ALBERT

Description: Objective: In this pilot study we evaluated the clinical activity, toxicity and mobilizing capacity of a new short-course (bi-weekly), dose intensive, cytoreductive/mobilizing salvage regimen (R-GIFOX) combining the cross-synergistic agents Gemcitabine (G), Ifosfamide (Ifo), Oxaliplatin (Ox) and Rituximab (R), in patients with relapsed and refractory CD20+ NHL. Based on the predicted clinical activity, tolerability and synergy among drugs, the R-GIFOX regimen may offer an effective and less toxic alternative to Cisplatin/ARA-C-based salvage regimens, also for patients aged or unfit for high-dose procedures. Patients and methods: Patients were scheduled to receive three courses of therapy followed by mobilization and ASCT or three more courses if ineligible for ASCT. Therapy was delivered on a compassionate basis after written informed consent. R-GIFOX consisted of R (375 mg/m2, d 1), G (1000 mg/m2, d 2), Ox (130 mg/m2, d 3) and Ifo (5 g/m2, d 3), as a 24-hour single infusion in patients aged ≤ 65 years, or fractionated over 3 days (dd 3–5) in older patients. Treatment was given every two weeks with G-CSF support (5 mcg/kg/day, dd 6–11; 10 mcg/kg/day at the 3rd course for stem cells mobilization). Responses were evaluated after three courses by the integrated FDG-PET/IWC criteria, and reassessed at the end of the entire program. Results: Fourteen patients (median age 63 years, range 37–78 years) with relapsed (n = 9) or primary progressive (n = 5) aggressive [diffuse large cell (n=7), mantle cell (n=4), follicular G3b (n=3)], advanced (stage IV = 71%), poor risk (IPI 3–5 = 50%; median number of previous therapy=2, r 1–4) NHL, were accrued. Forty-nine total courses were delivered (median 4, range 1–6); thirteen patients completed at least 3 courses of therapy and were evaluable for response. Actual dose intensity of the first 3 courses was 81%, 83.5%, and 86.5% for G, Ifo and Ox, respectively. CTCAE v3.0 G3/G4 thrombocytopenia was present in 26 % of courses, G3/G4 neutropenia in 22%; febrile neutropenia and infections in 8% and 6% of cycles, respectively. The ORR assessed after three courses of R-GIFOX was 77%, with 7 complete responses (54%; CR=5; CRu=2) and 3 partial; CRu converted to CR at BM biopsy after 6 courses. According to age, the ORR was 67% (4 CR, 2 PR) and 80% (3 CR, 1 PR) for patients aged ≤ 65 years and those older, respectively. According to disease status, the ORR was 40% (1 CR, 1 PR) and 89% (6 CR, 2 PR) for primary refractory and relapsed patients, respectively. Among CRs no patient has relapsed at a median time of 5 months (range, 2+ − 10+). Effective CD34+ cell mobilization was obtained in 4 out of 6 eligible patients and 2 already received ASCT. Failure free survival was 79.6%. In mantle cell lymphoma 2 CRs and 1 PR were obtained, including 2 molecular remissions (BM, PB). Conclusions: Based on the results of this pilot study, R-GIFOX was feasible, tolerable, effective and able to mobilize peripheral stem cells in patients with reccurrent aggressive NHL. It enabled the achievement of effective dose-intensities and high response rates also in the older patients. Finally, the R-GIFOX regimen showed clinical activity also in ’difficult’ histotypes such as mantle cell lymphomas.

