ALBERT

Description: Background: MPNs including Polycythemia Vera (PV), Essential Thrombocythemia (ET) and Primary Myelofibrosis (PMF), are clonal hematopoietic diseases in which the discovery of molecular driver mutations (JAK2, CALR, MPL) has deeply modified diagnostic approach in recent years. To date available data on epidemiology of MPNs and perspective analysis are rare. Our aim is to study the incidence of MPN Ph negative in a specific region of Italy named Latium and its variability across five years. Moreover we prospectively report the general features of our population. Method: We present here the prospective epidemiologic analysis of 1116 adult patients affected by MPNs (PV=289, ET=550, PMF=209) diagnosed according to 2008 WHO criteria, from January 2011 to December 2015 in 15 hematological Centers (5 academic and 10 community-based Hospitals) in Latium. A total of 289 PV, 550 ET and 209PMF were identified. The overall incidence rate of 289PV was 1.0/105 in 2011 and 2012, 1.1/105 in 2013, 0.9/105 in 2014 and 2015. The overall incidence rate of 550ET was 2.0/105 in 2011, 2.4/105 in 2012, 2.2/105 in 2013, 1.8/105 in 2014 and 1,2/105 in 2015 and the overall incidence rate of 209PMF was 0.7/105 in 2011 and 2012, 1.0/105 in 2013, 0.7/105 in 2014 and 0.5/105 in 2015. We have observed also 63 cases of MPNu (36M/32F) and the incidence rate was 0.3/105 in 2011 and 2012, 0.14/105 in 2013, 0.24/105 in 2014 and 0.22/105 in 2015. Baseline features of PV, ET and PMF patients are summarized in table 1. We have also analyzed the presence of comorbidities including obesity, arhythmia and neoplasia observed at the diagnosis in 1.6, 6.2 and 4% of all population, respectively; thirty-five percent of 1116 pts presented other comorbidities such as diabetes, inflammatory bowel disease, renal and liver failure. As thrombotic risk factors we considered diabetes, dislipidemia, smoke, essential hypertension and thrombophilia observed in 11,8, 16,2, 13,2, 51,7 and 3% of total pts, respectively. Conclusions: We confirm in our prospective observational protocol the overall incidence of MPN Ph negative, previously reported in the literature and the major incidence of male gender in PV and PMF, female in of ET. The annual incidence from 2011-2015 in Latium is remained substantially the same during the observation period. The decreasing trend observed in 2015 is probably due to the different update of some Centers that was done in October 2015 not including patients diagnosed in the last two months. Disclosures Latagliata: Novartis: Consultancy, Honoraria; Bristol Myers Squibb: Honoraria; Celgene: Honoraria; Janssen: Consultancy, Honoraria; Shire: Honoraria. Breccia:Pfizer: Honoraria; Novartis: Consultancy, Honoraria; Bristol Myers Squibb: Honoraria; Celgene: Honoraria; Ariad: Honoraria. Cimino:Celgene: Honoraria; Bristol-Mayer: Honoraria.

Description: Bendamustine has been demonstrated to be effective for the treatment of CLL, either alone compared with chlorambucil (Knauf et al, JCO 2009 and BJH 2012) or in combination with monoclonal antibodies such as rituximab both in second or more lines (Fischer et al, JCO 2011) and in first line treatment (Fischer et al, JCO 2012). However, the relationship between its activity with clinical and biological prognosticators has been addressed only in few studies. For this purpose, we evaluated the efficacy and safety of bendamustine, in a real-life contest, on 56 patients, median age 66 years (41-80), median number of previous regimens 1 (0-3, 32% previously untreated). Bendamustine was given for a median number of 6 cycles (70-90 mg/m2), in 82% of cases with rituximab at conventional doses. Overall (ORR) and complete response (CRR) rates were 73% and 44.6%, respectively. Obviously, CRR was higher (83.3%) for 18 patients treated in first line. A significant correlation was found between lower ORR and lymphocyte doubling time 30%) of alpha-4 integrin CD49d (OR 13.0; P=0.018), an important marker of bad prognosis in CLL (Bulian et al, JCO 2014). On the other hand, no significant correlations were found between ORR and CD38, ZAP-70 or IGHV mutational status. Similarly, no significant correlations were noted between ORR and FISH cytogenetics, excluding del(17)p, or NOTCH1 mutations, thus confirming the independence of response to bendamustine from some well-known important biologic prognostic factors. In fact, multivariate analysis confirmed a significant relationship only between ORR and TP53/del(17)p (OR 0.020; P=0.0015) and concomitant rituximab (OR 0.019; P=0.0074). The estimated 1-year OS and PFS were 57% and 86%, respectively. Side effects included grade 3-4 neutropenia, infections, thrombocytopenia and anemia which occurred in 21%, 12%, 12% and 5% of patients, respectively. Grade 3-4 non-hematologic toxicity, including infusion-related reactions, heart or kidney or liver failure were found almost exclusively in elderly patients treated with bendamustine after two or more lines of therapy (12.5%). In multivariate analisys of OS, calculated from the end of treatment with bendamustine, only response to bendamustine (P=0.008) was confirmed to be an independent prognostic factor, while both the number of previous therapies and the concomitant use of rituximab demonstrated no statistical significance. These our results confirm both the activity and safety of bendamustine, particularly in combination with rituximab, also in the setting of elderly patients, often affected by two or three comorbidities. Noteworthy, this effectiveness appears to be present also in patients with unfavorable clinical and biological features, excluding del(17)p or TP53 mutations, in which the employment either of modern oral BCR inhibitors or of BH3 mimetics anti-Bcl-2 will be definitely active, also in combination with the same bendamustine. Disclosures No relevant conflicts of interest to declare.

