ASCO 2014 : abstracts

juin 11, 2014
3:44

Un grand nombre d’équipe du CARPEM a présenté des abstracts à l’ASCO

Colorectal Cancer Subtyping Consortium (CRCSC) identification of a consensus of molecular subtypes.

Author(s): Rodrigo Dienstmann, Justin Guinney, Mauro Delorenzi, Aurelien De Reynies, Paul Roepman, Anguraj Sadanandam, Louis Vermeulen, Andreas Schlicker, Edoardo Missiaglia, Charlotte Soneson, Laetitia Marisa, Krisztian Homicsko, Xin Wang, Iris Simon, Pierre Laurent-Puig, Lodewyk F. A. Wessels, Jan Paul Medema, Scott Kopetz, Stephen Henry Friend, Sabine Tejpar, Colorectal Cancer Subtyping Consortium; Sage Bionetworks, Seattle, WA; SIB Swiss Institute of Bioinformatics, Ludwig Center for Cancer Research and Department of Oncology, University Lausanne, Lausanne, Switzerland; Ligue Nationale contre le Cancer, Paris, France; Agendia NV Research and Development Dpt, Amsterdam, Netherlands; The Institute of Cancer Research, London, United Kingdom; Cancer Research UK – Cambridge Institute, University of Cambridge, Cambridge, United Kingdom; Division of Molecular Carcinogenesis, The Netherlands Cancer Institute, Amsterdam, Netherlands; SIB Swiss Institute of Bioinformatics, Lausanne, Switzerland; Swiss Institute for Experimental Cancer Research, Swiss Federal Institute of Technology Lausanne, Department of Oncology Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland; Centre for Biomedical Informatics, Harvard Medical School, Boston, MA; Paris Descartes University, Paris, France; Department of Molecular Carcinogenesis, Netherlands Cancer Institute, Amsterdam, Netherlands; LEXOR, Academic Medical Center, Amsterdam, Netherlands; The University of Texas MD Anderson Cancer Center, Houston, TX; University of Leuven, KUL, Leuven, Belgium

Background: Recently, a number of independent groups reported novel molecular subtypes in colorectal cancer (CRC). A formal comparison across these classifiers is needed to reconcile findings and accelerate clinical translation. The CRCSC was formed to identify a consensus among the subtyping systems through large scale data sharing and meta-analysis. Methods: The CRCSC consists of 6 groups (15+ institutions) that analyzed more than 30 patient cohorts with gene expression data, spanning multiple platforms and sample preparation methods. Each of the 6 classifiers (with 3-6 subtypes) was applied to the collection of public and proprietary datasets encompassing over 4,000 samples, mostly stage II-III CRC. Concordance of subtype calls and associations with clinical, molecular and pathway features were assessed centrally by an independent team. Results: Despite heterogeneities in cohorts and methods, subtype concordance analysis readily yielded a clear consensus on 4 CRC molecular subtypes (CMS1-4), with significant interconnectivity among the calls from the participating groups. The remaining 16% of samples did not have a consensus assignment, which may be partly explained by an additional mixed subtype with variable epithelial-mesenchymal activation; further refinement is needed. Conclusions: This is the first example of a large-scale, community based comparison of cancer subtypes. Within the largest collection of CRC samples we identified recurrent signals of 4 biologically distinct subtype classes enriched for key clinical, pathway and molecular traits. Ongoing efforts are attempting to improve the granularity of these subtypes.

CMS1 14% MSI, immune pathway activation/expression, right-side tumors, older age at diagnosis, females, hypermutation, BRAF mut, intermediate survival
CMS2 41% High CIN, MSS, strong WNT/MYC pathway activation, left-side tumors, TP53 mut, EGFR amplification/overexpression, better survival
CMS3 8% Low CIN, moderate WNT/MYC pathway activation, KRAS mut, PIK3CA mut, IGFBP2 overexpression, intermediate survival
CMS4 20% CIN/MSI heterogeneous, mesenchymal/TGF-beta activation, younger age at diagnosis, NOTCH3/VEGFR2 overexpression, worse survival

Biomarker-driven access to crizotinib in ALK-, MET-, or ROS1-positive malignancies in adults and children: Feasibility of the French National Acsé Program

Citation: J Clin Oncol 32:5s, 2014 (suppl; abstr TPS2647)

Author(s): Gilles Vassal, Natalie Hoog Labouret, Marie-Cécile Le Deley, Marta Jimenez, Frederique Nowak, Denis Moro-Sibilot, Yann Godbert, Luc Taillandier, Jean-Yves Blay, Herve R. Bonnefoi, David Malka, Roch Houot, Thomas Aparicio, Bernard J. Escudier, Christophe Tournigand, Isabelle Ray Coquard, Pierre Laurent-Puig, Frederique Madeleine Penault-Llorca, Fabien M. Calvo, Agnes Buzyn, INCa Molecular Genetic Centers Taskforce; SFCE, Gustave Roussy, Cancer Campus, Villejuif, France; French National Cancer Institute (INCa), Boulogne-Billancourt, France; Gustave Roussy, Cancer Campus, Villejuif, France; UNICANCER, Paris, France; IFCT, Centre Hospitalier Universitaire de Grenoble, Grenoble, France; TUTHYREF Network, Institut Bergonié, Bordeaux, France; ANOCEF Group, Centre Hospitalier Universitaire Poitiers, Poitiers, France; NetSarc Intergroup, Centre Léon Bérard, Lyon, France; Breast Cancer Intergroup (UCBG), Institut Bergonié, Bordeaux, France; Unicancer GastroIntestinal Group (UCGI), Gustave Roussy, Cancer Campus, Villejuif, France; LYSA Group, Centre Hospitalier Universitaire Pontchaillou, Rennes, France; FFCD Group, Hôpital Avicenne, AP-HP, Bobigny, France; GETUG, Gustave Roussy, Cancer Campus, Villejuif, France; GERCOR Group, Hôpital Henri Mondor, AP-HP, Creteil, France; GINECO Group, Centre Léon Bérard, Lyon, France; Hôpital Européen Georges Pompidou, AP-HP, Paris, France; Centre Jean Perrin, Clermont-Ferrand, France

Background: Crizotinib (czb) is registered only for the treatment of patients (pts) with ALK+ lung cancer. Czb targets (ALK, MET, ROS1) are also altered (translocation, amplification, mutation) in a wide range of malignancies in adults and children. To avoid off label use and allow for a nationwide safe and controlled access to czb for pts with an ALK, MET or ROS1 positive tumor, the French National Cancer Institute (INCa) launched the AcSé program: access to tumor molecular diagnosis in the 28 INCa molecular genetic centers along with an exploratory phase II trial. Methods: Biomarker identification is proposed to pts ≥ 1 year with an advanced disease among more than 15 malignancies (such as colon, gastric, liver, thyroid, renal and breast cancers, cholangiocarcinoma, lymphoma, neuroblastoma, sarcomas, and ROS1 or MET lung cancer) (such as colon, gastric, liver, thyroid, renal and breast cancers, cholangiocarcinoma, lymphoma, neuroblastoma, sarcomas, and ROS1 lung cancer) known from literature to harbor a genomic alteration in a czb target. If not eligible for any other academic or industry trial targeting the same alteration, a patient with an ALK, MET or ROS1 positive tumor may enter one of the 22 specified cohorts defined as a disease and a type of target alteration, and receives czb (adult: 250 mg x 2; child: 280 mg/m2 x 2). Pts with an altered czb target as evidenced through a pangenomic tumor profiling program are also eligible. Tumor response is evaluated every 2 months using RECIST criteria. Three statistical 2-stage designs are considered for cohorts to anticipate 3 situations in terms of expected response rate and incidence. Accrual stops if 0 response / N1 pts; else N2 additional pts are recruited. 10,000 to 15,000 molecular tests and 490 pts treated in 150 centers are planned over 3 years. From Aug. 2013 to Jan. 2014, 22 pts have been accrued. The AcSé program is currently being expanded to other targeted drugs. Clinical trial information: NCT02034981.

Mir-31-3p as a predictive biomarker of cetuximab effects in a post hoc analysis of new EPOC phase III trial.

