ALBERT

Description: Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous category of lymphoid tumors that comprises different clinical forms not fully recognized in the WHO classification. In this regard, extranodal (EN) DLBCLs have particular clinicobiological features and outcome, sometimes related to the specific site where the lymphoma arises. Nowadays, rituximab plus chemotherapy (CT) is the gold-standard in the treatment of DLBCL. However, the superiority of rituximab-CT (R-CT) over CT alone has not been addressed for all the clinical subsets of the disease and, in fact, the clinical role of the new therapies might be different for primary nodal or EN DLBCLs. The aim of this study was to assess the impact of rituximab in patients suffering from nodal or EN DLBCL. Two-hundred and thirty non-immunocompromised patients (112M/118F; median age, 61 years) diagnosed with CD20+DLBCL in a single institution between 1997 and 2006 (five years before and after establishing R-CT as the standard treatment in DLBCL) and treated with adriamycin-containing regimens were the subject of the present study. The series included 148 primary nodal and 82 EN DLBCL. Patients with primary CNS lymphoma were excluded and lymphomas arising at Waldeyer ring were considered as nodal DLBCL. The main EN sites were GI (n=26), bone (n=13), soft tissue (n=13), lung/pleura (n=9), liver (n=9), and other (n=12). Main clinico-biological and evolutive variables were analyzed. One hundred nineteen patients received only CT and 111 R-CT. Eighty-seven cases with available information were assigned to germinal center B-cell-like (GCB) (41%) or non-GCB (59%) groups according to the Hans method (Blood2004;103:275–82) based on CD10, BCL6 and MUM1 expression. Main initial features, including the primary nodal or EN origin, international prognostic index (IPI), and GCB/non-GCB categories were similar for CT and R-CT groups. No correlation was observed between the GCB/non-GCB groups and the primary site of the tumor, although nodal lymphomas more frequently expressed MUM1 than EN (69% vs. 31%, respectively; p=0.01). CR rate and 5-year overall survival (OS) according to the treatment arm (CT vs. R-CT) is detailed for the whole series and for the nodal and EN groups in the table and OS curves depicted in the figure. In the whole series, variables predicting poor OS in the multivariate analysis were high-risk IPI (RR 2.5; p

Description: Diffuse large B-cell lymphomas (DLBCLs) can be divided into germinal-center B cell–like (GCB) and activated-B cell–like (ABC) subtypes by gene-expression profiling (GEP), with the latter showing a poorer outcome. Although this classification can be mimicked by different immunostaining algorithms, their reliability is the object of controversy. We constructed tissue microarrays with samples of 157 DLBCL patients homogeneously treated with immunochemotherapy to apply the following algorithms: Colomo (MUM1/IRF4, CD10, and BCL6 antigens), Hans (CD10, BCL6, and MUM1/IRF4), Muris (CD10 and MUM1/IRF4 plus BCL2), Choi (GCET1, MUM1/IRF4, CD10, FOXP1, and BCL6), and Tally (CD10, GCET1, MUM1/IRF4, FOXP1, and LMO2). GEP information was available in 62 cases. The proportion of misclassified cases by immunohistochemistry compared with GEP was higher when defining the GCB subset: 41%, 48%, 30%, 60%, and 40% for Colomo, Hans, Muris, Choi, and Tally, respectively. Whereas the GEP groups showed significantly different 5-year progression-free survival (76% vs 31% for GCB and activated DLBCL) and overall survival (80% vs 45%), none of the immunostaining algorithms was able to retain the prognostic impact of the groups (GCB vs non-GCB). In conclusion, stratification based on immunostaining algorithms should be used with caution in guiding therapy, even in clinical trials.

