Long-Term Safety and Effectiveness of Ponatinib Treatment in Patients with TKI Intolerance: Subgroup Analysis of the Observational Study of Ponatinib Treatment in Patients with CML in Italy (OITI)

Introduction: Although most patients (pts) with chronic myeloid leukemia (CML) have favorable outcomes with second-generation tyrosine kinase inhibitors (2G-TKIs), 20-30% change treatment (tx) due to resistance or intolerance. Pts who experience intolerance to 2G-TKIs develop tx-related adverse events (AEs) that often result in reduced tx adherence, leading to decreased dose or tx interruption. This is associated with a possible loss of response and an increased risk of disease progression. Ponatinib (PON), a 3G-TKI, has a well-established efficacy in resistant pts in later lines, with >10 years (y) of real-world (RW) clinical experience. However, limited RW data have been reported for PON tx in pts with CML intolerant to 2G-TKIs. The OITI trial evaluated RW safety and effectiveness of PON in pts with CML in Italy. Here, we present results of a post hoc subgroup analysis of OITI evaluating long-term safety and effectiveness of PON in pts with TKI intolerance.

Methods: OITI is a noninterventional study in pts with CML (≥18 y) who started PON in Italian clinical practices. The primary endpoint was complete cytogenetic response (CCyR) rate within 6 months (mo) from start of PON. In the absence of cytogenetic evaluation, molecular assessment was evaluated. Of 110 pts with chronic phase (CP) CML, 36 were intolerant to their previous TKIs and were included in this analysis.

Results: Thirty-six pts with CP-CML and intolerance to a previous TKI switched to PON; last tx before PON included dasatinib (n=26), bosutinib (n=5), nilotinib (n=4), and imatinib (n=1). Median (range) duration of last tx before PON was 2.8 (1.0-4.8) y. Most common AEs leading to discontinuation of prior TKI were pleural effusion (dasatinib [n=15], bosutinib [n=1]), hepatic toxicity (nilotinib [n=2], bosutinib [n=1]), hematologic toxicity (dasatinib, nilotinib, bosutinib [n=1 each]), and gastrointestinal toxicity (bosutinib [n=2]).

Median (range) age at start of PON was 61 (29-84) y. At diagnosis, of evaluable pts (ie, assessed with known result), 6 pts (23%) had low, 15 (58%) had intermediate, and 5 (19%) had high Sokal risk. Most pts presented with ≥1 cardiovascular (CV) risk factor, most commonly hypertension (n=26 [72%]), hypercholesterolemia (n=8 [23%]), and type 2 diabetes (n=2 [6%]). Starting doses of PON were 45 mg (n=7), 30 mg (n=14), and 15 mg (n=15). Fourteen pts (39%) received PON in second line (2L), 16 (44%) in 3L, and 6 (17%) in ≥4L. Median (range) duration of PON was 3.1 (2.6–4.7) y. Of 35 evaluable pts at the start of PON, 10 (29%) had no molecular response (MR), 4 (11%) had MR2, 12 (34%) had MR3, and 9 (25%) had DMR (MR4/MR4.5/MR5). Percent of evaluable pts with optimal molecular responses (MR3 and DMR) increased over time from 60% (21/35) at start of PON to 74% (26/35) at 6 mo, 80% (20/25) at 12 mo, 90% (18/20) at 24 mo, and 94% (17/18) at 36 mo. Progression-free survival (PFS) and overall survival (OS) rates at 36 mo were both 94%, with median (range) follow-up of 3.5 (0.1-7.6) y.

Thirty-one pts had ≥1 PON-related AE, most commonly hypertension (n=4), increased lipase, increased alanine aminotransferase, and increased aspartate aminotransferase (n=2 each). Six pts had CV events, including 1 arterial occlusion event. Dose modifications occurred in 19/36 pts due to medical decision (n=9), achieving at least a major CyR (n=5), and AEs (n=5). Median (range) time to dose modification was 1 (1-5) mo. Of 7 pts starting tx with 45 mg, 5 reduced to 30 mg and 1 reduced to 15 mg. Of 14 pts starting with 30 mg, 7 reduced to 15 mg, 1 increased to 45 mg for ineffectiveness, and 1 changed to Other dose. Of 15 pts starting with 15 mg, 3 increased to 30 mg and 1 changed to Other dose. Fifteen pts had permanent PON interruption due to AEs (n=7), progression/death (n=3), other reason (n=2), other tx, good molecular response, or medical decision (n=1 each). Of notable clinical relevance, only 1 pt who switched from bosutinib to PON had cross-intolerance (hepatic toxicity), which led to PON reduction but not permanent PON discontinuation.

Conclusion: Over a median follow-up of 3.5 y, PON demonstrated a manageable long-term safety profile in RW clinical practice among pts intolerant to previous TKIs. Median PON duration was 3.1 y, which allowed long-term maintenance or achievement of MR3 and DMR in 94% of evaluable pts, with high PFS and OS rates at 3 y. Results suggest dose optimization during tx represents a key strategy for disease management in this subgroup.