Effectiveness of Tyrosine Kinase Inhibitors in Patients with Chronic Myeloid Leukemia in Chronic Phase (CML-CP) after Two or More Prior Tyrosine Kinase Inhibitors

Introduction: Despite the availability of therapies that have broadened treatment choices and improved outcomes for patients with CML-CP, patients who have been treated with ≥2 prior tyrosine kinase inhibitors (TKIs) experience high failure rates due to resistance or intolerance, with outcomes worsening across lines of therapy.

Previously presented real-world data on TKI treatment patterns in patients with CML-CP from 3 registries across different countries did not identify a common TKI sequence. Between 12.8% and 35.7% of patients starting treatment with a second line TKI required third line therapy and beyond (3L+) and 32.8% (124/378) of patients in 3L+ discontinued treatment, with treatment intolerance among the most frequent reasons for discontinuation.

There is scarce information on response rates in patients treated with ≥2 prior TKIs and no established standard of care exists for these patients. Here we present data collected from CML registries on the effectiveness of different TKIs in patients with CML-CP initiating therapy after ≥2 prior TKI treatments.

Methods: This non-interventional, descriptive, retrospective cohort study is based on secondary use of data from 3 existing CML registries (Czech INFINITY, Dutch PHAROS, and Swedish CML register). The observation period was from January 1, 2008, until the last available date of follow-up for each registry (6‒12 years depending on the registry). Patients with CML-CP aged ≥18 years and initiating a TKI after being treated with ≥2 prior TKIs were included in the analysis. Study index date was defined as the start date of treatment for each TKI used in the 3L+ setting. Effectiveness endpoints (major molecular response [MMR, BCR::ABL1^IS ≤0.1%] rates by 6 and 12 months and overall survival [OS] rates) were analyzed separately for each TKI. Results for each registry are reported separately.

Results: Among patients enrolled in the INFINITY, PHAROS, and Swedish CML registries, 105, 48, and 172 patients, respectively, started 3L+ TKI therapy and were included in this analysis.

Across all 3 registries, the most frequently prescribed TKIs in the 3L+ setting were nilotinib (NIL; 52.4%, 58.3%, and 45.3%, respectively) and dasatinib (DAS; 38.1%, 16.7%, and 32.6%, respectively). NIL and DAS were mostly prescribed to patients who had received 2 prior TKIs at index, whereas bosutinib (BOS) and ponatinib (PON) were mostly prescribed in later treatment lines (BOS: 8/12 [66.7%], no data [ND], and 21/48 [43.7%] patients had received 3 prior TKIs at index, respectively; PON: 9/14 [64.3%], 3/8 [37.5%], and 7/19 [36.8%] patients had received 3 prior TKIs at index, respectively). At the time of TKI index, there were more patients already in MMR in the Swedish CML register compared with the other registries (Table 1).

For patients not in MMR at the time of TKI index, the highest MMR rates by 12 months across all registries were observed on DAS (12/32 [37.5%], 3/8 [37.5%] and 14/34 [41.2%], respectively), followed by NIL (14/47 [29.8%], 6/21 [28.6%], and 13/46 [28.3%], respectively). MMR rates by 12 months were in general lower for BOS (2/11 [18.2%], ND, 7/21 [33.3%]) and PON (2/11 [18.2%], 2/8 [25.0%], 5/13 [38.5%]) (Table 2). The OS analysis showed that the median survival time was longest for NIL (42.5, ND, and 45.0 months, respectively), followed by DAS (34.5, 68.0, and 40.9 months, respectively).

Conclusion: This is the most comprehensive study to date analyzing TKI response rates based on real-world data from patients with CML-CP treated in the 3L+ setting. The number of patients in each registry was low, which may explain the variability of response and limits comparability across TKIs. However, a common pattern of treatment was observed across the 3 countries, with BOS and PON being used more frequently in patients treated with ≥3 prior TKIs; this could explain some of the differences in response rates. In the Swedish CML register, the high number of patients in MMR at the time of TKI switch may suggest that intolerance is a frequent reason for switching TKI; this is consistent with the reasons for discontinuation reported in a previous analysis. This study contributes to the evidence on real-world response rates of TKI treatment in CML-CP.

This study was a research collaboration between Novartis and the 3 registries and was funded by Novartis.