ALBERT

Description: Abstract Background: CNS relapse is uncommon but challenging complication in patients with mantle cell lymphoma (MCL). Survival after CNS relapse is extremely poor that identification of high risk population is therefore critical in whom prophylactic chemotherapy may play a role. Patients and Methods: A total of 405 patients (median age, 66 years; range 22-85) with MCL newly diagnosed between 1994 and 2012 at 48 institutions in Japan were analyzed.Pathological specimens were centrally reviewed in all patients. We evaluated risk factors for CNS relapse in patients with MCL using competing risk regression analysis. Patients were excluded if they had evidences of CNS involvement at initial diagnosis. Overall, 81% of the patients received rituximab containing regimen with 222 patients received CHOP, 133 patients received high-dose cytarabine containing regimen and 50 patients received other regimen such as fludarabine. At a median follow up duration of 40.4 months, 25 patients (6.2%) experienced CNS relapse. Five patients presented with brain parenchymal mass, 13 patients presented with leptomeningeal disease, and 7 patients presented with both. The median time from initial diagnosis to CNS relapse was 18.5 months. Two-year cumulative incidence of CNS relapse was 4.7% (95%CI: 2.8-7.3). No difference was seen in the incidence of CNS relapse between the treatment with or without rituximab. Univariate analysis revealed blastoid or pleomoriphic variant, Ki-67 ≥ 30, serum total protein 〈 6.0g/dl, leukemic presentation and high risk group in simplified-MIPI at diagnosis as significant risk factors for CNS relapse. Within these five factors, Ki-67 was the strongest predictive factor for CNS relapse with hazard ration of 8.7 (95%CI: 3.4-22.2). Two-year cumulative incidence of CNS relapse in patients with Ki-67 ≥ 30 was 25.4% (95%CI: 13.5-39.1), while that in patients with Ki-67 〈 30 was 1.5% (95%CI: 0.2-2.9) (Figure). Ki-67 ≥ 30 was a significant risk factor for CNS relapse both in patients who received CHOP and hyper-CVAD/MA. Survival after CNS relapse was dismal with median survival time of 5.7 months. Summary: We identified four risk factors for CNS relapse in patients with MCL. Within these factors, Ki-67 ≥ 30 is the strongest predictive factor with 2-year cumulative incidence of 25.4%. Evaluation for CNS involvement at the diagnosis and assessment for value of prophylactic IT chemotherapy is mandatory in patients with Ki-67 ≥ 30. Figure 1: Cumulative incidence of CNS relapses by Ki-67 Figure 1:. Cumulative incidence of CNS relapses by Ki-67 Disclosures No relevant conflicts of interest to declare.

Description: Abstract 2682 Background: CD20 is a critical molecular target of a chimeric monoclonal antibody, rituximab, and expression of this protein can be detected in most of B-cell lymphoma cells by immunohistochemistry (IHC) and/or flow cytometry (FCM). Rituximab administration is considered when CD20 protein expression can be confirmed by IHC and/or FCM; however some patients show a CD20 IHC(−)/FCM(−) phenotype before and after rituximab treatment, and this phenomenon is thought to be strongly correlated with rituximab resistance (Hiraga J, Tomita A, et al., Blood, 2009). Recently, we encountered newly diagnosed diffuse large B-cell lymphoma (DLBCL) patients with a CD20(+) phenotype on IHC, but a CD20(−) phenotype on FCM. Neither the molecular mechanism for this phenotype nor its clinical significance, including rituximab resistance and poor prognosis, have been clarified for