Citation: J Clin Oncol 32:5s, 2014

Author(s): Pierre Laurent-Puig, John A. Bridgewater, John Neil Primrose, Sian Alexandra Pugh, Gareth Thomas, Karwan Moutasim, Francis Rousseau, Karine Fontaine, Celine Vazart, Virginie Decaulne, Anne Leclair, Berengere Genin, Sandrine Imbeaud, Francois Liebaert, Raphaële Thiébaut; UMR-S775 Bases Moléculaires de la Réponse aux Xénobiotiques, Paris, France; Department of Medical Oncology, University College London Cancer Institute, London, United Kingdom; University Hospital Southampton NHS Foundation Trust, Southampton, United Kingdom; University of Southampton, Southampton, United Kingdom; IntegraGen SA, Evry, France; IntegraGen S.A, Evry, France; INSERM UMR-U674, Université Paris-Descartes, Paris, France; IntegraGen, Evry, France; Integragen SA, Boulogne Billancourt, France

Background: miR31-3p expression level has been associated with progression free survival (PFS) in KRAS wild type metastatic colorectal cancer (mCRC) patients treated with anti-epidermal growth factor receptors (EGFRs). In this study we evaluated its predictive value on PFS in the new EPOC trial by measuring its expression in primary tumors and liver metastases. Methods: miR31-3p expression was determined by RT-QPCR in 125 formalin-fixed and paraffin-embedded (FFPE) primary tumor samples from patients. There were 63 patients (control arm) who received oxaliplatin or irinotecan-based chemotherapy and 62 received the above plus cetuximab. Correlations between miRNA expression and survival were performed using an adjusted Cox model. Liver metastasis were available for 58 of the 125 patients and comparison was done using a Pearson correlation on log2 transformed miR31-3p expression value. Results: A significant association was found between PFS and miR31-3p expression in the cetuximab arm (p=0.035; HR=1.2, CI95% CI [0.98 – 1.48]), and not in the control arm (p=0.36; HR=0.96, CI95% [0.75 – 1.23]).A predictive model was developed dichotomizing patients with high or low risk of progression. In the cetuximab arm PFS was significantly shorter in patients with high expression than in patients with low expression (p=0.033). In all patients with high miR31-3p expression, PFS was significantly shorter in the cetuximab treated arm than in the control arm (p=0.0177, median PFS: 49.6 and 64.9 weeks respectively), in all patients with low miR31-3p expression, PFS was not different between cetuximab and control arm. Study of miR-31-3p expression in primary tumor and matching metastasis showed a correlation in the control arm (p=0.00004) but not in the cetuximab arm (p=0.55). Conclusions: We demonstrated for the first time that miR31-3p expression is predictive of cetuximab effects and replicated association of its expression with PFS in patients receiving cetuximab. Furthermore miR31-3p allowed identification of a subgroup of patients in which cetuximab with chemotherapy had a detrimental effect on PFS. Eventually the correlation of miR31-3p expression in metastases and primary tumors in the control arm, but not in the cetuximab arm, suggests cetuximab has an effect on miR31-3p expression, supporting its involvement in the EGFR pathway.

Prognostic value of KRAS exon 2 gene mutations in stage III colon cancer: Post hoc analyses of the PETACC8 trial. Citation: J Clin Oncol 32:5s, 2014

Author(s): Julien Taïeb, Jean-François Emile, Karine Le Malicot, Aziz Zaanan, Josep Tabernero, Enrico Mini, Philippe Rougier, Jean Luc VAN Laethem, John A. Bridgewater, Gunnar Folprecht, Ramon Salazar, Ayman Zawadi, Eric Van Cutsem, Come Lepage, Pierre Laurent-Puig, Helene Blons; APHP and Paris Descartes University, Paris, France; Hospital Ambroise Pare, APHP, University Versailles-SQY, Boulogne, France; Fédération Francophone de la Cancérologie Digestive Faculté de Médecine, Dijon, France; Department of Gastroenterology, HEGP, Paris, France; Vall d’Hebron University Hospital, Barcelona, Spain; Unità di Chemioterapia, Dipartimento di Farmacologia, Università degli Studi di Firenze, Firenze, Italy; European Hospital George Pompidou, Paris, France; Hôpital Universitaire Erasme, Bruxelles, Belgium; Department of Medical Oncology, University College London Cancer Institute, London, United Kingdom; University Hospital Carl Gustav Carus, Dresden, Germany; Early Clinical Research Unit, Institut Català d’Oncologia, L´Hospitalet-Barcelona, Spain; Radiothérapie, Centre Hospitalier Départemental, La Roche Sur Yon, France; University Hospitals Gasthuisberg, Leuven, Belgium; Centre Hospitalier Universitaire Bocage, Dijon, France; UMR-S775 Bases Moléculaires de la Réponse aux Xénobiotiques, Paris, France; Hôpital Européen Georges Pompidou, Assistance Publique Hôpitaux de Paris, Paris, France

Background: Prognostic value of KRAS mutations in resected colon adenocarcinoma remains debated. We examined the prognostic impact of KRAS exon 2 mutations in stage III patients receiving adjuvant FOLFOX + /- cetuximab patients from the PETACC8 Phase III trial. Methods: KRAS exon 2 mutations in codon 12 (p.G12V, p.G12R, p.G12S, p.G12A, p.G12C, p.G12D) and codon 13 (p.G13D) were examined in all patients with available material and signed translational research informed consent enrolled in the PETACC8 trial. Analyses were restricted to BRAF wild type tumors, since the prognostic impact of BRAFV600E in this population has already been described. Because no benefit or deleterious effect from adjuvant cetuximab was reported in this trial, tumors from both study arms were pooled for analysis. Association between time to recurrence (TTR) and disease-free survival (DFS) and type of KRAS mutation was evaluated using Cox proportional hazard model. Results: KRAS mutations were found in 638/1657 tumors, including 502 codon 12 and 136 codon 13 alterations. KRAS mutations in codon 12 (HR: 1.67; 95% confidence interval [CI] [1.35-2.04]; P<0.001) but not in codon 13 (HR: 1.23; 95% confidence interval [CI] [0.85-1.79]; P=0.26) were significantly associated with shorter TTR as compared to patients with KRAS/BRAF wild-type tumors, independently of other covariates. Similar results were found for DFS. When anatomic sites were taken into account, the impact of KRAS mutations on TTR was only found for distal tumors (n=1043) with an increased risk of relapse (HR: 1.96; 95% confidence interval [CI] [1.51-2.56]; P<0.0001) for KRAScodon 12 mutated tumors and a trend for codon 13 (HR: 1.59; 95% confidence interval [CI] [1.00-2.56]; P=0.051), as compared to KRAS/BRAFwild type patients. Similar results were found for DFS. Conclusions: KRAS exon 2 mutations are independent predictors of TTR or DFS for patients with stage III distal colon cancer receiving adjuvant therapy. Future clinical trials in the adjuvant setting should consider tumor location and type of KRAS mutation as important stratification factors.

Strategic 1-multi-line therapy trial in unresectable wild-type RAS metastatic colorectal cancer: A gercor randomized open-label phase III study.

Author(s): Benoist Chibaudel, Christophe Tournigand, Franck Bonnetain, Marine Hug de larauze, Armand De Gramont, Pierre Laurent-Puig, Christophe Louvet, Thierry André, Aimery De Gramont, GERCOR; Hôpital Saint-Antoine, Paris, France; Hopital Henri Mondor, Créteil, France; Centre Hospitalier Universitaire, Besançon, Besançon, France; GERCOR – French Academic Research Group, Paris, France; Beaujon University Hospital, Clichy, France; UMR-S775 Bases Moléculaires de la Réponse aux Xénobiotiques, Paris, France; Department of Oncology, Institut Mutualiste Montsouris, Paris, France; Hospital Saint Antoine, Paris, France

Background: Several strategies using chemotherapy and molecular targeted drugs are available for the treatment of unresectable metastatic colorectal cancer. Recent and ongoing randomized trials evaluate chemotherapy with bevacizumab or epidermal growth factor receptor (EGFR) inhibitors, but the prior and/or subsequent lines are not fixed and the imbalanced cross-overs will not allow to interpret overall survival. This STRATEGIC-1 trial is a study designed to determine the best sequence of therapy and to define subset populations that will benefit most from one sequence. Methods: This is an ongoing randomized, two-arm, phase III study comparing two multi-line therapeutic strategies in patients wild-type RAS unresectable metastatic colorectal adenocarcinoma, ECOG PS 0-2 and age ≥ 18years. Randomization is stratified by center, GERCOR prognostic score (using PS and LDH level), prior use of oxaliplatin in adjuvant setting and the extent of metastatic disease. Patients (n=474) are randomized (1:1) to either (arm A) FOLFIRI/cetuximab, followed by an oxaliplatin-based chemotherapy with bevacizumab or (arm B) OPTIMOX/bevacizumab, followed by an irinotecan-based chemotherapy with bevacizumab, followed by anti-EGFR mab (cetuximab or panitumumab) with or without irinotecan. The primary endpoint is Duration of Disease Control (DDC). The sample size was planned for testing the primary variable DDC with a two-sided 5% α type one error and a 10% β type two error (Software: EAST 5.3) and a planned interim analysis. A 33% reduction in the risk of event (HR 0.67) was assumed under H₁ in the arm B. Secondary endpoints include OS, HR-QoL, TFS, ORR and PFS per sequence, salvage surgery rate, safety and correlations between biomarkers and clinical outcome. CRC tissue (primary or met.) and blood collection are mandatory for biomarker analyses. Enrolment began in October 2013. Clinical trial information: NCT01910610.

Impact of adjuvant chemotherapy with 5-FU or FOLFOX in colon cancers with microsatellite instability: An AGEO multicenter study.

