- Top of page
- Materials and Methods
- Disclosure Statement
This study aimed to clarify the clinicopathological prognostic parameters of de novo diffuse large B-cell lymphoma (DLBCL) in the rituximab era. We examined the correlation of 22 clinicopathological parameters with progression-free survival (PFS), overall survival (OS), and primary refractory disease in 285 DLBCL patients treated with rituximab-containing chemotherapy. Complete response rate was 87%, overall response rate was 91%, 5-year PFS rate was 72%, and 5-year OS rate was 91%. By log–rank test, higher International Prognostic Index (IPI) (P < 0.0001), Bcl-2 positivity (P = 0.0013), Bcl-6 negativity (P = 0.0112), and no irradiation (P = 0.0371) were significantly correlated with shorter PFS; higher IPI (P = 0.0107), starry sky pattern (P = 0.0466), and no irradiation (P = 0.0264) correlated with shorter OS. In multivariate analyses, higher IPI (P = 0.0006), Bcl-2 positivity (P = 0.0015), and Bcl-6 negativity (P = 0.04) were significantly correlated with shorter PFS; higher IPI (P = 0.0045) correlated with shorter OS. Bcl-2 (P = 0.0029), Bcl-6 (P = 0.002), and IPI (P < 0.0001) were significantly correlated with primary refractory disease. In conclusion, Bcl-2 positivity, Bcl-6 negativity, and higher IPI were indicators of shorter PFS and OS plus primary refractory disease in patients with DLBCL in the rituximab era.
Diffuse large B-cell lymphoma (DLBCL) represents the largest and most widely heterogeneous category of aggressive non-Hodgkin lymphomas. Several histopathological prognostic parameters of DLBCL have been reported. In the World Health Organization (WHO) Classification of Tumours of Haematopoietic and Lymphoid Tissues published in 2008, several DLBCL variants, subgroups, and subtypes were proposed. Morphological variants include centroblastic, immunoblastic, and anaplastic variants. Subtypes of DLBCL include T cell/histiocyte-rich large B-cell lymphoma and Epstein–Barr virus (EBV)-positive DLBCL of the elderly. Among these morphological variants and subtypes, immunoblastic variant DLBCL,[2, 3] T cell/histiocyte-rich large B-cell lymphoma, and EBV-positive DLBCL of the elderly have been reported to have poor prognoses. Immunohistochemical expression of Bcl-2[3, 6-9] and CD5 has been reported to be associated with an unfavorable prognosis; expression of Bcl-6 and CD10 are associated with a favorable prognosis. High Ki-67 index has also been reported to be a poor prognostic parameter. However, these results were obtained mainly in the pre-rituximab era. Therefore, these prognostic parameters should be re-evaluated in the rituximab era. As anti-CD30 mAb therapy was found to be effective for classical Hodgkin's lymphoma and anaplastic large cell lymphoma, the prognostic implication of CD30 expression should be examined in DLBCLs. cMYC rearrangement was reported to be a poor prognostic parameter of DLBCL in the rituximab era, and cMYC rearrangement and immunohistochemical cMyc expression were reported to be correlated. Therefore, immunohistochemical cMyc expression should also be evaluated.
Since 2000, DLBCL has been subdivided into germinal center B-cell (GCB) phenotype and non-GCB phenotype (including the activated B-cell phenotype and type 3 phenotype) using the cDNA microarray technique.[16, 17] For use in clinical practice, Hans et al. showed that a panel of immunohistochemical markers comprising CD10, Bcl-6, and MUM1 could be used on paraffin-embedded tissues to classify DLBCL into tumors with a GCB or non-GCB phenotype. The GCB phenotype showed a better outcome in the pre-rituximab era;[5, 16, 17] however, it was reported that the addition of rituximab to the cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) regimen eliminated the prognostic value of the GCB phenotype. Therefore, the use of the GCB phenotype as a prognostic marker is controversial in the rituximab era.
In the pre-rituximab era, the International Prognostic Index (IPI), which is based on clinical parameters such as age, stage, serum lactate dehydrogenase (LDH) level, performance status (PS), and extent of extranodal involvement (EN), proved to be highly valuable for the prediction of prognosis in patients with DLBCL. However, the IPI seems to have lost some of its high predictive value in the rituximab era.
The aim of this study was to clarify the clinicopathological prognostic parameters of de novo DLBCL in the rituximab era. Thirteen histopathological parameters including DLBCL morphological variant, necrosis, starry sky pattern, CD5, CD10, CD30, Bcl-2, Bcl-6, MUM1, GCB/non-GCB, cMyc, Ki-67, and EBV-encoded RNA (EBER)-1, as well as nine clinical parameters, including IPI, influencing progression-free survival (PFS) and/or overall survival (OS) were evaluated by log–rank tests and multivariate analyses. Correlation of primary refractory disease and these clinicopathological parameters was also examined.