these patients. Aims: To determine the frequency and the clinical features, including prognosis, of newly diagnosed DLBCL with the CD20 IHC(+)/FCM(−) phenotype, the molecular mechanisms of this phenotype in primary DLBCL cells, and the effectiveness of rituximab in vitro. Methods: Primary DLBCL patients (n = 100) were enrolled in this analysis. Anti-CD20 antibodies, L26 and B9E9, were used for the IHC and FCM analyses, respectively. Alexa488-labeled rituximab was also used for the FCM analysis. Genomic DNA, mRNA, and total protein were extracted from primary DLBCL cells showing the CD20 IHC(+)/FCM(−) phenotype, and subjected to DNA sequencing, quantitative RT-PCR, and immunoblotting. Primary DLBCL cells were used for an in vitro CDC assay with rituximab. Prognoses of the patients were statistically compared by Kaplan-Meier survival analysis. Results and Discussion: Both IHC and FCM analyses were performed for 38 of the 100 primary DLBCL patients; of these 38 cases, 8 cases (21.1%) showed CD79a(+)/L26(+) on IHC and CD19(+)/CD20(−)on FCM results. Quantitative RT-PCR indicated that MS4A1 (CD20) mRNA expression was significantly lower (p = 0.0005) in these patients cells than that in CD20 IHC(+)/FCM(+) primary DLBCL cells. Immunoblotting analysis showed that CD20 protein expression was relatively lower in these patient cells than that in the positive control cells, although no differences in length could be detected. No genetic mutations in the coding sequence of the MS4A1 gene were detected in any samples examined in the DNA sequencing analysis. FCM analysis using PE-labeled B9E9 antibody showed that the mean fluorescent intensity (MFI) of CD20 IHC(+)/FCM(−) cells was significantly lower than that of IHC(+)/FCM(+) cells (p = 0.03). When Alexa488-labeled rituximab was utilized, a similar result was obtained, but the difference in MFI between IHC(+)/FCM(+) and IHC(+)/FCM(−) cells was much smaller (p = 0.21), suggesting that the sensitivity of CD20 protein recognition by rituximab was much higher than that by B9E9. Primary DLBCL cells showing CD20 IHC(+)/FCM(+) (n = 13) and IHC(+)/FCM(−) (n = 5), and cell lines showing IHC(−)/FCM(−) after using rituximab (n = 2) were analyzed by in vitro CDC assay, partially confirming rituximab-induced cytotoxicity in IHC(+)/FCM(−) cells and showing that the effectiveness of rituximab on IHC(+)/FCM(−) cells was much lower than on IHC(+)/FCM(+) cells. No cytotoxicity of rituximab was detected in IHC(−)/FCM(−) cells. Prognoses of the IHC(+)/FCM(−) (n = 8) and IHC(+)/(+) (n = 28) patients were compared by Kaplan-Meyer analysis, and no significant differences were found (p = 0.40). These data suggest that rituximab is partially effective even in DLBCL cells with a CD20 IHC(+)/FCM(−) phenotype, and that combination chemotherapy with rituximab may overcome the lower effectiveness of rituximab on these cells in the clinical setting. Further accumulation of patient data and longer follow-up durations are required. Conclusion: Lower expression of MS4A1 mRNA may be one of the critical reasons for the CD20 IHC(+)/FCM(−) phenotype. FCM analysis using fluorescent-labeled rituximab may enable detection of cell populations with lower CD20 expression and partial sensitivity to rituximab. Utilization of rituximab may be recommended even for the DLBCL patients showing a CD20 IHC(+)/FCM(−) phenotype. Disclosures: Kinoshita: Chugai Pharmaceutical Co., LTD.: Honoraria, Research Funding; Zenyaku Kogyo: Honoraria. Naoe:Chugai Pharma.: Research Funding; Zenyaku-Kogyo.: Research Funding.

