New CML Risk Score Incorporates Modern Survival Probability

October 3, 2015

A new risk assessment score more accurately reflects the probability of dying from chronic myeloid leukemia than do existing scoring systems.

A new risk assessment score more accurately reflects the probability of dying from chronic myeloid leukemia (CML) than do existing scoring systems, according to a new study.

The use of tyrosine kinase inhibitors beginning with imatinib has yielded dramatically improved prognosis for CML patients, with an 8-year overall survival (OS) probability of over 80%. “A consequence of the improved survival induced by imatinib is the increased probability of dying of causes other than CML,” wrote study authors led by Markus Pfirrmann, PhD, of Ludwig-Maximilians Universität in Munich, Germany. “Investigators wonder to what extent survival probabilities still depend on CML.”

The new study included 2,290 patients from several European study cohorts. All patients had BCR-ABL1–positive CML in chronic phase, and received imatinib within 6 months of diagnosis. The results were published online ahead of print in Leukemia.

The median follow-up period was 6.3 years, and the full cohort had an 8-year OS probability of 89%. A total of 208 patients (9%) died during follow-up; the cause of death was CML (defined as death after disease progression) in 92 cases (44% of deaths), non-CML-related in 104 patients (50%), and unknown in 12 cases (6%). The 8-year cumulative incidence probability of CML-related death was 4%, and of non-CML-related death was 7%.

Of those who died of non-CML causes, 50 died of a second cancer, 21 due to a cardiovascular event, 12 due to an infection, and 21 from other causes.

The researchers examined how well existing scoring systems predicted death and its causes; the Sokal, Euro, and EUTOS scores all identified a high-risk group with a significantly higher incidence of CML-related death than other patients. But neither the Sokal nor the Euro score showed any difference between low- and intermediate-risk groups; the EUTOS score did show a difference.

They then created a new scoring system using variables found to be significantly associated with a higher risk of dying of CML: higher age, bigger spleen size below costal margin, higher percentage of peripheral blasts, and low platelet counts. The EUTOS Long-Term Survival (ELTS) score was then validated in 1,120 other patients.

Of that cohort, 61% were considered low risk. The intermediate-risk patients had an increased risk of dying of CML compared to the low-risk group (subdistribution hazard ratio [HR], 2.040 [95% confidence interval (CI), 1.039–4.005]; P = .035). The high-risk group had a subdistribution HR for death from CML of 6.746 (95% CI, 3.614–12.592; P < .001).

This is the first long-term scoring system that considered specifically CML-related death, the authors wrote. “We hope that the ELTS score will be considered for the risk-stratified planning, analysis, and outcome interpretation of clinical trials,” they wrote. “It provides a sound basis for the development of risk-adapted treatment.”