Treatment and Outcome Analysis of 2,904 Pateints from the EUTOS Population Based Registry

Introduction

Clinical studies clearly show that treatment with tyrosine-kinase inhibitors (TKI) greatly improve the prognosis of patients with CML. Detailed treatment recommendations have since been deduced from clinical trial results by the European LeukemiaNet (Blood 122:872-884).

However, little is known about whether these achievements can also be transferred to routine health care.

This work investigates if all CML patients are treated in agreement with current standards and wether they achieve outcomes at least as good as those in clinical trials.

Methods

The web-based registry aimed to collect all newly diagnosed adult patients with chronic phase Ph+ and/or BCR-ABL1+ CML in 20 countries or pre-specified regions covering ~ 92.5 million inhabitants. Overall 2904 patients were registered between 2008 and 2012.

Cytogenetic and molecular responses were analyzed using cumulative incidences considering death and progression as competing risks. Survival was analyzed using Kaplan-Meier curves and log rank tests.

Results

2342 patients were diagnosed in chronic phase and had follow-up data available. The median age was 55 years (range: 18 to 99 years) and 53% were male; 11% of the patients were at high risk of not achieving complete cytogenetic remission (CCyR) at 18 months according to the EUTOS score. According to the Euro score 10% of the patients were in the high risk group, 51% in the intermediate and 39% in the low risk group. 18% were included in clinical trials.

Treatment data of 1701 patients were contributed from 15 countries (85% of 2042 patients registered in those countries). As a first-line therapy 80% of patients received imatinib 15% nilotinib, 3% dasatinib and 3% hydroxyurea (HU). Of the patients receiving nilotinib or dasatinib 56% had been enrolled into clinical trials.

There were no significant differences between female and male patients regarding the first-line therapy. More than half of the patients who were treated with HU alone were aged 70 or older.

Time to first CCyR is known for 62% of patients. Median time to first CCyR was 10 months; after 12 months 57% (95% CI 54%-60%) and after 18 months 76% (95% CI 74%-79%) had achieved CCyR. Time to first CCyR differed significantly regarding the EUTOS score. Median time to first CCyR was 9 months for the low and 13 months for the EUTOS high risk group (p < 0.0001). After 18 months 78% (95% CI 75%-80%) of patients in the low and 69% (95% CI 60%-76%) of patients in the high risk group had achieved CCyR.

The patients' age did not have major influence on the time to first CCyR (p=0.8974, median time to first CCyR: 18 to 39 years: 9 months, 40-65 years: 9 months, older than 65 years: 11 months).

Time to first major molecular remission (MMR) could be calculated for 54% of patients. Median time to first MMR was 15 months. Cumulative incidence of MMR after 12 months was 41% (95% CI 38%- 44%).

The median observation time of living patients was 29 months. 187 patients died (8%). Probability of OS for all patients at 12, 24 and 30 months was 97% (95% CI 96% - 97%), 94% (95% CI 93% - 95%) and 92% (95% CI 90% - 93%), respectively.

108 patients progressed of whom 62 subsequently died, while 125 patients died without prior progression. Probability of PFS for all patients at 12, 24 and 30 months was 95% (95% CI 94% - 96%), 92% (95% CI 91% - 93%) and 90% (95% CI 88% - 91%), respectively.

Of the 187 patients who died, 33% died after progression. 67% died without prior progression. The probability of progression and subsequent death was 1% (95% CI 1% - 2%) after 12 months and 2% (95% CI 2% - 3%) after 24 months. The probability of dying without prior progression was 2% (95% CI 2% - 96%) after 12 months and 4% (95% CI 3% - 5%) after 24 months.

Discussion

The EUTOS population based registry provides the first unselected sample of adult Ph+ and/or BCR/ABL1+ adult CML patients in Europe. It shows that in Europe the success reported from commercial and academic studies is transferred to the general population. The majority of patients were treated first-line with imatinib, which was the only TKI approved for first-line use by the EMA during most of the registration period. Probabilities of PFS and OS in the registry are comparable to those in clinical trials. The importance of calculating both overall survival and leukemia-related survival is highlighted, since many patients die from different causes related or unrelated to leukemia and to treatment, but without progression.