Description: Abstract 1948 Poster Board I-971 Survival after treatment of diffuse large B-cell lymphoma (DLBCL) is influenced by differences in the tumor microenvironment. Gene expression profiling (GEP) studies have shown that the angiogenesis-related signature (stromal-2 signature) is prognostically unfavorable. However, the clinical and biological significance of angiogenesis quantified in tumor tissue sections of DLBCL from patients treated with rituximab plus chemotherapy (R-CT) is not yet fully explored. CD31, the platelet adhesion molecule PECAM1, is one of the genes included in the “stromal-2 signature” reported in the GEP studies. The objective of this study was to determine whether the microvessel density (MVD) and microvessel number (MVN) in DLBCL patients treated with R-CT were associated with the clinicopathological features of the tumors and related to the outcome of the patients. The MVD and MVN were assessed in a series of 160 patients with DLBCL from the Leukemia Lymphoma Molecular Profiling Project consortium (LLMPP) 86M /74F; median age 64 yrs. The GEP was investigated in 116 of these including 50 germinal center B (GCB), 55 activated B-cell (ABC) and 11 unclassifiable cases. An independent series of 129 patients from the Catalan Lymphoma-Study Group (GELCAB) (67M/62F; median age 64 yrs) was used to validate the results. Front-line treatment was R-CT in all cases of both series. Tissue Microarrays (TMA) were constructed from pretreatment biopsy specimens of de novo DLBCL. High grade B cell lymphoma otherwise unclassifiable, primary mediastinal B cell lymphoma, T-cell-rich B cell lymphoma, and tumors associated with immunodeficiency were excluded. All cases were stained in an automated immunostainer with an antibody against CD31 (DAKO). The MVD and MVN were quantified using digitalized images of the tumor using Olympus Cell B Basic Imaging Software. Microvessel areas were defined as vascular areas delineated by CD31+ staining. The MVD was calculated as the sum of all microvessel areas divided by the total area analyzed. The MVN was the sum of all identified individual vessels, divided by the total area analyzed. TMAs were independently scored by two observers and discrepancies were resolved over a double-headed microscope. To determine whether the angiogenic values scored using the TMA were representative of the tumor sample, whole tissue sections and TMA cores from the same tumor were evaluated in 40 cases and compared by a linear regression analysis. MVD and MVN were grouped in quartiles when necessary and considered high or low when above or below the 50th percentile, respectively. Linear correlation analysis between the CD31 (+) MVD results on TMA cores and on the corresponding whole tissue sections in 40 cases showed a good correlation (R2=0.81). In the LLMPP cohort, DLBCL with an ABC profile showed higher MVD than those with GCB profile (p=0.05). In addition, higher MVD was observed in patients with advanced stage (p

Description: Follicular lymphoma (FL) is typically a nodal disease. Primary extranodal FLs, that represent less than 10% of the cases, might have differentiated clinicobiological features. The aim of the present study was to analyze the main clinicobiological characteristics, response to therapy and outcome of a series of patients with FL primarily extranodal in origin, and compare them with nodal FLs. Twenty two patients (12M/10F; median age, 59 years) with FL with primary origin in extranodal location diagnosed in a single institution during a 24-year period, with primary origin in extranodal location were the subject of the study. The control group was constituted by 212 patients with nodal FL diagnosed during the same period of time. Main clinicobiological features were recorded and analyzed. The sites of the primary disease were: skin, 5 cases; Waldeyer’s ring, 4; GI tract, 3; bone marrow, CNS and parotid (two cases each); and pancreas, thyroid, kidney and orbit (one case each). Main histological and clinical features are listed in the table. Treatment was given without considering the nodal or extranodal origin of the disease and consisted of: monotherapy with alkylating agents (38 cases), polychemotherapy (149), and fludarabine alone or with other drugs (14) and others, including surgery and observation (33). CR rate was higher in