Table 4.

Univariate and multivariate Cox proportional hazards regression models for OS using a dataset of recurrent metastatic cases (n = 824)

Characteristic	Univariate analysis			Multivariate analysisa
	HR	95% CI	p value	HR	95% CI	p value
Age (ABC diagnosis), (≥ 60 vs. < 60 years)	1.66	1.40–1.98	< 0.001	1.42	1.19–1.70	< 0.001
Treatment line of fulvestrant (4th or more vs. 3rd vs. 1st and 2nd)	0.70	0.62–0.78	< 0.001	1.00	0.89–1.13	0.992
ABC diagnosis to fulvestrant use (≥ 3 vs. < 3 years)	0.16	0.13–0.19	< 0.001	0.14	0.11–0.18	< 0.001
Prior palliative chemotherapy (yes vs. no)	0.78	0.66–0.93	0.004	1.45	1.20–1.76	< 0.001
Histological or nuclear grade
(2 vs. 1)	1.73	1.41–2.12	< 0.001
(3 vs. 1)	2.58	2.03–3.27	< 0.001
Visceral metastasis (yes vs. no)	0.99	0.83–1.17	0.898
PgR (positive vs. negative)	1.14	0.94–1.38	0.200
HER2 (positive vs. negative)	0.87	0.65–1.17	0.360
DFI (≥ 5 vs. < 5 years)	1.00	0.84–1.18	0.950

OS overall survival, HR hazard ratio, CI confidence interval, ABC advanced breast cancer, PgR progesterone receptor HER2, human epidermal growth factor receptor 2, DFI disease-free interval

^an = 823 in the multivariate analysis (one patient was excluded because the date of recurrence was unknown. Histological/nuclear grade data were not included in this analysis because one-third of the data were missing