the Division of Hematology-Oncology, Department of Medicine, Department of Pathology, Samsung Medical Center, Sungkyunkwan University School of Medicine
    Division of Hematology-Oncology, Department of Diagnostic Pathology, Asan Medical Center, University of Ulsan College of Medicine
    Department of Hematology-Oncology, Korea Cancer Center Hospital
    Department of Hematology-Oncology, Dankook University School of Medicine
    Department of Internal Medicine, College of Medicine, Hallym University, Seoul, Korea
    Department of Hematology-Oncology, Gachun Medical School Gil Medical Center, Incheon
    Research Institute & Hospital, National Cancer Center, Goyang, Gyeonggi
    Dong-A Cancer Center, Dong-A University College of Medicine, Busan
    Department of Internal Medicine, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul Department of Medicine, Cheju National University College of Medicine, Jeju, Korea

    ABSTRACT

    PURPOSE: Patients with natural killer T (NK/T) -cell lymphomas have poor survival outcome, and for this condition there is no optimal therapy. The purpose of this study was to design a prognostic model specifically for extranodal NK/T-cell lymphoma, which can identify high-risk patients who need more aggressive therapy.

    PATIENTS AND METHODS: This multicenter retrospective study was comprised of 262 patients who were diagnosed with NK/T-cell lymphoma.

    RESULTS: After a median follow-up duration of 51.2 months, 5-year overall survival rate in 262 patients was 49.5%. Prognostic factors for survival were "B" symptoms (P = .0003; relative risk, 2.202; 95% CI, 1.446 to 3.353), stage (P = .0006; relative risk, 2.366; 95% CI, 1.462 to 3.828), lactate dehydrogenase (LDH) level (P = .0005; relative risk, 2.278; 95% CI, 1.442 to 3.598), and regional lymph nodes (P = .0044; relative risk, 1.546; 95% CI, 1.009 to 2.367). Of 262 patients, 219 had complete information on four parameters. We identified four different risk groups: group 1, no adverse factor; group 2, one factor; group 3, two factors; and group 4, three or four factors. The new model showed a superior prognostic discrimination as compared with the International Prognostic Index (IPI). Notably, the distribution of patients was balanced when a new model was adopted (group 1, 27%; group 2, 31%; group 3, 20%; group 4, 22%), whereas 81% of patients were categorized as low or low-intermediate risks using IPI.

    CONCLUSION: The newly proposed model for extranodal NK/T-cell lymphoma demonstrated a more balanced distribution of patients into four groups with better prognostic discrimination as compared with the IPI.

    INTRODUCTION

    Extranodal natural killer T (NK/T) -cell lymphoma, nasal-type, is a recently recognized distinct entity within the WHO classification of lymphoid tumors.1 Not common in Western countries, extranodal NK/T-cell lymphoma is more common in Asia2-9 accounting for 9% of all malignant lymphomas and 74% of lymphomas arising within nasal cavity and paranasal sinuses in Korea.10 Whereas most of sinonasal lymphomas are diffuse large B-cell lymphomas in Western population, 40% to 74% of sinonasal lymphomas are NK/T-cell lymphomas in Asia.10,11 Au et al recently reported that the high incidence of mature T-cell and natural killer (NK) -cell lymphomas in China is due to a high frequency of NK-cell, but not T-cell lymphomas.11 NK/T-cell lymphomas usually occur in middle-age patients, and their presenting features are characterized by a localized disease in about two thirds of the patients, frequent adjacent tissue invasion, a high frequency of "B" symptoms despite apparently limited disease, and rare bone marrow (BM) involvement.12-14 They are also characterized by predominantly involving extranodal sites rendering the prefix extranodal NK/T-cell lymphoma in the WHO classification.

