This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Niitsu, N. Articles by Hirano, M. Search for Related Content PubMed PubMed Citation Articles by Niitsu, N. Articles by Hirano, M.

Clinical Significance of Intracytoplasmic nm23-H1 Expression in Diffuse Large B-Cell Lymphoma

The Adult Lymphoma Treatment Study Group (ALTSG). ¹ Department of Hematology and Internal Medicine IV, Kitasato University School of Medicine, Kanagawa; ² Department of Clinical and Laboratory Medicine, Wakayama Medical University, Wakayama; ³ Department of Internal Medicine, Fujita Health University School of Medicine, Toyoake; and ⁴ Saitama Cancer Center Research Institute, Saitama, Japan

	ABSTRACT

Top ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Purpose: Recently, we established an ELISA technique for measuring nm23-H1 protein in serum and found that the serum nm23-H1 level is a potential prognostic factor for patients with non-Hodgkin’s lymphoma.

Experimental Design: We used immunohistochemistry to examine the expression of nm23-H1 by the lymphoma cells in patients with diffuse large B-cell lymphoma (DLBCL).

Results: By analyzing a consecutive series of 172 untreated DLBCL patients, we found that 100 (58.1%) were strongly positive. The cytoplasmic nm23 expression in lymphoma cells correlated significantly with the serum nm23-H1 level. There was a significant correlation between patients with cytoplasmic nm23-positive lymphoma and those with performance status 2–4, stage III/IV, bulky mass, B symptoms, elevated serum level of soluble interleukin 2 receptor, and elevated serum level of C-reactive protein. Overall and progression-free survival rates were significantly lower in patients with nm23-H1-positive lymphomas than in those with nm23-H1-negative lymphomas. Similar difference was seen between patients with high and low serum levels of nm23-H1. Thus, the correlation between presence or absence of cytoplasmic nm23-H1 expression and serum nm23-H1 levels suggests that serum nm23-H1 is produced directly by lymphoma cells.

Conclusion: We suggest that nm23-H1 expression is a prognostic factor for DLBCL, and that it is as important as serum nm23-H1, both of which are useful for planning a treatment strategy.

	INTRODUCTION

Top ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Diffuse large B-cell lymphoma (DLBCL) accounts for approximately 35–40% of non-Hodgkin’s lymphomas. Short-term cyclophosphamide-doxorubicin-vincristine-prednisolone chemotherapy coupled with localized radiotherapy and cyclophosphamide-doxorubicin-vincristine-prednisolone chemotherapy are considered standard treatment modalities for patients with localized disease and for those with advanced disease, respectively (1) . The treatment for DLBCL has undergone a remarkable change because of the development of anti-CD20 monoclonal antibody, as used in rituximab therapy (2) , and chemotherapy combined with rituximab may be in the mainstream of DLBCL treatment in the near future. It is important to predict the prognosis before initiating a selected treatment, such as high-dose chemotherapy combined with peripheral blood stem cell transplantation after rituximab-combined chemotherapy. International Prognostic Index (IPI) is generally used for this purpose (3) , because it can distinguish low-risk patients, to whom standard treatments can be applied, from high-risk patients, for whom such treatments may not be very effective. In addition, the IPI is based on simplified clinical characteristics, thereby allowing the collection of data from all patients before treatment. Thus, by using the IPI, it has become possible to predict a fairly accurate prognosis and, based on that prognosis, to make a stratified treatment plan appropriate for each patient.

Recently, DLBCL was categorized into two groups according to the pattern of gene expression. Prognosis was poorer in a group of patients with activated-B-cell-like (ABC) DLBCL than that with germinal-center-B-like (GCB) DLBCL, and this classification had a prognostic value even within a group of IPI-low-risk patients (4) . It is speculated that prognostication of patients with lymphoma would be based on such molecular characterization of neoplastic cells in the future, but it has, thus far, been difficult to obtain the data of gene expression before starting treatment. Also, because of aspects of cost and sample collection, it will take more time for such molecular characterization to be part of a routine pretreatment examination.

