Metastasectomy versus radiation of secondary sites in stage IV breast cancer.

Abstract

1094 Background: Prospective trials have yielded mixed results on the utility of surgery in metastatic breast cancer (mBC). Thus far, however, studies have focused primarily on the impact of lumpectomy or mastectomy. We previously showed that a combined approach involving resection of primary and secondary sites (i.e. ‘metastasectomy’) in patients with limited mBC was associated with improved overall survival (OS). We sought to evaluate the effect on OS of two approaches to loco-regional therapy (LRT) at secondary sites: metastasectomy versus radiation therapy. Methods: This is a retrospective analysis of patients diagnosed with mBC from 2010-2017 using the National Cancer Database. We identified 5 cohorts of patients by site of metastasis: mBC involving only 1) bone, 2) brain, 3) liver, or 4) lung; and 5) patients with metastasis involving >1 site. For each cohort, we used Kaplan-Meier (KM) models with log-rank testing to evaluate differences in OS, by the LRT approach at secondary sites (radiation versus metastasectomy). Prior to KM modeling, chi-squared statistics were used in each cohort to assess whether age, race, Charlson/Deyo score (CDS) for comorbidity, and receptor subtype were potential confounders of survival. The KM models were adjusted accordingly, as per the table below. Results: 53.4%) were between 50-70 years old, White (n=53,409, 78.9%), and had hormone receptor (HR)-positive/HER2 receptor-negative breast cancer. N=12,362 patients received radiation therapy at either the bone, brain, liver, or lung; while n=2674 underwent surgical resection of a metastatic site. Of patients with metastasis to 1 site (n=44,451), n=30,341(68.3%) involved the bone, n=1,119 (2.5%) involved the brain, n=5,227 (11.8%) involved the liver, and n=7,764 (17.5%) involved the lung. N=24,017 patients had metastatic disease involving > 1 site. KM modeling revealed superior OS of patients undergoing metastasectomy versus radiation of secondary sites in all 5 cohorts (p<0.05). The difference in median OS (ΔmOS) by LRT approach was more pronounced when metastasis involved only the liver (41.6 months) or lung (48.6 months), versus only the brain (9.7 months) or bone (8.7 months). Conclusions: Metastasectomy appears to confer a superior benefit for OS compared to radiation of secondary sites, particularly in patients with secondary site involvement limited to the liver or lung. More research is needed from prospective trials investigating surgical resection of metastatic sites.[Table: see text]