Description: Background: Multiple extranodal localizations, nodal spreading and bone marrow (BM) involvement have been associated with shorter overall and event-free survival in Extranodal Marginal Zone Lymphoma (EMZL). No optimal treatment has yet been defined and, in this setting, Rituximab has demonstrated only moderate single agent activity while the role of purine analogs is still controversial, although promising in terms of objective response rate. In order to verify a potential synergy of these agents we conducted a prospective phase II trial evaluating a sequential treatment with 2-CDA, a purine nucleoside analog with proapoptotic and cytotoxic activity on both proliferating and quiescent lymphoid cells, and Rituximab in patients (pts) with untreated disseminated EMZL. Patients and methods: Pts with histologically confirmed MZL, ECOG 〈 2 and Stage IV disease were enrolled. 2-CDA was administered at a dose of 0.12 mg/kg on 5 consecutive days as a 2-hour infusion, every 4 weeks for 6 courses on outpatient basis. Thereafter, 4 doses of Rituximab (375 mg/m2, i.v., every 2 weeks) were delivered. Prophylaxis with TMP-SMX was required and filgrastim introduced if grade 3–4 neutropenia had occurred. Responses were evaluated after the 4th course of 2-CDA and at the end of the entire program. Results: Twenty-three pts (median age 67 yrs; r 34–81) with advanced [multiple extranodal site (n=16), BM (n=15), distant/diffuse nodal involvement (n=14), orbit (n=8), multicentric skin (n=3), gastric (n=4), leukemic (n=2), lung and pleura (n=3), salivary glands (n=3), Waldeyer’s ring (n= 1), breast (n=1)] EMZL were accrued. B symptoms were present in 4 pts, abnormal LDH in 6. No pts had been irradiated and only 2 had received surgical resection due to emergency (ileum) and diagnostic (lung) procedures. All pts completed treatment program and were evaluable for response. The median number of delivered courses of 2-CDA was 6 (r 3–6). The overall response rate was 91 % [after the 4th course: CR=4, CRu=11, PR=6; at the end of treatment (i.e. including Rituximab): CR= 18 (78%), PR=3 (13%)]. Interestingly, a histological CR was achieved in 13 out of 15 pts with BM involvement (81%). Only 2 pts, with B symptoms at presentation, had a therapeutical diversion at the 3rd and 4th course due to minor response and persisting symptomatic disease. Extrahematological toxicity was negligible. CTCAE Grade 3 and 4 neutropenia occurred in 9 and 5 (22%) pts, respectively and febrile neutropenia occurred in 6 pts (26%). Grade 4 Thrombocytopenia and lymphocytopenia developed in 4 (17%) and 3 (13%) patients, respectively. Delayed grade 3 leuko- and thrombocytopenia occurred two months after 2-CDA withdrawal in three pts and slowly recovered in the following 6–8 weeks. No late opportunistic infection, nor disease recurrence have to date occurred. Failure free survival was 86 % at median follow up of 18 months (range, 4–34). Conclusions: Our prospective study shows that 2-CDA plus sequential Rituximab represents an effective strategy for the upfront treatment of disseminated EMZL. It is feasible in outpatient setting, and associated with favorable early and late toxicity profiles. While a longer follow-up will assess its impact on disease natural history, our results show for the first time that 2-CDA plus sequential rituximab may achieve very high CR rates even in poor prognosis pts with extragastric and widespread EMZL.

Description: Despite the mature results of the new intensive treatments, ABVD is still considered the standard therapy of the HL. Here we report the 2-year results of a phase 2 study which explore the possibility to ameliorate the performance of this (g)old schedule. Primary End-Points: CR and early cardio-toxicity. Secondary End-Points: survival (FFP,OS) and late cardio-toxicity. 23/25 CR (CI 90%) to validate the study (80% standard vs 90% DD-DI ABVD, CI +/− 10% one tail test, 5% sig stat level). Modifications of the standard (st) ABVD and strategy concepts are summarized as follow: Each patient received a total of 6 cycles without RT Adryamicin (ADM) was escalated from 50 to 70 mg/m2 in the cycles 1,2,3,4. The cumulative dose of ADM was 380 mg/m2. The intercycle period was shortened from 28 to 21 days for all 6 cycles; the 4 drugs were delivered at day 1 and 11 of each cycle G-CSF was given from d4 to d8, and from d14 to d18 of each cycle The therapy program was driven by dynamic indicators such as interim-PET and CT fusion images. Normalisation of PET images at the end of 2nd cycles was indicator of early CR, while the