PPARgamma and colon and rectal cancer: associations with specific tumor mutations, aspirin, ibuprofen and insulin-related genes (United States)

Cancer Causes Control. 2006 Apr;17(3):239-49. doi: 10.1007/s10552-005-0411-6.

Abstract

We hypothesize that the peroxisome proliferator-activated receptor-gamma (PPARgamma) is associated with colorectal cancer given its association with insulin, diabetes, obesity, and inflammation. In this study, we evaluated the association between colorectal cancer and specific tumor mutations and the Pro12Ala (P12A) PPARgamma polymorphism. We also evaluated interactions between the PPARgamma gene and other insulin-related genes and use of aspirin and non-steroidal anti-inflammatory drug use. Data were available from 1,577 cases of colon cancer that were matched to 1,971 population-based controls and 794 cases of rectal cancer that were matched to 1,001 population-based controls. Colon tumors from the case subjects were evaluated for p53 and Ki-ras mutations and microsatellite instability (MSI). Insulin-related genes evaluated were the Bsm1, polyA, and Fok1 polymorphisms of the VDR gene; the G972R IRS1 polymorphism; the G1057D IRS2 polymorphism; the 19CA repeat polymorphism of the IGF1 gene; and the -200A>C IGFBP3 polymorphism. The odds ratio (OR) between the PA/AA genotypes and proximal tumors was 0.83 (95% CI: 0.69-1.01); for distal tumors was 1.00 (95% CI: 0.83-1.21); and for rectal tumors was 1.04 (95% CI: 0.86-1.25). Evaluation of specific types of tumor mutations showed that colon cancer cases with the PA or AA genotypes were less likely to have p53 tumor mutations (OR 0.78; 95% CI: 0.62-0.99), specifically transition mutations (OR 0.74; 95% CI: 0.56-0.97). Colon cancer cases also were less likely to have a tumor with MSI if they had the PA or AA PPARgamma genotype (OR 0.68; 95% CI: 0.47-0.98); differences in Ki-ras mutations were not seen in colon tumors by PPARgamma genotype. Those who did not take ibuprofen-type drugs and had the PA or AA genotypes were at a significantly greater risk of rectal cancer (OR 2.11; 95% CI: 1.52-2.92; p interaction 0.03) than people with the PP genotype regardless of ibuprofen-type drug use. There was a significant interaction between the -200A>C IGFBP3 polymorphism and the Pro12Ala PPARgamma polymorphism and risk of colon cancer (p for interaction = 0.02) with individuals being at significantly lower risk if they had both the CC IGFBP3 genotype and the PA/AA PPARgamma genotype. For rectal cancer there was a significant interaction between the Bsm1/polyA polymorphisms (p = 0.001) of the VDR gene and the PA/AA Pro12Ala PPARgamma polymorphism with the highest risk group being those with both the PA/AA Pro12Ala PPARgamma and the BB/SS VDR genotypes. These data suggest that PPARgamma may be associated with many aspects of colorectal cancer including insulin- and inflammation-related mechanisms.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Anti-Inflammatory Agents, Non-Steroidal / therapeutic use*
  • Aspirin / therapeutic use*
  • California / epidemiology
  • Case-Control Studies
  • Colonic Neoplasms / epidemiology
  • Colonic Neoplasms / genetics*
  • Female
  • Genotype
  • Humans
  • Ibuprofen / therapeutic use*
  • Insulin / genetics
  • Male
  • Middle Aged
  • Minnesota / epidemiology
  • Mutation
  • Odds Ratio
  • PPAR gamma
  • Polymorphism, Genetic / genetics
  • Rectal Neoplasms / epidemiology
  • Rectal Neoplasms / genetics*
  • Utah / epidemiology

Substances

  • Anti-Inflammatory Agents, Non-Steroidal
  • Insulin
  • PPAR gamma
  • Aspirin
  • Ibuprofen