Comparison of Sokal, Hasford and EUTOS Scores in Terms of Long-Term Treatment Outcome According to the Risks in Each Prognostic Model: A Single Center Data Analyzed in 255 Early Chronic Phase Chronic Myeloid Leukemia Patients Treated with Frontline Imatinib Mesylate.

Abstract Abstract 2794 Background The main difference of recently developed new prognostic classification, EUTOS score, from the conventionally used Sokal and Hasford scores is the simplicity of adopting only 2 clinical parameters, basophil counts and size of spleen, to discriminate high vs. low risks in chronic myeloid leukemia (CML). Concerns may rise from possible inter-observer gap when measuring the size of spleen. While there is currently controversial debate on its prognostic power, we present our single center results of comparing EUTOS score with other two scores. Patients and Methods Among 380 adult patients, consecutively registered as early chronic phase (CP) CML at Seoul St. Mary's Hospital (Jan. 2004 – Apr. 2010), who received imatinib (IM) as frontline therapy, 255 were evaluable for analysis: subjects were strictly limited to those with documented clinical parameters measured at very early time of diagnosis by a single doctor, before any cytoreduction therapies. The median age was 42 years (range, 19–77) with a slight male predominance (57%). Cytogenetic and molecular responses were calculated at specific time points and cumulative rates of response were compared according to risk scores. Event-free survival (EFS) was analyzed by time from start of IM to the time of events, defined by death from any cause, primary or secondary resistance to IM, progression to advanced phases, intolerance to IM and switch to second line tyrosine kinase inhibitor (TKI) or allografting. Response assessment followed current European Leukemia Net recommendation. Results Median time from diagnosis to IM start was 14 days (range, 1–183). After a median follow-up duration of 56.6 months (range, 13.0–102.2) in survivors, there were 10 deaths and 3 patients with disease progression to accelerated phase of blastic crisis. Sixty-four patients had switched IM to second TKI due to either IM resistance or intolerance. Overall comparisons are described in Table 1. Of 42 EUTOS high risk patients, only 25 high risk patients by Sokal and 15 by Hasford were included. Almost 90% of either low or intermediate (INT) risk patients by all risk models achieved complete cytogenetic response (CCyR) at 12 months after IM, whereas high risk groups had lower rates, especially in EUTOS high risk (57%, p = <0.0001). Regarding major molecular response (MMR) at 18 months, significant differences were observed according to risk stratifications by Sokal and EUTOS, but not in Hasford risks, in which the same proportion of low and INT risks achieved MMR (48%, P = 0.308). This seemingly less prominent prognostic difference of low vs. INT groups of Sokal and Hasford scores were further demonstrated by insignificant difference between these groups in cumulative incidence (CI) of CCyR and MMR. All 3 high risk groups showed significantly lower incidence of both CCyR and MMR compared to low risk groups. INT risk groups by Sokal and Hasford models failed to show statistical significance in all three outcomes: CI of CCyR and MMR, and 5-year EFS. Of note, high risk stratified by EUTOS score demonstrated the least favorable outcomes with strongest p-values (Table 1). Conclusion By simply comparing the p-values, our data may implicate that high risk group discriminated by the EUTOS score is predicted with more unfavorable outcomes in achieving CCyR, MMR and survival after IM frontline therapy. Furthermore, the lack of statistically significant difference between outcomes of low and INT risk groups by conventional risk models may challenge the need for future classification of INT from low risk in this era of TKI. Disclosures: No relevant conflicts of interest to declare.