LAUSR

To the Editor: Imatinib has transformed chronic myeloid leukemia (CML) prognosis with a survival rate of almost 83% at 10 years. The second and third generations of tyrosine-kinase inhibitors (TKIs) have enlarged the therapeutic possibilities. Dasatinib and nilotinib induce higher complete cytogenetic response and major molecular response (MMR) rates than imatinib as well as fewer progression to blastic phase. Bosutinib was first studied in resistance/intolerance to imatinib situations Ponatinib, the only third line TKI, has a BCR-ABL1 pan-inhibitor activity and is maintaining efficacy against T315I mutation. However, TKIs are associated with side effects (SEs) and may impair patient observance. Actual practices about TKIs switches are based on the 2013 European Leukemia Net guidelines. However, few studies have considered the prognostic impact of TKIs switches in daily practice. The aim of our study was to analyze the number of switches on a global cohort of patients treated in first line with imatinib outside a clinical trial. In this retrospective multicentric study, the patient cohort was established using BCR-ABL1 transcript follow-up curves from a multi-center network, centralized in the molecular biology laboratory of Rennes University Hospital. Two cohorts were constituted, 1 comprising patients who experienced one or more switches during follow-up, and the other including patients maintaining imatinib therapy. Patients had to be aged 18 or over and received a first line therapy by imatinib, started between 2006 and 2014. Patients who underwent ASCTs were excluded, as well as pregnant patients during follow-up. All patients provided written informed consent. Overall survival (OS) of patients according to the existence of TKIs switches was analyzed, as well as survival based on MMR obtaining in first line. Besides, this study analyzed the incidence of SEs during the different TKIs lines. Groups were compared using Student test and Mann-Whitney Wilcoxon tests. For qualitative variable, groups were compared by χ2 and Fisher tests. OS was represented using a curve with Kaplan-Meier estimator and its 95% CI. Survival curves by group were compared by a log-rank test. In total 192 patients were included, 87 patients in the switch group and 105 in the no-switch group (Figure 1A). The median follow-up time was 5.1 years. It was 4.5 years (0.8-10.2) in the no-switch group and 5.2 years [0.8-9.4] in the switch group. Median age at diagnostic was 61 years [24-88]. Both groups were similar concerning Sokal score (P = .56). OS at 5 years was 95% and 93% at 7 years (Figure 1B). When comparing OS according to the existence of switches or not, no difference was found between the 2 groups, with an OS at 5 years of 96% in the switch group and 93% in the no-switch group (P = .166) (Figure 1C). Data analysis in first line setting showed that MMR rate was significantly higher in the no-switch group with 91% against 28% (P < .0001), but reaching or not MMR in first line had no impact on OS (P = .168). Loss of MMR was seen in first line in the 2 groups with 42% in the switch group versus 3% in the other group (P < .0001). Median time to MMR in first line was 12 months in both groups. The main reason for switching to second line were SEs (46%) followed by absence of MMR (43%) and then by the loss of MMR (11%). Mutations rarely occurred with 5 cases in the switch group and 1 in the no-switch group. No mutations were detected during the subsequent lines. Distribution in second line was as follows: 70 patients were treated with dasatinib (81%), 16 patients with nilotinib (18%) and 1 patient with bosutinib (1%). Global MMR rate in second line was 67%, ie, 63% with dasatinib and 88% with nilotinib (P = .11): median time to MMR was 5 months [1-45] with dasatinib and 4 months [2-12] with nilotinib. The reasons for switching to third line were as follow: loss of MMR (n = 5, 16%), toxicity (n = 25, 78%), and primary resistance (n = 2, 6%). SE reported in first line with imatinib concerned 83% of patients, with 81% in the switch group and 85% in the noswitch group. Grade 3/4 toxicity affected 26% of patients exclusively in the switch group. Cutaneous toxicity was more frequently observed in the switch group, 25% versus 14% in the no-switch group (P = 0.05). Significantly more hepatic SEs were also reported in the switch group, 12% compared with 2% (P = 0.006), including 80% of grade 3/4 in the switch group. Conversely, fluid retention was more often described in the noswitch group, 42% versus 24% (P = .01). In second line, overall toxicity was similarly observed among patients treated either with dasatinib or nilotinib. Liver toxicity was reported in 4 patients (25%) with nilotinib, and pleural effusion in 17 (28%) patients with dasatinib. Cardiovascular event was documented in 4 patients treated with nilotinib. In third line, 21 patients out of 39 experienced grade 1 or 2 toxicity (54%). Abbreviations: ABL, Abelson; ASCT, allogenic stem cell transplantation; BCR, breakpoint cluster region; BREHAT, Bretagne Réseau Expertise Hématologie; CCyR, complete cytogenetical response; CML, chronic myeloid leukemia; ELN, European leukemia net; MMR, major molecular response; OS, overall survival; Ph, chromosome Philadelphia [translocation t(9;22)]; TKI, tyrosine-kinase inhibitors; SEs, side effects Received: 1 July 2018 Accepted: 10 July 2018