Impact of absolute lymphocyte count on the efficacy of eribulin and capecitabine in patients with metastatic breast cancer.

Abstract

e13082 Background: In a phase 3 trial (Study 301), which compared eribulin and capecitabine for metastatic breast cancer did not show the superiority of one to the other on overall survival. Recently, a post-hoc analysis has revealed that efficacy of eribulin was greater in patients with high levels of absolute lymphocyte count (ALC), probably because the drug has immunologic mechanisms of action. Here we investigated whether the favorable impact of high ALC levels on the efficacy was specific to eribulin or shared with capecitabine. Methods: A total of 275 patients with metastatic breast cancer who were treated with eribulin (n = 125) or capecitabine (n = 150) in our hospital between July 2011–April 2019 were retrospectively analyzed. Progression-free survival (PFS) was compared between patients with higher ALC levels (≥1500/μL) and those with lower ALC levels ( < 1500/μL) in each treatment group. Then, we investigated how hazard ratio (HR) of PFS (eribulin vs. capecitabine) changed when we excluded patients with low ALC levels one by one; a scatter plot was made (X = cut-off value of ALC, Y = adjusted HR). PFS was compared by using the Cox proportional hazards model and covariates included age, hormonal and HER2 status, presence/absence of liver and other visceral metastasis, the number of involved organs and the number of previous chemotherapy regimens. Results: In eribulin group, PFS was significantly better in patients with higher ALC than those with lower ALC (adjusted HR, 0.50 [95% confidence interval (CI), 0.29–0.85]; p = .010). Contrarily, in capecitabine group, PFS was not significantly different between the two groups (adjusted HR, 0.86 [95% CI, 0.58–1.28], p = .456). According to the scatter plot, as the cut-off value of ALC (X) was increased, the hazard ratio of PFS (Y) changed biphasically, being almost flat when X was less than 1500, and a linear decrease when X was 1500 or more. Adjusted HR were (i) 1.07 [95% CI, 0.64–1.82], (ii) 0.79 [95% CI, 0.29–2.11] and (iii) 0.67 [95% CI, 0.18–2.41] when inclusion criteria of patients were (i) ALC ≥1500/μL, (ii) ALC ≥1900/μL and (iii) ALC ≥2200/μL, respectively. Conclusions: The efficacy of eribulin was significantly better in patients with higher ALC levels, whereas that of capecitabine was not significantly different. “The more, the better” relationship between ALC and the efficacy of eribulin (vs. capecitabine) was recognized when ALC was 1500/μL or more.