persistence of PET+ lesion(s) at the end of the 4th cycle was indicator of failure and consequently the treatment was shift to a high-dose CT/CD34+ rescue program. RT was eliminated. At the end of 6th cycle, PET- residual masses were monitored by PET/TC fusion images. An historical group of 94 HL pts treated with 6–8 cycles of st-ABVD +/− RT was used to compare response, survival and toxicity data. Compared to st-ABVD the theoretical relative dose-intensity (RDI) values of drugs were 1.86 for ADM, 1.33 for DTIC, Bleomycin and Vinblastine, and 1.46 for the entire schedule. Administered RDIs of the schedule (delivered cycles = 184) ranged between 1.10 to 1.55 (median value 1.39). Results: from June 2004 to August 2006 thirty-nine pts were enrolled (tab1). As the 8th August 2006 thirty-five out of 39 pts have performed the interim-PET-TC fusion images and thirty-two completed the treatment: 100% of valuable pts obtained the early CR (35/35) and CR (32/32). All pts are alive and free of disease (tab2 and fig.1). Cardiac functions were monitored by ECG, Ecography, MUGA scan, pro-BNP, and troponin T. Toxicity was mild and similar to standard ABVD. Dose-dense and dose-intense ABVD is feasible, well tolerated and highly active in newly diagnosed patients with advanced/intermediated Hodgkin’s lymphoma: six cycles without RT seem superior to 6–8 cycles of ABVD +/− RT (fig 1) and may be a promising program for optimal long-term results. tab1 presentation features Features total 39 100% Early unfavourable 9 23% Advanced 30 77% Age 〉 45 –yr 4 10% Bulky 19 44% Stage IV 13 33% B symptoms 25 64% Extranodal 9 23% VES 〉 50 mm 20 51% LDH Ratio 〉 1 15 38% IPS ≥ 3 10 26% N. Sites〉3 28 72% Male 13 33% tab.2 Response and Survival N. of patients percent *n°. of valuable pts as 8 August 2006 Early CR 35/35* 100% CR 32/32* 100% FFP-2yr 32/32* 100% OS-2yr 32/32* 100% Freedom From Progression Freedom From Progression

Description: To improve ABVD results we first developed a protocol which adds G-CSF to the standard ABVD treatment. From March 1997 to May 2004 69 patients with HL were treated with ABVD+G-CSF and 22 with a standard ABVD. Recently we designed a new dose-dense and dose-intensity ABVD scheme (escABVD-21) for advanced HL; in this new schedule the adriamycin was escalated from 25 to 35 mg/m2 (cycles 1,2,3,4) and the inter-cycle period was shortened from 28 to 21 days (for all 6 cycles); primary G-CSF was administered from d3 to d8 and drugs were delivered at d10 and d21 of every cycle. From June 2004, 19 patients were treated with this protocol. Relative dose-intensity (RDI) was calculated for any of these 110 newly diagnosed HL patients treated with ABVD. The results were also compared with a historical group of 70 patients who had undergone hybrid MOPP/ABVD. HL patients received from 4 to 8 cycles of CT +/− IF-RT. Patients were divided in 4 groups according to the RDI. The first group included 20 (11%) pts with RDI less than 0.80; the 2nd group, 64 (36%) pts with RDI values between 0.80 and 0.95, the 3rd group, 74 (42%) pts with RDI values between 0.96 and 1.10 and, finally, the 4th group included pts with RDI values of more than 1.10. In Tab 2 we report the CR, EFS and OS rates according to the 4 levels of RDI. Figures 1 shows EFS curve according to Kaplan-Meyer. Response and survival rates of groups 1,2,3 and 4 were: 50%vs91%vs97% vs 100%, for CR (p1.10. In particular, the new dose-dense and dose-intensity escABVD-21 protocol seems very promising in terms of complete response and toxicity profile: 19/19 pts (100%) obtained an early CR; (PET negative at the end of the 2nd cycle), and, as on 8th August 2005, all these19 patients were disease-free. The dose-escalation of adriamycin and the dose-density of the schedule were well-tolerated; toxicity was mild. These results show that subptimal RDI may compromise outcomes proportionally to the level of RDI reduction. On the contrary, Primary G-CSF permits to deliver dose-dense and dose intense schedules such as escABVD-21 maintaining the same profile of toxicity of standard ABVD, higher RDI levels, and consequently, a significant impact on complete response and survival rates. tab1 Presentation features n. of pts male sex age〉45 yrs advanced stage E sites〉1 bulky B symptoms IPI score ≥ 3 180 92 43 148 23 80 83 56 % 51% 24% 82% 13% 44% 46% 31% tab.2 Response and Survival according to 4 RDI levels RDI level (range 0.5–1.5) N. of pts. (%) CR rate Fisher’s Exact test (2-sided) EFS rate log-rank statistic 0S rate log-rank statistic overall 180(100%) 90% 80% 88% 1.10 20 (11%) 100% 100% 100% significance

