Presenter: Richard Charles Curry, MD Presenter's Affiliation: Memorial Sloan-Kettering Cancer Center, New York, NY
Primary central nervous system lymphoma (PCNSL) is a relatively uncommon variant of extranodal non-Hodgkin lymphoma (usually diffuse large B-cell lymphoma). The disease involves the central nervous system: brain, leptomeninges, eyes, or spinal cord, without evidence of systemic disease.
Typically, patients are treated with chemotherapy involving high-dose methotrexate followed by whole brain radiation therapy (WBRT), depending on performance status and age. Patients treated with WBRT typically receive doses of 45 Gy.
WBRT improves progression-free survival (PFS), but can negatively impact neurocognitive function, particularly with the high doses used in PCNSL.
The current study was designed to determine whether additional cycles of induction chemotherapy improve complete response (CR) rate. The study also examines effectiveness and toxicity of reduced-dose whole-brain radiotherapy (rdWBRT) after CR.
Shah et al. (JCO, 2007) reported the early, pilot phase results (N=30) of this study. In this study, 44% of patients achieved a CR after five or fewer cycles of chemotherapy, and 78% of patients achieved CR after seven cycles.
The current presentation reports on the final results of the multicenter phase II study (N=52) and addresses long-term disease control and cognitive outcomes in patients who actually received rdWBRT.
Materials and Methods
This is a prospective multicenter phase II trial
Age >18 years, any KPS, histologic confirmation of PCNSL, HIV negative, no evidence of systemic NHL with work-up including bone marrow biopsy and CT Chest/Abdomen/Pelvis, and adequate bone marrow, liver and renal function
Prior treatment for CNS lymphoma
Other active malignancies, excluding basal cell carcinoma of the skin or cervical carcinoma in-situ
Pts received 5 cycles of R-MPV (1 cycle= 14 days)
Rituximab: 500 mg/m2 on day 1
MTX: 3.5g/m2 over 2h on day 2
Vincristine: 1.4 mg/m2 on day 2
Procarbazine: 100 mg/m2 on days 2-8 during cycles 1, 3 and 5
MRI was performed after 5 cycles
Those with partial response (PR) received an additional 2 cycles of chemotherapy. Patients with a complete response (CR) after 5-7 cycles received rdWBRT (23.4 Gy). If patients did not achieve CR after chemotherapy, standard WBRT was offered (45 Gy).
2 cycles of cytarabine (3 mg/m2/day on day 1 and 2) followed WBRT (1 cycle= 28 days)
Primary Endpoint: 2-year progression-free survival (PFS) in patients receiving rdWBRT
Exploratory Endpoints: Neuropsychological performance (at baseline, post induction chemotherapy 12, 24, 36 months), white matter changes/ leukoencephalopathy (Fazekas scale: 0-6), prognostic value of apparent diffusion coefficient (ADC) on MRI.
A total of 52 patients were enrolled in the study and 43 patients completed 5 cycles of chemotherapy. In total, 31 (59%) pts were assessable for the primary endpoint (achieved a CR after induction and received rdWBRT) and 30 patients were actually assessed.
Patient characteristics: 22 (42%) women; median age = 60 (30-79); median KPS= 70 (50-100); MSKCC RPA Class I (Age < 50) = 9%, II (Age > 50, KPS > 60) = 46%, III (Age > 50, KPS <70) = 27%. Median follow-up was 5.6 years
The 2-y PFS for the rdWBRT patients was 78% (95% CI: 64%- 92%). The median PFS was 7.7 years. The median OS was not reached (median follow-up was 6 years). 3 yr-OS was 88% (CI: 70-95) and 5 yr-OS was 81% (CI: 62-91).
Because age is a significant prognostic variable in PCNSL, outcomes were analyzed by age group: < or > 60 years. No difference in PFS or OS was noted on intent to treat analysis.
Toxicity: 2 patients died of febrile neutropenia, 1 patient developed grade 3 renal toxicity, 1 patient developed grade 3 ileus and 1 patient experienced a perforated diverticulum.
Neuropsychological testing: Following induction chemotherapy, there was an improvement in all cognitive domains (executive functioning, memory and motor skills, p<0.01). No significant cognitive decline was observed during the subsequent follow-up period.
White Matter Changes on MRI: 36% of patients showed no change and 64% of patients developed minimal patchy white matter foci (Fazekas score 1) or start of confluence of white matter disease (Fazekas score 2).
Apparent diffusion coefficient (ADC) on MRI failed to predict response
The authors conclude that consolidation rdWBRT is a highly effective and safe treatment for newly diagnosed young and elderly PCNSL.
Induction chemotherapy resulted in a high CR rate thought to be due to an optimal chemotherapy drug regimen with an increased number or cycles
The long-term follow-up showed robust PFS and OS rates, comparable or superior to historical data of patients treated with full dose WBRT, with excellent cognitive outcomes and no significant white matter change on MRI over time.
The authors have initiated an RTOG randomized trial comparing R-MPV-A with vs. without rdWBRT to determine whether or not WBRT can be eliminated in patients who achieve CR with chemotherapy.
Given the high rates of progressive disease in PCNSL and the neurocognitive complications associated with WBRT, treatment strategies to either eliminate or reduce the dose in WBRT are enticing for investigation.
This study demonstrates the ability to substitute additional cycles of chemotherapy for higher doses of WBRT in order to increase the possibility of CR and rate of PFS, while reducing possible neurotoxicity.
With lower doses of WBRT, there is concern for increasing failures. PCNSL can relapse years after the completion of therapy. In addition, decline in neurocognitive function can continue for years after the completion of WBRT. Therefore, additional follow-up is needed to accurately assess the risk of failure, as well as true neurocognitive function after lower doses of WBRT.
Although the authors are investigating the ability to eliminate WBRT in patients with CR after chemotherapy, close and frequent follow-up is essential in these patients as their presumed risk of failure is high.
Further investigation: PRECIS and IELSG#32 are 2 phase II trials randomizing patients to WBRT (40 Gy) vs. Auto-SCT to see if autologous stem cell transplant can replace WBRT. While WBRT increases a patient's risk for neurocognitive decline, the morbidity and mortality of stem cell transplant is significant and cannot be overlooked.
Other strategies to manipulate radiation therapy to maintain or improve efficacy while decreasing the risk of toxicity include modifying fractionation, volume and modality. With conformal techniques and particle therapy, the ability to spare areas of the interest in the brain, such as the hippocampus, temporal lobes and other critical structures, is becoming more feasible. Additionally, it may be possible to treat patients who develop a partial response after chemotherapy with rdWBRT of 2340 cGy, followed by a boost to the areas of residual disease in order to reduce toxicity to the normal brain, while maintaining adequate dose to the areas at risk.
Dec 29, 2011 - Cytogenetic analysis at the time of diagnosis in patients with acute myeloid leukemia (AML) shows higher rates of chromosomal abnormalities for patients with central nervous system (CNS) involvement than for those with no CNS involvement, and survival is typically poor for patients with AML and CNS disease, according to a study published in the Jan. 1 issue of Cancer.