Author(s): David Tougeron, Gaelle Sickersen, Thierry Lecomte, Guillaume Mouillet, Isabelle Trouilloud, Romain Coriat, Thomas Aparicio, Gaetan Des Guetz, Cedric Lecaille, Pascal Artru, Estelle Cauchin, David Sefrioui, Tarek Boussaha, Aurelie Ferru, Julien Taïeb, Pierre Michel, Lucie Karayan-Tapon, Dewi Vernerey, Franck Bonnetain, Aziz Zaanan; Department of Gastroenterology, Poitiers University Hospital, Poitiers, France; Poitiers university Hospital, Poitiers, France; Centre Hospitalier Trousseau, Tours, France; Institut Regional du Cancer en Franche-Comté – University Hospital, Besançon, France; Ambroise Paré, Paris, France; Cochin Teaching Hospital, AP-HP, Paris Descartes University, Paris, France; Hôpital Avicenne, Assistance Publique Hôpitaux de Paris, Bobigny, France; Oncology Department, Hôpital Avicenne, HUPSSD, UCOG 93 (APHP), Bobigny, France; Clinique Bordeaux Nord, Bordeaux, France; Hôpital Privé Jean Mermoz, Lyon, France; Nantes University Hospital, Nantes, France; Digestive Oncology Unit, Department of Hepato-Gastroenterology, Rouen University Hospital, Rouen, France; Hôpital Saint-Antoine – Assistance Publique Hôpitaux de Paris, Paris, France; Department of Oncology, Poitiers University Hospital, Poitiers, France; APHP and Paris Descartes University, Paris, France; INSERM U935, University of Poitiers, Poitiers, France; CHRU Besançon, Besançon, France; Centre Hospitalier Universitaire, Besançon, Besançon, France; Department of Gastroenterology, HEGP, Paris, France

Background: Microsatellite instability (MSI) is found in 12% of colon cancers (CC) and associated with a low recurrence rate after curative surgery. Adjuvant chemotherapy with 5-FU seems ineffective, but small studies recently suggest that adjuvant chemotherapy with FOLFOX could be effective (Zaanan et al.). The aim of this study was to analyze efficacy of adjuvant chemotherapy with 5-FU or FOLFOX in relapse-free survival (RFS) of MSI CC. Methods: This multicenter retrospective study included patients with stage II or III MSI CC with curative surgery between 2000 and 2012. High-risk stage II CC were defined by one of these criteria: stage T4, bowel obstruction, tumoral perforation, vascular emboli, lymphatic invasion, perinervous invasion, or a number of lymph nodes examined inferior to 10. Prognostic factors of RFS were analyzed in univariate and multivariate analysis using Cox model. Results: A total of 433 MSI CC patients were analyzed, including 57% and 43% stage II and III, respectively. Median follow-up was 35 months. Mean age was 70 ± 17 years. Overall, 61%, 27%, 12% patients had a surgery alone (n=263), adjuvant FOLFOX (n=119) or 5-FU (n=51) respectively. Adjuvant chemotherapy was administered in 17% of stage II (n=41) and 70% of stage III (n=129). Recurrence rates were 6% (n=14) for stage II and 21% (n=39) for stage III. Adjuvant chemotherapy was associated with better RFS in univariate analysis, but only for FOLFOX (HR=0.46, 95%CI 0.23-0.79) and not for 5-FU (HR=1.02, 95%CI 0.60-1.73). Three-years RFS was 75% for surgery alone, 66% for 5-FU and 84% for FOLFOX (p=0.02). In multivariate analysis taking account the other prognosis factors, adjuvant chemotherapy with oxaliplatin remains significantly associated with better RFS (HR=0.29, 95%CI 0.13-0.65 ; p=0.003). In the subgroup analyses, benefit of FOLFOX, as compared to 5-FU and surgery alone, was significant in stage III (HR=0.38, 95%CI 0.21-0.69 ; p=0.0014) and a trend was observed for high-risk stage II (HR=0.141 95%CI 0.02-1.04 ; p=0.0549). Conclusions: This multicenter study confirms that adding oxaliplatin to 5-FU can restore the chemosensitivity of MSI CC.

Perioperative chemotherapy with FOLFOX in resectable gastroesophageal adenocarcinoma: Preliminary results of an AGEO multicentric retrospective study.

Author(s): Florence Mary, Aziz Zaanan, David Malka, Pascal Artru, Marouane Boubaya, Emmanuelle Samalin, Romain Coriat, Jean Baptiste Bachet, Mourad Benallaoua, Pauline Afchain, Christophe Locher, Isabelle Baumgaertner, Cedric Lecaille, Thomas Aparicio; Hôpital Avicenne, Bobigny, France; Department of Gastroenterology, HEGP, Paris, France; Institut Gustave Roussy, Villejuif, France; Hôpital Privé Jean Mermoz, Lyon, France; Digestive Oncology CRLC Val d’Aurelle, Montpellier, France; Cochin Teaching Hospital, AP-HP, Paris Descartes University, Paris, France; Centre Hospitalier Universitaire Pitié Salpétrière, Paris, France; Hôpital Saint Antoine, APHP, Paris, France; Department of Hepato-Gastroenterology, Meaux Hospital, Meaux, France; Hôpital Henri Mondor, Creteil, France; Clinique Bordeaux Nord, Bordeaux, France; Hôpital Avicenne, Assistance Publique Hôpitaux de Paris, Bobigny, France

Background: Perioperative 5-fluorouracile (5-FU) associated with cisplatin chemotherapy improved overall and recurrence free survival, the R0 resection rate, in resectable gastro-oesophageal junction and gastric adenocarcinoma. The aim of this study is to evaluate the feasibility, R0 resection rate, survival and tolerance of the 5-FU with oxaliplatin perioperative chemotherapy. Methods: We enrolled all the resectable gastric or gastro-oesophageal adenocarcinoma who had at least 3 cycles of pre-operative folfox based regimen (simplified folfox 6; or folfox 6: oxaliplatin 100 mg/m², 400 mg/m², 5-fluorouracile bolus, and 2400 mg/ m², 5-fluorouracile during 2 days). The following data were collected: age, weight, height, karnofsky index, toxicity, dose, surgery R0, histology, post-operative radiotherapy, recurrence and death. Overall and disease free survival was calculated with the Kaplan-Meyer method. Results: We enrolled 106 patients in 11 centers. There were 72 men, the median age was 66 years. Gastro-oesophageal junction represented 28.3%. The median number of chemotherapy cycle was 4 before surgery and 2.5 after. In univariate analysis the karnofsky index at inclusion was the only factor associated with the realization of 8 cycles chemotherapy. A grade 3-4 toxicity occurred in 13.8%. The R0 rate was achieved in 98 patients. The median overall and disease free survival was 41 and 28 month. The ACE rate was in univariate analysis a predictive factor for overall and disease free survival. The other predictive factor for overall survival was: Karnofsky index and the number of post surgery chemotherapy cycle. Conclusions: The oxaliplatine perioperative regimen is an alternative instead of cisplatin based regimen for the gastric and gastro-oesophageal junction resectable adenocarcinoma with a good tolerance and efficacy. A 12 cycle treatment is rarely achievable, 8 cycles seems to be more feasible.

Multidisciplinary risk assessment to reveal cancer treatments in unfit cancer patients.

Author(s): Pascaline Boudou-Rouquette, Olivier Huillard, Audrey Thomas-Schoemann, Anne Chahwakilian, Galdric Orvoen, Antoine Tesniere, Laure Cabanes, Julie Giroux, Anatole Cessot, Jean Stephanazzi, Helen Mosnier-Pudar, Jean-Philippe Durand, Vincent Montheil, Jerome Alexandre, Francois Goldwasser; Cochin Hospital, AP-HP, Paris Descartes University, Sorbonne Paris Cité, Paris, France; Broca Hospital, AP-HP, Paris Descartes University, Sorbonne Paris Cité, Paris, France

Background: Older age is a cause of disparity in cancer treatment decision. Treatment guidelines for patients with comorbidities, polypharmacy, denutrition or psycho-social frailty are needed. A pre-therapeutic multidimensional assessment might improve the unfit patient management. We developed an experimental program of integrated medicine called ARIANE. We report 18 months activity of this outpatient setting evaluation, its feasibility and impact on treatment decision-making. Methods: Unfit patients with predefined cancer treatment strategy entered into the program. A one-day evaluation combined consultations of cardiologist, geriatrician, diabetologist, anesthetist, pharmacist, pain specialist, dietician, psychologist and social worker. Evaluation of performance status, ECG, ejection fraction, ASA score, diabetes, social vulnerability and malnutrition was performed including a geriatric assessment, which focused on items like comorbidity (CIRS-G), dependence (ADL, IADL), fails (Up and Go Test), cognitive impairment (MMSE, Clock Drawing Test) and depression (GDS scale). A pharmacist assessed the risk of drug-drug interactions. Results: Eighty-seven pts, median age 81 years (range 25-94), 76% male, 51% PS 0-1, 77% grade 3 or 4 comorbidity were included. Genito-urinary, lung cancers and sarcoma represented 77% of pts. Eighty-two percent of pts were assessed by at least ≥ 7 participants. Identified factors of vulnerability were polypharmacy (n=65; 75%; >3 drugs), social distress or severe malnutrition (both n=21; 24%), depression (n=17; 19.5%) and cognitive impairment (n=13; 15%). We identified drug interactions in 18 pts (27%). The risk assessment resulted in anticancer treatment changes in 47/87 patients (54%): protocol adaptation (n=19/87; 22%), less aggressive treatment (n=15/87; 17.2%), or more intensive therapy (n=13/87; 15%). Conclusions: A one-day multidisciplinary risk assessment is an answer to the complexity of unfit cancer patients and improves the safety of anticancer treatments.

Association of sunitinib exposure with toxicity outcome in a real-life population of elderly patients with cancer.