    Because extranodal NK/T-cell lymphoma, nasal type is a relatively, newly recognized distinctive clinicopathologic entity in the WHO classification,1,14,15 optimal treatment strategies and prognoses have not been fully defined yet. At a consensus meeting in 1997, the importance of clinical features and location of the primary lesion was emphasized in determining the biologic behavior and definition of the disease.16 We, and a few other groups, have demonstrated that patients with NK/T-cell lymphomas have poor survival outcome, with the cumulative probability of survival at 5 years ranging from 37.9% to 45.3%.13,14,17 From radiotherapy alone to high-dose therapy with stem cell transplantation, several treatment options have been proposed for patients with extranodal nasal-type NK/T-cell lymphoma. Although validated in many lymphomas ranging from low to high grades, the prognostic impact of the International Prognostic Index (IPI) has been controversial in NK/T-cell lymphomas, with one previous study yielding positive correlations14 and four yielding negative correlations.17-20 When using the IPI, a small proportion (0% to 7%) of NK/T-cell lymphoma patients is categorized as high-risk group.12-14,18,20 Moreover, we have recently reported that two subtypes, namely upper aerodigestive tract NK/T-cell lymphoma (UNKTL) and extra-upper aerodigestive tract NK/T-cell lymphoma (EUNKTL), behave significantly different in regards to survival and prognosis.13,21 In retrospective analyses of a series of NK/T-cell lymphomas, several variables were associated with poor survival, such as regional lymph node involvement,20 elevated lactate dehydrogenase (LDH),13,20 poor performance,13 paranasal extension,20,22-24 the presence of B symptoms,13,22,25 and high Epstein-Barr virus (EBV) DNA titer.26 To design a prognostic model specifically for extranodal NK/T-cell lymphoma, we started a large, retrospective, multicenter study.

    PATIENTS AND METHODS

    Patients

    The criteria for case inclusion were as follows: (1) pathologically confirmed diagnosis of NK/T-cell lymphoma, according to the WHO classification; and (2) proven NK/T cell type by immunohistochemistry, flow cytometry, or EBV in situ hybridization. If EBV in situ hybridization was negative, patients were excluded from the analysis. Blastic NK-cell lymphoma/leukemia, aggressive NK-cell lymphoma/leukemia, and peripheral T-cell lymphoma, unspecified were excluded from the analysis. Two subtypes of extranodal NK/T-cell lymphoma, UNKTL and EUNKTL, were defined as previously described. Briefly, UNKTL includes all lymphomas confined to nasal cavity, nasopharynx, and the upper aerodigestive tract, whereas EUNKTL embraces lymphomas occurring at all other sites. Patients with primary lesion within the nasal cavity and secondary spread to other organs were categorized as UNKTL. (3) A complete set of clinical information was the third criterium. Patients with the following clinical data were considered eligible: patient demographics, complete blood count, Ann Arbor stage, computed tomography (CT) scan or magnetic resonance imaging (MRI) scan of the involved region, CT scan of abdomen, and BM findings. All patients provided informed consent in compliance with institutional guidelines.

    Local invasiveness was defined differently according to the two subtypes. For UNKTL, local invasiveness was defined in accordance with 2002 TNM classification of the American Joint Committee on Cancer: T1 tumor restricted to any one subsite, with or without bony invasion; T2 tumor invading two subsites in a single region or extending to involve an adjacent region within the nasoethmoidal complex, with or without bony invasion; T3 tumor extends to invade the medial wall or floor of the orbit, maxillary sinus, palate, or cribriform plate; T4 tumor invades any of the following: anterior orbital contents, skin of nose or cheek, minimal extension to anterior cranial fossa, pterygoid plates, sphenoidal or frontal sinuses, orbital apex, dura, brain, middle cranial fossa, nasopharynx, or clivus. Any UNKTL with T3 or greater were considered as local invasiveness in the analysis. For EUNKTL, the definition of local invasiveness differed according to primary sites. For gastrointestinal EUNKTL, local invasiveness referred to T4 lesion in TNM system. In EUNKTL primarily involving soft tissue, such as muscle or skin, invasion of neurovascular structure or bone invasion was considered as local invasion. Regional lymphadenopathies were defined as the invasion of lymph nodes corresponding to N1, N2, or N3 of the primary lesion at TNM staging system. Accordingly, M1 nodes at TNM system were not categorized as regional lymph nodes in the analysis. B symptoms were defined as unexplained fever with temperature above 38°C, night sweats, and unexplained weight loss of more than 10% of the usual body weight in the 6 months before diagnosis.

