ALBERT

Description: Key Points Increased PTK2 expression is associated with improved outcomes in patients with CLL treated with R-FC immunochemotherapy. PTK2 expression represents a useful, novel biomarker for selection of patients who will benefit from R-FC immunochemotherapy.

Description: As previously reported (Colombat, Blood2001;97:101), rituximab (4 weekly doses of 375mg/m²) can lead to high response rates (RR) and prolonged remissions with minimal toxicity as 1st line therapy for low tumor burden FL. We report the final analysis of a trial evaluating long term efficacy and safety of rituximab in untreated low tumor burden FL (GELF criteria). 49 patients (pts) were included in the initial trial (median age 52 yrs), 2 refused consent for the extended F/Up period, and 1 pt died at M12. Molecular bcl2-JH rearrangement was assessed throughout the study. The median F/up was 83.8 mths. Overall best RR, complete/unconfirmed RR and partial RR at D78 were 74%, 50% and 24% respectively. Median PFS was 23.5 mths for the study population. Median duration of response (34 responders at D78, i.e 6 weeks after the last rituximab dose) was 28.6 mths, but response was still maintained without any further treatment in 11 pts after 5 years (24%) and in 7 pts after 7 years (15%). 31/46 pts were bcl2 positive in blood and/or marrow samples before rituximab: 11 (35%) became negative at D50, and 20 remained positive (65%). Median PFS was 37 mths for bcl2-negative pts at D50, and 12 mths for patients remaining positive (p=0.018 Log-rank). Of the 7 pts with sustained response after 7 years, 5 were bcl2 positive at D0, 2/5 became negative at D50, and 5/5 were still negative at M84. At year 7, 4/46 pts have died (1 from myelodysplasia, 3 from NHL), 35/42 have progressed, and 7 have never progressed without any other treatment than the initial rituximab therapy. Time to progression was significantly longer in the bcl2-negative population at D50 (p= 0.018, Log-rank). Duration of response was not correlated with bcl2 status at D50, but was associated with ‘Best response CR/Cru’ (p=0.007 Log-rank). Long-term tolerance was good, with only 13 SAE observed in 13 pts during the additional 4 years of F/Up (4 surgeries for non NHL-related pathologies, 1 node biopsy, 1 sleep apnea syndrome, 1 ischemic cardiopathy, 2 deaths from NHL, 1 depression, 1 pneumonia, 1 erysipela, 1 bronchitis). This long-term update confirms that a single 4-dose rituximab treatment yields durable benefits without the toxicity of chemotherapy for pts with low burden FL : Median PFS of 23.5 mths for the cohort, 28.6 mths for responders and 37 mths for pts turning bcl2-negative at D50, 15% of pts have maintained their response after 7 years, (2bis) the quality (CR/Cru) of the initial response was associated with a longer response duration high overall survival is observed with 4 deaths/46 pts (8.6%).

Description: Key Points Only a complete disruption of TP53 function increases the risk for disease progression in previously treated CLL patients. MiR-34a expression significantly correlates with the predicted TP53 activity in previously treated CLL patients with TP53 abnormalities.

