Definition
Isolated central nervous system (CNS) relapse of diffuse large B-cell lymphoma (DLBCL) is a serious and potentially fatal complication of the disease. Delivery of chemotherapy with adequate CNS penetration as part of primary therapy provides an opportunity to prevent this complication. However, there is no universally agreed approach to CNS prophylaxis and there are open questions with respect to patient selection and prophylaxis choice. Nevertheless, appropriate CNS prophylaxis during primary therapy for DLBCL should be considered in patients with recognised risk factors for CNS relapse.
The eviQ Haematology Reference Committee issues this document to summarise risk factors for CNS relapse and provide guidance to patient selection and prophylactic treatment. Individual units may have equally valid approaches to this matter.
Patient selection
In the past, much of the data used to define groups at high risk of CNS relapse antedated modern immuno-histochemical diagnostic algorithms, PET-staging techniques, and the advent of routine chemo-immunotherapy with rituximab. Across all series, several or all factors in the International Prognostic Index (IPI) and revised International Prognostic Index (R-IPI) have conferred increased risk of CNS relapse.rrr The CNS International Prognostic Index (CNS-IPI) is a risk stratification tool developed by the German High Grade NHL Study Group (DSHNHL) and validated by the British Columbia Cancer Agency (BCCA), and can be used to predict risk of CNS involvement and recurrence.rr The following factors were identified as predictive of CNS relapse:
CNS-IPI risk factors:
- age > 60 years
- stage III/IV disease
- raised LDH
- >1 extranodal sites
- ECOG performance status >1
- kidney and/or adrenal involvement
The following 2-year rates of CNS relapse were reported:
Risk group |
Number of factors |
Risk of CNS relapse
(DSHNHL cohort n=2164) |
Risk of CNS relapse
(BCCA cohort n=1597) |
Low |
0-1 |
0.6% |
0.8% |
Intermediate |
2-3 |
3.4% |
3.9% |
High |
4-6 |
10.2% |
12% |
Greater than 2 extra nodal sites, irrespective of CNS-IPI, is an independent risk factor for CNS relapse.r The 3-year cumulative incidence is 15.2% for all patients, including 8% for those who did not have a CNS-IPI that was traditionally high risk.
Specific sites of disease involvement may be associated with increased risk of relapse (independent of CNS-IPI risk score).The following additional extranodal sites should be considered as high risk for CNS relapse:rrr
- testicularr
- intra-ocular
- base of skull
- breast
- epidural
- naso-pharyngeal.
Isolated large cell involvement of the bone marrow is not considered independently high risk for CNS disease, but typically marrow involvement is associated with other high risk factors.r
Further analysis from the BCCA using the CNS-IPI indicates that 90% of patients with DLBCL belong to the low- and intermediate-risk group with a CNS relapse risk of <5%. In the absence of specific symptoms, CNS-directed diagnostic and prophylactic intervention in this group remains the decision of the treating clinician (with the exception of patients with testicular, kidney or adrenal involvement). High-risk patients have a >10% risk of CNS relapse and are to be considered for CNS-directed investigations and treatment.r
Biological subtypes warranting consideration of CNS prophylaxis
Double/triple-hit lymphoma:
Double/triple-hit lymphoma (DHL/ THL) demonstrating rearrangements of MYC and BCL2 by FISH have a high risk of CNS relapse with a large series reporting a 13-20% of CNS relapse.r
Double expressors and cell of origin:
DLBCL that co-expresses MYC and BCL2 (or BCL6) by immunohistochemistry without associated translocations (“double-expressing lymphoma" or DEL) confers a 2-year risk of CNS relapse of between 11 and 22%in some studies.r Recent studies have indicated the integration of biomarkers such as activated B-cell (ABC) or unclassified cell of origin into the CNS-IPI risk model may improve CNS relapse prediction. A retrospective analysis of DEL by the BCCA found risk of CNS relapse to be approximately 9% (versus 2% in non-double-expressing DLBCL), however, this risk was further modified based on cell of origin, and CNS-IPI risk stratification.r The risk of CNS recurrence in DEL appears restricted to the ABC-like DLBCL (in which the CNS relapse risk is approximately 15%) with no apparent increased risk in double-expressing germinal centre B-cell (GCB)-like DLBCL. Increased risk also appears limited to the intermediate- and high-risk CNS-IPI groups. The addition of predictive biomarkers such as ABC or BCL2 and MYC expression could complement the CNS-IPI model in the identification of high-risk subgroups.r For example, DEL and high-risk CNS-IPI identified a group of patients with a 2-year risk of CNS relapse of 22.7%.
Intravascular lymphoma:
CNS involvement is common at baseline in this disease (up to 39%) and CNS prophylaxis is recommended.r
Both the British Society for Haematology (BSH)r and the Spanish Lymphoma Group (GELTAMO)r have recently published guidelines and algorithms that support the use of CNS prophylaxis in high risk patients as described above.
Figure 1. A practical algorithm on prophylaxis in high-risk aggressive CNS lymphoma patients - Spanish Lymphoma Group guidelinesr
DLBCL, diffuse large B-cell lymphoma; CNS, central nervous system; LDH, lactate dehydrogenase; CNS-IPI, central nervous system-International Prognostic Index; CC, conventional cytology; FCM, multiparameter flow cytometry; MRI, magnetic resonance imaging; CT, computerized tomography; IV, intravenous; HD-MTX, high dose-MTX; MTX, methotrexate; IT, intrathecal; ASCT, autologous stem cell transplant.
© Haematologica 2017
Choice of prophylaxis regimen
There is no agreed standard approach to CNS prophylaxis. CNS-penetrating intravenous antimetabolite or purine analogue therapy (methotrexate or cytarabine) is often used in patients thought to be at high risk of CNS disease. Intrathecal chemotherapy with methotrexate has long been a component of CNS prophylaxis and is beneficial in patients with testicular lymphoma.r However in other forms of DLBCL efficacy of IT MTX appears to provide little benefit. Systemic methotrexate is thought to be more effective than IT MTX at preventing parenchymal relapse, as opposed to leptomeningeal relapse. A local, multicentre, retrospective study of 269 patients explored three approaches to CNS prophylaxis (IT MTX alone with R-CHOP-like chemotherapy; IT MTX + IV MTX 3 g/m2 x 2 doses with R-CHOP-like chemotherapy; regimens containing high doses of ara-C and MTX (CODOX-M/IVAC or HyperCVAD)) and found CNS relapse rates of 18.4%, 6.9% and 2.3% (p=0.006).r
The consensus surrounding efficacy of IT MTX or HD-MTX as CNS prophylaxis on survival outcomes remains unclear given the lack of randomized controlled studies and poor prognosis of this cohort.rrr A recent retrospective study comparing the efficacy of HD-MTX prophylaxis in newly diagnosed patients based on initial treatment intent showed no significant difference in the CNS relapse rate, PFS or OS rates between the different arms.r Another multicentre retrospective study of 326 patients with DLBCL at high risk of CNS relapse demonstrated no association between HD-MTX and CNS relapse, PFS or OS.r However, another phase 2 study suggests early systemic CNS prophylaxis with HD-MTX followed by dose-dense immunochemotherapy can produce durable remissions with a low risk of CNS progression in high-risk patients (including those with BCL2/MYC double-hit lymphoma).r