Author(s): Pascaline Boudou-Rouquette, Audrey Thomas-Schoemann, Olivier Huillard, Anatole Cessot, Anne Chahwakilian, Galdric Orvoen, Julie Giroux, Jean-Philippe Durand, Michel Vidal, Benoit Blanchet, Francois Goldwasser; Cochin Hospital, AP-HP, Paris Descartes University, Sorbonne Paris Cité, Paris, France; Broca Hospital, AP-HP, Paris Descartes University, Sorbonne Paris Cité, Paris, France; Cochin Teaching Hospital, AP-HP, Paris Descartes University, Sorbonne Paris Cité, Paris, France

Background: Elderly people represent a growing population susceptible to be treated with sunitinib (SU). The aim of this retrospective study was to explore the relationship between SU exposure and the occurrence of dose-limiting toxicities (DLT) in a real-life population of elderly patients. Methods: Elderly patients(>70 years old) received sunitinib at a starting dose of 50 mg per day (4 weeks on/2 weeks off) (standard regimen, SR) or a continuous daily dose of 25 or 37.5 mg (adapted regimen, AR) after a multidisciplinary risk assessment. DLT during the first cycle (6 weeks) were assessed every 14 days, and SU and SU 12662 (active metabolite) plasma concentrations were measured by liquid chromatography. SU, SU 12662 and composite (SU+SU12662) exposure (AUC) were estimated from a bayesian approach. Results: Twenty-five patients, median age 75 years [25^th-75^th percentile 72-82 ], 13 (52%) men, 64% mRCC, median ECOG-PS=1 (0-3), median active comorbidities of 2 (0-4) and 15 (60%) pts with polypharmacy (> 3 drugs) were included. Eleven pts were included in the SR group, and 14 in AR group (7 pts with 37.5 mg and 7 pts with 25 mg). DLT occurred in 6/11 (54%) in SR group and 9/14 (64%) in AR group, resulting in either dose reductions (36% and 36%, respectively) or definitive SU discontinuation (9% and 21%, respectively). The most frequent grade 3-4 AE in SR and AR groups was hypertension (36% and 29%, respectively). Median composite AUC (n=19) on day 14 after treatment initiation was 1865 ng/mL.h (25^th-75^thpercentile 1287-2253). In SR group, pts with DLT exhibited a greater composite exposure than pts without DLT (3086 ± 982 vs 1621 ± 395 ng/mL.h, respectively p=0.0148). In contrast, in AR group, there was no significant difference in composite AUC between patients with DLT and those without (1542 ± 708 vs 1345 ± 745 ng/mL.h, p=0.70). Conclusions: This retrospective study shows a same frequency of DLTs in pts receiving SR compared to those with AR, and a lack of relation between composite exposure and DLT occurrence in patients under AR. This suggests that SR associated with drug monitoring would be a better option than an initial dose reduction in a real-life population of elderly cancer patients under sunitinib.

Low skeletal muscle density as predictive for febrile neutropenia in patients treated by doxorubicin/trabectedin/pegfilgrastim combination as a first-line treatment of advanced or metastatic leiomyosarcoma (LMS) (LMS02 study)

Author(s): Sami Antoun, Anne Floquet, Nicolas Penel, Corinne Delcambre, Christine Chevreau, Didier Cupissol, Jerome Alexandre, Florence Duffaud, Benjamin Lacas, Patricia Pautier; Gustave Roussy, Villejuif, France; Institut Bergonié, Bordeaux, France; Centre Oscar Lambret, Lille, France; Centre François Baclesse, Caen, France; Institut Claudius Regaud, Toulouse, France; Department of Medical Oncology, Centre Val d’Aurelle Paul Lamarque, Montpellier, France; Cochin Hospital, AP-HP, Paris Descartes University, Sorbonne Paris Cité, Paris, France; La Timone University Hospital, Marseille, France; Department of Medicine, Gustave Roussy, Villejuif, France

Background: Studies have shown that skeletal muscle mass (SMM) and skeletal muscle densities (SMD) are associated to chemotherapy toxicity. In 110 patients treated with a combination of doxorubicine (doxo) and trabectedin (trab), neutropenia and febrile neutropenia are still observed despite the use of granulocyte colony stimulating factor (G-CSF) (ASCO 2013; abstract 10505). Our aim was to analyze whether SMM or SMD are predictive of febrile neutropenia. Methods: SMM and SMD were assessed with computed tomography (CT) imaging before treatment by measuring cross-sectional areas of the tissues for SMM and the mean muscle Hounsfield Units (HU) for SMD. SMD assessed by this method reflects fatty muscle infiltration with lower mean HU reflecting lower density, and more fatty infiltration. Toxicity profile was collected for all cycles. Therefore severe toxicity is defined as any grade 3 or 4 toxicity. The cut-off level which predicts the occurrence of toxicity most accurately was deduced from the receiver operating characteristic curve (for SMD the value is: 37.1 HU). Pts received doxo 60 mg/m² followed by trab 1.1 mg/m²IV in 3-h at day 1, and pegfilgrastim 6 mg on day 2 every 21 days for a maximum of 6 cycles. Results: 55 pts were included (46 females and 9 males), with median age of 58 years, 27 pts with uterine LMS and 28 with soft tissue LMS were analyzed. Only 2 pts had an ECOG PS score >1. Pts received a total of 285 cycles with a median of 6 cycles per pt. Pts with a low density (SMD <37.1) had a higher probability of febrile neutropenia (8/19; 42%) than pts with a SMD > 37.1 (6/36; 17%) (p=0.05). No association between toxicity and SMM was found. Conclusions: Despite the use of GCSF, febrile neutropenia is observed in pts treated with doxo + trab. In this pilot study including a few pts, muscle density has been found to be associated with a high probability of febrile neutropenia. These interesting results need to be confirmed. They might highlight the concept of “frailty” i.e. a group of non-oncologic parameters associated with a higher susceptibility to events.

Randomized phase III trial of paclitaxel/carboplatin (PC) versus cisplatin/irinotecan (CPT-P) as first-line chemotherapy in patients with clear cell carcinoma (CCC) of the ovary: A Japanese Gynecologic Oncology Group (JGOG)/GCIG study.

Author(s): Aikou Okamoto, Toru Sugiyama, Tetsutaro Hamano, Jae Weon Kim, Byoung Gie Kim, Takayuki Enomoto, Daisuke Aoki, Yasuhisa Terao, Nao Suzuki, Mikio Mikami, Nobuo Yaegashi, Kiyoko Kato, Hiroyuki Yoshikawa, Sandro Pignata, Jerome Alexandre, John A. Green, Seiji Isonishi, Fumitoshi Terauchi, Keiichi Fujiwara, Kazunori Ochiai; Jikei University School of Medicine, Tokyo, Japan; Iwate Medical University, Morioka, Japan; Kitasato Academic Research Organization, Tokyo, Japan; Seoul National University, College of Medicine, Seoul, South Korea; Sungkyunkwan University, Seoul, South Korea; Niigata University, Niigata, Japan; Keio University, Tokyo, Japan; Juntendo University School of Medicine, Tokyo, Japan; St.Marianna University, Kawasaki, Japan; Tokai University, Kanagawa, Japan; Department of Gynecology, Tohoku University School of Medicine, Sendai, Japan; Kyushu University, Fukuoka, Japan; Department of Obstetrics and Gynecology, Faculty of Medicine, University of Tsukuba, Tsukuba, Japan; Istituto Nazionale Tumori di Napoli, Napoli, Italy; Hospital Hotel Dieu, Paris, France; University of Liverpool, Wirral, United Kingdom; Tokyo Medical University, Tokyo, Japan; Saitama Medical University International Medical Center, Saitama, Japan

Background: CCC is a histologic subtype of epithelial ovarian cancer showing different clinical and biological characteristics. CCC has become well known for its resistance to current standard chemotherapy (PC). Our previous trial demonstrated the potential benefit of CPT-P regimen on CCC. Methods: Patients (pts) with Stage I-IV CCC were randomized to receive paclitaxel 175mg/m² plus carboplatin AUC6 IV q3wk or irinotecan 60mg/m² IV (days 1, 8, 15) plus cisplatin 60mg/m² IV (day 1) q4wk for 6 cycles. International central pathologic review was performed for all cases. The primary endpoint was progression-free survival (PFS). Secondary endpoints were overall survival (OS), response rate per RECIST, and adverse events. Adverse events were graded according to NCI-CTCAE, version 3.0. Results: From November 2006 to March 2011, 667 patients were registered, and 619 pts were clinically and pathologically eligible for evaluation (305 pts in PC arm and 314 pts in CPT-P arm). Median age was 53 years. Baseline pt characteristics were not significantly different. Following 44.3 months median follow-up, 2-years PFS was 73.0% (95% CI:67.7-77.5) in the CPT-P arm vs. 77.6% (95% CI:72.4-81.9) in the PC arm, which was not significantly different (HR:1.171, 95% CI:0.867-1.581, p=0.303). Two-years OS was 85.5% in CPT-P arm (95% CI:81.1-89.0) and 87.4% in PC arm (95% CI:83.1-90.7), respectively (HR:1.133, 95% CI:0.796-1.613, p=0.486). Grade 3/4 leukopenia, neutropenia, thrombocytopenia, peripheral sensory neuropathy and joint pain occurred more frequently in the PC arm, whilst grade 3/4 anorexia, diarrhea, nausea, vomiting and febrile neutropenia occurred more frequently in the CPT-P arm. Conclusions: In this first CCC-specific international clinical trial, survival benefit was not observed by CPT-P. Since both regimens were well tolerated and the toxicity profiles were different, CPT-P can be an alternative regimen for CCC.

