Abstract
Abstract 3759
Data of 2060 patients from the in-study registry of the European Outcome and Treatment Study (EUTOS) for CML were used to develop and validate the EUTOS score [1]. All these patients were included in prospective controlled clinical trials. The EUTOS score aims to support clinical decision making within the first 18 months after initiation of treatment with imatinib. Patients who did not achieve complete cytogenetic remission (CCyR) within 18 months had a lower probability of achieving CCyR in the further course of therapy and were more likely to suffer from progressive disease. The EUTOS score is calculated by multiplying the percentage of basophiles by seven and the spleen size measured in centimeters below costal margin by four and adding both values. Both parameters have to be assessed before any therapy is started. If the resulting value is higher than 87 the patient is at high risk of not being in CCyR after 18 months of therapy, otherwise he is a low risk patient. In the In-study data the EUTOS score showed a sensitivity of 21%, a specificity of 92% and a positive predictive value of 34%. So every third patient in the high risk group eventually did not achieve CCyR.
As the development and validation of the EUTOS score used patient data from prospective clinical studies we now wanted to assess the score's performance on CML-patients in routine health care. The EUTOS Out-Study registry provides data of 1547 patients from Spain, Poland, Czech Republic, Romania, Slovakia, and Russia. Information on the EUTOS score and the status of CCyR at 18 months (+/− 3 months) was available for 316 patients.
The patients in the In-study registry were slightly older than in the Out-study registry (median (range): 52 years (18–83) vs 48 years (18–85)) and more men were involved (61% vs 52%). The Out-study data support the timeline of 18 months as patients without CCyR at 18 months progress more often than patients with CCyR (progression free survival after 36 months 99.2% vs 90.8%, p<0.0001).
As 316 of 1547 datasets might lead to a selection bias we compared the characteristics of both groups but medians of age, spleen size, platelets, percentage of blast cells, percentage of eosinophils, white blood cell count, hemoglobin and percentage of basophils were almost equal. So no selection bias is evident.
In the Out-study data the EUTOS score reached a sensitivity of 16%, a specificity of 90%, and a positive predictive value of 41%. These results were similar to the results of the In-study data and confirmed that the score defines a small high risk group with a high probability of not reaching CCyR. The cumulative incidence curve showed that high risk patients achieve CCyR significantly later and less often than low risk patients (Median 34.0 months vs 20.4 months, 32.6% vs 43.4% after 18 months of therapy, p<0.0001). In addition high risk patients have a significantly higher risk of progression (progression free survival after 5 years: 88.8% vs 80.7%, p=0.0235, median observation time 66 months) and death (overall survival after five years 89.9% vs 82.0%, p=0.0103, median observation time 66 months).
The results show that the EUTOS score is also valid in Out-study patients and is able to identify patients with a significantly higher risk of not achieving CCyR and of progression, after 18 months of therapy. As the score is easy to calculate with only two variables needed that are routinely measured it is a simple way to alert physicians to the need for closer monitoring of the patient.
Hoffmann:Novartis Pharma: Research Funding. Turkina:Bristol Myers Squibb: Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Novartis Pharma: Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau. Baccarani:Bristol Myers Squibb: Honoraria; Novartis Pharma: Research Funding.
