Recombinant DNA Advisoiy Committee - 2/10-11/92 
possible. If a gene could be inserted into a more primitive cell that produces educated, 
functional, peripheral T lymphocytes, this procedure would provide a broader-based 
repertoire with fewer immunological deficiencies. Dr. Anderson indicated that the largest 
safety issue is that there are deficiencies in the immune repertoire of these patients. 
A discussion commenced on whether the holes in the immune system can be identified and 
how the investigators will prove that this protocol is more effective than infusing the 
peripheral T lymphocytes. Dr. Culver responded that there are ways, both in vivo and in 
vitro , to measure further enhancement of immune function. All of the patients' 
specificities cannot be realistically measured and identified if specific deficiencies exist. In 
vitro proliferation studies on the first patient indicate that she did not make normal 
responses to any of the antigens that were a part of the standard assays. She exhibited 
delayed type hypersensitivity to Candida, diphtheria, and tetanus. However, there were a 
number of parameters which have been routinely followed that indicate that the patients 
are not immunologically normal. The revised protocol would help demonstrate that 
reconstitution is more significant as a result of CD34 selection techniques. 
Dr. Dunbar responded to the question of G-CSF toxicity. G-CSF would have to be 
discontinued if the white count reached 50,000. There is no data to indicate the best day 
for apheresis and what indicator to use, whether it should be a total white count, number 
of days, or CD34 positivity. Therefore, the investigators intend to search for 1% CD34 
positivity. If 1% CD34( + ) cells are not achieved by the seventh day, then G-CSF would 
be discontinued to avoid prolonged exposure, and three apheresis would be performed on 
consecutive days. In addition, if a patient developed any grade three or four 
nonhematologic toxicity, the investigators would discontinue therapy. If a lower toxicity 
developed, it would be mitigated, if possible, by lowering the dose. Overall, a relatively 
low dose of G-CSF will be used. 
Dr. Geiduschek addressed a fundamental issue of whether a protocol that has been 
approved by the RAC should be altered in mid-course to the extent that the conclusions 
from the original study were inconclusive. The investigators have a commitment to 
continue the original protocol to its conclusion unless there is a fundamental indication 
that patients are threatened. Conclusive results of the patients' treatment in the original 
protocol will never be determined if the protocol changes direction midway. Dr. Anderson 
emphasized that when one is considering human patients, rather than laboratory 
experiments, the primary criteria for treatment determination are what is in the best 
interest of the patients. If it is in the best interest of the patients to alter a protocol, it 
would not be appropriate for the RAC to insist that a scientific question be explored in the 
same manner that it had originated. Dr. Geiduschek responded that it is unclear that this 
amendment to the protocol is an obvious case of additional benefit to the patient, and that 
the patients appeared to be responding vigorously to treatment. Dr. Anderson reiterated 
[ 328 ] 
Recombinant DNA Research, Volume 15 
