Excellent Outcome and Good Tolerability of Long-Term Imatinib in Patients with Chronic Myeloid Leukemia (CML)

Introduction: The introduction of the tyrosine kinase inhibitors (TKIs) into clinical practice in the late 1990s has considerably improved both survival and quality of life for patients with CML. Imatinib was the only TKI available for several years with no useful drug treatment for patients with resistance and/or intolerance. Despite the lack of alternative agents the 8-year follow-up of the IRIS trialshowed that only 55% of patients were still on imatinib. The majority of those who discontinued did so for lack or loss of response rather than intolerance, suggesting that imatinib is very well tolerated in the long-term. This is particularly pertinent today as controversy persists as to the best agent for newly diagnosed patients. There is not only increasing evidence that the second and third generation TKIs are associated with more severe adverse events, but generic imatinib is now available in many countries at considerably less expense. We report our experience of treating 45 patients with continuous imatinib for more than 10 years.

Methods: We interrogated our single centre database of all patients treated with TKIs for CML at our centre from June 2000 to March 2015. From a total of 832 patients we identified 188 CML who had received only imatinib. Of these, 45 patients had received treatment for more than 10 years.

Results: The median duration of imatinib therapy was 6 years in the total cohort of imatinib only patients and 11 years (range 10-14.7) in the study group. All 45 patients were in chronic phase at diagnosis: the median age was 45.4 years (range 26-72). Forty patients were evaluable for Sokal scoring, with 19, 13 and 8 identified as low, intermediate and high risk respectively. The median imatinib starting dose was 400 mg daily. The proportions of patients who achieved optimal responses (OR), as defined by the ELN at 3, 6 and 12 months from start of imatinib, were 88.2%, 78.8 and 56.1% respectively. At 10 years the probabilities of CCyR, MR3, MR4, MR4,5 and MR5 were 100%, 100%, 100%, 100%, 75.6% respectively. The 10 year probability of obtaining a sustained (at least 2 years) molecular response was 100%, 64.4%, 35.6% and 15.6% for MR3, MR4, MR4.5 and MR5 respectively. In patients who were not optimal responders at one or more time points (n=21), the median dose of imatinib was ≥400 mg in the first 12 months of treatment; for 13/21 higher dosages (range 600-800 mg daily) were prescribed. We found a significant correlation between a low or intermediate Sokal score at diagnosis and OR at 3 months (p=0.012). No correlation was found between Sokal score and OR at 6 or 12 months. No statistically significant association was found between an optimal response at 6 or 12 months and the future depth of responses. In fact, the overall rates of sustained MR4.5 for patients optimal responders at 6 and 12 months were 52% and 52% versus 41.6% and 50% for non optimal responders at the same time points.

Grade 4 toxicities and secondary malignancies were not observed during the follow-up. Seven pts (15.5%) experienced grade 3 events, including 1 each of supraventricular tachycardia and anemia, and neutropenia, fatigue and hypophosphataemia were each seen in 2 patients. The most frequent adverse event of any grade was fatigue (36% of pts), followed by anemia (27%) and neutropenia (18%). The cumulative probability of common side effects increased over the time. Cardiovascular events were mostly grade 1-2 palpitations and hypertension. At last follow-up, all pts were alive.

Conclusions: Our patient cohort analysis confirms long term safety and tolerability of imatinib after 10 years of therapy. The majority of side effects were grade 1-2 and some increased in incidence over the time. The most frequent adverse events were hematological. Imatinib continues to provide an excellent therapeutic outcome granting deep molecular responses even in some patients deemed to be poor risk at diagnosis. ELN optimal response status at 6 and 12 months was not associated with prediction of the future depth of response, in this very good risk population (majority of patients in optimal response at 3 months).