Fax: (713) 794-1800
Proliferation predicts failure-free survival in mantle cell lymphoma patients treated with rituximab plus hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone alternating with rituximab plus high-dose methotrexate and cytarabine
Article first published online: 23 JAN 2009
Copyright © 2009 American Cancer Society
Volume 115, Issue 5, pages 1041–1048, 1 March 2009
How to Cite
Garcia, M., Romaguera, J. E., Inamdar, K. V., Rassidakis, G. Z. and Medeiros, L. J. (2009), Proliferation predicts failure-free survival in mantle cell lymphoma patients treated with rituximab plus hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone alternating with rituximab plus high-dose methotrexate and cytarabine. Cancer, 115: 1041–1048. doi: 10.1002/cncr.24141
- Issue published online: 18 FEB 2009
- Article first published online: 23 JAN 2009
- Manuscript Accepted: 29 SEP 2008
- Manuscript Received: 28 JUL 2008
- mantle cell lymphoma;
It has been demonstrated that the tumor proliferation index has prognostic significance in patients with mantle cell lymphoma (MCL). Patients in most of studies, however, have been treated with relatively traditional chemotherapy regimens. At the authors' institution, patients with MCL received an aggressive chemotherapy regimen: rituximab plus hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone alternating with rituximab plus high-dose methotrexate and cytarabine (R-hyper-CVAD).
The authors assessed the proliferation rate of MCL using immunohistochemistry and an antibody specific for Ki-67 in 71 untreated patients who subsequently received R-hyper-CVAD. The study group included 59 patients who had classic MCL and 12 patients who had the blastoid variant of MCL.
For the entire study group and for the group of patients with classic MCL, a proliferation index of >20% Ki-67-positive cells was correlated significantly with shorter failure-free survival. There was no correlation between the proliferation index and overall survival.
The current results indicated that the proliferation index in patients with MCL predicted prognosis in those who uniformly received the R-hyper-CVAD chemotherapy regimen. Cancer 2009. © 2009 American Cancer Society.
Mantle cell lymphoma (MCL) is a distinct type of B-cell non-Hodgkin lymphoma that accounts for approximately 6% of all non-Hodgkin lymphomas.1 Histologically, most cases of MCL, known as the classic or common variant, are composed of a monotonous proliferation of small-to-intermediate sized, neoplastic lymphocytes with scant cytoplasm and more or less irregular nuclei. Blastoid variants of MCL also are described; these variants have aggressive histologic features and a high mitotic rate.1 In both the classic and blastoid variants of MCL, the neoplastic cells are positive for pan-B-cell antigens and CD5 and are negative for CD10, CD23, and BCL-6. The t(11;14)(q13;q32) translocation is a hallmark of MCL, juxtaposing the cyclin-D1 gene CCND1 at 11q13 with the immunoglobulin heavy chain gene (IgH) on the derivative chromosome 14, resulting in cyclin D1 over-expression. Several other molecular alterations have been identified in subsets of MCL cases, most of which dysregulate the cell cycle or impair the DNA damage-response pathway.2
It has been demonstrated in several clinical studies3-5 that the prognosis for patients with MCL is poor. The clinical course of MCL usually is aggressive, and the median survival is 3 to 5 years. In most of these studies, MCL patients were treated in a heterogeneous manner, mostly with standard chemotherapy regimens, and the most commonly used regimen was combined cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP). Attempts to predict the prognosis for patients with MCL using clinical and laboratory parameters, such us the International Prognostic Index (IPI), have been suboptimal, although a newly developed MCL IPI (MIPI) may prove to be more useful.6 The poor prognosis for patients with MCL led investigators at our institution and others to develop dose-intensified treatment approaches to treat MCL; the results are promising and suggest that dose intensification is superior to CHOP chemotherapy.7-9 At our institution, Romaguera and colleagues7 have used rituximab plus hyperfractioned cyclophosphamide, vincristine, doxorubicin, and dexamethasone alternating with rituximab plus high-dose methotrexate-cytarabine (R-hyper-CVAD) as frontline therapy. This approach has been proven as an effective cytoreductive regimen in patients with MCL, achieving an overall complete response/unconfirmed complete response rate of 87% and a 3-year failure-free survival (FFS) rate of 64% after a median follow-up of 40 months.7