Description: Background Mantle cell lymphoma (MCL) is an aggressive B-cell non-Hodgkin lymphoma (NHL) which is characterized by the t(11;14)(q13;q32) translocation leading to overexpression of cyclin D1 that consists of around 3% of NHL. Most cases are diagnosed at advanced stage and despite the progress in the treatment, the prognosis remains poor. Appropriate prognostification of the disease is essential to consider risk stratified treatment to improve the outcome. Although MCL international prognostic index (MIPI) has been reported as the specific prognostic index of MCL, most of the patients included in the study received chemotherapy without rituximab. Validation of MIPI in rituximab era has been done by several studies but the result remains controversial. Patients and Methods To evaluate the treatment outcome and prognostic factors of MCL in Japan, we conducted a multicenter retrospective study. Medical records of 633 patients with MCL newly diagnosed and treated from 63 hospitals between 1992 and 2012 were reviewed and analyzed. Pretreatment characteristics and treatment regimens were evaluated for their association with overall survival (OS) and progression free survival (PFS) by hazard ratio (HR). Since the patients received various chemotherapy regimens, treatment were adjusted to analyze prognostic factors. Results The median age of patients was 66 years old (range 22-92). Ninety percent of the patients were in stage III/IV, 70% had bone marrow involvement and 12% had poor performance status. According to the simplified MIPI (s-MIPI), 191 patients (32%) were classified as low risk, 233 patients (39%) were classified as intermediate risk and 169 patients (29%) were classified as high risk. Among the patients, 504 patients (82%) received rituximab containing chemotherapy, 55% received CHOP-like regimen, 32% received high-dose Ara-C containing regimen (HDAC) and 17% received autologous stem cell transplantation (ASCT) at the first-line treatment. With a median follow up duration of 43.4 months, overall 3-years PFS and OS was 36% and 73%, respectively. By limiting the analysis to the patients who received rituximab at the first-line treatment, s-MIPI was still a valid prognostic index for both PFS and OS even in the rituximab era (Figure), with 3-years PFS for low, intermediate and high MIPI of 62%, 36% and 14%, respectively. Multivariate analysis by Cox model revealed that Hb

Description: Background: Burkitt lymphoma (BL) is a B-cell lymphoma with high proliferating ability, derived from germinal center B-cells (GCB). The MUM-1/IRF4 gene has been identified as a myeloma-associated oncogene which is activated at the transcription level as a result of t(6;14)(p25;q32) chromosomal translocation. In normal lymphohematopoietic tissue, MUM1 protein has an important role in lymphocyte activation and in terminal B cell differentiation, and is expressed in plasma cells and in a small percentage of GCB. It is known that some of BL express MUM1, suggesting late GCB origin, but its characteristics still remain controversial. Patients and methods: A total of 91 patients previously diagnosed as BL were retrospectively analyzed. Two patients were excluded from the study, because translocation or amplificationof BCL-2 was detected by FISH in addition to translocation of MYC. The other 89 cases were confirmed as BL based on morphology, immunophenotype and the result of FISH analysis for translocation of MYC and BCL-2. The expression of MUM-1 was examined by immunohistochemistry. Furthermore, to characterize MUM1-positive (MUM1+) BL, we have compared the clinicopathologic characteristics of MUM1+ BL and MUM1-negative (MUM1-) BL. Results: Forty out of the 89 cases showed positivity for MUM1. The clinical characteristics of 40 MUM1+ BL were as follows. There were 20 men and 20 women, with a median age of 28 years, ranging from 3 to 83 years old. Thirty three cases (83%) showed extranodal involvement at presentation. The most frequent sites of extranodal involvement were bone marrow (n=14), lower gastrointestinal tract (n=14), central nervous system (n=5) and peripheral blood (n=5). Eight (20%) of MUM1+ BL patients had a bulky mass, 27 (69%) were categorized as stage III/IV, and 14 (36%) had B symptoms. With regard to laboratory data at presentation, 35 patients (90%) had elevated level of LDH. According to international prognostic index, 27 cases (69%) were identified into high/high-intermediate. Compared with MUM1- BL, patients with MUM1+ BL showed significantly younger onset (p=0.0053) and a higher ratio of females (p=0.007). The MUM1+ BL was also featured by higher percentage of elevated level of LDH, and tended to have anemia (hemoglobin level 20 years old), MUM1+ cases showed more significantly worse prognosis (p=0.0096). (Figure 1) Conclusion: MUM1+ BL showed significantly worse prognosis, particularly in adult cases, compared with MUM1- BL. In addition, the difference of the onset age, sex ratio, and involved sites between the two groups was highlighted. Our results demonstrate that MUM1 expression might predict worse prognosis of BL, and MUM1+ BL should be distinguished from MUM1- BL. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 1549 Poster Board I-572 Classical Hodgkin lymphoma (CHL) is characterized by Hodgkin and Reed Sternberg (H-RS) cells, which are B-cell origin in many cases. Recently, we highlighted an adverse prognostic significance of cytotoxic molecule (CM) expression among CHL patients (Asano N, et al. J Clin Oncol. 2006). However, the clinical characteristics of CM-positive CHL still remain controversial. We here document the clinicopathologic profiles of 35 patients with CM-positive CHL, consisting of 23 men and 12 women with a median age of 50 years (range, 16 - 84 years). All patients had lymphoadenopathy, and 14 cases showed mediastinal involvement at presentation. Physical findings included hepatomegaly and splenomegaly in six and five patients, respectively. Four patients had a bulky mass, and nine showed stage IV disease. As for laboratory data, four patients had elevated white blood cell counts (greater than 15.0 × 103/mm3). Anemia (hemoglobin level; less than 10.5 g/dl) was also present in six patients. The pathological diagnoses were nodular sclerosis type (NS) in 22 cases and mixed cellularity (MC) in 12 cases. The H-RS cells of CM-positive CHL had a prototypic immunophenotype of CD15+ CD30+ and Fascin+. All of the cases completely lacked CD20 positivity on H-RS cells, while the expression of CD3e and CD45RO was found in 2 and 1, respectively. The H-RS cells were further positive for EBV RNA transcripts in 14 of 32 (44%) cases studied by in situ hybridization method. All except three cases were negative for Pax5. A clonal TCR-g chain gene arrangement was undetected in any of the cases with successful amplification of control GAPDH DNA by PCR analysis. No cases show B-cell clonality. 27 of 35 patients received systemic multi-agent chemotherapy consisting of first-line treatment regimens as follows: doxorubicin, bleomycin, vinblastine, and dacarbacin (ABVD) (22 patients); cyclophosphamide, doxorubicin, vincristin, and prednisone (CHOP) (2 patients); two with C-MOPP; and one BEACOPP regimens. Three patients died within 6 months, before completing induction treatment because of disease progression. Overall, 22 patients responded to first-line treatment: 13 with complete response and 9 with partial response. 13 of them had relapses, and 8 died with a clinical course ranging from 7 to 142 months. Effective therapeutic approaches should be explored for CM-positive CHL patients, who resist standard treatment for CHL. Comparing CM-positive nodal peripheral T-cell lymphomas of not otherwise specified type (PTCL-N)(n=55) with CM-positive CHL(n=35), no significant differences were detected in clinical parameters except for the frequency of elevated lactate dehydrogenase level in CM-positive PTCL-N. Immunophenotypically, CM-positive CHL cases showed significantly higher rates for CD15, CD30, and Fascin expression, while CM-positive PTCL-N cases showed significantly higher CD3e, CD8, CD45RO, and CXCR3 positivity. The survival curve of CM-positive CHL showed poorer prognosis than that of CM-negative CHL (P = .0002) and better than that of CM-positive PTCL-N (P = .002) (Figure). These findings suggest that CM-positive CHL is characterized by an unfavorable clinical feature, although their histologic and phenotypic features are more suggestive of CHL. Moreover, one case in CM-positive CHL had presented the skin lesion as a CD30-positive lymphoproliferative disorder or an ALK-negative anaplastic large cell lymphoma (ALCL) at the time of recurrence. These findings suggest that CM-positive CHL represent a distinct variant form based on clinicopathologic and phenotypic traits beyond the framework of CHL. Significant overlaps in biologic and morphologic features has been identified in CHL and non-Hodgkin lymphoma. CM-positive CHL cases may be ‘intermediate lymphoma‘ between CHL and PTCL or ALK-negative ALCL. Further studies, including array CGH profiling analysis are needed to clarify the origin of CM-positive CHL. Figure Figure. Disclosures No relevant conflicts of interest to declare.

