3.0 
STUDY DESIGN AND TREATMENT PLAN 
3.1 OVERVIEW OF RESEARCH OBJECTIVES 
The main objective of this trial is to evaluate the safety of RCC cell 
injections prepared with and without GM-CSF gene transduction. Emphasis is 
placed on documenting the adverse effects of vaccinations on tissues at or 
near the site of injection, as well as acute and longer-term systemic 
toxicities. As discussed in the previous section, such an independent 
assessment of maximum tolerated cell and cytokine dose will be critical for 
the selection of an optimal RCC cell preparation and schedule of 
administration for phase II trials of therapeutic efficacy. The two armed 
study design is employed specifically to allow distinction of toxicities due 
to cell administration from those due to local GM-CSF production by the 
transduced tumor cells. 
Also critical is evaluation of whether or not tumor vaccines non- 
transduced or transduced with the GM-CSF gene can induce a specific, antitumor 
immune response in patients. The parallel dose escalation arms in the study 
design may additionally aid in the evaluation of the contribution of GM-CSF 
gene transfer to antitumor immune response induction. A number of points bare 
on this issue. First, at higher cell doses irradiated tumor vaccines have 
significant anti-tumor immunogenicity in some pre-clinical tumor models 
without GM-CSF gene transduction, as discussed in Section 2. The short-term 
culture technique developed for this study does not employ any drug selectable 
marker and thus maintains the potential representation of antigenic diversity 
while expanding cell numbers up to doses not previously studied in humans. 
Second, non-transduced RCC tumor cells in these short-term culture conditions 
also secrete low levels of GM-CSF as discussed in Section 2 — albeit far below 
those levels set for GM-CSF gene transduced cells in this trial. Contribution 
of GM-CSF gene transfer itself to toxicities that follow can best be studied 
by this seperation of cell dose from cell dose plus high-level GM-CSF 
secretion. Third, although the patient groups are small and are not intended 
to be equalized for many variables (see Biostatistics, Section 5), large 
differences in biological outcome between the two treatments in a pilot study 
could shape phase II trial design if equal patient safety is established. 
Lastly, this human trial is intentionally designed to test the relevance 
of the murine gene therapy tumor models that engendered it. If predictions 
created by these models are confirmed in this trial, specifically that the 
local expression of GM— CSF by gene transduced, irradiated tumor vaccines 
induce human tumor specific immune responses more readily than non— transduced 
vaccines, then this experimental research direction can be welcomed as 
relevant to clinical cancer therapy. 
Recombinant DNA Research, Volume 17 
[309] 