Outcomes of surgical management for borderline (BR)/locally advanced (LA) pancreatic adenocarcinoma (PAC) after neoadjuvant FOLFIRINOX chemotherapy: Results of an AGEO/French multicenter study.

Author(s): Daniel Pietrasz, Lysiane Marthey, Mathilde Wagner, Jean-Frédéric Blanc, Eric Terrebonne, Antonio Sa Cunha, Romain Coriat, Isabelle Trouilloud, Nicolas Regenet, Pascal Artru, Eric Francois, Lilian Schwarz, David Malka, Jean-Christophe Vaillant, Jean Baptiste Bachet, Mehdi Karoui; Pitié Salpêtrière Hospital, Paris, France; Beclere Hospital, Clamart, France; Pitié Salpetrière Hospital, Paris, France; Hôpital Saint-André, Bordeaux, France; CHU de Bordeaux, Pessac, France; Paul Brousse Hospital, Villejuif, France; Cochin Hospital, Paris, France; Ambroise Paré, Paris, France; Nantes university hospital, Nantes, France; Hôpital Privé Jean Mermoz, Lyon, France; Department of Medical Oncology, Centre Antoine-Lacassagne, Nice, France; Rouen University Hospital, Rouen, France; Institut Gustave Roussy, Villejuif, France; Pitie Salpetriere Hospital, Paris, France; Centre Hospitalier Universitaire Pitié Salpétrière, Paris, France

Background: FOLFIRINOX has been shown to increase survival in patients with metastatic PAC with an objective response rate of 32%. In BR and LA PAC, a resection rate of 36% has been reported in a retrospective study. The aim of our study was to analyse post-operative outcomes in patients resected for BR or LA PAC who received neoadjuvant treatment based on FOLFIRINOX. Methods: Data of patients who underwent a curative intent resection after neoadjuvant FOLFIRINOX in 18 French centers were reviewed in this retrospective study. BR or LA PAC were classified following the National Comprehensive Cancer Network classification using CT-scan by investigators. Post-operative morbidity was graded according to the Dindo classification. Disease free survival (DFS) and overall survival (OS) were analysed using Kaplan Meier method from the date of surgical resection. Results: From November 2010 to December 2013, 58 patients (36 men, median age of 59 years, 34 BR and 24 LA at baseline CT-scan staging) underwent a pancreatic resection after a median number of 6 cycles (1-30) of FOLFIRINOX. Additional preoperative chemoradiotherapy was administred in 42 patients (72%). Initial median tumor size was 31 mm (16-66) with 79% located in the pancreatic head. 24% (n=14) of tumors required en bloc vascular resection. Post-operative mortality and severe morbidities (grade 3-4) were 3.4% and 19%, respectively. R0 resection was achieved in 86% of patients. The histological complete response (ypT0N0) rate was 19%. After a median follow-up of 11.0 months (0.4-37.2) from surgery, a recurrence occurred in 15 patients (local n=5; metastatic n=9; both n=1) including 9 deaths during follow-up. Median DFS and OS were not reached. One year DFS and OS rates were 53% (16/34) and 71% (9/34), respectively. Conclusions: In this retrospective study, pancreatic resection after neoadjuvant FOLFIRINOX was safe even when combined with radiochemotherapy. Rates of R0 resection and histological complete response were high suggesting that this strategy should be further evaluated in a prospective manner in patients with BR/ LA PAC.

Second-line chemotherapy for advanced biliary tract cancer after failure of gemcitabine plus platinum: Results of an AGEO multicenter retrospective study.

Author(s): Bertrand Brieau, Laetitia Dahan, Yann De Rycke, Tarek Boussaha, Philippe Vasseur, David Tougeron, Thierry Lecomte, Romain Coriat, Jean Baptiste Bachet, Pierre Claudez, Aziz Zaanan, Pauline Soibinet, Jérôme Desramé, Anne Thirot-Bidault, Isabelle Trouilloud, Florence Mary, Christophe Locher, Lysiane Marthey, Wulfran Cacheux, Astrid Lievre; Hôpital Cochin, Paris, France; La Timone, Marseille University Hospital, Marseille, France; Department of Biostatistics, Institut Curie, Paris, France; Hôpital Saint-Antoine – Assistance Publique Hôpitaux de Paris, Paris, France; CHU Poitiers, Poitiers, France; Department of Gastroenterology, Poitiers University Hospital, Poitiers, France; Centre Hospitalier Trousseau, Tours, France; Cochin Teaching Hospital, AP-HP, Paris Descartes University, Paris, France; Centre Hospitalier Universitaire Pitié Salpétrière, Paris, France; CHU Saint Etienne, Saint Etienne, France; Department of Gastroenterology, HEGP, Paris, France; CHU Reims, Reims, France; Hôpital Privé Jean Mermoz, Lyon, France; Department of Hepato-Gastroenterology, Bicêtre Hospital, Kremlin-Bicêtre, France; Ambroise Paré, Paris, France; Hôpital Avicenne, Bobigny, France; Department of Hepato-Gastroenterology, Meaux Hospital, Meaux, France; Hopital Bicetre, Le Kremlin Bicetre, France; Institut Curie, Paris, France; Department of Medical Oncology, Institut Curie, Saint-Cloud, France

Background: First-line chemotherapy (CT1) with the combination of gemcitabine (gem) + platinum has become a new standard in advanced biliary tract cancer (ABTC) but data on second-line CT (CT2) are lacking. The aim of this study was to evaluate the efficacy and tolerability of CT2 in patients (pts) with ABTC who received gem-platinum in CT1. Methods: We retrospectively reviewed data of consecutive patients who received CT2 for ABTC after failure to gem-platinum in 17 French institutions from November 2002 to December 2013. Progression-free survival (PFS) and overall survival (OS) were estimated from the start of L2 CT using Kaplan Meier method. Cox models were applied for multivariate analyses. Results: Among 603 pts who were treated by gem-platinum in CT1, 196 pts (median age, 63 years, range: 28-82; male, 51.5 %) received a CT2. CT1 included gem + cisplatin (7%) or oxaliplatin (ox) (93%), with a median PFS of 9.7 months (mo) and an ORR of 31%. Characteristics at the beginning of CT2 were: metastatic disease, 94% ; 1-2 metastatic sites, 68%; ECOG PS 0-1, 68%. CT2 CT was 5FU-irinotecan (iri) (n=62), 5FU-ox (n=17), 5FU-cisplatin (n=37), 5FU/capecitabine (CAP) (n=39) or other various regimens (n=41). Among the 186 evaluable pts, there were 22 PR (12%) and 70 SD (38%). After a median follow-up of 26.4 months, median PFS and OS were 3.2 and 6.7 mo respectively. There was no significant difference between CT regimens in terms of PFS (5FU-iri, 2.6 mo; 5FU-ox/5FU-cisplatine, 4.0 mo; 5FU/CAP, 3.2 mo and others, 3.7 mo; p=0,27) and OS (6.0 mo, 6.3 mo, 5.6 mo and 9.7 mo respectively; p=0.27) In multivariate analysis, PS 2-3, bilirubine > 17 µmol/L and CA19.9 > 400 UI/mL were significantly associated with a shorter PFS while PS 2-3, CA19.9 > 400 UI/mL and non-response to CT1 with a shorter OS. A grade 3-4 toxicity was observed in 32% of pts (neutropenia, 33%; diarrhea, 17%) and a toxic death occurred in 1.4% of pts. Conclusions: CT2 is associated with a disease control in a half of pts with ABTC who received gem-platinum in CT1. Nevertheless, the short median PFS observed in this study should encourage the evaluation of new treatments in pts with good clinical conditions and an adequate biliary drainage

Pulmonary metastasectomy in colorectal cancer patients with previously resected liver metastasis: Pooled analysis.