    1991 to 2004, 304 patients were accrued from 10 tertiary hospitals in Korea. Of 304 patients, we excluded 42 patients for the following reasons: 12 patients did not have sufficient clinical data available at the time of analysis; 13 patients showed EBV negativity for in situ hybridization; five patients had NK/T cell leukemia; four patients had blastic NK lymphoma/leukemia; three patients had peripheral T-cell lymphoma unspecified; and five patients for other reasons, such as inadequate quality of pathologic specimen for definitive diagnosis.

    Histology

    All pathologic specimens were reviewed and reclassified by central review in accordance with the WHO criteria for pathologic diagnosis.1 Immunophenotyping was performed using a panel of monoclonal antibodies, including antibodies against cytoplasmic CD3 (Dakopatts, Copenhagen, Denmark), CD20 (Dakopatts), and CD56 (Monosan, Uden, the Netherlands). EBV RNA was detected by an in situ hybridization technique. Briefly, paraffin sections were pretreated with xylene, followed by treatment with proteinases K and hybridized with fluorescein isothiocyanate-conjugated EBV oligonucleotides (Dakopatts) complementary to the nuclear RNA portion of the EBER1 and EBER2 genes.

    Treatment

    Patients received one of the following initial treatment modalities: (1) an anthracycline-containing chemotherapeutic regimen followed by radiotherapy (RT); (2) a non–anthracycline-containing chemotherapeutic regimen followed by RT; (3) anthracycline-based chemotherapy; (4) non–anthracycline-based chemotherapy; (5) involved-field RT as the primary treatment; (6) surgery alone; and (7) supportive care only. The anthracycline-based regimens used were as following: cyclophosphamide, doxorubicin, vincristine, prednisolone (CHOP); dose-escalated CHOP (deCHOP); cyclophosphamide, vincristine, prednisone, bleomycin, doxorubicin, procarbazine (COPBLAM); and (etoposide, doxorubicin, vincristine, cyclophosphamide, prednisolone (EPOCH). The non–anthracycline-containing regimens used were ifosfamide, methotrexate, etoposide (IMEP); dexamethasone, ifosfamide, cisplatin, etoposide (DICE); etoposide, methylprednisolone, cisplatin, cytarabine (ESHAP); dexamethasone, cytarabine, cisplatin (DHAP); etoposide, ifosfamide, cisplatin, dexamethasone (VIPD); and cyclophosphamide, vincristine, prednisone (CVP). In patients with localized disease, involved-field radiotherapy was given at the physician's discretion following chemotherapy. The treatment response was assessed according to standard response criteria.27

    Statistical Analysis

    Overall survival (OS) and relapse-free survival (RFS) were estimated using the Kaplan-Meier product-limit method. OS was measured from the date of diagnosis to the date of death or the last follow-up visit. RFS was calculated from the date of diagnosis to the first documented relapse in patients who attained complete remission (CR). Survival rates were compared for statistical differences by using log-rank analysis. Continuous biologic variables were dichotomized. A prognostic model was established by fitting all variables that significantly influenced OS at a level of P =  .05 in the univariate analysis. A forward stepwise Cox regression analysis28 was then performed and prognostic index was derived. P = .05 were considered statistically significant and all P values correspond to two-sided significance tests.

    RESULTS

    Patient Characteristics

    In all, 262 patients from 10 centers fulfilled the inclusion criteria and were registered for the study. The clinical characteristics of the 262 patients are outlined in Table 1. The majority of the patients were categorized as UNKTL (222 of 262, 85%). EUNTKLs (40 of 262, 15%) included primary lesions at the following sites: soft tissue (n = 10), gastrointestinal tract (n = 10), skin (n = 7), liver (n = 4), lung (n = 4), orbit (n = 2), and other (n = 3). The IPI scores were distributed as 212 of 262 patients (81%) had IPI scores between 0 to 2, and 41 patients (19%) had IPI scores between 3 or 4.

