TABLE 3

Hazard Ratios, 95% CIs, and P Values from Cox Regression Models on THTF, Grouping Patients According to Quartiles

						Univariate^*		Adjusting for clinical variables^†
Variable	n	Quartile range	Events (n)	Median THTF (mo)	% patients with no events at 3 y (±SE)	HR	P^‡	HR	P^‡
Baseline sum of SUV_max							<0.001		<0.001
First quartile	20	0, 0.6	11	64	75 ± 10	Reference		Reference
Second quartile	18	0.6, 4.2	10	38	56 ± 12	1.4 (0.57, 3.2)		1.4 (0.57, 3.3)
Third quartile	19	4.2, 14.4	12	29	42 ± 11	1.8 (0.79, 4.1)		2.1 (0.85, 5.2)
Fourth quartile	19	14.4, 181	18	7	11 ± 7	6.2 (2.8, 13.6)		7.4 (3.0, 18.2)
Baseline maximum SUV_max							<0.001		<0.001
First quartile	19	0, 0.5	10	85	79 ± 9	Reference		Reference
Second quartile	20	0.5, 2.4	12	29	50 ± 11	1.6 (0.70, 3.8)		1.5 (0.64, 3.6)
Third quartile	18	2.4, 5.0	14	12	33 ± 11	2.9 (1.3, 6.6)		4.0 (1.6, 9.8)
Fourth quartile	19	5.0, 21.9	15	13	21 ± 9	3.8 (1.7, 8.5)		5.3 (2.1, 13.2)
Baseline average SUV_max							0.007		<0.001
First quartile	19	0, 0.3	10	85	74 ± 10	Reference		Reference
Second quartile	19	0.3, 1.4	13	29	47 ± 11	1.9 (0.84, 4.4)		1.6 (0.69, 3.9)
Third quartile	19	1.4, 2.8	14	12	37 ± 11	2.5 (1.1, 5.8)		3.2 (1.2, 8.0)
Fourth quartile	19	2.8, 18.1	14	14	26 ± 10	2.9 (1.3, 6.7)		3.8 (1.5, 9.7)
Baseline disease sites (n)							<0.001		<0.001
First quartile	28	0, 1	15	59	64 ± 9	Reference		Reference
Second quartile	12	1, 2	6	64	67 ± 14	1.2 (0.46, 3.1)		1.8 (0.59, 5.3)
Third quartile	19	2, 8	14	20	42 ± 11	2.2 (1.03, 4.5)		2.6 (1.1, 6.0)
Fourth quartile	17	8, 10	16	7	6 ± 6	6.4 (3.0, 13.5)		6.7 (2.5, 17.8)
Overall	76		51	26.5 (15.5, 46.6)	46 ± 6

↵* Univariate analysis of association between each PET variable (grouped into quartiles) and THTF.
↵† Assessment of association between each PET variable and THTF, adjusting for clinical variables including age, baseline Gleason score, prostate-specific antigen level, and alkaline phosphatase level.
↵‡ P values were from test for trend of 4 quartiles.
HR = hazard ratio.
Data in parentheses are 95% CIs.