Description: Design/methods: Rituximab added to 8 cycles of CVP (R-CVP) chemotherapy improves time to progression and duration of response in previously untreated patients with stage III/IV CD20 positive follicular NHL compared with CVP alone (Marcus et al Blood2005;105:1417–23). A protocol pre-planned analysis of this study with a median follow-up of 53 months has now been performed. Results: A total of 321 patients (median age 53 years) were recruited. Eighty-three percent of patients in both arms had intermediate to high-risk disease according to the Follicular Lymphoma International Prognostic Index (FLIPI, score 2–5). Median time to progression or death (TTP) in the R-CVP arm was 34 months compared with 15 months in the CVP arm, p

Description: Abstract 1649 Poster Board I-675 Non-conventional γσ T lymphocytes are innate immunity effectors bearing potent anti-tumoral activity, particularly against malignant B cells. IPH1101 is an agonist of γσ T cells, which in the presence of low doses of IL-2 potentiates their direct cytotoxic activity. ADCC is a major molecular mechanism underlying rituximab's efficacy. Increasing the number and the activation state of killer lymphocytes mediating ADCC is therefore believed to be beneficial for therapeutic potency. Since γσ T cells have been found to be capable of mediating ADCC, modulating γσ T cells in the context of rituximab is worth being tested in a clinical trial. The main purpose is to assess the clinical efficacy of IPH1101 with low doses of IL-2, combined with a standard rituximab treatment, in patients (pts) with follicular lymphoma. This is an open label, multinational study consisting of a dose escalation Phase (ph) I-like part followed by a ph II part. The ph I part has shown a good safety and immuno-biological efficacy profile for the highest dose of IL-2. Consequently, the pts of the phase II part were treated with the combination of rituximab (375 mg/m2) administered 4 times weekly, IPH1101 (750 mg/m2) administered i.v. 3 times (every 3 weeks) and IL-2 (8 MIU) administered daily s.c. for 5 days starting on the day of each IPH1101 administration. All pts presented low grade FL which had relapsed after 1 to 4 lines of previous therapy including at least one rituximab-containing line. Inclusion criteria set forth that pts should have no lesion 〉 7 cm. Results We report here the end of recruitment and updated data on 45 patients and more than 100 cycles of IPH1101. Overall safety was good, and most of the drug-related adverse events were, as expected, flu like symptoms of grade 1 or 2. The SAEs reported were 2 hypotensions, 1 bronchospasm, 1 allergic reaction (back pain), 1 glomerular filtration decrease, 1 ALAT elevation, 1 hypertension and 1 asthenia. The immuno-biological follow up demonstrated the very good pharmacodynamic profile of IPH1101 in these patients and these data are presented in details in another abstract from Lucas et al. To date, in the first set of evaluable patients (12 pts) and after at least 3 months post treatment, investigators reported about 75% of response and 50% of CR. Most of the responses are already confirmed by an independent panel. Conclusion These observations confirm the good safety and the biological rationale of this approach. Furthermore, the response rate in this first set of pts is encouraging in the context of previously treated patients. Updated results will be presented at the meeting. Disclosures Laurent: Innate pharma: Honoraria. Lafaye de Micheaux:Innate Pharma: Employment. Solal-Celigny:Innate Pharma: Research Funding. Soubeyran:Innate Pharma: Research Funding. Delwail:Innate Pharma: Honoraria. Ghesquières:innate pharma: Honoraria. Thieblemont:Innate Pharma: Honoraria. Jourdan:Innate Pharma: Honoraria. Beautier:Innate Pharma: Employment. Squiban:Innate pharma: Employment. Rossi:Innate Pharma: Honoraria.