    Survival Analysis

    After a median follow-up duration of 51.2 months, an estimated 5-year OS rate in 262 patients was 49.5% (Fig 1). Forty-nine percent of the deaths occurred in the first 6 months. The initial treatment strategies were as follows: 103 (39%) patients received anthracycline-based chemotherapy followed by involved field RT; 17 (7%) patients non–anthracycline-based chemotherapy followed by RT; 99 (38%) patients anthracycline-based chemotherapy without RT; 20 patients (8%) non–anthracycline-based chemotherapy; 16 patients (6%) received radiotherapy alone; five patients (2%) had supportive care and; two patients (1%) received surgery alone. Of the 246 patients (94%) who were evaluated for treatment response, 56% achieved CR. Of the 138 patients with CR, 51 (37%) patients relapsed. The 5-year RFS among patients who achieved CR was 60%. Sixteen patients received autologous hematopoietic stem-cell transplantations. Patients with localized UNTKL showed the highest 5-year OS rate (76%) while disseminated EUNTKL had the lowest 5-year OS (0%). Survival curves of both localized and disseminated EUNTKL did not reach plateau (Fig 2).

    Univariate Analysis

    The clinical factors predicting poor survival at univariate analysis were as follows: poor performance status (Eastern Cooperative Oncology Group performance status 2 to 4, P = <.0001), advanced Ann Arbor stage (stage III/IV, P = <.0001), elevated LDH level (P = <.0001), intermediate-high to high IPI scores (P = <.0001), EUNKTL, regional lymphadenopathy (P = <.0001), local invasiveness (P = .0097), the presence of B symptoms (P = .0002), and BM involvement (P = .0106).

    Prognostic Model

    Clinical parameters that were included in the multivariate analysis were Eastern Cooperative Oncology Group performance status, stage, LDH, subtypes, regional lymphadenopathy, local invasiveness, B symptoms, and BM involvement. The forward conditional Cox regression model was used. Prognostic factors for survival were B symptoms (P = .0003; relative risk, 2.202; 95% CI, 1.446 to 3.353), stage (P = .0006; relative risk, 2.366; 95% CI, 1.462 to 3.828), LDH level (P = .0005; relative risk, 2.278; 95% CI, 1.442 to 3.598), and regional lymph nodes (P = .0044; relative risk, 1.546; 95% CI, 1.009 to 2.367). Of 262 patients, 219 patients had complete information on four parameters and thus included in the prognostication. The prognostic grouping of the 219 patients was performed according to the following criteria: group 1, no adverse factors; group 2, one factor; group 3, two factors; and group 4, three or four factors. The survival curves according to the new prognostic index for NK/T-cell lymphomas are shown in Figure 3. The new index categorized four groups with significantly different survival outcomes (P < .0001). The proportion of patients in each group and associated hazard ratios are presented in Table 2. Of 219 patients, 215 patients had complete information available for the parameters of the IPI. The IPI separated the patients into four risk groups with different survival outcomes (Fig 4). However, the IPI did not discriminate well between the low and low-intermediate risk groups (P = .0614) or between the high-intermediate and high risk groups (P = .8304). As shown in Figure 5, the new prognostic model was efficient in discriminating the patients with low to low-intermediate risk according to the IPI (P = <.0001). The model differentiated high-intermediate risk group (n = 4) from high-risk patients (n = 4) according to the IPI, with borderline significance (P = .0742).