Description: Azacitidine (AZA) is the standard of care for higher-risk myelodysplastic syndromes (MDS). Cytopenias are a major problem during the first treatment cycles due to MDS itself and to the mechanism of action of the drug, which includes not only demethylation, but also cytotoxicity and apoptosis. We evaluated the incidence of infectious complication during AZA treatment in 200 “real life” patients, collected retrospectively by the italian cooperative groups Gruppo Romano MDS (GROM) and Basilicata MDS Registry. Patients gave informed consent and the study was approved by the Ethycal Committes of partecipating Centers. Patient started AZA between 3/2006 and 3/2014. Median age at treatment start was 71 years (range 33-93 yrs, 66 females, 134 males). There were 182 MDS (up to 20% BM-Blasts) and 18 AML (median BM-blasts 23%, range 21-60% BM-blasts). In MDS, IPSS stratification (n=178) was: low: 4%, Int-1: 19%, Int-2: 60%, high 17%, and according to R-IPSS (n=123 pts): low: 4%, intermediate: 20%, high: 51%, very-high: 20%. MDS-specific comorbidity index (MDS-CI) was available for 154 patients: 50 patients were classified as low-risk (0), 85 as intermediate (1-2), and 19 as high risk (〉2). Azacitidine was started at a median of 1.4 months from initial diagnosis. One-hundred-sixty-six patients received AZA at the standard dose of 75 mg/sqm (7 days continuously: 22%, 5-2-2 days: 64%, 6-1-1: 4%) while 34 patients received AZA at 50 mg/sqm for 5- 7 days. Response, according to IWG 2006 criteria, was evaluable in 188 patients: 27 obtained complete remission (CR, 14%), marrow CR was achieved in 6 (3%), partial remission (PR) in 30 (16%), hematological improvement (HI) in 40 (21%), disease was reported stable (SD) in 54 (29%), whereas 31 patients presented progressive disease (16%). A median of 10 (range 1-62) cycles of AZA were delivered. Probability of response was independent from age, IPSS, IPSS-R, MDS-CI, and was associated to a shorter time from diagnosis to therapy initiation (p=0.04) and to the 75 mg/sqm through 7 days vs 50 mg/sqm through 5 -7 days schedule (60 vs 30% CR/PR/HI, p=0.007). With a median follow-up of 14.8 months (range 0-100 months), 57 patients are alive, resulting into a median overall survival of 17.2 months. IPSS and MDS-CI did not predict survival, while grouping according to IPSS-R was associated to a significantly different survival probability (p=0.0007). Across delivery of a total of 1547 AZA cycles, 213 (13%) febrile events were recorded. Fever of unknown origin was diagnosed in 20 patients (10%). Of 193 clinically documented events, a positive microbiologic test was available in 49 cases (25%, 44 for bacteria, 5 for viruses). A single episode of infection was recorded in 124 patients (62%) at a median of 3.8 months from therapy initiation (0-53 months), 53 (26.5%) patients suffered from a second infectious event and 28 (14%) patients from a third one (at 5.6 and 8.3 months from AZA start, respectively). Infections were the attributable cause of death in 30 patients (15%, 3 in CR, 17 with progressive disease and 10 SD). Most frequent infection sites were lungs (45%), cutis (9%), mouth (9%), bowel or peri-anal region (7%). The risk of infectious complications was lower in IPSS low-risk (p=0.05) or IPSS-R low/int MDS (p=0.01), and in patients with MDS-CI 0-2 (p=0.0001). Our data indicate that treatment with AZA is associated with a relatively high probability to develop infectious complications, especially pneumonia, which is however rarely the cause of death. The risk to develop infections is the highest during the first courses of AZA delivery and correlates to baseline patients’ characteristics, including disease stage and comorbidities. Once the infectious episode has occurred, the outcome depends on the disease status. Disclosures Voso: Celgene: Consultancy. Venditti:celgene: Consultancy. Breccia:novartis: Consultancy; BMS: Consultancy; Celgene: Consultancy.

