Attachment IV - Page 33 
- 11 - 
DRAFT 
NOVEMBER 7, 1983 
1. Pyrogen Contamination 
Pyrogenlclty testing should be conducted by traditional methods of 
endotoxin detection with limulus amebocyte lysate (LAL) and direct pyrogenlclty 
testing by injection of rabbits with the final product. Criteria comparable to 
those adopted for acceptance of the natural product should be used for the DNA 
product . 
Certain biological pharmaceuticals may be pyrogenic in humans 
despite having passed the LAL test and the rabbit pyrogen test. This phenomenon 
may be due to nonendotoxin contaminants which appear to be pyrogenic only in 
humans. To attempt to predict whether human subjects will experience a pyrogenic 
response, human blood mononuclear cells can be cultured lu vitro with the final 
product and the cell culture fluid injected into rabbits. A fever in the rabbits 
indicates that the product contains pyrogenic substances. 
2. Viral Contamination 
Tests for viral contamination should be appropriate to the cell 
substrate and culture conditions employed. Absence of contamination of the final 
product by adventitious viruses should be demonstrated. 
3 . Nucleic Acid Contamination 
Removal of nucleic acid at each step in the purification process 
should be demonstrated in pilot experiments by examining the extent of elimination 
of added radiolabeled host cell DNA. This analysis will provide the theoretical 
extent of the removal of nucleic acid during purification. 
A direct analysis of nucleic acid in the final product should be 
performed by hybridization analysis of immobilized contaminating nucleic acid 
utilizing appropriate probes, such as both nick-translated host cell and vector 
DNA. This method ought to provide sensitivity to the level of 10 picograms per 
[ 60 ] 