Description: Abstract 267 Introduction: The tumor cells of HL, the Hodgkin and Reed-Sternberg (H-RS) cells, derive from germinal center B-cells with a deranged B-cell transcription program due to epigenetic silencing and acquired genetic lesions. Tissue H-RS cells are surrounded by a preponderant infiltrate of mixed non-malignant reactive cells, including B-lymphocytes, which provide essential signals for their survival and proliferation. Small percentages of B-cells, clonally related to H-RS cells, were found in the blood of HL patients (pts), suggesting they may represent putative HL-initiating cells (Jones, 2009). Since the serum free light chain (sFLC) assay detects and quantifies monoclonal and polyclonal B-cell populations expanding in lymphohemopoietic tissues, we exploited the sFLC testing to further explore the biologic significance of B-lymphocytes in HL. Patients and Methods: Frozen (-80°C) serum samples from 119 untreated cHL pts (48% males), with normal renal function and serum immunochemistry, were assayed by immunonephelometry (Freelite, The Binding Site, Ltd., UK). After quantization of free κ and λ concentrations (normal ranges, κ: 3.3-19.4 mg/L; λ: 5.71-26.3 mg/L), sFLC κ/λratio was calculated (reference range 0.26-1.65). The median age was 31 years (r, 15–70), with 22% aged ≥ 45 years. Histology was nodular sclerosis in 85 pts (71.4%) and 72 (60.5%) were in stage I–II. According to GHSG criteria, 16% had early favorable, 29% early unfavorable and 55% advanced disease. The International Prognostic Score (IPS) was of 0-2 and ≥ 3 in 66.4% and 33.6% of pts, respectively. Following ABVD (4-6 courses ± RT), the median Event-free survival (EFS) was 78 mo.s [95% CI, 68-88] for the entire population and 75 mo.s [95% CI, 60-90] and 64 mo.s [95% CI, 50-77] for pts in early and advanced stage, respectively. Results: Elevated κor λsFLC concentrations were found in 47% (median 29.55 mg/L; r, 19.44 - 64.70) and 29.4% (median 32.70 mg/L; r, 26.60 - 79.77) of pts, respectively. In 30 pts (25.2%) levels of polyclonal κ and λ sFLC were concurrently elevated. The sFLC κ/λratio was abnormal in only 7.5% of pts (clonal λ: 8/119; clonal λ: 1/119). The presence of high sFLC levels correlated with lymphopenia (〈 0.6 × 109/L; p= 0.04), leukocytosis (WBC 〉 15 × 109/L; p=0.03), ESR (〉 50; p=0.04) and unfavorable IPS (≥ 3; p=0.03), but not with EBV status and risk factors such as stage, B symptoms, bulky and extra nodal disease, LDH and albumin. The association with leukocytosis may in part result from inhibition of spontaneous neutrophils apoptosis exerted by FLC (Cohen, 2003). Most interestingly, while we found no significant association with response rate, baseline elevation in sFLC predicted for EFS in 51 evaluable pts with early stage disease. Patients were divided into tertiles and best break point value for predicting EFS coincided with the upper limit of the highest tertile for both κ (〉25.53 mg/L) or λsFLC (〉26.66 mg/L) (Figure 1). The pts in the top tertile had the worst outcome compared with the 2 lower tertiles (κ, p=0.015; λ, p=0.002). Outcomes for the 2 lower tertiles were comparable. Data remained significant after stratification for the IPS. In contrast, baseline sFLC levels, κ or λ, were not predictive for EFS in pts with advanced disease. Interestingly, sFLC levels in a control group of 30 pts in continuous CR from 2 years, were within the normal ranges. Conclusions: We have shown that about 50% of cHL pts displays elevated levels of sFLC mirroring the presence of a consistent polyclonal B-cell expansion at diagnosis. The role of reactive B-cells in HL is poorly understood. While some data indicate that high intratumoral B-cell counts may predict for a better outcome, the activity of rituximab in cHL, regardless of CD20 expression on H-RS cells, was suggested to result by depletion from the HL microenvironment of normal B lymphocytes required for tumor cell growth (Younes, 2003). Our study support that elevated sFLC levels may reflect an increased polyclonal B cell activity in cHL microenvironment which appears to negatively influence the outcome of pts in early stage disease. This effect is lost in advanced disease, suggesting that rituximab might result more active for pts in early than advanced stages. That putative HL-initiating small B-cells may emerge from the expanded polyclonal B-cell population present from the early phases of disease development, is an intriguing possibility. Disclosures: Marchei: Radim, Italy: Employment. Amoroso:The Binding Site, Ltd: Consultancy.