Description: Abstract 3906 Today the treatment target of CLL is the attainment of an optimal disease control combining chemotherapy with monoclonal antibodies (MoAbs). This approach produces more complete molecular remissions and longer response duration (RD), remaining often a minimal residual disease (MRD) detectable by flow cytometry. In addition, consolidation and maintenance therapy with MoAbs might provide a further RD and overall survival (OS) benefit in CLL, as it has been already clearly demonstrated in indolent non-Hodgkin lymphomas. We treated in first line 145 CLL symptomatic patients (pts), median age 63 years (37–80), with six monthly courses of intravenous (25 mg/m2) or oral fludarabine (30–40 mg/m2) and then, after a median time of 30 days, with four weekly doses (375 mg/m2) of rituximab (rtx). Before treatment, 15 pts had a modified low Rai stage, 127 an intermediate stage and only 3 a high stage. We defined as high risk pts having at least two of these markers: unmutated IgVH, CD38〉30%, ZAP-70〉20%, intermediate/poor cytogenetics (trisomy 12 or del11q or del17p). Sixty-three pts (43.5%) belonged to the high risk subset. For MRD flow cytometric study, the threshold was set at 〉1% CD19+CD5+CD79b+/− bone marrow (BM) CLL cells. Based on NCI criteria (Cheson, 1996), 111/145 (76%) pts achieved a complete remission (CR), 27/145 (19%) a partial remission (PR) and 7/145 (5%) no response or progression. Phenotypic CR (CD19+CD5+CD79b- BM cells

Description: B-cell chronic lymphocytic leukemia (B-CLL) follows heterogeneous clinical courses, and several biological parameters need to be added to the current clinical staging systems to predict which patients will experience an indolent or an aggressive outcome. This study analyzed CD38 expression by flow cytometry and soluble APO1/Fas (sAPO1/Fas), Bcl-2 (sBcl-2), and CD23 (sCD23) proteins by immunoenzymatic methods to evaluate their effect on the clinical course of 168 unselected B-CLL patients. Intermediate/high risk modified Rai stages were characterized by a higher CD38+ B-cell number (P = .0002) and higher sCD23 levels (P 

Description: B-cell chronic lymphocytic leukemia (CLL) is a heterogeneous disease with highly variable clinical courses which can be at least in part foreseen by investigating the expression of known prognosticators, including the IGHV gene mutational status or CD38/ZAP-70 expression. By analysing the coordinated expression of several surface antigens in a series of CLL with known clinical courses, we identified the simultaneous over-expression of CD38 and CD49d as part of the signature characterizing the subgroup with the worst outcome. The aim of the present study was to investigate the relationship between CD49d and other well-established biologic prognosticators (CD38 and ZAP-70 expression, IGHV gene mutational status), or markers of tumor burden (Rai clinical stage, beta2-M, sCD23), and to define the independent prognostic impact of CD49d in predicting overall survival (OS) and disease progression (evaluated as time-to-treatment, TTT) in CLL patients. The study includes samples from 303 patients affected by CLL according to the current diagnostic criteria (median age: 63.5 years, range 32–97). The entire cohort of patients was utilized to investigate the impact of CD49d and other prognosticators (CD38, ZAP-70, IGHV gene mutational status) on OS. TTT information and additional laboratory parameters (beta2-M, sCD23) were available for 232/303 patients, whose therapies were established according to NCI-WG criteria. CD49d expression was determined by three-color flow cytometry combining anti-CD49d, anti-CD19 and anti-CD5 mAbs; its expression was demonstrated to be stable over-time and the 30% of positive CD5+CD19+CLL cells was chosen as cut-off to discriminate CD49dlow from CD49dhigh cases. CD49d, whose expression was strongly associated with that of CD38 (p=2.2×10exp-16) and ZAP-70 (p=2.6×10exp-5), or with IGHV gene status (p=1.1×10exp-6), was independent predictor for OS (HR=4.39; p=0.0081) along with IGHV status (HR=5.54; p=0.0005) or, if this parameter was omitted, with ZAP-70 (HR=2.90; p=0.0092). CD49d also effectively predicted TTT and refined the prognostic relevance of all the investigated prognosticators. Notably, a CD49dhigh phenotype, while not changing the outcome of good prognosis (ZAP-70low, mutated-IGHV) CLL, was necessary to correctly predict the bad clinical courses of ZAP-70high (HR=3.12; p=0.023) or unmutated-IGHV (HR=2.95; p=0.002) cases. These findings support the introduction of CD49d detection in routine prognostic assessment of CLL patients, and suggest both pathogenetic and therapeutic implications for CD49d expression in CLL, e.g. envisioning the use of a humanized anti-CD49d monoclonal antibody, currently employed in multiple sclerosis (Natalizumab), for selcted CLL patients.

