Table 3. Tumor LINE-1 hypomethylation and patient survival according to colorectal cancer location and MSI status.

	No. of cases	Colorectal cancer-specific mortality			Overall mortality
		No. of events	Univariable HR (95% CI)	Multivariable HR (95% CI)a	No. of events	Univariable HR (95% CI)	Multivariable HR (95% CI)a
MSI-low/MSS proximal colon cancer
LINE-1 hypomethylation (20% decrease as a unit)	440	153	1.77 (1.26-2.50)	1.44 (1.04-2.01)	259	1.41 (1.07-1.84)	1.27 (0.97-1.66)
Pb			0.001	0.030		0.014	0.08
MSI-high proximal colon cancer
LINE-1 hypomethylation (20% decrease as a unit)	181	23	4.51 (1.78-11.4)	6.14 (2.27-16.6)	85	1.62 (0.96-2.74)	2.30 (1.30-4.06)
Pb			0.002	0.0003		0.07	0.004
MSI-low/MSS distal colon cancer
LINE-1 hypomethylation (20% decrease as a unit)	388	108	1.21 (0.84-1.76)	1.13 (0.76-1.66)	203	0.97 (0.73-1.28)	1.06 (0.80-1.41)
Pb			0.31	0.55		0.82	0.69
MSI-high distal colon cancer
LINE-1 hypomethylation (20% decrease as a unit)	21	3	3.05 (0.74-12.6)	2.76 (0.75-10.1)	8	2.38 (0.89-6.36)	2.00 (0.87-4.60)
Pb			0.12	0.13		0.08	0.10
Rectal cancer
LINE-1 hypomethylation (20% decrease as a unit)	287	95	0.92 (0.60-1.40)	0.87 (0.57-1.34)	162	0.79 (0.57-1.09)	0.72 (0.52-1.00)
Pb			0.69	0.52		0.15	0.052

CI, confidence interval; HR, hazard ratio; LINE-1, long interspersed nucleotide element-1; MSI, microsatellite instability; MSS, microsatellite stable.

^aThe multivariable Cox regression model initially included sex, age, year of diagnosis, family history of colorectal cancer in parent or sibling, disease stage, tumor differentiation, CpG island methylator phenotype, and KRAS, BRAF, and PIK3CA mutations. A backward elimination with a threshold of P = 0.05 was used to select variables in the final models.

^bP value was calculated by the Wald test (two-sided).