Previously, we established an ELISA technique to determine the serum level of nm23-H1 protein (5) . Originally, nm23-H1 was identified as a protein that was expressed at a lower level in metastatic cancer cells (6) . Since then, low expression or mutation of nm23-H1 and nm23-H2 has been implicated in cancer prognosis or metastasis in a variety of tumors and malignant transformations. nm23 genes play critical roles in cellular proliferation, differentiation, oncogenesis, and tumor metastasis (7 , 8) . The mechanisms responsible for these pleiomorphic effects are not well understood. We previously reported that serum levels of nm23-H1 in aggressive non-Hodgkin’s lymphomas and acute myelogenous leukemia were significantly higher than those in controls, and that high nm23-H1 levels correlated with poor prognosis in both aggressive non-Hodgkin’s lymphomas and acute myelogenous leukemia (5 , 6) . In a multi-institutional study involving a large number of patients, we found that serum nm23-H1 was an important and independent prognostic factor for patients with DLBCL and with peripheral T-cell lymphoma (9) . Although there were still several unknowns, including the question of whether serum nm23-H1 is produced by lymphoma cells directly, and whether there is a correlation between serum nm23-H1 and intracellular nm23-H1 expression by the tumor cells, our subsequent studies (10 , 11) indicated that these factors are possible. In the present study, we, therefore, evaluated intracellular nm23-H1 protein expression using immunohistochemistry.

	PATIENTS AND METHODS

Top ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patients.
We collected consecutive and untreated patients with B-cell lymphoma (n = 244) and Hodgkin’s lymphoma (HL; n = 18), who were managed by the Adult Lymphoma Treatment Study Group in Japan from 1997 to 2002. Twelve patients in our previous report on the serum nm23-H1 levels (9) were included in the present study. The lymphomas were classified according to the WHO classification as defined elsewhere (12) . Of the 244 patients with B-cell neoplasms, 15 had mucosa-associated lymphoid tissue (MALT) lymphoma, 18 had mantel cell lymphoma (MCL), 4 had B-cell chronic lymphocytic leukemia (CLL), 29 had follicular lymphoma (FL; including 3 with grade 1, 14 with grade 2, and 12 with grade 3), 172 had DLBCL, and 6 had lymphoblastic lymphoma (LBL). Of the 18 patients with HL, 8 had mixed cellularity HL and 10 had nodular sclerosis HL. The median age of the patients was 64 years (range, 18–90 years). Clinical staging was performed according to the Ann Arbor classification system (13) . The patients with DLBCL were classified according to the IPI, and the risk was low in 35 patients, low-intermediate in 44, high-intermediate in 50, and high in 43. The patients with age equal to or more than 70 years or those having stage I disease with normal serum lactate dehydrogenase levels, irrespective of their age, (n = 11) were treated with the standard cyclophosphamide-doxorubicin-vincristine-prednisolone regimen. Among other patients, those who were enrolled earlier (in 1997 and 1998; n = 44) were treated with CAMBO-VIP (cyclophosphamide, doxorubicin, methotrexate, bleomycin, vincristine, etoposide, ifosfamide, and prednisolone; Ref. 14 ), whereas those enrolled later (from 1999 to 2002; n = 117) were treated with CyclOBEAP (cyclophosphamide, vincristine, bleomycin, etoposide, doxorubicin, and prednisolone; Ref. 15 ). There were no significant differences in the background characteristics of the patients between the latter two groups. Most of the patients with indolent lymphoma were treated with a single agent or a combination of cyclophosphamide, vincristine, and prednisone (COP). Patients were followed at intervals of several months. Re-evaluation of clinical records included physical findings, complete blood counts, biochemistry data, and computed tomograms of the chest, abdominal, and pelvic areas. The median follow-up time was 34 months (range, 8 to 62 months). All of the patients gave informed consent for both treatment and sample collection in accordance with institutional policy.

