Primary central nervous system lymphoma: a single-centre experience of 55 unselected cases

Current treatment for primary central nervous system lymphoma (PCNSL) involves high-dose methotrexate (HDMTX) with or without radiotherapy. Many published studies describing this approach include a highly selected group of patients. We report a single-centre experience of unselected cases of PCNSL.

We retrospectively reviewed the case notes of 55 consecutive patients diagnosed with biopsy-proven PCNSL between 1995 and 2003 at Addenbrooke's Hospital Cambridge, UK. We describe the treatment and outcome, including survival, treatment-related toxicity and long-term functional disability.

At diagnosis, 45% of patients were considered unfit to receive treatment with HDMTX, owing to poor performance status or comorbidity. These patients had a median survival of 46 days and may not have been included in other published studies. The remaining patients were treated with a chemotherapy regimen, which included HDMTX. Patients who received at least one cycle of a chemotherapy containing HDMTX had a median survival of 31 months. Forty per cent did not complete planned chemotherapy owing to toxicity, disease progression or death. The median survival of patients treated with HDMTX aged 60 years compared with patients aged under 60 years was 26 months vs 41 months (P = 0.07), respectively. Younger patients treated with HDMTX, who achieved complete remission with chemotherapy, had a median survival of 56 months. We identified a high incidence of functional disability among survivors, resulting from a combination of the tumour itself, the neurosurgical procedure required for diagnosis and the late neurotoxicity of combined chemoradiotherapy.

The treatment of PCNSL is associated with significant early and late toxicity. Further attempts to improve treatment should address mechanisms to reduce this toxicity. In particular, the benefit of radiotherapy in patients who achieve complete remission with HDMTX will remain uncertain until it is addressed in a multicentre, randomised trial.