Identifying Risk Factors for Progression to Blastic Phase in Chronic Myeloid Leukemia Patients Under Imatinib Therapy: A Cohort Analysis

Background

After the introduction of tyrosine-kinase inhibitors (TKIs), prognosis of chronic myeloid leukemia (CML) has improved dramatically, although a small subset of patients progresses to blastic phase (BP), which is associated with an adverse outcome. Even though second and third generation TKIs have shown higher-response rates and less BP-progression rates compared to imatinib, these therapies are not as readily available in lower- and middle-income countries, and imatinib is still used as first-line TKI in many settings worldwide.

Aim

To identify factors associated with progression to BP in patients with CML treated with imatinib as first-line TKI.

Methods

This is a case-cohort study from a referral center in Mexico City that included all patients with the diagnosis of CML treated with imatinib as first-line TKI between 2001 and 2023. Patients with BP at diagnosis were excluded. We retrospectively collected clinical and laboratory variables at diagnosis, including additional chromosomal abnormalities (ACAs) at diagnosis and during follow-up, and molecular and cytogenetic responses.

Results

The median age at diagnosis was 41 years (range, 13-84), with 46.6% being female. Laboratory findings at diagnosis were as follows: median leukocyte count of 206.2 x 10³/µL (range, 16.4-820), median peripheral blood (PB) basophil count 5% (range, 0-28), median PB blasts 2% (range, 0-17), median bone marrow blasts 0% (range, 0-12). Only 9.5% presented with ACAs at diagnosis. The median follow-up of the cohort was 112 months (range, 4-318).

A total of 87.2% of patients were in the chronic phase (CP), while 12.8% were diagnosed in the accelerated phase (AP). Patients were considered high risk, based on different scales in the following percentages: 37.1% with EUTOS, 10.6% with Hasford, 20.2% with Sokal, and 22.7% with ELTS.

During follow-up, ACAs developed in 24.3% of patients with a median time since diagnosis of 19.5 months (range, 3.5-240.5). Treatment changes were required in 22.6%, primarily due to treatment failure or resistance (87.5%) and intolerance (12.5%), with no significant difference between those who progressed to the BP.

Out of a total of 148 patients, 10.8% of the cohort progressed to blast phase (BP) with a median time of 47.4 months since diagnosis (range, 4.9-145.3 months). Median progression to BP free survival (PFS) was not reached, with a 10-year PFS of 87.2%. Factors associated to PFS in an univariate analysis were: PB basophils ≥4% (HR 3.87 [95% CI, 1.09-13.8], p=0.037), PB blasts ≥4% (HR 5.29 [95% CI, 1.89-14.85], p=0.002), AP phase at diagnosis (HR 3.48 [95% CI, 1.10-11.04], p=0.034), and a complex karyotype (HR 17.26 [95% CI, 1.79-166.69], p=0.014). In a multivariate analysis, only PB blasts ≥4% (HR 3.86 [95% CI 1.18-12.58], p=0.025) and a complex karyotype at diagnosis (HR 10.86 [95% CI 1.05-112.1], p=0.045) remained statistically significant for progression to BP. The development of ACAs during follow-up did not impact PFS.

The scales with the best performance to predict PFS were a high EUTOS, with a 10-year PFS of 76.2% vs. 92.4% (p=0.019), a high Hasford, with a 10-year PFS of 53.8% vs. 89.7% (p=0.019) and a high ELTS, with a 10-year PFS of 78.5% vs. 93.2% (p=0.036). Excluding patients who progressed within a year, achieving major molecular response (MMR) within one year was a protective factor against BP progression at 10 years (98.8% vs. 62.6%, HR 0.028, 0.004-0.224, p=0.001).

The overall 10-year overall survival (OS) for the entire cohort was 84.1%, with a median OS of 6.3 months (0.0-12.7) after BP development. The prognostic factors for OS in the multivariate analysis were progression to BP (HR 43.04, 7.77-238.48, p<0.001), age ≥55 years, and PB basophils >4%.

Conclusions

In a large cohort of CML patients treated with first-line imatinib, we observed a 10% progression to BP. Our findings identified several risk factors for progression to BP, including basophil count (≥4%), peripheral blood blasts (≥4%), a complex karyotype, failure to achieve MMR at one year, and AP at diagnosis. This study also underscores the utility of established risk scores like EUTOS, Hasford and ELTS in predicting long-term outcomes. It is crucial to identify high-risk patients who may benefit from more intensive monitoring or alternative therapies, particularly in resource-limited settings where second or third generation TKIs are not readily available, and most patients are treated with imatinib.

Demichelis:Astellas: Consultancy, Honoraria; AMGEN: Honoraria; Pfizer: Consultancy, Honoraria; TEVA: Consultancy, Honoraria; Servier: Consultancy, Honoraria; Abbvie: Consultancy, Honoraria. Tuna Aguilar:Abbvie: Speakers Bureau; Novartis: Speakers Bureau; Bristol Myers Squibb: Speakers Bureau; Pfizer: Speakers Bureau.