Description: Introduction Adult T-cell leukemia/lymphoma (ATLL) is a mature T-cell neoplasm caused by human T-cell leukemia virus type 1 (HTLV-1). Many studies on HTLV-1-related mRNA including HTLV-1 bZIP factor (HBZ) and Tax have been performed, mainly using cell lines, patient-derived cells, and mice. However, there are scant data concerning HTLV-1-related mRNA in formalin-fixed, paraffin-embedded (FFPE) tissue samples. We detected HBZ and Tax mRNA on FFPE tissue samples using in situ hybridization (ISH), and investigated the association with clinicopathological characteristics. Materials and methods Eighty-seven biopsy samples from newly diagnosed ATLL patients were examined. The same samples were used in a previous study (Miyoshi et al. Blood, 2016). Use of patient materials and clinical information was approved by the Ethical Committee of Kurume University, in accordance with the Declaration of Helsinki. ISH was performed on FFPE tissue samples using RNAscope 2.5 HD Reagent Kit-BROWN (Advanced Cell Diagnostics, Hayward, CA). HBZ-and Tax-specific probes were used. MT-4 (HTLV-1 immortalized cell line) and Jurkat (T-cell acute lymphoblastic leukemia cell line) were used as positive and negative controls, respectively. Dot-like signals were counted at high magnification (40 diameters) on 10 randomly selected fields and the number of signals per 1,000 ATLL cells were calculated. The results were reviewed by two experienced hematopathologists. High expression was indicated when more than the median value of HBZ or Tax signals was stained. The antibodies for immunohistochemistry targeted CD4, CD30, Ki-67(MIB-1), CCR4, FoxP3, GATA3, IRF4, HLA class I, β2-microglobulin (β2M), PD-1, and PD-L1. Clinicopathological characteristics of ATLL patients were compared by Fisher's exact test (2-sided), Mann-Whitney's U test, and Spearman's rank correlation analysis. Overall survival was estimated by the Kaplan-Meier method and compared by the Log-rank test. P〈 .05 was considered statistically significant. EZR ver. 1.32 was used for all statistical analyses. Results In the histogram and scatter plot of HBZ-ISH and Tax-ISH, the median values of HBZ signals and Tax signals were 806/1000 ATLL cells (range 0.4 - 4013.1) and 5.0/1000 ATLL cells (range 0.1 - 891.2), respectively (Figure 1). Representative samples of HBZ-ISH and Tax-ISH are presented in Figure 2A and 2B, respectively. Notably, the high-expression group of HBZ displayed significant reductions in skin lesions (P = .025), Ann Arbor stage (P = .021), and peripheral blood involvement (P = .028). The high-expression group of Tax displayed significant increases in lactate dehydrogenase activity (P = .0020), splenomegaly (P = .0070), CD 30 (P = .015), and PD-1 in tumor-infiltrating lymphocytes (PD-1_TIL; P

Description: Age-related Epstein-Barr virus–associated B-cell lymphoproliferative disorder (aEBVLPD) is a disease group characterized by EBV-associated large B-cell lymphoma in elderly without predisposing immunodeficiency. In nearly one- third of cases, aEBVLPD occurs as a polymorphous subtype with reactive cell-rich components, bearing a morphologic similarity to classic Hodgkin lymphoma (cHL). The aim of this study was to clarify clinicopathologic differences between the polymorphic subtype of aEBVLPD (n = 34) and EBV+ cHL (n = 108) in patients aged 50 years or older. Results showed that aEBVLPD was more closely associated with aggressive clinical parameters than cHL, with a higher age at onset (71 vs 63 years); lower male predominance (male-female ratio, 1.4 vs 3.3); and a higher rate of involvement of the skin (18% vs 2%), gastrointestinal tract (15% vs 4%), and lung (12% vs 2%). aEBVLPD was histopathologically characterized by a higher ratio of geographic necrosis, greater increase (〉 30%) in cytotoxic T cells among background lymphocytes, higher positivity for CD20 and EBNA2, and absence of CD15 expression. As predicted by the clinical profile, aEBVLPD had a significantly poorer prognosis than EBV+ cHL (P 〈 .001). The polymorphous subtype of aEBVLPD constitutes an aggressive group with an immune response distinct from EBV+ cHL, and requires the development of innovative therapeutic strategies.

