[698] Hereditary Predisposition to LINE-1 Hypomethylated Colorectal Cancer: Potential Clinical Utility of Tumor LINE-1 Methylation Test for Familial Cancer Risk Assessment

Reiko Nishihara, Charles Fuchs, Shuji Ogino. Dana-Farber Cancer Institute, Boston, MA; Brigham and Women's Hospital, Boston, MA

Background: Beyond known familial colorectal cancer (CRC) syndromes, the mechanisms underlying the elevated CRC risk associated with CRC family history (seen in ∼20% of all newly diagnosed CRCs) remain largely unknown. Because LINE-1 hypomethylated CRCs are associated with young age of onset, we hypothesized that some LINE-1 hypomethylated CRCs may represent a new familial CRC predisposition syndrome, and that LINE-1 methylation test may help familial CRC risk assessment (beyond MSI).
Design: Using a population of 134,079 individuals with 3,184,415 person-years of follow-up, we prospectively examined the association between CRC family history and the risk of CRC (N=1,224) according to tumor LINE-1 methylation level. We excluded individuals with polyposis and those with inflammatory bowel diseases. We excluded MSI-high CRC from outcome, to eliminate the influence of Lynch syndrome (i.e., hereditary susceptibility to MSI-positive cancer).
Results: CRC family history significantly increased the risk of LINE-1 methylation-low CRC (P<0.001), and to a lesser degree, that of LINE-1 methylation-intermediate CRC (p=0.002).

Family History of CRC and Risk of Developing CRC according to Tumor LINE-1 Methylation Level
 No. of first-degree relatives with CRC01>=2P value for trend
 Person-years of follow-up2,787,643370,08826,684 
LINE-1 M-low (<55% methylated)Multivariate HR (95% CI)1 (reference)1.68 (1.19-2.38)3.48 (1.59-7.63)P <0.001
LINE-1 55-65% methylatedMultivariate HR (95% CI)1 (reference)1.50 (1.17-1.93)1.42 (0.64-3.14)P = 0.002
LINE-1 M-high (>=65% methylated)Multivariate HR (95% CI)1 (reference)1.11 (0.83-1.48)1.43 (0.59-3.49)P = 0.35
CI, confidence interval; HR, hazard ratio; M, methylation. P value for trend indicates stastistcal trend for CRC risk from those with no family history, to those with 1 affected first-degree relative, to those with >=2 affected first-degree relatives.

In contrast, CRC family history did not elevate the risk of LINE-1 methylation-high CRC (P=0.35). The association between CRC family history and CRC risk significantly differed by tumor LINE-1 methylation level (P for heterogeneity =0.020).
Conclusions: CRC family history is associated with higher risk of LINE-1 hypomethylated CRC, implying a previously unrecognized familial CRC trait characterized by heritable predisposition to epigenetic alterations. Our data suggest that, similar to the MSI testing, LINE-1 methylation can serve as a tumor biomarker for familial cancer risk assessment.
Category: Gastrointestinal

Monday, March 4, 2013 8:00 AM

Proffered Papers: Section D, Monday Morning


Close Window