Immunohistochemical Staining.
The expression of nm23 by lymphoma cells was examined immunohistochemically on formalin-fixed, paraffin-embedded sections by the avidin-biotin-peroxidase complex method. Two monoclonal antibodies to nm23-H1 [clone 37.6 (Novocastra Laboratories, Ltd, Newcastle upon Tyne, United Kingdom); clone H1–229 (Seikagaku Corporation, Tokyo, Japan)] and a monoclonal antibody to nm23-H2 (clone H2–206; Seikagaku Corporation) were used as primary reagents. Working dilutions of these antibodies and tissue distribution of positive cells were checked by using lymph nodes with reactive hyperplasia. Consequently, clone 37.6 among the above antibodies gave satisfactory staining at a range of 1:50 to 1:200 dilutions. Weak cytoplasmic staining was also seen in sinus histiocytes and endothelial cells. Because the two antibodies (clone H1–229 and clone H2–206) showed faint staining even at a 1:50 dilution by our method, clone 37.6 was used as the antihuman nm23 antibody for the study, and it was used at a 1:100 dilution. A given lymphoma cell that showed distinct immunostaining was judged to be positive for nm23-H1. The nm23-H1 expression was analyzed semiquantitatively according to the proportion of positively stained lymphoma cells. Two hundred lymphoma cells were analyzed. The degree of nm23-H1 positivity was categorized as "weak/none" if 30% or less of the lymphoma cells were positively stained; "moderate" if the proportion of the positively stained cells were in a range between 31 and 70%; or "strong" if more than 70% of the cells were positively stained (16) . For univariate and multivariate analyses, staining of 31% or more of the tumor cells was defined as positive, and staining of 30% or less of the tumor cells was defined as negative.

ELISA for Human nm23-H1.
We previously established an ELISA procedure to determine nm23-H1 protein levels in the serum (5) . Briefly, 96-well plates (Corning Co., Corning, NY) were coated with a monoclonal anti-nm23-H1 antibody (Seikagakukougyo Co., Tokyo, Japan), washed four times with PBS, and incubated with 25% Block Ace solution (Dainihon Seiyaku, Osaka, Japan). Serum samples were diluted 2-fold with PBS, and 50-µl aliquots were added to the wells. After incubation at room temperature for 1 h, the wells were washed four times with PBS containing 0.05% Tween 20 (T-PBS). Samples were then incubated at room temperature for 1 h with a polyclonal rabbit anti-nm23-H1 antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA), washed four times with T-PBS, and incubated with alkaline phosphatase-conjugated antirabbit IgG (Bio-Rad Lab, Richmond, CA). After four washes with T-PBS, alkaline phosphatase activity was detected with diethanolamine as a substrate and an alkaline phosphatase-detecting kit (Bio-Rad Lab). The reaction was stopped with 50 µl of 0.4 N NaOH. Absorbance was measured at 405–415 nm with a correction wavelength of 620–630 nm using a microplate reader.

Statistical Analyses.
The patients’ characteristics were compared using ² tests. Complete remission was defined as the absence of detectable disease based on clinical, radiological, and histological criteria. Judgment criteria used for the analysis were overall survival (OS) and progression-free survival (PFS). Progression was defined as either a progression of the lymphoma in nonresponding patients or partial response patients, a relapse for complete response patients, or death from any cause without progression. PFS was calculated as the duration from randomization to the date of the first progression. Patients who did not experience a progression at the time of analysis were censored at the most recent date of disease assessment or at the stop date if the most recent date was later. OS was measured from the date of randomization to the date of death, regardless of cause of the death. Patients who were still living at the time of analysis were censored at the most recent date of disease assessment or at the stop date if the most recent date was later. Survival analysis was performed according to the Kaplan-Meier method (17) . The statistical significance of the differences in survival was determined by the log-rank and generalized Wilcoxon’s tests (18) . Differences between groups were evaluated by the Mann-Whitney U test (nonparametric analysis; Ref. 19 ), and P < 0.05 was taken to indicate significance. Multivariate analysis of the prognosis was performed using Cox’s proportional-hazards regression model (20) . All of the statistical analyses were performed with SAS software (version 6.10, SAS Institute, Cary, NC).

	RESULTS

Top ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Studies of Cytoplasmic nm23-H1 Protein Expression in Each Histopathological Type.
As has been reported (21) , the reaction products for nm23-H1 were found primarily in the cytoplasm of centroblasts in the germinal centers, transformed lymphocytes in the interfollicular areas, and plasma cells in the medullary cords of a lymph node with reactive hyperplasia (Fig. 1A) . Weak cytoplasmic staining was also seen in sinus histiocytes and endothelial cells (Fig. 1B) . Positively stained examples of HL, FL, MCL, and DLBCL are shown in Fig. 1, C–F . The staining was primarily cytoplasmic in these lymphoma cells. Of the 262 B-cell lymphoma and HL, 150 (57.3%) were strongly positive for nm23-H1 (Table 1) . According to histopathological type, 75% of HL (mixed cellularity), 90% of HL (nodular sclerosis), 33.3% of MALT, 66.7% of MCL, 25% of CLL, 41.4% of FL (none of grade 1, 35.7% of grade 2, 58.3% of grade 3), 58.1% of DLBCL, and none of LBL were nm23-H1-strongly positive. The strongly positive rate of nm23-H1 was higher in HL, MCL, and DLBCL, and lower in MALT, CLL, and LBL. However, there was no statistically significant difference. The strongly positive rate of nm23-H1 was significantly higher in HL than in LBL (P = 0.035). In FL, the rate of nm23-H1 expression increased as the histological grade increased from 1 to 3, but the difference was not statistically significant. Moderately positive staining (i.e., positive staining in 31–70% of lymphoma cells) was observed in 40 (16.8%) of the 238 mature B-cell neoplasms and in 36 (20.9%) of the 172 DLBCLs.

