91ème dossier Asco 2011
2091-Chemin radiographique de progression pour glioblastome récurrent sous Avastin (bevacizumab) avec ou sans Campto (iritonecan)
Author(s): P. Dory-Lautrec, E. Tabouret, M. Barrie, C. Boucard, M. Matta, D. Autran, N. Girard, O. L. Chinot; University Hospital Timone, Marseille, France

Résumé :
Bevacizumab (BEV) avec ou sans irinotecan (IRI) ont démontré une activité chez les malades avec glioblastome GBM récurrent. Cependant, l'augmentation des résultats est incompatible avec l'implication de BEV dans la progression infiltrante associée avec une survie plus courte. Pendant que Pope et all. (Neuro-Oncol, 2009) propose 4 modèles de progression de la maladie. Il apparait qu'aucun de ces 4 modèles ne puisse identifier complètement la progression du phénotype de gliomatose cérébrale. Nous avons donc testé une analyse anatomique plus étendue de la gliomatose cérébrale comme décrit par Peretti, J Neuro-Oncol, que 2002) 
Méthodes :
Des IRMs ont été effectuées tous les 2 mois chez 25 malades avec glioblastome GBM récurrent traités avec BEV+IRI, par analyse rétrospective. Le statut (local, multifocal, diffus), le nombre de lobes impliqués, et l'infiltration de la tumeur dans le corps calleux (CC), les noyaux gris centraux ou le thalamus (BG) et le tronc cérébral ont été analysé au temps d'initiation BEV-+IRI jusqu'à progression. 
Resultats : 
Au temps d'initiation BEV+IRI, 9/25 patients (36%), 11/25 patients (44%), et 5/25 patients (20%) ont présenté un phénotype local (L1), multifocal (M1), et diffus (D1) respectivement. Selon les critères RANO, 7/25 patients (28%) ont présenté une réponse objective. Au temps de la progression, sur les 9 patients phénotype local L1, 2 ont progressé comme diffus (DP), 2 comme multifocal (MP) et 5 sont restés local (LP). Sur les 11 multifocal M1, 3 ont progressé en diffus (DP), et 8 ont gardé leur phénotype multifocal. Tous les patients avec le phénotype diffus D1 à l'initiation du traitement ont conservé leur phénotype diffus. Dans l'ensemble, 10/25 patients ont progressé avec un phénotype diffus qui a impliqué 4 à 6 lobes et, dans 8/10 cas, les noyaux gris centraux ou le thalamus et le tronc cérébral ou le corps calleux. La survie au temps de la progression sous BEV était inférieure à 3 mois chez 6/10 patients, pendant que la survie de M1 ou L1 était <3 mois chez 4/15 patients.
Conclusions : 
Le glioblastome récurrent GBM présentent fréquemment un phénotype diffus ou multifocal alors que le changement du phénotype local vers celui de diffus sous Avastin n'est pas commun. Quand il est observé, le phénotype diffus s'étend aux deux lobes, traverse la ligne médiane et peut être associé à une survie plus courte. L'analyse de l'infiltration anatomique avant l'administration BEV et au temps de la progression devra être considérée dans les essais ultérieurs.

92ème dossier Asco 2011
2092-Résultats cliniques avec radio-chimiothérapie de Temodal et Temodal adjuvant sur les gliomes de haut grade en Colombie
Author(s): L. D. Ortiz, A. F. Cardona, C. E. Fadul, A. Londono, H. A. Becerra, E. Jimenez-Hakim, C. J. Yepes, H. Carranza, R. E. Bruges Maya, C. J. Castro, C. A. Vargas; Instituto de Cancerología, Clínica las Américas, Medellin, Colombia; Fundación Santa Fe de Bogotá, Bogotá, Colombia; Darthmouth-Hitchcock Medical Center, Lebanon, NH; Latin American Neuro-Oncology Network - RedLANO, Bogota, Colombia; Fundacion Santa Fe de Bogota, Bogota, Colombia; Instituto de Cancerologia, Clinica las Americas, Medellin, Colombia; Instituto Nacional De Cancerología, Bogota, Colombia

