Ultra Deep Sequencing (UDS) Allows More Sensitive Detection Of Tyrosine Kinase Inhibitor (TKI)-Resistant BCR-ABL Mutations That Would Influence Therapeutic Decision At The Time Of Switchover To Second- Or Third-Line Therapy

According to the 2013 European LeukemiaNet recommendations, BCR-ABL kinase domain (KD) mutation screening by conventional Sanger sequencing (SS) is recommended in all patients (pts) with progression, failure or warning, since detection of second-generation (2G) TKI-resistant mutations (T315I, F317L/V/I/C, V299L, Y253F/H, E255K/V, F359V/I/C) helps in the selection of the most suitable alternative therapy. However, SS has limited sensitivity. A recent study (Parker et al, J Clin Oncol 2011) has shown that low level mutations (i.e., undetectable by SS) may be identified by mass spectrometry and this may aid in more appropriate selection of the second-line treatment strategy for imatinib-resistant pts.

We sought to determine i) whether an UDS-based approach for BCR-ABL KD mutation screening might allow more sensitive detection of 2GTKI-resistant mutations at the time of switchover to second or third-line therapy and ii) whether low level mutations identified by UDS undergo clonal expansion, thus predicting for subsequent failure, if the 2GTKI to which they are insensitive to happens to be chosen.

To this purpose, we used UDS to scan the BCR-ABL KD in a cohort of 75 chronic myeloid leukemia (n=50) or Philadelphia chromosome-positive acute lymphoblastic leukemia (n=25) pts who were treated with a 2GTKI (dasatinib or nilotinib) after having failed first-line (n=62) or second-line (n=13) TKI therapy. A total of 235 samples collected at the time of switchover and at subsequent timepoints during second-or third-line treatment were analyzed. All the samples had already been evaluated by SS. UDS was performed on a Roche GS Junior instrument, according to an amplicon sequencing design and protocol set up and validated in the framework of the IRON-II international study. Runs were designed to achieve high sequencing depth. However, raw data analysis with Amplicon Variant Analyzer software (Roche Applied Science) was performed filtering out variants <1% to reduce the likelihood of amplification artifacts and sequencing errors.

Forty-three imatinib-resistant pts failed subsequent second- (n=37) or third-line (n=6) treatment with 2GTKIs (dasatinib, n=28; nilotinib, n=15). By routine SS analysis, 32 BCR-ABL KD mutations had been identified in 29/43 (67%) pts at the time of switchover - informing subsequent-line treatment selection in 12 pts who were found to harbor 2GTKI-resistant mutations. At the time of treatment failure (after a median of 9 months; range, 2-30), SS had detected newly acquired 2GTKI-resistant mutations in all 43 pts. We thus wondered how many of these mutations could have been detected at the time of switchover using a more sensitive approach. By UDS re-analysis, 73 mutations were identified in 35/43 (81%) pts at switchover – UDS detected all the 32 mutations previously identified by SS plus 41 low level mutations (i.e., with an abundance between 1 and 15%). Twenty-four of the 41 low level mutations were resistant to the 2GTKI that happened to be selected (T315I=10; E255K=2; E255V=1; Y253H/F=4; F359V=1; F317L=4; V299L=1) and all invariably expanded becoming detectable by SS at relapse, alone or in combination with pre-existing mutations. Thus, mutations that would have influenced therapeutic decision after first or second TKI failure could have been detected in 17 more pts by UDS (p<0.001).

Thirty pts who achieved a stable response to second- (n=26) or third-line (n=4) treatment with 2GTKIs after having failed first- or second-line treatment, respectively, were also analyzed for comparison. By UDS, a total of 13 mutations were detected – including 6 mutations (6 pts) that had already been identified by SS plus 7 low level mutations (7 pts). No low level mutation resistant to the 2GTKI the pts actually received was detected.

In comparison to SS, UDS could identify 2GTKI-resistant mutations at the time of switchover in a significantly greater proportion of pts. Low level mutations (down to 1% abundance) detected by UDS were consistently found to expand when the 2GTKI they are insensitive to happened to be chosen. UDS-based screening of the BCR-ABL KD at switchover might thus have enabled more effective therapeutic tailoring. Further evaluation of whether this technology may be implemented in a diagnostic setting is highly warranted.

Supported by PRIN 2009 (prot.2009JSMKY), Fondazione CARISBO, AIL, AIRC, FP7 ‘NGS-PTL’, IGA NT 11555 and 13899.

Soverini:Novartis: Consultancy; Bristol-Myers Squibb: Consultancy; ARIAD: Consultancy. Machova Polakova:Novartis: Honoraria, Research Funding; Bristol-Myers Squibb: Honoraria, Research Funding. Castagnetti:Novartis: Consultancy; Bristol-Myers Squibb: Consultancy. Gugliotta:Novartis: Consultancy, Honoraria; Bristol-Myers Squibb: Consultancy, Honoraria. Rosti:Novartis: Consultancy, Speakers Bureau; Bristol Myers Squibb: Consultancy, Speakers Bureau; Ariad: Consultancy, Speakers Bureau; Roche: Speakers Bureau; Pfizer: Speakers Bureau. Baccarani:Novartis: Consultancy, Membership on an entity’s Board of Directors or advisory committees, Speaker fees Other; Bristol-Myers Squibb: Consultancy, Membership on an entity’s Board of Directors or advisory committees, Speaker fees, Speaker fees Other; Ariad: Consultancy, Membership on an entity’s Board of Directors or advisory committees, Speaker fees, Speaker fees Other; Pfizer: Consultancy, Membership on an entity’s Board of Directors or advisory committees, Speaker fees, Speaker fees Other; Teva: Consultancy, Membership on an entity’s Board of Directors or advisory committees, Speaker fees Other. Martinelli:Novartis: Consultancy, Speaker fees Other; Bristol-Myers Squibb: Consultancy, Speaker fees, Speaker fees Other; Pfizer: Consultancy, Speaker fees, Speaker fees Other; Ariad: Consultancy, Speaker fees, Speaker fees Other.