2). Not all human TIL are cytotoxic In vitro . However, if the TIL population 
demonstrates cytotoxicity for autologous tumor, vector infection and G418 
selection does not alter either the magnitude of cytolytic capacity or the target 
specificity of the TIL (Table 3). Further, the types of cytokine produced by TIL 
(determined by Northern Blot) are not altered by vector infection or G418 
selection (data not shown) . Southern Blot analysis of TIL populations from two 
patients also show variable patterns. In patient 1, T cell receptor beta chain 
gene rearrangements of both untreated and vector treated/selected TIL show the 
identical oligoclonal pattern of T receptor utilization and the probe for vector 
insertions shows a smear with no predominant clones. In patient 5, probes for 
TCR beta also shows an oligoclonal pattern of receptor utilization which is 
unchanged by vector infection/selection. In this N7 marked TIL population which 
had been maintained in culture for a prolonged period, in addition to the smear 
found when probing for vector inserts, two distinct minor bands were also visible 
which indicate some degree of clonal preference had occurred with prolonged 
culture (Table 4) . 
Although there is variability from patient to patient in the effect that 
vector insertion/selection has had on the TIL population, the gene marked TIL are 
sufficiently representative of the whole TIL population to permit us to answer 
these primary questions: 
(1) How long do TIL persist in vivo ? 
(2) Where do they persist: lymph nodes, tumor, blood stream? 
(3) Does the location and/or longevity of TIL correlate with the 
clinical effect? 
If we are also able to isolate and grow (in G418) vector-marked TIL from a 
patient's tissues as we can grow vector-marked murine T cells isolated from 
recipient mice, then we can also ask another critical question: 
(4) What functional TIL cell types (i.e., cytotoxic, 
proliferative, cytokine producing, etc.) correlate with the 
clincial effect? 
b) Our data indicate that vector-marked human TIL are free of helper 
virus. Of primary importance, as pointed out by the Subcommittee, is that the 
assays used to detect infectious retrovirus are the most sensitive available. The 
assays we are using to evaluate TIL for helper virus are listed below. Detailed 
protocols, as requested by the Subcommittee, are described in Appendix A. 
(1) In our own laboratory we will perform the S+/L- assay to determine if helper 
virus is present. In addition, 3T3 cells will be infected with viral supemate 
and cultured so that virus in the test material will be greatly amplified and 
easily detected by the S+/L- assay. Besides testing retroviral supernate, 
transduced TIL will also be studied. Medium from TIL cultures will be tested by 
the S+/L- assay and by 3T3 amplification. Transduced TIL will also be cultured 
directly on 3T3 cells. Viral envelope sequences will also be looked for using the 
polymerase chain reaction. Finally, the reverse transcriptase assay will be 
performed to detect possible, though very unlikely, recombination between human 
Recombinant DNA Research, Volume 13 
[ 211 ] 
