Search

PURPOSE The standard of care (SOC) for elderly or frail glioblastoma (GBM) patients is 40 Gy in 15 fraction radiotherapy. However, this regimen has a lower BED compared to the Stupp regimen, 60 Gy in 30 fractions. We hypothesize that isoeffective hypofractionated radiation of 52.5 Gy in 15 fractions (BED equivalent to Stupp) will have superior survival compared to standard of care. METHODS Elderly GBM patients treated with 52.5 Gy in 15 fractions were pooled from 2 phase II studies, 1 phase 1 and a prospective observation study. Overall survival (OS) and progression free survival (PFS) were defined as the time elapsed between surgery/biopsy and death from any cause or progression. Univariate and multivariate analyses were performed. RESULTS 62 newly-diagnosed patients were eligible for this analysis. Median follow-up was 10 months. The median OS and PFS was 10.3 and 6.9 mos, respectively. Patients with KPS ≥ 70 and < 70 had a median OS of 15.3 and 9.5 mos. No survival difference was seen between unmethylated and methylated patients with a median OS of 10.2 and 10.3 months, respectively. Multivariable analysis demonstrated that concurrent chemotherapy was an independent prognostic factor for improved PFS and OS. Grade 3 neurologic toxicity was seen in 2 patients (3.2%). CONCLUSION This is the first pooled, prospective analysis of elderly/frail GBM patients treated with dose-escalated hypofractionated radiation. Treatment was well tolerated and demonstrated excellent OS and PFS, exceeding that from prior elderly trials (Roa; 6.5 mo [poor KPS]/Perry; 9.3mo [good KPS]). This treatment regimen gives the elderly population an alternative to Stupp that is not de-escalating therapy. Future prospective trials are needed to validate these results.

INTRODUCTION Current adjuvant treatment for patients with newly diagnosed glioblastoma includes concurrent chemoradiation and maintenance temozolomide with Tumor Treating Fields (TTFields). We report our clinical trial evaluating feasibility and tolerability of scalp-sparing radiation with concurrent temozolomide and TTFields. METHODS Adult patients (age ≥ 18 years) with newly diagnosed glioblastoma with a KPS of ≥ 60 were eligible. All patients received concurrent scalp-sparing radiation (60 Gy in 30 fractions) with temozolomide (75 mg/m2 daily) and TTFields (200 kHz). Maintenance therapy included temozolomide and continuation of TTFields. Radiation treatment was delivered through TTFields arrays. The primary endpoint was safety and toxicity of tri-modality treatment within 30 days of completion of chemoradiation treatment. RESULTS Thirty patients were enrolled. Twenty were male and ten were female, with a median age of 58 years (range 19 to 77 years). Median follow-up was 10.8 months (range 1.6 to 21.3 months). Twenty (66.7%) patients had unmethylated MGMT promotor and ten (33.3%) patients had methylated promoter. Scalp dose constraints were achieved for all patients. Skin adverse events (erythema, dermatitis, irritation, folliculitis) were noted in 83.3% of patients, however, these were limited to Grade 1 or 2 events, which resolved spontaneously or with topical medications. No patient had radiation treatment interruption due to skin AEs. Other Grade 1 events included pruritus (33.3%), fatigue (30%), nausea (13.3%), headache (10%), dizziness (6.7%), and cognitive impairment (3.3%). Other Grade 2 events included headache (3.3%). The median PFS for the entire cohort was 9.1 months (at least 8.5 months, 95% confidence). The median PFS for patients with MGMT promoter methylation status was 11.4 months (at least 9.5 months, 95% confidence). Overall survival was not reached. CONCLUSIONS Concurrent TTFields with scalp-sparing chemoradiation is feasible treatment option with limited toxicity. Future randomized prospective trials are warranted to define therapeutic advantages of concurrent TTFields with chemoradiation.