extranodal than in nodal FL (82% vs. 53%, respectively; p=0.02), but no differences were found in overall survival. FLIPI score was the most significant variable predicting overall survival in the global series as well as in either in nodal or extranodal FL. In conclusion, extranodal FL had some peculiar clinicobiological features with respect to nodal cases. Regarding the outcome, although patients with extranodal FL showed a higher CR rate, the overall survival was similar in both groups. Extranodal FL (N=22) Nodal FL (N=212) p Age (median, range) 59 (28–82) years 55 (24–93) years NS Sex (M/F) 12/10 100/112 NS Histological grade 3 (%) 12 10 NS CD10+(%) 85 90 NS Bcl2+ (%) 67 91 NS Bcl2/JH (%) 37 65 0.03 Stage IV (%) 50 64 NS Bone marrow + (%) 36 62 0.02 LDH 〉 450 IU/L (%) 15 24 NS B2-microglobulin 〉 2.3 mg/L (%) 7 41 0.058 High-risk FLIPI (%) 19 35 NS CR rate (%) 82 53 0.02 5-year OS (%) 75 74 NS 5-year FFS (%) 67 27

Description: BACKGROUND The UPR is a prosurvival pathway activated in cells under ER stress induced by the accumulation of unfolded proteins. UPR activation in B cells normally occurs during the differentiation to antibody secreting plasma cells and requires XBP1activation. XBP-1 is a member of the TREB family of transcription factors that exists in the endoplasmic reticulum (ER) as a 33kDa protein, and in the nucleus as an active 50kDa transcription factor. The UPR stimulates two different ER proteins, ATF-6 and Ire-1, to increase XBP-1 transcription and XBP-1 mRNA splicing resulting in the accumulation of the active 50kDa nuclear protein. Moreover XBP1 is a target of proteosome inhibitors and is related to the aggressive behaviour of some carcinomas. The role of the activation of XBP-1 in lymphomas is still unknown. DESIGN: Reactive lymphoid tissues and 25 neoplastic human B-cell lines representing different stages of B-cell development were studied for XBP-1 expression by western blot and XBP-1, PAX-5, Blimp-1/prdm1, MUM-1/IRF-4 and ICSBP1/IRF-8 by immunohistochemistry. XBP-1 activation was assessed in 225 B-cell lymphomas from the archives of the laboratory of pathology by western blot, RT-PCR and immunohistochemistry . To further evaluate whether XBP-1 activation was related to the plasmacytic program or to ER stress signals we analyzed the cell lines by Western blot for XBP-1 and ATF-6 expression. RESULTS We characterize XBP-1 expression in reactive lymphoid tissues, 25 human cell lines and 225 B-cell tumors. In nearly all tonsillar lymphoid cells XBP-1 was detected as a cytoplasmic protein with a paranuclear dot pattern. Nuclear positivity was observed only in scattered centrocytes in the light zone of the germinal centers and in plasma cells, always coexpressed with plasma cell related transcription factors as MUM-1/IRF-4 and Blimp1/prdm1. Active p50XBP-1 was found in 24/25 cell lines by western blot regardless ATF-6 expression and confirmed by immunohistochemistry . Moreover p50XBP1 was found in 27/31(87%) plasmacytomas, 36/64(56%) DLBCL-ABC and in 3/10(30%) DLBCL-GCB and 22/43(51%) plasmablastic lymphomas. Intriguingly, p50XBP1 was detected also in 2/11(18%)BL and 4/25(16%)MCL with blastic features. CONCLUSIONS.XBP-1 is activated in a subset of follicular centre cells committed to plasma cell differentiation and in plasma cells.UPR prosurvival pathways in the neoplastic cell lines are activated independently of the extent of the ATF-6 activation.p50XBP1 is mostly activated in aggressive B-cell lymphomas regardless to the plasmacytic differentiation of the tumours. Thus, p50XBP-1 may be a new molecular target in the treatment of aggressive B-cell malignancies.

Description: Introduction: MYC rearrangements (MYCr) occur in 5 to 15% of diffuse large B-cell lymphomas (DLBCL) and 20 to 35% of high-grade B-cell lymphoma, NOS (HGBL-NOS), are a defining criterion of the category HGBL with rearrangements of MYC and/or BCL2/BCL6 (HGBL, with MYCr and BCL2/BCL6), and may be present in 90% of Burkitt lymphoma. The current WHO classification considers cytogenetic techniques as the appropriate tool to detect MYCr but does not define how to approach to the identification of such alteration. As the global incidence of MYCr in large B-cell lymphomas (LBCL) is low, it is necessary to clarity whether FISH or other cytogenetic methods have to be applied to all LBCL or only in