ALBERT

Description: Introduction: Diffuse large B-cell lymphoma (DLBCL) is the largest subtype of lymphoma, but consists of heterogeneous groups with different prognosis. Many studies conducted to determine prognosis factors of DLBCL. Recently, several studies have been reported that the elderly Epstein-Barr virus (EBV)-positive DLBCL patients showed worse prognosis than the elderly EBV-negative DLBCL. (Oyama T et al. Clin Cancer Res 2007;13:5024-5032). On the other hand, other studies reported that EBV-positivity is not a prognosis factor in young-adult DLBCL patients (Hong JY et al. Annals of Oncology 2015;26:548-555, Nicolae A et al. BLOOD 2015;126:863-872). In the present study, retrospective analyses were carried out for patients with DLBCL diagnosed between 1990 and 2007 to analyze the clinical and prognostic significance of EBV. Patients: Those who met the following criteria were included in the study: (i) pathology confirmed de novo DLBCL, according to the WHO classification; (ii) adequate amount and quality of paraffin-embedded biopsy specimens or unstained slides for EBV-encoded RNA in situ hybridization. Statistical methods: Intergroup comparisons were carried out with Fisher's exact test for categorical variables and the Mann-Whitney test for age. For survival analysis, we carried out Kaplan-Meier and multivariate proportional hazard (Cox) analyses. Statistical analyses were carried out using the software Stata SE version 14.0. Results: A total of 456 patients were included. The median follow up duration of survivor was 100 months (range, 1.9-363). There were 236 male (51.75%) and 220 female patients (48.25%) with a median age of 67 (range, 22-94) years. 246 patients (54.0%) were in advanced (III/IV) stage, 134 (29.4%) presented with poor Eastern Cooperative Oncology Group Performance Status (ECOG-PS), 275 (60.3%) with elevated lactate dehydrogenase (LDH) level, 263 (57.7%) with extranodal involvement of 1 or more sites, and 64 (14.0%) with bulky mass. As a consequence, 211 patients (46.3%) were in high/high-intermediate International Prognostic Index (IPI) categories. 5 years overall survival (OS) was 66.2%, in total. EBV was detected in samples from 47 patients (10.3 %). The positivity was 10.6% (15 / 142) in patients of 60 years or younger, and 10.2 % (32 / 314) in those older than 60 years (p = 0.51). No differences in back ground date were found between EBV - positive and negative DLBCLs regarding age (median 66.5 vs. 64.8, P = 0.17), male sex (88.1% vs. 11.9%, P = 0.16), advanced stage (88.2% vs. 11.8%, P = 0.17), poor ECOG-PS (90.3% vs. 9.7%, P = 0.47), elevated LDH (89.5% vs. 10.6%, P = 0.48), extranodal sites of one or more (90.5% vs. 9.5%, P = 0.31), bulky mass (92.2% vs. 7.8%, P = 0.33), and higher IPI categories (88.6% vs. 11.4%, P = 0.29). The prognosis was also not different between EBV-positive and negative DLBCLs. The 5 years OS was 42.6% for EBV-positive DLBCLs, and 56.4% for EBV-negative DLBCLs (P = 0.11). The same results were found for the younger (53.3% vs. 66.9%, P = 0.30) and the elderly patients (42.9% vs. 51.8%, P = 0.18). Conclusion: In this practice based, uncontrolled study, EBV-positively wad not different in younger and elderly DLBCLs. Although EBV-positive DLBCL showed worse trend of prognosis, the difference was not statistically significant. The significance of EBV in DLBCL should be reconsidered in unbiased, large-scale patient date. Figure Figure. Disclosures Suzuki: Chugai: Honoraria; Bristol-Myers Squibb: Honoraria; Kyowa Hakko kirin: Honoraria. Suzumiya:Chugai: Research Funding, Speakers Bureau; Astellas: Research Funding; Toyama Chemical: Research Funding; Eizai: Research Funding; Takeda: Speakers Bureau; Kyowa Hakko Kirin: Research Funding.

