)]. The 6m-PFS probability was 29.25%. The median OS (75% events) was 7.5 m[IC 95% (5.98 – 9.11)]. No significant association with MGMT status was found in terms of OS or PFS. BEV related AEs have been reported in 56.2% of the population being most of them mild or moderate. Grade 3-4 most frequent toxicity: lymphopenia 31% and fatigue 12.5%. Six of 32 pts were long term survivors, in this population the median PFS was 9.8 m [IC 95% (8.2 – 24.4)] and median OS (50% events) was 15.9 m [IC 95 %( 9.2 – NA)], no differences in baseline characteristics were identified in comparison with total population. The median number of TMZ cycles administered was 4 and median BEV cycles were 9. Conclusions: Although the combination don’t met the previous reported activity of BEV, 19% of patients had longer survivals which suggest the need to identify patients that benefit for this treatment. Clinical trial information: NCT01115491.