selected cases. We previously identified LMO2 as a potential surrogate marker of MYCr in LBCL (Colomo L, Am J Surg Pathol 2017). Our aim with this study is to confirm this observation and evaluate the clinical impact of this marker in the survival of patients with LBCL. Methods: We have prospectively studied between September 2014 and July 2019 a new series of 180 LBCL including patients with DLBCL, HGBL, with HGBL, with MYCr and BCL2/BCL6, HGBL-NOS and transformed low-grade lymphomas into DLBCL (tDLBCL) diagnosed according to WHO criteria. LMO2 (clone 1A9-1), MYC (clone Y69) and a common immunohistochemistry (IHC) panel of B and T-cell markers have been used for the histological categorization of the cases, using whole tissue sections. The cutoff for LMO2 and MYC were 30% and 40%, respectively. MYC and BCL6 genes were studied using break apart probes, and BCL2 gene using dual-color dual-fusion probes (IGH/BCL2), all from Vysis-Abbott. We have statistically correlated the loss of expression of LMO2 and the overexpression of MYC with the presence or absence of MYCr. Moreover, we performed survival analyses assessing the clinical impact of LMO2 in a series of 162 LBCL patients (112 DLBCL, 20 HGBL, with MYCr and BCL2/BCL6, 4 HGBL-NOS and 26 tDLBCL). The survival series included cases diagnosed before 2014 with IHC and FISH data. Results: The prospective series included 132 patients with DLBCL (78M/52F; median age 67 years, range 35-95), 9 HGBL, with MYCr and BCL2/BCL6 (5M/4F; median age 67 years, range 42-85), 4 HGBL-NOS (2M/2F; median age 58 years, range 42-89), and 35 tDLBCL (31 transformed follicular lymphomas, 3 marginal zone lymphoma and 1 lymphoplasmacytic lymphoma; 23M/20F; median age 64 years, range 40-82). LMO2 and MYC were expressed as follows, respectively: 84/130 (65%) and 46/132 (35%) in DLBCL; 1/9 (11%) and 8/9 (89%) in HGBL, with MYCr and BCL2/BCL6; 0/4 and 3/4 (75%) HGBL-NOS; 25/34 (73%) and 7/33 (21%) tDLBCL. MYCr were identified in 9/132 (7%) DLBCL; all HGBL, with MYCr and BCL2/BCL6; 4/4 HGBL-NOS; 7/35 (20%) tDLBCL. The table shows the comparisons between LMO2 and MYC protein expression for the identification of the presence of MYCr in the series of LBCL. Whereas in the whole series LMO2 and MYC had similar results, among CD10-positive cases, LMO2 had better results than MYC and identified better the presence of MYCr than MYC protein expression. The 5-year progression-free survival (PFS) according the diagnostic categories was 59% for DLBCL, 28% for HGBL, with MYCr and BCL2/BCL6, 25% for HGBL-NOS and 22% for tDLBCL (P=0.015). In addition, PFS was significantly lower for the presence of MYCr (26% vs 53%, P=0.02) and MYC IHC expression (35% vs 53%, P=0.005), and showed a positive trend for LMO2 loss of expression (39% vs 52%, P=0.1). The 5-year overall survival (OS) according the diagnostic categories was 67% for DLBCL, 23% for HGBL, with MYCr and BCL2/BCL6, 50% for HGBL-NOS and 77% for tDLBCL (P

Description: Biologic factors that predict the survival of patients with a diffuse large B-cell lymphoma, such as cell of origin and stromal signatures, have been discovered by gene expression profiling. We attempted to simulate these gene expression profiling findings and create a new biologic prognostic model based on immunohistochemistry. We studied 199 patients (125 in the training set, 74 in the validation set) with de novo diffuse large B-cell lymphoma treated with rituximab and CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) or CHOP-like therapies, and immunohistochemical stains were performed on paraffin-embedded tissue microarrays. In the model, 1 point was awarded for each adverse prognostic factor: nongerminal center B cell–like subtype, SPARC (secreted protein, acidic, and rich in cysteine) 〈 5%, and microvascular density quartile 4. The model using these 3 biologic markers was highly predictive of overall survival and event-free survival in multivariate analysis after adjusting for the International Prognostic Index in both the training and validation sets. This new model delineates 2 groups of patients, 1 with a low biologic score (0-1) and good survival and the other with a high score (2-3) and poor survival. This new biologic prognostic model could be used with the International Prognostic Index to stratify patients for novel or risk-adapted therapies.

