Clinical Features, Treatment, and Survival of Secondary Central Nervous System Lymphoma

Background

Secondary central nervous system lymphoma (SCNSL) is a rare complication of non-Hodgkin lymphoma (NHL) and is almost always fatal. In a review of 14 studies, SCNSL occurs in 2.3% to 10% of NHL patients and in 5% of diffuse large B-Cell lymphoma (DLBCL) patients. Median survival is 2 to 6.5 months from time of central nervous system (CNS) involvement. In this study, we sought to determine the factors that influence survival in SCNSL.

Methods

In a retrospective chart review with Institutional Review Board approval, all patients with SCNSL were identified from pathology reports and ICD-9 codes (198.3) using inpatient and outpatient clinic visits at the University of Rochester Medical Center, Rochester, New York from January 1^st, 2005 to September 30^th, 2013. Pathology reports were used to determine the total number of lymphoma patients diagnosed at URMC during this time period, which was used to calculate the incidence of SCNSL. Descriptive statistics and Kaplan-Meier curves were estimated using SAS 9.3 (Cary, NC).

Results

Forty-nine patients (median age = 60) developed SCNSL from DLBCL (32), transformed disease (5), marginal zone (3), Burkitts (3), lymphoplasmacytic (1), mantle cell (1), chronic lymphocytic leukemia (1), PTLD-DLBCL (1), or T-cell lymphoma (1). The cumulative incidence of secondary CNS disease in DLBCL was 3% during the study period. The majority of patients had stage IV disease (68.9%) and an elevated LDH at diagnosis. CNS prophylaxis was prescribed in 10.6% of all SCNSL patients and 12.5% in the DLBCL subset. Eighty-five percent of patients had other sites of metastases, with the majority of metastases occurring in the bone marrow, bone, lung, and liver. Patients developed CNS disease in the CSF (61%) or in the brain parenchyma (55%) or both (16%). Of the 27 patients with parenchymal involvement, 9 did not have a biopsy, but 5 had biopsy proven CSF involvement, resulting in a total of 4 unconfirmed cases, 3 of which were DLBCL. 12% of patients were asymptomatic; the other 88% presented with symptoms including headaches (33%), visual changes (29%), motor weakness (20%), cognitive problems (16%), and speech disturbances (14%). Treatment of SCNSL included a combination of treatments: methotrexate (63%), Rituxan (41%), cytarabine (39%), radiation (27%), and CHOP (20%). The median survival from diagnosis of SCNSL was 3.1 months in all patients and 2.6 months in DLBCL patients. In patients with leptomeningeal involvement only, patients that developed this involvement within 30 days of their initial NHL diagnosis were more likely to survive past 1 year (55.6%; 95% CI 0.23, 0.88) than patients who developed leptomeningeal involvement after 30 days (7.7%; 95% CI -0.07, 0.22) (p-value 0.02).

Conclusion

Over a period of 8.5 years, 49 NHL patients developed secondary CNS disease at a medical center in Western New York. Thirty-two out of the 49 NHL patients had DLBCL with secondary CNS disease. The cumulative incidence of secondary CNS disease in DLBCL is 3%, which is similar to what was reported in previous studies. The majority of SCNSL patients presented with neurologic symptoms. Our population closely resembles previously reported cohorts in characteristics and survival which remains poor.