Author(s): Samer Salah, Francesco Ardissone, Michel Gonzalez, Pascal Gervaz, Marc Riquet, Kazuhiro Watanabe, Jon Zabaleta, Dalia Al-Rimawi, Ehab Massad, Elena Lisi, Osama H Hamed; Department of Medical Oncology, King Hussein Cancer Center, Amman, Jordan; Department of Clinical and Biological Sciences-Thoracic Surgery, San Luigi Hospital, Orbassano, Italy; Department of Thoracic Surgery, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland; Department of Surgery, Geneva University Hospital and Medical School, Geneva, Switzerland; Department of Thoracic Surgery, Hôpital Européen Georges Pompidou (HEGP), Assistance Publique Hôpitaux de Paris (APHP), Paris, France; Department of Surgery, Tohoku University, Sendai, Japan; Thoracic Surgery Service, Hospital Donostia, San Sebastian, Spain; Office of Scientific Affairs and Research, King Hussein Cancer Center, Amman, Jordan; Department of Surgery, King Hussein Cancer Center, Amman, Jordan; Department of Clinical and Biological Sciences-Thoracic Surgery, San Luigi Hospital, Torino, Italy

Background: Data addressing the outcomes and patterns of recurrence following pulmonary metastasectomty (PM) in patients with colorectal cancer (CRC) and previously resected liver metastasis are very limited. Methods: We searched the PubMed for studies assessing PM in CRC and gathered individual data for patients who had PM and a previous curative liver resection. The influence of potential factors on overall survival (OS) was analyzed through univariate and multivariate analysis. Furthermore, the influence of patterns of recurrence following PM on post-recurrence survival was investigated. Results: Between 1983 and 2009, 146 patients from five studies underwent PM and had previous liver resection. The median interval from resection of liver metastasis until detection of lung metastasis and the median follow up from PM were 23 and 48 months respectively. Sixty seven (46%) underwent intra-operative thoracic lymph node (LN) sampling; nine of them (13%) had pathologically confirmed LN involvement, and 48 (33%) received perioperative chemotherapy. Five-year OS and recurrence free survival following PM were 54% and 29% respectively. Factors predicting inferior OS in univariate analysis included thoracic LN involvement and size of largest lung nodule ≥ two cm. In multivariate analysis, thoracic LN involvement was the only independent factor (HR 4.86, 95% CI: 1.56-15.14, P = 0.006). Ninety-three patients had recurrence following PM. Data on dates and pattern of recurrence was available for 74 patients, the five-year post- recurrence survival for patients with lung only recurrence treated with repeated PM (n=21), lung only recurrence treated with chemotherapy (n=20), liver only recurrence (n=10), and all other patterns of recurrence (n=23) were 60%, 0%, 18%, and 5% respectively, p < 0.0001. Conclusions: PM offers a chance for long term survival in selected patients with CRC and previously resected liver metastasis. Thoracic LN involvement predicted poor prognosis; therefore, significant efforts should be undertaken for adequate staging of the mediastinum prior to PM. Also, adequate intra-operative LN sampling allows proper prognostic stratification and enrollment in novel adjuvant therapy trials.

In-patient supportive care unit in oncology: A 3-year assessment (2011-2013).

Author(s): Maria Kfoury, Pauline Leroy, Stephane Hans, Christian Herve, Daniel Brasnu, Stephane Oudard, Clarisse Sauvajot, Christophe Aubaret, Ludovic Tripault, Anne Guillou, Florian Scotte; Supportive Care Functionnal Unit, Medical Oncology Department, Georges Pompidou European Hospital, Paris, France; Supportive Care Unit, Medical Oncology Department, Georges Pompidou European Hospital, Paris, France; Hôpital Européen Georges Pompidou, APHP, Paris, France; University Rene Descartes, Paris, France; Service de Chirurgie ORL, Paris, France; Department of Medical Oncology, Hôpital Européen Georges Pompidou, René Descartes University, Paris, France

Background: An in-patient unit for supportive care in oncology (USSO) was created in 2011 at the Georges Pompidou European Hospital. Its primary objective was to improve patient care management and complications during their cancer treatment. A yearly evaluation was conducted to monitor the unit’s activity. Methods: An observational study was performed in the 4-bed unit from January 1 – December 31, 2013. Results: A total of 244 patients were included in the study (158 in 2011 versus 192 in 2012). In 2013, 67.6% of patients came directly from home, which resulted in 6.5% emergency admissions. The predominant diseases observed were head and neck cancer 37.7% and lung cancer 24.1%. The main reasons for hospitalization were bad performance status 30.3%, pain 13.5%, invasive procedures 12.7%, and nutritional disorders 9.1%. In USSO, an interdisciplinary approach remains the cornerstone of treatment with a daily collaboration of dieticians, psychologists, physiotherapists, and the palliative care team. Based on this policy, the patients average length of stay was reduced to 8.3 days for an occupancy rate of 100%. Finally, 61% of the patients were discharged home. With the help of a social worker, 9% of the patients were transferred to a recovery unit and 12.7% to a palliative care unit (total 9% deaths recorded). Conclusions: With the opening of the in-patient USSO, and the development of its interdisciplinary strategy, most of the patients were able to recover faster. By anticipating admissions at an early stage of the disease or treatment complications, we greatly amended patient management. These positive results enabled us to increase USSO capacity from 4 to 8 beds, which not only improved the supportive care organization but also the patient’s quality of life.

A randomized phase II study of GDC-0980 versus everolimus in metastatic renal cell carcinoma (mRCC) patients (pts) after VEGF-targeted therapy (VEGF-TT).

Author(s): Thomas Powles, Stephane Oudard, Bernard J. Escudier, Janet Elizabeth Brown, Robert E. Hawkins, Daniel E. Castellano, Alain Ravaud, Michael D. Staehler, Brian I. Rini, Wei Lin, Bridget O’Keeffe, Michelle Byrtek, Mark Lackner, Jill M Spoerke, Joseph A. Ware, Rui Zhu, Robert J. Motzer, on behalf of the ROVER Study Group; St. Bartholomew’s Hospital, London, United Kingdom; Department of Medical Oncology, Hôpital Européen Georges Pompidou, René Descartes University, Paris, France; Institut Gustave Roussy, Villejuif, France; Cancer Research UK Clinical Centre, Leeds, United Kingdom; The Christie Hospital NHS Foundation Trust, Manchester, United Kingdom; Hospital Universitario 12 de Octubre, Madrid, Spain; Hôpital Saint-André CHU, Bordeaux, France; Department of Urology, University Hospital Munich-Grosshadern, Ludwig Maximilian University, Munich, Germany; Cleveland Clinic, Taussig Cancer Institute, Cleveland, OH; The University of Texas MD Anderson Cancer Center, Houston, TN; Genentech, Inc., South San Francisco, CA; Genentech Inc., South San Francisco, CA; Memorial Sloan Kettering Cancer Center, New York, NY

Background: GDC-0980 (G) is a potent oral dual pan-PI3K and mTOR (TORC1/2) inhibitor that has been evaluated in multiple solid tumors in Phase I and II. Everolimus (E) inhibits TORC1 and is active in mRCC post VEGF-TT. This study is the first to dirctly test if a PI3K/mTOR inhibitor may improve efficacy over a TORC1 inhibitor. Methods: Clear cell mRCC pts who progressed on or after VEGF-TT were randomized (1:1) to G (40 mg QD) or E (10 mg QD), stratified by MSKCC score and time to progression on first VEGF-TT (≤ or > 6 months). The primary endpoint was progression-free survival (PFS); secondary endpoints included safety, overall survival (OS), objective response rate (ORR); and pharmacokinetic (PK) and biomarker correlation with safety and efficacy were exploratory. Results: Eighty-five pts were randomized (G 42:E 43). After 62 events (G 32:E 30), stratified analysis revealed the median PFS was significantly shorter for G than E (3.7 vs. 6.1 mo: HR 2.04 [95%CI: 1.18-3.54; p<0.01]) and did not favor G for any stratification subgroup. Median OS was not significantly different but trended in favor of E (11.9 vs. 14.6 mo: HR 1.73 [95%CI: 0.87-3.43; p=0.12]). ORR was 7.1% for G and 11.6% for E. Patients treated with G had greater incidence of Grade 3-4 adverse events (AEs) and were more likely to discontinue treatment because of an AE (G 31%:E 12%). G was associated with substantially more high-grade hyperglycemia (G 40%:E 7%) and rash (G 24%:E 5%). PK analyses of G suggest a relationship between exposure and safety (rash and hyperglycemia), but no clear exposure-efficacy relationship. Retrospective biomarker analyses from archived tissue revealed a relationship between VHL mutation status (by NGS) and outcome with E but not G. High HIF1A protein expression was associated with better outcome in both arms. Conclusions: This study was unable to demonstrate that inhibition of PI3K/TORC1/TORC2 with GDC-0980 provides any benefit over inhibition of TORC1 alone with everolimus. This may be due to the high rate of AEs associated with potent pathway inhibition. VHL mutation and HIF1A expression may be predictive of mTOR inhibitor benefit, though future prospective validation is required. Clinical trial information: NCT01442090.

Regional differences observed in the phase 3 trial (ELM-PC 5) with orteronel (TAK-700) plus prednisone in patients with metastatic castration-resistant prostate cancer (mCRPC) that has progressed during or following docetaxel.