PURPOSE Multi-parametric MRI and artificial intelligence (AI) methods were previously used to predict peritumoral neoplastic cell infiltration and risk of future recurrence in glioblastoma, in single-institution studies. We hypothesize that important characteristics of peritumoral tissue heterogeneity captured, engineered/selected, and quantified by these methods relate to predictions generalizable in the multi-institutional ReSPOND (Radiomics Signatures for PrecisiON Diagnostics) consortium. METHODS To support further development, generalization, and clinical translation of our proposed method, we trained the AI model on a retrospective cohort of 29 de novo glioblastoma patients from the Hospital of the University of Pennsylvania (UPenn) (Male/Female:20/9, age:22-78 years) followed by evaluation on a prospective multi-institutional cohort of 84 glioblastoma patients (Male/Female:51/33, age:34-89 years) from Case Western Reserve University/University Hospitals (CWRU/UH, 25), New York University (NYU, 13), Ohio State University (OSU, 13), University Hospital Río Hortega (RH, 2), and UPenn (31). Features extracted from pre-resection MRI (T1, T1-Gd, T2, T2-FLAIR, ADC) were used to build our model predicting the spatial pattern of subsequent tumor recurrence. These predictions were evaluated against regions of pathology-confirmed post-resection recurrence. RESULTS Our model predicted the locations that later harbored tumor recurrence with sensitivity 83%, AUC 0.83 (99% CI, 0.73-0.93), and odds ratio 7.23 (99% CI, 7.09-7.37) in the prospective cohort. Odds ratio (99% CI)/AUC(99% CI) per institute were: CWRU/UH, 7.8(7.6-8.1)/0.82(0.75-0.89); NYU, 3.5(3.3-3.6)/0.84(074-0.93); OSU, 7.9(7.6-8.3)/0.8(0.67-0.94); RH, 22.7(20-25.1)/0.94(0.27-1); UPenn, 7.1(6.8-7.3)/0.83(0.76-0.91). CONCLUSION This is the first study that provides relatively extensive multi-institutional validated evidence that AI can provide good predictions of peritumoral neoplastic cell infiltration and future recurrence, by dissecting the MRI signal heterogeneity in peritumoral tissue. Our analyses leveraged the unique dataset of the ReSPOND consortium, which aims to develop and evaluate AI-based biomarkers for individualized prediction and prognostication, by moving from single-institution studies to generalizable, well-validated multi-institutional predictive biomarkers.

PURPOSE Radiation treatment planning for meningiomas conventionally involves MRI contrast enhanced images to define residual tumor. However, the gross tumor volume may be difficult to delineate for patients with a meningioma in the skull base, sagittal sinus, or post resection. Advanced PET imaging using 68(GA)DOTATATE PET, which has been shown to be more sensitive and specific than MRI imaging, can be used for target volume delineation in these circumstances. We hypothesize that 68(GA)DOTATATE PET scan-based treatment planning will lead to smaller radiation volumes and will detect additional areas of disease compared to standard MRI alone. METHODS Our data evaluated retrospective, deidentified, and blinded gross tumor volume (GTV) contour delineation with 7 CNS specialists (3 neuroradiologists, 4 CNS radiation oncologists) for 26 patients diagnosed with a meningioma who received both a 68(GA)DOTATATE PET and an MRI for radiation treatment planning. Both the MRI and the PET were non-sequentially contoured by each physician for each patient. RESULTS The mean MRI volume for each physician ranged from 24.14-35.52 ccs. The mean PET volume for each physician ranged from 10.59-20.54 ccs. The PET volumes were significantly smaller for 6 out of the 7 physicians. In addition, 7/26 (27%) patients had new non-adjacent areas contoured on PET by at least 6 of the 7 physicians that were not contoured by these physicians on the corresponding MRI. These new areas would not have been in the traditional MRI based volumes. CONCLUSION Our study supports that 68(GA)DOTATATE PET imaging can help radiation oncologist create smaller and more precise radiation treatment volumes. Utilization of 68(GA)DOTATATE PET may find undetected areas of disease which in turn can improve local control and progression free survival. 68(GA)DOTATATE PET guided treatment planning should be studied prospectively.