Résumé :
L'utilisation de la radiothérapie et du Temodal pendant et après la radiothérapie (Stupp ) dans les gliomes de haut grade HGG a été montré pour accroître la survie à 5 ans. Pour comparer ces résultats avec nos pratiques courantes, un régistre bidirectionnel a été monté par RedLANO. 
Méthodes :
171 patients avec gliome de haut grade HGG ont été inscrits. Le temps à progression (TTP) et la survie totale ont été estimés.
Résultats :
L'âge médian est de 56 ans, (gamme 17-84 ans) and 97/171 patients sont des hommes (56%). 82% de glioblastome GB dont 62% de tumeurs primaires. 55% des patients ont subi la chirurgie, 23% une biopsie. 77% des malade ont complété le protocole Stupp SP, avec une médiane de 58,2 Grays (gamme 18-64 grays) et suivi un nombre moyen de cycles de Temodal TMZ de 5 (SD + / -3). La pseudoprogression a été trouvée chez 20% des patients. La survie totale médiane était de 15,8 mo (gamme 11,9-19,7 mois) et le TTP était de 9,5 mois (gamme 8,3-12,7 mois). Les taux de survie à 1 an et 2 ans étaient de 69% et 31%. Dans l'analyse rétrospective, les malades qui avaient un KPS postopératoire supérieur à 70 ont montré une survie meilleure aussi bien qu'une meilleure classification RPA (p=0.04) et sans plusieurs foyers. 71/171 patients (41%) ont été traités en 2de ligne, à la récurrence par Avastin (Bev) pour 32 patients, Temodal journalier TMZ pour 24 patients; BICNU pour 14 et divers pour 2.
Conclusions : 
Cette étude est la plus grande étude de radio-chimiothérapie avec Temodal concomitant et adjuvant TMZ sur gliome de haut grade HGG dans le monde latino-américain. Les résultats moyens de survie dans cette étude sont comparables aux résultats observés dans les études antérieures.
térieurs.

93ème dossier Asco 2011
2093-Facteurs pronostiques pour patients âgés avec gliomes malins anaplasiques de grade III
Author(s): A. Satra, N. Hashemi-Sadraei, H. S. Bawa, D. M. Peereboom, G. Stevens, L. A. Rybicki, J. H. Suh, M. A. Vogelbaum, R. Weil, G. Barnett, M. S. Ahluwalia; Cleveland Clinic Foundation, Cleveland, OH; Cleveland Clinic, Cleveland, OHIO Etats-Unis

Background: Anaplastic astrocytoma (AA), anaplastic oligodendroglioma (AO), and anaplastic oligoastrocytoma (AOA) are the major histological categories of WHO grade 3 gliomas. There is very limited data on specific prognostic factors in elderly patients (pts) with grade 3 gliomas, therefore we undertook this project to identify prognostic factors in elderly grade 3 glioma pts. Methods: After obtaining IRB approval, the Cleveland Clinic Brain Tumor and Neuro-Oncology Center's database was used to identify pts with histologically confirmed grade 3 glioma ≥65 years of age at time of diagnosis. Multivariable analysis was conducted with use of a Cox proportional hazards model and a stepwise selection algorithm that used p<.10 as the criteria for entry and p<0.05 as retention in the model to identify independent predictors of survival. Results: Charts of 96 pts (52% of whom were men) diagnosed between 1994 and 2009 were reviewed. The median age at presentation was 71 years (range: 65-88 years). 65% of pts had biopsy only, 15 % had gross total resection (GTR), 1% had near total resection (NTR), and 19% had subtotal resection (STR). 92% of pts underwent radiation, 48% had concurrent chemotherapy. Median overall survival (OS) of AA, AOA, AO was 5.7 months, 22.5 months and 65.2 months respectively. Median OS for all pts was 6.7 months. On multivariable analysis, two factors were identified as independent predictors of overall survival, histology (AA vs. AO, p<.022) and Karnofsky Performance Status (p<.018). Conclusions: Poor performance status and AA were associated with higher risk of mortality.

94ème dossier Asco 2011
2094-Thérapie photodynamique (PDT) traitement multidisciplinaire des tumeurs de cerveau, étude de faisabilité
Author(s): M. S. Gonzalez, V. Vanaclocha, I. Azinovic, J. Rebollo, R. Cañon, E. Calvo, A. Brugarolas; Plataforma de Oncología, USP-Hospital San Jaime, Torrevieja, Spain; Hospital 9 de Octubre, Valencia, Spain; START Madrid, Centro Integral Oncológico Clara Campal, Madrid, Spain; Plataforma de Oncología, Torrevieja, Spain

