2.5.3. Immune Responses 
The patient's peripheral blood mononuclear cells and serum were analyzed to assess the development of 
cellular and humoral anti-tumor immune responses against autologous cultured tumor cells. As shown in 
Figure 2, peripheral blood mononuclear cells obtained between the 3thd and 8th injections frequently 
demonstated 3 to 4 fold greater tumor lytic activity compared to the pre-treatment control (there were 
insufficient viable cells to assess the effects of treatments 9 an 10). Tumor lytic activity following the 8th 
treatment could be completely inhibited by incubation of the effector cells with autologous tumor and 
partially inhibited by K562 cells (Figure 3). Partial inhibition of tumor lytic activity was observed with 
anti-CD8 but not with anti-CD4 antibody. These findings are consistent with the generation of a cellular 
anti-tumor immune response. The data suggest that this cellular immune response is comprised of both 
natural killer and CD8+ cytotoxic T cell components. 
Since the last RAC submission, we have performed additional assays with cryopreserved samples from 
our previously treated patient to confirm the reliability of the assay results and to provide data on cellular 
immune responses at additional time points during the course of observation. Intra-assay coefficients of 
variation ranged from 0.8 to 6% indicating the reliability of the cellular immune response assays. Overall, 
repeat evaluations were performed with 10 specimens. Discordant values were observed for a single 
specimen— the inhibition of tumor lytic activity by K562 cells tested for the 8th treatment (Figure 3). The 
results 22.8 +/- 0.6 % and 3 +/- 3 % were obtained in different assays. We have chosen to plot the 22.8 
+/- 0.6% value in Figure 3 as this value is most consistent with the concomittent demonstration of partial 
inhibition of tumor lytic activity by anti-CD8 antibody. The intra-class correlation coefficient was 
determined as a measure of inter-assay variation. The intra-class correlation coefficient was .65 for all 
data points and .90 if the discordant specimen was excluded from the analysis. 
Additional assays were performed with peripheral blood obtained following 2 and 3 month no treatment 
intervals between treatments 9 and 10. These samples also exhibited low baseline levels of anti-tumor 
lytic activity similar to the levels obtained prior to the initiation of therapy and following injections with 
transduced cells secreting very low doses of EL-2 (less than 10 units IL-2/24 hours). In summary, negative 
results were obtained at 5 different time points spanning the first six weeks of observation (pre-treatment 
and post-treatments 1 and 2 with very low doses of IL-2 ) and following 2 and 3 month no treatment 
intervals between treatments 9 and 10. Hence, there was no evidence of anti-tumor cellular immune 
responses in the absence of treatment with genetically modified cells secreting > 20 units EL-2/24 hours. 
Furthermore, partial inhibition of cytolytic activity by anti-CD8 antibody following the 8th treatment with 
higher doses of IL-2 secretion suggests the generation cytotoxic CD8+ T cells as a component of the 
cellular anti-tumor immune response. These data support our contention that injection of the genetically 
modified cells was associated with the generation of a cellular anti-tumor immune response and that the 
observed cellular immunity was not due to variability in the assay procedures nor solely the result of 
natural fluctuations in the patient’s anti-tumor cellular immune responses. 
2.5.4. Clinical Course 
There have been no significant adverse reactions at the immunization sites and no treatment related 
abnormalities have been observed on monitoring of complete blood counts, serum chemistries and 
urinalyses. Transient, mild erythema (<24 hrs) was observed at the injection site with immuninzations at 
[ 620 ] 
Recombinant DNA Research, Volume 19 