PURPOSE Multi-parametric MRI based radiomic signatures have highlighted the promise of artificial intelligence (AI) in neuro-oncology. However, inter-institution heterogeneity hinders generalization to data from unseen clinical institutions. To this end, we formulated the ReSPOND (Radiomics Signatures for PrecisiON Diagnostics) consortium for glioblastoma. Here, we seek non-invasive generalizable radiomic signatures from routine clinically-acquired MRI for prognostic stratification of glioblastoma patients. METHODS We identified a retrospective cohort of 606 patients with near/gross total tumor resection ( >90%), from 13 geographically-diverse institutions. All pre-operative structural MRI scans (T1,T1-Gd,T2,T2-FLAIR) were aligned to a common anatomical atlas. An automatic algorithm segmented the whole tumors (WTs) into 3 sub-compartments, i.e., enhancing (ET), necrotic core (NC), and peritumoral T2-FLAIR signal abnormality (ED). The combination of ET+NC defines the tumor core (TC). Quantitative radiomic features were extracted to generate our AI model to stratify patients into short- (< 14mts) and long-survivors ( >14mts). The model trained on 276 patients from a single institution was independently validated on 330 unseen patients from 12 left-out institutions, using the area-under-the-receiver-operating-characteristic-curve (AUC). RESULTS Each feature individually offered certain (limited but reproducible) value for identifying short-survivors: 1) TC closer to lateral ventricles (AUC=0.62); 2) larger ET/brain (AUC=0.61); 3) larger TC/brain (AUC=0.59); 4) larger WT/brain (AUC=0.55); 5) larger ET/WT (AUC=0.59); 6) smaller ED/WT (AUC=0.57); 7) larger ventricle deformations (AUC=0.6). Integrating all features and age, through a multivariate AI model, resulted in higher accuracy (AUC=0.7; 95% C.I.,0.64-0.77). CONCLUSION Prognostic stratification using basic radiomic features is highly reproducible across diverse institutions and patient populations. Multivariate integration yields relatively more accurate and generalizable radiomic signatures, across institutions. Our results offer promise for generalizable non-invasive in vivo signatures of survival prediction in patients with glioblastoma. Extracted features from clinically-acquired imaging, renders these signatures easier for clinical translation. Large-scale evaluation could contribute to improving patient management and treatment planning. *Indicates equal authorship.

PURPOSE We investigated the long term impact of SRS and WBRT in two large prospective phase III trials. METHODS Patients with 1–4 BMs +/- resection were randomized to SRS or WBRT. Cognitive deterioration was a drop of >1 standard deviation from baseline in >2/6 cognitive measures (CM). Quality of life (QOL) scores were scored 0–100 point scale. CM and QOL scores were modeled using baseline adjusted Linear Mixed Models (LMM) with uncorrelated random intercept for subject and random slopes for time. Differences over time between groups and the effect of >2 cognitive scores with >2 SD change from baseline were assessed. RESULTS 88 patients were included with median follow up of 24 months. We observed decreasing CM over time (SRS: 4/6; WBRT: 5/6). Mean CM was significantly higher in SRS for Total recall and Delayed Recall at 3, 6, 9, 12 months. More patients in WBRT arm declined 1 SD in >1 and >2 CM at the 3, 6, 9, and 12 months. A 1 SD decline in >3 CM at 1 year was 21% SRS vs 47% WBRT (p=0.02). SRS had fewer patients with a 2 SD decline in >1 CM at every time point. SRS had fewer patients with a 2 SD decline at >2 and >3 CM. WBRT had lower QOL at 3 months, but switched to SRS having lower QOL at 24 months for PWB, EWB, FWB, FactG, BR, and FactBR (p< 0.05). A 2 SD decline in cognition decreased mean FWB by 6.4 units (95% CI: -11, -1.75; p=0.007) and decreased QOL by 5.1 units (95% CI: -7.7, -2.5; p< 0.001). CONCLUSIONS We report the first pooled prospective study demonstrating the long term outcomes of patients with BMs after cranial radiation. WBRT was associated with worse cognitive outcomes. Impaired cognition is associated with worse QOL.