    DISCUSSION

    To improve risk-based stratification for therapy, we attempted to establish a prognostic model specifically devised for patients with this particular subset of lymphomas. Recently, the Intergruppo Italiano Linfomi has proposed a new prognostic index for PTCLu,29 comprising age, performance status, LDH, and BM involvement, in 385 cases. The prognostic index for PTCLu model had a superior predictive capacity as compared to IPI. The Follicular Lymphoma IPI has been suggested to be more discriminant than the IPI in follicular lymphomas.30 We initially collected 304 patients from 10 institutions and excluded 42 patients, resulting in 262 patients for the analysis. Since we excluded EBV-negative NK/T-cell lymphomas as well as other mature T- and NK-cell neoplasms, our series consisted of homogeneous group of patients. In this study, an overall CR rate and 5-year OS were 56% and 49.5%, respectively. These results were similar to those of the largest study reported by Chim et al.14

    Although this subset of lymphomas is known for its poor prognosis (5-year OS rate, 49.5%, Fig 1), 81% of the patients had IPI scores less than 2 (Table 1). Furthermore, the IPI did not distinguish between low-risk and low-intermediate-risk groups with sufficient statistical power (P = .0614). This may be partly explained by peculiar clinical features of the disease such as high frequency of localized disease (76%), rare involvement of BM (6%), and high occurrence of constitutional symptoms in localized disease (29%). BM involvement is markedly low in NK/T-cell lymphomas as compared to PTCLu (20% to 40%).31 In our analysis, performance status, stage, LDH, EUNKTL subtype, regional lymphadenopathies, local invasiveness, B symptoms, and BM involvement were significant prognostic parameters for survival at univariate analysis (Table 3). However, only four factors, B symptoms, stage, LDH, and regional lymph node invasion were retained at a multivariate level (Table 4). Various treatment modalities were not included in the prognostic analysis since the aim of the model was to identify risk groups from initial clinical variables to aid in determining therapeutic approach. Notably, the distribution of patients was balanced when a new model was adopted (group 1, 27%; group 2, 31%; group 3, 20%; group 4, 22%; Table 5) .

    We have recently analyzed clinical features of UNKTL and EUNKTL and found that EUNKTL pursued more aggressive clinical course.13 In this larger series, the subtype EUNKTL was a significant adverse factor for survival (P = .0009), but was not retained in multivariate analysis. However, there was a notable difference in survival pattern according to the subtypes (Fig 2). Interestingly, UNTKL patients, regardless of the stage, showed a plateau in their survival curves. Conversely, survival curves of EUNTKL patients, localized and disseminated, continue to decline. This finding may suggest that more aggressive initial treatment followed by consolidation therapy should be pursued in EUNKTL patients. From a retrospective analysis, Au et al have recommended early consideration of allogeneic transplantation for relapsed, refractory, or extranasal disease.32

    Few groups have suggested that local invasiveness, such as paranasal extension, is an important adverse factor for survival in nasal NK/T-cell lymphomas.20,22-24 Kim et al has analyzed the prognostic significance of local tumor invasiveness in 114 stage I/II nasal NK/ T-cell lymphoma patients and demonstrated that the presence of local tumor invasiveness was a significant adverse factor for CR, disease-free survival, or OS.33 Local tumor invasiveness was defined as bony invasion and/or perforation or invasion of the skin based on CT and physical findings in their study. However, in our analysis, statistical significance of local invasiveness shown at univariate level was not maintained at multivariate analysis. Although there was a significant survival difference according to local invasiveness in patients without lymph node invasion (P = .0234, Fig 6A), the survival difference was abrogated in patients with lymph node invasion (P = .8482, Fig 6B). Therefore, invasion of regional lymph nodes may be a more powerful predictive factor for poor survival outcome than local tumor invasiveness.

    In conclusion, the newly proposed model for extranodal NK/ T-cell lymphoma demonstrated balanced distribution of patients into four groups with better prognostic discrimination as compared to IPI. It is crucial to identify high-risk patients early and accurately in order to improve survival in this subset of lymphoma patients.

    Authors' Disclosures of Potential Conflicts of Interest

    The authors indicated no potential conflicts of interest.