Description: Abstract 2284 Poster Board II-261 Background: To evaluate the role of autologous and allogeneic SCT in the treatment of PTCLs, Japan Study Group for Cell Therapy and Transplantation conducted a multicenter retrospective survey in Japan and Korea. Methods: After excluding patients with adult T-cell leukemia/lymphoma and NK-cell tumors, patient data were newly collected from 330 patients (222 male and 108 female) with a median age of 49 years (range, 13–71) who underwent SCT between 9/1991 and 12/2008 (196 autologous and 134 allogeneic including 31 patients with previous autograft). Allogeneic SCT (53 BM, 54 PB, 1 BM+PB, 26 CB) was performed using a reduced-intensity conditioning (RIC) in 84 patients (63%). While a pathologic central review will be performed, currently there were 159 (48%) patients with PTCL, not otherwise specified, 63 (19%) with angioimmunoblastic T-cell lymphoma, 47(14%) with anaplastic large cell lymphoma (23 ALK-negative, 14 ALK-positive and 10 unknown), 12 (4%) with enteropathy-associated T-cell lymphoma, and others. The disease status at transplant in the allo-group was significantly worse than that in the auto-group (Table 1). The median number of chemotherapy regimens was 2 (range, 1–7), and the median duration between diagnosis and transplant was 267 days (range, 120-4889 days). Results: The median follow-up for surviving patients was 45 mo (range, 2.3–141 mo). There was no significant difference in overall survival among different groups, including histological subtypes, RIC and myeloablative conditioning in the allo-group and high-dose chemotherapy regimens in the auto-group. Early survival rate after transplant was significantly better for auto-group than allo-group (Wilcoxon P=0.001), but the difference was marginal in the total course (Logrank P=0.06) (Figure). The non-relapse mortality (NRM) in the auto-group was significantly lower than that in the allo-group (P1), cell source (CB/BM+PB), and performance status (PS; 〉1), stage, chemorefractory disease, international prognostic index (IPI; H-I/H risk) and prognostic index for PTCL, unspecified (PIT; group 3/4) at transplant. The risk factors in the allo-group were bulky mass at diagnosis, age (〉50 years), cell source, and PS, stage, IPI and PIT at transplant, while those in the auto-group were age (〉40 years), recurrence after frontline therapy, number of chemotherapy regimens, and stage, chemorefractory disease, IPI and PIT at transplant. Conclusions: Despite a worse disease status at transplant in the allo-group, the overall survival was comparable to that in the auto-group. This supports the notion that early allogeneic SCT is a valuable treatment option for PTCLs, although a large-scale randomized trial to identify a suitable upfront-transplant type for chemosensitive patients with PTCLs is warranted. Disclosures: No relevant conflicts of interest to declare.

Description: De novo CD5-positive diffuse large B-cell lymphoma (CD5+ DLBCL) is well known to have clinicopathologically and genetically different characteristics from CD5-negative (CD5−) DLBCL and mantle cell lymphoma. To obtain more clinicopathologic information of CD5+ DLBCL, we reviewed specimens from 128 cases of CD5+ DLBCL and analyzed the relationship between the morphologic features and a long-term survival. The current series includes the 109 CD5+ DLBCL cases described in our previous study (Yamaguchi M, Blood 2002). All patients were diagnosed between 1984 and 2002 as having DLBCL according to the WHO classification. They had no past history of any other lymphoproliferative disorders and were immunohistochemically confirmed to be cyclin D1-negative. Four morphologic variants were identified: common (monomorphic) (n=97), pleomorphic (n=15), giant-cell rich (n=14) and immunoblastic (n=2). Intravascular or intrasinusoidal infiltration was seen in 41% of the cases, and bi-nucleated snowman-like cells were seen in 80% of the cases. BCL2 protein expression in CD5+ DLBCL was more frequent than that in CD5− DLBCL (P=0.0009). Immunohistochemical analysis in a serial 45 cases of CD5+ DLBCL revealed that 80% of CD5+ DLBCL cases (36/45) were classified as non-germinal center B-cell type DLBCL. CD10 was positive in 8 cases (18%), and MUM1 was positive in 93% of the cases (42/45). The estimated 5-year overall survival rate was 36% after a median observation time of 79 months. Patients with the common variant showed better prognosis than those with the other three morphologic variants (P=0.018). Multivariate analysis showed that age more than 60 at diagnosis [hazard ratio (HR) 2.25, 95% confidence interval (CI) 1.36–3.73], advanced stage (HR 1.76, 95%CI 1.06–2.93), high serum LDH level (HR 2.2, 95%CI 1.21–4.00), and histopathologic categorization (not common) (HR 1.6, 95%CI 1.00–2.59) were independent prognostic factors for survival. In all, 16 patients (13%) developed central nervous system (CNS) recurrence. Our study revealed the morphologic spectrum and a high incidence of CNS recurrence in CD5+ DLBCL, and confirmed that CD5+ DLBCL shows frequent BCL2 protein expression and it is mainly regarded as non-germinal center B-cell type of DLBCL.