Background: Usual treatment of brain tumors includes surgery, radiotherapy and chemotherapy, but the general prognosis is bad, with local relapse as the main cause of death. PDT is a local anticancer treatment that has showed promising results in other tumors (GI, H&N, etc) and can be administered in addition to the classical modalities. Methods: Patients (pts) diagnosed of primary, relapsed or metastatic (mets) brain tumors received an endovenous photosensitizer agent (Photofrin or Foscan). After a 2-4 days period, the tumor was removed. The residual cavity was exposed to a special laser during a predetermined time, in order to achieve the optimal activation of the photosensitizer that, once activated, kills the residual tumor cells. Complementary treatment with radio and chemotherapy was also administered after surgical recovery. Results: From December 2000 to January 2011, 51 pts (34 M/ 17 F; age 13-73, median 51) were treated with a total of 61 procedures (22 Photofrin, 39 Foscan). Diagnoses were glioblastoma (20), gliosarcoma (1), anaplastic astrocitoma (9), grade 2 astrocitoma (7), oligodendroglioma (6, 2 anaplastic), malignant meningioma (1), lung ca. mets (4), breast ca. mets (3). 18/44 primary CNS tumor pts had PDT at the initial resection procedure and 32 procedures were performed at the time of relapse. In 10 cases a second procedure was performed after relapse. Toxicity: 2 pts with thalamic gliomas died in the first 24 h; 6 pts presented postoperative neurological alterations, recovering in 3 to 17 days, and 4 pts presented cutaneous ulcerations related to the PDT, one of them requiring plastic surgery. Median OS of the treated primary brain tumors was 13 m (0-120+) and 8 pts are alive at 1+, 71+, 92+, 93+, 96+, 98+, 104+, and 120+ m (1 GBM, 1 AA, 3 grade 2 A, 3 OD).Conclusions: PDT can be safely used in the multidisciplinary treatment of brain tumors. In this group of heavily treated pts, it clearly benefits some of them, showing an improvement in OS over expected from historical data. Further studies are warranted.

95ème dossier Asco 2011
2095-Le rôle du plasma GFAP comme biomarqueur des glioblastomes
Author(s): H. Husain, W. Savage, A. Everett, X. Ye, C. Blair, K. E. Romans, C. Bettegowda, P. Burger, S. A. Grossman, M. Holdhoff; The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University, Baltimore, MD; The Johns Hopkins University School of Medicine, Baltimore, MD; Johns Hopkins University School of Medicine, Baltimore, MD; Ludwig Center for Cancer Genetics and Therapeutics, The Johns Hopkins University School of Medicine, Baltimore, MD

Background: Magnetic resonance imaging (MRI) of the brain primarily assesses blood brain barrier (BBB) dysfunction which provides indirect information regarding tumor size. The limitations of MRI have become evident with the use of temozolomide (pseudo-progression) and VEGF based therapies (pseudo-response). As a result, a blood-based marker to assess tumor burden is increasingly important. Previous studies suggested that plasma glial fibrillary acidic protein (GFAP) levels are sensitive and specific for glioblastoma (GBM). This pilot study examines the utility of GFAP as a biomarker in pre- and post-operative gliomas. Methods: Plasma samples were collected in patients pre-operatively (N=34) and 24-48 hours post-operatively (N=24) for what was post-operatively confirmed to be newly diagnosed glioma. Plasma GFAP was detected using an electrochemiluminescent immunoassay with a detection threshold of >0.04ng/ml. Presence or absence of contrast enhancement (CE) on the pre-op MRI was compared with the pre-op plasma GFAP level using Fisher’s exact test.Results: The histologic, radiographic, and detection rates and pre- and post-op plasma GFAP levels are presented below. Higher grade tumors more often had CE on MRI than lower grade tumors (p=0.002). Plasma GFAP was detectable in 80% of patients with CE and 21% without CE (p=0.001). Conclusions:Higher grade tumors had more CE and higher plasma GFAP levels pre-operatively. However, post-operative GFAP levels dramatically increased in patients with lower grade gliomas that previously had undetectable levels. These data strongly suggest that GFAP is not a specific biomarker for GBM or an accurate measure of tumor burden. In addition, injury to the brain or BBB from surgery results in striking elevations of plasma GFAP independent of histologic grade or prior contrast enhancement.

