The most important determination of cell purity is histologic evaluation. 
Malignant renal tumors have histologic characteristics that distinguish them 
as malignant cells of epithelial origin. Their cytoplasm often contains many 
granules (granular cell) or lipid deposits (clear cell) that differentiates 
them from cells of mesenchymal origin such as fibroblasts. Although it is 
sometimes difficult to distinguish normal from malignant cells in culture, it 
is unusual to have normal contaminating epithelial cells within the excised 
mass. Furthermore, when normal renal epithelium growth is compared with 
malignant renal epithelium growth in vitro, the normal exhibits contact 
inhibition whereas the malignant does not. 
Besides normal renal cells of epithelial origin, other cells of non- 
epithelial origin can also contaminate the primary tumor cell cultures. 
Lymphocytes, macrophages and red blood cells are often present in the initial 
cultures especially when the tumor is necrotic. But these cells do not 
propagate under the renal tumor growth conditions. Fibroblasts are a concern 
in many types of primary tumor cultures, but not a great concern in renal 
tumors due to the minimal amount of reactive tissue present at the site of 
these tumors. However, histochemical staining is performed routinely with 
anticytokeratin antibodies to distinguish cells of epithelial origin from 
cells of mesenchymal origin. Ten consecutive primary renal tumor were 
evaluated for cytokeratin expression at the time of enzymatic digestion and 
again after in vitro passage. Positive cytokeratin staining ranged from 40 to 
95% initially, and increased to between 60 and 100% by in vitro passage 4, 
with the majority of cultures demonstrating greater than 80% cytokeratin 
expression (Table 10) . These results strongly confirm that all of the primary 
cultures are composed of a majority of cells of epithelial origin. 
Establishing the safety of autologous GM-CSF-secretinq human tumor 
vaccines . 
Also critical to the development of a human cytokine-secreting tumor 
vaccine is safety after re-injection of the tumor cells back into the patient. 
We therefore evaluated the primary human renal cell tumor explants for 
in vitro growth and GM-CSF expression following lethal doses of irradiation. 
Initially 5 fresh human renal tumor cultures were evaluated for the optimal 
dose above which none of the tumor cells proliferated in culture. The results 
are shown in Table 11. A radiation dose of 15,000 rads or greater resulted in 
100% inhibition of proliferation by day 3 following radiation. All of the 
tumor cells in culture were noted to die by day 28 following irradiation. 
Next we evaluated GM-CSF secretion by the transduced tumor vaccine cells after 
receiving 5,000, 10,000 or 15,000 rads, and compared the results with GM-CSF 
production by the unirradiated vaccine cells. The results for one renal tumor 
vaccine are shown in Figure 5. Despite even 15,000 rads, all of the tumors 
evaluated were able to secrete at least as much human GM-CSF as the 
unirradiated, transduced vaccine cells when comparisons were made 1 day and 4 
days after irradiation. Therefore, it is possible to inhibit proliferation of 
the RCC vaccine cell population without affecting expression of the GM-CSF 
gene. Other pre-clinical toxicity data is presented in Appendix 14. A. 2. 
Recombinant DNA Research, Volume 17 
[301] 