BACKGROUND About one-third of children with primary CNS MMGCT experience incomplete responses to initial induction chemotherapy prior to irradiation, many of whom will subsequently relapse. Such high-risk patients are variably defined as having initial alpha-fetoprotein (AFP) elevations exceeding 1,000ng/mL, predominant histopathologies of malignant non-germinomatous GCT and incomplete responses to induction chemotherapy. Drugs targeting GCT-specific molecular markers have been identified for non-germinomatous GCT elements but have yet to be incorporated into prospective clinical trials. Four children with clearly identified HR-MMGCT characteristics have been treated on an innovative pilot regimen incorporating intensified chemotherapy and molecularly targeted agents, with avoidance or minimization of irradiation. METHODS Four children (two with pure suprasellar embryonal carcinoma (EC) - one with Down syndrome and the other with pre-diagnosis cognitive dysfunction; one with initial serum AFP exceeding 7,000ng/mL and yolk sac tumor (YST)+EC+Teratoma pathology; one with initial serum AFP exceeding 1,000ng/mL) were treated with 3 cycles of “standard” induction chemotherapy (ACNS1123), followed by 1–3 transplant cycles (thiotepa/carboplatin) each with complete radiographic and tumor marker responses. Two children with pure EC subsequently received six cycles of brentuximab-vedotin without irradiation and remain disease-free off therapy for 2–4 years. One child with YST+EC+Teratoma has subsequently received reduced dose craniospinal irradiation and pineal region boost, and will receive oral everolimus, erlotinib, palbocyclib and intravenous brentuximab-vedotin. The fourth child with YST+MT will commence everolimus, erlotinib and palbocyclib without irradiation. CONCLUSION This treatment strategy for HR-MMGCT patients provides preliminary tolerance and response data justifying extension to a multi-center trial.

INTRODUCTION Treatment-induced cerebral necrosis (TN) is a challenging complication encountered in neuro-oncology. Diagnosis and treatment of TN remains poorly defined. METHODS In this single institution, retrospective study, consecutive patients with gliomas and TN between 01/01/2012 and 04/20/2020 at the OSUCCC were identified. Details of the tumor treatment, molecular markers, radiological and pathological findings of TN, as well as treatment, recurrence rate and management upon recurrence were collected. RESULTS Of the 53 patients analyzed, 37 had glioblastoma, 7 had anaplastic oligodendroglioma and 9 had grade II or III astrocytoma. MGMT promoter hypermethylation was present in 31/50 (59%) and IDH mutation in 17/53 (32%). Diagnosis of TN was based on histology in 43/53 (81%) or clinical/radiographic features in 10/53 (19%). Worsening of focal weakness (36%), seizures (9%) or being (30%) were common presentations at TN diagnosis. Patient with right compared to left hemisphere involvement were more symptomatic at TN diagnosis. (p=0.049). Bevacizumab (BEV) (51%), resection (28%), steroids only (9%) or Laser Interstitial Thermal Therapy (6%) were used to treat TN. Steroids were weaned off in 20/27 (74%) after receiving BEV. Among all treatments, BEV was significantly associated with a better outcome (resolution or partial improvement of enhancement in 84.6%) (p=0.0006, Bonferroni corrected p< 0.005). TN Recurrence occurred in 36%, 70% and 100% of the patients treated with BEV, resection and LITT respectively. The median duration to TN recurrence was 10 weeks (range: 3–70 weeks). Initial treatment used for TN, MGMT methylation and IDH mutation status did not predict TN recurrence. (p=0.074; p=0.819; p=0.607 respectively). CONCLUSIONS BEV appears to be a superior treatment to control TN overall. Recurrence of TN in patients previously treated with BEV was 36%. There was no difference in the risk of developing recurrent TN based on MGMT or IDH status.

