Treatment Model of CNS Lymphoma Using Complement-Dependent Cytotoxicity Induced by Rituximab and Autoserum.

Primary central nervous system lymphoma (PCNSL) is almost exclusively CD20-positive non-Hodgkin lymphoma (NHL). Although rituximab (R) is widely used for CD20-positive NHL, it is not considered to reach brain lesions effectively beyond the blood brain barrier. Intraventricule administration (ivt) of R is reported to be effective in meningeal lymphoma but the effect on lesions in the brain parenchyma seems to be limited. Recently, a case of refractory PCNSL that was successfully treated with ivt of R with autologous serum was reported (Takami A, et al. Cancer Science, 2006). Because the cerebrospinal fluid does not contain complements which exists in the serum, induction of complement-dependent cytotoxicity by ivt of R plus autoserum was speculated. To investigate this effect, we developed an animal treatment model of CNSL.

Raji, CD20-positive Burkitt lymphoma cell line, was inoculated into the deep frontal lobe of the brain of 8-week old F344 (nru-/nru-) nude rats, using brain stereotaxic apparatus. At the same time, a cannula was placed into the ipsilateral lateral ventricle. After several days, R or control immunoglobulin (cIg), plus human serum or saline, was administrated into the lateral ventricle.

The brain was extracted 24 hours after the last administration and frozen section was made. Human CD20-positive Raji cell tumor was also positively stained with FITC-conjugated anti-human IgG antibody when R but not cIg was administrated. Consequently, R in the lateral ventricle was considered to penetrate ependymal cells and brain parenchyma, and bound to lymphoma cells. Next, these rats were treated with ivt of R plus serum (R + Serum), cIg plus serum (cIg + Serum), or R plus saline (R + saline). These were administrated once a day from day 5 to day 9 after inoculation of Raji, and then survival was monitored. When an obvious weakness, such as marked and consecutively loss of activity or weight, was observed, these rats were euthanized and this is defined as dead day. In each case, the brain was extirpated and examined whether lymphoma existed or not. Death without lymphoma or from technical problem was excluded from the analysis. Survival of each group was analyzed by Kaplan-Meier method and log-lank test. R + Serum group had longer survival than cIg + Serum (p = 0.049). Long-term survivors were only seen in R + Serum and this group seemed to be superior to R + saline but statistical difference was not detected (p = 0.083). There were no difference between cIg + Serum and R + saline (p =0.382) and neither group had long-term survivor.

The possibility of novel treatment of CNSL with ivt of R and autoserum was shown in the rat CNSL model. To confirm this approach, clinical trials are warranted.

No relevant conflicts of interest to declare.