By using gene expression profiling methods, Rosenwald and colleagues10 demonstrated that MCL has a unique signature. They also observed that the expression of a relatively small number of genes involved in proliferation subdivided MCL patients into 4 groups that correlated with prognosis. Others have used Ki-67 expression, assessed by immunohistochemical methods, to assess the proliferation index (PI) in MCL tumors and demonstrated that this approach has prognostic value.11-15 Most patients in those Ki-67 studies received various traditional chemotherapy regimens, most commonly CHOP. There is also a recent study in which a subset of 96 patients with MCL received R-CHOP.15 Currently, however, it is unknown whether the PI also has prognostic significance in patients with MCL who receive dose-intensified chemotherapy regimens, such us the R-hyper-CVAD developed at our institution. In the current study, we assessed the prognostic value of the PI, as assessed by Ki-67 immunohistochemical analysis, in a group of patients with MCL who uniformly received R-hyper-CVAD.
MATERIALS AND METHODS
The study group included 76 patients who were previously untreated, had a diagnosis of MCL confirmed at the M. D. Anderson Cancer Center, and had available pretreatment paraffin blocks or unstained slides for immunohistochemical analysis. At our institution, patients who have MCL with a purely mantle zone pattern are not treated with R-hyper-CVAD and, thus, were not included in this study. All patients were treated uniformly according to the protocol, as described previously.7 The diagnosis of MCL was established using the criteria of the World Health Organization classification.1 The survival analysis included 71 patients; 5 patients were excluded because of inadequate length of follow-up.
The type of biopsy specimens reviewed included 53 lymph node, 9 gastrointestinal tract (GI), 5 tonsil, 3 spleen, 1 bone marrow, 1 nasopharynx, 1 submandibular mass, 1 palatine tongue, 1 base of tongue, and 1 eyelid mass. All specimens histologically met the criteria for either classic or blastoid MCL, had a B-cell immunophenotype, and either overexpressed cyclin D1 or carried the t(11;14)(q13;q32). The latter was confirmed by either conventional cytogenetics or fluorescence in situ hybridization, as described previously.16 In most specimens, both cyclin D1 overexpression and the t(11;14) were demonstrated.
Staging and Therapy
All patients received a pretreatment staging evaluation that included a complete physical examination, chest radiograph, upper and lower GI endoscopy, bilateral bone marrow aspirations and biopsies, whole body computerized tomography scans, and gallium and/or positron emission tomography scans when indicated, as described previously.7 Patients also had blood drawn for a complete blood count with differential, serum levels of lactate dehydrogenase (LDH) and β2-microglobulin, and other routine chemistry studies. Flow cytometric immunophenotypic analysis for lymphoid cell surface markers was performed on peripheral blood and bone marrow aspirate specimens from each patient. A serum LDH level >618 IU/L (normal range, 313-618 IU/L) was considered high. A β2-microglobulin level >3.0 mg/L (normal range, 0.7-1.8 mg/L) was considered high, as determined in an earlier study.7 After signing of informed consent approved by the Institutional Review Board, all patients were enrolled into a prospective clinical trial that included administration of the R-hyper-CVAD chemotherapy regimen for 6 to 8 cycles.7
Immunohistochemical analysis was performed using previously published methods.17 Briefly, formalin-fixed, paraffin-embedded tissue sections (4 μm thick) were deparaffinized in xylene and rehydrated in graded series of alcohol solutions. A monoclonal antibody specific for Ki-67 (DAKO, Carpinteria, Calif) was used for immunostaining. Heat-induced antigen retrieval was carried out using a preheated target retrieval solution (DAKO). The PI was calculated by evaluating the entire slide stained for Ki-67, selecting a representative area, and manually counting 1000 cells. The number of Ki-67-positive cells was divided by the total cells counted and was expressed as a percentage for the PI.