View larger version (187K): [in this window] [in a new window]   Fig. 1. nm23-H1 immunostaining. In a reactive lymph node, centroblasts in the germinal centers, transformed lymphocytes in the interfollicular areas, and plasma cells in the medullary cords were positively stained (A). At a higher magnification, weak cytoplasmic staining was also seen in sinus histiocytes and endothelial cells (B). The neoplastic cells of Hodgkin’s lymphoma (C), follicular lymphoma (D), mantle-cell lymphoma (E), and diffuse large B-cell lymphoma (F) showed cytoplasmic staining. Immunoperoxidase stain with hematoxylin counterstain. A, x10; B and F, x50; C, x100; D and E, x25.

Expression of nm23-H1 in Relation to Extranodal Sites of DLBCL.
The nm23-H1 expression in DLBCL was evaluated according to the primary site of involvement (Table 2) . Fifty-five (60.4%) of 91 patients with nodal lymphoma and 45 (55.6%) of 81 patients with extranodal lymphoma were nm23-H1-strongly positive, but there was no statistically significant difference between these groups. Seventeen (81%) of 21 patients with primary tonsillar lymphoma and 28 (46.7%) of 60 patients with extranodal lymphomas other than tonsillar lymphoma were nm23-H1-strongly positive. The nm23-H1-strongly positive rate was significantly higher in primary tonsillar lymphoma than in primary nodal lymphoma (P = 0.03) and in extranodal lymphoma other than tonsil (P = 0.003).

View larger version (30K): [in this window] [in a new window]   Fig. 2. Relationship between the serum nm23-H1 level and cytoplasmic nm23-H1 expression in patients with diffuse large-cell lymphoma.

View larger version (29K): [in this window] [in a new window]   Fig. 4. Overall survival curves (A and C) and progression-free survival curves (B and D) of patients with diffuse large B-cell lymphoma according to serum nm23-H1 level. The patients were divided into high and low serum nm23-H1 levels by two cutoff values, i.e., 80 ng/ml (A, B) and 25 ng/ml (C, D). In both cutoff values, high nm23-H1 patients had shorter overall and progression-free survival periods than low nm23-H1 patients.

	DISCUSSION

Top ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

Recently, therapeutic effects for DLBCL have been improved, at least partly because the choices of treatment have broadened to include high-dose chemotherapy combined with peripheral blood stem cell transplantation, allogeneic bone marrow transplantation, mini-transplantation, and antibody treatment with rituximab or ibritumomab. However, it is important to predict prognosis and make a stratified treatment plan that attains a long remission period with the initial treatment. In other words, it is important to choose a treatment that is adequate without causing secondary cancers, infertility, or organ failure such as cardiopulmonary toxicity in the long survivors, and to choose chemotherapy combined with a molecular targeting drug. The treatment strategy for DLBCL at present is determined generally by the stage and IPI. By using IPI, a fairly accurate prognostication could be made, and, hence, it would be possible to make a stratified treatment plan for each patient. However, IPI is a marker based primarily on clinical findings. In this study, a group of patients with cytoplasmic nm23-H1-positive DLBCL had significantly shorter OS and PFS than did the nm23-H1-negative group. Multivariate analysis also suggested that cytoplasmic nm23-H1, like serum nm23-H1, may be a significant prognostic factor for DLBCL. However, multivariate analysis indicated that there was a higher relative risk associated with elevated serum nm23-H1 than with positive cytoplasmic nm23-H1, suggesting that serum nm23-H1 may be a more useful prognostic factor. In the study of acute myelogenous leukemia (6) , serum nm23-H1 levels were correlated with intracellular nm23 mRNA level and the WBC counts, but not with the mRNA level alone. In the background of this correlation, there seems to be a participation of other biological factors, such as cell proliferation and escape from apoptosis. In lymphoma, a portion of serum nm23 appears to reflect the nm23 expression level of each lymph node or lymphoid tissue, because the serum nm23 level shows a correlation with cytoplasmic nm23 or cell surface nm23. It is suggested by its correlation with advanced stage disease and presence of a bulky mass that cytoplasmic nm23 expression is associated with tumor burden. Furthermore, cytoplasmic nm23 expression or the resultant serum nm23 may stimulate certain substances that cause fever and worsen the general condition of the patients, because cytoplasmic nm23 expression was correlated with poor performance status and presence of B symptoms. It may be associated with treatment resistance and escape from apoptosis, because it is correlated with treatment responsiveness and prognosis.