Description: Abstract 2868 Background: GA101 is the first type II, glycoengineered and humanized monoclonal anti-CD20 antibody with Phase I results (NHL/CLL) (Salles, ASH 2008, 2009; Cartron, EHA 2009; Morschhauser, ASH 2009; Sehn, ASH 2009) and phase II results (Indolent NHL) previously reported (EHA 2010). Based on an additional 1 year follow-up period, we report updated efficacy and safety results with single agent GA101, presenting new and encouraging progression free survival (PFS) data indicating a survival advantage of those patients randomized to the high dose cohort (1600/800mg). Methods: 40 eligible patients were randomized to receive GA101 in a low-dose (LD, n=18) or a high dose (HD, n=22) cohort. GA101 was given on d1, d8, d22 and q21 days for total of 9 infusions. In the LD cohort, GA101 was given 400mg all infusions; in the HD cohort, d1 and d8 at 1600mg and 800mg for subsequent infusions. Primary endpoint was end of treatment response (EOR), assessed 4 weeks after last infusion (44 weeks after treatment start). Secondary objectives were safety, pharmacokinetics and PFS. Results: Patients (Table 1) were heavily pre-treated (median 3 prior therapies) with 55% of patients not responding to or relapsing within six months after a previous rituximab-containing regimen (rituximab refractory). There were no significant differences in demographics and baseline tumour burden between the two cohorts and 75% of patients completed all scheduled 9 infusions. EOR was 17% (3 PR, 6 SD, 7 PD, 2 UNK) in the LD cohort, and 55% (2 CR, 10 PR, 6 SD, 4 PD) in the HD cohort. Of note, 6/22 rituximab-refractory patients (5 HD, 1 LD) responded, with a EOR response of 50% in rituximab-refractory patients in the HD cohort (5/10). Responses occurred across all FcγIIIR genotypes in both cohorts: of 3 responders in LD, 2 patients with F/F genotype, other unknown; of 12 responders in HD; 5 F/F, 7 F/V. Responding patients from both LD and HD groups appeared to have higher GA101 plasma concentrations compared to non-responding patients. Median PFS was 6 months [1.1-16.9+ months] and 11.3 months [1.8-14.2+ months] for the LD and HD cohorts respectively (Hazard ratio 0.55 [95% CI 0.24;1.27]). Of 15 responding patients at EOR, 9 have an ongoing response in follow-up (LD=2, HD=7), with 2 PRs converting to CR (LD=1, HD=1) and another PR to CRu (HD). In addition 2 patients (LD=1, HD=1) in follow-up converted from EOR SD to PR, one patient with an ongoing response and the other subsequently relapsing, therefore 10 patients currently have an ongoing response. GA101 was well tolerated in both cohorts with the most common AEs being infusion related reactions (LD 72%, HD 73% of patients), mostly of G1-2. During treatment, related G3-4 hematological AEs were transient neutropenia (n = 3 in HD), febrile neutropenia (n=1 in HD) and thrombocytopenia (n = 1 in HD), four patients experienced at least one G3-4 Infection (LD n=1, HD n=3). Nine patients experienced a total 12 SAEs during treatment period, with 4 related to GA101 (HD n=4; herpes zoster, febrile neutropenia, pancreatitis, neutropenia) and 2 patients in the additional follow-up period, with SAEs of pyrexia (LD) and bacteraemia (HD), both unrelated to GA101. In addition, no B-cell recovery has been observed to date. Conclusion: GA101 single agent is well tolerated, with promising efficacy and provides very encouraging PFS data, in this group of heavily pre-treated relapsed and refractory iNHL patients, indicating a survival advantage for those patients in the HD cohort (1600/800mg). Disclosure: Salles: Roche: Consultancy, Honoraria. Morschhauser:Roche: Consultancy, Honoraria. Wenger:Roche: Employment. Birkett:Roche: Employment. Cartron:Roche: Consultancy, Honoraria; GSK: Honoraria.

