Impact of exclusion criteria for the DASISION and ENESTnd trials in the front-line treatment of a 'real-life' patient population with chronic myeloid leukaemia.

Both Dasision and ENESTnd trials had many exclusion criteria, with a possible selection bias compared with the real-life. To address the impact of this bias on the first-line treatment in the current clinical practice, we revised 207 unselected newly diagnosed chronic phase chronic myeloid leukaemia (CML) patients [M/F 108/99, median age 58.8 years, interquartile range 42.3-70.2] treated with front-line imatinib from June 2002 to June 2013 at our Institution, and evaluated how many of them would have been excluded from enrolment in the two trials. Twenty-eight patients (13.5%) should have been excluded by both trials because of polycomorbidities (12), severe cardiomyopathy (five), age > 80 with frailty (three), drug abuse (two) or other severe concomitant diseases (six). In addition, eight patients should have been excluded by Dasision due to isolated chronic obstructive broncopulmonar disease, and 19 patients should have been excluded by ENESTnd due to isolated diabetes (10), arrhythmia (four), acute myocardial infarction > 6 months before CML diagnosis (two), chronic pancreatic disease (two) and peripheral arterial obstructive disease (one). On the whole, 36 patients (17.4%) would have been excluded by Dasision trial and 47 (22.7%) by ENESTnd trial. The patients potentially not eligible for both trials were significantly older and with imatinib had a worse outcome compared with patients potentially eligible. Our data highlight that an automatic transposition of results available in clinical controlled trials into the frontline real-life management of CML patients should be regarded with caution.