Monozygotic Twins with Childhood ALL Demonstrate Prenatal Origin of the Philadelphia Chromosome and Post-Natal Ikaros Deletion.

Abstract 87

Studies on twin pairs with concordant ALL have demonstrated the pre-natal occurrence of chromosomal aberrations associated to leukemia either for the rare ALL subsets with short latency, such as t(4;11), but also for more prevalent, later and less aggressive subtypes, such as t(12;21) or hyperdiploidy. SNP-array studies indicated that additional cooperative lesions occur in ALL, including hyperdiploidy, TEL/AML1 and t(9;22) subsets, and they functionally complement the pre-leukemic hit for the conversion to overt ALL. This seems to be true also for Ph+ ALL; although the very dismal prognosis indicates that the BCR/ABL fusion gene could be sufficient to drive the leukemia, SNP-array showed a frequent association with the deletion of the Ikaros gene (IKZF1), among others. However, so far a formal proof for the succession of events in Ph+ ALL was not available.

We report a monozygotic twins pair (T1 and T2) with Ph+ ALL, diagnosed at the age of 32 and 36 months, respectively, and treated according to the AIEOP-BFM ALL2000 protocol. This twin pair was a model to demonstrate for the first time the pre-natal occurrence of the t(9;22) translocation and the post-natal origin of Ikaros gene deletion.

A different pattern of Ig/TcR clonal rearrangements was found at diagnosis, although a ‘cross-test' identified a marker shared by the twins, being present in a small subpopulation in T2. Formal demonstration of common pre-natal origin of Ph+ ALL came from the cloning of the BCR/ABL genomic breakpoint, which was identical in both twins.

The two twins had a different early response to treatment, being T1 prednisone good responder and T2 poor responder, both received BMT from the same sibling donor, and had different outcome: while T2 died, T1 is in CR after 6 years.

Early MRD monitoring by using Ig/TcR markers in parallel with a specific genomic RQ-PCR assay indicated persistently high MRD levels in T2 by both methods. By contrast, a persistently high level of BCR/ABL genomic signal was present in the follow up of T1, despite morphological remission and Ig/TcR and immunophenotypic negativity at both day 33 and 78.

With the aim to assess the nature of Ph+, Ig/TcR negative cells present in T1 remission sample, we performed FISH analysis on sorted bone marrow cells. We found that 51% of CD3+ and 45% of CD34-/19+/20+ (40% and 6% of total CD45 sorted population, respectively) were BCR-ABL+ and disomic, thus demonstrating that Ph+ cells were mature lymphoid cells. By contrast, the myeloid compartment was not affected, since CD34-/33+ and CD34+/19-/33+/38+ cells were all BCR-ABL negative.

The Affymetrix GeneChip mapping analysis by 500K SNP array identified several and different additional abnormalities: T1 showed a subclone with high hyperdiploid karyotype, while T2 displayed deletion of IKZF1 (Δ4-7) and EBF1 genes. In addition, both twins showed identical deletions affecting ABL and BCR exons, confirming the common origin of the translocation. A patient-specific genomic RQ-PCR assay did not reveal the presence of the IKZF1 deletion in T1, demonstrating the post-natal occurrence of this lesion as a second genetic event in Ph+ cells of T2.

In conclusion, the Ph chromosome originated in utero in a lymphoid committed precursor cell, generating the pre-natal non-leukemic Ph+ clone shared by the twins. After birth, a second independent hit transformed the Ph+ clone in T1 and T2, inducing hyperdiploidy in T1 and IKZF1 deletion in T2, with consequent different outcome. Residual pre-natal, non leukemic Ph+ cells persisted during normal lymphopoiesis.

This twin pair represents a life model for the prenatal origin of the Ph chromosome as well as the occurrence of additional post-natal genetic events.