Description: Abstract 4134 Gene expression profile (GEP) allows to distinguish two groups with different origin in patients with diffuse large B-cell lymphoma (DLBCL): germinal-center (GC) and activated (ABC), with the latter having a significantly poorer outcome. However, GEP is a technique not available in current clinical practice. For this reason, attempts to reproduce GEP data by immunophenotyping algorithms have been made. The aim of this study was to apply the most popular algorithms in a series of patients with DLBCL homogeneously treated with immunochemotherapy, in order to assess the correlation with GEP data and their usefulness to predict response and outcome of the patients. One hundred fifty seven patients (80M/77F; median age 65 years) diagnosed with DLBCL in 5 institutions of the Grup per l'Estudi dels Limfomes de Catalunya I Balears (GELCAB) during a 5-year period, treated with Rituximab-containing regimens (in most cases, R-CHOP), in whom histological material to construct a tissue microarrays (TMA) was available, constituted the subjects of the present study. Four algorithms were applied: Colomo (Blood 2003, 101:78) using CD10, bcl-6 and MUM1/IRF4; Hans (Blood 2004, 103:275) using CD10, bcl-6 and MUM1/IRF4; Muris (J Pathol 2006, 208:714) using CD10 and MUM1/IRF4, and Choi (Clin Cancer Res 2009, 15:5494), using CD10, bcl-6, GCET1, FOXP1 and MUM1/IRF4. The thresholds used were those previously described. GEP studies were performed in 62 patients in whom fresh frozen material was available. Main clinical and evolutive data were recorded and analyzed. The proportion of positive cases for the different single antigens was as follows: CD10 26%, bcl-6 64%, GCET1 46%, FOXP1 78% and MUM1/IRF4 28%. The distribution of cases (GC vs. non-GC) according to the algorithms is detailed in the table. In 88 of 110 patients (80%) with all the antigens available, the patients were allocated in the same group (either GC or non-GC). When the immunochemistry was compared with GEP data, the sensitivity in the GC group was 59%, 52%, 70% and 40% for Colomo, Hans, Muris and Choi algorithms, respectively. The sensitivity in the non-GC group was 81%, 85%, 62% and 84%, respectively. On the other hand, the positive predictive value (PPV) in the GC group was 81%, 83%, 72% and 77%, respectively. In non-GC subset the PPV for the different algorithms was 59%, 55%, 72% and 52%, respectively. We observed a higher percentage of misclassified cases in the GC-phenotype subset than in the non-GC subgroup. None of the immunohistochemical algorithms showed a significant superiority as surrogate of GEP information among the others. The ability of GEP groups as well as of groups defined by the algorithms to predict complete response (CR) rate, progression-free survival (PFS) and overall survival (OS) of the patients is showed in the table. Thus, whereas the GEP groups showed significant prognostic value for CR rate, PFS and OS, none of the immunohistochemical algorithms were able to predict the outcome. In conclusion, in a homogeneous series of DLBCL patients treated with immunochemotherapy, the different immunohistochemical algorithms were not able to mimic the GEP information. The prognostic impact of the groups defined by immunohistochemistry (GC vs. non-GC) was particularly low. N (%) CR rate N (%) 5-year PFS (%) 5-year OS (%) Colomo algorithm GC 53 (44) 39 (74) 48 54 Non-GC 68 (56) 53 (78) 55 62 Hans algorithm GC 61 (41) 47 (77) 54 60 Non-GC 88 (59) 67 (76) 52 59 Muris algorithm GC 87 (57) 63 (72) 48 57 Non-GC 65 (43) 51 (78) 56 63 Choi algorithm GC 45 (33) 32 (71) 48 54 Non-GC 90 (67) 70 (78) 52 61 Gene expression profile 30 (58) 25 (83) 76* 80** GC Activated 22 (42) 17 (77) 31* 45** * p=0.005, ** p=0.03. