Low Rates of CNS Relapse in High Risk DLBCL Patients Treated with R-CODOX-M and R-IVAC: Results from a Phase 2 UK NCRI/Bloodwise Trial
Elizabeth H Phillips,Amy A Kirkwood,Anthony Lawrie,Simon Rule,Russell Patmore,Ruth Pettengell,Cathy H Burton,Kirit M Ardeshna,Silvia Montoto,Shankaranarayana Paneesha,Paul Smith,Katharina Wanek,Laura Clifton-Hadley,David C. Linch,Andrew McMillan
DOI: https://doi.org/10.1182/blood.v128.22.1855.1855
IF: 20.3
2016-12-02
Blood
Abstract:Abstract Introduction: Central nervous system (CNS) relapse of diffuse large B-cell lymphoma (DLBCL) represents a major clinical challenge and is fatal in most patients. Recently Schmitz et al (J ClinOncol 2016), defined an effective risk model, the CNS-IPI, to identify those at highest risk of CNS relapse, based on the international prognostic index (IPI) score and presence of renal or adrenal involvement. For DLBCL patients receiving R-CHOP-like regimens +/- intrathecal methotrexate, the risk of CNS relapse for low, intermediate and high-risk patients was <1%, 3-4% and 10-12%, respectively. The optimum strategy for CNS prophylaxis, however, has yet to be defined. Aim: To assess CNS relapse rates in an intermediate-high risk cohort of patients with DLBCL treated with the R-CODOX-M R-IVAC regimen, incorporating multiple CNS-penetrating agents. Methods: Patients with newly diagnosed DLBCL and an IPI score ≥3 were enrolled in a prospective, multi-centre, phase 2 trial (McMillan et al, Hematol Oncol 2015; 31(S1), 130a) and treated with modified CODOX-M and IVAC, including high dose intravenous methotrexate, cytarabine, ifosfamide and etoposide with 8-12 intrathecal injections (Mead et al, AnnOncol 2002; 23(8):1264-74); plus 8 doses of rituximab. The primary endpoint was progression-free survival (PFS). CNS involvement was diagnosed according to neurological signs, radiological findings and/or demonstration of malignant lymphocytes within the cerebrospinal fluid. Involvement ofextranodal sites was prospectively documented at registration and at relapse. Presence of CNS, adrenal and renal involvement was confirmed using case report forms prior to this post hoc analysis. Results: 108 patients were treated at 32 UK sites between May 2008 and April 2013. Median age was 50 years (18-65 years). Eight patients (7.4%) had CNS involvement at baseline. Eighty-two patients (75.9%) received 4 cycles of treatment. At a median follow-up of 45 months, PFS and overall survival were 65.5% (95% CI: 55.5 - 73.8) and 73.7% (64.0 - 81.2), respectively. Progression or relapse within the CNS occurred in 5 patients (4.6%; Table 1) at a median of 5.5 months after registration (0.9-9.1 months). All patients died within 9 months of CNS relapse, 4 due to DLBCL and one treatment-related death. Excluding those with CNS involvement at baseline or incomplete information (n=4; 2 with missing baseline information (no CNS relapse) and 2 awaiting confirmation of CNS status at relapse), CNS-IPI was evaluable in 96 patients, of which 95% had an elevated LDH, 57% had a performance status of ≥2, and 8% were ≥60 years. All patients had stage III-IV disease, 76% had >1 extranodalsite and 27% had renal or adrenal involvement. Forty-one patients (43%) were intermediate risk (2-3 factors) and 55 (57%) were high risk (4-6 factors) for CNS relapse. 2-year CNS relapse rates were 0% for intermediate risk and 6.2% (2.0 - 18.1) for high risk patients (Figure 1). Of the 3 CNS relapses in high risk patients, 2 occurred concurrently with systemic relapse; there was only one episode of isolated CNS relapse. Of the 8 patients with CNS involvement at baseline, 2 (25%) developed CNS relapse, including 1 isolated CNS relapse. One further patient died of refractory DLBCL whilst 5 (62.5%) are alive and progression free with a minimum of 28 months follow-up. Conclusions: Inclusion of CNS-directed therapy intrinsic to the R-CODOX-M IVAC regimen resulted in very low rates of CNS relapse. Although patient numbers and low event rates make direct comparison difficult, results appear promising alongside historical results with R-CHOP chemotherapy. CNS relapse rates for both intermediate and high risk patients in this trial were below the 95% confidence intervals for CNS relapse reported in large training and validation cohorts by Schmitz et al (0% vs 2.2 - 4.4 and 2.3 - 5.5 for intermediate risk patients and 6.2% vs 6.3 - 14.1 and 7.9 - 16.1 for high risk). Of note, only 2 patients in the whole cohort progressed with isolated CNS disease, one of whom had CNS disease at diagnosis. Thus, where systemic disease was fully treated, treatment failure due to inadequate CNS penetration was rare. Reasonable outcomes were achieved in patients with CNS involvement at diagnosis but greater patient numbers are required to further evaluate this regimen in secondary CNS lymphoma. Table 1 PFS events and CNS relapse rates Table 1. PFS events and CNS relapse rates Figure 1 CNS relapse rates according to CNS-IPI and presence of CNS disease at baseline Figure 1. CNS relapse rates according to CNS-IPI and presence of CNS disease at baseline Disclosures Phillips: Roche: Consultancy. Patmore:Roche: Honoraria; Janssen Cilag: Honoraria. Ardeshna:Roche: Membership on an entity's Board of Directors or advisory committees, Other: Conference Expenses, Research Funding. Montoto:Roche: Honoraria; Gilead: Research Funding.
hematology