1 Service Hématologie, 2 Service Informatique Médicale, Hôpitaux de Brabois, Nancy; 3 Centre Henri Becquerel, Rouen; 4 Service Hématologie, Hôpital Henri Mondor, AP-HP, Créteil; 5 Service Hématologie, Centre Hospitalier Universitaire, Bordeaux; 6 Service Hématologie, Centre Hospitalier Universitaire, Caen; 7 Service Hématologie, Centre Hospitalier Universitaire, Limoges; 8 Service Hématologie, Centre Hospitalier Universitaire, Toulouse; 9 Service Hématologie, Centre Hospitalier, Chambéry; 10 Service Anatomo-Pathologie, Hôtel Dieu, AP-HP, Paris; 11 Service Hématologie, Hospices Civils, Lyon, France
Received 19 March 2003; revised 17 June 2003; accepted 20 August 2003
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The incidence of secondary central nervous system (CNS) occurrences in diffuse large-B-cell lymphoma is not sufficiently high to warrant the use of CNS prophylaxis in all patients. The addition of rituximab increases the complete response rate and prolongs event-free and overall survival in elderly patients with such lymphoma.
Patients and methods:
We analyzed a cohort of 399 elderly patients with lymphoma prospectively treated with eight cycles of CHOP with or without rituximab in order to assess if rituximab decreases the risk of CNS localization. Prophylaxis of CNS disease was not part of the treatment protocol.
Results:
We observed 20 CNS occurrences: 12 on therapy, four after partial remission and four following complete remission. In three patients, the CNS was the only site of relapse. In a multivariate analysis, increased age-adjusted International Prognostic Index (IPI) was the only independent predictive factor of CNS recurrence. Only three of 20 patients are alive with a follow-up of 24 months.
Conclusions:
Rituximab did not influence the risk of CNS occurrence, possibly because of low rituximab diffusion. Direct intrathecal administration of rituximab could overcome this problem. We also confirmed that CNS occurrence is related to IPI as well as very poor prognosis of relapses occurring on therapy.
Key words: aggressive non-Hodgkins lymphoma, central nervous system occurrence, risk factors, rituximab
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The majority of non-Hodgkins lymphomas which involve the CNS are B-cell neoplasms which express CD20. The Groupe dEtude des Lymphomes de lAdulte (GELA) has recently shown that the addition of rituximab monoclonal antibody to the CHOP regimen (cyclophosphamide/doxorubicin/vincristine/prednisone) increases the CR rate and prolongs event-free and overall survival in elderly patients with diffuse large-B-cell lymphoma [1]. The longer survival in the CHOP + rituximab group was due to a lower rate of disease progression during therapy and fewer relapses among patients who had a complete response. Rituximab is a chimeric anti-CD20 human IgG1 monoclonal antibody in which the CD20-binding region was derived by genetic engineering from a mouse monoclonal antibody. Little is known about its CNS diffusion.
We took advantage of this multicenter study to re-analyze data; the main objective was to assess if rituximab decreases the risk of secondary CNS occurrence. We also evaluate the influence of recognized CNS relapse risk factors in this homogeneous cohort of patients. Secondarily, we focused on the patients who achieved CR and developed isolated CNS relapse.
![]() |
Patients and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Prophylaxis of CNS disease was not part of the treatment protocol. All the patients (n = 399) were initially analyzed for the occurrence of CNS localization. Lumbar puncture and cerebrospinal fluid (CSF) cytopathological analysis were carried out (before randomization) in all the patients who entered the trial. Patients were not eligible if they had CNS involvement. However, we cannot formally exclude that subclinical CNS involvement might have existed in a proportion of patients at the time of diagnosis and before randomization, especially in patients who relapsed quickly after the diagnosis. Presumably, the randomization process would have allocated such patients equally between the two groups.
Diagnosis of CNS disease
The diagnosis of CNS localization, carried out after standardized staging evaluation, was based on the presence of malignant cells on cytocentrifuge preparations of spinal fluid in seven cases and on brain biopsy in two cases. In 10 cases, the diagnosis of CNS disease was based on symptoms and radiological findings and in one case it was based on clinical symptoms only.