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 1945 Poster Board I-968 T-cell Subpopulations Quantified by Flow Cytometry in Lymph Node Cell Suspensions Identify a Group of Patients with Follicular Lymphoma with Good Prognosis Neus Villamor, Gonzalo Gutiérrez, Joaquim Carreras, Eva Giné, Gabriela Ghita, Marta Aymerich, Montserrat Torrebadell, Antonio Martínez, Lluís Colomo, Emili Montserrat, Elías Campo, Armando López-Guillermo Hematopathology Unit and Department of Hematology, Hospital Clínic, IDIBAPS, Barcelona, Spain. Introduction: Tumor microenvironment plays an important role in the behavior of follicular lymphoma (FL), as demonstrated by gene expression profile analysis. Using this technique, an increase in macrophages has been associated with poor outcome, whilst an increase in T-cells is associated with better prognosis. Immunohistochemical analysis has been performed as alternative method to gene expression profile, but the quantification is time consuming and poorly standardized. Patients and methods: T-cell populations from lymph nodes of 68 patients (35M/33F, median age 59, range 29 to 81) with FL at diagnosis were quantified by flow cytometry in cell suspensions. The percentage of CD3, CD4, CD8, CD57, and germinal center (GC) CD4 cells (CD4+CD57+), as well as the ratio B/T, CD4/CD8, CD4/CD3, CD8/CD3 and GC-CD4/CD4 were correlated with the main initial features and the clinical outcome of the patients. The distribution according to the histological grade was: grade 1 and 2, 53 patients; grade 3a, 14; grade 3b, 1. Fifty-one percent of patients had low-risk FLIPI. 62 patients have received polychemotherapy, including rituximab in 33, whereas in 6 a watchful waiting policy was established. After a median follow-up of 4 years, 17 patients have died, with a 5-year overall survival (OS) of 77%. Results: The mean (±SD) percentage of B-cells, CD3, CD4, CD8 and GC-CD4 was 58.6% (±14.2), 36.1% (±15.2), 27.1% (±12.3), 8.7% (±5.1), and 3.4% (±2.4). No correlation was found between percentages of T cell subpopulations, B/T, CD4/CD3 and CD8/CD3 cell ratios and the clinical characteristics, failure-free survival (FFS) and OS. The median value CD4/CD8 and GC-CD4/CD4 ratios was 3.4 (range: 0.51 to 1) and 0.11 (n=56) (range: 0.02 to 0.34), respectively. Grade 3 histology was more frequently observed in patients with CD4/CD8 ratio

Description: Abstract 3677 Lymphoplasmacytic lymphoma (LPL) is a very heterogeneous disease from the clinical stand point, including the fact that Waldenström macroglobulinemia (WM) can be recognized in a significant proportion of LPL cases. Whether LPL with and without criteria for WM differ in the clinical features and outcome is not well known and is the aim of the present study. For this purpose, 50 patients (median age, 67 years (range, 19 to 91; 50% males) with a tissue biopsy diagnostic of LPL or a bone marrow infiltration by LPL were included in the present study. Main clinic-biological characteristics and outcome were recorded and analysed. Bone marrow infiltration and presence of a serum paraprotein were observed in 42 cases (89%) and 37 cases (86%), respectively. WM according to WHO criteria was identified in 26 patients (60%). Thirty-four patients received treatment for LPL (69%), including rituximab at any time in 74% of treated patients. The main clinical features of the series according to the WM/LPL or non-WM/LPL criteria are listed in the table. No relevant differences were identified when comparing WM cases with remaining LPL cases, except for those determined by their definition. Six patients eventually developed solid neoplasms with no differences between both groups. After a median follow-up of 35 months (range 0,3 to 209), 17 patients have died, including 10 patients by disease progression, 2 by secondary malignancies, 2 by heart failure, and 3 by unknown causes. The median survival of the whole series was 133 months (CI 95%: 40–226). Among the ten patients who died as a result of disease progression, two different patterns were observed. In six cases disease progression was characterized by general symptoms and lymph node growth, whereas the other 4 cases showed severe cytopenias unresponsive to treatment. These patterns were both observed indistinctly in WM and the non-WM/LPL cases. OS was similar in both groups of patients (median OS 133 vs. 216 months in WM and non-WM, respectively; p=NS). In conclusion, no major differences were observed in terms of initial features and outcome among LPL patients according to the definition of WM or non-WM/LPL.FeaturesAll patients (n=50)WM/LPL patients (n=26)Non-WM /LPL patients (n= 24)Age (median, yrs)676668Gender (male, %)25 (50%)12 (46%)13 (54%)Stage IV (%)46 (94%)26 (100%)20 (87%)Bone marrow infiltration44 (90%)26 (100%)*18 (78%)*Serum paraprotein37 (86%)26 (100%)*11 (65%)*IgM type30 (81%)26 (100%)*4 (36%)*M size (median, g/L)181815.3ECOG ≥ 2 (%)8 (18%)2 (9%)6 (29%)Enlarged lymph nodes17 (37%)8 (33%)9 (41%)Splenomegaly12 (26%)5 (20%)7 (33%)Bulky disease7 (15%)3 (12%)4 (18%)Hb (