INTRODUCTION Radiation necrosis (RN) is a potential complication after radiation therapy to primary brain tumors and brain metastases. The pathophysiology of RN is not well understood but it is hypothesized that vascular endothelial growth factor (VEGF) plays an important role. Bevacizumab, a monoclonal antibody against VEGF-A, is often successful in the management of RN. The objective of this study is to assess whether VEGF receptor (VEGFR) inhibitors, a group of oral tyrosine kinase inhibitors (TKIs), can prevent or reverse cerebral radiation necrosis METHODS We retrospectively studied a cohort of 102 patients with renal cell carcinoma (RCC) and brain metastases seen at The Ohio State University James Cancer Center between 01/01/2011 and 04/30/2019. We identified those who developed RN and analyzed the temporal relationship between the use of VEGFR TKIs and the development of RN. RESULTS The cumulative incidence of RN in our cohort is 13.7% after radiation treatments that included LINAC-based stereotactic radiosurgery, fractionated stereotactic radiotherapy, or Gamma Knife radiosurgery. There was no statistically significant difference in the cumulative incidence of RN between patients taking TKIs and patients who were off TKIs (9.9% and 11.5% respectively, p= 0.741). The median time to development of RN was only numerically shorter in patients taking TKIs (151 versus 315 days, p=0.315). One patient developed RN after stopping cabozantinib. Three other patients developed RN while on cabozantinib. Two patients developed RN while on pazopanib, and 3 patients developed RN while on sunitinib. One patient was started on axitinib during active RN without significant improvement subsequently. CONCLUSIONS VEGFR TKIs do not consistently prevent or reverse cerebral radiation necrosis and do not seem to have the efficacy that bevacizumab has against RN.

Trastuzumab/pertuzumab improved outcomes for patients (pts) with HER2+ breast cancer. Increased survival coupled with limited blood-brain barrier (BBB) penetration of these agents apparently contributes to increased incidence of brain metastases (BM). Lapatinib (L) crosses the BBB and demonstrates activity against BM. Based upon pre/early clinical data, it’s hypothesized that L + whole brain radiotherapy (WBRT) or Stereotactic Radiosurgery (SRS) would improve intracranial control compared to WBRT/SRS (RT) alone. This randomized phase II trial included HER2+ breast cancer pts with ≥ 1 measurable, unirradiated parenchymal BM. Pts were randomized 1:1 to WBRT (37.5 Gy/3 weeks) or SRS +/- concurrent L (1000 mg dailyx6 weeks), andwere stratified by graded prognostic assessment (GPA), use of non-CNS penetrating anti-HER2 therapies, and previous SRS/resection. The primary endpoint was intracranial complete response (CR) rate 12weeks (wk) after RT. Secondary endpoints included objective response rate (ORR), lesion-specific response rate, CNS progression-free survival, and overall survival. 114 of 143 pts were evaluable for 12-wk CR (52 RT, 62 RT+L). Rate of grade 3 and 4 adverse events were 8% and 0% for RT and 29% and 6% for RT+L. 92% and 90% pts received concurrent and adjuvant L per protocol/acceptable variation. There were no significant differences in 12 or 4wk CR rates (p=0.97 and p=0.77); 5.8% and 0% at 12wk and 3.6% and 1.5% at 4wk for RT and RT+L, respectively. The ORR at 4 wk was 42% and 56% (p=0.059, RECIST), 40% and 58% (p=0.027, WHO) in the RT and RT+L, respectively. Although the addition of lapatinib to WBRT/SRS did not improve the 12wk CR rate, it showed meaningful clinical benefit by demonstrating a trend toward improvement of 4 wk ORR. Results of ongoing central review of MRIs using RECIST1.1, WHO and RANO-BM criteria including secondary endpointswill be reported. Support: NCI grants U10CA180868, U10CA180822, UG1CA189867, and Novartis

