A Large Majority of Donor NeoR-Expressing Gene Modified T-Cells Circulating More Than 9 Years after Infusion with allo-BMT, and Capable of Being In-Vitro Expanded and Cloned, Contain a Large Deletion of the HS-tk Gene and Arise from One T-Cell Clone in Each Patient.

We performed a clinical trial of HS-tk-expressing gene-modified T cell (GMTC) infusion at time of geno-identical allo-BMT. Among 12 patients (pts) initially included, 4 are still alive, off immunosuppression, free of chronic graft-versus-host disease (GvHD). More than 9 years after BMT, circulating GMTC can be found in all 4 patient’s mononuclear cells (PBMC), as detected by HS-tk or NeoR quantitative PCR assays. Although a similar sensitivity for both assays, GMTC were detected at a lower level with the HS-tk vs NeoR PCR assay, suggesting HS-tk gene deletions. The aims of our study were to

1. develop a molecular tool in order to precisely identify the areas of deletion,

2. characterize the mechanism of deletion and

3. determine if the deletions were present in the packaging cell line genome or occurred ex-vivo during GMTC preparation or subsequently in-vivo.

PBMC, harvested from the 4 pts (70, 106, 67 & 76 months post BMT) were polyclonally expanded, G418-selected and cloned. Of 32 NeoR+ clones, 31 (12, 3, 2 and 14 clones for pts #6, 7, 8 & 9, respectively) were HS-tk PCR negative, confirming the presence of HS-tk transgene deletions in a large majority (31/32) of long-term circulating Neo-R expressing GMTC. By contrast, all clones generated from freshly produced GMTC were found to be both NeoR+/HS-tk+, excluding an impact of the cloning process. Using a newly designed PCR-based transgene “walking” method (from the NeoR gene to the 5′LTR), we identified, among the 31 NeoR+HS-tk- clones, 4 different deletions, each one patient specific, involving the total HS-tk transgene and including partial flanking SV40 or packaging signal y sequences. These newly identified deletions differed from the spliced HS-tk form previously described (Garin et al., Blood 2001). Junction deletion areas were sequenced and involved, for 3 out of 4 patients, homologue sequence motifs (CCGCCC, AATTC and GATC respectively for pt# 6, 7 & 8), suggesting recombination mechanisms within the transgene sequences. With the help of deletion-specific PCR assays, we then confirmed that each deletion was patient-specific and not detected in the 3 other pts. Using more sensitive deletion-specific (nested) PCR assays, no deletions within packaging cell line DNA were detected. Nevertheless, these patient-specific deletions were found in GMTC products, albeit at a very low frequency, prior to infusion at time of BMT as well as 1, 3 and 6 years post infusion, independently of GCV treatment or anti-HS-tk immune response. Lastly, by means of LAM-PCR, we determined that all the clones of a same patient carried the same patient-specific insertional site, demonstrating that all the G418-reselected/cloned T cells of a patient derived from the same original GMTC. Overall, we establish that

1. patient-specific HS-tk gene deletions occur within the integrated retroviral sequence in GMTC prior to infusion;

2. HS-tk gene deletions are found in almost all Neo-R expressing long term circulating GMTC, thus strongly in favour of a in vivo selection advantage, by escaping an anti- HS-tk immune response and/or GCV treatment and

3. unique patient-specific HS-tk gene deletions and retroviral insertion site in GMTC capable of in-vitro expansion/cloning.

This last finding suggests an extremely rare initial gene deletion event and/or a possible in vivo growth/survival advantage of rare GMTC clones.