Description: We report 28 patients (pts) who experienced late onset severe neutropenia (NCI/CTC grade III/IV) after rituximab. Rituximab had been given for the treatment of DLCL(N=15), follicular lymphoma (N=9), mantle cell lymphoma (N=2), CLL (N=2). Rituximab was administered as a front line treatment (N=11), for recurrent disease (N=17), before ASCT (N=1) or after ASCT (N=5). Rituximab was given as a single agent (N=14), with CHOP (N=9) or with other regimen (N=5). Rituximab was given at a dose of 375mg/m2 at weekly intervals over a period of 4 weeks when used as a single agent, or as a single dose of 375mg/m2 with chemotherapy. Characteriscs of neutropenia are summarized in Table. At the time of the neutropenia, all the pts were in CR or VGPR. In 3 pts, the neutropenia had relapsed. All 3 pts were still in CR, and had not been retreated with rituximab. 3 patients were retreated with rituximab for a recurrence of their NHL after the outbreak of neutropenia and 1 pt experienced a further episode of neutropenia after the reintroduction of rituximab. Tests for anti PMN antibodies were performed in 6 pts. In 4 pts, antibodies bound to the surface of neutrophils were detected by the direct neutrophil immunofluorescence test. No antibody could be detected in the serum. A direct toxic effect of rituximab can be ruled out as granulocytes and uncommitted hematopoietic precursor stem cells do not express CD 20. We propose the following hypothesis. The rituximab-induced depletion of the normal B-lymphocyte population was followed by the acquisition of a new immune repertoire under non physiological condition which could promote the transient production of autoantibodies. Some of these antibodies might target either neutrophils or hematopoietic precursors. Some of these antibodies might also be directed against other cells since other delayed-onset cytopenia have been reported after the administration of rituximab, pure red aplasia in particular. An other possible mechanism can be proposed. Rituximab administration could lead to the production of antibodies directed against the complex formed when rituximab is bound to the FcγRIIIb receptor on the PMN. This hypothesis does not account for the delay between the administration of rituximab and the onset of the neutropenia. Characteristics of neutropenic episodes. Characteristics N *: interval between the last infusion of rituximab and the nadir of neutropenia. **: 1 pt was lost to follow-up for 4.5 months but had a normal blood count subsequently. Median time to neutropenia (range) (weeks)* 15 (4–33) Nadir PMN 10x9/L median (range) 0.135 (0–0.760) Bone marrow aspiration 14 Hypocellular marrow 7 Normocellular marrow with severe reduction in mature neutrophils 6 Normal 1 Fever/sepsis 6/1 Patients treated with G-CSF/duration (range) (days) 12/6 (3–21) Duration of neutropenia (range) (days) Patients treated with G-CSF N=12 4 (2-53) Patients not treated with G-CSF N=16 12 (4–105)** Follow-up from the nadir of neutropenia (range) (month) 6 (0.5–36) NHL/CLL status at follow-up CR or CRu 22 VGPR 1 Progression 5 Status of PMN count at follow-up Normal 26 Neutropenia 2

Description: Abstract 2796 Background: Diffuse large B-cell lymphoma (DLBCL) relapse prognosis is poor, and the optimal salvage treatment is not known. In a previous pilot study, vinorelbine, ifosfamide, mitoxantrone, and prednisone (NIMP) without rituximab has shown a promising efficacy in the setting of relapsed DLBCL. Aims: To evaluate the efficacy and tolerability of the combination of rituximab and NIMP in the treatment of DLBCL in first relapse. Methods: This multicentric, single-arm phase II study included patients 18 to 75 years old, with CD20-positive DLBCL in first relapse (defined as having obtained at least a PR of more than 50% to an anthracycline-based front-line regimen) occurring more than 30 days after the last chemotherapy cycle or more than one year after an autologous stem cell transplantation (ASCT) in first line. Other inclusion criteria were a performance status ≤ 2, absence of CNS involvement, and having signed an informed consent form. Patients with evidence of transformation from indolent lymphoma, primary refractory disease, or positive HIV tests were excluded. Initial and relapse biopsies were centrally reviewed. Patients received intravenous (IV) rituximab 375 mg/m2 D1, ifosfamide 1000 mg/m2 as a continuous infusion from D1 to D5, IV vinorelbine 25 mg/m2 D1 and D15, IV mitoxantrone 10 mg/m2 D1, and oral prednisone 1 mg/kg D1 to D5, repeated every 28 days for three cycles. Pegfilgrastim support (6 mg at D7) was recommended. The primary endpoint was CR/CRu after 3 cycles, and was assessed by computed tomography