The criteria to define chronic (CP), accelerated (AP) and blast phase (BP) CML vary in the literature. The WHO recently proposed, based on the literature and collective experience of the clinical advisory committee, new criteria for myeloid malignancies including CML in an attempt to provide more uniform criteria (

Blood 2002; 100:2292
). Imatinib is the current standard therapy for CML, and the criteria used in most studies using imatinib are shown below in contrast to the WHO proposal:

PhaseCriteriaImatinibWHO
*Unrelated to therapy (Rx), **Unresponsive to Rx, Plts=platelets, BM=bone marrow 
AP Blasts 15–29% 10–19% 
 Blasts + promyelocytes ≥ 30% NA 
 Basophils ≥ 20% ≥ 20% 
 Plts <100 x 109/L* Yes Yes 
 Plts >1000 x 109/L** No Yes 
 Increasing spleen size and WBC** No Yes 
 CE At any time Not at diagnosis (Dx) 
BP Blasts ≥ 30% ≥ 20% 
 Extramedullary disease Yes Yes 
 Large clusters of blasts in BM NA Yes 
PhaseCriteriaImatinibWHO
*Unrelated to therapy (Rx), **Unresponsive to Rx, Plts=platelets, BM=bone marrow 
AP Blasts 15–29% 10–19% 
 Blasts + promyelocytes ≥ 30% NA 
 Basophils ≥ 20% ≥ 20% 
 Plts <100 x 109/L* Yes Yes 
 Plts >1000 x 109/L** No Yes 
 Increasing spleen size and WBC** No Yes 
 CE At any time Not at diagnosis (Dx) 
BP Blasts ≥ 30% ≥ 20% 
 Extramedullary disease Yes Yes 
 Large clusters of blasts in BM NA Yes 

We investigated the clinical validity of the WHO proposal among 809 patients (pts) with CML in all stages treated with imatinib at MDACC since 1999. According to the imatinib classification, 537 (66%) pts were in CP, 196 (24%) in AP, and 76 (9%) in BP. When analyzing the specific subsets of pts where the classifications differ, major findings are: 1) Pts with CE at Dx (n=14) have similar cytogenetic (CG) complete remission (CR) rate (86%) than pts (n=477) with imatinib CP (75%, p=.53), but it is lower when CE develops after Dx (34/64, 53%; p=.0005). Similar results are found for overall survival (OS) and progression-free survival (PFS). 2) CG CR rate among pts with 20%–29% blasts (4/19, 21%) is more similar to that of AP (37/99, 37%)(p=0.19) than to pts with ≥30% blasts (5/76, p=0.07). This trend is more significant for OS and PFS. 3) Pts with increasing WBC and spleen, or plts >1000 x109/L unresponsive to Rx have a CG CR rate (9/42, 21%; p=.07) lower than others with AP, but there is no difference in OS or PFS. 4) Pts with plts >1000 x109/L without prior Rx have similar outcome as those with CE developing on therapy. In conclusion, based on results imatinib therapy, AP can be defined as blasts ≥10%, basophils ≥20%, plt <100 x109/L (unrelated to Rx) or >1000 x109/L (unresponsive to Rx), increasing WBC and spleen (unresponsive to Rx), and CE not at diagnosis; BP is defined by blasts ≥30% or extramedullary disease; pts who develop CE during the course of Rx of with plts >1000 x109/L not related to Rx constitute an intermediate group between CP and AP; all others are CP.

Author notes

Corresponding author

Sign in via your Institution