Description: [Background] Mantle cell lymphoma (MCL) is an aggressive B-cell lymphoma, characterized by the overexpression of cyclin D1 derived from t(11;14)(q13;q32) and poor prognosis. Most MCLs show nodal presentation, but also accompany extranodal involvement, such as bone marrow, peripheral blood or gastrointestinal tract. As a result, many MCLs present with advanced stage disease. Since only a small portion of patients show limited-stage disease, minimal data exist on treatment of patients diagnosed with limited stage disease. Nevertheless, the treatment strategy of MCL is recommended according to the clinical stage of limited- (stage I or non-bulky II) vs. advanced-stage, as well as other types of lymphoma. [Patients and methods] We recently collected 633 patient data of MCL (Chihara, et al. Ann Oncol 2015). Information of clinical stage was available in 626 patients. The patient data were retrospectively analyzed the by the clinical stage at initial presentation. [Results] The clinical stage was I in 24 patients (4%), II in 33 (5%), III in 70 (11%), and IV in 499 (80%). Only one patient presented with bulky stage II. Detailed demographic information by the clinical stage are listed in Table. Age and sex were not significantly different by clinical stage. Limited stage patients were associated with better performance status (PS), less B symptoms, no extranodal involvement, and lower lactate dehydrogenase (LDH) level and white blood cell (WBC) count. Most patients in any stage were treated with cytotoxic chemotherapy, but more patients in limited stage received radiotherapy. The proportion of high-dose cytarabine (HDCA)-containing regimen over CHOP/CHOP-like was higher in advanced stage patients. Complete and overall response rates were 92% and 96% in stage I, 58% and 94% in stage II, 66% and 86% in stage III, and 52% and 82% in stage IV, respectively (P = 0.02). However, the higher response rate in limited stage patients did not translate into better prognosis. The median survival was 11.0 years in stage I, 13.4 years in stage II, 11.5 years in stage III, and 5.6 years in stage IV (Figure). The prognosis was not significantly different among patients with stage I, II, and III (P = 0.33). [Conclusion] Prognosis of limited-stage MCL was almost similar to that of stage III MCL. Although the present study includes several limitations including a retrospective nature and limited number of patients, prognosis of patients with limited-stage MCL was not satisfactory. The significance of radiotherapy, as well as the optimal choice of chemotherapy, for limited-stage MCL needs re-evaluation. Table Table. Figure Figure. Disclosures Suzuki: Chugai: Honoraria; Kyowa Hakko kirin: Honoraria; Bristol-Myers Squibb: Honoraria. Asano:Jannsen: Honoraria; Chugai: Honoraria. Kinoshita:Ono: Research Funding; Gilead: Research Funding; Zenyaku: Honoraria, Research Funding; Takeda: Research Funding; Chugai: Honoraria, Research Funding; Eisai: Honoraria, Research Funding; Solasia: Research Funding; Janssen: Honoraria; Kyowa Kirin: Honoraria. Suzumiya:Chugai: Honoraria, Research Funding; Astellas: Research Funding; Eisai: Honoraria, Research Funding; Takeda: Honoraria; Toyama Chemical: Research Funding; Kyowa Hakko kirin: Research Funding. Ogura:SymBio Pharmaceuticals: Consultancy, Honoraria; Celltrion, Inc.: Consultancy, Honoraria.

