Background Long‐term treatment‐free remission (TFR) represents a new goal for chronic myeloid leukemia (CML). In clinical practice, tyrosine kinase inhibitor (TKI) dose reductions can be considered a means of preventing adverse effects and improving quality of life. We hypothesized that administration of low‐dose TKIs before treatment discontinuation does not impair TFR in patients with CML who have a deep molecular response (DMR, ≥MR4). Methods We conducted a retrospective analysis of 77 patients with CML who discontinued treatment with TKIs. Twenty‐six patients had been managed with low‐dose TKIs before stopping treatment. Patients were to be exposed to TKIs for ≥5 years and to low‐dose TKIs for ≥1 year and in DMR for ≥2 years. The loss of major molecular response (MMR) was considered a trigger for restarting therapy. Results In the low‐dose group, 61.5% of patients received second‐generation TKIs, and dose reduction was ≥50% for 65.4% of patients. With a median follow‐up of 61.5 months, TFR at 12 months was 56.8% in the full‐dose TKI group and 80.8% in the low‐dose group, and TFR at 60 months was 47.5% and 58.8%, respectively. The median time to molecular recurrence (≥MMR) from TKI discontinuation in the entire cohort was 6.2 months. All patients quickly achieved MMR after resuming TKI therapy. Results appear independent of both dose reduction and potential pretreatment with interferon‐α. Conclusion This retrospective study shows that TFR was not impaired by low‐dose TKI regimens before TKI cessation in Patients with CML. Nevertheless, prospective randomized clinical trials must be undertaken to analyze the probability of successful TFR in patients managed with TKI dose de‐escalation strategies before TKI discontinuation.

In diffuse large B-cell lymphoma (DLBCL), the age-adjusted International Prognostic Index (aaIPI) score is currently used to predict patient outcomes and to choose the best therapeutic treatment. Body mass index (BMI) and gender are occasionally sited as prognostic factors; however, their value has never been studied in a large series of patients included in prospective clinical trials in the rituximab era. To assess the impact of BMI and gender on OS and PFS independently of the aaIPI score, we pooled 985 patients that were prospectively included in GELA studies and uniformly treated with R-CHOP. Univariate analysis indicated that high aaIPI and male gender were associated with a worse PFS, whereas high (>25) or low (<18.5) BMI scores were not. High aaIPI score was the only factor predictive for OS. In a multivariate analysis, including aaIPI score, gender, BMI, and interaction between BMI and gender, aaIPI remained the strongest predictive factor, and BMI < 18.5 was significantly associated with a worse OS but not PFS. In conclusion, in the rituximab era, the aaIPI score remains the major predictor of outcome in DLBCL patients; however, male gender and low BMI seem to impact outcome.

No abstract

796 Background: The LNHCP93 trial is an intensive high-dose methotrexate and cytarabine containing chemotherapy (CT) derived from CT regimens used for Burkitt lymphomas (C5R protocol) followed by brain radiotherapy (RT) showing favourable long-term survival in Primary CNS Lymphoma (PCNSL) patients younger than 60 years: the complete response (CR) rate was 33%, the 5-year progression-free survival (PFS) and overall survival (OS) rates were 31% and 42%, respectively with however 9% of toxic death. Rituximab improves outcome of systemic Diffuse Large B-Cell Lymphoma (DLBCL) but it is unclear if its addition is useful in PCNSL with DLBCL histology. Intrathecal (IT) liposomal cytarabine (Depocyt) which shows a long-CSF half-life and better efficacy compared to IT methotrexate and free cytarabine in lymphomatous meningitis is also an attractive treatment for PCNSL. Methods: In order to improve antitumor activity of C5R protocol, we prospectively evaluated the addition of intravenous rituximab and IT liposomal cytarabine to CT before RT for 18 to 60 years old immunocompetent patients with PCNSL. The primary objective of this study is the complete and unconfirmed complete response rate (CR/CRu) after immuno-CT, according to International Primary CNS Lymphoma Collaborative Group response criteria. The number of patients was calculated in order to improve the 30% CR/CRu rate of LNHCP93 study up to 50%. The secondary objectives were side effect profile including acute and late neurological toxicities, PFS and OS. Patients received two courses of methotrexate, cyclophosphamide, doxorubicin, vincristine, prednisone (COPADEM) followed by two courses of methotrexate, cytarabine (CYM) administred at 21-day intervals from day 1 to day 21. At each day 1, intravenous rituximab 375mg/m2and at each day 3 IT liposomal cytarabine 50mg were infused. After immuno-CT, a brain RT was planned for all patients. A new response assessment was performed after RT. Patients were followed every 4 months for the first two years and then every 6 months. Neurotoxicity was evaluated by a Mini-Mental State Examination (MMSE) test. Results: Fifty-three PCNSL patients with a confirmed diagnosis of CD20 positive DLBCL were included between August 2007 and September 2011: median age was 55 years (range, 36–60), 57% were male, 42% had a performance status >1, 55% had an involvement of deep structures of brain, 45% a high CSF protein level and 36% a high LDH level. Median MMSE score assessed for 42 patients at baseline was 26 (range 3–30). After immuno-CT, 22 patients achieved a CR and 13 a CRu (CR + CRu, 66%), 7 a partial response (13%), 6 had a stable or progressive disease (11%) and 5 (10%) were not evaluated due to early death (n=2) or toxicities (n=3). For 45 patients who completed the fourth cycles, CR/CRu rate was 75.6%. Thirty-one patients (58%) presented at least one serious adverse event, mainly infections (25 patients, 47%). At least one red blood cell and platelet transfusions were given in 94% and 68% of patients, respectively and 57% of patients experienced a febrile neutropenia. Forty-two patients (79%) underwent RT, 23 of them with a reduced whole brain RT (median 26 Gy) and a boost to the tumor bed. With a median follow-up of 22 months, 15 patients progressed or relapsed (28.3%) and 13 patients died (24.5%), 3 (5.7%) of them from acute toxicity. The 2-year PFS and OS rates are 67% (95%CI, 50.0% to 79.4%) and 82% (95%CI, 67.0% to 90.7%), respectively. Median MMSE score was 27 (range, 26–30) for the 7 patients evaluated between 12 and 18 months after the beginning of treatment. Conclusions: With the addition of intravenous rituximab and intrathecal liposomal cytarabine, C5R protocol provided higher CR/CRu rate and promising outcome results with a 2-year PFS of 67% as compared to historical controls. No additional severe toxicities were observed compared to C5R protocol and with short follow-up no neurocognitive decline was observed. Disclosures: Off Label Use: Depocyt (intrathecal liposomal cytarabine) is off-label used in first line therapy of Primary CNS Lymphoma but approval for treatment of lymphomatous meningitis. Morschhauser:Roche: Honoraria. Blay:Roche: Honoraria, Research Funding.

No abstract