according to the IWG criteria. Mobilization, consolidation or subsequent salvage therapy was decided at the discretion of the investigator. All the monitoring and data management were performed by the GOELAMS clinical research assistants, with a database lock on July 27, 2010. Results: Fifty patients (21[42%] women and 29[58%] men) were included in 18 centers between December 2004 and accrual closure in April 2010. All patients received at least 1 cycle of R-NIMP. Forty-five patients were available for central pathology review, toxicity and response. The central review of all patient samples confirmed DLBCL histology. Median age at study entry was 62.9 years (range: 34.8–75.6). Median time between first diagnosis of DLBCL and relapse was 18.0 months (range: 2.4–208). The following tumor responses were observed: 67.9 % overall response rate with 20 CR/CRu (43.5%), 11 PR (24.4%), 2 SD (4.4%), and 12 (26.7%) progressed under therapy. Toxicity information was available for 109/120 (91 %) of the first 3 cycles of R-NIMP administered. The following toxicities were observed (all grades, ≥ grade 3 for all cycles): anemia (87%, 8%), neutropenia (66%; 46%), thrombopenia (65%, 14%), elevated liver tests (39%, 0%), constipation (25%, 0%), kidney failure (7%, 0%), nausea (14%, 0%), vomiting (6%, 0%), allergic reactions (5%, 0%), and mucositis (5%, 0%). Twenty-nine infectious events (27%) were observed needing hospitalization in 9 cases. Twenty-nine patients received consolidation therapy at the discretion of the investigator. Of the 11 patients who received 3 additional cycles of R-NIMP, 3 remained in CR/CRu, 1 remained in PR, 4 converted from PR to CR/CRu, and 3 progressed. Among the 11 patients who underwent ASCT, 9 were in CR at the end of the procedure, one patient died of toxicity and 1 progressed. For the 12 patients mobilized after a R-NIMP cycle, a median of 1 apheresis (range 1–4) was necessary to harvest a median of 3.85 × 106 CD34+ cells/kg. The median time to second progression or relapse (TTP2) was 11.4 months, and the median survival of 55.5 months. On multivariate analysis, the variable associated with a longer TTP2 was the achievement of CR/Cru (RR: 0,12; CI95%: 0.03–0,39; p=0.0006). Within the subgroup of patients having received a consolidation treatment, having received an ASCT was associated with a longer TTP2 (RR: 0.20; CI95%: 0.04–0.98; p=0.047). Time to first relapse or previous rituximab exposure did not affect TTP2 nor OS, whereas relapse-IPI (as a continuous variable, by 1 additional risk factor) was associated with a poor survival (RR: 2.59; CI95%: 1.25–4.45; p=0.008). Conclusions: R-NIMP is a well-tolerated regimen, yields a high complete response rate, and allows for successful mobilization of CD34+ cells. This regimen is a suitable salvage treatment for relapsed DLBCL prior to appropriate consolidation. Further investigation is warranted. ClinicalTrials.gov number: NCT00842595 Disclosures: Off Label Use: Vinorelbine in Non-Hodgkin's Lymphoma (Off-label in France).

Description: Chronic lymphocytic leukemia (CLL) is a slowly progressing but incurable disease with high unmet medical needs. As of 2014, there were few reliable publication sources of epidemiology data for CLL in France, particularly with regard to relapsed patient population. In addition, epidemiology registries provide useful data on disease incidence but no information can be directly obtained on the number of relapsed patients. At the same time, several new and effective drugs will be made available to the Hematologist/Oncologist community. As a consequence, there will be an increase in the complexity of therapeutic algorithms. We undertook an innovative modeling approach to address the need for a greater understanding of the number of relapsed patients available to these new therapies. The objective of the model was to determine the yearly incidence of first line and later lines of therapy patients from 2014 to 2018 and in consequence, provide relevant data to help determine which population might be in need for the new compounds and bring accurate data that could be implemented into a prospective economic model. First, we used the Registre Régional des Hémopathies Malignes de Basse-Normandie (RRHMBN) to collect real world patient characteristics, such as tumor staging (based on Binet Classification), fitness status, and deletion 17p. The RRHMBN is dedicated to Hematological malignancies and supported by official Health Authorities: Institut National du Cancer(INCA) and Institut National de Veille Sanitaire (INVS). Following that, we applied these data to the French National Network of Registers FRANCIM incidence and survival data. 27.6% of patients were under 65 years. 