Clonal Evolution and Lack of BCR-ABL1 Mutations in Pediatric Ph+ ALL Patients Resistant/Refractory to Imatinib Treatment

Philadelphia positive acute lymphoblastic leukemia (Ph+ ALL) is characterized by the translocation t(9;22), resulting in the BCR-ABL1 fusion gene. It occurs in 2-3% of pediatric ALL and is associated with poor prognosis. Despite the introduction of tyrosine kinase inhibitors (TKIs), such as Imatinib, resulting in a significant increase in the cure rate, a consistent number of patients show resistance to treatment and subsequently relapse. In an adult cohort, more than 70% of relapses during TKI treatment involve selection of tyrosine kinase domain (TKD) mutations.

In order to better understand the mechanism of Imatinib resistance in pediatric Ph+ ALL, we selected as our cohort a subgroup of patients treated with Imatinib on top of chemotherapy, according to the EsPhALL clinical trial and enrolled in Italy from January 2004 to December 2014 by the AIEOP study group. Patients (n=10) either didn't achieve remission or experienced early relapse by the end of 2014. Patients' samples have been screened for mutations in the ABL1 tyrosine kinase domain (TKD). Nine out of ten BM samples collected at relapse showed wild type TKD. Accordingly, no subsequent relapses showed the appearance of an ABL1 mutated clone. Only one patient carried the tyrosine 243 to phenylalanine substitution (Y253F) at the first and second relapses, but not at diagnosis, demonstrating the expansion of a resistant minor clone at diagnosis, following TKI treatment.

In addition, we monitored the BCR-ABL1 transcript level by RQ-PCR at diagnosis, relapse and multiple follow up points. In nine out of ten patients the expression of BCR-ABL1 was highly positive at the time of the relapse, and in five out of six cases the increase of the transcript level was predictive of a subsequent clinical relapse.

Moreover, we investigated the cooperative copy number alterations by MLPA analysis in 6 matched diagnosis and relapse(s) samples. MLPA data showed a heterogeneous scenario of clonal evolution. In three patients, the relapse clone shared only a minor genetic identity with the predominant clone at diagnosis and wasn't the product of its direct evolution. In one case, the clone at relapse was identical to that at diagnosis, while in two cases the predominant clone at relapse was the product of the direct clonal evolution of the main clone at diagnosis. In 4 cases, the dynamics of clonal evolution were confirmed by the analyses of the IG/TCR rearrangements.

In conclusion, our data depict a very different scenario of resistance in childhood Ph+ ALL compared to the adult cohort. Indeed, Imatinib resistance doesn't seem to be dependent on ABL1 mutations in pediatric Ph+ ALL. Moreover, the MLPA results contribute to the description of clonal heterogeneity in Ph+ ALL, and may partially explain the different response to treatment.