Description: Abstract 1664 Poster Board I-690 Background Delivering of dose-dense R-CHOP represents a critical factor for improving response rate and survival outcomes in poor risk DLBCL patients (pts) but it may turn a very challenging goal to achieve in elderly pts with severe cardiac comorbidities. In this subset of pts, tachyarrhythmias, conduction disturbances, ventricular dysfunction, and a pericarditis-myocarditis syndrome, may occur during or immediately after the administration of doxorubicin hydrochloride, independently from total cumulative dose. This may lead to early treatment discontinuation and even death, strongly hampering the chance of receiving adequate chemotherapy. We have conducted a pilot study first exploring the feasibility of a dose-dense biweekly R-CHOP-like regimen including non-pegylated liposome-encapsulated doxorubicin in elderly pts at a high risk for both disease presentation and cardiac conditions. Liposomal anthracyclines, beyond the potential of a more selective uptake by lymphoma cells, may broaden the feasibility of a dose-effective treatment through a reduction of early and late cardiotoxicity. Patients and Methods Untreated pts with poor R-IPI risk DLBCL and moderate to high ‘life threat’ impact cardiopathy (NIA/NCI index) were scheduled to receive six courses of a R-COMP-14 regimen in which an equal dose (50 mg/sqm) of non-pegylated liposome-encapsulated doxorubicin (Myocet®) was substituted for doxorubicin hydrochloride within the R-CHOP platform. Treatment was given every two weeks with G-CSF support. Responses and endpoints were evaluated according to the International Harmonization Project with treatment failure defined as discontinuation of R-COMP-14 for any reason (e.g., disease progression, toxicity, initiation of new treatment without documented progression, or death). Results Thirty-seven pts, median age 72 years (r, 62-82) with pre-existing heart disease [history of myocardial infarction (n=10), atrial fibrillation (n=8), valve disease (n=6), conduction disturbances (n=9), hypertensive cardiomyopathy (n=4)] and R-IPI score ≥ 3, were accrued in this prospective study. Thirty-three pts were in advanced stage [stage IV (n=25) + stage III (n=8): 89%], 18 (49%) had B symptoms, 29 (78%) displayed abnormal LDH and 11 (30%) presented involvement of 〉 2 extra nodal sites. According to Charlson, median Comorbidity Index and Score were 5.5 (r, 2-8) and 8.5 (r, 5-11), respectively. Pre-treatment left ventricular ejection fraction (LVEF) ranged between 42% and 67% (median, 57%). One-hundred eighty-four total courses were delivered (median 6, range,1-8). The median time to recycling was 15.9 days (r, 13-29) and 67% of pts received the 6 scheduled courses. Hematologic toxicity was tolerable, with CTCAE v3.0 G3/G4 thrombocytopenia and G3/G4 infections being recorded in 6 (16%) 5 (12%) pts, respectively. Seventeen failures were documented; they were due to progression (n=2), early death (n=3), treatment discontinuation (n=6) and disease recurrence (n=6). Early deaths occurred at day +12, +13 of the 1st course and day + 18 of the 2nd course, accounting for a treatment mortality rate of 8%. Treatment was discontinued because of a 〉 15% reduction in LVEF occurring at the 3rd and 4th courses (n=3) and the occurrence of symptomatic arrhythmia (n=3). The overall response rate was 71 % (95% CI: 53-84) with 24 complete responses and 2 partial responses. Time to Treatment Failure and Time to Progression at 28 mo.s were 51% and 63%, respectively. Disease Free Survival was 69% at a median follow-up of 14 mo.s. Discussion The present study documented the attractive therapeutic potential of the R-COMP-14 regimen in the setting of poor-risk elderly DLBCL pts with moderate and high impact cardiac comorbidity, in terms of both response rate and time to progression. Survival endpoints were more than satisfactory in a such unfavourable cohort of subjects for which treatment recommendations are still controversial, also due to lack of international guidelines on cardioprotective agents and strategies. Although cardiac events have occured in a fraction of pts, very early and independently from cumulative doses, they did not substancially affect the feasibility of the regimen, provided that a close clinical monitoring was ensured. R-COMP-14 represents an active dose-dense therapeutic platform for pts with cardiac comorbidity. Disclosures No relevant conflicts of interest to declare.

