Clinical Characteristics of Elderly Chronic Myeloid Leukemia Patients in Korea; Based on the National Health Information Database

Objectives

Aging involves progressive impairments in the functional reserve of multiple organs, which might also affect drug metabolism and pharmacokinetics. Elderly might experience more drug adverse events than younger patients. Thus, we evaluated clinical characteristics and tyrosine kinase inhibitor (TKI) prescription pattern in Korean elderly chronic myeloid leukemia (CML) patients, using National Health Information Database "NHID".

Methods

This study included patients 70 and older who were diagnosed with Philadelphia chromosome positive CML (ICD code C92.1) from 2005 to 2013 and prescribed TKIs (which included imatinib, dasatinib, nilotinib and radotinib which were covered by the National Health Insurance Service (NHIS), currently). We excluded following cases; ①TKI was prescribed just once ② TKI was started with a higher dose than the recommended for chronic phase (imatinib, 400mg QD; dasatinib, 100mg QD; nilotinib, 300mg BID; radotinib 300mg BID). ③ interferon was treated prior to TKI ④ TKI dosage data was missing. We extracted age, sex, medical insurance premium (which is imposed by proportionate to income), prescribed medication list, and date of death from NHID. Medication possession ratio (MPR) was used for adherence. Each patients' laboratory result, disease phase (chronic phase, accelerate phase, or blastic crisis), or the reasons for TKI change or dose reduction were not available.

Overall survival (OS) was calculated by Kaplan-Meier methods. Additionally, Univariate and multivariate analysis were done with Cox proportional hazard regression, with following variables using SAS; sex (male vs female), age (70 ≤age<75, 75≤age<80, age≥80) medical insurance premium (high vs low), frontline TKI (imatinib vs other TKIs), and MPR (≥0.9 vs <0.9).

Result

A total of 385 patients were followed up median 52 (1-133) months until the end of 2016. The patients consisted of 54% male (n=208), median age of 75 (70-95), and 64 patients (16.6%) who were diagnosed after the age of 80. About 85% of the patients experienced imatinib. Dasatinib was prescribed to 107 patients (27.8%), nilotinib to 71 (18.4%), and radotinib to 16 (4.2%). Frontline TKI was maintained in 279 (72.5%) patients, and the others were treated with more than one TKI; 2TKIs, n=80 (20.8%); 3TKIs, n=22 (5.7%); 4TKIs n=4 (1%). Imatinib was selected as frontline TKI in 318 patients (81.8%). Median MPR to TKI was 0.89 (0.00-1.84), and 51% of patients' MPR was less than 0.9. Patients were treated median 42 (0-146) months, and 186 (48.3%) were expired during the follow up period.

Among 327 patients, 42 (12.8%) started with reduced dose of imatinib, and 285 (87.2%) started with imatinib 400mg QD. After median 2 (0-118) months, full dose imatinib reduced to mean 263mg±63 in 165 patients. Only 111 patients (34%) maintained imatinib. Dasatinib 100 mg was initiated in 89.9% (n=94) of 107 patients. Thirty patients (28%) did not change dasatinib dose. Others, reduced dose to mean 57±13mg after median 3 (0-33) months. Patients who were taking nilotinib or radotinib, full dose without dose reduction, were 31 (43%) and 7 (43%) respectively. Nilotinib was reduced to mean 325±62 after median 2(0-29) months, and radotinib to mean 357±162mg after median 4 (0-43) months.

The estimated OS at 5-year was 60.3% and 10-year was 24.6%. For both univariate and multivariate analysis, age and MPR associated OS significantly; Patients aged 80 and older and from 75 to 80 showed poor OS than those aged from 70 to 75 through adjusted HR 2.67[95% CI.I, 1.81-3.93] (p<0.0001) and 1.81[95% C.I., 1.29-2.53] (p=0.0006), respectively. Patients with high MPR (≥0.9) survived longer than those with low MPR (<0.9), and its adjusted HR was 2.34 [95% C.I. 1.70-3.23] (p<0.0001).

Conclusion

Based on NHID, 10-50% of TKIs were initiated with reduced dose in age 70 and older CML patients. Full dose was maintained only 28-43% of prescribed TKIs. Despite of this, the outcome was favorable showing that because the estimated OS at 5-year was 60.3%. Age and adherence to TKI were associated with OS in elderly CML patients.