Raji human Burkitt lymphoma cells (American Type Culture Collection, Manassas, VA) were grown in RPMI-1640 (Gibco BRL, Grand Island, NY) supplemented with 10% FBS, 2 mM L-glutamine, penicillin (100 U/ml), and streptomycin (100 μg/ml). MTX and RTX (Mabthera) were purchased from Mayne Pharmaceuticals and Roche Pharmaceuticals, respectively.

Anesthetized 6-week-old male Balb/c-nu mice were secured in a rodent stereotactic frame, a hollow guide screw was implanted into a small drill hole made 2 mm right and 1 mm anterior to the bregma, and 5×10⁴, 5×10⁵ and 5×10⁶ Raji cells in 5 μl HBSS were injected through this guide screw into the white matter at a depth of 3 mm [anterior/posterior (AP) +1.0 mm, medial/lateral (ML) −1.7 mm, dorsal/ventral (DV) −3.2 mm]. All of the experiments where mice received antitumor therapy were performed using an inoculum of 5×10⁵ Raji cells. Animal experiments were approved by the appropriate Institutional Review Boards of the Samsung Medical Center, Seoul, Korea and conducted in accordance with the ‘National Institute of Health Guide for the Care and Use of Laboratory Animals’ (NIH publication no. 80-23, revised in 1996).

Twelve days after the intracerebral injections of Raji cells, mice were randomly assigned to receive sham treatment (Group A), MTX administration followed by RTX treatment (Group B), or RTX administration followed by MTX treatment (Group C) (Fig. 3A). Twenty-one days after the injection of Raji cells, the mice were sacrificed. Brains were harvested and processed for paraffin embedding. The tumor volume was calculated by measuring the section with the largest tumor portion applying the formula: (width)² × length × 0.5.

To evaluate the penetration of RTX into the mouse brains bearing CNS lymphoma, RTX conjugated with Alexa Fluor 680 (RTX-AF680) was utilized. RTX was conjugated with Alexa Fluor 680 according to the manufacturer's protocol (SAIVI Alexa Fluor 680 antibody/protein 1 mg labeling kit; Invitrogen, CA, USA). Three days after i.v. injection of RTX-AF680 (Fig. 3A), mice were anesthetized with 2–3% isoflurane. The signal from RTX-AF680 was detected in the region of the brain using a prototype Xenogen IVIS^® Spectrum in vivo imaging system (Caliper Life Science). Fluorescence intensity was analyzed as photons per second (p/s) by Living Image 3.1 software (Caliper Life Science).

Statistical comparisons of the RTX penetration and tumor regression between groups were analyzed by one-way analysis of variance (ANOVA) followed by the least significant difference (LSD) test. All data were presented as means ± SEM. A values of P<0.05 was considered statistically significant.

To establish the CNS lymphoma animal model, Raji human Burkitt lymphoma cells were stereotactically injected into the brains of immune compromised Balb/c-nu mice with a range of cell doses (0.5, 5 and 50×10⁵ cells/mouse), and mice were observed for the development of irreversible neurological symptoms before euthanization. Nude mouse survival times were dependent on the cell dose (Fig. 1). The optimal cell dose of Raji cells was determined by the length of survival. A cell dose of 5×10⁵ resulted in 100% mortality within 21 days (19–23 days) after implantation and was used in all of the subsequent in vivo experiments. Huge infiltrative intracerebral tumors were observed in the brain three weeks after the tumor cell injection (Fig. 2A). Pathologic characteristics of CNS lymphoma such as extensive leptomeningeal seeding (arrows in Fig. 2A) and perivascular cuffing at the periphery of tumors (arrow heads in Fig. 2A) were obvious. In addition to cerebral involvement, lymphoma cells were clearly observed in the leptomeninges of the spinal cord but not in the parenchyma (arrows in Fig. 2B and D, thoracic spinal cord). Most lymphoma cells highly expressed CD20, a B-cell marker (Fig. 2C for perivascular cuffing in the brain, Fig. 2D for leptomeningeal seeding in the spinal cord).

To evaluate our hypothesis, we administrated MTX and RTX to the mice bearing lymphoma cells in their brains according to different schedules (Fig. 3A). Administered RTX was labeled with AF680 to visualize the distribution in the brain in vivo. Each mouse was examined 3 days after RTX treatment (Group B; Day 17, Group C; Day 15), while mice from Group A were examined twice at Day 15 and 17 (Fig. 3A). We did not observe fluorescent signals from the brains of the mice not administered RTX (Group A, Fig. 3B). When mice received MTX treatment followed by RTX (Group B), RTX was not accumulated in the brain (Group B, Fig. 3B). In contrast, penetration of RTX across the BBB or the blood-tumor barrier was significantly increased by changing the treatment order from MTX + RTX to RTX + MTX (Group C, Fig. 3B and C).

Next, we analyzed whether RTX penetration is associated with therapeutic effects by measuring tumor volumes at Day 21 (Fig. 3A). The RTX and MTX combination treatment, in which MTX administration was followed by RTX injection (Group B), reduced tumor volume by 32%, but this was not statistically significant (Fig. 4A and B; Group A, 62±8.0 mm³; Group B, 41.7±15.1 mm³). In contrast, mice that received RTX treatment followed by MTX administration showed a significantly reduced tumor volume (Fig. 4A and B; Group C, 20.1±10.2 mm³, 68% tumor volume reduction; P<0.05).