Statistical methods
Comparisons among groups were analyzed by the chi-square test and the Fisher exact test when appropriate. In order to determine the most appropriate factor predicting for CNS occurrences, a logistic regression analysis was carried out to assess the effect of pretreatment prognostic factors [age, sex, number of extranodal sites, ß2-microglobulin level, serum albumin level, B symptoms (weight loss, fever and night sweats) and type of treatment]. A second multivariate logistic regression analysis was carried out including the age-adjusted International Prognostic Index (aa IPI) score. Survival curves and actuarial risk of patients with CNS occurrence was calculated by the KaplanMeier method.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
Risk factors for CNS occurrence
We analyzed whether the risk of CNS occurrence was related to the treatment group (CHOP versus CHOP + rituximab) as well as the usual risk factors described in CNS relapse. As shown in Table 3 and in univariate analysis, an increased risk of CNS occurrence was associated with an advanced stage (P = 0.014), an elevated lactate dehydrogenase (LDH) level (P = 0.005), a poor performance status (P = 0.018) and an increased aa IPI (P <0.001). The presence of more than one extranodal site and the localization of extranodal sites as well as the age, sex and B symptoms were not related to the risk of CNS occurrence. Treatment group did not influence the risk of CNS localization. We observed 11 CNS occurrences in the CHOP + rituximab group and nine in the CHOP group without differences between patients in CR (two in both groups), PR (three in the CHOP + rituximab group, one in the CHOP group) or developing CNS localization on therapy (six in both groups). In a first multivariate analysis, all the parameters found to be significant (P <0.10), except aa IPI, were introduced in a logistic regression model. Poor performance status and elevated LDH were the two independent predictive factors of CNS occurrence. In a second logistic regression analysis including aa IPI, aa IPI was identified as the only independent factor associated with a higher risk of CNS occurrence (aa IPI 0 and 1 versus 2 and 3: odds ratio = 3.05 and confidence limits 1.585.88).
|
Treatment of CNS occurrence and outcome
CNS occurrence has a poor prognosis especially when observed on therapy or after PR and if it is part of systemic relapse. Treatment of the 20 patients with CNS occurrence was conducted according to each centers policy and therefore was heterogeneous. A total of 18 of 20 patients received a variable number of intrathecal injections of methotrexate, ara-C and dexamethasone associated with a systemic chemotherapy: high-dose methotrexate, COP-COPADM (including prednisone/vincristine/cyclophosphamide/ara-C/methotrexate/doxorubicin) [3], VIM2-ARAC (including mitoxantrone/etoposide/ifosfamide/ara-C) [4] or DHAP (including mitoxantrone/etoposide/ifosfamide/ara-C) [5]. Four patients presenting with cranial nerve palsies or intraparenchymal lesions in the brain or epidural lesion were given additional radiotherapy (two associated with chemotherapy). Only one patient presented an epidural tumor but was considered as CNS occurrence since he had associated leptomeningeal disease. Finally, two patients did not receive any treatment because of diffuse relapse and poor performance status. At the time of analysis, only three of 20 patients were alive: two presented as isolated CNS relapse after CR, one after PR (with a 20-month follow-up). All the patients with relapse on therapy died within 4 months whatever the administered treatment.
Isolated CNS relapse following CR
The CNS was the only site of relapse in only three patients in CR (three of 276 = 1.1%). All three patients received systemic chemotherapy (one was treated with high-dose methotrexate and two with COP-COPADM regimen) associated with a variable number of intrathecal injections of ara-C, methotrexate and dexamethasone. All the patients attained a second CR. Two patients were alive in second CR with a follow-up of 20 and 34 months and one died 4 months after systemic recurrence.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Most reported studies include isolated and diffuse localizations, occurring on therapy or after PR or CR has been obtained. In such series of patients, the overall incidence of CNS localizations is around 5%, similar to that which we observed. We show that the risk of CNS localization is related to the aa IPI; patients with a higher index have higher rates of CNS occurrence, confirming reported studies showing that secondary CNS occurrence is related to IPI or aa IPI (if analyzed) or to the factors forming part of it [810]. However, it is admitted that CNS localization associated with systemic relapse is an expression of end-stage disease and is usually refractory to all types of treatment. Indeed, in our population, all these patients died within 4 months. On the contrary, isolated CNS localization occurring after CR may be potentially treatable. In our series, we observed three (1.1%) isolated relapses among 276 CR patients. The incidence and risk factors for isolated CNS relapse have been assessed in recent studies of homogeneous and prospectively treated diffuse lymphoma patients who achieved CR and who did (ACVBP regimen) [11] or did not (CHOP regimen) [12] receive intrathecal chemoprophylaxis. Similar isolated CNS incidence was observed after CR (1% and 1.6%, respectively) and IPI was also identified as a unique independent risk factor for isolated CNS relapse in multivariate analysis. In contrast, Zinzani et al. [13] reported a higher incidence (5.2%) of isolated CNS occurrence in an unselected series of 175 patients with non-lymphoblastic, non-Burkitt aggressive lymphoma in CR following an anthracycline-containing regimen without intrathecal CNS prophylaxis.
The ACVBP regimen (four cycles every 3 weeks of doxorubicin/cyclophosphamide/vindesine/bleomycin/prednisone with four intrathecal injections of methotrexate followed by sequential consolidation therapy including high-dose methotrexate, ifosfamide, etoposide and ara-C) has been used by the GELA since 1980. The 93-5 GELA study compared ACVBP with the standard CHOP regimen in elderly patients [14]. The ACVBP regimen was found to be associated with longer event-free survival and prolonged survival. CNS occurrences were more frequent in the CHOP group (26 of 312 versus nine of 323) but the majority occurred on therapy (21 in the CHOP group versus only six in the ACVBP group); the number of isolated CNS relapses was similar in the two groups (two of 172 in the CHOP versus two of 177 in the ACVBP group). These data suggest that the risk of CNS disease is related to the risk of systemic recurrence. Administration of more intensive induction chemotherapy (ACVBP) as well as use of drugs crossing the bloodbrain barrier (methotrexate, VP16 and aracytine) could explain the lower number of systemic and isolated CNS occurrences. The higher number of secondary CNS occurrences (26 of 312 = 8%) in the 93-5 CHOP arm compared with the 98-5 study could be partially related to the number of T-cell lymphomas (15%) included in the 93-5 study as well as the absence of aa IPI = 0 patients.