Description: Introduction: Multiple myeloma is still incurable and optimize existing chemotherapeutic strategies and development of novel agents are necessary to improve the outcome of patients. Cotylenin A, a fusicoccane diterpene glycoside with a complex sugar moiety, was isolated as a plant-growth regulator. Cotylenin A modulates the 14-3-3 intracellular signaling pathway and has been shown to inhibit the growth of several cancer cells. Herein, we examined the antitumor effects of cotylenin A to develop a novel treatment against myeloma. Methods: Five human myeloma cell lines, RPMI8226, KMS-11, KMS-26, KMS-12 PE and KMS-12 BM were cultured with cotylenin A alone or in combination with several anticancer drugs. To measure the effects of various drugs on the growth of myeloma cells, the number of viable cells was determined by the MTT assay after 6 days of exposure to various concentrations of drugs with or without 2 μg/ml cotylenin A. The growth-inhibiting effects of the drugs were examined by determining the concentrations of drugs required to reduce the cell number to one-half of that in untreated cells (IC50). Xenografts in six-week-old female (Fox Chase SCID C.B-17/Icr-scid Jcl) mice were used to determine the in vivoefficacy of cotylenin A. Results: Cotylenin A alone inhibited the growth of myeloma cells in dose dependent manner, but the effect of growth inhibition was relatively weak. Cotylenin A and vincristine synergistically inhibited the growth and induced apoptosis in myeloma cells. While other microtubule-disturbing agents also showed co-operative effects with cotylenin A, other anticancer agents, such as doxorubicin, cisplatin, camptothecin, methotrexate, gemcitabine and 5-fluorouracil, did not show such co-operation with cotylenin A (Table 1). Cotylenin A had synergistic effects with vincristine and the results were confirmed by an isobologram analysis. These differences might be attributed to the effects on autophagic responses. Combined treatment with cotylenin A and vincristine induced autophagy (formation of LC3-II and degradation of p62 protein). However, doxorubicin did not enhance the autophagy induced by cotylenin A. The induction of apoptosis was confirmed by an analysis of the DNA histogram and the expression of annexin V. The expression of cleaved caspase-3 was increased in treatment with 2 ng/ml vincristine and enhanced by 2 μg/ml cotylenin A by western blot analysis. An invasion assay using transwell chamber revealed that low concentration cotylenin A (0.6 μg/ml) effectively inhibited the invasive activity of RPMI 8226 myeloma cells without inhibiting growth. A colony-forming assay indicated that 2 μg/ml cotylenin A preferentially inhibited the formation of large colonies, which have high self-renewal activity. The combined treatment with 3 μg/ml cotylenin A and 3 ng/ml vincristine more effectively suppressed the formation of large colonies than vincristine alone. Expression of pluripotency-associated transcription factor Sox2 mRNA in RPMI 8226 myeloma cells was significantly suppressed by treatment with 4 μg/ml cotylenin A. Combined treatment with 5 mg/kg cotylenin A and 0.5 mg/kg vincristine significantly inhibited the growth of KMS-26 myeloma cells as xenografts. In addition, mice with cotylenin A and vincristine showed the reduction of body weight loss comparing with those of mice by vincristine alone. This result supported that the combination of cotylenin A and vincristine might be safe. Conclusions: Cotylenin A and vincristine synergistically inhibited the growth of myeloma cells in vitro and in vivo, and the invasion of myeloma cells. Our results suggest that the combination of cotylenin A and vincristine may have therapeutic value for myeloma. Table 1. Potentiation of the growth-inhibitory activities of various anticancer agents in RPMI8226 myeloma cells by cotylenin A. Growth inhibition (IC50) Anticancer agent (ng/ml) - Cotylenin A + Cotylenin A Ratio(-/+) Doxorubicin 5.5 ± 0.6 5.3 ± 0.5 1.03 Cisplatin 246 ± 30.2 237 ± 28.4 1.03 Camptotecin 1.42 ± 0.16 1.38 ± 0.14 1.03 Methotrexate 2.76 ± 0.3 1.81 ± 0.2 1.52 Gemcitabine 4.3 ± 0.4 3.1 ± 0.3 1.39 5-Fluorouracil 56.5 ± 5.1 57.8 ± 6.3 0.98 Vincristine 6.3 ± 0.6 1.6 ± 0.1 3.94 Vinblastine 0.91 ± 0.11 0.42 ± 0.05 2.17 Paclitaxel 16.4 ± 1.9 7.7 ± 0.9 2.13 Ratio (-/+), IC50 without cotylenin A: IC50 with cotylenin A. The values are the mean ± SD of four determinations. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 4594 Background: B-chronic lymphocytic leukemia (B-CLL) is rare in Japan. Therefore, only a few data are available on the clinical characteristics of B-CLL for Japanese patients (pts). We reported that the rate of atypical CLL appeared to be higher in Japanese pts than that of Caucasians. Patients and Methods: This nation-wide study was performed from Oct. 2003 to March 2008, and was supported by Japanese Elderly Leukemia and Lymphoma Study Group. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Boards of the core members' hospitals. The pts with CLL and CLL-like disease were registered. Precursor leukemia and adult T-cell leukemia/lymphoma were excluded. The smears of peripheral blood and bone marrow of all the registered pts were examined by the Diagnosis Committee by using the 4th edition of WHO classification and FAB classification. FAB classification defines typical CLL as prolymphocytes≤10% and atypical CLL including CLL/PLL as 11%≤prolymphocyte