    Author Contributions

    Conception and design: Jeeyun Lee, Cheolwon Suh, Young H. Ko, Won Seog Kim

    Collection and assembly of data: Jeeyun Lee, Yeon Hee Park, Young H. Ko, Soo Mee Bang, Jae Hoon Lee, Dae Ho Lee, Jooryung Huh, Sung Yong Oh, Hyuk-Chan Kwon, Hyo Jin Kim, Soon Il Lee, Jung Han Kim, Jinny Park, Seok Joong Oh, Kihyun Kim, Chulwon Jung, Keunchil Park, Won Seog Kim

    Data analysis and interpretation: Jeeyun Lee, Yeon Hee Park, Soo Mee Bang, Won Seog Kim

    Manuscript writing: Jeeyun Lee, Cheolwon Suh

    Final approval of manuscript: Jeeyun Lee, Cheolwon Suh, Yeon Hee Park, Keunchil Park, Won Seog Kim

    NOTES

    J.L., C.S., and Y.H.P. contributed equally to the work presented here.

    Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.

    REFERENCES

    Jaffe ES, Harris NL, Stein H, et al: World Health Organization Classification of Tumours. Pathology and Genetics of Tumours of the Hematopoietic and Lymphoid Tissues. Lyon, Frances, IARC Press, 2001, pp 204-207

    Aozasa K, Ohsawa M, Tajima K, et al: Nationwide study of lethal mid-line granuloma in Japan: Frequencies of Wegener's granulomatosis, polymorphic reticulosis, malignant lymphoma and other related conditions. Int J Cancer 44:63-66, 1989

    Chan JKC, Ng CS, Lau WH, et al: Most nasal/nasopharyngeal lymphomas are peripheral T-cell neoplasms. Am J Surg Pathol 11:418-429, 1987

    Ferry J, Sklar J, Zukerberg L, et al: Nasal lymphoma: A clinicopathologic study with immunophenotypic and genotypic analysis. Am J Surg Pathol 15:268-779, 1991

    Harabuchi Y, Yamanaka N, Kataura A, et al: Epstein-Barr virus in nasal T-cell lymphomas in patients with lethal midline granuloma. Lancet 335:128-130, 1990

    Ho FCS, Choy D, Loke SL, et al: Polymorphic reticulosis and conventional lymphomas of the nose and upper aerodigestive tract: A clinicopathologic study of 70 cases, and immunophenotypic studies of 16 cases. Hum Pathol 21:1041-1050, 1990

    Liang R, Todd D, Chan TK, et al: Treatment outcome and prognostic factors for primary nasal lymphoma. J Clin Oncol 13:666-670, 1995

    Kim GE, Cho JH, Yang WI, et al: Angiocentric lymphoma of the head and neck: Patterns of systemic failure after radiation treatment. J Clin Oncol 18:54-63, 2000

    Siu LL, Chan JK, Kwong YL: Natural killer cell malignancies: Clinicopathologic and molecular features. Histol Histopathol 17:539-554, 2002

    Ko YH, Kim CW, Park CS, et al: REAL classification of malignant lymphomas in the Republic of Korea: Incidence of recently recognized entities and changes in clinicopathologic features. Cancer 83:806-812, 1998

    Au WY, Ma SY, Chim CS, et al: Clinicopathologic features and treatment outcome of mature T-cell and natural killer-cell lymphomas diagnosed according to the World Health Organization classification scheme: A single center experience of 10 years. Ann Oncol 16:206-214, 2005

    Kim K, Kim WS, Jung CW, et al: Clinical features of peripheral T-cell lymphomas in 78 patients diagnosed according to the Revised European-American lymphoma (REAL) classification. Eur J Cancer 38:75-81, 2002

    Lee J, Park YH, Kim WS, et al: Extranodal nasal type NK/T-cell lymphoma: Elucidating clinical prognostic factors for risk-based stratification of therapy. Eur J Cancer 41:1402-1408, 2005

    Chim CS, Ma SY, Au WY, et al: Primary nasal natural killer cell lymphoma: Long-term treatment outcome and relationship with the International Prognostic Index. Blood 103:216-221, 2004