Statistical analysis was performed comparing patients who had tumors with high versus low PI. The cutoffs that initially were considered were the median percentage of Ki-67-positive tumor cells for the entire study and, separately, for the group of patients who had the classic MCL variant. Four additional cutoffs for the entire study group also were assessed based on a review of histograms: 5%, 10%, 15%, and 20% Ki-67-positive tumor cells (Fig. 1). Failure-free survival (FFS) was defined as the time from the beginning of treatment to initial treatment failure or to last follow-up. Treatment failure was defined as recurrence or progression of disease, death caused by disease or toxic effects, and death caused by treatment-related second malignancies.7 Overall survival (OS) was measured from the onset of therapy to the date of last follow-up or death from any cause, whichever occurred first. The Kaplan-Meier method was used to measure actuarial survival using the log-rank test. The statistical independence of individual prognostic factors was evaluated by multivariate analysis using the Cox proportional hazards model. The PI was compared with a variety of clinical and laboratory features using the chi-square test and the Fisher exact test. For part of this analysis, we assessed both the traditional IPI and the recently proposed MIPI.6 Statistical analysis was carried out using Statview software (Abacus Concepts, Berkeley, Calif).
There were 60 men and 16 women, and the median patient age was 62 years (range, 42-82 years). The median follow-up was 51 months for the entire study group. Sixty-three tumors were identified as the classic variant of MCL, and 13 tumors were identified as blastoid variants of MCL. The patterns of growth were diffuse in 30 tumors, nodular and diffuse in 27 tumors, nodular in 18 tumors, and nodular and mantle zone in 1 tumor.
Association of the Proliferation Index With Clinical, Histologic, and Laboratory Parameters in Mantle Cell Lymphoma
Table 1 summarizes the association of the PI with various clinical, histologic, and laboratory parameters. The median PI for the entire study group was 11.4%. Within the group of patients with the classic MCL variant, the median PI was 10% (range, 1%-55%), whereas patients with the blastoid variant MCL had a median PI of 22% (range, 12%-87%; P=.0038; Mann-Whitney test). Figure 1A provides examples of the classic MCL variant with either a low or high PI. The PI was not correlated with presenting clinical or laboratory features in the entire study group (Table 1). Similarly, the PI was not correlated with the IPI (score range, 0-4) (Table 1).
|Characteristic||No. of Patients||Ki-67 Expression||P*|
|Serum β2-microglobulin level†||.35|
|High (>3.0 mg/L)||41||18||23|
|Serum LDH level‡||.42|
|High (>618 IU/L)||17||10||7|
|Peripheral blood involvement‡||.82|
A new prognostic index has been proposed for assessing the prognosis of MCL patients, the so-called MIPI.6 For the current analysis, we could assess only 64 patients, because data were incomplete for 7 patients. The percentage of Ki-67-positive cells (the PI) was not correlated with low risk versus intermediate/high risk according to the MIPI (Table 1). Only the blastoid variant of MCL had a positive, significant correlation with a high PI using the 11.4% cutoff (Fisher exact test; P=.01).
Univariate Survival Analysis
Complete clinical and laboratory parameters and follow-up data were available on 71 patients for survival analysis. The significance of the established prognostic parameters initially was evaluated by using univariate statistical analysis and the Kaplan-Meier method. Age, sex, histology (classic vs blastoid variant), serum β2-microglobulin or LDH levels, the presence of B-symptoms, the presence of peripheral blood involvement, and IPI score (0-2 vs >2) were not correlated with OS or FFS.
