Observed 
Estimated 
Exact 95% Confidence 
Outcome 
Incidence 
Interval 
MTD (0/3) 
0 % 
0 -63 % 
MTD (1/6) 
17 % 
0 -66 % 
DLT (2/6) 
33 % 
2 - 70 % 
DLT (2/5) 
40 % 
3 - 77 % 
DLT (2/4) 
50 % 
6 - 89 % 
DLT (2/3) 
67 % 
10 - 99 % 
DLT (2/2) 
j 100 % 
22 - 100 % 
10.0 STUDY INTERPRETATION 
Since this is a Phase I study the main aim will be to collect information about dosage 
and immunomodulatory and virological efficacy for a future Phase II study. Potential 
problems in this preliminary Phase I study would include toxicity with development of 
GVHD or lack of efficacy as evidenced by absence of immunomodulatory effects. 
10.1 Excessive Toxicity 
If administration of CTL was associated with an unacceptably high risk of 
GVHD we would modify the study in two ways. First we would clone the CTLs 
and administering only clones that lysed EBV infected cells but not other 
allogeneic or autologous cells. Second we would transduce the CTLs with the Tk 
vector, so that the CTLs may be killed by administration of Acyclovir should 
toxicity occur. This approach presents some logistic difficulties as all CMV + 
recipients are at present on Ganciclovir prophylaxis for CMV. However in the 
next year we plan to start a trial of Foscamet prophylaxis for CMV so this 
approach would be possible if required. 
10.2 Lack of Efficacy 
If patients on the lower dose schedules developed EBV lymphoproliferation, this 
would not necessarily be regarded as a failure as outgrowth of EBV infected B 
cells may reflect insufficient numbers of adoptively transferred CTLs. If however 
patients receiving higher doses developed lymphoproliferation or failed to show 
immunomodulatory effects we would aim to enhance the activity of the 
transferred CTLs Possible approached would include transducing the CTLs with 
cytokine genes such as IL1 or cytokine receptor genes. Such manipulations have 
Recombinant DNA Research, Volume 17 
[ 211 ] 
