Treatment and Disease Characteristics of Chronic Myeloid Leukemia in Blast Crisis: The European Leukemianet Blast Crisis Registry

Introduction

Despite improvements in the treatment of chronic myeloid leukemia (CML), an important minority of patients (pts) will still develop blast crisis (BC). BC is highly aggressive, and if not treated promptly, will inevitably lead to patient’s death. Due to its rarity and heterogeneous clinical presentation, no standard treatment is available for BC, and specifically addressed clinical trials are difficult to perform. Furthermore, it is unclear if the biology of BC has changed since the advent of tyrosine kinase inhibitors (TKI). To better understand BC and its treatment we sought to collect clinical and biological data of pts diagnosed with BC and started an international registry within the European LeukemiaNet (ELN) in 2018.

Materials and methods

The European Blast Crisis Registry with participants in 9 countries (Germany, Poland, Czech Republic, France, Italy, Sweden, Austria, Russia, and Armenia) gathers routinely collected clinical data in an online database. Pts are eligible for the registry if they had a diagnosis of BCR::ABL1 positive BC after January 1^st, 2015 and were at least 18 years old at the time of BC. Both de novo BC and BC developing from a precedent CML in chronic phase are included. The Mann-Whitney and exact Fisher tests were used to compare variables. Overall survival (OS) was calculated according to Kaplan Meier. Significance was set at p<.05.

Results

At the cutoff date of July 2^nd, 2022, 200 patients (60% male) with confirmed BC were recruited. Median age at diagnosis of CML and at diagnosis of BC was 45 years (yrs, range, 13-86) and 48 yrs (range, 18-86), respectively. BC occurred de novo in 37% of pts. At the time of CML diagnosis 47 pts had at least one additional cytogenetic abnormality (ACA) beside the t(9;22)(q34;q11), with 5 pts having 3 different ACAs. The most frequent ACA was +8 (10 pts) followed by +der(22) (6 pts). For pts diagnosed with a previous history of CML, median time between CML and BC was 31.3 months (range, 1-378 months). Before BC, 76% of patients were treated with TKI, most frequently with imatinib (78%). Median number of TKI treatments for chronic phase CML were 2 (range, 1-5). At the time of BC, mutations of BCR::ABL1 occurred in 25% of pts. The most frequent mutations were E255K and T315I, seen in 9 pts each. Mutations occurred more frequently in pts with evolution to BC after chronic phase (25% vs 8%, p<.001). The morphology of BC was myeloid (mBC) in 58% of the pts and lymphoid (lyBC) in 35%; 7% had a mixed or megakaryoblastic phenotype. Thirty-five and 16 pts had extramedullary manifestations (EMM) and central nervous system (CNS) involvement, respectively. Treatment data of BC are available for 191 pts and show the expected heterogeneity. Median lines of therapy for BC were 3. Most common was TKI+chemotherapy (Ctx)+allogeneic stem cell transplantation (alloTx) (49%), 19% received TKI+Ctx, 10% TKI alone, 5% TKI+alloTX, 1% Ctx+alloTx, 6% Ctx alone and 1 pt each alloTx upfront or no treatment. Pts receiving dasatinib (p=.004), Ctx (p=.005) and alloTx (p<.001) were younger. Imatinib was more used in de novo BC (p<.001), whilst ponatinib in pts with a previous CML (p=.016). Shorter time between CML and BC correlated significantly with a higher use of Ctx (p=.005). The type of therapy was not significantly associated with EMM/CNS involvement or BC phenotype, although slightly more pts with lyBC received Ctx (93% vs 82% p=.054).

Best responses within the first 6 months of treatment are available in 82 pts (4 received an alloTx) and are: 18% MR5, 6% in MR4.5, 2% MR4 and 10% MMR; 24% had no response.

With a median follow up of 27.0 months from diagnosis of BC, median OS was 23.7 months. Pts with EMM and CNS involvement had a similar outcome compared with pts with no EMM/CNS (median OS 46.0 vs 60.0 months for EMM/CNS vs none respectively, p=.56). The morphology of BC had a significant influence on survival. LyBC had a 2-yrs survival of 62% (95%-confidence interval (CI): 46-72%), mBC had a 2-yrs survival of 39% (95%-CI: 28-50%), p=.01). Sex had no influence on survival, morphology of BC or treatment received.

Conclusions

This analysis on a large cohort of CML pts in BC demonstrates a high degree of heterogeneity with regard to presentation and treatments. The outcome of lyBC is more favorable as compared to mBC, whilst the presence of CNS involvement and EMM had no significant impact on survival. Focused trials are warranted to identify the best treatment option for this still difficult disease.