Phenome-Wide Scan Finds Potential Orofacial Risk Markers for Cancer

Abstract

Cancer is a disease caused by a process that drives the transformation of normal cells into malignant cells. The late diagnosis of cancer has a negative impact on the health care system due to high treatment cost and decreased chances of favorable prognosis. Here, we aimed to identify orofacial conditions that can serve as potential risk markers for cancers by performing a phenome-wide scan (PheWAS). From a pool of 6,100 individuals, both genetic and epidemiological data of 1,671 individuals were selected: 350 because they were previously diagnosed with cancer and 1,321 to match to those individuals that had cancer, based on age, sex, and ethnicity serving as a comparison group. Results of this study showed that when analyzing the individuals affected by cancer separately, tooth loss/edentulism is associated with SNPs in AXIN2 (rs11867417 p = 0.02 and rs2240308 p = 0.02), and leukoplakia of oral mucosa is associated with both AXIN2 (rs2240308 p = 0.03) and RHEB (rs2374261 p = 0.03). These phenotypes did not show the same trends in patients that were not diagnosed with cancer, allowing for the conclusion that these phenotypes are unique to cases with higher cancer risk.

Figure 1

Plot representing the phenome-wide association analysis in the total sample. The horizontal red line indicates the threshold of p = 0.002; the horizontal blue line indicates the threshold of p = 0.05, phenotypes found below the blue line (p > 0.05 – not associated) are not annotated in the plots to avoid noise. The triangle tip direction represents the odds ratio direction of each association, upward triangles indicate OR ≥ 1; downward triangles indicate a protective effect (OR < 1.0); different triangle colors indicate different disease groups (from left to right – dark green = neoplasms, dark blue = neurological system, bright red = circulatory system, brown = respiratory, green = digestive, dark red = dermatologic and light blue = congenital anomalies). (a) AXIN2 - rs11867417 and its association with glossitis (p < 0.002). (b) AXIN2 - rs2240308 and its protective effect towards having gingivitis, chronic periodontitis, and leukoplakia of the oral mucosa (p < 0.05). (c) RHEB - rs1109089 and its association with both disorders of tooth development (p < 0.05) (p < 0.05), and tooth fracture (p < 0.002) (p < 0.002), and its protective effect towards anomalies of jaw size/ symmetry (p < 0.05).

Figure 2

Plot representing the phenome-wide association analysis in the cancer-affected sample. The horizontal red line indicates the threshold of p = 0.002; the horizontal blue line indicates the threshold of p = 0.05, phenotypes found below the blue line (p > 0.05 – not associated) are not annotated in the plots to avoid noise. The triangle tip direction represents the odds ratio direction of each association, upward triangles indicate OR ≥ 1; downward triangles indicate a protective effect (OR < 1.0); different triangle colors indicate different disease groups (from left to right – light red=circulatory system, green=digestive, dark red=dermatologic and light blue=congenital anomalies). (a) AXIN2 - rs11867417 and its association with loss of teeth/edentulism, and its protective effect towards gingivitis (p < 0.05). (b) AXIN2 - rs2240308 and its association with loss of teeth/edentulism and its protective effect against leukoplakia of the oral mucosa (p < 0.05). (c) RHEB - rs1109089 and its protective effect against periodontitis (p < 0.05).

Figure 3

Overall study design.