BACKGROUND/OBJECTIVE: Gliosarcoma is a WHO grade IV histologic variant of glioblastoma that occurs very rarely among pediatric patients, even more so amongst Neurofibromatosis Type 1 (NF1) individuals. Patients with NF1-associated high-grade gliomas (HGGs) reportedly have better clinical outcomes than others despite their germline tumor-suppressor mutation. Since molecular data to explain this are lacking, we compared tumor exomes from two pediatric gliosarcoma patients, one with NF1 (P1) and one without (P2). PATIENTS/METHODS: Two patients with gliosarcoma (ages 12 and 11 years), were treated with resection (gross total for P1 and near-total for P2), bevacizumab, temozolomide and irradiation. We performed whole-exome sequencing (WES, Agilent SureSelect V6 + Cosmic) on tumor and blood to determine both germline and somatic mutations. Tumor mutations were filtered based on predicted deleterious effects and a list of genes was compared to online pediatric tumor databases (PeCan and PedcBioPortal). RESULTS: P1 is alive without progression after 5.8 years. P2 developed tumor recurrence after 2 years, expiring 4.6 years from diagnosis. Tumor DNA revealed a lower mutational burden in P1 (NF1) than in P2 including some genes which have been previously reported in pediatric HGGs (e.g. IKBKB in P2). However, there was no overlap in specific gene mutations. CONCLUSIONS: This is the first report of WES in pediatric gliosarcoma. The differences in mutational burden between these patients might explain improved survival outcomes for HGG patients with NF1 as well as shedding light on the different mutational processes leading to their tumors. Further molecular analyses including methylation profiling are ongoing.

BACKGROUND: Recurrent glioma has a poor prognosis and largely recurs within the irradiated volume. Improvements in radiosensitization are needed. The utility of concurrent temozolomide with hypofractionated radiotherapy for recurrent glioma is undefined. METHODS: 103 consecutive patients treated with re-irradiation (re-RT) for recurrent glioma between 2010-2014 were analyzed. Patients were excluded if they had incomplete records (n = 20) or initially treated for LGG or AO (n = 21). 62 met inclusion criteria and separated into two groups: re-RT plus concurrent TMX (n = 17) and Re-RT alone (n = 45). OS and survival from re-RT were estimated using Kaplan Meier plots and compared using Log-rank tests. RESULTS: The median age was 55.5 yo for re-RT alone and 56 yo for the re-RT + TMZ. The median KPS was 85 and 90, for the re-RT alone and re-RT + TMZ groups, respectively. Gr3 histology comprised 35% of patients in the re-RT group and 12% in the re-RT + TMZ group. Additional therapy (chemotherapy and/or bevacizumab) was given in 32 % of patients in the re-RT group and 47% of the RT + TMZ group. 51% had repeat surgery in RT alone group and 6% in re-RT + TMZ group. The median dose of RT was 35 Gy in 10 fractions. Oral temozolomide was administered 75 mg/m2 for fourteen days during RT. OS for patients receiving re-RT alone was 21.8 months compared to 24.7 months for patients in the re-RT + TMZ group (p = 0.03, HR 0.5 [0.3-0.9]). Survival from re-RT was 17.5 months for the re-RT +TMZ group and 7 months for the re-RT group (p = 0.0025, HR 0.4 [0.2-0.7]. CONCLUSIONS: Survival was significantly higher in the group receiving re-RT + TMZ compared to re-RT alone. The cohort of patients receiving re-RT + TMZ was small and there may be selection bias. Re-RT with concurrent TMZ may be a viable option for well selected patients. Further studies are needed.