Modelling the Development of Precancerous Polyps: Utility of Social History, Comorbidities, FOBT/FIT Testing, and NSAID Use

Abstract

Introduction: Studies have attempted to ascertain the risk factors (RF) for developing colorectal cancer (CRC), but few have attempted to elucidate the causes of precancerous (PC) polyposis. Early detection of PC polyps is essential to prevent progression to CRC. Evaluation of RF such as family history (FH), tobacco and alcohol abuse, and glycemic control may be useful in determining which patients may be more likely to have PC polyps. Methods: Data were collected on 4200 patients at the Robley Rex VAMC in Louisville, Ky. 2758 were excluded based on exclusion criteria: diagnosis of IBD and pre-existing malignancies, birth prior to 1948, poor bowel preparation, and lack of pathologic/histologic data from first colonoscopy. Of the 1442 patients included in the study, a subset had: FOBT or FIT testing, positive family history of CRC, positive smoking history, COPD, history of NSAID use, and hemoglobin A1c within 6 months of colonoscopy. A Poisson regression was used for analysis of relative risk (RR). P<0.05, was deemed statistically significant. Results: Delaying colonoscopy had a RR of 1.03 (P<0.001) per year and males had an increased RR of 1.32 (P=0.011) when compared to females. Family history of CRC (first degree relative) had a RR of 1.179 (P=0.016). Hemoglobin A1c >6.5% had a RR of 1.147 (P=0.031). A positive FOBT/FIT also demonstrated an increase in RR of 1.677 (P<0.001). Current smokers had a RR of 1.215 (P=0.003) when compared to those who never smoked. Patients with COPD had a RR of 1.348 (P<0.001). NSAID users had a decreased RR of 0.688 (P<0.001), if NSAIDs were used within 6 months preceding initial colonoscopy. Conclusion: Development of PC polyps is multifactorial. Increased risk occurs in those with a positive family history, smoking, history of COPD and suboptimal glycemic control. Results indicate that positive FOBT and FIT are also associated with an increased relative risk for PC polyps. This may illustrate a greater ability of these testing modalities to detect PC polyps. Interestingly, the use of NSAIDs appears to have a protective effect, thereby decreasing the RR. The major limitation of this study is the primarily male VA population. In the future, utilization of the factors identified in this study may allow for further risk stratification, resulting in specific target years for initial screening colonoscopies.Table: Table. Model for precancerous colonic polyp risk