Description: Background Clinical impact of chronic hepatitis B virus (HBV) and hepatitis C virus (HCV) infections is not well studied in a large series of patients with lymphoma undergoing autologous stem cell transplantation (ASCT). Retrospective analysis was performed to evaluate clinical outcomes of patients with lymphoma receiving ASCT with or without HBV and HCV infections. Patients and Methods Among 13218 adult patients registered in the Adult Lymphoma Working Group of the Japan Society for Hematopoietic Cell Transplantation (JSHCT) database, we selected patients who had the information on HBV and HCV infection status and received ASCT from 1989 to 2010. Patients diagnosed as Hodgkin lymphoma, mature or precursor T- and B-cell lymphoma were included. Histological diagnosis of natural killer (NK)-cell lymphoma and adult T-cell leukemia/lymphoma were excluded in this analysis. Patients who had human immunodeficiency virus (HIV) infection and history of previous transplantation were also excluded. Survival analysis was performed to compare patients positive for HBV and HCV with those negative for both. Prognostic indicator for survival was also investigated in patients positive for HBV and HCV. Results A total of 4641 patients with 2819 male were analyzed. The median age of all patients was 54 years (range: 16-81 years). The patient characteristics were summarized in Table 1. HBV and HCV infections were present in 162 (3.5%) and 104 (2.2%) patients, respectively. Ten patients (0.2%) had both HBV and HCV infections. Detailed data on anti-HBV antigen or antibody and genetic data were not available in the analysis. Patients with HBV and HCV infections were older and more diagnosed with advanced stage than those without infections. With a median follow-up of 2.8 years, the 2-year overall survival (OS) rate of all patients was 75% (95%CI: 73 to 76). Cumulative incidence of treatment-related mortality (TRM) at one year after ASCT was 5.8% (95%CI: 5.2 to 6.6). According to HBV and HCV positivity, the estimated 2-year overall survival (OS) rates were 74% (95%CI: 65 to 81), 77% (95%CI: 66 to 84), 60% (95%CI: 25 to 83) and 75% (95%CI: 73 to 76) in patients with HBV-positive, HCV-positive, both virus positive and negative patients, respectively. Cumulative incidence of TRM at one year was 11% (95%CI: 6.2 to 16) 6.6% (95%CI: 2.7 to 13) and 5.7 (95%CI: 5.0 to 6.4) in patients with HBV-positive, HCV-positive and both virus negative patients, respectively. In patients with or without HBV and HCV infections, there was no statistically significant difference in rates of OS (p=0.82, Figure 1) and TRM (p=0.63). In multivariate analysis of patients infected with HBV, factors associated with shorter OS were male sex (HR: 3.1, 95%CI: 1.2 to 7.6), non-remission/relapse status at ASCT (HR: 2.7, 95%CI: 1.2 to 5.9) and ASCT before 2005 years (HR: 3.6, 95%CI: 1.6 to 8.4). ASCT before 2005 years was associated with higher TRM in patients with HBV infection (HR: 4.4, 95%CI: 1.5 to 13). In patients with HCV infection, multivariate analysis revealed that partial remission (HR: 3.5, 95%CI: 1.2 to 10.6), non-remission/relapse (HR: 5.3, 95%CI: 1.9 to 14.6) status at ASCT and age≥ 50 years at ASCT (HR: 7.0, 95%CI: 1.5 to 31.9) were significantly associated with shorter OS. Non-remission/relapse status at ASCT was associated with higher TRM in patients infected with HCV (HR: 4.4, 95%CI: 1.1 to 18.3). Difference in histology was an adverse factor for outcomes in the HBV- and HCV-negative groups, however, the factor could not be identified as adverse indicator in the HBV- and HCV-positive groups. Conclusions Prognosis of patients with HBV and HCV infections would be comparable to patients without those infections. Disease status at ASCT could be one of useful landmarks to predict outcomes in patients positive for HBV and HCV undergoing ASCT. After careful consideration, ASCT might be a feasible option for patients with HBV and HCV infections. Disclosures: No relevant conflicts of interest to declare.