It is generally considered that nm23-H1 is present in the nucleus (about 15%), in the cytoplasm (about 65%), and in the cell membrane (about 20%), and the cytoplasmic nm23-H1 expression was immunohistochemically examined in the present study. Our results suggest that some parts of the serum nm23-H1 molecules were derived from the cytoplasm and the cell membrane of lymphoma cells by secretion and cleavage, and that other parts of it were derived from the background immune cells, such as T cells and/or natural killer cells. Therefore, it can be expected that the serum nm23-H1 level is still higher in lymphoma patients than in normal volunteers even if the patients’ lymphomas are immunohistochemically negative for cytoplasmic nm23.

In summary, serum nm23-H1 is a rather stable protein that is easily and rapidly measurable using only a small serum volume before initiating treatment. Cytoplasmic nm23-H1 can be more easily examined at the time of conventional phenotypic examinations for diagnosis of lymphoma, but quantitative evaluation of immunohistochemical staining is difficult. Thus, the methods to examine serum nm23-H1 levels and cytoplasmic nm23-H1 expression have their own merits and demerits. In the future, we hope that more useful prognostic factors can be defined by studying more cases. It is also important to study nm23-H1 in lymphomas for further analysis of its functions, because there are still many questions with regard to the functions of nm23-H1.

View larger version (15K): [in this window] [in a new window]   Fig. 3. Overall survival curve (A) and progression-free survival curve (B) of patients with diffuse large B-cell lymphoma according to the expression of cytoplasmic nm23-H1.

	FOOTNOTES

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Requests for reprints: Nozomi Niitsu, Department of Internal Medicine IV, Kitasato University School of Medicine, 1-15-1 Kitasato, Sagamihara-shi, Kanagawa, 228-8555, Japan. Phone: 81-42-778-8111; Fax: 81-42-778-9978; E-mail: nniitsu{at}med.kitasato-u.ac.jp

Received 9/ 2/03; revised 12/14/03; accepted 12/31/03.

