Sudden blast crisis in chronic myeloid leukemia treated with tyrosine kinase inhibitors

Since the introduction of imatinib (IM) therapy, the overall survival (OS) of patients with chronic myeloid leukemia–chronic phase (CML-CP) has improved significantly, with the 6-year OS being 88% and transformation (accelerated phase and blast crisis; AP/BC)-free survival of 93% [1]. As the median survival of patients with BC even in the era of tyrosine kinase inhibitor (TKI) therapy remains poor, the best strategy remains the prevention of BC. In the International Randomized Study of Interferon vs. STI571 (IRIS) trial, progressions per year averaged 1.9% in the first 3 years and 0.7% in the next 3 years, suggesting that the highest risk of transformation to advanced phase is within the first few years of therapy [1]. Transformation to BC occurs predominantly in the setting of inadequate disease control, which could stem from poor drug adherence and/or IM resistance. Hence, optimal disease control could reduce the incidence of BC. However, some patients with CML-CP progress to “sudden blast crisis” (SBC) after having achieved an optimal response [2–5]. The definition of SBC has gradually evolved with changes in therapy. In the interferon era, it was defined as onset of BC within 3 months of previously documented complete hematological response [2], while in the IM era, SBC was defined as the onset of blast phase after documented complete cytogenetic response (CCyR) in the immediately preceding bone marrow (BM) and within 3 months of normal complete blood counts (CBCs) [3,5]. In that study, 4% of patients (23/541) developed BC after a median follow-up of 48 months, and 4/541 cases (17% of patients with BC and 0.7% of all patients) were classified as having SBC [3]. In this issue of Leukemia and Lymphoma, Tantiworawit and colleagues report that in their CML series, 13.1% of IM-treated patients with CML-CP (n = 213) developed BC [4]. Nine of these patients (4% of all patients and 32% of patients with BC) were classified as having SBC. The incidence of BC and of SBC was higher in this Canadian study [4] compared to the M. D. Anderson study. This could partially be attributed to differences in definitions. The M. D. Anderson group defined BC as BM/peripheral blood (PB) blasts ≥ 30% [2,3], while Tantiworawit and colleagues defined BC as BM/PB blasts ≥ 20%. Another difference was the definition of sudden BC. The M. D. Anderson study mandated documented CCyR before the diagnosis of sudden BC [3], while Tantiworawit and colleagues defined it as the occurrence of BC in patients who had an optimal response, as per European LeukemiaNet (ELN) recommendations [6]. The M. D. Anderson’s definition is more stringent than that proposed by Tantiworawit et al. As only 60% of patients achieve a CCyR during first the 6 months of IM therapies, 30–40% patients who did not achieve CCyR but achieved optimal response (ELN guidelines) would have been excluded from the M. D. Anderson analysis.

Despite the retrospective nature of these studies and their differences, they raise some interesting issues. Classifying patients with blast crisis into SBC and gradual blast crisis may not be biologically appropriate: SBC indicates rapid kinetics of blast crisis, while “gradual blast crisis” suggests slow kinetics. However, there are no data to suggest that the kinetics of blast crisis differ in these cases. Before widely using these definitions, understanding the biology of these blast crises and their kinetics is critical.

Retrospective studies from the interferon and IM eras suggest that patients presenting with SBC have a predominantly lymphoid blast morphology (56–67%) [2], are more likely to have a low Sokal score at diagnosis [2–5] and tend to present within a short time of starting IM therapy [2–5]. Lymphoid blast crisis generally presents without a preceding accelerated phase, and thus it presents without warning, while myeloid blast crisis presents with or without a preceding accelerated phase. Hence, it would be interesting to know whether lymphoid BC is predominantly seen in the first few years of IM therapy and whether myeloid blast crisis is predominant in later years. Published data from major randomized studies of imatinib [1], dasatinib [7] and nilotinib [8] grouped patients with AP and BC together, and hence it is rather difficult to determine the frequency and type of blast crisis during each year of TKI therapy.

The Canadian studies suggest that patients presenting with SBC have better prognosis than patients presenting with gradual blast crisis. However, this could be due to differences in therapy between these two groups. In the Canadian study, most of the patients with gradual blast crisis presented before the availability of second-generation TKI therapy and were not suitable for allogeneic stem cell transplant. On the other hand, the majority of the patients with SBC received dasatinib as second-line therapy and underwent allogeneic stem cell transplant. Moreover, the survival in both groups of patients is not satisfactory.

These studies also suggest that the majority of blast crisis cases occur in the setting of inadequate disease control, raising the possibility that with the availability of effective therapies this can be reduced significantly. However, despite more effective TKI therapies, a small number of patients still progress into blast crisis. In the IRIS study, patients who achieved a BCR–ABL ratio (International scale; IS) < 10% within 6 months, < 1% within 12 months or < 0.1% by 18 months of IM therapy have a significantly lower risk of transformation to AP/BC compared to patients who did not achieve similar responses (1–4% vs. 20–25%) [9]. Similarly, two phase III studies demonstrated that, in patients with newly diagnosed CML-CP, both dasatinib [7] and nilotinib [8] induce a deeper and faster molecular response than IM, and reduce progression to AP/BC. However, in all these studies, 2–3% patients still progress to AP/BC within 2–3 years of starting TKI therapy, suggesting that with the current strategy we cannot completely prevent transformation to BC.

Currently, none of the clinical markers or early response criteria can predict SBC. This is the situation that clinicians fear most for their patients–the development of BC without warning and the very limited options for ongoing disease control in these patients. Research efforts should be focused on understanding the mechanism of transformation and developing molecular markers which can predict these patients at diagnosis, leading to the early use of potentially more effective strategies such as combination therapy and/or allogeneic stem cell transplant.

Potential conflict of interest

Disclosure forms provided by the authors are available with the full text of this article at www.informahealthcare.com/lal.