Description: Abstract 2677 Background: Classical Hodgkin lymphoma (CHL), which is characterized by the presence of Hodgkin and Reed Sternberg (H-RS) cells in a background of non-neoplastic inflammatory cells, is divided into four histological subgroups, nodular sclerosis (NSCHL), mixed cellularity (MCCHL), lymphocyte-rich, and lymphocyte depletion. While NSCHL in young adults is characterized by a mediastinal mass and good prognosis, the clinicopathological characteristics of NSCHL in the elderly (NSCHL-e) remain uncertain. Patients and methods: Enrolled patients were diagnosed with CHL between 1986 and 2006 as part of the Hodgkin Lymphoma's Multicenter Study Group. To better characterize NSCHL-e, we compared the clinicopathological profiles of 84 NSCHL-e patients aged 50 or over with 237 NSCHL-y patients aged 49 or younger and 302 with MCCHL. Results: The total of 743 CHL patients consisted of 496 men and 247 women with a median age of 48 years (range, 15– 89 years). The pathological diagnoses were NSCHL in 324 patients (43%) and MCCHL in 303 (41%). NSCHL patients showed a bimodal age distribution, with an initial peak in their 20s and a second small peak in their 60s. We categorized the former as NSCHL-y (49 or younger) and the latter as NSCHL-e (50 and over). NSCHL-e patients were characterized by male predominance and a more advanced clinical stage (53%) than NSCHL-y. Immunophenotypically, H-RS cells had the prototypic immunophenotype of CD15+ CD30+ and Pax5+. NSCHL-e cases showed a significantly higher rate of CD20 (24%) than NSCHL-y (8%, P = 0.001). Furthermore, H-RS cells in 29 of 75 (39%) patients with NSCHL-e were positive for EBV RNA transcripts by in situ hybridization, whereas only 7% of NSCHL-y cases were EBER-positive (P 〈 0.0001) (Table). Regarding NSCHL-e and MCCHL, no significant difference between these patients was seen in clinical characteristics. Immunophenotypically, NSCHL-e patients showed significantly higher rates for CD3 and TIA-1, while MCCHL patients showed higher EBV positivity (75%). Fifty-five of 63 patients received systemic multi-agent chemotherapy as first-line treatment, consisting of doxorubicin, bleomycin, vinblastine, and dacarbacin (ABVD) in 38 patients; CHOP in 8; C-MOPP in 8; and BEACOPP in 1. Overall, 51 patients responded to first-line treatment, 39 with complete response and 12 with partial response. Disease-specific survival of NSCHL-e was poorer than that of NSCHL-y (P 〈 0.001) but similar to that of MCCHL (P = 0.43) (Figure). Conclusion: NSCHL-e is characterized by an unfavorable prognosis and different clinicopathological features to NSCHL-y, which is considered as typical NSCHL. A number of cases of NSCHL-e might have been associated with MCCHL, with most being EBV-positive. These results suggest the limitations of current histological subgroupings for CHL. Disclosures: Matsushita: Pfizer CO.: Membership on an entity's Board of Directors or advisory committees, Research Funding; Baxter Co.: Membership on an entity's Board of Directors or advisory committees, Research Funding.