For the assessment of PI and prognosis, we initially used a median of 11.4% Ki-67-positive cells as a cutoff for the entire study group. After a median follow-up of 51 months, 20 of 36 patients (56%) who had MCL with a high PI failed therapy compared with 12 of 35 patients (44%) patients who had MCL with a low PI. At 5 years, the FFS rate was 32.2% for patients who had MCL with a high PI versus 62% for patients who had MCL with a low PI (P=.14; log-rank test). Regarding survival, 12 of 36 patients (33.3%) who had MCL with a high PI died compared with 13 of 35 patients (37%) who had MCL with a low PI. The OS rate at 5 years was 50% for patients who had MCL with a high PI versus 66.5% for patients who had MCL with a low PI (P=.51; log-rank test). For the subset patients with the classic MCL variant, using a median of 10% Ki-67-positive cells as a cutoff, we observed a significant difference in the FFS rate. At 5 years, the FFS rate was 37% for patients who had MCL with a high PI versus 70% for patients who had MCL with a low PI (P=.028; log-rank test) (Fig. 2A). There was no difference in OS using the 10% cutoff.
We also assessed other Ki-67 percentages as cutoffs for the entire group of patients with MCL and for the subset of patients with the classic MCL variant. Cutoffs of 15% and 20% both were statistically significant, but the 20% cutoff yielded the lowest P value. By using this cutoff, the 5-year FFS rate was 28.6% for patients who had MCL with a high PI (>20%) compared with 53.1% for patients who had MCL with a low PI (≤20%; P=.003) (Fig. 2B). Similarly, in the group with the classic MCL variant, the 5-year FFS rate was 14% for patients who had MCL with a high PI compared with 58% for patients who had MCL with a low PI (P=.0005) (Fig. 2C). Table 2 summarizes the differences in 5-year FFS depending on the percentage of Ki-67-positive cells (PI) chosen as a cutoff. Regarding OS, none of the PI cutoffs that we tested yielded a significant difference at 5 years. By using the 20% cutoff, the OS rate was 51.4% for patients who had MCL with a high PI versus 64.3% for patients who had MCL with a low PI (P=.19; log-rank test).
|Ki-67 Cutoff, % Positive Cells||% Patients Above the Cutoff||No. of Events||Total||P|
Multivariate Survival Analysis
Multivariate analysis using the Cox proportional hazards model was applied to the entire study group using PI (Ki-67 expression) as the standard covariate and introducing all other prognostic variables 1 by 1 as covariates into the model. This analysis revealed that a high PI was an independent, adverse prognostic factor associated with a shorter FFS (P=.03) along with age ≥60 years (P=.033) and high serum levels of β2-microglobulin (P=.055) (Table 3). These prognostic parameters were not significant for OS.
|Age >60 y||2.5||1.1-5.7||.033|
Patients with MCL have a poor prognosis and are incurable if they are treated with traditional chemotherapeutic approaches. Although response rates to single-agent and combination chemotherapy regimens range from 40% to 70%, they are short-lived, disease progression develops in 6 to 18 months, and the median survival is 30 to 40 months.7-9 At our institution, the poor prognosis for patients with MCL lead to the use of dose-intensive chemotherapy regimens for these patients as a part of clinical trials; currently, the regimen that we use is R-hyper-CVAD.7 This regimen has yielded a response rate of 93%, which increased to 100% after the addition of high-dose cyclophosphamide and total body irradiation followed by autologous or allogenic stem cell transplantation.7
Molecular studies of MCL have revealed several genetic alterations that dysregulate the cell cycle.2 The cytogenetic hallmark of MCL, the t(11;14)(q13;q32), results in the overexpression of cyclin D1, which accelerates G1/S phase cell cycle transition. Rosenwald and colleagues demonstrated that10 the proliferation gene expression signature of MCL has prognostic significance. Ki-67 is a well accepted marker of cell proliferation that can be tested easily by immunohistochemical methods in routinely processed archival tumor specimens. Ki-67 is a nuclear antigen that is expressed in all phases of the cell cycle except in resting cells (G0 phase). Others investigators have used Ki-67 immunohistochemistry as an alternative to gene expression profiling to assess MCL proliferation.11-15 In every study, the immunohistochemical assessment of Ki-67 has demonstrated prognostic significance in patients with MCL. In the largest study of 304 patients by Tiemann and colleagues, the percentage of Ki-67-positive cells subdivided patients with MCL into 3 prognostic groups in which higher percentages of Ki-67-positive cells predicted shorter OS.13 In earlier studies, however, patients with MCL were treated with traditional chemotherapy regimens, most often CHOP chemotherapy; and, in most studies, patients with MCL were not treated in a uniform manner. In a recent study by Determann and colleagues,15 Ki-67 also had prognostic significance in 96 patients who received R-CHOP chemotherapy. To our knowledge, the prognostic significance of Ki-67 expression, which has been evaluated by several groups,7-9 has not yet been assessed in patients with MCL who received intensive chemotherapy regimens With the high complete remission rate and improved survival of patients with MCL who receive the R-hyper-CVAD regimen at our institution, we wondered whether Ki-67 expression would retain prognostic significance in this group of patients.
