Retrospective Survival of Patients with Primary Central Nervous System Lymphoma (PCNSL) Treated with Upfront Chemoimmunotherapy (R+MVP) and without Radiation: North Shore - LIJ University Hospital From 2006 to 2011

Abstract 4958

Treatment of primary CNS lymphoma (PCNSL) typically consists of high dose methotrexate (MTX)-based chemotherapy with or without the addition of whole-brain radiotherapy (WBRT). Although this treatment approach results in high rates of disease remission, many patients eventually progress. The addition of WBRT significantly increases the risk of treatment-related neurotoxicity, especially in elderly patients. Therefore, a more effective and less toxic treatment is required.

To address this issue, Shah et al (JCO 2007; v25; 4730–4735) reported on the combination of Rituximab and high-dose methotrexate (M), vincristine (V), procarbazine (P) plus reduced whole-brain radiation therapy for 30 patients with newly diagnosed PCNSL. They concluded that the addition of Rituximab to MVP led to a 2 year OS of 67% and PFS of 57%.

However, the role of WBRT has remained controversial. Delayed neurotoxicity limits its acceptance as a standard of care. The results of the G-PCNSL-SG1 randomized phase IV trial were presented at the ASCO 2010 Annual Meeting (Thiel et al JCO 2010; v28; supplement; abstract 8008). Patients who achieved a CR after 6 cycles of high-dose MTX were randomized to receive either 45 Gy WBRT or no further therapy. Median overall survival was 32 months for the chemotherapy + WBRT patients and 37 months in the chemotherapy alone group.

Moreover, patients with recurrent disease may be rechallenged with high-dose MTX. In a retrospective analysis of 48 patients, complete responses were seen in 58% of patients, and 54% of patients were alive at least one year after the WBRT (Hottinger et al Neurology 2007; v69; 1178–1182). The author concluded that reserving WBRT until tumor recurrence is a reasonable strategy to minimize or delay the risk of treatment-related toxicity.

In 2006, we began to use the R+MVP and cytarabine consolidation as previously described above by Shah et al. Due to well-known toxicities of WBRT and the efficacy and tolerability of high dose MTX re-challenge at relapse, we decided to omit radiotherapy and reserve its use only for MTX-refractory patients. We are now reporting this experience in a retrospective manner. Institutional IRB approval was obtained. All patients for whom the intention to treat with R+MVP at North Shore University and Long Island Jewish Hospitals between 2006 and 2011 for at least 5 cycles are included.

Patients will be included in the study in an intention-to-treat fashion provided that their planned treatment regimen consisted of at least 5 cycles that began prior to January 1, 2011. The aim is to evaluate overall survival (OS) and progression-free survival (PFS). Descriptive statistics (mean, median, standard deviation, interquartile range, frequencies, and proportions) were calculated for demographic and clinical factors. The Kaplan-Meier (KM) product-limit method was used to estimate PFS and OS. Subjects in which the outcomes of interest (progression or death) were not observed were considered censored using their last known date of follow-up.

There were 13 subjects included in our study. The mean age was 71 years (standard deviation (sd) = 8.48) and the mean Karnofsky performance status (KPS) was 72 (sd = 10.90). The sample consisted of eight males (62%) and five females (38%). Three subjects progressed (23%) and two subjects died (15%). Three of the thirteen subjects were also diagnosed with leptomeningeal disease. Twelve subjects (92.31%) had at least 5 cycles of R+MVP treatment. One subject (8%) only had three cycles of treatment and died three months after diagnosis. 5/13 patients developed grade <=2 peripheral neuropathy. Grade 3 infections were noted in 3 patients including sepsis and pneumonia. 3 patients had delayed clearance of methotrexate; one of which received carboxypeptidase. All toxicities completed resolved to baseline. There was one death from acute respiratory death syndrome (ARDS) possibly related to treatment. The estimated median PFS is 53 months. The median survival is 78 months. The 95% confidence intervals for the KM curves could not be estimated.

The activity of R+MVP chemoimmunotherapy without WBRT for PCNSL in this older patient population compares very favorably with prior reports where WBRT was included. Omission of upfront WBRT appears to be an acceptable alternative for these patients providing excellent survival and the option to retreat with the same regimen at progression.