Table 3.

Multivariate Proportional Hazards Analyses: Overall Survival for Patients With Node-Negative Disease in either the RTOG 9704 or UCLA Stages I and II Pancreatic Cancer TMAs

Histone Modification Group	RTOG 9704 TMA				UCLA Stages I and II TMA
	Hazard Ratio	95% CI	P*	No. of Patients	Hazard Ratio	95% CI	P*	No. of Patients
Low H3K4me2	2.70	1.27 to 5.74	.01	69	4.46	2.07 to 9.59	< .001	65
Low H3K9me2	1.35	0.75 to 2.43	.32	70	2.26	1.04 to 4.94	.040	63
Low H3K18ac	1.89	1.05 to 3.43	.04	69	2.70	1.37 to 5.31	.004	65
H3K4me2 and H3K18ac†	3.61	1.57 to 8.30	.003	68	5.00	2.25 to 11.1	< .001	65

NOTE. Separate analyses were performed on either the RTOG 9704 TMA or the UCLA stage I and II TMA for only the subset of patients with node-negative cancer. HR of 1 indicates no difference between the two groups of patients for the listed histone variable, whereas an HR > 1 indicates an increased risk of death/failure for histone group listed. In addition to histone status, the following covariates were used in the analysis of each TMA: tumor differentiation (not poor v poor), T stage (T1-T2 v T3-T4), age (continuous), sex, and surgical margins (negative/unknown v positive). Karnofsky performance status (90-100 v 60-80) and treatment arm (gemcitabine v fluorouracil) were additional covariates used in the model for RTOG 9704.

Abbreviations: RTOG, Radiation Therapy Oncology Group; UCLA, University of California, Los Angeles; H3K4me2, histone H3 lysine 4 dimethylation; H3K9me2, histone H3 lysine 9 dimethylation; H3K18ac, histone H3 lysine 18 acetylation.

^*P value from χ² test that used the Cox proportional hazards model.

^†One or both low level.