50% of patients over 65 years had unfit status. 77% of patients had stage A, 14% stage B and 8% stage C. 2/3 of stage A were not treated and 1/3 evolved to stage B or C. 5% of patients had a 17p deletion at diagnosis (Table 1 below). Table 1. Yearly number of patients newly diagnosed Figure 1 Figure 1. Finally, we applied the results from medical literature to progression free survival for each standard of care treatment in our epidemiological model, Fludarabine (F) – Cyclophosphamide (C) – Rituximab (R) regimen with a PFS equal to 51.8 months for patients under 65 years. (R) – Bendamustine (B) regimen with a PFS equal to 33.9 months for patients fit and over 65 years. For unfit patients over 65 years, ofatumumab (O) – Chlorambucil (C) regimen with a PFS equal to 27 months. The table below (Table 2) shows the yearly number of patients eligible for treatment in first and subsequent lines of therapy as produced by our model. Table 2. Yearly number of patients in need for treatment in first and subsequent lines Figure 2 Figure 2. Our modeling approach compensates for a lack of real world information on relapsed CLL. We have shown that there is a significant number of relapsed patients who are in need of new treatments in France. The model is able to answer the question of how many patients are in need of novel therapies in later lines through 2018 and provide for CLL investigators efficient prospective data that can be integrated into medico-economic model development. Furthermore, these data address a current medical need for the development of a new CLL algorithm for patient management, necessitated by the expected arrival of new treatment options for patients in the relapsed setting. Disclosures No relevant conflicts of interest to declare.

Description: As of 2014, there were few reliable publication sources of epidemiology data for Non-Hodgkin Lymphoma (NHL) in France, particularly with regard to indolent forms such as Follicular Lymphoma (FL), Waldenström Macroglobulinemia (WM) and Marginal Zone Lymphoma (MZL). In addition, epidemiology registries provide useful data on disease incidence but no information can be directly obtained on the number of relapsed patients, especially in the rare iNHL subtypes. At the same time, many new and effective drugs will be made available to the Hematologist/Oncologist community. As a consequence, there will be an increase in the complexity of therapeutic algorithms. We undertook an innovative modeling approach to address the need for a greater understanding of the number of relapsed patients available to these new therapies. The objective of the model was to determine the yearly incidence of first line and later lines of therapy patients from 2014 to 2018 and in consequence, provide relevant data to determine precisely which population might be in need for the new compounds and bring accurate data that could be implemented into a prospective economic model. First, we used the Registre Régional des Hémopathies Malignes de Basse-Normandie (RRHMBN) to collect real world patient characteristics, such as patient subgroups and tumor staging. The RRHMBN is dedicated to Hematological malignancies and supported by official Health Authorities: Institut National du Cancer(INCA) and Institut National de Veille Sanitaire (INVS). Then, we applied these data to the French National Network of Registers FRANCIM incidence and survival data (table 1 below). Table 1 Yearly number of patients newly diagnosed Table 1. Yearly number of patients newly diagnosed Finally, we applied the results from medical literature to progression free survival for each standard of care treatment in our epidemiological model. The table below shows yearly the number of patients in need for treatment in first and subsequent lines of therapy as produced by our model. Table 2 Yearly number of patients in need for treatment in first and subsequent lines Table 2. Yearly number of patients in need for treatment in first and subsequent lines Our modeling approach compensates for a lack of real world information on relapsed iNHL. We have shown that there is a significant number of relapsed patients who are in need of new treatments in France. The model is able to answer the question of how many patients are in need of novel therapies in later lines through 2018 and provide for iNHL investigators efficient prospective data that can be integrated into medico-economic model development. Furthermore, these data address a current medical need for the development of a new iNHL algorithm, necessitated by the expected arrival of new treatment options for patients in the relapsed setting. Disclosures No relevant conflicts of interest to declare.