Description: Idiopathic hypereosinophilic syndrome (HES) is a rare hematological disorder characterized by persistent eosinophilia presenting often with organ involvement. The increased blood and tissue levels of eosinophils (eos) appear to produce tissue damage via local release of toxic substances producing inflammation and fibrosis. Data from the literature are fragmentary and deduced from several historical series. The most commonly involved organs reported are cardiovascular, pulmonary, cutaneous and neurologic systems. Cardiac disease is described as the major cause of morbidity and mortality. Due to the new knowledge about myeloproliferative variant of HES and PDGFRA rearrangement, we tried to identify different clinical picture. The aim of this retrospective analysis is to evaluate the prevalence of HES-related organ damage, its relation to F/P status, and response to imatinib therapy. A prospective multicenter study of the HES began in 2001. Patient enrolled had HES diagnosis, independently of presence of FIP1L1-PDGFRA (F/P) rearrangement. 72 patients were treated with IM 100 to 400 mg daily with a median observation time of 28 months (r 12–68). 33 patients (46%) were positive for the F/P rearrangement (F/P+), while 39 (54%) were negative (F/P−). At the time of enrolment HES patients were systematically screened for organ damage with instrumental evaluation (chest radiography, echocardiogram, abdomen ultrasonography). Symptoms were considered too, so if respiratory symptoms were presented or a reduction of ejection fraction was found organ involvement was established. Results. Organ involvement was recorded in 42% of F/P+ and in 51% of F/P−. Skin involvement was only recorded in F/P−, whereas splenomegaly was reported in 7 F/P+ and only in one case of F/P−. To date, soft tissue was peculiar site of F/P+ patients. All the 33 F/P+ patients achieved a complete hematologic remission (CHR) and molecular negativity, and together they became negative for organ localization and free of symptoms. In the two patients with soft tissue involvement (temporal/parietal and retro-orbital masses) imaging response was documented with positron emission tomography (PET) and with CT scan one month after the beginning of the therapy. Favourable response was recorded also in patients with cardiac involvement, apiece to grade of fibrosis. 5 out of 39 F/P− patients achieved a transitory CHR, but no durable effects on organ involvement. All these patients are alive. Conclusion. Organ involvement do not seems to be a constant characteristic of HES, irrespective to F/P status, but there are differences between F/P+ and F/P− patients. Organ damage in F/P+ subset is reversible before fibrosis development. In the whole population observed, no deaths were recorded in more than five years. F/P+(33 pt) F/P−(39 pt) Statistic Organ damage 42% 51% p=NS More than 1 organ involvement 18% 3% p=0,01 Heart 15% 5% p=NS Lung 18% 26% p=NS Spleen 21% 3% 0,001 Skin 0 15% p=0,001 Liver 3% 3% p=NS Soft tissue 6% 0 p=0,1 Skeleton 3% 0 p=NS Gut 0 3% p=NS

