Abstract P5-18-12: Comparison of treatment patterns and outcomes in metastatic breast cancer patients initiated on trastuzumab vs. lapatinib; a retrospective analysis

Background: Trastuzumab-or lapatinib-based therapies are recommended by the National Comprehensive Cancer Network (NCCN) guidelines as preferred agents for metastatic HER2-positive breast cancer. Few studies have compared treatment patterns, healthcare resource utilization (HRU), and costs in metastatic breast cancer (MBC) patients receiving trastuzumab vs. lapatinib. The objective of this study was to compare discontinuation rates, HRU, and costs in MBC patients (pts) initiated on trastuzumab vs. lapatinib. Methods: MBC adult women initiated on trastuzumab or lapatinib on/after 03/13/2007 (lapatinib FDA-approval date) were selected from the US-based PharMetrics® Integrated Database (2000–2011). Selected pts were required to be continuously eligible in their healthcare plan ≥6 months prior to and ≥30 days following trastuzumab or lapatinib initiation date. Pts were observed over the period spanning from trastuzumab or lapatinib initiation date (without restriction on line of therapy) up to the end of continuous health plan enrollment or data availability, whichever occurred first. Trastuzumab or lapatinib discontinuation rates (gap ≥45 consecutive days) were compared using multivariate Cox proportional-hazards models and reported as hazard ratios (HRs). Incremental HRU was estimated using negative binomial regression models and reported as incidence rate ratios (IRRs). Monthly cost difference (2010 USD; measured from a payer perspective) were estimated using multivariate generalized linear or two-part models. Since a large proportion of lapatinib users were previously treated with trastuzumab, a sensitivity analysis was conducted to control for the impact of prior trastuzumab use. Results: Among the 643 pts selected, 381 and 262 pts were initiated on trastuzumab and lapatinib, respectively. Of the 262 pts initiated on lapatinib, 171 (65%) were previously treated with trastuzumab. After adjustment for confounders (demographics, index year, prior MBC therapies, comorbidities, baseline HRU, and MBC duration), compared to trastuzumab users, lapatinib users had a higher rate of treatment discontinuation (HR = 1.57; p Sensitivity analyses showed that pts who were initiated on lapatinib without prior trastuzumab use had a higher incidence of inpatient admissions (IRR = 1.74; p = .002) and greater overall medical service costs compared to pts who were initiated on trastuzumab (adjusted difference=$3,059; p Conclusion: Overall, while MBC pts initiated on trastuzumab or lapatinib had similar healthcare costs, pts initiated on lapatinib had higher rates of treatment discontinuation and outpatient visits (not treatment administration related). Medical costs were significantly higher among pts initiated on lapatinib without prior trastuzumab use. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P5-18-12.