Our data show that rituximab (at the dose of 375 mg/m2 on day 1 of each eight 3-week CHOP courses) did not influence the risk of secondary CNS occurrence in elderly patients with diffuse large-B-cell lymphoma, possibly because of low rituximab diffusion across the bloodbrain barrier. Direct intrathecal administration of rituximab might overcome this problem, since preliminary pharmacokinetic studies in the monkey suggest that such administration results in an initially high concentration before distribution into the CSF without toxicity [15]. We also confirmed other reported studies showing that CNS occurrence is related to aa IPI as well as very poor prognosis of relapses occurring on therapy or in a systemic disease. Our observations suggest that the risk of secondary CNS occurrence is related to disease control, since most of the patients developed CNS disease either on therapy or in PR. It can be hypothesized that more effective systemic treatment could decrease the number of patients with CNS disease occurring on therapy. This strategy has been included in the ongoing 01-5B trial designed for patients aged from 60 to 65 years comparing the CHOP + rituximab with an ACVBP + rituximab regimen.
![]() |
FOOTNOTES |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2. Oinonen R, Franssila K, Elonen E. Central nervous system involvement in patients with mantle cell lymphoma. Ann Hematol 1999; 78: 145149.[CrossRef][ISI][Medline]
3. Patte C, Michon J, Frappaz D et al. Therapy of Burkitt and B-cell acute lymphoblastic leukaemia and lymphoma: experience with the LMB protocols of the SFOP in children and adults. Bailleres Clin Haematol 1994; 7: 339348.[Medline]
4. Hopfinger G, Heinz R, Koller E et al. Ifosfamide, mitoxantrone and etoposide (VIM) as salvage therapy of low toxicity in non-Hodgkins lymphoma. Eur J Haematol 1995; 55: 223227.[ISI][Medline]
5. Velasquez WS, Cabanillas F, Salvador P et al. Effective salvage therapy for lymphoma with cis-platin in combination with high-dose Ara-C and dexamethasone (DHAP). Blood 1988; 71: 117122.[Abstract]
6. Harjunpaa A, Wiklund T, Collan J et al. Complement activation in circulation and central nervous system after rituximab (anti-CD20) treatment of B-cell lymphoma. Leuk Lymphoma 2001; 42: 731738.[ISI][Medline]
7. Rubenstein JL, Rosenberg J, Damon L. High-dose methotrexate plus rituximab (Anti CD20) monoclonal antibody in the treatment of primary CNS lymphoma. Society for Neuro-Oncology Fourth Annual Meeting, Scottsdale, AZ, USA, November 1999 (Abstr).
8. Bashir RM, Bierman PJ, Vose JM et al. Central nervous system involvement in patients with diffuse aggressive non-Hodgkins lymphoma. Am J Clin Oncol 1991; 14: 478482.[ISI][Medline]
9. Hollender A, Kvaloy S, Nome O et al. Central nervous system involvement following diagnosis of non-Hodgkins lymphoma: a risk model. Ann Oncol 2002; 13: 10991107.
10. Van Besien K, Ha CS, Murphy S et al. Risk factors, treatment, and outcome of central nervous system recurrence in adults with intermediate-grade and immunoblastic lymphoma. Blood 1998; 91: 11781184.
11. Haioun C, Besson C, Lepage E et al. Incidence and risk factors of central nervous system relapse in histologically aggressive non-Hodgkins lymphoma uniformly treated and receiving intrathecal central nervous system prophylaxis: a GELA study on 974 patients. Ann Oncol 2000; 11: 685690.[Abstract]
12. Bos GM, Van Putten WL, Van der Holt B et al. For which patients with aggressive non-Hodgkins lymphoma is prophylaxis for central nervous system disease mandatory? Ann Oncol 1998; 9: 191194.[Abstract]
13. Zinzani PL, Magagnoli M, Frezza G et al. Isolated central nervous system relapse in aggressive non-Hodgkins lymphoma: the Bologna experience. Leuk Lymphoma 1999; 32: 571576.[ISI][Medline]
14. Tilly H, Lepage E, Coiffier B et al. A randomized comparison of ACVBP and CHOP in the treatment of advanced aggressive non Hodgkins lymphoma: the LNH93-5 study. Blood 2000; 96: 832a (Abstr 3596).
15. Rubenstein JL, Combs D, Rosenberg J et al. Rituximab therapy for CNS lymphomas: targeting the leptomeningeal compartment. Blood 2003; 101: 466468.