Philadelphia-Positive Acute Lymphoblastic Leukemia Patients Already Harbor Bcr-Abl Kinase Domain Mutations at Low Levels at the Time of Diagnosis - a Report by the GIMEMA ALL Working Party

In Philadelphia-positive (Ph+) acute lymphoblastic leukemia (ALL) patients (pts) treated with tyrosine kinase inhibitors (TKIs), responses are generally short-lived and relapse is most often accompanied by selection of point mutations in the Bcr-Abl kinase domain (KD). In order to assess whether mutations are already present at the time of diagnosis, we analyzed RNA samples from 13 pts with newly diagnosed Ph+ ALL enrolled on the GIMEMA LAL1205 protocol – designed to assess activity, safety and tolerability of dasatinib monotherapy as first-line treatment for adult Ph+ ALL. The pts investigated (males/ females: 6/7; median age: 56 years, range 30–73) included 6 pts who subsequently relapsed with dasatinib-resistant mutations, 2 pts who relapsed without evidence of mutations as assessed by direct sequencing and 5 pts in persistent remission. Screening for low level mutations was performed by cloning the entire Bcr-Abl KD (amino acid 206 through 524) in a bacterial vector and sequencing 150 to 200 independent clones for each pt. All pts, irrespective of their subsequent response to dasatinib, had evidence of aberrant KD sequences due to point mutations (13/13 pts), small insertions (2/13 pts) or small deletions (1/13 pts). Two to five point mutations were detectable for each pt. In some cases, multiple mutations could be found to co-exist in the same cloned fragment. A total of 45 different point mutations (including 18 silent mutations, 4 nonsense mutations and 23 missense mutations) were observed. Base substitutions were scattered all over the KD and there were no mutation hotspots. Thirty-nine out of 45 (87%) mutations were transitions: G>A (n=14), A>G (n=10), C>T (n=9), T>C (n=6). Such a high prevalence of transitions (which normally occur 1.4 times more frequently than transversions) suggests that a specific mechanism generating mutations is predominant. Activation-induced cytidine deaminase (AID) has been recently shown to be aberrantly expressed in a proportion of Ph+ ALLs and has been implicated in the development of at least some Bcr-Abl mutations; correlation with AID expression will be presented. Mutations were detected in no more than three independent clones, suggesting that they were present at low levels. Mutations detected in less than two clones were considered only if independently validated by a specifically designed ASO-PCR assay run on the original cDNA sample. The majority of point mutations detected have never been reported in association with TKI resistance. However, three pts were found to harbor known imatinib- or dasatinib-resistant mutations, including two T315I mutations that were later on selected and led to relapse. This supports the theory that mutations randomly arise as a consequence of a high genetic instability and very few of them will confer a growth advantage under the selective pressure of TKIs. Six samples from newly diagnosed chronic phase CML (CP-CML) pts were screened for comparison with the same cloning approach, including 2 pts who later progressed with evidence of KD mutations after 3 and 6 months from the start of imatinib, respectively, and 4 pts who achieved a stable molecular response on imatinib. Both pts who experienced early relapse on imatinib had low level mutations detectable at the time of diagnosis, whereas none of the clones from CP-CML imatinib responders had evidence of KD mutations. The results herein presented support the concept that Ph+ ALL cells have a particularly high genomic instability that allows early escape from TKI inhibitor therapy. Only a fraction of CML cases share this feature. Our data highlight the importance of understanding the mechanism(s) responsible for this ‘mutator’ phenotype, as well as how to interfere therapeutically with it.