96ème dossier Asco 2011
2096-
Non publié
97ème dossier Asco 2011
2097-O6-méthylguanine-méthyltransférase (MGMT) dans le glioblastome, analyse à la 1er chirurgie et suivantes
Author(s): F. Bertolini, C. Baraldi, E. Zunarelli, A. Valentini, R. Depenni, A. Falasca, F. Bertoni, F. Cavalleri, A. Chiari, A. Fontana, P. F. Conte, G. Pinna, Gruppo Neuro-Oncologico Modena (GNO-MO); Oncology, University Hospital, Modena, Italy; Pathology, University Hospital, Modena, Italy; Neurosurgery, Nuovo Ospedale Civile S. Agostino-Estense, Modena, Italy; Radiotherapy, University Hospital, Modena, Italy; Neuroradiology, Nuovo Ospedale Civile S. Agostino-Estense, Modena, Italy; Neurology, Nuovo Ospedale Civile S. Agostino-Estense, Modena, Italy

Background: Glioblastoma (GBM) is the most common malignant primary brain tumor in adults. Outcome has recently improved with the introduction of post-operative concomitant chemoradiotherapy (CRT) followed by 6 cycles of temozolomide (TMZ). MGMT expression has already been encoded as an independent prognostic factor and predictive for TMZ sensitivity, but few data are available on the modifications of MGMT status after CRT. Methods: From January 2005 to November 2010 at Modena University Hospital and Nuovo Ospedale Civile S. Agostino-Estense, 15 consecutive patients (pts) with GBM and treated with at least 2 subsequent surgeries were included in this retrospective study. Aim of the present analysis is to evaluate if methylation status may change between first and following determinations. Results: Fifteen pts (M/F=10/5; median age: 48, range: 40-67 years) underwent a first surgery for GBM. MGMT status was: methylated in 5 (33%); unmethylated in 10 pts (77%). Two unmethylated pts presented an early relapse and underwent a second surgery after 41 and 46 days, respectively. MGMT analysis confirmed the absence of methylation. One pt with multifocal GBM was treated after the first surgery on right frontal lobe with TMZ, operated on left frontal lobe, submitted to CRT, then re-operated for a relapse in the first site of surgery (right frontal lobe). In this case MGMT status evaluated in the site of relapse (right frontal lobe) changed from unmethylated to methylated. MGMT on the lesion of left frontal lobe was unmethylated. The remaining 12 pts underwent post-operative concomitant CRT followed by monthly TMZ (median number of cycles: 8.5; range: 2-24). MGMT status on second surgery was: 5 methylated (42%) and 7 unmethylated (58%), with a 100% concordance. Two pts received a third surgery: in one case MGMT status changed from unmethylated to methylated. Overall median survival was 15.8 months. The two patients with “MGMT switch” have a median survival of 35.8 months. Conclusions: in 2 on 13 cases (15%) after CRT, MGMT status changed from unmethylated to methylated and these pts have a median overall survival similar to methylated pts. This observation may be confirmed

Background: The prognosis for newly-diagnosed GBM remains poor. Adding temozolomide to radiation therapy improved the median event-free survival (EFS) to 6.9 months and median overall survival to 14.6 months. The five year survival remains < 10%. GBM’s have abundant neo-vascularization and the highest level of vascular endothelial growth factor (VEGF). Bevacizumab is an antibody to VEGF and is the most active agent for recurrent GBM. Hypoxia inducing factor-1 α (HIF-1 α) is an important regulator of VEGF, and topotecan may inhibit HIF-1 α. We performed a phase II trial in newly diagnosed GBM patients, adding bevacizumab and topotecan to standard therapy. Methods: Eighty newly diagnosed GBM patients were enrolled after their craniotomy between December 2009 and December 2010. Patients received standard radiation therapy and temozolomide at 75 mg/m2/d beginning between 2-6 weeks post-craniotomy. Bevacizumab, 10 mg/kg every 14 days was added a minimum of 4 weeks post-op. Two weeks after radiation therapy was completed, 12 monthly cycles of temozolomide at 150-200 mg/m2/d days 1-5, oral topotecan at 1.5 mg/m2/day days 2-6 for patients not on an enzyme inducing anti-epileptic drug (EIAED) and 2.0 mg/m2/day days 2-6 for patients on an EIAED bevacizumab at 10 mg/kg on days 1 and 15. Results: Adding bevacizumab and topotecan to standard therapy resulted in a 6-month EFS of 90%. The median overall survival has not been reached, but 96% of the patients are alive at 8 months. The regimen was tolerable, 96% completed radiation therapy. Eight patients came off due to toxicity, there were 2 CNS hemorrhages, 2 grade 4 thrombocytopenias, 1 pulmonary embolus, 1 colon perforation, 1 craniotomy bone flap infection, and 1 CMV pneumonitis. There were 2 toxic deaths, 1 from CMV pneumonits and 1 from CNS hemorrhage. There are 10 progressions. A comprehensive analysis of tumor biomarkers is underway. Conclusions: Adding bevacizumab to temozolomide and radiation followed by temozolomide, bevacizumab and oral topotecan is tolerable. The 6-month EFS is encouraging. Randomized phase III trials of the addition of bevacizumab to the treatments of newly diagnosed GBM patients are essential.

