Baseline Characteristics of CML Patients Accross Europe - Comparing Real-World Patients with Patient Collectives Included in Clinical Trials

Introduction:

Most of the knowledge about treatments and outcome of CML patients originates from clinical studies. To get new and unbiased insights in the epidemiology, treatment and outcome of CML, the EUTOS population-based registry of newly diagnosed CML patients was established, - as part of the European Treatment and Outcome Study (EUTOS) for CML.

The aim was to collect the data of all adults with newly diagnosed CML, irrespective of treatment and of enrolment in studies.

Patients and Methods:

The EUTOS population-based registry collected data of newly diagnosed CML patients, 18 years or older, over a specified period of time from 2008 till 2012 living in defined regions. The data were collected by 22 study groups in 20 European countries. Data were gathered via a web-based CRF-system.

For comparison we used the already published data from five Company-sponsored registration studies IRIS (O’Brien et.all, NEJM, 2003), TOPS (Cortes et al, JCO, 2009) ENESTnd (Saglio et al, NEJM, 2010), DASISION (Kantarjian et al, NEJM, 2010) and BELA (Cortes et al, JCO, 2012), from three Investigator-sponsored studies GIMEMA (Castagnetti et al, JCO, 2010 and Gugliotta et al, Blood, 2011), French SPIRIT (Preudhomme et al, NEJM, 2010) and German CML IV (Hehlmann et al, JCO, 2011) and from two single referral centers HAMMERSMITH (De Lavallade et al, JCO, 2008) and MDA (Jain et al, Blood, 2013).

Results:

Till 15.05.2014 2978 patients were registered in the EUTOS Population-based registry. 94.3% of the patients were diagnosed in chronic phase (CP), 3.6% in accelerated phase (AP), and 2.2% in blastic phase (BP). For the calculation of the prognostic scores 361 patients had to be excluded because they were pretreated. For the comparison we used 2350 patients in Chronic Phase with laboratory values before any treatment.

54% of the patients in the EUTOS Population-based registry were male, less than in all studies (56.6 - 60.6%). The median age at diagnosis was 56 years, higher than in all studies (46 - 55). In EUTOS the proportion of patients more than 60 years and more than 65 years old was 40.4 % and 21.9 % respectively. Similar data were rarely reported in all other studies. Median value of the spleen size below costal margin was 0. 46.1% of the patients had a palpable spleen and 15.2% had a spleen size ≥ 10 (spleen size is always reported in cm under costal margin in this abstract). The % of palpable spleen is only reported by IRIS, 25.0% and by the FRENCH Spirit group, 49.8%. The median spleen is only reported by GIMEMA, 2.0. Spleen size ≥ 10 is reported by IRIS, 6.0%, ENESTnd, 12.4% and HAMMERSMITH 25.5%. While the median values for Platelets and Hemoglobin show no big differences, the median WBC in EUTOS is 83.9 x10⁹/l and in the Company-sponsored registration studies: IRIS 18-20 x10⁹/l , in ENESTnd 23-26 x10⁹/l, in DASISION 23-25 x10⁹/l , and in BELA 22-23 x10⁹/l, in the Investigator-sponsored studies: GIMEMA 55 x10⁹/l , in the FRENCH SPIRIT 83-104 x10⁹/l , in the GERMAN CML IV 75-91 x10⁹/l , and in the single referral center study HAMMERSMITH 140 x10⁹/l, clearly indicating that in company-sponsored, registration studies, the reported values of the WBC were not recorded prior to any treatment. The median values for Blasts, Basophils and Eosinophils show also not so big differences. The % of Sokal low risk patients is in EUTOS with 34.5% lower than in all studies (35.2 - 60%) with the exception of HAMMERSMITH 28.9%.

Discussion:

The EUTOS Population-based registry provides the first European wide real-world series of patients with newly diagnosed Ph+, BCR-ABL+ CML. The age and sex distribution and some baseline characteristics such as Sokal Score as well as median WBC count in the EUTOS population-based registry are different from many prospective studies. This should be taken in due consideration before extrapolating the results of treatment studies to real life. Spleen size, which is known as an important value for prediction, is only very rarely reported in clinical studies. With further follow-up, this registry will provide a population-based insight on treatment, survival, and causes of death.