Author(s): Karim Fizazi, Robert Jones, Stephane Oudard, Eleni Efstathiou, Fred Saad, Ronald De Wit, Johann Sebastian De Bono, Felipe Cruz, George Fountzilas, Albertas Ulys, Flavio Carcano, Neeraj Agarwal, David B. Agus, Joaquim Bellmunt, Daniel Peter Petrylak, Connie Lee, Bindu Tejura, Niels Geert Borgstein, Robert Dreicer; Institut Gustave Roussy, University of Paris Sud, Villejuif, France; University of Glasgow, Institute of Cancer Sciences, Glasgow, United Kingdom; HEGP, Paris, Université Paris Descartes, Paris, France; Alexandra General Hospital of Athens, Oncology Department, Department of Clinical Therapeutics, Medical School, University of Athens, Athens, Greece; University of Montréal Hospital Center, Montréal, QC, Canada; Erasmus University Medical Center, Rotterdam, Netherlands; The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust, London, United Kingdom; ABC Foundation School of Medicine, Santo André, Brazil; Aristotle University of Thessaloniki School of Medicine, Thessaloniki, Greece; Faculty of Medicine, Vilnius University, Vilnius, Lithuania; Hospital de Cancer de Barretos, Barretos-SP, Brazil; Huntsman Cancer Institute, University of Utah, Salt Lake City, UT; University of Southern California, Los Angeles, CA; Dana-Farber Cancer Center Institute/Brigham and Women’s Cancer Center, Boston, MA; Columbia University Medical Center, New York, NY; Takeda Pharmaceuticals International Co., Cambridge, MA; Cleveland Clinic, Cleveland, OH

Background: Orteronel is an investigational, nonsteroidal, selective 17,20-lyase inhibitor. ELM-PC 5 (NCT01193257) did not meet the primary endpoint of OS (P = 0.1898) despite improvement in radiographic PFS (rPFS; P = 0.00038) (Dreicer, ASCO GU 2014, Abstract #7). Here we analyze baseline characteristics and outcomes across study regions (Europe [E], North America [NA], and non-E/NA). Methods: Men with progressive mCRPC who had received prior docetaxel but not orteronel, abiraterone or ketoconazole were randomized 2:1 to orteronel 400 mg BID plus prednisone 5 mg BID (O+P) or placebo plus prednisone (P). Enrollment (N = 1099) was stratified by region (E, n = 590; non-E/NA, n = 397; NA, n = 112). The primary endpoint was OS; other key endpoints were rPFS, ≥50% prostate-specific antigen (PSA50) decrease and pain response. Results: Baseline characteristics were generally balanced, except for pain score, ≥2 prior chemotherapies, PSA and lacate dehydrogenase (LDH). Substantial regional differences were noted in OS, rPFS and PSA50 rates.Regional data are summarized below (Table). Conclusions: Substantial differences in efficacy were seen across regions, with differences in OS possibly being related to subsequent therapies. Clinical trial information: NCT01193257.

Association of hypertension and proteinuria with overall survival in solid-tumor patients treated with anti-VEGF drugs in the MARS study

Author(s): Vincent Launay-Vacher, Nicolas Janus, Isabelle Laure Ray-Coquard, Philippe Beuzeboc, Catherine Daniel, Joseph Gligorov, Frédéric Selle, Francois Goldwasser, Olivier Mir, Jean-Philippe Spano, Jean Christophe Thery, Jean-Baptiste Rey, Christelle Jouannaud, Jean F. Morere, Stephane Oudard, Florian Scotte, Michel Azizi, Richard Dorent, Gilbert Deray; Service ICAR – Pitie-Salpetriere Hospital, Paris, France; Centre Léon Bérard, Lyon, France; Medical Oncology Department, Institut Curie, Paris, France; Assistance Publique–Hôpitaux de Paris, Tenon APREC, CancerEst, University Paris VI, Paris, France; Universite Pierre et Marie Curie, Oncology, GHU-Est Tenon, Paris, France; Cochin Hospital, AP-HP, Paris Descartes University, Sorbonne Paris Cité, Paris, France; Hopital Cochin, Paris, France; Department of Medical Oncology, Pitié-Salpêtrière University Hospital, Paris, France; Medical Oncology Department, AP-HP, Salpetriere Hospital, University Paris VI, Paris, France; Pharmacy Department, Institut Jean Godinot, Reims, France; Medical Oncology, Institut Jean Godinot, Clermont-Ferrand, France; Oncology Department, Bobigny, France; Department of Medical Oncology, Hôpital Européen Georges Pompidou, René Descartes University, Paris, France; Supportive Care Functionnal Unit, Medical Oncology Department, Georges Pompidou European Hospital, Paris, France; Arterial Hypertensive Department, Hôpital Européen Georges Pompidou, Paris, France; Cardiology Department, Hôpital Tenon, Paris, France; Nephrology Department, Pitie-Salpetriere Hospital, Paris, France

Background: MARS is a multicentric, noninterventional, prospective study which first reported the high prevalence of both baseline and de novo hypertension (HTN) and proteinuria (Pu) in patients receiving bevacizumab (Bev) or sunitinib (Su) (Gligorov J et al. SABCS 2013; Ray-Coquard I et al. ASCO GI 2014; Ray-Coquard I et al. ASCO GU 2014; Goldwasser F et al. ECC 2013; Launay-Vacher V et al. ASCO 2013). Methods: It included 1,124 patients, all naive of any previous anti-VEGF treatment. A First Renal Assessment was performed before the anti-VEGF was started with periodic follow-up for 1 year (Renal Follow-up Plan). Univariate (UA) and multivariate analyses (MA) tested the associations of HTN and Pu, at baseline or de novo, with overall survival (OS) (pre-planned). Results: De novo Pu was not associated with reduced OS, in any group. De novo Pu was a statistically significant prognostic factor for better OS in Bev-treated CRC, both in UA and MA. Baseline Pu was prognostic of reduced OS in Bev-treated LC, in UA but not in MA. HTN, either at baseline or de novo, was not found to be associated with OS, in any group. Conclusions: HTN was not associated with OS in our study, even in RCC Su-treated patients. De novo Pu, determined with a urinary dipstick, which is neither invasive nor costly, could be an easily accessible prognostic factor of better OS in Bev-treated CRC. Bev treatment in LC patients experiencing de novo Pu should be discussed since it can be prognostic of poorer OS.

Severe skin rash during vemurafenib treatment: A predictive factor of early positive response in metastatic melanoma?

Author(s): Nora Kramkimel, Lilia Sakji, Audrey Thomas-Schoemann, Jean-Louis Golmard, Elodie Regnier-Rosencher, Christine Theodore, Francois Goldwasser, Michel Vidal, Eve Maubec, Laurent Mortier, Marie-Françoise Avril, Nicolas Dupin, Benoit Blanchet; Assistance Publique–Hôpitaux de Paris, Hôpital Cochin, Paris, France; CHRU Lille – Dermatology Department, Lille, France; Cochin Hospital, AP-HP, Paris Descartes University, Sorbonne Paris Cité, Paris, France; GH Pitié-Salpétrière, Paris, France; Hopital Cochin, Paris, France; Hospital Foch, Suresnes, France; Assistance Publique–Hôpitaux de Paris, Hôpital Bichat, Paris, France; Dermatology, CHRU of Lille, Lille, France; Cochin Teaching Hospital, AP-HP, Paris Descartes University, Sorbonne Paris Cité, Paris, France

Background: Interindividual variability in pharmacokinetics may influence clinical benefit or toxicity of vemurafenib (VMF) in metastatic melanoma (MM) and identification of predictive factors of response and toxicity are mandatory. Methods: From Nov. 2011 to Dec. 2013, VMF plasma concentration was assayed at different times within the first 65 days of treatment, after informed consent, in 55 patients treated with VMF for MM in 3 French Dermatologic Departments. Baseline clinical and biologic characteristics, average VMF plasma concentration per patient for tumor response analysis and absolute exposure at the day of the toxicity, tumor response (PFS, OS, early response rate evaluated after 2 months of VMF) according to RECIST criteria and toxicities (grade ≥2 rash and renal failure, any grade≥3 adverse events) graded using the NCI 4.0 scale, were tested univariately. Candidate variables with pResults: Mean age was 55 years (25-84). 92.7% had stage IV MM, 25.5% had brain metastasis and 76% were PS 0-1. 90.7% of tumor had BRAF V600E mutation and 2.5% V600K. VMF was the first line treatment in 66.7%. Early response rate (i.e CR, PR and SD) was 61.5%. Mean PFS was 5.4 months (0.3-17) and mean OS was 9.0 months (0.3-24+). Mean VMF concentration (n=118) was 47.5 ± 20.4 mg/L. Interindividual variability in dose-normalized VMF exposure was 43%. In multivariate analysis, higher average VMF dose was associated with longer PFS (OR: 0.998 [0.997-0.999], p=0.0025). Grade ≥2 rash, observed in 29.6%, (OR: 5.64 [1.004-31.696], p =0.0495) and daily VMF dose regimen (OR: 1.29 [1.032-1.612],p=0.0256) were associated with early 2-month response. BMI (OR: 0.75 [0.592-0.95], p=0.0173) and VMF plasma concentration (OR: 1.061 [1.017-1.108], p =0.0065) were associated with grade ≥2 rash. Conclusions: This is the first study that highlights correlation between grade≥2 rash and early MM response under VMF. Moreover, relationship between PFS and high VMF average dose underlines importance to keep highest VMF dose even after adverse event. Finally, low BMI and high VMF plasma exposure were identified as risk factors for development of grade≥2 rash

Sentinel node biopsy in the initial evaluation of 87 patients with Merkel cell carcinoma.