Description: Figure Figure The immunoglobulin (Ig) beta protein (CD79b) dimerizes with Ig alpha (CD79a) to form the signaling component of the B-cell antigen receptor complex (BCR). Given the critical role of Ig beta in BCR signaling, it has been suggested that its variable expression in CLL may induce either cell proliferation or apoptosis. High CD79b expression has been associated with atypical morphology, strong expression of surface Ig, advanced clinical stage and short overall survival in B-CLL (Garcia Vela, 1999; Zucchetto, 2006). Moreover, high CD38 expression was correlated with high CD40, CD69 and CD79b which are consistent with an activation phenotype (Damle, 2002). Furthermore, we have demonstrated that the ZAP-70+ CLL subgroup shows a rapid disease progression and an inferior overall survival (Del Principe, 2006). Since it has been described that BCR engagement has significant effects on B-CLL cell survival, activation and cell cycle progression (Deglesne, 2006), we decided to evaluate the real impact of CD79b expression on B-CLL prognosis. The primary aims of our research were: 1) to determine progression-free survival (PFS) and overall survival (OS) upon CD79b expression; 2) whether CD79b could predict varied outcome within ZAP-70+ and ZAP-70 negative subgroups; and finally 3) whether CD79b was an independent prognostic factor. Therefore, we investigated 401 patients (pts), median age 65 years (range 33–89), 213 males and 188 females. With regard to modified Rai stages, 123 pts had a low stage, 261 an intermediate stage and 17 a high stage. CD79b was determined by multicolor flow cytometry, using the monoclonal antibody anti-CD79 beta-FITC (clone SN8, Dako) and fixing a cut-off value of 30%. CD79b+ B-CLL pts were 207/401 (52%). CD79b〉30% was significantly associated with the intermediate/high Rai stage (p=0.00001), beta-2 microglobulin 〉2.2 mg/dl (p

Description: Clinical trials of monoclonal antibodies in combination with chemotherapy have reported previously unattained response rates in CLL because this approach reduces disease burden to levels detectable only by flow cytometry or molecular methods. The latest studies from CALGB have demonstrated that rituximab added concurrently or sequentially to fludarabine (Flu) for symptomatic, untreated CLL allows to achieve higher remission rates, longer progression free and overall survival. Nevertheless, the analysis of different biologic parameters could better explain the discordant outcome independent of treatment observed in these studies. Recent literature data indicate that unmutated VH genes, CD38 and/or ZAP-70 protein tyrosine kinase overexpression may predict both a lower response and a shorter survival. We performed at our Institution a phase II study that added rituximab sequentially to Flu as initial therapy for symptomatic, untreated CLL in order to evaluate either the toxicity or the clinical response or outcome. Complete remission (CR) was also assessed by a multiparametric flow cytometric method. ZAP-70 protein and CD38 antigen were determined before chemotherapy on mononuclear cells by flow cytometry using an anti-ZAP-70 and an anti-CD38 antibody, respectively. Forty-nine B-CLL patients, median age 59 years (range 37–74) received six monthly courses of Flu (25 mg/m2 for 5 days) and four weekly doses of rituximab (375 mg/m2) starting on an average of thirty days (range 5–180) after completion of the Flu therapy. According to modified Rai stages, 4 pts had a low stage, 42 an intermediate stage and 3 a high stage. Three out of 49 pts experienced fever, chills and rigors, during the first infusion of rituximab and only 1 patient presented grade 3 infective lung toxicity according to WHO. Hematologic toxicity included neutropenia (grade 1 and/or 2 in 12 pts, grade 3 and/or 4 in 22 pts), thrombocytopenia (grade 1 and/or 2 in 4, grade 3 and/or 4 in 3 pts) and anemia (grade 2 in 3 pts). Based on the NCI criteria, 45/49 (91.8%) pts achieved a CR, 3/49 (6.1%) a partial remission (PR) and 1/49 (2%) no response (NR). The median follow-up duration was 29 months. Median duration of CR and PR has not been reached. Noteworthy, our B-CLL pts treated sequentially with Flu and rituximab experienced a very long progression-free survival (PFS) from treatment (72% at 3 years). ZAP-70 and CD38 were positive (〉20%) in 19/48 (39.5%) and in 13/48 (27%) pts, respectively. Minimal residual disease (MRD) performed on bone marrow by flow cytometry was positive (〉5% CD19+CD5+CD79b- CLL cells) in 7/37 (19%) analysed pts. A significant shorter PFS was observed in ZAP-70+ pts (38% vs 100% at 3 years; P=0.003), in CD38+ pts (39% vs 92% at 3 years; P=0.007) and also in pts with higher MRD after treatment (50% vs 81% at 2 years; P=0.04). Therefore, the addition of monoclonal antibodies, such as rituximab, to chemotherapy allowed a better outcome in CLL, exerting a key role to eradicate MRD. Moreover, the stratification of pts within different risk classes using novel biologic predictive factors, such as ZAP-70 and CD38, might allow us to offer more tailored treatment strategies, reserving experimental approaches and/or transplantation procedures only to CLL subsets with proved adverse biologic and clinical features.