Description: Introduction: Human T-cell lymphotropic virus (HTLV-1) carriers are often found in coastal areas of the Kyushu-Okinawa region in Japan. Of the patients infected with the virus, only 2-5% will develop adult T-cell leukemia/lymphoma (ATLL). ATLL is classified into four clinical subtypes as follows: acute, lymphoma, chronic, and smoldering. Chronic-type ATLL has a better prognosis compared with acute and lymphoma types. However, in half of patients, chronic-type ATLL progresses to leukemia/lymphoma crisis. Abnormal expressions of CD58 and beta-2 microglobulin (b2M) were recently reported to be involved in the immune escape mechanism of tumor cells in diffuse large B-cell lymphoma (Pasqualucci et. al. Cancer Cell 2014). In our previous study, acute-type ATLL had significantly higher rates of CD58 gene deletion, which was associated with leukemia/lymphoma crisis in chronic-type ATLL (Yoshida et. al. Cancer Res 2014). However, the clinicopathological significance of HLA expression in ATLL patients remains controversial. In addition, we previously reported that the expression of programmed cell death ligand 1 (PD-L1), one of the regulatory factors of tumor immune responses, influenced disease prognosis in ATLL patients. Especially PD-L1 expression in stromal cells in tumor microenvironments is associated with better prognosis of ATLL (Miyoshi et al. Blood 2016). The present study evaluated the clinical relevance of HLA, b2M, and PD-L1 expressions, with particular emphasis on predicting overall survival of ATLL patients. Methods: We report the clinicopathological profiles of 123 patients diagnosed as having ATLL at Kurume University (Kurume, Japan) according to the 2008 World Health Organization classification system. Among these patients, 45 had the acute type and 52 had the lymphoma type, with a median age of 66 years (range, 13-90 years). We examined the immunohistological expressions of HLA and b2M proteins in all the 123 patients by using tissue microarray. The expression patterns and clinicopathological features were also examined statistically. All data were analyzed by using the STATA software version 12.0 (StataCorp, College Station, Texas). Results: In the immunohistochemical study, 91 (74%) of the 123 patients were positive for HLA, including 60 who were cell membrane positive (HLAm+), 31 who demonstrated cytoplasmic localization (HLAcy+), and 32 who were negative for HLA (HLA−). In addition, 89 (72%) of the 123 patients were positive for b2M, including 54 who were cell membrane positive (b2Mm+), 35 who demonstrated cytoplasmic localization (b2Mcy+), and 34 who were b2M negative (b2M−). The HLAm+b2Mm+ group, which showed membrane staining for both HLA and b2M, accounted for 48 cases (39%) in the present study (Table 1). In terms of clinical features, no significant differences were found between the HLAm+b2Mm+ group and the other groups. Immunophenotypical evaluation revealed significantly higher rates of CD30-positive lymphoma cells (P = 0.004) and PD-L1-positive stromal cells in microenvironments (PD-L1+stromal cells ≥50/HPF: miPD-L1high) in the HLAm+b2Mm+ patients (P = 0.008) than in the patients in other groups. Disease-specific survival (DSS) curves showed that the HLAm+b2Mm+ group had significantly better prognosis than the other groups (Figure 1-a). Especially the HLAm+b2Mm+ group with miPD-L1high had an excellent prognosis, with a 5-year survival rate of 77% (Figure 1-b). The multivariate analysis demonstrated that the membrane expressions of both HLA and b2M in lymphoma cells with increased number of PD-L1-positive stromal cells were an independent factor of good prognosis in ATLL (Table2). Conclusions: The results of our study demonstrated the following three points: 1) The expression patterns of HLA and b2M in the patients with ATLL varied. 2) The HLAm+b2Mm+ group showed significantly better prognosis than the other groups. Especially the HLAm+b2Mm+ group with miPD-L1high demonstrated the most favorable prognosis. These results suggest that the host immune surveillance mechanism for lymphoma cells in the HLAm+b2Mm+ group with mPD-L1high is still functioning. 3) The membrane expressions of HLA and b2M in ATLL cells are useful as a predictive marker of good prognosis and increase in the number of PD-L1-positive stromal cells in tumor microenvironments. Figure. Figure. Disclosures Asano: Chugai: Honoraria; Jannsen: Honoraria. Ohshima:Chugai: Research Funding, Speakers Bureau; Kyowa Kirin: Research Funding, Speakers Bureau.