- Top of page
- Materials and Methods
- Disclosure Statement
In this study, clinicopathological prognostic parameters of patients with de novo DLBCL in the rituximab era were examined. Histopathological parameters studied included DLBCL morphological variant, necrosis, starry sky pattern, CD5, CD10, CD30, Bcl-2, Bcl-6, MUM1, GCB/non-GCB, cMyc, Ki-67, and EBER-1. The results revealed that higher IPI, Bcl-2 positivity, and Bcl-6 negativity were significantly correlated with shorter PFS and primary refractory disease in the rituximab era.
Concerning the outcome of patients with DLBCL, Récher et al. reported that the 3-year PFS rate was 73% and the 3-year OS rate was 84%. Sehn et al. reported that the 4-year PFS rate was 70% and the 4-year OS rate was 70%. Patients in the present study might have had better outcomes than the patients in these reports: 5-year PFS rate was 72% and 5-year OS rate was 91%, which might be partly associated with the low median age (55 years) and high incidence of low/low intermediate IPI and stage I/II disease in the current study.
Bcl-2 protein, an antiapoptotic molecule, is expressed on resting B and T cells, but not on normal germinal center cells. Bcl-2 is expressed in 22–80% of DLBCLs, and Bcl-2 positivity has been reported to be associated with an unfavorable prognosis.[3, 6-9] Bcl-6 protein is expressed in B and CD4+ T cells within the germinal center, and is expressed in 47–84% of DLBCLs. The Bcl-6 rearrangement observed in 30–40% of DLBCLs leads to deregulation of Bcl-6 gene expression.[27, 28] Bcl-6 rearrangement was reported to correlate with a favorable outcome. However, the level of Bcl-6 protein expression is not correlated with the presence or absence of Bcl-6 gene rearrangement and mutation.[30, 31] Bcl-6 protein expression was reported to be associated with favorable prognosis. The addition of rituximab to treatment regimens has considerably improved the survival of patients with DLBCL and was reported to have eliminated the negative impact of Bcl-2 expression and the positive impact of Bcl-6 expression on clinical outcome.[33-35] However, in the present study, Bcl-2 positivity and Bcl-6 negativity were found to be parameters predicting a significantly shorter PFS (particularly Bcl-2 in the L/LI IPI group, and Bcl-6 in the HI/H IPI group) and primary refractory disease even in the rituximab era.
CD10 shows restricted expression in the germinal center B cells of reactive lymphoid tissue and is expressed in 30–60% of DLBCLs. MUM1 is a lymphoid-specific member of the interferon regulatory factor family of transcription factors. Normally expressed in plasma cells and a minor subset of germinal center B cells, MUM1 has been reported to be expressed in 35–65% of DLBCLs. CD10 has been reported to be a favorable prognostic parameter. CD10, Bcl-6, and MUM1 are included in the panel of markers used to assess GCB or non-GCB phenotype. Some previous studies examining the difference in prognosis between patients with GCB phenotype and those with non-GCB phenotype DLBCL revealed that the former group had a more favorable prognosis.[16, 17] However, Colomo et al. found no prognostic difference between these groups; thus, this has recently become a controversial issue. In the present study, GCB versus non-GCB was not a significant prognostic factor of DLBCL and neither were CD10 or MUM1.
Ki-67 index, cMyc index, starry sky pattern, and necrosis are considered to be correlated immunohistochemical and histopathological findings, and all of them are associated with proliferation activity of tumors. High Ki-67 index and high cMyc index reflecting cMYC rearrangement were reported to be poor prognostic parameters.[12, 14] The predictive value of Ki-67 index was reported in the pre-rituximab era. In the present study, only starry sky pattern was a marginally significant predictor of OS by log–rank test; however, this result was not maintained in multivariate analysis. Our results suggested that the predictive values of these factors are limited in the rituximab era.
Our results suggested that CD5 was not a significant poor prognostic factor in the rituximab era. Yamaguchi et al. reported that CD5 was a significant poor prognostic factor of OS in the pre-rituximab era, but not in the rituximab era. In addition, expression of CD30, and EBER-1 and morphological DLBCL variant were not significant prognostic parameters in the rituximab era.
In the pre-rituximab era, IPI proved to be highly valuable in predicting the prognosis of DLBCLs; however, IPI seems to have lost some of its predictive value in the rituximab era. In the present study, conventional IPI was a significant prognostic parameter for predicting PFS, OS, and primary refractory disease. Several parameters comprising IPI, such as stage, LDH, PS, and EN, were also significant prognostic parameters predicting PFS, OS, or primary refractory disease.
In group 2, histopathology of rebiopsied material at the time of relapse revealed DLBCL in five patients and low-grade B-cell lymphoma in five patients. The latter could have represented transformed low-grade B-cell lymphoma from initial presentation. Therefore, it was speculated that approximately 50% of late relapsed DLBCLs had transformed from low-grade B-cell lymphomas.
In conclusion, our study shows that Bcl-2 positivity, Bcl-6 negativity, and higher IPI are significant indicators of shorter PFS, that IPI is a significant indicator of shorter OS, and that Bcl-2 positivity, Bcl-6 negativity, and higher IPI are indicators of primary refractory disease. Our results clarify the significant clinicopathological prognostic parameters of DLBCL in the rituximab era.