    Jaffe ES: Classification of natural killer (NK) cell and NK-like T-cell malignancies. Blood 87:1207-1210, 1996

    Harris NL, Jaffe ES, Diebold J, et al: World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues: Report of the clinical advisory committee meeting- Airlie House, Virginia, November 1997. J Clin Oncol 17:3835-3849, 1999

    Cheung MM, Chan J, Lai W, et al: Early stage nasal NK/T-cell lymphoma: Clinical outcome, prognostic factors, and the effect of treatment modality. Int J Radiat Oncol Biol Phys 54:182-190, 2002

    Cheung MM, Chan J, Lau W, et al: Primary non-Hodgkin's lymphoma of the nose and nasopharynx: Clinical features, tumor immunophenotype, and treatment outcome in 113 patients. J Clin Oncol 16:70-77, 1998

    Aviles A, Diaz NR, Neri N, et al: Angiocentric nasal T/natural killer cell lymphoma: A single centre study of prognostic factors in 108 patients. Clin Lab Haematol 22:215-220, 2000

    You JY, Chi KH, Yang MH, et al: Radiation therapy versus chemotherapy as initial treatment for localized nasal natural killer (NK)/T-cell lymphoma: A single institute survey in Taiwan. Ann Oncol 15:618-625, 2004

    Lee J, Kim WS, Park YH, et al: Nasal-type NK/T-cell lymphoma: Clinical features and treatment outcome. Br J Cancer 92:1226-1230, 2005

    Robbins KT, Fuller LM, Vlasak M, et al: Primary lymphomas of the nasal cavity and paranasal sinuses. Cancer 56:814-819, 1985

    Logsdon MD, Ha CS, Kavadi VS, et al: Lymphoma of the nasal cavity and paranasal sinuses: Improved outcome and altered prognostic factors with combined modality therapy. Cancer 80:477-488, 1997

    Li YX, Coucke PA, Li JY, et al: Primary non-Hodgkin's lymphoma of the nasal cavity: Prognostic significance of paranasal extension and the role of radiotherapy and chemotherapy. Cancer 83:449-456, 1998

    Kim WS, Song SY, Ahn YC, et al: CHOP followed by involved field radiation: Is it optimal for localized nasal natural killer/T-cell lymphoma Ann Oncol 12:349-352, 2001

    Au WY, Pang A, Choy C, et al: Quantification of circulating Epstein-Barr virus (EBV) DNA in the diagnosis and monitoring of natural killer cell and EBV-positive lymphomas in immunocompetent patients. Blood 104:243-249, 2004

    Cheson BD, Horning SJ, Coiffier B, et al: Report of an international workshop to standardize response criteria for non-Hodgkin's lymphoma. J Clin Oncol 17:1244-1253, 1999

    Cox DR: Regression models and life-tables. J Stat Soc 34:187-220, 1982

    Gallamini A, Stelitano C, Calvi R, et al: Peripheral T-cell lymphoma unspecified (PTCL-U): A new prognostic model from a retrospective multicentric clinical study. Blood 103:2474-2479, 2004

    Solal-Céligny P, Roy P, Colombat P, et al: Follicular lymphoma international prognostic index. Blood 104:1258-1265, 2004

    Hanson CA, Brunning RD, Gajl-Peczalska KJ, et al: Bone marrow manifestation of peripheral T-cell lymphoma: A study of 30 cases. Am J Clin Pathol 86:449-460, 1986

    Au WY, Lie AKW, Liang R, et al: Autologous stem cell transplantation for nasal NK/T-cell lymphoma: A progress report on its value. Ann Oncol 14:1673-1679, 2003

    Kim TM, Park YH, Lee SY, et al: Local tumor invasiveness is more predictive of survival than international prognostic index in stage IE/IIE extranodal NK/T-cell lymphoma, nasal type. Blood 106:3785-3790, 2005

查询更多Lymphoma相关信息在本站>>

《临床肿瘤学医学期刊》2006年2月第24卷第2期