Table 3.

Results of univariate Cox regression and final multivariate Cox model assessing predictors of local recurrence

Variable	Univariate			Multivariate
	HR	95% CI	P	HR	95% CI	P
Age, y	0.99	0.95-1.03	.64
Female sex	1.24	0.49-3.11	.65
Infratentorial location	1.10	0.32-3.76	.88
Radio-resistant	0.56	0.19-1.68	.30
Prior radiation to same lesion	2.13	0.28-16.0	.46
Parallel chemotherapy	0.49	0.14-1.67	.25
Parallel immunotherapy	0.65	0.09-4.86	.67
Parallel chemotherapy or immunotherapy	0.48	0.16-1.45	.19	0.33	0.09-1.22	.10
Time from surgery to SRS, mo	0.98	0.58-1.65	.94
Subtotal resection	1.73	0.40-7.50	.46	4.03	0.84-19.3	.08
PTV volume, cm3	1.02	0.99-1.05	.18
PTV volume >12 cm3	3.10	1.12-8.57	.03	3.28	1.09-9.83	.03
PTV coverage, %	0.99	0.93-1.05	.67
Conformity index	0.96	0.76-1.21	.72	0.97	0.91-1.05	.49
NCI	0.97	0.79-1.20	.80
Number of intact brain metastases treated concurrently	0.95	0.74-1.21	.68

Abbreviations: CI = confidence interval; HR = hazard ratio; NCI = new conformity index; PTV = planning target volume; SRS = stereotactic surgery.

All variables with P < .7 on univariate analysis were included in the initial multivariate model. Variables that did not improve the Akaike Information Criterion were excluded from the final multivariate model.

Bold indicates statistical significance at P < .05.