ORLANDO ­– For prognosis of long-term survival outcomes in imatinib-treated adult patients with chronic myeloid leukemia (CML), the European Treatment and Outcomes Study (EUTOS) long-term survival (ELTS) score is recommended, according to a study presented at the American Society of Hematology annual meeting.1

The EUTOS long-term survival score was developed by using data from 2205 in-study patients with CML treated with imatinib in order to classify patients into 1 of 3 risk groups with clinically significant different probabilities of dying from CML. This score was validated in 1120 out-study patients.

However, many investigators still apply the Sokal score when determining the prognosis of patients with CML treated with tyrosine kinase inhibitors. The Sokal score allocated more than 20% of patients with chronic phase CML to the high-risk group while the new ELTS score designated 12% to the high-risk group. Therefore, researchers sought to compare risk group allocations and prognosis of patients with CML between the 2 scores.

“The EUTOS long-term survival score has 4 variables,” Markus Pfirrman of the EUTOS group explained during his presentation. “For younger patients, getting 1 year older has a low impact, but for older patients, getting 1 year older has a larger impact.”

From the EUTOS registry, researchers identified 3325 patients with a median observation time of 6.1 years. Results showed that the 6-year overall survival probability was 91% (95% CI, 89 – 92) with CML-related death occurring in 142 of 309 deceased patients. The 6-year cumulative incidence of probabilities (CIP) of dying of CML was 4% (95% CI, 4 – 5) vs 5% (95% CI, 4 – 6) for death unrelated to CML.

Researchers found that the of 758 patients allocated to high risk by the Sokal score, the ELTS score designated 165 as low risk and 265 as intermediate risk. The 6-year CIP of dying was 5% (95% CI, 2 – 9) for the 165 low-risk patients and 5% (95% CI, 3 – 9) for the 265 intermediate-risk patients compared with the 6-year CIP of dying of 13% (95% CI, 9 – 17) for the 328 high-risk patients according to both scores (P = .0062 and P = .0050, respectively). In addition, overall survival probabilities were significantly higher for the 430 Sokal high but ELTS non-high-risk patients than the 328 Sokal and ELTS high-risk patients (P = .030).

“Regarding the probabilities of dying of CML and overall survival compared to the ELTS score, the Sokal score was too pessimistic allocating too many to the high-risk group,” Dr Pfirrman said.

However, results also showed that the overall survival probabilities of the 768 non-low-risk Sokal patients were not significantly different from the 1262 low-risk patients allocated by both scores.

RELATED: Spliceosome Modulators Could Be New Thereapeutic Targets in Myeloid Malignancies

“On the other hand, patients with low-risk Sokal but non-low-risk ELTS score had significantly worse probabilities of dying of CML than patients with low risk according to both scores,” Dr Pfirrman concluded.

The findings suggest that the allocation by the Sokal score was not appropriate for patients with CML treated with tyrosine kinase inhibitors.


  1. Pfirrmann M, Hasford J, Saussele S, et al. The EUTOS survival score is preferable over the Sokal score for prognosis of long-term survival of patients with chronic myeloid leukemia. Oral presentation at: 57th American Society of Hematology (ASH) Annual Meeting & Exposition; December 5-8, 2015; Orlando, FL.