Population Based Study of Cytogenetic Abnormalities in Addition to Philadelphia (Ph) Chromosome in Patients with Chronic Myeloid Leukaemia (CML) and Impact on Survival

Abstract

Ph chromosome is the hallmark of CML. However there are few reports of additional chromosomal abnormalities at the time of diagnosis and the impact this has on overall survival (OS). [NT1] The Cytogenetic laboratory at this hospital provides a regional service as a single facility. The data from this laboratory was combined with survival information to evaluate the impact of additional chromosomal changes on outcomes in patients with CML.Methods: This is a retrospective population based study, it was not possible to obtain consent from individual patients and details about haematological parameters or treatments delivered were not available. The aim was to evaluate if cytogenetic changes should be considered in addition to established risk scoring systems. Patients were classified as complex-Philadelphia (Ph) if they had t(9;22)(q31;q34) with additional chromosomal abnormalities. Impact of individual additional abnormalities was analysed and then the effect was stratified according to presence of chromosomal gains, deletions or translocations. Few cases who had normal cytogenetics on their first sample as they were initially treated elsewhere.Results: 1129 patients were diagnosed with CML between 1985 and 2013, with 4760 samples analysed. The median age of the patient was 52.4 years (4.3-103, 602 male; 511 female; 16 unknown). Median follow up was 6.4 years [0-26.8 years, 725/1129 (64.2%) had follow-up more than 10 years]. End point for analysis was probability of survival at 10yr. 194/1129 (17.2%) had complex-Ph at diagnosis, 759/1129 (67.2%) had standard Ph, 77/1129 (6.8%) had negative cytogenetics and in 34/1129 (3%) cytogenetic analysis failed at diagnosis. Patients with standard Ph translocation had significantly better chance of achieving cytogenetic CR than those with complex-Ph (23.4% vs. 13.4%, p<0.001). OS was significantly better in patients below the age of 45 (65% vs 25% p <0.0001). OS was also better in patients diagnosed after 2000 (67 % vs 40 %, p<0.0001). In univariate analysis OS was significantly lower with trisomy 8 (10% vs 50%, p<0.0001), del(5q) [NT2] (20% vs 50%, p=0.001), other deletions (12% vs 48%, p=0.0005), del 17( 48% vs. 0%, p<0.0001), add 21 (50% vs. 0%, p<0.0001), any translocations (50% vs. 22%, p<0.0001), der 22 or iso 17 (48% vs. 10%, p<0.0001), any deletions (50% vs. 20%, p<0.0001) and variant Ph translocations (50% vs 22%, p=0.004). In multivariate analysis, excluding year of diagnosis, age group (HR 1.93, 95% CI:1.6-2.4 P=<0.0001), complex t(9;22;v) (HR 1.8, 95% CI:1.0-3.1, P=0.035), del(17q) (HR 3.8, 95% CI:1.1-12.6, P=0.033), translocations (HR:1.6, 95% CI: 1.1-2.25, p=0.013), trisomy 8 (HR: 1.76, 95% CI: 1.19-2.65, p=0.005), add 21 (HR: 3.21, 95% CI: 1.65-6.25, p=0.001) and der 22 or iso17 (HR: 1.57, 95% CI: 1.1-2.25, p=0.013) were independently associated with inferior OS. Number of risk factors in individual patients was used to design a scoring system. Patients with 0 risk factor (70% OS at 10 yr.), 1 risk factor (40% OS at 10 years), 2 risk factors (22% OS at 10 years) or more than 3 risk factors (12% OS at 10 years) had incrementally reduced OS. This was true of patients diagnosed before and after 2000.This analysis suggests that incorporating nature of karyotype at diagnosis can refine established scoring systems. However this data needs to be analysed with larger patient population to include all established risk factors and the effect of therapeutic measures. DisclosuresCavet:Novartis: Research Funding; BMS: Research Funding.