F. Autoimmune Disease . A wholly theoretical risk exists that expression of the protein 
neomycin phosphotransferase could lead to an immune response that would be 
generalized to other (non-transduced cells) in the body. The extensive existing 
literature includes no evidence for such a phenomenon occurring in animals, nor has 
such an effect been observed in humans in gene therapy trials to date (2,42,55). 
G. Risks related to harvesting of cells and re-infusion . The risks of venipuncture include 
bruising and minor bleeding. The amount of blood that will be phlebotomized should not 
cause volume depletion hemodynamically significant enough to result in syncope or to 
significantly exacerbate symptoms of mild chronic anemia attendant to HIV infection 
(initial eligibility hematocrit cut-off is 30%). Nevertheless, patients will be observed 
for at least one hour following phlebotomy and examined for clinically significant 
orthostatic blood pressure changes. 
Since harvested patient cells will undergo expansion and manipulation in vitro, there is 
a risk of microbial contamination with bacteria, mycoplasma, fungi or viruses. All 
procedures will be optimized to maintain sterility of patient cells. Monitoring of cells 
prior to re-infusion will include: viability (at least 70% trypan blue exclusion), 
culture for helper virus as indicated above, surveillance for aerobic and anaerobic 
bacterial and fungal contamination by culture in thioglycolate broth and on blood agar 
and chocolate agar. A probe hybridization test will be used to test for mycoplasma and 
other prokaryotic organisms. The standard limulus assay is planned to test for endotoxin. 
A gram's stain of the final cell pellet will be done prior to infusion to confirm the 
absence of organisms. Finally, to insure that all components of the protocol are in place, 
we plan to dry-run the procedure at least once on human PBLs before the first patients 
are treated. 
H. Confidentiality . Every attempt will be made to preserve strict confidentiality and 
privacy throughout the study. No patient identifiers will be included in case report 
forms or other communications or in laboratory labelling. Patients will be assigned 
confidential study identifiers which will be used in reports and publications as well as in 
laboratory processing of patient material. Standard procedures for any communication 
with the press will be followed including no identification of the patient or details of 
their medical problems other than those obvious from the study entry criteria. 
Recombinant DNA Research, Volume 18 
[397] 
