The Incidence of Bcr-Abl Mutations and Their Impact on Clinical Outcome in Newly Diagnosed Chronic Phase Chronic Myeloid Leukemia Patients Treated with Bosutinib Versus Imatinib in the BELA Trial

Abstract 2769

In the phase 3 BELA study, bosutinib was associated with superior rates of major molecular response (MMR) at 12 months, faster times to MMR and complete cytogenetic response (CCyR), fewer on-treatment events of transformation to accelerated/blast phase CML, and fewer overall deaths on study when compared with imatinib; each agent was associated with an acceptable but distinct safety profile. However, clinical resistance to TKIs commonly results from development of Bcr-Abl mutations. Therefore, we investigated the incidence of Bcr-Abl mutations and their impact on clinical outcome in the BELA trial.

Newly diagnosed patients (pts) with CP CML were randomized to bosutinib 500 mg/day (n = 250) or imatinib 400 mg/day (n = 252). Direct sequencing of Bcr-Abl kinase domain (10%-20% sensitivity) in a central laboratory was employed at baseline (bosutinib, n = 237; imatinib, n = 241) as well as at treatment discontinuation (bosutinib, n = 62; imatinib, n = 49).

The median age was 48 years (range, 19–91 years) in the bosutinib arm and 47 years (range, 18–89 years) in the imatinib arm. Based on a minimum of 18 months from the last pt randomized, median treatment duration was 22.0 months for both treatment arms; 33% (n = 81) of pts in the bosutinib arm and 26% (n = 66) of pts in the imatinib arm have discontinued treatment. The primary reason for discontinuation of bosutinib was treatment-related toxicity (22% vs 6% with imatinib), while the primary reason for discontinuation of imatinib was disease progression (13% vs 4% with bosutinib); all other reasons for discontinuation were reported for ≤3% of pts in either arm.

At baseline, already present Bcr-Abl polymorphisms were equally distributed among the 2 arms (bosutinib, n = 24; imatinib, n = 17). The most common baseline polymorphisms were E499E (bosutinib, n = 18; imatinib, n = 10), 499E/E or 499E/E variant (bosutinib, n = 3; imatinib, n = 1), and T83T (bosutinib, n = 0; imatinib, n = 4); additional polymorphisms detected in at least one of the treatment arms included E197K, K247R, T240T, and Y320C.

As of the database snapshot, 81 pts had discontinued treatment in the BOS arm compared with 66 pts in the IM arm, of which 62 (77%) pts in the BOS arm and 49 (74%) pts in the IM arm had a sample collected at treatment discontinuation. Of these, 37/62 (60%) pts in the BOS arm did not have a detectable mutation and 28/49 (57%) pts in the IM arm did not have a detectable mutation.

Four (6%) pts had a detectable mutation in the BOS arm (T315I [n = 2], E255K, and V299L [n = 1 each]) compared with 10 (21%) pts in the IM arm (G250E [n = 2], M244V [n = 2], Q252H, Y253H, E255K, M244V/M351T, D276G/H396R, and D276G/T277A/T315I [n = 1 each]) at the time of treatment discontinuation. In the BOS arm, 1 pt with the V299L mutation achieved CCyR (time to CCyR of 254 days). In the IM arm, 1 pt with the G250E mutation achieved both CCyR and MMR (time to response of 252 days for both) and 2 pts with the M244V and D276G/T277A/T315I mutations achieved CCyR (times to CCyR of 254 and 340 days, respectively). Treatment failure was experienced by 2/4 pts in the BOS arm with a mutation (E255K and T315I) compared with 7/10 pts in the IM arm (M244V [n = 2], G250E, Q252H, Y253H, M244V/M351T, and D276G/H396R). An on-treatment event of transformation to accelerated/blast phase CML was experienced by 2/4 pts in the BOS arm (E255K and T315I) versus 5/10 pts in the IM arm (G250E, Q252H, Y253H, M244V/M351T, and D276G/H396R). Among pts with a new mutation detected at treatment completion, 2 deaths were reported in each treatment arm; all 4 deaths occurred after discontinuation of study treatment.

In conclusion, baseline and emergent Bcr-Abl kinase domain mutations were detected in both treatment arms, with polymorphisms at the E499 position the most common. In this analysis, imatinib was associated with a higher rate of progressive disease compared with bosutinib among pts with new mutations at treatment discontinuation; a similar trend has been observed in the overall study population. Additional follow-up will allow for more complete mutational analysis.