	REFERENCES

Top ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Miller TP, Dahlberg S, Cassady JR, et al Chemotherapy alone compared with chemotherapy plus radiotherapy for localized intermediate- and high-grade non-Hodgkin’s lymphoma. N Engl J Med, 339: 21-6, 1998.[Abstract/Free Full Text]
2. Coiffier B, Lepage E, Briere J, et al CHOP chemotherapy plus rituximab compared with CHOP alone in elderly patients with diffuse large-B-cell lymphoma. N Engl J Med, 346: 235-42, 2002.[Abstract/Free Full Text]
3. Shipp MA. A predictive model for aggressive non-Hodgkin’s lymphomas. The International Non-Hodgkin’s Lymphoma Prognostic Factors Project. N Engl J Med, 329: 987-94, 1993.[Abstract/Free Full Text]
4. Alizadeh AA, Eisen MB, Davis RE, et al Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature (Lond), 403: 503-11, 2000.[CrossRef][Medline]
5. Niitsu N, Okabe-Kado J, Kasukabe T, Yamamoto-Yamaguchi Y, Umeda M, Honma Y. Prognostic implications of the differentiation inhibitory factor nm23–H1 protein in the plasma of aggressive non-Hodgkin’s lymphoma. Blood, 94: 3541-50, 1999.[Abstract/Free Full Text]
6. Niitsu N, Okabe-Kado J, Nakayama M, et al Plasma levels of the differentiation inhibitory factor nm23–H1 protein and their clinical implications in acute myelogenous leukemia. Blood, 96: 1080-6, 2000.[Abstract/Free Full Text]
7. Steeg PS, Bevilacqua G, Kopper L, et al Evidence for a novel gene associated with low tumor metastatic potential. J Natl Cancer Inst (Bethesda), 80: 200-4, 1988.[Abstract/Free Full Text]
8. Lascu I, Gonin P. The catalytic mechanism of nucleoside diphosphate kinases. J Bioenerg Biomembr, 32: 237-46, 2000.[CrossRef][Medline]
9. Niitsu N, Okabe-Kado J, Okamoto M, et al Serum nm23–H1 protein as a prognostic factor in aggressive non-Hodgkin’s lymphoma. Blood, 97: 1202-10, 2001.[Abstract/Free Full Text]
10. Okabe-Kado J. Serum nm23–H1 protein as a prognostic factor in hematological malignancies. Leuk Lymphoma, 43: 859-67, 2002.[CrossRef][Medline]
11. Niitsu N, Honma Y, Iijima K, et al Clinical significance of nm23–H1 proteins expressed on cell surface in non-Hodgkin’s lymphoma. Leukemia (Baltimore), 17: 196-202, 2003.
12. Jaffe ES Harris NL Stein H Vardiman JW eds. . WHO Health Organization Classification of Tumours. Pathology and Genetics. Tumours of Haematopoietic and Lymphoid Tissues, IARC Press Lyon 2001.
13. Carbone PP, Kaplan HS, Musshoff K, Smithers DW, Tubiana M. Report of the Committee on Hodgkin’s Disease Staging Classification. Cancer Res, 31: 1860-1, 1971.[Free Full Text]
14. Hirano M, Okamoto M, Maruyama F, et al Alternating non-cross-resistant chemotherapy for non-Hodgkin’s lymphoma of intermediate-grade and high-grade malignancy. A pilot study. Cancer (Phila), 69: 772-7, 1992.
15. Niitsu N, Okamoto M, Kuraishi Y, Nakamura S, Kodama F, Hirano M. CyclOBEAP (cyclophosphamide, vincristine, bleomycin, etoposide, doxorubicin, prednisolone) regimen with granulocyte colony-stimulating factor (G-CSF) for patients with aggressive non-Hodgkin’s lymphoma: a pilot study. The Adult Lymphoma Treatment Study Group. Eur J Haematol, 65: 188-94, 2000.[CrossRef][Medline]
16. Dursun A, Akyurek N, Gunel N, Yamac D. Prognostic implication of nm23–H1 expression in colorectal carcinomas. Pathology, 34: 427-32, 2002.[CrossRef][Medline]
17. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc, 53: 457-81, 1958.[CrossRef]
18. Wilcoxon F. Individual comparisons by ranking methods. Biometrics Bull, 1: 80-3, 1945.[CrossRef]
19. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep, 50: 163-70, 1966.[Medline]
20. Cox DR. Regression models and life tables. J Stat Soc Br, 34: 187-202, 1972.
21. Aryee DN, Simonitsch I, Mosberger I, et al Variability of nm23–H1/NDPK-A expression in human lymphomas and its relation to tumour aggressiveness. Br J Cancer, 74: 1693-8, 1996.[Medline]

This article has been cited by other articles:

				N. Niitsu, H. Nakamine, M. Okamoto, J.-i. Tamaru, and M. Hirano A clinicopathological study of nm23-H1 expression in classical Hodgkin's lymphoma Ann. Onc., July 22, 2008; (2008) mdn413v1. [Abstract] [Full Text] [PDF]

				N Niitsu, M Okamoto, N Nakamura, H Nakamine, M Bessho, and M Hirano Clinicopathologic correlations of stage IE/IIE primary thyroid diffuse large B-cell lymphoma Ann. Onc., July 1, 2007; 18(7): 1203 - 1208. [Abstract] [Full Text] [PDF]

				I. S. Lossos and D. Morgensztern Prognostic Biomarkers in Diffuse Large B-Cell Lymphoma J. Clin. Oncol., February 20, 2006; 24(6): 995 - 1007. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Niitsu, N. Articles by Hirano, M. Search for Related Content PubMed PubMed Citation Articles by Niitsu, N. Articles by Hirano, M.