Within the group with classic MCL, using a median of 10% Ki-67-positive cells as a cutoff, a high Ki-67 PI was correlated with worse 5-year FFS. By using a higher cutoff of 20% Ki-67-positive cells, a high Ki-67 PI was correlated with poorer 5-year FFS both for the group with classic MCL and for the entire study group. Therefore, our results indicate that tumor PI has prognostic importance even in patients with MCL who are treated with intensive chemotherapy regimens, such as the R-hyper-CVAD regimen used at our institution.
The current results for patients with the blastoid variant of MCL in this study agree with previous data from our institution. Previously, Romaguera et al18 compared patients who had the classic MCL variant with patients who had the blastoid variant. The rates of complete remission in that study were slightly lower for patients with the blastoid variant (79% vs 89%), but the difference was not statistically significant (P=.72). At a median follow-up of 57 months, there was a plateau in the FFS curve for patients with blastoid MCL, and the median OS was not reached. The improved outcome of patients who have MCL with blastoid cytologic features suggests the potential for long-term remission after intense, nonmyeloablative chemotherapy in this subgroup of patients.
In summary, the data presented here demonstrate that the PI has prognostic significance in patients with MCL. The data support the results of earlier studies but also supplement those prior studies by demonstrating that the MIPI has prognostic value in patients who uniformly receive an aggressive chemotherapy regimen, R-hyper-CVAD, which is used to treat patients with MCL at our institution and elsewhere.
Conflict of Interest Disclosures
The authors made no disclosures.
- 1Mantle cell lymphoma. In: Jaffe ES, Harris NL, Stein H, Vardiman JW, eds. World Health Organization Classification of Tumours. Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissues. Lyon, France: IARC Press; 2001: 168-170., , , et al.
- 8Early consolidation by myeloablative radiochemotherapy followed by autologous stem cell transplantation in first remission significantly prolongs progression-free survival in mantle cell lymphoma—results of a prospective randomized trial of the European MCL Network. Blood. 2005; 105: 2677-2684., , , et al.
- 9The hyper-CVAD-rituximab chemotherapy programme followed by high-dose busulphan, melphalan, and autologous stem cell transplantation produces excellent event-free survival in patients with previously untreated mantle cell lymphoma. Ann Hematol. 2007; 86: 101-105., , , et al.
- 13Histopathology, cell proliferation indices and clinical outcome in 304 patients with mantle cell lymphoma (MCL): a clinicopathological study from the European MCL Network. Br J Haematol. 2005; 13: 29-38., , , et al.
- 18Blastoid mantle cell lymphoma (MCL): 5-yr failure-free survival (FFS) rate of 50%, without failures after 2.5 years following treatment with rituximab (R)-hyperCVAD alternating with R-methotrexate/cytarabine (M/A) [ASH Annual Meeting Abstracts]. Blood. 2006; 108. Abstract 2749., , , et al.