99ème dossier Asco 2011
2099-Impact des traitements sur la sexualité, la fertilité et la fonction sexuelle chez les adultes avec tumeurs cérébrales
Author(s): E. M. Dunbar, S. Savona, N. Ferree, P. A. Pumphrey; University of Florida, Gainesville, FL

Background: There is an unmet need to understand how anti-neoplastic treatments impact the sexuality, fertility, & sexual function of adult brain cancer patients in both the immediate & short-term future. For ethical & practical reasons, the paucity of literature is predominantly retrospective & addresses either non-CNS cancers or adult sequela of childhood cancer treatment. The authors sought information immediately impacting “real-time” treatment planning for adult brain cancer patients.Methods: Search criteria included English-written PubMed articles before 12-2010 regarding adults >13yo. Subject/ key terms were combined to refine parameters: “neoplasm/cancer”, “infertility/fertility”, “sexual dysfunction/function/behavior”, “sexuality”, “brain”, “drug therapy/anti-neoplastic/chemotherapeutic/biologic/anti-angiogenic”, & common CNS chemotherapies. Articles focusing on “pituitary neoplasm/cancer” & surgery/radiation were excluded. Cross-reference with CINAHL & Web of Science provided no significant differences. Results: (manual when needed): 300+ citations (0 Meta/Systematic Reviews, 2 RCTs, 18 Reviews, 198 Cases/Other. Articles predominantly addressed sexual organ or other systemic cancers, childhood cancer survivors, & long-term sequela. No articles focused on adult brain cancer patients treatments nor the immediate/short-term impact of common CNS chemotherapies used in adult non-CNS cancers. Major themes included the scope of the problem, barriers to evaluation/treatment, & need for more research. 84 addressed spontaneous fertility after treatment, 86 addressed fertility-sparing/restitution success, & 13 addressed emotional sequela. Only 1 focused-group & 1 survey addressed information needs/barriers of patients at diagnosis.Conclusions: This search highlights the need for interdisciplinary research regarding information immediately impacting “real-time” treatment planning for adult brain cancer patients. Practical patient & provider resources, updated evidence, including pharmacology/toxicology cross-

Background: There are minimal treatment options for advanced inoperable neurofibromatosis type 2(NF2), a rare and debilitating disorder presenting with multiple vestibular schwannomas, meningiomas, and ependymomas. Since NF2 patients harbor an aberration in a single gene (merlin), whose protein product impacts multiple signals, including PI3-kinase/Akt, Raf/MEK/ERK and mTOR, it is conceivable that one of the many agents targeting these pathways that have already shown antitumor activity in malignancies, will also cause regression of NF2-related tumors. Because of its extreme rarity, large clinical trials are challenging; we therefore enrolled patients in a variety of rationally targeted trials and sought response signals. Methods: We reviewed the records of consecutive NF2 patients referred to the Clinical Center for Targeted Therapy (Phase I Clinical Trials Program clinic) who were treated on ≥1 trial starting in January 2007. Patients were evaluated every 6-8 weeks for response using RECIST criteria and CT / MRI. Results: Six patients (men: N = 3) were enrolled on an early clinical trial. Median age at diagnosis was 26 (range 8-47). Two pts were treated with a RAS inhibitor (salirasib), and both achieved stable disease (SD) for 10 and 48+ months. One patient had SD on a MEK inhibitor for 7 months and was subsequently enrolled on several target agent-based studies including multi-kinase+HDAC, EGFR + HDAC, or EGFR inhibitor alone and has ongoing SD with a VEGF antibody (bevcizumab) for 22+ months. Two patients treated with combined bevacizumab and mTOR inhibitor show SD for 4+ and 5+ months, with the latter individual demonstrating, so far, a 9% decrease in tumor size on scans, improvement in neurologic symptoms and almost complete flattening of subcutaneous lesions. Conclusions: Referral of NF2 patients to a clinical trials center for targeted therapy demonstrates an acceptable safety profile, preliminary evidence of activity, and is pragmatic in this rare disease setting. Further evaluation of several of these therapies is warranted.

FIN