Initial management of primary central nervous system lymphoma in Spain in the last decade. The experience of the GELTAMO and S panish neuro‐oncology groups

Abstract

Introduction: Primary central nervous system lymphoma (PCNSL) is a challenging tumour frequently misdiagnosed and associated with a considerable diagnostic delay. Outcome of PCNSL patients not included in clinical trial may be substantially different. We describe the clinical outcome of a non-selected, representative series of PCNSL diagnosed in the last decade in Spain. Method(s): We present a retrospective series of 314 immunocompetent adults diagnosed of PCNSL between 2005 and 2014 from 22 institutions from Spain. Result(s): 314 patients (M/F, 164/150) with a median age of 65 years [range 18-82] were included in the study. Seventy-one percent of patients had an Eastern Cooperative Oncology Group (ECOG) >=2 at diagnosis. Ninety-six percent were diffuse large B-cell lymphoma. Most frequent symptoms at presentation were focal deficit (45.5%) and cognitive impairment /or psychiatric (39%). Visual deficit was presented in 29 (9.2%). Magnetic resonance imaging (MRI) was performed at a median of 22 [1-708] days after the first symptom. Atypical contrast enhancement (ring or non-homogeneous) on MRI was present in 33.8% of patients. Steroids were administered before surgery in 52.5% of patients and withheld before surgery in 32% of them. Diagnosis was done by biopsy or resection in 79.6% and 20%, respectively. Median time from clinical onset to pathological diagnosis was 45 [4-817] days, being it longer in those patients with ECOG >=2 comparing with ECOG <2 (53 vs 31 days, p = 0.003). Median time from clinical onset to start initial treatment was 63 [7-846] days. Time from MRI was performed to pathological diagnosis was longer in those who had received steroids than other not (40 vs 18 days, p = 0.003). Nineteen percent of patients never received any treatment, most of them due to old age and/or ECOG >=2 at diagnosis. In the case of patients receiving treatment, first line included chemotherapy alone based in high dose methotrexate in 189 patients (74.7%) or associated with radiotherapy, 41 patients (16.2%) and radiotherapy alone, 23 (9.1%). Moreover, 61 patients (24.1%) received autollogous stem cell transplantation as consolidation. Median overall survival for patients treated with radiotherapy alone was 12.5 months (IC 95%, 5.9-19), if received chemotherapy alone was 13.9 months (IC 95%, 9.9-19.9), for those treated with chemotherapy plus radiotherapy was 30.2 months (IC 95%, 14.8-45.6) and those who received an autollogous transplant was 72.9 months (IC 95%, 66.7-79.1). For the whole series, univariate analysis showed the absence of cognitive impairment /or psychiatric symptoms (p = 0.001), absence of visual deficit (p = 0.026), age < 0.0001), male sex (p = 0.012) and ECOG <2 (p < 0.0001) as the most important variables predicting better survival at diagnosis. However, in the multivariate analysis, only age < 0.005), male sex (p = 0.04) and ECOG <2 (p < 0.0001) were prognostic factors for better outcome. Conclusion(s): Steroid administration was associated with a diagnostic delay. Younger patients, ECOG < 2 and male gender were independent prognostic factors of better survival. Prospective studies are necessary to validate these findings.