Treatment and Management of Chronic Myelogenous Leukemia in Patients Cared for in a Large HMO

Introduction

CML management has rapidly changed in the last decade. Data are critically needed to characterize disease treatment, management, and drug use (particularly TKIs, tyrosine kinase inhibitors) to ultimately inform how to improve outcomes in patients with CML. Our objective was to describe practice trends in a large community based health plan serving over 3.2 million members in southern California, Kaiser Permanente (KPSC).

Methods

Patients were identified from the health plan’s NCI-SEER affiliated cancer registry. The cohort consisted of 257 CML adult patients (18+ years, chronic phase, no prior cancer history) all diagnosed between January 2001 and December 2012 and followed through December 31, 2013. We collected data from electronic health records and chart reviews including demographics, TKI use and adherence, healthcare utilization and clinical outcomes. Medication possession ratio (MPR) was used to measure adherence, calculated as the number of days supplied divided by the number of days between first and last dispense date of any TKI. We classified patients into MPR>90% (indicates good adherence) and MPR<90%. Discontinuation was ascertained from providers’ notes in the medical charts.

Results

Of the 257 patients with newly diagnosed chronic phase CML, 111 were female (43%). About 58% (n=148) were non-Hispanic Whites, and 23% (n=59) were Hispanic. About half (48%, n=124) had one or more existing comorbidities. Over 90% of the cohort initiated TKI therapy within 3 months of diagnosis, and mean MPR was 88% (s.d. 18%). About two-thirds of the patients (63%, N=157) had an MPR>90% for TKI use. Virtually all patients (96%) started on imatinib. Of the 59 cases diagnosed 2010-2012, only 14 used dasatinib or nilotinib as first line therapy. Half of the cohort (n=107) discontinued first line treatment (mainly imatinib). Reasons for discontinuation included general adverse effects (e.g., skin rash, muscle cramps), cardiovascular, lung or abdominal effects (n=41); incomplete cytogenetic or molecular response (n=38); drug non-response, provider recommended, not compliant (n=21); or BMT/stem cell transplant (n=7).

The cohort was followed a maximum of 13 years (mean 5.1 years, s.d. 3.1 years). The total number of hospitalizations was 351 among patients diagnosed 2001-2006; 113 diagnosed 2007-2009; and 71 diagnosed 2010-2012. The median length of each hospitalization was 10 days (s.d. 30.6) during the follow-up years. Among imatinib users, the median number of overall outpatient visits per year varied from 20 visits/year for patients diagnosed 2001-2006 to 11 visits/year for patients diagnosed 2010-2012. Overall, 18 patients underwent bone marrow and/or stem cell transplant (all diagnosed 2001-2006). We examined progression and overall survival in patients diagnosed in 2001-2006 and exposed to imatinib. The rate of progression to accelerated phase or blast crisis was lower in patients with MPR>90% (10.0/1,000 person-years) than with MPR<90% (14.2/1,000 person-years). Similarly, patients with MPR>90% had a lower mortality rate (25.9/1,000 person-years) than patients with MPR<90% (39.2/1,000 person-years). Kaplan Meier curves for overall survival also demonstrate better survival, but not significantly (log rank P=0.35), among those with MPR>90%.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal

Discussion

Most patients initiated TKI therapy within 90 days of diagnosis in this health plan. Our results show median length of each hospitalization was 10 days across all follow-up time. Despite having full pharmacy coverage, the percent of subjects with MPR>90% was moderate (64%). Our results also suggest that subjects with MPR>90% for TKIs had lower progression rates and greater overall survival, but the results were not significant, and need to be confirmed in a larger study. Next steps include examining the frequency of cytogenetic and molecular testing and how this might impact outcomes, and factors associated with lower TKI medication possession ratios.