Table 3. Multivariate analyses.

Modeling according to clinical factors significant on univariate analysis showed that low-to-moderate EGFR amplification was an independent adverse prognostic factor for overall survival, both in all cases with known adjuvant therapy status (upper) and the subset treated with TMZ (lower). HR = hazard ratio, CI = confidence interval, EGFR = epidermal growth factor receptor; CEP7 = centromeric enumeration probe for chromosome 7, TMZ = temozolomide.

All cases with known adjuvant therapy status (N = 465)
Variable	HR	95% CI	P
Patient age > 45 years versus ≤ 45 years	2.46	1.79-3.38	< 0.0001
Resection versus biopsy	0.42	0.34-0.52	< 0.0001
TMZ versus no TMZ	0.39	0.30-0.46	< 0.0001
EGFR:CEP7 ratio = 2-20 versus non-amplified or ratio > 20	1.42	1.07-1.89	0.017
Only cases treated with TMZ (N = 312)
Variable	HR	95% CI	P
Patient age > 45 years versus ≤ 45 years	2.09	1.41-3.09	0.0002
Resection versus biopsy	0.39	0.29-0.51	< 0.0001
EGFR: CEP7 ratio = 2-20 versus non-amplified or ratio > 20	1.56	1.10-2.22	0.013