Description: Background: We evaluated the clinical activity, toxicity, and stem cells mobilizing capacity of a short-course (bi-weekly), dose intensive, salvage regimen (R-GIFOX) in patients with relapsed and refractory Hodgkin Lymphoma. The agents Gemcitabine, Ifosfamide, Oxaliplatin and Rituximab employed in this combination have been all accounted of cross-synergistic activity in both preclinical and early clinical studies in Hodgkin lymphoma. Patients and methods: Patients were scheduled to receive three courses of therapy followed by mobilization and ASCT or three more courses if ineligible for ASCT. R-GIFOX consisted of Rituximab (375 mg/m2 on day 1), Gemcitabine (1000 mg/m2 on day 2), Oxaliplatin (130 mg/m2 on day 3) and Ifosfamide (5 g/m2 on day 3), as a 24-hour single infusion. Treatment was given every two weeks with G-CSF support (5 mcg/kg/day or 10 mcg/kg/day at the end of the third course for stem cells mobilization). Responses were evaluated by the integrated FDG-PET/IWC criteria, after the third course and at the end of the entire program. Results: Fifteen patients (median age 36 years, range 24–64 years) with relapsed (n = 12) [post-ABMT (n=4), 〈 12 mo.s (n=6), 〉 12 mo.s (n=2)] or primary progressive (n = 3) Hodgkin Lymphoma, were accrued in this prospective study. Forty percent of them had received 2 or more previous treatment lines. Stage IV at recurrence was found in 9 patients (60%), B symptoms in 7 (46%). Sixty-two total courses were delivered (median 3, range, 3–6). Hematologic toxicity was tolerable, but CTCAE v3.0 G4 thrombocytopenia was found in 15% of courses and infection occurred in 9% of cycles. Ifosfamide was withdrawn in two patients; both aged 64 years, at the fourth course due to the occurrence of tachyarrhythmia and encephalopathy, respectively. Actual dose intensity of the first 3 courses was 84%, 87%, and 91% for Gemcitabine, Ifosfamide, and Oxaliplatin, respectively. All the patients completed at least 3 courses of therapy and were valuable for response. The overall response rate assessed after three courses of R-GIFOX was 87%, with 12 complete responses (CRs) (80%; CR=8, CRu=4) and 1 partial response. Interestingly 3 CRs were achieved among the four patients with post-ABMT relapses. Effective CD34+ cell mobilization was obtained in 6 out of 11 eligible patients and all of them proceeded to subsequent ASCT. The median number of collected CD34+ cells was 4,16 × 106/kg body weight (range, 3.52–10.95) In all instances a single leukapheresis was performed. Among the five ‘bad mobilizers’, two patients had previously undergone radiation therapy and one had received immunoradiotherapy with Ibritumomab Tiuxetan. Failure Free Survival was 46% at 32 mo.s and Disease Free Survival 53% at 30 mo.s. with a median follow-up of complete responders of 11 mo.s. Conclusions: The present study confirms the attractive therapeutic potential of Gemcitabine/Ifosfamide-based combination in the setting of recurrent Hodgkin Lymphoma and further suggests additional contribution from Rituximab and Oxaliplatin. R-GIFOX may represent a less toxic, cytoreductive, and stem cell mobilizing alternative to toxic cisplatin-based salvage regimes, useful for pre-ABMT cytoreduction but also for a full salvage treatment program to be safely delivered to patients unfit for high-dose procedures.

