Radiotherapy of primary central nervous system lymphoma in patients with and without human immunodeficiency virus. Ten years of treatment experience at the University of California San Francisco

Cancer. 1994 May 15;73(10):2570-82. doi: 10.1002/1097-0142(19940515)73:10<2570::aid-cncr2820731019>;2-1.


Background: Classic human immunodeficiency virus (HIV) negative primary central nervous system lymphoma (PCNSL) is a relatively uncommon occurrence, whereas the incidence of HIV positive PCNSL has increased dramatically during the past 5 years. Between 2% and 6% of all patients with acquired immune deficiency syndrome (AIDS) will have PCNSL develop clinically, and as many as 11% will have PCNSL found at autopsy. In the United States, some of the earliest and most extensive experience with HIV positive PCNSL has occurred in San Francisco. This article reports on PCNSL in patients with and without HIV.

Methods: Using our diagnosis-based computer retrieval system, we identified all patients treated with radiotherapy for PCNSL between January 1982 and January 1992 and reviewed their medical records. Population characteristics, risk factors and length of survival (LOS), were analyzed. Statistical methodology included the Cox proportional hazards model and Kaplan-Meier survival curves.

Results: Fifty-six patients were identified with PCNSL, of which 41 were HIV positive and 15 were HIV negative. There was a fourfold increase in the total number of PCNSL cases during the time period 1987-1991 as compared with the preceding 5 years; these cases primarily were in the HIV positive group. The average Karnofsky performance score (KPS) in the HIV positive group was 50, whereas in the HIV negative group, the KPS was 35. However, the median LOS in the HIV positive group was 3 months, whereas in the HIV negative group it was 20 months. Within each group, there was a statistically significant correlation between KPS and LOS.

Conclusions: Patients with HIV and PCNSL have a much shorter median LOS than do patients with PCNSL but without HIV; however, the outcome for both groups is dismal. Randomized clinical trials are needed to determine which patients with HIV might benefit from more aggressive therapy and to determine the optimal choice and dosages of chemoradiation therapy for patients with PCNSL without HIV.

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Brain Neoplasms / mortality
  • Brain Neoplasms / radiotherapy*
  • Female
  • Humans
  • Lymphoma / mortality
  • Lymphoma / radiotherapy*
  • Lymphoma, AIDS-Related / mortality
  • Lymphoma, AIDS-Related / radiotherapy*
  • Male
  • Middle Aged
  • Proportional Hazards Models
  • Retrospective Studies
  • Risk Factors