Author(s): Eve Maubec, Amandine Servy, Francoise Boitier, Pierre Emmanuel Sugier, Florent Grange, Sandrine Mansard, Thierry Lesimple, Benoit Couturaud, Sebastien Albert, Xavier Sastre, Eduardo Marinho, Agnes Carlotti, Lydia Deschamps, Angelique Girod, Jamal Kassouma, Astrid Blom, Nicolas Dupin, Beatrice Crickx, Marie-Françoise Avril; Assistance Publique–Hôpitaux de Paris, Hôpital Bichat, Paris, France; APHP Hôpital Cochin Department of Dermatology, Paris, France; Department of Dermatology, Hospital University of Cochin, Paris, France; UMR-946 INSERM, Paris, France; Hôpital Robert Debré, Reims, France; Dermatology CHU, Clermont Ferrand, France; Centre Eugène Marquis, Rennes, France; Institut Curie, Paris, France; Head and neck surgery department, Bichat University Hospital, Paris, France; Assistance Publique–Hôpitaux de Paris, Hôpital Cochin, Paris, France; Hopital de Reims, Reims, France; University of Washington, Seattle, WA

Background: Merkel cell carcinoma (MCC) is a rare cutaneous neuroendocrine tumor. The objective was to determine the role of sentinel lymph node biopsy (SNB) in the management of patients with MCC. Methods: A SNB procedure was proposed to all patients, from 5 hospitals, referred for a MCC without palpable regional lymph node from 1999 to 2013. Features between positive (SNB+) and negative (SNB-) SNB patients (pts) were compared using a Fisher’s exact test. Results: Eighty-seven pts with a F/M ratio=0.85, a median age =74 yrs (31-90 yrs; 63% aged at least 70 yrs), a stage T1 (N=54), T2 (N=30) or T3 (N=3), a location on extremities (56%), head and neck (32%) or trunk (12%) were included. Among pts, 8% were immunosuppressed. The frequency of SNB metastasis (20/83=24% of successful SNB) did not differ significantly in T1 and T2 stages (p= 0.4). 19/20 SNB+ pts had a radical lymph node dissection and additional positive lymph nodes were found in 6/19 (31%) pts. Adjuvant radiotherapy was delivered to the primary tumour site in 71/87 pts and to the involved lymph node basin in 18/20 SNB+ pts. During a median follow-up period of 25 mths (4-158 mths) (25 mths in SNB+ and 23 mths in SNB- pts) recurrence developed in 23% of all pts within a median period of 13 mths (3-115 mths). Fifteen per cent of pts died of MCC. Although higher in SNB+ pts (20% vs 9.5%) the frequency of specific death did not differ significantly in the 2 groups (p=0.19). SNB- pts who did not received radiotherapy to the primary tumour site had a worse course of the disease with more recurrences (42% vs 14%; p= 0.04) and a trend to a higher death rate (25% vs 6%; p= 0.08). Conclusions: SNB positivity is not associated with T stage and is high in T1 stage. All pts without palpable lymph node should be considered for SNB. Negative SNB might predict for better outcome A randomized, open-label, phase II study assessing the efficacy and the safety of bevacizumab in neoadjuvant therapy in patients with FIGO stage IIIc/IV ovarian, tubal, or peritoneal adenocarcinoma, initially unresectable. Author(s): Roman Rouzier, Philippe Morice, Anne Floquet, Frédéric Selle, Eric Lambaudie, Virginie Fourchotte, Christophe Pomel, Raffaele Fauvet, Pierre Emmanuel Colombo, Elsa Kalbacher, Philippe Follana, Sandrine Martin, Anne Lesoin, Fabrice Lecuru, Patricia Pautier, Frederic Guyon, Paul H. Cottu, Florence Joly, Youssef Ghazi, Elisabeth Chereau; Institut Curie, Saint-Cloud, France; Institut Gustave Roussy, Villejuif, France; Institut Bergonié, Bordeaux, France; Universite Pierre et Marie Curie, Oncology, GHU-Est Tenon, Paris, France; Institut Paoli-Calmettes, Marseille, France; Institut Curie, Department of Surgical Oncology, Paris, France; Centre Jean Perrin, Clermont-Ferrand, France; CHU Amiens, Amiens, France; Institut régional du Cancer de Montpellier – Val d’Aurelle, Montpellier, France; CHU Besançon, Besançon, France; Département d’Oncologie Médicale, Centre Antoine Lacassagne, Nice, France; Centre François Baclesse, Caen, France; Centre Oscar Lambret, Lille, France; European Hospital George Pompidou, Paris, France; Department of Medicine, Gustave Roussy, Villejuif, France; Institut Curie, Paris, France; Comite Uro-Gynecologie, Centre François Baclesse, Caen, France; Roche, Boulogne-Billancourt, France; Institut Paoli Calmette, Marseille, France

Background: We hypothesize that improving the response rate of stage IIIC or IV and non-optimally resectable ovarian cancer patients to neoadjuvant chemotherapy would improve the optimal debulking rate at interval debulking surgery (IDS) and ultimately the survival. In the ICON7 and OCEANS trials, addition of bevacizumab to chemotherapy has been shown to improve the response rates. We assume that its administration in the neoadjuvant setting would improve the response rate and consequently will help to achieve optimal debulking rate at IDS. Methods: This study, named ANTHALYA, is a multicenter, open-label, randomised phase II study, conducted in 15 sites in France. 90 patients with FIGO stage IIIC/IV ovarian, tubal or peritoneal adenocarcinoma, initially unresectable are to be enrolled. At inclusion, patients are randomised (2:1) to receive 4 cycles of neoadjuvant carboplatin and paclitaxel chemotherapy either combined to 3 cycles of bevacizumab in the treatment arm (not given the cycle before surgery) or alone in the control arm. The control arm will be used to assess the complete resection rate in the arm treated without bevacizumab in the neoadjuvant setting. The primary objective for this study is to evaluate the efficacy of neoadjuvant bevacizumab and chemotherapy measured by the complete resection rate after IDS. Complete resection is defined as the removal of all macroscopic residual tumour at IDS (CC score = 0). The secondary objectives for this study are as follows: (1) to evaluate the safety profile of bevacizumab when added to carboplatin and paclitaxel in the neoadjuvant setting; (2) to assess the efficacy of bevacizumab measured by Objective Response Rate (ORR) for neoadjuvant period and after all courses of treatment, assessed according to RECIST criteria and CA-125 levels; and (3) to evaluate progression-free survival (PFS). Exploratory analysis are conducted to evaluate the biomarkers profile and to explore prognosis and predictive markers. Clinical trial information: NCT01739218.

Prognostic factors of overall survival of stage III or IV adrenocortical carcinomas (ACC): A multicenter ENS@T study

Author(s): Rossella Libe, Isabelle Borget, Cristina Lucia Ronchi, Arianna Ardito, Matthias Kroiss, Jerome Bertherat, Marcus Quinkler, Margarita Bala, Felix Beuschlein, Delphine Vezzosi, Harm Haak, Sophie Leboulleux, Barbara Zaggia, Martin Fassnacht, Eric Baudin; Hopital Cochin, Paris, France; Institut Gustave Roussy, Villejuif, France; Department of Endocrinology Uniklinikum, Wuerzburg, Germany; University Division of Internal Medicine, Orbassano, Italy; Department of Endocrinology, Hôpital Cochin, Paris, France; Department of Endocrinology and Metabolic Diseases, Charité Universitätsmedizin, Berlin, Germany; Medizinische Klinik – Campus Innenstadt. Klinikum der Ludwig-Maximilian-Universität, Munich, Germany; Department of Endocrinology, CHU Toulouse, Toulouse, France; Maxima Medical Center, Internal Medicine, Eindhoven, Netherlands; University Division of Internal Medicine, San Luigi Hospital, Orbassano, Italy; University of Würzburg, Würzburg, Germany

Background: Several reports suggest heterogeneity of advanced ACC.The primary objectives of our study were to provide evidence for an improved staging and grading system and refine the prognostic classification of stage III-IV ACC. Methods: 444 patients registered with the ENS@T ACC database (diagnosed 2000-2009) were enrolled (210 stage III; 234 stage IV). Inclusion criteria were: age>18 yrs, pathological review, synchronous stage III-IV, follow-up available. Primary endpoint was overall survival. Parameters captured were: age, modality of diagnosis, TNM, pathological grading (Weiss score,Ki 67%), R status at first surgery. Univariate and multivariate analyses were performed according two models. Model 1 “prior surgery”; model 2 « post surgery. » Results: 301 patients (68%) died due to ACC (median follow-up: 55.2 mo). Median overall survival was 24 mo. N positive status, but not venous invasion, was confirmed to behave like stage IV. A modified ENSAT (mENSAT) classification was therefore used: stage III in case of invasion in surrounding tissues/organs or vena renalis/cava; stage IVa, IVb, IVc if the number of metasatic organs (including N positive status) was 2, 3 or >3, respectively. ACC tumors with Ki 67>20% and/or Weiss score>6 was confirmed to behave significantly more aggressively. At multivariate analysis within model 1: age>50 years (HR:1.6,P<0.0001), presence of tumor- (HR:1.6, P=0.01) or hormone-related symptoms (HR: 1.6, p=0.03), mENS@T stage, (HR:2.6, HR:3.8 or HR:4.9 for stage IVa, IVb, or IVc respectively, all P<0.0001) were found significant. Within multivariate model 2: age>50 years (HR:1.1,P<0.01), presence of tumoral- (HR:1.7,P=0.01) or hormone-related symptoms (HR: 1.6,P=0.04), mENS@T stage, (HR:2.1,HR:2.7 or HR:3.7 for stage IVa, IVb, IVc respectively, all P<0.0001), R status (for R1/2/x HR:1.7,P=0.0006), grade (Weiss >6 and Ki67> 20%, HR:1.3,P=0.06) were found significant. By combining the different parameters 2 years-OS for mENSAT stage IV ranges between 17-67% in model 2. Conclusions: This largest prognostic study on advanced ACC patients allows to refine the prognostic stratification and future therapeutic management of advanced ACC patients.