TABLE 2.

Results of Cox Regression Models

		Bivariate Cox regression						Multivariate Cox regression
		PSA PFS			OS			PSA PFS			OS
Cox model	Parameter	HR	95% CI	P	HR	95% CI	P	HR	95% CI	P	HR	95% CI	P
Cycle 2	TTV	1.42	1.10–1.83	0.006	2.11	1.50–2.95	<0.001	1.45	1.11–1.89	0.006	2.37	1.62–3.47	<0.001
	NLs	3.17	0.82–12.3	0.1	5.81	1.93–17.5	0.002	3.30	0.86–12.7	0.083	7.67	2.77–21.3	<0.001
Cycle 3	TTV	1.99	1.38–2.87	<0.001	1.98	1.22–3.20	0.005	1.99	1.36–2.89	<0.001	2.19	1.24–3.88	0.007
	NLs	3.34	0.84–13.3	0.087	4.92	1.3–18.6	0.019	2.10	0.53–8.28	0.3	5.50	1.46–20.7	0.012

Data are fitted bivariate Cox regressions for PSA PFS and OS conditional on TTV or NLs at start of cycle 2 or 3, and fitted multivariate Cox regressions for PSA PFS and OS conditional on both TTV and NLs combined. All regressions are adjusted for changes in PSA level measured at start of corresponding cycle relative to baseline.