Mechanism of drug resistance in relation to site of metastasis: Meta-analyses of randomized controlled trials in advanced breast cancer according to anticancer strategy.

BACKGROUND: Breast cancer is heterogeneous at different levels: biologic subtypes, intratumoral areas, and sites of metastases. Randomized controlled trials (RCTs) classify metastatic sites as visceral or non-visceral, but this has little influence in treatment decisions, particularly in the absence of clinical urgency. Indeed, it is unclear if response to treatments differs among sites of metastases.

PATIENTS AND METHODS: RCTs investigating 3 different anticancer strategies in metastatic breast cancer were identified: (1) new hormonal therapy, (2) new targeted therapies in hormone receptor positive tumours (everolimus or palbociclib), and (3) new anti-HER2 therapies. RCTs reporting hazard ratios (HR) for Progression Free Survival (PFS) and Overall Survival (OS) for sub-groups based on sites of metastases were weighted using generic inverse variance approach, and pooled in meta-analyses using Revman 5.3. Subgroup difference was tested with Chi2 statistics.

RESULTS: Eleven RCTs (6701pts.) qualified. There was a significant difference in PFS between women with visceral versus non-visceral metastases when two endocrine strategies were compared, with benefits limited to women with visceral metastases [Pooled HR 0.85; 95% CI, 0.77-0.95 versus 1.02 (0.88-1.18) for non-visceral; p(difference)=0.05]. However, combination of an endocrine therapy and a targeted therapy was associated with better PFS compared to endocrine therapy alone for both groups [HR 0.51 (0.43-0.60) versus 0.45 (0.36-0.56) for non-visceral; p(difference)=0. 36]. Novel HER-2 targeted therapies were associated with significantly better PFS and OS only in visceral metastases [HR 0.59 (0.52-0.66) versus 0.71(0.44-1.13) for non-visceral, p(difference)=0.45, for PFS; and 0.64 (0.56-0.73) versus 0.82 (0.57=1.19) for non-visceral, p(difference)=0.20, for OS].

CONCLUSION: Combination of targeted agents and endocrine therapy results in concordant, superior PFS suggesting targetable endocrine resistance across metastatic sites. Discordant responses with endocrine strategy alone support use of targeted therapy, rather than change in endocrine agent at disease progression. HER2 targeted therapies may be less effective in areas of poor vascularization.