Description: Abstract 3716 Poster Board III-652 Introduction Intrathecal (IT) chemotherapy is an integral component of treatment for Burkitt lymphoma (BL), together with intensive systemic chemotherapy including blood brain barrier crossing agents such as high dose cytarabine (Ara-C) and methotrexate (MTX). However, the optimal IT treatment is yet to be established. Since cytotoxic concentrations of conventional IT agents (Ara-C and MTX), are maintained in the cerebrospinal fluid (CSF) for only a few hours, repeated lumbar punctures are necessary which may turn cumbersome for patients (pts) and pose technical difficulties in some instances. The availability of a sustained-release formulation of Ara-C (liposome-encapsulated Ara-C; Depocyte®) may offer the opportunity of reducing the total number of IT administrations by maintaining/enhancing, the efficacy of CNS prophylaxis. In this regard, a single 50 mg IT injection of Depocyte is able to achieve cytoxic concentrations of free-Ara-C in the CSF for 10-14 days. We report the results of a prospective phase II study aimed at evaluating the safety/activity profile of IT Depocyte in pts with BL and atypical (a)BL. Patients and Methods The study was designed to assess, in untreated pts with BL and aBL, the safety and feasibility of a 50% reduction (from 8 to 4) of the number of IT injections required by substituting IT Ara-C (4 doses) and MTX (4 doses) with 4 administrations (50 mg) of IT Depocyte. IT injections were planned on days (d) 1 or 2 of each rituximab (R)-CODOX-M courses and on d 8 of each R-IVAC course. Pts with aBL received 2 additional R-CODOX courses without any further IT therapy. Primary study endpoints were safety, Depocyte-related extra-hematologic adverse events ≥G3, and CNS failure, i.e. progression/relapse at leptomeningeal and/or parenchymal sites; secondary endpoints included event- and disease-free survival (EFS, DFS). Results A total of 30 HIV negative pts (15 BL and 15 aBL) were enrolled and treated with a dose-modified Magrath regimen (Lacasce, 2004). The median age of pts (M/F: 22/8) was 53 years (r, 25-78), 10 pts (33%) were considered as a low risk (LR) category by displaying ≥ 3 of the following factors: normal LDH, WHO PS 0-1, Ann Arbor stage I-II, and ≤ 1 extra nodal sites. All remaining cases (67%) were considered as high-risk (HR). At diagnosis, 3 pts (10%) had a positive CSF for lymphoma, 6 (20%) had bone marrow involvement and 12 (40%) bulky (〉10 cm) disease. Each pt received a median of 4 (r, 1-6) IT injections of Depocyte at the a median day of 0.5 (r, -1 to 1) for R-CODOX-M1 (course1), 4.0 (r, -2 to 17) for R-IVAC1 (course 2), 1 (r, -1 to 1) for R-CODOX-M2 (course 3) and 6.5 (r, -3 to 14) for R-IVAC2 (course 4). On a total of 111 applications, the following IT injection-related adverse events (NCI-CTCAE v 3.0) of G1-G2 severity were recorded (pts experiencing toxicity): headache 26.6%, nausea 6.6%, vomiting 3.3%, fever 10%, lumbar pain 10%, fatigue 26.6%, somnolence 6.6% and sinus bradycardia 3.3%. A G3 headache episode, accompanied by a transient loss in visual acuity, led to refusal of further IT Depocyte by a single pt. At 51 mo.s, the EFS was 70% with a DFS of 90% at a median observation of 24 mo.s (r, 1-49). Among the 27 complete responders, no isolated leptomeningeal relapses occurred. In particular, none of the 3 responders with CSF involvement at presentation, displayed any form of CNS progression (leptomenigeal and/or parenchymal). In contrast, a parenchymal CNS involvement, with a negative CSF, was observed, at 4.0 mo.s, as a part of the systemic progression in a single pt (3.3%) with chemorefractory aBL. This pt presented with unfavorable features (hi-LDH, stage IVB, PS 2, bulky retroperitoneal adenopathy and 4 extra nodal sites including liver, pancreas, kidney and spleen) and skipped both the first IT and systemic MTX administrations. Conclusions Within the limits of a single arm study, our results show that substitution of 8 IT injections (Ara-C/MTX) with 4 IT administrations of Depocyte within the R-CODOX-M/IVAC regimen is feasible and devoid of severe and/or life-threatening/invalidating neurotoxicity. The CNS progression/recurrence rate was at least super imposable to historical results (CNS relapse rate: 6% to 11%) achieved by the Magrath regimen including double-agent IT treatment. Based on these results, it appears that Depocyte can be safely incorporated into the Magrath regimen to provide adequate single-agent CNS prophylaxis with a reduced burden of IT applications. Disclosures: Off Label Use: Liposome-encapsulated Ara-C for CNS prophylaxis. Vitolo:Mundipharma: Lecture fees. Pinto:Mundipharma: Lecture fees.