Molecular Analysis of T-Cell Receptor Gene Rearrangements in Children with T-Cell Acute Lymphoblast Leukemia on DFCI ALL Consortium Protocol 95-01.

During normal T-cell differentiation, T-cell receptor (TCR) genes undergo recombination in an ordered way. TCRd genes rearrange first followed by TCRg genes, which may result in TCRgd+ T lymphocytes. TCRab T lymphocytes develop via a separate differentiation lineage, with TCRb preceding TCRa gene rearrangements. Unlike in normal T-cells, TCRd, TCRg and TCRb genes rearrangements can be detected in T-cell acute lymphoblastic leukemia (T-ALL) and these rearranged gene sequences can be used as molecular targets to detect minimal residual disease (MRD). Knowledge of the association between rearrangement patterns and disease pathogenesis of pediatric T-ALL is relatively limited. We analyzed 45 sequences of the TCR gene rearrangements (17 TCRd and 28 TCRg) obtained at presentation from 37 children with T-lineage ALL and these sequences were used as molecular targets to perform real-time PCR (RQ-PCR). Seven children had both TCRd and TCRg gene rearrangements. Sequence analysis revealed a higher frequency of Vd1-Jd1 rearrangements (88%) compared to Vd2-Jd1 (6%) and Vd2-Dd3 (6%). Of the 28 TCRg sequences, Vg2 gene segment was identified in 7 sequences, followed by Vg4 (5), Vg3 (4), Vg8 (4), Vg1 (2), Vg7 (2), Vg10 (2), Vg5 (1), and Vg9 (1). In keeping with previous findings the median number of nucleotides in the junctional regions was higher in TCRd (28bp) than in TCRg sequences (7.5bp, p<0.0001) suggesting random nucleotide insertions and that Dd3 genes contribute a longer junctional region in TCRd gene rearrangement. Detailed analysis of junctional region of TCRg sequences revealed that this region is larger in patients who also have a TCRd rearrangement (n=7, median 11bp) compared to those patients for whom only a TCRg sequences was found (n=21, median 5bp, p=0.02), perhaps explained by a recent finding that lack of expression of TdT was associated with the absence or shorter N sequences which contributed to shorter CDR3 regions in immature lymphocytes. Sequence specific junctional region primers were designed and amplified with J gene consensus primers and RQ-PCR performed using J regionconsensus probes. 13 of 14 TCRd targets (93%) reached a high reproducible sensitivity of 10⁻⁵ and 10⁻⁶, whereas 10 of 18 TCRg targets (56%) showed such sensitivity (p=0.044, Fisher’s exact test). The junctional features of the TCR rearrangements for cases with only TCRg sequences and short N-regions were identified suggest that leukemic transformation is initiated at an early stage of T-cell ontology . The high junctional region diversity in TCRd rearrangements provides a better target to design patient specific primers (about 20bp) providing a higher sensitivity to detect MRD. RQ-PCR analysis was performed on 31 patients in whom BM (31) or PB samples (26) at day 30 were available for analysis. At this time point18 patients (58%) had no PCR detectable disease, six had MRD levels between 10⁻⁶ and 10⁻³ and 7 has MRD levels > 10⁻³. However, there was no difference in EFS in these three groups, and only one patient relapsed. We conclude that TCRg/d rearrangements provide suitable markers for RQ-PCR MRD analysis, but that in this treatment protocol high levels of MRD at the end of induction are not necessarily predictive of relapse in T-lineage ALL as they are in B-lineage ALL