Analysis of Glucose Profiles in Imatinib-Resistant or Intolerant Chronic Myelogenous Leukemia (CML) Patients (pts) Treated with Nilotinib: Lack of Correlation between Glucose Levels and Nilotinib Efficacy.

Background: Nilotinib is an oral selective BCR-ABL inhibitor, more potent than imatinib (IC50 <30 nM) and active against 32/33 BCR-ABL imatinib-resistant mutants. Recent phase I/IIa study results demonstrated that nilotinib is well tolerated and has significant efficacy in imatinib-resistant or intolerant Ph+CML pts. Although nilotinib and imatinib have similar chemical structure and mechanism of action, differences in safety profiles, including effects on glucose metabolism, have been observed. In the pivotal phase II nilotinib study, the incidence of Grade 3/4 hyperglycemia was 10.7% in CP and 4.4% in AP pts. No pts required dose adjustment and/or interruption/discontinuation due to hyperglycemia. A recent publication involving small number of pts (n=13) in the phase I trial, suggested that blood glucose (BG) levels could be used as an indirect predictor for response to nilotinib (1).

Methods: Based on this observation, a post-hoc analysis of BG levels in the pivotal phase II study investigating safety and efficacy of nilotinib in imatinib-resistant or -intolerant CML pts was conducted.

Results: This analysis included 447 pts [51% male; median age 58 (21–85) yrs]. BG levels were collected once weekly for the first 2 mos and then once every 2 wks for the remainder of the study period. Only 20% were fasting samples. The median treatment durations were 341 (CP) and 190 days (AP). Although some pts has increase in non-fasting BG levels during nilotinib treatment, mean BG for the entire cohort (n=477) were within normal fasting range (mean 82.5±61.0 mg/dL). At baseline, 2 (0.6%) CP and 1 (0.8%) AP pt had Grade 3/4 hyperglycemic events (BG >250 mg/dL). During the nilotinib treatment period, 12.5% of CP and 6.7% of AP pts had Grades 3/4 hyperglycemic events. No episodes of diabetes-related serious adverse events occurred (ketoacidosis, hyperosmolar coma or diabetes-related hospitalization) and no pt discontinued nilotinib due to hyperglycemia. CP pts who presented with elevated BG had higher body mass index (BMI >30), p=0.0442. There was no significant increase in mean body weight during the treatment period in either CP or AP pts. There was no correlation between high BG levels (Grade 3/4) and response to nilotinib treatment in either the CP or AP cohorts.

Conclusion: There were no clinically significant changes in mean BG levels of pts receiving nilotinib. The highest BG levels occurred in pts with increased BMI. Although an increase in the incidence of Grade 3/4 hyperglycemia was observed, no hyperglycemia complications occurred. Importantly, no correlation was found between BG levels and nilotinib response. Based on the results from this large phase II study, BG levels cannot be used to predict hematological or cytogenetic responses to nilotinib. Breccia M, et al. Leuk Res. 2007 Mar 21; [Epub ahead of print]