Description: Introduction: Empiric antifungal therapy (EAT) is recommended for persistent or recurrent febrile neutropenia based on an old randomized controlled trial, but such treatment is apparently overtreatment for the majority of patients. On the other hand, preemptive therapy triggered by positive blood tests for fungal antigens and/or imaging study findings was shown to increase the incidence of invasive fungal infection, and thus, a risk-based approach is important. The D-index, which is defined as the area over the neutrophil curve during neutropenia and hence reflects both the duration and depth of neutropenia (Figure 1A), enables real-time monitoring of the risk of invasive fungal infection. Previous studies showed that the cumulative D-index (c-D-index), which was calculated as cumulative D-index from the onset of neutropenia (Figure 1B), had high negative predictive values for invasive mold infection or pulmonary infection with cutoff values of 5,800 or 5,500 in high-risk neutropenic patients [J Clin Oncol 2009; 27: 3849-54. Biol Blood Marrow Transplant 2010; 16: 1355-61]. Methods: We investigated a novel approach, called D-index-guided early antifungal therapy (DET) and compared it to EAT in high-risk neutropenic patients. In the EAT group, empiric antifungal therapy was started for persistent (〉=4 days) or recurrent febrile neutropenia. For patients with persistent or recurrent febrile neutropenia in the DET group, preemptive antifungal therapy was applied until c-D-index reached 5,500, but antifungal agent was initiated after c-D-index exceeded 5,500, even if there was no significant finding in serum fungal makers or imaging studies, to prevent excessive invasive fungal infection. Micafungin at 150 mg/day was administered as EAT or DET in this study. We randomized 423 patients who underwent chemotherapy or hematopoietic stem cell transplantation for hematological malignancies, in which predicted period of neutropenia exceeded 7 days, into the EAT group or the DET group, and 413 were eligible for intent-to-treat analyses (201 patients in the EAT group, 212 patients in the DET group). The prophylactic use of fluconazole or itraconazole was allowed. Primary endpoint was the development of proven/probable invasive fungal infection. Results: Backgrounds of the patients were similar between the 2 groups (Table 1). Invasive fungal infection (proven/probable/possible) was observed in 12 patients (6.0%) of the EAT group and 5 patients (2.4%) of DET group, respectively. Proven/probable invasive fungal infection was identified in 5 patients (2.5%) of the EAT group and 1 patient (0.5%) of DET group, which fulfilled the predetermined criteria of non-inferiority of the DET group. Regarding the pathogens, the EAT group included 1 case of candidemia and 4 cases of invasive pulmonary aspergillosis, and the DET group included one fusariosis. The survival rate of the EAT and DET group was 98.0% vs. 98.6% at day 42 and 96.4% vs. 96.2% at day 84, respectively. During the observation period, 31 patients died due to disease progression (n=19), infection (n=5) or other causes (n=7). Causes of infection related mortality included Pseudomonas aerginosa infection (n=2), fusariosis (n=1), toxoplasmosis (n=1) and septic shock by unknown pathogen (n=1). The frequency of micafungin use was significantly lower in the DET group than the EAT group (32.5% vs. 60.2%, P

Description: Background: The immunological graft-versus-leukemia effects associated with chronic myeloid leukemia (CML) are well-established. Both regulatory T cells (Treg) and natural killer (NK) cells play a role in the maintenance of CML. In this study, we examined changes in the number of circulating Treg and NK cells and the NK activity in CML patients before and after treatment with a tyrosine kinase inhibitor (TKI). Our goal was to evaluate the clinical significance of this treatment on these two different cell types. Moreover, the levels of these cells suggest that there is a need to make the TKI more effective as an immunotherapy. Patients and Methods: Treg and NK cells were characterized and quantified by flow cytometry in 27 newly diagnosed CML patients, 19 newly diagnosed patients with other myeloproliferative neoplasms (MPN), and 15 healthy controls (HC-group). CML patients, who were diagnosed as chronic phase and initially treated with a TKI between September 2007 and July 2017 and were followed up for more than 12 months were enrolled in the study. The molecular response was analyzed by quantitative PCR of the BCR-ABL fusion genes at approximately 6-month intervals after the initiation of treatment. After obtaining informed consent, peripheral blood (PB) was drawn from all patients at their initial diagnosis and at 6-month intervals thereafter. CD4+CD25+Foxp3+ Treg cells, CD3+CD4+T cells, CD3+CD8+ T cells, and CD3-CD56+ NK cells were analyzed using flow cytometry, and the absolute numbers of these cells were calculated. The NK activity was measured by the standard 51Cr release assay at an effector:target (E:T) ratio of 20:1. Results: The median age for CML patients was 65 years (range: 44-85), and 17 out of 27 of the patients were male. The median period of observation was 656 days. After 12 months, 12 CML patients (MMR-group) had a major molecular response (MMR), and 15 patients (NR-group) did not have a MMR. Treg% values in the PB before treatment in the CML-group (0.39±0.30%) were significantly lower than that in the HC-group (0.70±0.29%, p

Description: Abstract 3093 Allogeneic hematopoietic stem cell transplantation (allo SCT) is potentially curative treatment for relapsed follicular lymphoma (FL) that remains non-curative disease even with modern immunochemotherapy. Number of potential candidate of allo SCT for FL has been increasing with the development of reduced intensity conditioning (RIC) regimens and increased donor availability due to use of hematopoietic stem cell from unrelated donor including cord blood (CB). However, short-term and long-term transplantation-related complications are still major obstacles in applying this treatment. Recently, EBMT and CIBMTR reported a prognostic score of allo SCT for FL (Ann Oncol 2011;22 :Suppl 4, iv94). However, this score awaits validation. To elucidate prognostic factors for OS after allo SCT for FL, we conducted a retrospective study using the national registry data of the Japan Society of Hematopoietic Cell Transplantation (JSHCT). In total, 472 cases of allo SCT for FL performed from 2000 to 2008 were identified. 220 (46.6%) were male and the median age at allo SCT was 50 yo (range: 27–75). Forty six (9.7%) patients were 60 yo or older. Eighty two (17.3%) patients had previous history of autologous transplantation (ASCT). Stem cell source was BM or PB from related donor in 215 (45.5%), unrelated donor BM in 180 (38.1%), and unrelated donor umbilical cord blood (CB) in 77 (16.3%). Conditioning regimen was myeloablative (MAC) in 20.7% and reduced intensity (RIC) in 79.3%, respectively. Patients undergoing MAC were younger than those with RIC (45 vs 50 yo, P

Description: Epstein-Barr virus (EBV)–DNA was prospectively analyzed in plasma and mononuclear cells (MNCs) from peripheral blood in patients with extranodal natural killer (NK)/T-cell lymphoma, nasal type, to evaluate the clinical significance for diagnosis, monitoring the tumor burden, and prognostication. Thirty-three patients were enrolled, and 32 were evaluable. Pretreatment plasma and MNC EBV-DNA was detectable in 14 (range, 50-71 000 copies/mL) and 6 patients (range, 20-780 copies/μg DNA), respectively, and both were well correlated (r = 0.8741, P 〈 .0001). Detectable plasma EBV-DNA was associated with higher clinical stage (P = .02), presence of B symptoms (P = .02), worse performance status (P = .02), and higher serum soluble IL-2 receptor level (P 〈 .0001). Twenty-two patients attained complete response. Plasma EBV-DNA level was significantly higher in nonresponders than in responders (mean, 16 472 vs 2 645 copies/mL; P = .02). Multivariate analysis showed clinical stage (hazard ratio, 9.0; 95% confidence interval, 1.8%-45.0%) and pretreatment plasma EBV-DNA (hazard ratio, 10.6; 95% confidence interval, 1.3%-87.0%) were significant prognostic factors. Three-year overall survival of plasma EBV-DNA positive and negative patients was 42.9% and 94.4%, respectively (P = .0009). Plasma was a preferable sample for this purpose in NK/T-cell lymphoma, nasal type, and EBV-DNA level was a good indicator for response and overall survival.