Methotrexate and/or Rituximab in a New Nude Rat Central Nervous System Lymphoma Model: Efficacy, Toxicity and Magnetic Resonance Imaging.

The aim of this study was to determine the efficacy and toxicity of methotrexate (MTX) and/or rituximab treatment in a novel central nervous system (CNS) lymphoma model, using either intravenous (IV) or intraarterial (IA) drug administration in conjunction with osmotic blood-brain barrier disruption (BBBD). Female athymic nude rats (rnu/rnu) were inoculated with human MC116 B-cell lymphoma cells in the caudate putamen. On the day of treatment and first magnetic resonance imaging (MRI) (day 16–20), rats were randomized into 7 groups:

1. control group with saline IV (n = 15) or

2. IA/BBBD (n = 7);

3. rituximab 375 mg/m² IV (n = 6) or

4. IA/BBBD (n = 3);

5. MTX 1000 mg/m² IV (n = 6) or

6. IA/BBBD (n = 6); and

7. rituximab 375 mg/m² IV + MTX 1000 mg/m² IV (n = 6).

A second MRI scan was done 1 week after treatment. Study end points included tumor response on MRI (defined as ≤75% tumor volume on post-treatment MRI), toxicity, and tumor volumes on histology. In the IV groups, the percent change in tumor volume on T2/FLAIR images in the control versus rituximab group was statistically significant (p = 0.0051). In the IA/BBBD groups, there was a statistically significant difference between control and rituximab (p = 0.0167) for absolute and percent changes in tumor volumes on both T2/FLAIR and T1 with gadolinium sequences. For T1 with gadolinium, the difference between control and MTX was also significant for both absolute change (p = 0.0066) and percent change (p = 0.0043). Response rates in the IV groups differed significantly between controls and rituximab (p = 0.0017 on T2/FLAIR and p = 0.0049 on T1 with gadolinium). A good correlation between tumor volumes (in log mm³) on histology and fluid-attenuated inversion recovery (FLAIR)/T2 (R² = 0.83) or contrast-enhanced T1 weighted images (R² = 0.73) for tumors located in the caudate putamen was observed. Results of complete blood counts at sacrifice did not differ significantly between the groups. The MC116 CNS lymphoma model is valuable for preclinical testing of efficacy and toxicity of various treatment regimens. Rituximab and MTX were particularly effective in this model, regardless of the route of administration.